From ramkumar at intotoinc.com Fri Mar 1 02:21:24 2002 From: ramkumar at intotoinc.com (Ram Kumar K V) Date: Thu, 28 Feb 2002 23:21:24 -0800 (PST) Subject: [uClinux-dev] uClinux-mips Message-ID: <200203010721.g217LO720920@intotoinc.com> Hi, How to set the date and time for the MIPS board running uClinux 2.4.14? I tried with date MMDDhhmmYY, but failed to set the date. Please coment on this. thanks, Ramkumar. This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From miles at lsi.nec.co.jp Fri Mar 1 02:48:05 2002 From: miles at lsi.nec.co.jp (Miles Bader) Date: 01 Mar 2002 16:48:05 +0900 Subject: [uClinux-dev] Re: BusyBox `msh' doesn't work correctly with vfork / uClibc In-Reply-To: References: Message-ID: The following patch to msh.c (made with very little understanding of the source, though...) seems to fix the immediate problem: diff -up user/busybox/msh.c.\~1\~ user/busybox/msh.c --- user/busybox/msh.c.~1~ Thu Feb 14 11:42:15 2002 +++ user/busybox/msh.c Fri Mar 1 16:42:16 2002 @@ -2604,17 +2604,15 @@ forkexec( register struct op *t, int *pi signal(SIGQUIT, SIG_DFL); } if (t->type == TPAREN) - exit(execute(t->left, NOPIPE, NOPIPE, FEXEC)); + _exit(execute(t->left, NOPIPE, NOPIPE, FEXEC)); if (wp[0] == NULL) - exit(0); + _exit(0); cp = rexecve(wp[0], wp, makenv()); prs(wp[0]); prs(": "); warn(cp); if (!execflg) trap[0] = NULL; - leave(); - /* NOTREACHED */ - exit(1); + _exit (1); } /* -Miles -- Run away! Run away! This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From miles at lsi.nec.co.jp Fri Mar 1 03:03:33 2002 From: miles at lsi.nec.co.jp (Miles Bader) Date: 01 Mar 2002 17:03:33 +0900 Subject: [uClinux-dev] Re: BusyBox `msh' doesn't work correctly with vfork / uClibc In-Reply-To: References: Message-ID: Oh, also, I should note that this is the version of busybox from uClinux-dist-20020220. -Miles -- Suburbia: where they tear out the trees and then name streets after them. This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From Dieter.Ripp at gmx.de Fri Mar 1 04:21:52 2002 From: Dieter.Ripp at gmx.de (Dieter Ripp) Date: Fri, 01 Mar 2002 09:21:52 -0000 Subject: [uClinux-dev] UClinux into the flash In-Reply-To: <200203010721.g217LO720920@intotoinc.com> Message-ID: Dear all, I am able to download the Kernel linux.bin via TFTP into the ram of the Motorola M5272C3 development board. The GO 20000 command starts the uclinux kernel. So far so good. But what should be done, to put the kernel into Flash of Motorola M5272C3 development. For your help in advance Dieter Ripp This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From xue_pengyu at anyka.com.cn Fri Mar 1 06:04:19 2002 From: xue_pengyu at anyka.com.cn (xue_pengyu) Date: Fri, 1 Mar 2002 19:04:19 +0800 Subject: [uClinux-dev] m68k assemble language Message-ID: <8173EA33E7F5D511898500E04C57D7C6EFC7@SERVER> Hi, all : i write a simple source file like this test.S : ......init CPU.... jsr mytest .................... test.c void mytest(void) { return; } then i compile them together m68k-elf-gcc -m68000 test.S test.c -o test -msep-data -T/usr/local/m68k-elf/lib/elf2flt.ld -nostdlib m68k-elf-objcopy test -O srec test.sre then i load the test.sre to the board, but i found that the instruction jsr mytest failed, mytest(void) can't be exected,can someone help me to solve the problem? If i am not using jsr mytest, the program runs well. many thanks! xue pengyu This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From carlos at embedded.cl Fri Mar 1 07:15:41 2002 From: carlos at embedded.cl (Carlos Manuel Duclos Vergara) Date: Fri, 1 Mar 2002 09:15:41 -0300 Subject: [uClinux-dev] ColdFire patch In-Reply-To: <3C7EBC37.69EBD83@snapgear.com> References: <20020228113409.3e492305.carlos@embedded.cl> <3C7EBC37.69EBD83@snapgear.com> Message-ID: <20020301091541.72da84a9.carlos@embedded.cl> > > Hi Carlos, > Hi Greg! > > These changes are already in CVS, have been for a couple > of weeks now... > if you look inside, there is only a change of position.... the original config.in won't work for me (maybe some weird thing in my ws, but in this order it works just fine) the second is a "semi automatic" clock frequency.... if you specify auto clock frequency it will fill that address with the lowest speed available for that processor... is not a big thing, but after reading the code i realize that were no management for the auto option! bye -- "Solo me arrepiento de unos * de menos y unos ++ de sobra" Carlos Manuel Duclos Vergara This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From julienb at actimage.fr Fri Mar 1 07:47:40 2002 From: julienb at actimage.fr (Julien Boibessot) Date: Fri, 1 Mar 2002 13:47:40 +0100 Subject: [uClinux-dev] UClinux into the flash References: Message-ID: <02c001c1c11f$4646b330$8947ec95@bruker.fr> I'm trying to do the same thing....:-) I'm currently following Bernard Kuhn How To (http://www.uclinux.org/pub/uClinux/archive/5095.html). More news as soon as I manage to make something work....:-) Julien ----- Original Message ----- From: "Dieter Ripp" To: Sent: Friday, March 01, 2002 10:21 AM Subject: [uClinux-dev] UClinux into the flash > > Dear all, > > I am able to download the Kernel linux.bin via TFTP into the ram of the Motorola M5272C3 development board. The GO 20000 command starts the uclinux kernel. > So far so good. > > But what should be done, to put the kernel into Flash of Motorola M5272C3 development. > > For your help in advance > > Dieter Ripp > > > > > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From julienb at actimage.fr Fri Mar 1 07:56:36 2002 From: julienb at actimage.fr (Julien Boibessot) Date: Fri, 1 Mar 2002 13:56:36 +0100 Subject: [uClinux-dev] 2.4 kernel problem with network/NFS on Motorola 5272 devt board References: Message-ID: <02c401c1c120$85a805a0$8947ec95@bruker.fr> Hello, I have successfuly manage to compile and run a 2.4 kernel for my devt board (:-)), but there seems to be something wrong with the network layer. During the boot linux try to do a "route add -net 127.0.0.0 lo" and the following error occur: "SIOCADDRTO: Invalid argument". I can't make a "route add ....." myself too....(for mounting an NFS filesystem for example). For info the red D2 led on my board (Motorola DNA 5272) is always on (does not occur with 2.0.x kernel). Does anybody knows what's going wrong ?? (I'm using 20011112 kernel snapshot.....) Thanks Julien This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From philip at engarts.com Fri Mar 1 07:14:14 2002 From: philip at engarts.com (Philip Nye) Date: Fri, 1 Mar 2002 12:14:14 -0000 Subject: [uClinux-dev] UClinux into the flash References: Message-ID: <004901c1c11a$9aa1f3e0$5801a8c0@localdomain> > From: "Dieter Ripp" > But what should be done, to put the kernel into Flash of Motorola > M5272C3 development. Start by having a look at Bernard Kuhn's mini Howto: http://www.rcs.ei.tum.de/~kuhn/uclinux/coldfire/tarifa/20011119/README Philip Nye This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From MGroden at hxi.com Fri Mar 1 09:33:11 2002 From: MGroden at hxi.com (Mike Groden) Date: Fri, 1 Mar 2002 09:33:11 -0500 Subject: [uClinux-dev] UClinux into the flash Message-ID: <879EA880A0FED411996B00B0D0B082EC1007DD@hxi_exch01.hxi.com> Before seeing the B. Kuhn's instructions, I worked through this myself using the Colilo bootloader and a flash programmer from P&E Microsystems. The flash programmer attaches via the BDM port and the "wiggler" supplied with the eval kit - it's relatively inexpensive. I had to use m68k-elf-objcopy to move the binary image and create an srec file that the programmer uses. I can send you detailed steps if this option looks interesting. -----Original Message----- From: Dieter Ripp [mailto:Dieter.Ripp at gmx.de] Sent: Friday, March 01, 2002 4:22 AM To: uclinux-dev at uclinux.org Subject: [uClinux-dev] UClinux into the flash Dear all, I am able to download the Kernel linux.bin via TFTP into the ram of the Motorola M5272C3 development board. The GO 20000 command starts the uclinux kernel. So far so good. But what should be done, to put the kernel into Flash of Motorola M5272C3 development. For your help in advance Dieter Ripp This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From philw at lineo.com Fri Mar 1 06:26:37 2002 From: philw at lineo.com (Phil Wilshire) Date: Fri, 01 Mar 2002 06:26:37 -0500 Subject: [uClinux-dev] Minicom is a hassle to use with the uCsimm References: <20020220102224.GA5740@codepoet.org> <20020226150833.GE724@0xd6.org> <20020226153158.GA14957@codepoet.org> <20020226195943.6cdeb1d0.andersn@isy.liu.se> <20020227081933.B11625@beast.internal.moreton.com.au> Message-ID: <3C7F656D.E762B36C@lineo.com> Hi all, Here is a script I use to download to the uCSimm. It seems to work sometimes..... You will have to set up the filename etc.. As usual YMMV regards Phil David McCullough wrote: > > Jivin andersn at isy.liu.se lays it down ... > > > > Hi. > > > > Is there anyone successfully using something else but minicom for talking to the serial port of the uCsimm? > > > > I find minicom to be really troublesome for this. Downloading a new image (if I'm not able to use flashloader via the TP-connection) with rx only works like every tenth time or so. The other times there is always a NAK problem. > > The delete button is not forwarded correctly to the uCsimm (how do I set this up properly?). > > > > Any suggestions for other applications that could be used instead of minicom? > > I've used "cu" to do this on uCsimm/uCdimm, I think I just use "~+cat file" > to send the image, > > Cheers, > Davidm > > -- > David McCullough: Ph: +61 7 3435 2815 http://www.SnapGear.com > davidm at snapgear.com Fx: +61 7 3891 3630 825 Stanley St., W'gabba QLD 4102, Oz > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ -- Phil Wilshire Wizard Trainer -- Lineo Inc. -- Where Open Meets Smart Tel: (703) 863-0715 -------------- next part -------------- A non-text attachment was scrubbed... Name: download.sh Type: application/x-sh Size: 530 bytes Desc: not available URL: From Fabio.Ricciato at TILAB.COM Fri Mar 1 11:19:11 2002 From: Fabio.Ricciato at TILAB.COM (Ricciato Fabio) Date: Fri, 01 Mar 2002 17:19:11 +0100 Subject: [uClinux-dev] uClinux on ARM7TDMI and Integrator AP: who succeeded ? Message-ID: <6ECEC1E214F2E342814ABB1ED10E7955040369@EXC2K01B.cselt.it> Hi Fabrice, thanks a lot for your notes. I used your patches and after some other little tuning I achieved a really basic kernel compilation. Then, I tried to add the serial console support but the compilation requires me to have serial_core.o. The question is: should I have to unset the CONFIG_SERIAL_CORE or shoud I have to find that missing file on original kernel? I'm using kernel 2.4.17. Thanks Bye Fabio -----Original Message----- From: Fabrice Gautier [mailto:Fabrice_Gautier at sdesigns.com] Sent: Tuesday, February 26, 2002 8:57 PM To: 'uclinux-dev at uclinux.org' Subject: RE: [uClinux-dev] uClinux on ARM7TDMI and Integrator AP: who succeeded ? I have it running here, with a 740T core module. With 2.4 kernel. First I copied the mach-integrator directory from arch/arm and added this patch: --- arm/mach-integrator/arch.c Fri Sep 28 03:19:04 2001 +++ armnommu/mach-integrator/arch.c Wed Dec 5 12:07:24 2001 @@ -62,7 +62,7 @@ MACHINE_START(INTEGRATOR, "ARM-Integrator") MAINTAINER("ARM Ltd/Deep Blue Solutions Ltd") - BOOT_MEM(0x00000000, 0x16000000, 0xf1600000) + BOOT_MEM(0x00000000, 0x16000000, 0x16000000) BOOT_PARAMS(0x00000100) FIXUP(integrator_fixup) MAPIO(integrator_map_io) Then i had to add support for the 740T, i forked it from arch/armnommu/mm/proc-arm720.S And then i modified the makefiles etc, for all that. The kernel should be loaded at 0x8000 with r0=0, and it will try to mount a romfs starting @ 0x25000000 in flash (that's block 128 when using arm flash utility) arm-720.S and integrator.diff attached. I doubt the integrator.diff patch will apply cleanly since I edited it manually. If you dont use 740T youll have to write the code supporting your core module. For the 740T i've configured the MPU so that RAM and ROM are cacheable. With a split cache (1 bank data, 3 bank instructions) The bad thing is that I've very bad performance, seemingly due to the memory bus. Even with cache enabled (if I did it right) the speed is directly related to the speed of the memory bus. If cpu clock is 160Mhz or 50Mhz I still have the same performance (wether it is BogoMIPS or some stupid memory transfert loop) -- Fabrice Gautier, Fabrice_Gautier at sdesigns.com > -----Original Message----- > From: Ricciato Fabio [mailto:Fabio.Ricciato at TILAB.COM] > Sent: Tuesday, February 26, 2002 8:30 AM > To: uclinux-dev at uclinux.org > Subject: [uClinux-dev] uClinux on ARM7TDMI and Integrator AP: who > succeeded ? > > > Hi to all, > > I read in the archive that many of you tried to port uclinux on > Integrator AP with ARM7TDMI but I cannot find the porting on the web. > My question is: has anybody succeeded ? > I approached the problem but I got really a lot of troubles during my > porting work. > I'd like to understand if it's a feasible work or not, and to > understand > if there is some code already ported for such platform configuration. > Thanks a lot > > Fabio > This message resent by the uclinux-dev at uclinux.org list > server http://www.uClinux.org/ > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From ken at reasonability.net Fri Mar 1 12:18:06 2002 From: ken at reasonability.net (Ken Treis) Date: Fri, 01 Mar 2002 09:18:06 -0800 Subject: [uClinux-dev] 2.4 kernel problem with network/NFS on Motorola 5272 devt board References: <02c401c1c120$85a805a0$8947ec95@bruker.fr> Message-ID: <3C7FB7CE.2020108@reasonability.net> Julien Boibessot wrote: >Hello, > >I have successfuly manage to compile and run a 2.4 kernel for my devt board >(:-)), but there seems to be something wrong with the network layer. >During the boot linux try to do a "route add -net 127.0.0.0 lo" and the > This has been fixed (for at least the boards which use the generic "big" rc file) in the most recent distribution (20020220). Under 2.4, you need to also add the netmask to this command. So the command becomes: route add -net 127.0.0.0 netmask 255.0.0.0 lo Have a look at vendors/Generic/big/rc. If it doesn't do what you need it to, fix it. :) > >following error occur: "SIOCADDRTO: Invalid argument". >I can't make a "route add ....." myself too....(for mounting an NFS >filesystem for example). > Probably related to the same problem above; check that your netmask works. > >For info the red D2 led on my board (Motorola DNA 5272) is always on (does >not occur with 2.0.x kernel). > I've noticed this under 2.4 as well; it's a light that's controlled by the PHY IIRC. The Coldfire FEC driver is pretty significantly different under 2.4. However, we haven't seen any noticeable problems under 2.4 when this light is on. Greg might be better able to address this one. >Does anybody knows what's going wrong ?? >(I'm using 20011112 kernel snapshot.....) > -- Ken Treis ken at reasonability.net This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From Fabrice_Gautier at sdesigns.com Fri Mar 1 14:30:17 2002 From: Fabrice_Gautier at sdesigns.com (Fabrice Gautier) Date: Fri, 1 Mar 2002 11:30:17 -0800 Subject: Missing file in CVS ? RE: [uClinux-dev] uClinux on ARM7TDMI and I ntegrator AP: who succeeded ? Message-ID: Humm... Thats very strange... I've serial_core.c in my source, but it appears it has disapeared from the cvs. cvs log on serial core.c tell me the server doesnt know anything about it. It seems it has disapeared from the kernel release as well... I think serial_core.c com from arm-linux, so that would explain ... I may have take it directly from arm linux. -- Fabrice Gautier, Fabrice_Gautier at sdesigns.com > -----Original Message----- > From: Ricciato Fabio [mailto:Fabio.Ricciato at TILAB.COM] > Sent: Friday, March 01, 2002 8:19 AM > To: uclinux-dev at uclinux.org > Cc: Fabrice_Gautier at sdesigns.com > Subject: RE: [uClinux-dev] uClinux on ARM7TDMI and Integrator AP: who > succeeded ? > > > Hi Fabrice, > thanks a lot for your notes. > I used your patches and after some other little tuning I achieved a > really basic kernel compilation. > Then, I tried to add the serial console support but the compilation > requires me to have serial_core.o. > The question is: > should I have to unset the CONFIG_SERIAL_CORE or shoud I have to find > that missing file on original kernel? > I'm using kernel 2.4.17. Thanks > Bye > > Fabio > > -----Original Message----- > From: Fabrice Gautier [mailto:Fabrice_Gautier at sdesigns.com] > Sent: Tuesday, February 26, 2002 8:57 PM > To: 'uclinux-dev at uclinux.org' > Subject: RE: [uClinux-dev] uClinux on ARM7TDMI and Integrator AP: who > succeeded ? > > > I have it running here, with a 740T core module. With 2.4 kernel. > > First I copied the mach-integrator directory from arch/arm and added > this > patch: > > --- arm/mach-integrator/arch.c Fri Sep 28 03:19:04 2001 > +++ armnommu/mach-integrator/arch.c Wed Dec 5 12:07:24 2001 > @@ -62,7 +62,7 @@ > > MACHINE_START(INTEGRATOR, "ARM-Integrator") > MAINTAINER("ARM Ltd/Deep Blue Solutions Ltd") > - BOOT_MEM(0x00000000, 0x16000000, 0xf1600000) > + BOOT_MEM(0x00000000, 0x16000000, 0x16000000) > BOOT_PARAMS(0x00000100) > FIXUP(integrator_fixup) > MAPIO(integrator_map_io) > > > Then i had to add support for the 740T, i forked it from > arch/armnommu/mm/proc-arm720.S > And then i modified the makefiles etc, for all that. The kernel should > be > loaded at 0x8000 with r0=0, and it will try to mount a romfs > starting @ > 0x25000000 in flash (that's block 128 when using arm flash utility) > > arm-720.S and integrator.diff attached. I doubt the integrator.diff > patch > will apply cleanly since I edited it manually. If you dont use 740T > youll > have to write the code supporting your core module. > > For the 740T i've configured the MPU so that RAM and ROM are > cacheable. > With > a split cache (1 bank data, 3 bank instructions) > > The bad thing is that I've very bad performance, seemingly due to the > memory > bus. Even with cache enabled (if I did it right) the speed is directly > related to the speed of the memory bus. If cpu clock is > 160Mhz or 50Mhz > I > still have the same performance (wether it is BogoMIPS or some stupid > memory > transfert loop) > > > > -- > Fabrice Gautier, > Fabrice_Gautier at sdesigns.com > > > > -----Original Message----- > > From: Ricciato Fabio [mailto:Fabio.Ricciato at TILAB.COM] > > Sent: Tuesday, February 26, 2002 8:30 AM > > To: uclinux-dev at uclinux.org > > Subject: [uClinux-dev] uClinux on ARM7TDMI and Integrator AP: who > > succeeded ? > > > > > > Hi to all, > > > > I read in the archive that many of you tried to port uclinux on > > Integrator AP with ARM7TDMI but I cannot find the porting > on the web. > > My question is: has anybody succeeded ? > > I approached the problem but I got really a lot of troubles > during my > > porting work. > > I'd like to understand if it's a feasible work or not, and to > > understand > > if there is some code already ported for such platform > configuration. > > Thanks a lot > > > > Fabio > > This message resent by the uclinux-dev at uclinux.org list > > server http://www.uClinux.org/ > > > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From andersen at codepoet.org Fri Mar 1 16:12:17 2002 From: andersen at codepoet.org (Erik Andersen) Date: Fri, 1 Mar 2002 14:12:17 -0700 Subject: [uClinux-dev] Re: [uClibc]POSIX threading support In-Reply-To: <20020226175506.GG724@0xd6.org> References: <20020220102224.GA5740@codepoet.org> <20020226150833.GE724@0xd6.org> <20020226153158.GA14957@codepoet.org> <20020226175506.GG724@0xd6.org> Message-ID: <20020301211217.GA14801@dillweed.codepoet.org> On Tue Feb 26, 2002 at 11:55:06AM -0600, M. R. Brown wrote: > > I fixed that earlier today.. > > Yes, this was fixed, but the SH port was missing enough files for pthreads > not to build at all. Patch is attached... Thanks, applied. Sorry for the delay. Took me a few days to get my include/bits reorg all figured out. Hopefully I didn't break anything... > Also, I was able to build SDL and it's test programs to test pthread > functionality, but when running libpthread-linked apps I get: > > ./testhread: relocation error: /lib/libpthread.so.0.9.9: undefined symbol: __udivsi3_i4 > > which makes me think libpthread isn't linked to libgcc properly (the SH > cross-compiler is gcc-3.0.3). Any ideas? Very odd. Could you try again with the latest and greatest. We went through several rounds of changes to the gcc wrapper this week, and so things were broken for several days... -Erik -- Erik B. Andersen http://codepoet-consulting.com/ --This message was written using 73% post-consumer electrons-- -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 232 bytes Desc: not available URL: From hamilton at SEDSystems.ca Fri Mar 1 16:52:07 2002 From: hamilton at SEDSystems.ca (Kendrick Hamilton) Date: Fri, 1 Mar 2002 15:52:07 -0600 (CST) Subject: [uClinux-dev] JFFS error Message-ID: Has anybody seen anything like this? I have JFFS as my root partition. I was adding and removing files and tried to remove the link cat. I got the following errors: I am using uClinx 2.0.39.1 rm cat flash: write timeout 0x01500000 Flash: verify error 0x01500000. (flash_write() 1358) *programAddress = 0xe7e7e6e6, *theData = 0x34383931 flash: write timeout 0x01500000 Flash: verify error 0x01500000. (flash_write() 1358) *programAddress = 0x00000000, *theData = 0x34383931 flash: write timeout 0x01500000 Flash: verify error 0x01500000. (flash_write() 1358) *programAddress = 0xe7e7e6e6, *theData = 0x34383931 flash: write timeout 0x01500000 Flash: verify error 0x01500000. (flash_write() 1358) *programAddress = 0x00000000, *theData = 0x34383931 flash: write timeout 0x01500000 Flash: verify error 0x01500000. (flash_write() 1358) *programAddress = 0xe7e7e6e6, *theData = 0x34383931 FATAL FLASH ERROR (1) JFFS: jffs_write_node: Failed to write raw_inode. rm: unable to remove `cat': Operation not permitted # pwd /bin # pwd /bin # ls -l -rwxr-xr-x 1 0 0 205492 busybox lrwxrwxrwx 1 0 0 7 cat -> ??|???| lrwxrwxrwx 1 0 0 7 chmod -> ???|??? lrwxrwxrwx 1 0 0 7 chown -> ??|???| lrwxrwxrwx 1 0 0 7 cp -> ???|??? lrwxrwxrwx 1 0 0 7 date -> ??|???| lrwxrwxrwx 1 0 0 7 dd -> ???|??? lrwxrwxrwx 1 0 0 7 echo -> ??|???| lrwxrwxrwx 1 0 0 7 hostname -> ???|??? lrwxrwxrwx 1 0 0 7 kill -> ??|???| lrwxrwxrwx 1 0 0 7 ln -> ???|??? lrwxrwxrwx 1 0 0 7 ls -> ??|???| lrwxrwxrwx 1 0 0 7 mkdir -> ???|??? lrwxrwxrwx 1 0 0 7 mknod -> ??|???| lrwxrwxrwx 1 0 0 7 more -> ???|??? lrwxrwxrwx 1 0 0 7 mount -> ??|???| lrwxrwxrwx 1 0 0 7 msh -> ???|??? lrwxrwxrwx 1 0 0 7 ping -> ??|???| lrwxrwxrwx 1 0 0 7 ps -> ???|??? lrwxrwxrwx 1 0 0 7 pwd -> ??|???| lrwxrwxrwx 1 0 0 7 rm -> ???|??? lrwxrwxrwx 1 0 0 7 rmdir -> ??|???| lrwxrwxrwx 1 0 0 7 sh -> ???|??? lrwxrwxrwx 1 0 0 7 sleep -> ??|???| lrwxrwxrwx 1 0 0 7 stty -> ???|??? lrwxrwxrwx 1 0 0 7 umount -> ??|???| lrwxrwxrwx 1 0 0 7 uname -> ???|??? lrwxrwxrwx 1 0 0 7 usleep -> ??|???| lrwxrwxrwx 1 0 0 7 vi -> ???|??? -rwsr-xr-x 1 0 0 68300 tinylogin lrwxrwxrwx 1 0 0 9 login -> ??|???|?? lrwxrwxrwx 1 0 0 7 touch -> ???|??? -rwxr-xr-x 1 0 0 102064 ftp -- Kendrick Hamilton E.I.T. SED Systems, a division of Calian Ltd. 18 Innovation Blvd. PO Box 1464 Saskatoon, Saskatchewan Canada S7N 3R1 Hamilton at sedsystems.ca Tel: (306) 933-1453 Fax: (306) 933-1486 This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From ken at reasonability.net Fri Mar 1 18:48:49 2002 From: ken at reasonability.net (Ken Treis) Date: Fri, 01 Mar 2002 15:48:49 -0800 Subject: [uClinux-dev] C++ name mangling Message-ID: <3C801361.1060802@reasonability.net> Seems like all I do is post C++ questions here; hope I can serve as a warning to others. :) I notice that -fname-mangling-version-0 is used in the CXXFLAGS for everything under vendors/. This forces an old mangling scheme that's compatible with pre-2.8 versions of g++. Is this needed this to link against certain libraries? It prevents us from using some kinds of templates. So far, I've found it safe (for our apps) to remove this option and let g++ default to version 1. Is this setting something that's a holdover from days gone by (IOW, should I bother submitting a patch for it), or is it something that's there for a perfectly good reason? -- Ken Treis ken at reasonability.net This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From eric at corpbuilders.com Fri Mar 1 20:09:36 2002 From: eric at corpbuilders.com (Eric deLangis) Date: Fri, 1 Mar 2002 18:09:36 -0700 Subject: [uClinux-dev] Motorola COldfire dhcpcd-new problem Solved In-Reply-To: <8CCBDD5583C50E4196F012E79439B45C1D1973@atl-ms1.megatrends.com> Message-ID: To All: Is there a repository with all the "most current" revisions for the MCF5272C3 development system. I'm running into the same problems (such as this dhcpcd issue and the serial port problem)with old source that was distributed with the development system. I went on to the CVS site to pull down the fec.c/.h files, and I found "many" other files that are newer than what I am using, but I don't know which files are need for proper operation of my system. I'd like to locate site that I can download a complete set of "current" source from for this board. Thanks Eric -----Original Message----- From: owner-uclinux-dev at uclinux.org [mailto:owner-uclinux-dev at uclinux.org]On Behalf Of Subash Kalbarga Sent: Wednesday, February 20, 2002 1:51 PM To: uclinux-dev at uclinux.org Subject: RE: [uClinux-dev] Motorola COldfire dhcpcd-new problem Solved Hi Dhcpcd-new now works fine. Apaparently, fec.c was updated after the 12th Nov 2001 distribution. I used the new fec.c (version 1.4) and it solves my problem Thanks a lot Greg Subash -----Original Message----- From: Greg Ungerer [mailto:gerg at snapgear.com] Sent: Tuesday, February 19, 2002 11:48 PM To: uclinux-dev at uclinux.org Subject: Re: [uClinux-dev] Motorola COldfire dhcpcd-new problem Hi Subash, Subash Kalbarga wrote: > I tried it on a M5272C3 board , with the 2.4 kernel and dhcpcd-new selected > in the config > > In fact if I do a ifconfig to set a IP statically while the network cable is > not plugged in, it will set it but then when I plug the network cable in it > is not pingable either and never recovers! Which may mean it is not > something to do with the DHCP client? Hmm, pretty strange. I just tried the latest uClinux snapshot with 2.4.17 kernel and using dhcp-new and it works ok. What version source/snapshot are you using? Regards Greg ------------------------------------------------------------------------ Greg Ungerer -- Chief Software Wizard EMAIL: gerg at snapgear.com SnapGear PHONE: +61 7 3435 2888 825 Stanley St, FAX: +61 7 3891 3630 Woolloongabba, QLD, 4102, Australia WEB: www.snapgear.com This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From gerg at snapgear.com Fri Mar 1 22:37:47 2002 From: gerg at snapgear.com (gerg) Date: Sat, 02 Mar 2002 13:37:47 +1000 Subject: [uClinux-dev] 2.4 kernel problem with network/NFS on Motorola 5272 devt board References: <02c401c1c120$85a805a0$8947ec95@bruker.fr> <3C7FB7CE.2020108@reasonability.net> Message-ID: <3C80490B.844EF437@snapgear.com> Hi Ken, Julien, Ken Treis wrote: > Julien Boibessot wrote: > >For info the red D2 led on my board (Motorola DNA 5272) is always on (does > >not occur with 2.0.x kernel). > > > I've noticed this under 2.4 as well; it's a light that's controlled by > the PHY IIRC. The Coldfire FEC driver is pretty significantly different > under 2.4. > > However, we haven't seen any noticeable problems under 2.4 when this > light is on. Greg might be better able to address this one. I hadn't noticed this before. It almost certainly has something to do with the initial setup of the PHY. I have probably got code in the 2.0 fec driver to init the PHY in a certain way (to make the LEDs have certain functions). I suspect this code hasn't been carried over to the 2.4 fec driver. Regards Greg ------------------------------------------------------------------------ Greg Ungerer -- Chief Software Wizard EMAIL: gerg at snapgear.com Snapgear PHONE: +61 7 3279 1822 825 Stanley St, FAX: +61 7 3279 1820 Woolloongabba, QLD, 4102, Australia WEB: www.snapgear.com This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From gerg at snapgear.com Fri Mar 1 22:40:37 2002 From: gerg at snapgear.com (gerg) Date: Sat, 02 Mar 2002 13:40:37 +1000 Subject: [uClinux-dev] ColdFire patch References: <20020228113409.3e492305.carlos@embedded.cl> <3C7EBC37.69EBD83@snapgear.com> <20020301091541.72da84a9.carlos@embedded.cl> Message-ID: <3C8049B5.BC82446C@snapgear.com> Hi Carlos, Carlos Manuel Duclos Vergara wrote: > > These changes are already in CVS, have been for a couple > > of weeks now... > > if you look inside, there is only a change of position.... the original > config.in won't work for me (maybe some weird thing in my ws, but in > this order it works just fine) OK. It looked from the patch like an inclusion of the CLK FREQ config, but there was no removal part. > the second is a "semi automatic" clock frequency.... if you specify auto > clock frequency it will fill that address with the lowest speed > available for that processor... > is not a big thing, but after reading the code i realize that were no > management for the auto option! No, that is true. AUTO is there as a default for CPU's that don't need to, or want to, have configurable CLK FREQ support. Regards Greg ------------------------------------------------------------------------ Greg Ungerer -- Chief Software Wizard EMAIL: gerg at snapgear.com Snapgear PHONE: +61 7 3279 1822 825 Stanley St, FAX: +61 7 3279 1820 Woolloongabba, QLD, 4102, Australia WEB: www.snapgear.com This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From gerg at snapgear.com Fri Mar 1 22:42:58 2002 From: gerg at snapgear.com (gerg) Date: Sat, 02 Mar 2002 13:42:58 +1000 Subject: [uClinux-dev] Motorola COldfire dhcpcd-new problem Solved References: Message-ID: <3C804A42.A0BCAE08@snapgear.com> Hi Eric, Eric deLangis wrote: > Is there a repository with all the "most current" revisions for the > MCF5272C3 development system. You can always get the latest code for ColdFire platforms from: http://www.uclinux.org/ports/coldfire/source.html > I'm running into the same problems (such as > this dhcpcd issue and the serial port problem)with old source that was > distributed with the development system. I went on to the CVS site to pull > down the fec.c/.h files, and I found "many" other files that are newer than > what I am using, but I don't know which files are need for proper operation > of my system. The code snapshots at the above web page contain the latest from CVS when they are generated. > I'd like to locate site that I can download a complete set of "current" > source from for this board. Regards Greg ------------------------------------------------------------------------ Greg Ungerer -- Chief Software Wizard EMAIL: gerg at snapgear.com Snapgear PHONE: +61 7 3279 1822 825 Stanley St, FAX: +61 7 3279 1820 Woolloongabba, QLD, 4102, Australia WEB: www.snapgear.com This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From gerg at snapgear.com Sat Mar 2 09:53:18 2002 From: gerg at snapgear.com (gerg) Date: Sun, 03 Mar 2002 00:53:18 +1000 Subject: [uClinux-dev] uClinux 2.5.5! Message-ID: <3C80E75E.113AAC57@snapgear.com> Hi All, I have ported the uClinux 2.4.17 patches upto 2.5.5! Enough things have changed in the kernel that it took a little efort to get going :-) It is only running on one target so far, the Motorola M5272C3 (ColdFire) board. But others should quickly follow. I want to track the 2.5.x series pretty closely from now on. So expect patches against most 2.5 series kernels. I am going to keep the patches as minimal as possible, that will give us the best shot of getting into the mainline linux kernels. So the linux-2.5.XucY series has begun! You can get the patch at: http://www.uclinux.org/pub/uClinux/uClinux-2.5.x/linux-2.5.5uc0.patch.gz Regards Greg ------------------------------------------------------------------------ Linux version 2.5.5uc0 (gerg at goober) (gcc version 2.95.3 20010315 (release)(ColdFire patches - 20010318 from http://fiddes.net/coldfire/)(-msep-data patches)) #255 Sat Mar 2 23:53:04 EST 2002  uClinux/COLDFIRE(m5272) COLDFIRE port done by Greg Ungerer, gerg at snapgear.com Flat model support (C) 1998,1999 Kenneth Albanowski, D. Jeff Dionne On node 0 totalpages: 1024 zone(0): 0 pages. zone(1): 1024 pages. zone(2): 0 pages. Kernel command line: Calibrating delay loop... 43.62 BogoMIPS Memory available: 2572k/4096k RAM, 0k/0k ROM (747k kernel code, 215k data) kmem_create: Forcing size word alignment - task_struct kmem_create: Forcing size word alignment - vm_area_struct kmem_create: Forcing size word alignment - mm_struct kmem_create: Forcing size word alignment - buffer_head kmem_create: Forcing size word alignment - filp Dentry-cache hash table entries: 512 (order: 0, 4096 bytes) Inode-cache hash table entries: 512 (order: 0, 4096 bytes) kmem_create: Forcing size word alignment - inode_cache Mount-cache hash table entries: 512 (order: 0, 4096 bytes) kmem_create: Forcing size word alignment - bdev_cache kmem_create: Forcing size word alignment - cdev_cache Buffer-cache hash table entries: 1024 (order: 0, 4096 bytes) Page-cache hash table entries: 1024 (order: 0, 4096 bytes) kmem_create: Forcing size word alignment - proc_inode_cache POSIX conformance testing by UNIFIX Linux NET4.0 for Linux 2.4 Based upon Swansea University Computer Society NET3.039 kmem_create: Forcing size word alignment - sock kmem_create: Forcing size word alignment - sock_inode_cache Initializing RT netlink socket Starting kswapd kmem_create: Forcing size word alignment - bio BIO: pool of 256 setup, 13Kb (54 bytes/bio) biovec: init pool 0, 1 entries, 12 bytes biovec: init pool 1, 4 entries, 48 bytes biovec: init pool 2, 16 entries, 192 bytes biovec: init pool 3, 64 entries, 768 bytes biovec: init pool 4, 128 entries, 1536 bytes biovec: init pool 5, 256 entries, 3072 bytes kmem_create: Forcing size word alignment - file lock cache kmem_create: Forcing size word alignment - ext2_inode_cache kmem_create: Forcing size word alignment - nfs_inode_cache ColdFire internal UART serial driver version 1.00 ttyS0 at 0x10000100 (irq = 73) is a builtin ColdFire UART ttyS1 at 0x10000140 (irq = 74) is a builtin ColdFire UART kmem_create: Forcing size word alignment - blkdev_requests block: 64 slots per queue, batch=16 eth0: FEC ENET Version 0.2, 00:cf:52:72:c3:01 fec: Phy @ 0x0, type 0x001378e1 Blkmem copyright 1998,1999 D. Jeff Dionne Blkmem copyright 1998 Kenneth Albanowski Blkmem 1 disk images: 0: 110E0C-16D60B [VIRTUAL 110E0C-16D60B] (RO) RAMDISK driver initialized: 16 RAM disks of 4096K size 1024 blocksize PPP generic driver version 2.4.1 NET4: Linux TCP/IP 1.0 for NET4.0 kmem_create: Forcing size word alignment - tcp_sock kmem_create: Forcing size word alignment - udp_sock kmem_create: Forcing size word alignment - raw4_sock IP Protocols: ICMP, UDP, TCP, IGMP kmem_create: Forcing size word alignment - ip_dst_cache IP: routing cache hash table of 512 buckets, 4Kbytes TCP: Hash tables configured (established 512 bind 512) NET4: Unix domain sockets 1.0/SMP for Linux NET4.0. kmem_create: Forcing size word alignment - unix_sock VFS: Mounted root (romfs filesystem) readonly. Freeing unused kernel memory: 24k freed (0xf8000 - 0xfd000) Shell invoked to run file: /etc/rc Command: hostname uClinux Command: /bin/expand /etc/ramfs.img /dev/ram0 Command: mount -t proc proc /proc Command: mount -t ext2 /dev/ram0 /var Command: mkdir /var/tmp Command: mkdir /var/log Command: mkdir /var/run Command: mkdir /var/lock Command: ifconfig lo 127.0.0.1 Command: route aedd -net 127.0.0.0 netmaskt 255.0.0.0 lo Command: dhhcpcd -p -a eth0 & 0: config: auto-negotiation on, 100FDX, 100HDX, 10FDX, 10HDX. [11] Command: cat /etc/motd Welcome to ____ _ _ / __| ||_| _ _| | | | _ ____ _ _ _ _ | | | | | | || | _ \| | | |\ \/ / | |_| | |__| || | | | | |_| |/ \ | ___\____|_||_|_| |_|\____|\_/\_/ | | |_| For further information check: http://www.uclinux.org/ Execution Finished, Exiting Sash command shell (version 1.1.1) /> eth0: status: link up, 100MBit Half Duplex, auto-negotiation complete. ps PID PORT STAT SIZE SHARED %CPU COMMAND 1 S 18K 0K 9.5 init 2 S 0K 0K 0.9 keventd 3 R 0K 0K 10.3 ksoftirqd_CPU0 4 S 0K 0K 0.2 kswapd 5 S 0K 0K 0.0 bdflush 6 S 0K 0K 0.0 kupdated 11 S 35K 0K 0.3 dhcpcd -p -a eth0 12 S0 R 20K 0K 1.5 /bin/sh 13 S 11K 0K 0.9 /bin/inetd /> ------------------------------------------------------------------------ Greg Ungerer -- Chief Software Wizard EMAIL: gerg at snapgear.com Snapgear PHONE: +61 7 3279 1822 825 Stanley St, FAX: +61 7 3279 1820 Woolloongabba, QLD, 4102, Australia WEB: www.snapgear.com This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From hamilton at SEDSystems.ca Sat Mar 2 10:48:28 2002 From: hamilton at SEDSystems.ca (Kendrick Hamilton) Date: Sat, 2 Mar 2002 09:48:28 -0600 (CST) Subject: [uClinux-dev] C++ name mangling In-Reply-To: <3C801361.1060802@reasonability.net> Message-ID: Ken, We have been using C++. We are not using the -fname-mangling-version-0 flag. The only problem I have (and it may be fixed now, we use an old version of uClibc) is that I need to put extern "C" around the includes of the C libraries. Just a couple of other problems, We setup try/catch blocks around code that throws exceptions but the exceptions are not being caught. For our code this is not a major problem, but it would be nice if the catchs worked. On Fri, 1 Mar 2002, Ken Treis wrote: > Seems like all I do is post C++ questions here; hope I can serve as a > warning to others. :) > > I notice that -fname-mangling-version-0 is used in the CXXFLAGS for > everything under vendors/. This forces an old mangling scheme that's > compatible with pre-2.8 versions of g++. Is this needed this to link > against certain libraries? It prevents us from using some kinds of > templates. > > So far, I've found it safe (for our apps) to remove this option and let > g++ default to version 1. Is this setting something that's a holdover > from days gone by (IOW, should I bother submitting a patch for it), or > is it something that's there for a perfectly good reason? > > -- > Ken Treis > ken at reasonability.net > > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ > -- Kendrick Hamilton E.I.T. SED Systems, a division of Calian Ltd. 18 Innovation Blvd. PO Box 1464 Saskatoon, Saskatchewan Canada S7N 3R1 Hamilton at sedsystems.ca Tel: (306) 933-1453 Fax: (306) 933-1486 This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From andersen at codepoet.org Sat Mar 2 10:59:40 2002 From: andersen at codepoet.org (Erik Andersen) Date: Sat, 2 Mar 2002 08:59:40 -0700 Subject: [uClinux-dev] C++ name mangling In-Reply-To: References: <3C801361.1060802@reasonability.net> Message-ID: <20020302155940.GA14008@codepoet.org> On Sat Mar 02, 2002 at 09:48:28AM -0600, Kendrick Hamilton wrote: > Ken, > We have been using C++. We are not using the > -fname-mangling-version-0 flag. The only problem I have (and it may be > fixed now, we use an old version of uClibc) is that I need to put extern > "C" around the includes of the C libraries. Yup, that stuff is all fixed in uClibc these days... -Erik -- Erik B. Andersen http://codepoet-consulting.com/ --This message was written using 73% post-consumer electrons-- This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From pbleyer at embedded.cl Sat Mar 2 12:41:07 2002 From: pbleyer at embedded.cl (Pablo Bleyer Kocik) Date: Sat, 02 Mar 2002 13:41:07 -0400 Subject: [uClinux-dev] uClinux 2.5.5! In-Reply-To: <3C80E75E.113AAC57@snapgear.com> Message-ID: <5.1.0.14.0.20020302133838.020c16b0@mail.embedded.cl> This looks very nice, Greg. Congratulations! However, I didn't know that Linux changed its name to Liunx in 2.5 ;^) Cheers! --- liunx-2.5.5uc0.patch.gz 02-Mar-2002 09:51 588k GZIP compressed document --- At 12:53 AM 03-03-2002 +1000, you wrote: >Hi All, > >I have ported the uClinux 2.4.17 patches upto 2.5.5! >Enough things have changed in the kernel that it took a >little efort to get going :-) > >It is only running on one target so far, the Motorola >M5272C3 (ColdFire) board. But others should quickly follow. > >I want to track the 2.5.x series pretty closely from now on. >So expect patches against most 2.5 series kernels. >I am going to keep the patches as minimal as possible, that >will give us the best shot of getting into the mainline >linux kernels. > >So the linux-2.5.XucY series has begun! > >You can get the patch at: > > >http://www.uclinux.org/pub/uClinux/uClinux-2.5.x/linux-2.5.5uc0.patch.gz > >Regards >Greg > > > >------------------------------------------------------------------------ >Linux version 2.5.5uc0 (gerg at goober) (gcc version 2.95.3 20010315 >(release)(ColdFire patches - 20010318 from >http://fiddes.net/coldfire/)(-msep-data patches)) #255 Sat Mar 2 >23:53:04 EST 2002 > > >uClinux/COLDFIRE(m5272) >COLDFIRE port done by Greg Ungerer, gerg at snapgear.com >Flat model support (C) 1998,1999 Kenneth Albanowski, D. Jeff Dionne >On node 0 totalpages: 1024 >zone(0): 0 pages. >zone(1): 1024 pages. >zone(2): 0 pages. >Kernel command line: >Calibrating delay loop... 43.62 BogoMIPS >Memory available: 2572k/4096k RAM, 0k/0k ROM (747k kernel code, 215k >data) >kmem_create: Forcing size word alignment - task_struct >kmem_create: Forcing size word alignment - vm_area_struct >kmem_create: Forcing size word alignment - mm_struct >kmem_create: Forcing size word alignment - buffer_head >kmem_create: Forcing size word alignment - filp >Dentry-cache hash table entries: 512 (order: 0, 4096 bytes) >Inode-cache hash table entries: 512 (order: 0, 4096 bytes) >kmem_create: Forcing size word alignment - inode_cache >Mount-cache hash table entries: 512 (order: 0, 4096 bytes) >kmem_create: Forcing size word alignment - bdev_cache >kmem_create: Forcing size word alignment - cdev_cache >Buffer-cache hash table entries: 1024 (order: 0, 4096 bytes) >Page-cache hash table entries: 1024 (order: 0, 4096 bytes) >kmem_create: Forcing size word alignment - proc_inode_cache >POSIX conformance testing by UNIFIX >Linux NET4.0 for Linux 2.4 >Based upon Swansea University Computer Society NET3.039 >kmem_create: Forcing size word alignment - sock >kmem_create: Forcing size word alignment - sock_inode_cache >Initializing RT netlink socket >Starting kswapd >kmem_create: Forcing size word alignment - bio >BIO: pool of 256 setup, 13Kb (54 bytes/bio) >biovec: init pool 0, 1 entries, 12 bytes >biovec: init pool 1, 4 entries, 48 bytes >biovec: init pool 2, 16 entries, 192 bytes >biovec: init pool 3, 64 entries, 768 bytes >biovec: init pool 4, 128 entries, 1536 bytes >biovec: init pool 5, 256 entries, 3072 bytes >kmem_create: Forcing size word alignment - file lock cache >kmem_create: Forcing size word alignment - ext2_inode_cache >kmem_create: Forcing size word alignment - nfs_inode_cache >ColdFire internal UART serial driver version 1.00 >ttyS0 at 0x10000100 (irq = 73) is a builtin ColdFire UART >ttyS1 at 0x10000140 (irq = 74) is a builtin ColdFire UART >kmem_create: Forcing size word alignment - blkdev_requests >block: 64 slots per queue, batch=16 >eth0: FEC ENET Version 0.2, 00:cf:52:72:c3:01 >fec: Phy @ 0x0, type 0x001378e1 >Blkmem copyright 1998,1999 D. Jeff Dionne >Blkmem copyright 1998 Kenneth Albanowski >Blkmem 1 disk images: >0: 110E0C-16D60B [VIRTUAL 110E0C-16D60B] (RO) >RAMDISK driver initialized: 16 RAM disks of 4096K size 1024 blocksize >PPP generic driver version 2.4.1 >NET4: Linux TCP/IP 1.0 for NET4.0 >kmem_create: Forcing size word alignment - tcp_sock >kmem_create: Forcing size word alignment - udp_sock >kmem_create: Forcing size word alignment - raw4_sock >IP Protocols: ICMP, UDP, TCP, IGMP >kmem_create: Forcing size word alignment - ip_dst_cache >IP: routing cache hash table of 512 buckets, 4Kbytes >TCP: Hash tables configured (established 512 bind 512) >NET4: Unix domain sockets 1.0/SMP for Linux NET4.0. >kmem_create: Forcing size word alignment - unix_sock >VFS: Mounted root (romfs filesystem) readonly. >Freeing unused kernel memory: 24k freed (0xf8000 - 0xfd000) >Shell invoked to run file: /etc/rc >Command: hostname uClinux >Command: /bin/expand /etc/ramfs.img /dev/ram0 >Command: mount -t proc proc /proc >Command: mount -t ext2 /dev/ram0 /var >Command: mkdir /var/tmp >Command: mkdir /var/log >Command: mkdir /var/run >Command: mkdir /var/lock >Command: ifconfig lo 127.0.0.1 >Command: route aedd -net 127.0.0.0 netmaskt 255.0.0.0 lo >Command: dhhcpcd -p -a eth0 & >0: config: auto-negotiation on, 100FDX, 100HDX, 10FDX, 10HDX. >[11] >Command: cat /etc/motd >Welcome to > ____ _ _ > / __| ||_| > _ _| | | | _ ____ _ _ _ _ > | | | | | | || | _ \| | | |\ \/ / > | |_| | |__| || | | | | |_| |/ \ > | ___\____|_||_|_| |_|\____|\_/\_/ > | | > |_| > >For further information check: >http://www.uclinux.org/ > >Execution Finished, Exiting > >Sash command shell (version 1.1.1) >/> eth0: status: link up, 100MBit Half Duplex, auto-negotiation >complete. >ps > PID PORT STAT SIZE SHARED %CPU COMMAND > 1 S 18K 0K 9.5 init > 2 S 0K 0K 0.9 keventd > 3 R 0K 0K 10.3 ksoftirqd_CPU0 > 4 S 0K 0K 0.2 kswapd > 5 S 0K 0K 0.0 bdflush > 6 S 0K 0K 0.0 kupdated > 11 S 35K 0K 0.3 dhcpcd -p -a eth0 > 12 S0 R 20K 0K 1.5 /bin/sh > 13 S 11K 0K 0.9 /bin/inetd >/> > >------------------------------------------------------------------------ >Greg Ungerer -- Chief Software Wizard EMAIL: gerg at snapgear.com >Snapgear PHONE: +61 7 3279 1822 >825 Stanley St, FAX: +61 7 3279 1820 >Woolloongabba, QLD, 4102, Australia WEB: www.snapgear.com >This message resent by the uclinux-dev at uclinux.org list server >http://www.uClinux.org/ PabloBleyerKocik/ pbleyer /"Rintrah roars & shakes his fires in the burdend air; @embedded.cl/ Hungry clouds swag on the deep" ? William Blake This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From danielp at cse.unsw.edu.au Sun Mar 3 06:00:36 2002 From: danielp at cse.unsw.edu.au (Daniel Potts) Date: Sun, 3 Mar 2002 22:00:36 +1100 (EST) Subject: [uClinux-dev] 2nd UART patch for 68VZ328 In-Reply-To: <3C7B37CD.6BD9A72F@snapgear.com> Message-ID: On Tue, 26 Feb 2002, Greg Ungerer wrote: > http://www.cse.unsw.edu.au/~danielp/uclinux/patches/linux-2.4.6.0pre1-mc68328touch_screen.diff > > - touchscreen on VZ + ucDimm board, again see the .readme file and > > docs/touch_screen.README. > > Applied. I fixed up the mc68328digi_init to use the __init tag. > No need to modify misc.c to initialize it anymore. Hi, Does __init work under m68k? I've tried upgrading my compilers, yet this still does not appear to work. Is anything else needed? Cheers, Daniel This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From mark_mcchrystal at yahoo.com Sun Mar 3 16:16:31 2002 From: mark_mcchrystal at yahoo.com (Mark McChrystal) Date: Sun, 3 Mar 2002 13:16:31 -0800 (PST) Subject: [uClinux-dev] 2nd UART patch for 68VZ328 In-Reply-To: Message-ID: <20020303211631.5703.qmail@web14304.mail.yahoo.com> I ran into the same problem--I think the fix was(for modules or compiled in): module_init(mc68328digi_init); module_exit(mc6832digi_cleanup); You may also need to include linux/config.h Mark > Does __init work under m68k? I've tried upgrading my > compilers, yet this > still does not appear to work. Is anything else > needed? > > Cheers, > Daniel > > This message resent by the uclinux-dev at uclinux.org > list server http://www.uClinux.org/ __________________________________________________ Do You Yahoo!? Yahoo! Sports - sign up for Fantasy Baseball http://sports.yahoo.com This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From gerg at snapgear.com Sun Mar 3 18:48:39 2002 From: gerg at snapgear.com (Greg Ungerer) Date: Mon, 04 Mar 2002 09:48:39 +1000 Subject: [uClinux-dev] uClinux 2.5.5! References: <5.1.0.14.0.20020302133838.020c16b0@mail.embedded.cl> Message-ID: <3C82B657.2040200@snapgear.com> Oops :-) Slip of the keyboard there. Fixed now. Regards Greg Pablo Bleyer Kocik wrote: > > This looks very nice, Greg. Congratulations! > > However, I didn't know that Linux changed its name to Liunx in 2.5 ;^) > > Cheers! > > --- > liunx-2.5.5uc0.patch.gz 02-Mar-2002 09:51 588k GZIP compressed document > --- > > At 12:53 AM 03-03-2002 +1000, you wrote: > >> Hi All, >> >> I have ported the uClinux 2.4.17 patches upto 2.5.5! >> Enough things have changed in the kernel that it took a >> little efort to get going :-) >> >> It is only running on one target so far, the Motorola >> M5272C3 (ColdFire) board. But others should quickly follow. >> >> I want to track the 2.5.x series pretty closely from now on. >> So expect patches against most 2.5 series kernels. >> I am going to keep the patches as minimal as possible, that >> will give us the best shot of getting into the mainline >> linux kernels. >> >> So the linux-2.5.XucY series has begun! >> >> You can get the patch at: >> >> >> http://www.uclinux.org/pub/uClinux/uClinux-2.5.x/linux-2.5.5uc0.patch.gz >> >> Regards >> Greg >> >> >> >> ------------------------------------------------------------------------ >> Linux version 2.5.5uc0 (gerg at goober) (gcc version 2.95.3 20010315 >> (release)(ColdFire patches - 20010318 from >> http://fiddes.net/coldfire/)(-msep-data patches)) #255 Sat Mar 2 >> 23:53:04 EST 2002 >> >> >> uClinux/COLDFIRE(m5272) >> COLDFIRE port done by Greg Ungerer, gerg at snapgear.com >> Flat model support (C) 1998,1999 Kenneth Albanowski, D. Jeff Dionne >> On node 0 totalpages: 1024 >> zone(0): 0 pages. >> zone(1): 1024 pages. >> zone(2): 0 pages. >> Kernel command line: >> Calibrating delay loop... 43.62 BogoMIPS >> Memory available: 2572k/4096k RAM, 0k/0k ROM (747k kernel code, 215k >> data) >> kmem_create: Forcing size word alignment - task_struct >> kmem_create: Forcing size word alignment - vm_area_struct >> kmem_create: Forcing size word alignment - mm_struct >> kmem_create: Forcing size word alignment - buffer_head >> kmem_create: Forcing size word alignment - filp >> Dentry-cache hash table entries: 512 (order: 0, 4096 bytes) >> Inode-cache hash table entries: 512 (order: 0, 4096 bytes) >> kmem_create: Forcing size word alignment - inode_cache >> Mount-cache hash table entries: 512 (order: 0, 4096 bytes) >> kmem_create: Forcing size word alignment - bdev_cache >> kmem_create: Forcing size word alignment - cdev_cache >> Buffer-cache hash table entries: 1024 (order: 0, 4096 bytes) >> Page-cache hash table entries: 1024 (order: 0, 4096 bytes) >> kmem_create: Forcing size word alignment - proc_inode_cache >> POSIX conformance testing by UNIFIX >> Linux NET4.0 for Linux 2.4 >> Based upon Swansea University Computer Society NET3.039 >> kmem_create: Forcing size word alignment - sock >> kmem_create: Forcing size word alignment - sock_inode_cache >> Initializing RT netlink socket >> Starting kswapd >> kmem_create: Forcing size word alignment - bio >> BIO: pool of 256 setup, 13Kb (54 bytes/bio) >> biovec: init pool 0, 1 entries, 12 bytes >> biovec: init pool 1, 4 entries, 48 bytes >> biovec: init pool 2, 16 entries, 192 bytes >> biovec: init pool 3, 64 entries, 768 bytes >> biovec: init pool 4, 128 entries, 1536 bytes >> biovec: init pool 5, 256 entries, 3072 bytes >> kmem_create: Forcing size word alignment - file lock cache >> kmem_create: Forcing size word alignment - ext2_inode_cache >> kmem_create: Forcing size word alignment - nfs_inode_cache >> ColdFire internal UART serial driver version 1.00 >> ttyS0 at 0x10000100 (irq = 73) is a builtin ColdFire UART >> ttyS1 at 0x10000140 (irq = 74) is a builtin ColdFire UART >> kmem_create: Forcing size word alignment - blkdev_requests >> block: 64 slots per queue, batch=16 >> eth0: FEC ENET Version 0.2, 00:cf:52:72:c3:01 >> fec: Phy @ 0x0, type 0x001378e1 >> Blkmem copyright 1998,1999 D. Jeff Dionne >> Blkmem copyright 1998 Kenneth Albanowski >> Blkmem 1 disk images: >> 0: 110E0C-16D60B [VIRTUAL 110E0C-16D60B] (RO) >> RAMDISK driver initialized: 16 RAM disks of 4096K size 1024 blocksize >> PPP generic driver version 2.4.1 >> NET4: Linux TCP/IP 1.0 for NET4.0 >> kmem_create: Forcing size word alignment - tcp_sock >> kmem_create: Forcing size word alignment - udp_sock >> kmem_create: Forcing size word alignment - raw4_sock >> IP Protocols: ICMP, UDP, TCP, IGMP >> kmem_create: Forcing size word alignment - ip_dst_cache >> IP: routing cache hash table of 512 buckets, 4Kbytes >> TCP: Hash tables configured (established 512 bind 512) >> NET4: Unix domain sockets 1.0/SMP for Linux NET4.0. >> kmem_create: Forcing size word alignment - unix_sock >> VFS: Mounted root (romfs filesystem) readonly. >> Freeing unused kernel memory: 24k freed (0xf8000 - 0xfd000) >> Shell invoked to run file: /etc/rc >> Command: hostname uClinux >> Command: /bin/expand /etc/ramfs.img /dev/ram0 >> Command: mount -t proc proc /proc >> Command: mount -t ext2 /dev/ram0 /var >> Command: mkdir /var/tmp >> Command: mkdir /var/log >> Command: mkdir /var/run >> Command: mkdir /var/lock >> Command: ifconfig lo 127.0.0.1 >> Command: route aedd -net 127.0.0.0 netmaskt 255.0.0.0 lo >> Command: dhhcpcd -p -a eth0 & >> 0: config: auto-negotiation on, 100FDX, 100HDX, 10FDX, 10HDX. >> [11] >> Command: cat /etc/motd >> Welcome to >> ____ _ _ >> / __| ||_| >> _ _| | | | _ ____ _ _ _ _ >> | | | | | | || | _ \| | | |\ \/ / >> | |_| | |__| || | | | | |_| |/ \ >> | ___\____|_||_|_| |_|\____|\_/\_/ >> | | >> |_| >> >> For further information check: >> http://www.uclinux.org/ >> >> Execution Finished, Exiting >> >> Sash command shell (version 1.1.1) >> /> eth0: status: link up, 100MBit Half Duplex, auto-negotiation >> complete. >> ps >> PID PORT STAT SIZE SHARED %CPU COMMAND >> 1 S 18K 0K 9.5 init >> 2 S 0K 0K 0.9 keventd >> 3 R 0K 0K 10.3 ksoftirqd_CPU0 >> 4 S 0K 0K 0.2 kswapd >> 5 S 0K 0K 0.0 bdflush >> 6 S 0K 0K 0.0 kupdated >> 11 S 35K 0K 0.3 dhcpcd -p -a eth0 >> 12 S0 R 20K 0K 1.5 /bin/sh >> 13 S 11K 0K 0.9 /bin/inetd >> /> >> >> ------------------------------------------------------------------------ >> Greg Ungerer -- Chief Software Wizard EMAIL: gerg at snapgear.com >> Snapgear PHONE: +61 7 3279 1822 >> 825 Stanley St, FAX: +61 7 3279 1820 >> Woolloongabba, QLD, 4102, Australia WEB: www.snapgear.com >> This message resent by the uclinux-dev at uclinux.org list server >> http://www.uClinux.org/ > > > PabloBleyerKocik/ > pbleyer /"Rintrah roars & shakes his fires in the burdend air; > @embedded.cl/ Hungry clouds swag on the deep" -- William Blake > > This message resent by the uclinux-dev at uclinux.org list server > http://www.uClinux.org/ > -- ------------------------------------------------------------------------ Greg Ungerer -- Chief Software Wizard EMAIL: gerg at snapgear.com SnapGear PHONE: +61 7 3435 2888 825 Stanley St, FAX: +61 7 3891 3630 Woolloongabba, QLD, 4102, Australia WEB: www.snapgear.com This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From eric at corpbuilders.com Sun Mar 3 18:48:56 2002 From: eric at corpbuilders.com (Eric deLangis) Date: Sun, 3 Mar 2002 16:48:56 -0700 Subject: [uClinux-dev] Unable to build any kernels - 2.0.x/2.4.x/uClibc/uC-libc In-Reply-To: <3C5E0CCE.8F61D255@tele-ip.com> Message-ID: To All: I've just downloaded the latest source files and M68K/ColdFire Binary Tools (m68k-elf tool chain) from http://www.uclinux.org/ports/coldfire/source.html and tried to build kernels for the the Motorola 5272C3 target board for the following four configurations, and got the following results: (Default configutation options were used for "Kernel Settings" and "Vendor/User Settings", i.e. only the kernel versions and uC(-)libc options were changed between the builds.) 2.0.x/uClibc => build error = "syscalls.c:767: sys/io.h: No such file or directory MAKE[6]: *** [ioperm.o] Error 1" 2.0.x/uC-libc => load error = "unable to open initial console" 2.4.x/uClibc => build error = "../../../include/asm/errno.h:1:asm.m68k/errno.h: no such file or directory. MAKE[5] ***[closedir.o] Error 1" 2.4.x/uC-libc => load error = "unable to open initial console" Notes: I am using Red Hat 7.2, Source = 20020220, m68k Tool chain = 20020218. I have also added the "-Os" switch to the /uClinux-dist/config.arch file as specified in earlier thread for "Subject: RE:[uClinux-dev] uClibc befuddlement in dtostr.c". This switch got rid of the "dtostr.c" error, but a new error has developed as you can see above. Also, this switch was specified in the /vendor/Motorola/M5272C3/config.arch file. Why didn't this switch take effect? Why was it necessary to specify it in the top config.arch file? So far, I am unable to build any kernel using the "latest" Source and Tool versions. Am I missing something here???? Are there patches for this build? If so, where is this documented? Any suggestions will be greatly appreciated. Eric De -----Original Message----- From: owner-uclinux-dev at uclinux.org [mailto:owner-uclinux-dev at uclinux.org]On Behalf Of Bruce Paterson Sent: Sunday, February 03, 2002 9:24 PM To: uclinux-dev at uclinux.org Subject: Re: [uClinux-dev] Unable to open initial console? Tonny Tzeng wrote: > > Do you have /dev/ttyS0 or /dev/tty1 in your root filesystem? > The kernel thread 'init' (init/main.c) will try to open the above devices. > You will get the error "Unable to open initial console" If no such devices > found in the root filesystem, or got error in opening the console device. > > Best regards, > Tonny > > > ok, I've gotten to the stage of successfully mounting root romfs > partition, > > but I get the Warning: Unable to open initial console problem. > > I'm not sure how the console driver works with the serial driver with > printk, > > but I was able to get printk messages through the serial line (minicom), > > until this Warning. Well having just been there ! In my caes it turned out to be a corrupted ROMFS. It thought it had mounted ok since it could happily find the start of the filesystem, but it turned out that the part containing /bin/sh (or /bin/init) had been corrupted by the boot process. This in turn resulted from a variable setup in crt0_ram.S being overwritten by some incorrect old code (after it had been written correctly of course). Bruce ------------------------------------------------------------------------- This email and any files transmitted with it are confidential and intended solely for the use of the individual or entity to whom they are addressed. If you have received this email in error please notify the system manager. /\\\/\\\/\\\ / / Bruce Paterson VK3TJN / \\\ \\\ \\\ / / Senior Design Engineer / /\\\/\\\/\\\/ / 87 Peters Ave, Mulgrave, Vic, 3170 / / \\\ \\\ \\\ / PO Box 4112, Mulgrave, Vic, 3170, AUSTRALIA / / \\\/\\\ \\\/ Ph: +61 3 8561 4232 Fax: +61 3 9560 9055 Tele-IP Ltd. Email: bruce at tele-ip.com Icq: #32015991 WWW: http://www.tele-ip.com -------------------------------------------------------------------------- This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From gerg at snapgear.com Sun Mar 3 18:52:44 2002 From: gerg at snapgear.com (Greg Ungerer) Date: Mon, 04 Mar 2002 09:52:44 +1000 Subject: [uClinux-dev] 2nd UART patch for 68VZ328 References: Message-ID: <3C82B74C.6000306@snapgear.com> Hi Daniel, Daniel Potts wrote: > Does __init work under m68k? I've tried upgrading my compilers, yet this > still does not appear to work. Is anything else needed? Yes, it works under m68k. No compiler support is required. You need to make sure you include in your source though. Regards Greg ------------------------------------------------------------------------ Greg Ungerer -- Chief Software Wizard EMAIL: gerg at snapgear.com SnapGear PHONE: +61 7 3435 2888 825 Stanley St, FAX: +61 7 3891 3630 Woolloongabba, QLD, 4102, Australia WEB: www.snapgear.com This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From gerg at snapgear.com Sun Mar 3 19:59:56 2002 From: gerg at snapgear.com (Greg Ungerer) Date: Mon, 04 Mar 2002 10:59:56 +1000 Subject: [uClinux-dev] Re: BusyBox `msh' doesn't work correctly with vfork / uClibc References: Message-ID: <3C82C70C.8070208@snapgear.com> Hi Miles, Applied, in the distribution at least. Did Erik (maintainer of busybox) give any feedback? Regards Greg Miles Bader wrote: > The following patch to msh.c (made with very little understanding of the > source, though...) seems to fix the immediate problem: > > diff -up user/busybox/msh.c.\~1\~ user/busybox/msh.c > --- user/busybox/msh.c.~1~ Thu Feb 14 11:42:15 2002 > +++ user/busybox/msh.c Fri Mar 1 16:42:16 2002 > @@ -2604,17 +2604,15 @@ forkexec( register struct op *t, int *pi > signal(SIGQUIT, SIG_DFL); > } > if (t->type == TPAREN) > - exit(execute(t->left, NOPIPE, NOPIPE, FEXEC)); > + _exit(execute(t->left, NOPIPE, NOPIPE, FEXEC)); > if (wp[0] == NULL) > - exit(0); > + _exit(0); > > cp = rexecve(wp[0], wp, makenv()); > prs(wp[0]); prs(": "); warn(cp); > if (!execflg) > trap[0] = NULL; > - leave(); > - /* NOTREACHED */ > - exit(1); > + _exit (1); > } > > /* > > -Miles > -- ------------------------------------------------------------------------ Greg Ungerer -- Chief Software Wizard EMAIL: gerg at snapgear.com SnapGear PHONE: +61 7 3435 2888 825 Stanley St, FAX: +61 7 3891 3630 Woolloongabba, QLD, 4102, Australia WEB: www.snapgear.com This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From sheeliang at lit.org.sg Sun Mar 3 21:35:48 2002 From: sheeliang at lit.org.sg (sheeliang at lit.org.sg) Date: Mon, 4 Mar 2002 10:35:48 +0800 Subject: [uClinux-dev] uClinux on ARM7TDMI and Integrator AP: who succeeded ? In-Reply-To: <6ECEC1E214F2E342814ABB1ED10E7955040369@EXC2K01B.cselt.it>; from Fabio.Ricciato@tilab.com on Fri, Mar 01, 2002 at 05:19:11PM +0100 References: <6ECEC1E214F2E342814ABB1ED10E7955040369@EXC2K01B.cselt.it> Message-ID: <20020304103548.A6634@localhost.localdomain> On Fri, Mar 01, 2002 at 05:19:11PM +0100, Ricciato Fabio wrote: > The question is: > should I have to unset the CONFIG_SERIAL_CORE or shoud I have to find > that missing file on original kernel? I think u should patch the serial_core.c, serial_amba.c and serial_core.h from the armlinux. They worked for me. This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From miles at lsi.nec.co.jp Sun Mar 3 21:47:11 2002 From: miles at lsi.nec.co.jp (Miles Bader) Date: 04 Mar 2002 11:47:11 +0900 Subject: [uClinux-dev] Re: BusyBox `msh' doesn't work correctly with vfork / uClibc In-Reply-To: <3C82C70C.8070208@snapgear.com> References: <3C82C70C.8070208@snapgear.com> Message-ID: Greg Ungerer writes: > Applied, in the distribution at least. Did Erik (maintainer of busybox) > give any feedback? He pointed me to an old post containing a patch that purports to do a more complete job of making msh vfork-safe, but my initial attempt to apply it result in lots of failed hunks, so I'm going to have to look at it more carefully. FYI, the above-mentioend patch is: http://uclibc.org/lists/uclibc/2001-December/004491.html Thanks, -Miles -- Love is a snowmobile racing across the tundra. Suddenly it flips over, pinning you underneath. At night the ice weasels come. --Nietzsche This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From gerg at snapgear.com Sun Mar 3 23:55:49 2002 From: gerg at snapgear.com (Greg Ungerer) Date: Mon, 04 Mar 2002 14:55:49 +1000 Subject: [uClinux-dev] ColdFire patch References: <20020228113409.3e492305.carlos@embedded.cl> <3C7EBC37.69EBD83@snapgear.com> <20020301091541.72da84a9.carlos@embedded.cl> Message-ID: <3C82FE55.10303@snapgear.com> Hi Carlos, Carlos Manuel Duclos Vergara wrote: > if you look inside, there is only a change of position.... the original > config.in won't work for me (maybe some weird thing in my ws, but in > this order it works just fine) It makes more sense to have the clock frequency config strait after the CPU choice, so I have moved it to there (as your patch did). Regards Greg ------------------------------------------------------------------------ Greg Ungerer -- Chief Software Wizard EMAIL: gerg at snapgear.com SnapGear PHONE: +61 7 3435 2888 825 Stanley St, FAX: +61 7 3891 3630 Woolloongabba, QLD, 4102, Australia WEB: www.snapgear.com This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From stefan.jonsson at faltcom.se Mon Mar 4 03:14:33 2002 From: stefan.jonsson at faltcom.se (Stefan Jonsson) Date: Mon, 4 Mar 2002 09:14:33 +0100 Subject: [uClinux-dev] ARM7TDMI uCbootstrap Message-ID: Hello I have an EB40LS card from Lineo with a very "easy to use"-bootloader. I also have an extra EB40 card with the memory module AT91MEC. I wonder: is it possible to somehow retrieve the bootloader from EB40LS, or is it maybe provided somewhere? And in that case, is it "allowed" to use the "uCbootstrap" on my other card. Regards, Stefan Jonsson Student @ University of Umea, Sweden. This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From julienb at actimage.fr Mon Mar 4 04:58:06 2002 From: julienb at actimage.fr (Julien Boibessot) Date: Mon, 4 Mar 2002 10:58:06 +0100 Subject: [uClinux-dev] Problem when adding NFS support to kernel 2.4 References: Message-ID: <007801c1c363$16dbb280$8947ec95@bruker.fr> Hi all ! I'm currently having some difficulties adding NFS support to uClinux kernel 2.4. First question: why isn't it a default option ? I'm using 20011112 snapshot on Motorola 5272C3 devt board. When adding this option (NFS support), I get the following error message at compile time: "m68k-elf-ranlib libc.a m68k-elf-ranlib: libc.a: No space left on device" I've also selected MTD support option but removing it doesn't prevent me to get this error message. When I add NFS client v3 support in addition to NFS support (and MTD), I get the following error message: "/usr/local/bin/m68k-elf-ld.real: address 0x10428e of linux section .bss is not within region ram" (at the end of the compilation when ld is trying to link all obj files) Does anybody know what's happening ?? Thanks ! Julien ----- Original Message ----- From: "Stefan Jonsson" To: Sent: Monday, March 04, 2002 9:14 AM Subject: [uClinux-dev] ARM7TDMI uCbootstrap > Hello > > I have an EB40LS card from Lineo with a very "easy to use"-bootloader. I > also have an extra EB40 card with the memory module AT91MEC. > > I wonder: is it possible to somehow retrieve the bootloader from EB40LS, or > is it maybe provided somewhere? And in that case, is it "allowed" to use the > "uCbootstrap" on my other card. > > Regards, > Stefan Jonsson > Student @ University of Umea, Sweden. > > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From danielp at cse.unsw.edu.au Mon Mar 4 06:16:29 2002 From: danielp at cse.unsw.edu.au (Daniel Potts) Date: Mon, 4 Mar 2002 22:16:29 +1100 (EST) Subject: patch - 68328 touch screen (Re: [uClinux-dev] 2nd UART patch for 68VZ328) In-Reply-To: <3C82B74C.6000306@snapgear.com> Message-ID: Hi, Attached are patches to make this work. It as the lack of module_init(..). Cheers, Daniel On Mon, 4 Mar 2002, Greg Ungerer wrote: > > Hi Daniel, > > Daniel Potts wrote: > > > Does __init work under m68k? I've tried upgrading my compilers, yet this > > still does not appear to work. Is anything else needed? > > > Yes, it works under m68k. No compiler support is required. > You need to make sure you include in your > source though. > > Regards > Greg -------------- next part -------------- --- linux-2.4.x/include/asm-m68knommu/MC68VZ328.h Sun Mar 3 00:41:50 2002 +++ linux-2.4.x-new/include/asm-m68knommu/MC68VZ328.h Sun Mar 3 00:49:34 2002 @@ -251,6 +251,8 @@ #define SAM_IRQ_NUM 22 /* Sampling Timer for RTC */ #define EMIQ_IRQ_NUM 23 /* Emulator Interrupt */ +#define SPI_IRQ_NUM SPI2_IRQ_NUM + /* '328-compatible definitions */ #define SPIM_IRQ_NUM SPI_IRQ_NUM #define TMR1_IRQ_NUM TMR_IRQ_NUM @@ -259,7 +261,7 @@ /* * Here go the bitmasks themselves */ -#define IMR_MSPI (1 << SPI _IRQ_NUM) /* Mask SPI interrupt */ +#define IMR_MSPI (1 << SPI_IRQ_NUM) /* Mask SPI interrupt */ #define IMR_MTMR (1 << TMR_IRQ_NUM) /* Mask Timer interrupt */ #define IMR_MUART (1 << UART_IRQ_NUM) /* Mask UART interrupt */ #define IMR_MWDT (1 << WDT_IRQ_NUM) /* Mask Watchdog Timer interrupt */ @@ -288,7 +290,7 @@ #define ISR_ADDR 0xfffff30c #define ISR LONG_REF(ISR_ADDR) -#define ISR_SPI (1 << SPI _IRQ_NUM) /* SPI interrupt */ +#define ISR_SPI (1 << SPI_IRQ_NUM) /* SPI interrupt */ #define ISR_TMR (1 << TMR_IRQ_NUM) /* Timer interrupt */ #define ISR_UART (1 << UART_IRQ_NUM) /* UART interrupt */ #define ISR_WDT (1 << WDT_IRQ_NUM) /* Watchdog Timer interrupt */ @@ -317,7 +319,7 @@ #define IPR_ADDR 0xfffff30c #define IPR LONG_REF(IPR_ADDR) -#define IPR_SPI (1 << SPI _IRQ_NUM) /* SPI interrupt */ +#define IPR_SPI (1 << SPI_IRQ_NUM) /* SPI interrupt */ #define IPR_TMR (1 << TMR_IRQ_NUM) /* Timer interrupt */ #define IPR_UART (1 << UART_IRQ_NUM) /* UART interrupt */ #define IPR_WDT (1 << WDT_IRQ_NUM) /* Watchdog Timer interrupt */ @@ -822,6 +824,8 @@ } utx; volatile unsigned short int umisc __attribute__((packed)); volatile unsigned short int nipr __attribute__((packed)); + volatile unsigned short int hmark __attribute__((packed)); + volatile unsigned short int unused __attribute__((packed)); } m68328_uart; -------------- next part -------------- --- uClinux-2.4.x.orig/drivers/char/mc68328digi.c Tue Feb 26 18:37:40 2002 +++ linux-2.4.x/drivers/char/mc68328digi.c Mon Mar 4 22:11:02 2002 @@ -1580,3 +1580,6 @@ /* Unregister device driver */ misc_deregister(&mc68328_digi); } + +module_init(mc68328digi_init); +module_exit(mc68328digi_cleanup); From jody.pearson at cern.ch Mon Mar 4 06:42:54 2002 From: jody.pearson at cern.ch (Jody Pearson) Date: Mon, 04 Mar 2002 12:42:54 +0100 Subject: [uClinux-dev] Rom problems for Palm (Xcopilot). Message-ID: <3C835DBE.FEE174C3@cern.ch> Hi all, I have successfully compiled the 2.0.x kernel, with the userland tools, and created a rom image as below; # ls -lh romdisk* -rw-r--r-- 1 jpearson users 310k Aug 11 1999 romdisk.htdocs.tar.gz -rw-r--r-- 1 root root 1008k Mar 4 11:33 romdisk.img -rw-r--r-- 1 root root 11k Mar 4 11:33 romdisk.map -rw-r--r-- 1 root root 2.6M Mar 4 11:33 romdisk.s19 romdisk: total 684k drwxr-xr-x 2 jpearson users 4.0k Sep 13 21:38 bin drwxr-xr-x 2 jpearson users 4.0k Sep 15 00:20 dev drwxr-xr-x 2 jpearson users 17 Sep 13 21:38 etc drwxr-xr-x 2 jpearson users 103 Aug 11 1999 htdocs drwxr-xr-x 2 jpearson users 25 Jan 6 1999 lib -rwxr-xr-x 1 root root 678k Mar 4 11:32 linux drwxr-xr-x 2 jpearson users 6 Apr 24 1998 proc -rw-r--r-- 1 jpearson users 3.3k Oct 8 1998 ramfs.img lrwxrwxrwx 1 jpearson users 8 Mar 1 15:53 tmp -> var/tmp/ drwxr-xr-x 2 jpearson users 6 Jun 3 1998 usr drwxr-xr-x 2 jpearson users 6 Mar 16 1998 var # Also, please see attached romdisk.map file. The problem is that xcopilot simply hangs. Does anyone have any ideas please. Thanks Jody -------------- next part -------------- 0 ROM Disk [0xffffffff, 0xffffffff] 37777777777, sz 65535, at 0x0 1 . [0x307 , 0x0 ] 61364363, sz 16877, at 0x0 1 .. [0x307 , 0x0 ] 52520623, sz 16877, at 0x0 [link to 0x20 ] 1 etc [0x307 , 0x0 ] 70237054, sz 16877, at 0x0 2 . [0x307 , 0x0 ] 70237054, sz 16877, at 0x0 [link to 0x60 ] 2 .. [0x307 , 0x0 ] 61364363, sz 16877, at 0x0 [link to 0x20 ] 2 init [0x307 , 0x0 ] 70237055, sz 33252, at 0x3488 1 dev [0x307 , 0x0 ] 1061722, sz 16877, at 0x0 2 . [0x307 , 0x0 ] 1061722, sz 16877, at 0x0 [link to 0x3570 ] 2 .. [0x307 , 0x0 ] 61364363, sz 16877, at 0x0 [link to 0x20 ] 2 console [0x307 , 0x0 ] 1061723, sz 8612, at 0x0 2 cua0 [0x307 , 0x0 ] 1061724, sz 8612, at 0x0 2 cua1 [0x307 , 0x0 ] 1061725, sz 8612, at 0x0 2 cua2 [0x307 , 0x0 ] 1061726, sz 8612, at 0x0 2 cua3 [0x307 , 0x0 ] 1061727, sz 8612, at 0x0 2 cua4 [0x307 , 0x0 ] 1061730, sz 8612, at 0x0 2 cua5 [0x307 , 0x0 ] 1061731, sz 8612, at 0x0 2 cua6 [0x307 , 0x0 ] 1061732, sz 8612, at 0x0 2 cua7 [0x307 , 0x0 ] 1061733, sz 8612, at 0x0 2 cua8 [0x307 , 0x0 ] 1061734, sz 8612, at 0x0 2 cua9 [0x307 , 0x0 ] 1061735, sz 8612, at 0x0 2 fb0 [0x307 , 0x0 ] 1061736, sz 8612, at 0x0 2 fb1 [0x307 , 0x0 ] 1061737, sz 8612, at 0x0 2 kmem [0x307 , 0x0 ] 1061740, sz 8612, at 0x0 2 mem [0x307 , 0x0 ] 1061741, sz 8612, at 0x0 2 null [0x307 , 0x0 ] 1061742, sz 8612, at 0x0 2 ptyp0 [0x307 , 0x0 ] 1061743, sz 8612, at 0x0 2 ptyp1 [0x307 , 0x0 ] 1061744, sz 8612, at 0x0 2 ptyp2 [0x307 , 0x0 ] 1061745, sz 8612, at 0x0 2 ptyp3 [0x307 , 0x0 ] 1061746, sz 8612, at 0x0 2 ptyp4 [0x307 , 0x0 ] 1061747, sz 8612, at 0x0 2 ptyp5 [0x307 , 0x0 ] 1061750, sz 8612, at 0x0 2 ptyp6 [0x307 , 0x0 ] 1061751, sz 8612, at 0x0 2 ptyp7 [0x307 , 0x0 ] 1061752, sz 8612, at 0x0 2 ptyp8 [0x307 , 0x0 ] 1061753, sz 8612, at 0x0 2 ptyp9 [0x307 , 0x0 ] 1061754, sz 8612, at 0x0 2 ptypa [0x307 , 0x0 ] 1061755, sz 8612, at 0x0 2 ptypb [0x307 , 0x0 ] 1061756, sz 8612, at 0x0 2 ptypc [0x307 , 0x0 ] 1061757, sz 8612, at 0x0 2 ptypd [0x307 , 0x0 ] 1061760, sz 8612, at 0x0 2 ptype [0x307 , 0x0 ] 1061761, sz 8612, at 0x0 2 ptypf [0x307 , 0x0 ] 1061762, sz 8612, at 0x0 2 ram0 [0x307 , 0x0 ] 1061763, sz 24996, at 0x0 2 ram1 [0x307 , 0x0 ] 1061764, sz 24996, at 0x0 2 ram2 [0x307 , 0x0 ] 1061765, sz 24996, at 0x0 2 ram3 [0x307 , 0x0 ] 1061766, sz 24996, at 0x0 2 rom0 [0x307 , 0x0 ] 1061767, sz 24868, at 0x0 2 rom1 [0x307 , 0x0 ] 1061770, sz 24996, at 0x0 2 rom2 [0x307 , 0x0 ] 1061771, sz 24996, at 0x0 2 rom3 [0x307 , 0x0 ] 1061772, sz 24996, at 0x0 2 tty [0x307 , 0x0 ] 1061773, sz 8612, at 0x0 2 tty0 [0x307 , 0x0 ] 1061774, sz 8612, at 0x0 2 tty1 [0x307 , 0x0 ] 1061775, sz 8676, at 0x0 2 tty2 [0x307 , 0x0 ] 1061776, sz 8676, at 0x0 2 tty3 [0x307 , 0x0 ] 1061777, sz 8676, at 0x0 2 tty4 [0x307 , 0x0 ] 1073700, sz 8676, at 0x0 2 tty5 [0x307 , 0x0 ] 1073701, sz 8612, at 0x0 2 tty6 [0x307 , 0x0 ] 1073702, sz 8676, at 0x0 2 tty7 [0x307 , 0x0 ] 1073703, sz 8612, at 0x0 2 tty8 [0x307 , 0x0 ] 1073704, sz 8612, at 0x0 2 tty9 [0x307 , 0x0 ] 1073705, sz 8612, at 0x0 2 ttyS0 [0x307 , 0x0 ] 1073706, sz 8612, at 0x0 2 ttyS1 [0x307 , 0x0 ] 1073707, sz 8612, at 0x0 2 ttyS2 [0x307 , 0x0 ] 1073710, sz 8612, at 0x0 2 ttyS3 [0x307 , 0x0 ] 1073711, sz 8612, at 0x0 2 ttyS4 [0x307 , 0x0 ] 1073712, sz 8612, at 0x0 2 ttyS5 [0x307 , 0x0 ] 1073713, sz 8612, at 0x0 2 ttyS6 [0x307 , 0x0 ] 1073714, sz 8612, at 0x0 2 ttyS7 [0x307 , 0x0 ] 1073715, sz 8612, at 0x0 2 ttyS8 [0x307 , 0x0 ] 1073716, sz 8612, at 0x0 2 ttyS9 [0x307 , 0x0 ] 1073717, sz 8612, at 0x0 2 ttyp0 [0x307 , 0x0 ] 1073720, sz 8640, at 0x0 2 ttyp1 [0x307 , 0x0 ] 1073721, sz 8640, at 0x0 2 ttyp2 [0x307 , 0x0 ] 1073722, sz 8640, at 0x0 2 ttyp3 [0x307 , 0x0 ] 1073723, sz 8640, at 0x0 2 ttyp4 [0x307 , 0x0 ] 1073724, sz 8640, at 0x0 2 ttyp5 [0x307 , 0x0 ] 1073725, sz 8612, at 0x0 2 ttyp6 [0x307 , 0x0 ] 1073726, sz 8612, at 0x0 2 ttyp7 [0x307 , 0x0 ] 1073727, sz 8612, at 0x0 2 ttyp8 [0x307 , 0x0 ] 1073730, sz 8612, at 0x0 2 ttyp9 [0x307 , 0x0 ] 1073731, sz 8612, at 0x0 2 ttypa [0x307 , 0x0 ] 1073732, sz 8612, at 0x0 2 ttypb [0x307 , 0x0 ] 1073733, sz 8612, at 0x0 2 ttypc [0x307 , 0x0 ] 1073734, sz 8612, at 0x0 2 ttypd [0x307 , 0x0 ] 1073735, sz 8612, at 0x0 2 ttype [0x307 , 0x0 ] 1073736, sz 8612, at 0x0 2 ttypf [0x307 , 0x0 ] 1073737, sz 8612, at 0x0 2 zero [0x307 , 0x0 ] 1073740, sz 8612, at 0x0 2 digi [0x307 , 0x0 ] 1073741, sz 8612, at 0x0 1 bin [0x307 , 0x0 ] 10053625, sz 16877, at 0x0 2 . [0x307 , 0x0 ] 10053625, sz 16877, at 0x0 [link to 0x3fb0 ] 2 .. [0x307 , 0x0 ] 61364363, sz 16877, at 0x0 [link to 0x20 ] 2 version [0x307 , 0x0 ] 10053626, sz 33252, at 0x1f70 2 ls [0x307 , 0x0 ] 10053627, sz 33252, at 0x3160 2 ps [0x307 , 0x0 ] 10053630, sz 33252, at 0x2d5c 2 shutdown [0x307 , 0x0 ] 10053631, sz 33252, at 0x1bfc 2 free [0x307 , 0x0 ] 10053632, sz 33252, at 0x1174 2 hexdump [0x307 , 0x0 ] 10053633, sz 33252, at 0x23b0 2 df [0x307 , 0x0 ] 10053634, sz 33252, at 0x1eb8 2 pppd [0x307 , 0x0 ] 10053635, sz 33252, at 0x954c 2 sh [0x307 , 0x0 ] 10053636, sz 33252, at 0x6708 2 rootloader [0x307 , 0x0 ] 10053637, sz 33252, at 0x2bb4 2 slattach [0x307 , 0x0 ] 10053640, sz 33252, at 0x44c0 2 loattach [0x307 , 0x0 ] 10053641, sz 33252, at 0x232c 2 httpd [0x307 , 0x0 ] 10053642, sz 33252, at 0x357c 2 telnetd [0x307 , 0x0 ] 10053643, sz 33252, at 0x4ed4 2 ping [0x307 , 0x0 ] 10053644, sz 33252, at 0x5d74 2 lissa [0x307 , 0x0 ] 10053645, sz 33252, at 0x37bc 2 mount [0x307 , 0x0 ] 10053646, sz 33252, at 0xad3c 1 ramfs.img [0x307 , 0x0 ] 61364364, sz 33188, at 0xd0c 1 proc [0x307 , 0x0 ] 21263137, sz 16877, at 0x0 2 . [0x307 , 0x0 ] 21263137, sz 16877, at 0x0 [link to 0x49470 ] 2 .. [0x307 , 0x0 ] 61364363, sz 16877, at 0x0 [link to 0x20 ] 1 lib [0x307 , 0x0 ] 30043507, sz 16877, at 0x0 2 . [0x307 , 0x0 ] 30043507, sz 16877, at 0x0 [link to 0x494d0 ] 2 .. [0x307 , 0x0 ] 61364363, sz 16877, at 0x0 [link to 0x20 ] 2 liberror.txt [0x307 , 0x0 ] 30043510, sz 33188, at 0xd04 1 usr [0x307 , 0x0 ] 40671722, sz 16877, at 0x0 2 . [0x307 , 0x0 ] 40671722, sz 16877, at 0x0 [link to 0x4a260 ] 2 .. [0x307 , 0x0 ] 61364363, sz 16877, at 0x0 [link to 0x20 ] 1 var [0x307 , 0x0 ] 52520624, sz 16877, at 0x0 2 . [0x307 , 0x0 ] 52520624, sz 16877, at 0x0 [link to 0x4a2c0 ] 2 .. [0x307 , 0x0 ] 61364363, sz 16877, at 0x0 [link to 0x20 ] 1 htdocs [0x307 , 0x0 ] 61364365, sz 16877, at 0x0 2 . [0x307 , 0x0 ] 61364365, sz 16877, at 0x0 [link to 0x4a320 ] 2 .. [0x307 , 0x0 ] 61364363, sz 16877, at 0x0 [link to 0x20 ] 2 index.html [0x307 , 0x0 ] 61364366, sz 33188, at 0x520 2 logo.gif [0x307 , 0x0 ] 61364367, sz 33188, at 0x13dd 2 rt-control-i-net-ready-logo-m.gif [0x307 , 0x0 ] 61364370, sz 33188, at 0x8ae 2 uc68ez328-anim.gif [0x307 , 0x0 ] 61364371, sz 33188, at 0x5c5a 1 tmp [0x307 , 0x0 ] 61364372, sz 41471, at 0x8 1 linux [0x307 , 0x0 ] 61565020, sz 33261, at 0xa9988 From gerg at snapgear.com Mon Mar 4 07:36:51 2002 From: gerg at snapgear.com (gerg) Date: Mon, 04 Mar 2002 22:36:51 +1000 Subject: patch - 68328 touch screen (Re: [uClinux-dev] 2nd UART patch for68VZ328) References: Message-ID: <3C836A63.C2F6EAC8@snapgear.com> Hi Daniel, Applied. Thanks Greg Daniel Potts wrote: > > Hi, > > Attached are patches to make this work. > > It as the lack of module_init(..). > > Cheers, > Daniel > > On Mon, 4 Mar 2002, Greg Ungerer wrote: > > > > > Hi Daniel, > > > > Daniel Potts wrote: > > > > > Does __init work under m68k? I've tried upgrading my compilers, yet this > > > still does not appear to work. Is anything else needed? > > > > > > Yes, it works under m68k. No compiler support is required. > > You need to make sure you include in your > > source though. > > > > Regards > > Greg > > ------------------------------------------------------------------------ > Name: linux-2.4.17-uc0_MC68VZ328_h.diff > linux-2.4.17-uc0_MC68VZ328_h.diff Type: Plain Text (TEXT/PLAIN) > Encoding: BASE64 > > Name: linux-2.4.17-uc0_touch_screen.diff > linux-2.4.17-uc0_touch_screen.diff Type: Plain Text (TEXT/PLAIN) > Encoding: BASE64 -- ------------------------------------------------------------------------ Greg Ungerer -- Chief Software Wizard EMAIL: gerg at snapgear.com Snapgear PHONE: +61 7 3279 1822 825 Stanley St, FAX: +61 7 3279 1820 Woolloongabba, QLD, 4102, Australia WEB: www.snapgear.com This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From gerg at snapgear.com Mon Mar 4 08:05:41 2002 From: gerg at snapgear.com (gerg) Date: Mon, 04 Mar 2002 23:05:41 +1000 Subject: [uClinux-dev] Problem when adding NFS support to kernel 2.4 References: <007801c1c363$16dbb280$8947ec95@bruker.fr> Message-ID: <3C837125.445F6F0E@snapgear.com> hi Julien, Julien Boibessot wrote: > I'm currently having some difficulties adding NFS support to uClinux kernel > 2.4. > First question: why isn't it a default option ? No reason. Whoever set up the default wasn't using it... > I'm using 20011112 snapshot on Motorola 5272C3 devt board. > > When adding this option (NFS support), I get the following error message at > compile time: > "m68k-elf-ranlib libc.a > m68k-elf-ranlib: libc.a: No space left on device" > I've also selected MTD support option but removing it doesn't prevent me to > get this error message. > > When I add NFS client v3 support in addition to NFS support (and MTD), I get > the following error message: > "/usr/local/bin/m68k-elf-ld.real: address 0x10428e of linux section .bss is > not within region ram" > (at the end of the compilation when ld is trying to link all obj files) This is the memory size set too small in the linker script. Get the latest from cvs, cvs.uclinux.org, it is fixed there. You want the file: ~arch/m68knommu/platform/5272/MOTOROLA/ram.ld Regards Greg ------------------------------------------------------------------------ Greg Ungerer -- Chief Software Wizard EMAIL: gerg at snapgear.com Snapgear PHONE: +61 7 3279 1822 825 Stanley St, FAX: +61 7 3279 1820 Woolloongabba, QLD, 4102, Australia WEB: www.snapgear.com This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From julienb at actimage.fr Mon Mar 4 08:56:03 2002 From: julienb at actimage.fr (Julien Boibessot) Date: Mon, 4 Mar 2002 14:56:03 +0100 Subject: [uClinux-dev] uClinux 5272 QSPI support in 20011112 snapshot References: <007801c1c363$16dbb280$8947ec95@bruker.fr> Message-ID: <002f01c1c384$5291c9b0$8947ec95@bruker.fr> Hi all ! I've seen that there seems to have a 5272 qspi support for Coldfire in uClinux kernel (/drivers/char/mcf_qspi.c). I'd like to know if the author of this driver could say us a little bit more about his work. Why is this driver not present in the 2.4.x tree ?? Thanks !! Julien ----- Original Message ----- From: "Julien Boibessot" To: Sent: Monday, March 04, 2002 10:58 AM Subject: [uClinux-dev] Problem when adding NFS support to kernel 2.4 > Hi all ! > > I'm currently having some difficulties adding NFS support to uClinux kernel > 2.4. > First question: why isn't it a default option ? > > I'm using 20011112 snapshot on Motorola 5272C3 devt board. > > When adding this option (NFS support), I get the following error message at > compile time: > "m68k-elf-ranlib libc.a > m68k-elf-ranlib: libc.a: No space left on device" > I've also selected MTD support option but removing it doesn't prevent me to > get this error message. > > When I add NFS client v3 support in addition to NFS support (and MTD), I get > the following error message: > "/usr/local/bin/m68k-elf-ld.real: address 0x10428e of linux section .bss is > not within region ram" > (at the end of the compilation when ld is trying to link all obj files) > > Does anybody know what's happening ?? > > Thanks ! > Julien > > > ----- Original Message ----- > From: "Stefan Jonsson" > To: > Sent: Monday, March 04, 2002 9:14 AM > Subject: [uClinux-dev] ARM7TDMI uCbootstrap > > > > Hello > > > > I have an EB40LS card from Lineo with a very "easy to use"-bootloader. I > > also have an extra EB40 card with the memory module AT91MEC. > > > > I wonder: is it possible to somehow retrieve the bootloader from EB40LS, > or > > is it maybe provided somewhere? And in that case, is it "allowed" to use > the > > "uCbootstrap" on my other card. > > > > Regards, > > Stefan Jonsson > > Student @ University of Umea, Sweden. > > > > This message resent by the uclinux-dev at uclinux.org list server > http://www.uClinux.org/ > > > > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From gerg at snapgear.com Mon Mar 4 09:25:47 2002 From: gerg at snapgear.com (gerg) Date: Tue, 05 Mar 2002 00:25:47 +1000 Subject: [uClinux-dev] linux-2.5.5uc1 Message-ID: <3C8383EB.AD322635@snapgear.com> Hey ALl, I have put a linux-2.5.5uc1 patch at: http://www.uclinux.org/pub/uCliunx/uClinux-2.5.x/linux-2.5.5uc1.patch.gz Some minor fixups. Most 68k/ColdFire targets should be good now (although I have still only tested the M5272C3 board). Regards Greg ------------------------------------------------------------------------ Greg Ungerer -- Chief Software Wizard EMAIL: gerg at snapgear.com Snapgear PHONE: +61 7 3279 1822 825 Stanley St, FAX: +61 7 3279 1820 Woolloongabba, QLD, 4102, Australia WEB: www.snapgear.com This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From bkuhn at lineo.com Mon Mar 4 09:28:34 2002 From: bkuhn at lineo.com (Bernhard Kuhn) Date: Mon, 04 Mar 2002 15:28:34 +0100 Subject: [uClinux-dev] Rom problems for Palm (Xcopilot). References: <3C835DBE.FEE174C3@cern.ch> Message-ID: <3C838492.FB6439DD@lineo.com> Jody Pearson schrieb: > I have successfully compiled the 2.0.x kernel, with the userland tools, > and created a rom image as below; [...] > -rw-r--r-- 1 root root 1008k Mar 4 11:33 romdisk.img You probably will have to align the rom image to exactly 1024 Kbyte with i.e dd if=/dev/zero if=zero.img bs=1024 count=1024 cat romdisk.img zero.img | dd of=image.tom bs=1024 count=1024 best regards Bernhard -- Bernhard Kuhn, Software Engineer, Lineo Inc. (Where Open Meets Smart) This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From jody.pearson at cern.ch Mon Mar 4 10:14:32 2002 From: jody.pearson at cern.ch (Jody Pearson) Date: Mon, 04 Mar 2002 16:14:32 +0100 Subject: [uClinux-dev] Rom problems for Palm (Xcopilot). References: <3C835DBE.FEE174C3@cern.ch> <3C838492.FB6439DD@lineo.com> Message-ID: <3C838F58.5A6AB3BE@cern.ch> Hi Bernhard Well, I tried this, and the problem changes (slightly!) Instead of E - Error loading ROM file I get a core dump from xcopilot. I think there is some error in the ${TOPDIR}/arch/m68knommu/platform/PalmVx/fixup.pl script. But I'm no perl wizard 8^) Thanks Jody > > > I have successfully compiled the 2.0.x kernel, with the userland tools, > > and created a rom image as below; > > [...] > > -rw-r--r-- 1 root root 1008k Mar 4 11:33 romdisk.img > > You probably will have to align the rom image to exactly > 1024 Kbyte with i.e > > dd if=/dev/zero if=zero.img bs=1024 count=1024 > cat romdisk.img zero.img | dd of=image.tom bs=1024 count=1024 > > best regards > > Bernhard > > -- > Bernhard Kuhn, Software Engineer, Lineo Inc. (Where Open Meets Smart) > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From jody.pearson at cern.ch Mon Mar 4 10:29:06 2002 From: jody.pearson at cern.ch (Jody Pearson) Date: Mon, 04 Mar 2002 16:29:06 +0100 Subject: [uClinux-dev] Linux rom from 2.0.40 dist... Message-ID: <3C8392C2.658AB726@cern.ch> Hi all, I'm getting teasingly closer to getting this thing to work.... I can now generate a linux.rom which is exactly 1024k, the fixup.pl gives me this. If i invoke xcopilot with the rom like this # cp linux.rom ~/.xcopilot/pilot.rom # xcopilot I get; E - Error loading ROM file If i invoke it with # xcopilot -nocheck I get the screen full of random pixels and; Bus error: wrote a long to undefined memory address 0x003f8186 Bus error: wrote a word to undefined memory address 0x003f8184 Illegal instruction: 000f Bus error: wrote a long to undefined memory address 0x003f8180 Bus error: wrote a word to undefined memory address 0x003f817e Bus error: wrote a long to undefined memory address 0x003f817a Bus error: wrote a word to undefined memory address 0x003f8178 Illegal instruction: 000f Bus error: wrote a long to undefined memory address 0x003f8174 Bus error: wrote a word to undefined memory address 0x003f8172 Bus error: wrote a long to undefined memo Scrolling past at a rate of knots. Any ideas of this ? Thanks for your support. Jody Pearson This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From olivierp at actimage.fr Mon Mar 4 12:30:15 2002 From: olivierp at actimage.fr (Olivier Pierrat) Date: Mon, 04 Mar 2002 18:30:15 +0100 Subject: [uClinux-dev] kernel version ,tools chain versions, ... Message-ID: <3C83AF27.3090705@actimage.fr> Hi, First, I am a little confused with uClinux and the different distro, and tools chains version.Is there somewhere a document discribing what is include in the kernel version 2.4.x, because some fonctionality existe in the 2.4.x but not in the 2.0.x, or not the same way ? It seems the best way is to use a 2.4.x kernel, but what is the influence of the libc version ? I use the uC-libc, is it compatible with the kernel 2.4.X, because I have a lot of problems to compile a kernel 2.4.x ? I think the uClibc is newer, and it's API is compatible with the POSIX, is it right ? Is it better to use the uClibc with the 2.4.x version of the kernel ? Thanks, for your explanations. Olivier. This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From mark_mcchrystal at yahoo.com Mon Mar 4 15:49:02 2002 From: mark_mcchrystal at yahoo.com (Mark McChrystal) Date: Mon, 4 Mar 2002 12:49:02 -0800 (PST) Subject: [uClinux-dev] Linux rom from 2.0.40 dist... In-Reply-To: <3C8392C2.658AB726@cern.ch> Message-ID: <20020304204902.38488.qmail@web14307.mail.yahoo.com> For 2.4.x I had to use David McCullough's hacked xcopilot to get things working and I think I also made some changes to the linker script to make the system stable. I can send you a copy of my xcopilot which is a slightly hacked version of David M's xcopilot if you're interested. Mark --- Jody Pearson wrote: > Hi all, > > I'm getting teasingly closer to getting this thing > to work.... > > I can now generate a linux.rom which is exactly > 1024k, the fixup.pl > gives me this. > > If i invoke xcopilot with the rom like this > > # cp linux.rom ~/.xcopilot/pilot.rom > # xcopilot > > I get; > > E - Error loading ROM file > > If i invoke it with > > # xcopilot -nocheck I get the screen full of random > pixels and; > > Bus error: wrote a long to undefined memory address > 0x003f8186 > Bus error: wrote a word to undefined memory address > 0x003f8184 > Illegal instruction: 000f > Bus error: wrote a long to undefined memory address > 0x003f8180 > Bus error: wrote a word to undefined memory address > 0x003f817e > Bus error: wrote a long to undefined memory address > 0x003f817a > Bus error: wrote a word to undefined memory address > 0x003f8178 > Illegal instruction: 000f > Bus error: wrote a long to undefined memory address > 0x003f8174 > Bus error: wrote a word to undefined memory address > 0x003f8172 > Bus error: wrote a long to undefined memo > > Scrolling past at a rate of knots. > > Any ideas of this ? > > Thanks for your support. > > Jody Pearson > > > This message resent by the uclinux-dev at uclinux.org > list server http://www.uClinux.org/ __________________________________________________ Do You Yahoo!? Yahoo! Sports - sign up for Fantasy Baseball http://sports.yahoo.com This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From miles at lsi.nec.co.jp Mon Mar 4 19:57:03 2002 From: miles at lsi.nec.co.jp (Miles Bader) Date: 05 Mar 2002 09:57:03 +0900 Subject: [uClinux-dev] Re: linux-2.5.5uc1 In-Reply-To: <3C8383EB.AD322635@snapgear.com> References: <3C8383EB.AD322635@snapgear.com> Message-ID: gerg writes: > I have put a linux-2.5.5uc1 patch at: What's the mainline uClinux CVS going to do with 2.4 vs. 2.5 (i.e., is it going to continue following 2.4.x)? Thanks, -Miles -- I have seen the enemy, and he is us. -- Pogo This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From gerg at snapgear.com Mon Mar 4 22:08:47 2002 From: gerg at snapgear.com (Greg Ungerer) Date: Tue, 05 Mar 2002 13:08:47 +1000 Subject: [uClinux-dev] Re: linux-2.5.5uc1 References: <3C8383EB.AD322635@snapgear.com> Message-ID: <3C8436BF.5090403@snapgear.com> Hi Miles, Miles Bader wrote: > gerg writes: > >>I have put a linux-2.5.5uc1 patch at: >> > > What's the mainline uClinux CVS going to do with 2.4 vs. 2.5 > (i.e., is it going to continue following 2.4.x)? CVS will continue with 2.4.x (just as we maintain the 2.0.x). We should keep that as our stable base. 2.5.x is about tracking mainline linux more closely, with a view to getting uClinux into it. It may be a pipe dream on our part. Who knows. But there is no way we can get in without staying upto date with each 2.5.x release. Regards Greg ------------------------------------------------------------------------ Greg Ungerer -- Chief Software Wizard EMAIL: gerg at snapgear.com SnapGear PHONE: +61 7 3435 2888 825 Stanley St, FAX: +61 7 3891 3630 Woolloongabba, QLD, 4102, Australia WEB: www.snapgear.com This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From miles at lsi.nec.co.jp Mon Mar 4 23:21:54 2002 From: miles at lsi.nec.co.jp (Miles Bader) Date: 05 Mar 2002 13:21:54 +0900 Subject: [uClinux-dev] Re: linux-2.5.5uc1 In-Reply-To: <3C8436BF.5090403@snapgear.com> References: <3C8383EB.AD322635@snapgear.com> <3C8436BF.5090403@snapgear.com> Message-ID: Greg Ungerer writes: > 2.5.x is about tracking mainline linux more closely, with a view to > getting uClinux into it. It may be a pipe dream on our part. Who > knows. But there is no way we can get in without staying upto date > with each 2.5.x release. What I wonder is if I should bother making my v850 port work on 2.5, as the changes from 2.4 seem not to be entirely trivial. Will there be a separate publicly accessable uClinux CVS tree for 2.5? Thanks, -Miles -- P.S. All information contained in the above letter is false, for reasons of military security. This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From shiyas at globaledgesoft.com Mon Mar 4 23:39:25 2002 From: shiyas at globaledgesoft.com (shiaz) Date: Tue, 5 Mar 2002 10:09:25 +0530 Subject: [uClinux-dev] aplio toolchain In-Reply-To: <003001c1c0c0$06e7e880$1700a8c0@wiscore.com> References: <02022812264802.04902@shiaz> <02022821095000.23146@shiaz> <003001c1c0c0$06e7e880$1700a8c0@wiscore.com> Message-ID: <02030510092500.00976@shiaz> I've downloaded the gcc-3.0.1 and binutils and newlibc. I've built the toolchain and tried compiling some sample programs trying some system calls, but to my surprise when using vfork this error it gives. /tmp/ccRU5l0N.o: In function `main': /tmp/ccRU5l0N.o(.text+0x38): undefined reference to `vfork' collect2: ld returned 1 exit status Please comment on this. Thankz in advance.. regards shiaz On Friday 01 March 2002 06:54 am, you wrote: > Can you compile the "Hello world" sample program by typing the > following command? > arm-elf-gcc -o hello hello.c > If the C runtime has been installed properly, the ld should be invoked > by the gcc to create the final executable. Otherwise, you have to link > the object files with runtime explicitly. As I know, the "C compiler > cannot create executables." error was caused by the above test > command. > > Best regards, > Tonny > > > I've given make install in gcc dir... > > In my install directory crt?.o is installed . After this i tried to make > > gcc > > > in build dir it's givinme this error.. > > > > On Thursday 28 February 2002 01:39 pm, you wrote: > > > The installation of GCC will try to build a dummy executable by using > > > $(CC) directly, but it is impossible until the crt?.o have been > > installed. > > > > So you could jump into the gcc directory and 'make install', then > > compile > > > > and install the C library, then re-compile the gcc toolchain should > > > pass the error. > > > > > > Best regards, > > > Tonny > > > > > > > Can anyone tell me the cause of the error while building the > > > > toolchain for arm uclinux using the aplio Armtools(got from > > > > ftp://ftp.wiscore.com/ARMTools/Aplio/arm-tools-src.tgz.) > > > > Configuring in arm-elf/libf2c > > > > loading cache ./config.cache > > > > checking if compiler f771 has been built... yes > > > > checking for gcc... /home/shiaz/Aplio/apliotools/build-egcs/gcc/xgcc > > > > -B/home/shiaz/Aplio/apliotools/build-egcs/gcc/ > > > > -B/home/shiaz/Aplio/apliotools/build-egcs/arm-elf/newlib/ > > > > -B/home/shiaz/Aplio/apliotools/instdir/arm-elf/bin/ -idirafter > > > > /home/shiaz/Aplio/apliotools/build-egcs/arm-elf/newlib/targ-include > > > > -idirafter /home/shiaz/Aplio/apliotools/egcs/newlib/libc/include > > > > -nostdinc checking whether the C compiler > > > > (/home/shiaz/Aplio/apliotools/build-egcs/gcc/xgcc > > > > -B/home/shiaz/Aplio/apliotools/build-egcs/gcc/ > > > > -B/home/shiaz/Aplio/apliotools/build-egcs/arm-elf/newlib/ > > > > -B/home/shiaz/Aplio/apliotools/instdir/arm-elf/bin/ -idirafter > > > > /home/shiaz/Aplio/apliotools/build-egcs/arm-elf/newlib/targ-include > > > > -idirafter > > > > > > /home/shiaz/Aplio/apliotools/egcs/newlib/libc/include -nostdinc -g > > > > > > > -O2 ) works... no > > > > > > > > configure: error: installation or configuration problem: C compiler > > > > cannot create executables. > > > > > > This message resent by the uclinux-dev at uclinux.org list server > > > http://www.uClinux.org/ > > > > This message resent by the uclinux-dev at uclinux.org list server > > http://www.uClinux.org/ > > > This message resent by the uclinux-dev at uclinux.org list server > http://www.uClinux.org/ This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From shiyas at globaledgesoft.com Mon Mar 4 23:45:13 2002 From: shiyas at globaledgesoft.com (shiaz) Date: Tue, 5 Mar 2002 10:15:13 +0530 Subject: [uClinux-dev] arm-uclinux Message-ID: <02030510151301.00976@shiaz> Hi list, I've been trying to build uclinux kernel 4 arm7 from last 3 days in vain. These r the steps i followed to build. Built the gcc-3.0.1, binutils & newlibc Got the arm-uclinux kernel from uclinux.org. I've built the toolchain successfully. Now i'm stuck at building the above kernel. It's giving me broken build. the errors are as follows ====================ERROR======================================== make[1]: Entering directory `/home/shiaz/arm/linux/fs' if [ -n "bad_inode.c binfmt_aout.c binfmt_elf.c binfmt_em86.c binfmt_flat.c binfmt_java.c binfmt_script.c block_dev.c buffer.c dcache.c devices.c dquot.c exec.c fcntl.c fifo.c filesystems.c file_table.c inode.c ioctl.c locks.c namei.c nls.c nls_cp437.c nls_cp737.c nls_cp775.c nls_cp850.c nls_cp852.c nls_cp855.c nls_cp857.c nls_cp860.c nls_cp861.c nls_cp862.c nls_cp863.c nls_cp864.c nls_cp865.c nls_cp866.c nls_cp869.c nls_cp874.c nls_iso8859_15.c nls_iso8859_1.c nls_iso8859_2.c nls_iso8859_3.c nls_iso8859_4.c nls_iso8859_5.c nls_iso8859_6.c nls_iso8859_7.c nls_iso8859_8.c nls_iso8859_9.c nls_koi8_r.c noquot.c open.c pipe.c readdir.c read_write.c select.c stat.c super.c" ]; then \ /home/shiaz/arm/linux/scripts/mkdep *.[chS] > .depend; fi binfmt_aout.c needs config but has not included config file make[1]: *** [fastdep] Error 135 make[1]: Leaving directory `/home/shiaz/arm/linux/fs' make: *** [dep-files] Error 2 This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From kj_lin at accton.com.tw Tue Mar 5 01:29:54 2002 From: kj_lin at accton.com.tw (kj_lin at accton.com.tw) Date: Tue, 5 Mar 2002 14:29:54 +0800 Subject: [uClinux-dev] Is it the problem of C++ library or kernel? Message-ID: Hi all, My kernel version is uclinux-2.4.10-u2 for ARM and the toolchain(arm-elf) is from http://www.uclinux.org/pub/uClinux/m68k-elf-tools/. I encountered a weird problem on C++ code. I tried to test a simple C++ program, Hello.cxx, like this: #include int main() { cout<<"Hello World!\n"; return 0; } But the result is as following: #./hello fault-common.c 386 fault-common.c 386 fault-common.c 386 .. .. .. .. Does any body know what is going on? Is it the problem of C++ library or my uclinux kernel? Thanks! KJ This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From gerg at snapgear.com Tue Mar 5 02:28:25 2002 From: gerg at snapgear.com (Greg Ungerer) Date: Tue, 05 Mar 2002 17:28:25 +1000 Subject: [uClinux-dev] Re: linux-2.5.5uc1 References: <3C8383EB.AD322635@snapgear.com> <3C8436BF.5090403@snapgear.com> Message-ID: <3C847399.8070600@snapgear.com> Hi Miles, Miles Bader wrote: > Greg Ungerer writes: > >>2.5.x is about tracking mainline linux more closely, with a view to >>getting uClinux into it. It may be a pipe dream on our part. Who >>knows. But there is no way we can get in without staying upto date >>with each 2.5.x release. >> > > What I wonder is if I should bother making my v850 port work on 2.5, as > the changes from 2.4 seem not to be entirely trivial. > > Will there be a separate publicly accessable uClinux CVS tree for 2.5? No plan to do this. I expect 2.5 will be a moving target for some time. I would like to see all architecturs in it. So if don't mind some regular effort to get it working it would be good to have v850 support in there. Regards Greg ------------------------------------------------------------------------ Greg Ungerer -- Chief Software Wizard EMAIL: gerg at snapgear.com SnapGear PHONE: +61 7 3435 2888 825 Stanley St, FAX: +61 7 3891 3630 Woolloongabba, QLD, 4102, Australia WEB: www.snapgear.com This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From miles at lsi.nec.co.jp Tue Mar 5 02:57:35 2002 From: miles at lsi.nec.co.jp (Miles Bader) Date: 05 Mar 2002 16:57:35 +0900 Subject: [uClinux-dev] Re: BusyBox `msh' doesn't work correctly with vfork / uClibc In-Reply-To: <20020301080856.GA26459@codepoet.org> References: <20020301080856.GA26459@codepoet.org> Message-ID: Erik Andersen writes: > Could you check out this patch from Stefan Soucek > http://uclibc.org/lists/uclibc/2001-December/004491.html ... > Anyway, if you try it and it does indeed work (or if you have the > time to improve it), I'll be glad to include it for 0.60.3 on your > recommendation. I looked at that patch, and found that most of it had already been applied in the version of busybox/msh.c that I have (which is from uClinux-dist-20020220, and claims to be version 0.60.0). Those parts seems sane. However, Stefan's patch doesn't address the problem with `exit' that I reported, so my patch (that changes things to use `_exit') or something similar is still necessary. The parts of Stefan's patch which _aren't_ in the current sources are: (1) Some ifdefs to allow either fork or vfork to be used; the current sources just always use vfork, which seems fine to me. (2) A change to the `grave' function to avoid problems with vfork; since that function appears to do pretty complicated things within the child, Stefan's patch just does (where cp is the command): execlp("/bin/sh", "/bin/sh", "-c", cp, NULL); Not sure if that works, but the current code in `grave' is certain to screw things up when used with vfork. So, I'd recommend: Apply my earlier patch to use _exit Think about doing something with the `grave' function similar to what Stefan did. -Miles -- 80% of success is just showing up. --Woody Allen This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From sureshsingh.keisam at analog.com Tue Mar 5 07:33:52 2002 From: sureshsingh.keisam at analog.com (Suresh Singh K.) Date: Tue, 5 Mar 2002 18:03:52 +0530 Subject: [uClinux-dev] Porting Proprietary TCP/IP stack to uclinux Message-ID: <009001c1c442$03021bc0$d7204789@skeisamd01> Hi All, I am currently on a project which requires to port our own TCP/IP to uclinux. I want to know whether it's possible to PLUG IN A PROPRIETORY TCP/IP into the uclinux kernel without voilating the GPL conditions ?. I want to use the uclinux kernel (without it's native TCP/IP code) with my company's TCP/IP . And my company doesn't want it's "IP code & related modules like QOS,SNDCP " to come under the GPL conditions. I also want to know if I can plug our proprietory "IP code " just below the linux native TCP/IP code without violating the GPL conditions of the linux kernel ? If our "IP Code" only access ( doesn't modify) the interface (both function & data ) provided by the native linux TCP/IP code , Does it voilates the GPL ?. In linux, every ethernet receive interrupt call calls "netif_rx( )" passing the receive packet inside an sk_buff . Can the ethernet driver access ( doesn't modify) other linux TCP/IP interfaces (like the functions in dev.c ) other than the registeration , module interfaces & net_device structure without violating the GPL conditions ?. That's, I want my driver code OR any code below "IP of linux" to remain as PROPRIETORY. But still using the various functions & datas provided by the "linux TCP/IP" ( only reading & not modifying). I would be very thankful to anyone who can Guide me on the above issues . Thanks in Advance, :-) Regards, Suresh Singh K. ( sureshsingh.keisam at analog.com ) -------------- next part -------------- An HTML attachment was scrubbed... URL: From stefan.jonsson at faltcom.se Tue Mar 5 08:37:59 2002 From: stefan.jonsson at faltcom.se (Stefan Jonsson) Date: Tue, 5 Mar 2002 14:37:59 +0100 Subject: [uClinux-dev] CGI and uClinux in ARM7-EB40LS ? Message-ID: Hello list, 1) I am in need of running CGI-scripts on a webserver on EB40LS. That it will not be I Perl but I for instance C, I have understood. But what do I need to make this possible? What web-server (if it is relevant) and do I need a /home/httpd/cgi-bin ... or something? And where it might be possible to find this (in that case) needed things. 2) I always make an image.bin and download that to target. Is it possible to just download an application and get that to work as expected? I mean, how do I compile this application? How do I safely download that to my card to get it to work? I am most greatful for any answers to my problem. Regards, Stefan Jonsson, University of Umea, Sweden. This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From andersen at codepoet.org Tue Mar 5 09:23:29 2002 From: andersen at codepoet.org (Erik Andersen) Date: Tue, 5 Mar 2002 07:23:29 -0700 Subject: [uClinux-dev] Unable to build any kernels - 2.0.x/2.4.x/uClibc/uC-libc In-Reply-To: References: <3C5E0CCE.8F61D255@tele-ip.com> Message-ID: <20020305142329.GA2009@codepoet.org> On Sun Mar 03, 2002 at 04:48:56PM -0700, Eric deLangis wrote: > 2.0.x/uClibc => build error = "syscalls.c:767: sys/io.h: No such file or > directory MAKE[6]: *** [ioperm.o] Error 1" Fixed now in CVS. Not all arches have sys/io.h, and I was able to simply eliminate it... -Erik -- Erik B. Andersen http://codepoet-consulting.com/ --This message was written using 73% post-consumer electrons-- This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From paul.robertson at dgw.co.uk Tue Mar 5 09:58:26 2002 From: paul.robertson at dgw.co.uk (Paul Robertson) Date: Tue, 5 Mar 2002 14:58:26 -0000 Subject: [uClinux-dev] Porting Proprietary TCP/IP stack to uclinux References: <009001c1c442$03021bc0$d7204789@skeisamd01> Message-ID: <038c01c1c456$345ff520$741e4c0a@paulr> As I understand it (read Linus' preamble to the GPL in COPYING in the root of the kernel distribution), you can write closed kernel modules on the condition that they only make kernel calls through the established APIs. nVidia do this for their closed source video card drivers. If you modify kernel code in any way, that modification is covered by the GPL. I should point out that I am not a legal guy, so all I can tell you is my understanding of the situation. You might try the linux kernel mailing list FAQ: http://www.tux.org/lkml/ There are a number of archives linked from there (see FAQ question 4). Paul. ----- Original Message ----- From: "Suresh Singh K." To: Cc: ; Sent: Tuesday, March 05, 2002 12:33 PM Subject: [uClinux-dev] Porting Proprietary TCP/IP stack to uclinux Hi All, I am currently on a project which requires to port our own TCP/IP to uclinux. I want to know whether it's possible to PLUG IN A PROPRIETORY TCP/IP into the uclinux kernel without voilating the GPL conditions ?. I want to use the uclinux kernel (without it's native TCP/IP code) with my company's TCP/IP . And my company doesn't want it's "IP code & related modules like QOS,SNDCP " to come under the GPL conditions. I also want to know if I can plug our proprietory "IP code " just below the linux native TCP/IP code without violating the GPL conditions of the linux kernel ? If our "IP Code" only access ( doesn't modify) the interface (both function & data ) provided by the native linux TCP/IP code , Does it voilates the GPL ?. In linux, every ethernet receive interrupt call calls "netif_rx( )" passing the receive packet inside an sk_buff . Can the ethernet driver access ( doesn't modify) other linux TCP/IP interfaces (like the functions in dev.c ) other than the registeration , module interfaces & net_device structure without violating the GPL conditions ?. That's, I want my driver code OR any code below "IP of linux" to remain as PROPRIETORY. But still using the various functions & datas provided by the "linux TCP/IP" only reading & not modifying). I would be very thankful to anyone who can Guide me on the above issues . Thanks in Advance, :-) Regards, Suresh Singh K. ( sureshsingh.keisam at analog.com ) This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From mathias.fritzson at mecel.se Tue Mar 5 11:30:47 2002 From: mathias.fritzson at mecel.se (mathias.fritzson at mecel.se) Date: Tue, 5 Mar 2002 17:30:47 +0100 Subject: [uClinux-dev] How do I get exception handling vectors > ramvec Message-ID: <200203051627.g25GR0J13133@clinux.org> Hello We are trying to build our kernel image directly for ram but the .ramvec ends up empty. And thats no good right =) Here's what we do. The .ld file seems correct (at the bottom) and when I look at the linux image in the linux directory it's all empty from 800000 to 800400. screening through the diffrent start up files and there is no copying of the .ramvec there I use this objcopy lines to strip the linux file $(CROSS_COMPILE)objcopy -O srec \ $(ROOTDIR)/$(LINUXDIR)/linux $(IMAGEDIR)/image.s19 but no trace of any .ramvec anywhere The interrupt handlers are defined in entry.S but where does the vectors go and how do I get them to the .ramvec ? Should there be any sections in the head.S or how do we do it ?? TIA /Mathias Fritzson And the .ld file... ----->8------------------------------------- MEMORY { romvec : ORIGIN = 0x00000000, LENGTH = 0x000400 flash : ORIGIN = 0x00000400, LENGTH = 0x080000 - 0x00400 ramvec : ORIGIN = 0x00800000, LENGTH = 0x000400 ram : ORIGIN = 0x00800400, LENGTH = 0x100000 - 0x00400 eram : ORIGIN = 0x00900000, LENGTH = 0 } SECTIONS { .romvec : { _romvec = . ; *(.romvec) } > ram .text : { text_start = . ; *(.text) _etext = . ; __data_rom_start = ALIGN ( 4 ) ; } > ram .ramvec : { _ramvec = . ; } > ramvec .data : { __data_start = . ; *(.rodata) *(.data) _edata = . ; edata = ALIGN( 0x10 ) ; __data_end = ALIGN( 0x10 ) ; } > ram .bss : { __bss_start = ALIGN( 0x10 ) ; *(.bss) *(COMMON) end = ALIGN( 0x10 ) ; _end = ALIGN( 0x10 ) ; } > ram .eram : { _ramend = . ; } > eram } -----8<-------------------- This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From PaulMBanasik at eaton.com Tue Mar 5 11:28:21 2002 From: PaulMBanasik at eaton.com (Banasik, Paul M) Date: Tue, 5 Mar 2002 11:28:21 -0500 Subject: [uClinux-dev] sanity check -- anonymous cvs Message-ID: <2A8FE75B9F9D92449D13A4A6E391A7D2030F6E8F@pitpasmb04.ch.etn.com> Is anonymous CVS up and running? I'm trying to determine if it's a problem with my company's firewall, or something else. $ cvs -d:pserver:anonymous at cvs.uclinux.org:/var/cvs login (Logging in to anonymous at cvs.uclinux.org) CVS password: cvs [login aborted]: connect to cvs.uclinux.org:2401 failed: Connection refused Thanks, Paul B. This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From mirco_cremasco at hotmail.com Tue Mar 5 11:29:03 2002 From: mirco_cremasco at hotmail.com (mirco cremasco) Date: Tue, 05 Mar 2002 17:29:03 +0100 Subject: [uClinux-dev] problems building elf tool Message-ID: I've successfully rebuild-ed gcc-2.95.3, binutils-2.10, genromfs-0.5.1 and now I'm able to compile uClinux-dist-20020220 for a m5272c3 board. But there is still a problem: the generated bin image is too big!! images/romfs.bin is 3.9M !! (using m68k-els-tools-20020218 it is 1M) looking into romfs/bin directory I've found that telnet (as example) is 326K in place of 70K compiled with m68k-els-tools-20020218. I didn't changed any compiler option uClinux-dist makefiles. There are any option for compile gcc?? I used ./configure --target=m68k-elf --prefix=/opt/uClinux \ --with-headers=/opt/uClinux/uClinux-dist/lib/libc/include make make install Thaks, best regards Mirco Cremasco Imagos s.a.s. _________________________________________________________________ Prova MSN Messenger per conversare in linea con i tuoi amici: http://messenger.msn.it This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From chri at infis.univ.trieste.it Tue Mar 5 12:01:31 2002 From: chri at infis.univ.trieste.it (Christian Pellegrin) Date: Tue, 5 Mar 2002 18:01:31 +0100 (CET) Subject: [uClinux-dev] Porting Proprietary TCP/IP stack to uclinux In-Reply-To: <009001c1c442$03021bc0$d7204789@skeisamd01> Message-ID: On Tue, 5 Mar 2002, Suresh Singh K. wrote: > Hi All, > I am currently on a project which requires to port our own TCP/IP to uclinux. I want to know whether it's possible to PLUG IN A PROPRIETORY TCP/IP into the uclinux kernel without voilating the GPL conditions ?. I want to use the uclinux kernel (without it's native TCP/IP code) with my company's TCP/IP . And my company doesn't want it's "IP code & related modules like QOS,SNDCP " to come under the GPL conditions. > I also want to know if I can plug our proprietory "IP code " just below the linux native TCP/IP code without violating the GPL conditions of the linux kernel ? If our "IP Code" only access ( doesn't modify) the interface (both function & data ) provided by the native linux TCP/IP code , Does it voilates the GPL ?. > In linux, every ethernet receive interrupt call calls "netif_rx( )" passing the receive packet inside an sk_buff . Can the ethernet driver access ( doesn't modify) other linux TCP/IP interfaces (like the functions in dev.c ) other than the registeration , module interfaces & net_device structure without violating the GPL conditions ?. That's, I want my driver code OR any code below "IP of linux" to remain as PROPRIETORY. But still using the various functions & datas provided by the "linux TCP/IP" ( only reading & not modifying). > I would be very thankful to anyone who can Guide me on the above issues . > > Thanks in Advance, :-) > I guess that the most portable way is using PF_PACKET socket. It's also quite good in speed if you use the new mmap interface. There is a wonderful article on the last 2 number of LJ about PF_PACKET. I'm absolutely not an expert but I guess that if you hook to internal functions it will be a nigtmare to keep on with kernel changes. And this will look very ugly: a GPL redirector to a non-GPL module / user level program. So also Analog is going to release Linux support for something like a VoIP processor like TI and Aplio :-p. Unfortunatly they didn't understand the GNU manifesto that it's 100% interest of hardware manufactures to do Open Source :-(((((. Bye! This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From ferrin at bingdata.es Tue Mar 5 12:27:19 2002 From: ferrin at bingdata.es (Agustin Ferrin Pozuelo) Date: 05 Mar 2002 18:27:19 +0100 Subject: [uClinux-dev] Unable to build kernel 2.4 for m68k (Dragonball VZ - uCdimm) Message-ID: <1015349240.11145.11.camel@id13.bingdata.es> I am afraid I am missing some patch in order to build kernel 2.4 for the M68VZ328/uCdim. I am using "uClinux-dist-20020220" and the m68k-elf toolchain from Bernhard Kuhn. I get this when it goes to compile 'a.out' execution: -- binfmt_aout.c: In function `load_aout_binary': binfmt_aout.c:302: `SEGMENT_SIZE' undeclared (first use in this function) binfmt_aout.c:302: (Each undeclared identifier is reported only once binfmt_aout.c:302: for each function it appears in.) binfmt_aout.c:307: structure has no member named `mmap' -- What am I doing wrong? ...I just got fed up with the uClinux-2.0.38 that Lineo sent me along with the M68VZ328/uCdimm development kit. Second uart support is wrong and I have not the time to mend it. So I am trying to migrate to kernel version 2.4, but I don't even manage to build the kernel. -- Agustin Ferrin Pozuelo Embedded Systems Engineer at BINGdata Espa?a ferrin at bingdata.es This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From chri at infis.univ.trieste.it Tue Mar 5 12:35:20 2002 From: chri at infis.univ.trieste.it (Christian Pellegrin) Date: Tue, 5 Mar 2002 18:35:20 +0100 (CET) Subject: [uClinux-dev] Porting Proprietary TCP/IP stack to uclinux In-Reply-To: <009001c1c442$03021bc0$d7204789@skeisamd01> Message-ID: On Tue, 5 Mar 2002, Suresh Singh K. wrote: > Hi All, > I am currently on a project which requires to port our own TCP/IP to uclinux. I want to know whether it's possible to PLUG IN A PROPRIETORY TCP/IP into the uclinux kernel without voilating the GPL conditions ?. I want to use the uclinux kernel (without it's native TCP/IP code) with my company's TCP/IP . And my company doesn't want it's "IP code & related modules like QOS,SNDCP " to come under the GPL conditions. > I also want to know if I can plug our proprietory "IP code " just below the linux native TCP/IP code without violating the GPL conditions of the linux kernel ? If our "IP Code" only access ( doesn't modify) the interface (both function & data ) provided by the native linux TCP/IP code , Does it voilates the GPL ?. > In linux, every ethernet receive interrupt call calls "netif_rx( )" passing the receive packet inside an sk_buff . Can the ethernet driver access ( doesn't modify) other linux TCP/IP interfaces (like the functions in dev.c ) other than the registeration , module interfaces & net_device structure without violating the GPL conditions ?. That's, I want my driver code OR any code below "IP of linux" to remain as PROPRIETORY. But still using the various functions & datas provided by the "linux TCP/IP" ( only reading & not modifying). > I would be very thankful to anyone who can Guide me on the above issues . > Sorry for my previous mail. There is already an API for pluggin in another socket type that can live in a module, there is no need to add code to the kernel. Have a look in net/unic/af_unix.c. Bye! This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From bkuhn at lineo.com Tue Mar 5 12:44:41 2002 From: bkuhn at lineo.com (Bernhard Kuhn) Date: Tue, 05 Mar 2002 18:44:41 +0100 Subject: [uClinux-dev] Unable to build kernel 2.4 for m68k (Dragonball VZ - uCdimm) References: <1015349240.11145.11.camel@id13.bingdata.es> Message-ID: <3C850409.909FFA17@lineo.com> Agustin Ferrin Pozuelo schrieb: > > I get this when it goes to compile 'a.out' execution: Maybe you have activated the following option in your kernel setup: General setup ---> Kernel support for a.out binaries ? best regards Bernhard -- Bernhard Kuhn, Software Engineer, Lineo Inc. (Where Open Meets Smart) This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From olivierp at actimage.fr Tue Mar 5 12:33:50 2002 From: olivierp at actimage.fr (Olivier Pierrat) Date: Tue, 05 Mar 2002 18:33:50 +0100 Subject: [uClinux-dev] compilation process Message-ID: <3C85017E.9040705@actimage.fr> Dear, I have not understand all details about the difference bettween PIC and non-pic user application, and the -msep-data option. Can you give me some explanations, please ? Using the following command to generate a simple application, the link operation is performed by invoking /usr/local/lib/gcc-lib/m68k-elf/2.95.3/collect2 with some default option appended with the options specifies in the command line. ==> m68k-elf-gcc -m5200 -msep-data -Wl,-elf2flt -o test test.c -lc But when I invoke coolect2 it seems to be the same as m68k-elf-ld.real, according to the help display by collect2. Is it right ? I think, it is wrong, beaucause I can't use the -elf2flt option with m68k-elf-ld.real. What's the difference ? Why collect2 is invoked instead of m68k-elf-ld.real ? When I compile my test.c which perform a printf, with the command line given in the previous paragraph, al works fine. But when I try to compile with separate commande, printf doesn't work. What's happen ? Can you give me some clue ? commands line : 1. m68k-elf-gcc -c -m5200 -msep-data -o test.o test.c 2. m68k-elf-ld.real -o test.elf /usr/local/m68k-elf/lib/m5200/msep-data/crt0.o -L/usr/local/lib/gcc-lib/m68k-elf/2.95.3/m5200/msep-data -L/usr/local/lib/gcc-lib/m68k-elf/2.95.3 -L/usr/local/m68k-elf/lib/m5200/msep-data -L/usr/local/m68k-elf/lib -V -T elf2flt.ld test.o -lc -lgcc -lgcc 3. m68k-elf-elf2flt -v -o test test.elf Thanks for your help, I try to understand what happens. Olivier. This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From hamilton at SEDSystems.ca Tue Mar 5 13:01:48 2002 From: hamilton at SEDSystems.ca (Kendrick Hamilton) Date: Tue, 5 Mar 2002 12:01:48 -0600 (CST) Subject: [uClinux-dev] Unable to build kernel 2.4 for m68k (Dragonball VZ - uCdimm) In-Reply-To: <1015349240.11145.11.camel@id13.bingdata.es> Message-ID: Two things 1. Last I checked a.out does not work. You don't need it anyway. Select Kernel support for flat binaries. I have never tried Enable ZFLAT support. You don't need support for other types of binaries other then flat. 2. Have you checked the uClinux-2.0.x kernel on the uClinux cvs web page. There have been patches going to it that said they were fixing serial port problems with M68VZ328. On 5 Mar 2002, Agustin Ferrin Pozuelo wrote: > I am afraid I am missing some patch in order to build kernel 2.4 for the > M68VZ328/uCdim. > > I am using "uClinux-dist-20020220" and the m68k-elf toolchain from > Bernhard Kuhn. > > I get this when it goes to compile 'a.out' execution: > -- > binfmt_aout.c: In function `load_aout_binary': > binfmt_aout.c:302: `SEGMENT_SIZE' undeclared (first use in this > function) > binfmt_aout.c:302: (Each undeclared identifier is reported only once > binfmt_aout.c:302: for each function it appears in.) > binfmt_aout.c:307: structure has no member named `mmap' > -- > What am I doing wrong? > > ...I just got fed up with the uClinux-2.0.38 that Lineo sent me along > with the M68VZ328/uCdimm development kit. Second uart support is wrong > and I have not the time to mend it. > So I am trying to migrate to kernel version 2.4, but I don't even manage > to build the kernel. > > > -- Kendrick Hamilton E.I.T. SED Systems, a division of Calian Ltd. 18 Innovation Blvd. PO Box 1464 Saskatoon, Saskatchewan Canada S7N 3R1 Hamilton at sedsystems.ca Tel: (306) 933-1453 Fax: (306) 933-1486 This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From ssoucek at coactive.com Tue Mar 5 13:17:39 2002 From: ssoucek at coactive.com (Stefan Soucek) Date: Tue, 5 Mar 2002 10:17:39 -0800 Subject: [uClinux-dev] RE: [BusyBox] Re: BusyBox `msh' doesn't work correctly with vfork / uClibc Message-ID: Thanks for looking through. This has been sleeping for a while and I might put some work into it to synch up our msh version. > I looked at that patch, and found that most of it had already been > applied in the version of busybox/msh.c that I have (which is from > uClinux-dist-20020220, and claims to be version 0.60.0). Those parts > seems sane. > > However, Stefan's patch doesn't address the problem with `exit' that I > reported, so my patch (that changes things to use `_exit') or > something > similar is still necessary. I obviously missed those parts. I don't know why our shell doesn't crash on them but maybe they never get activated. > The parts of Stefan's patch which _aren't_ in the current sources are: > > (1) Some ifdefs to allow either fork or vfork to be used; > the current > sources just always use vfork, which seems fine to me. The ifdefs are rather there to switch back to fork() if the vfork() patch breaks msh. > (2) A change to the `grave' function to avoid problems with vfork; > since that function appears to do pretty complicated things > within the child, Stefan's patch just does (where cp is the > command): > > execlp("/bin/sh", "/bin/sh", "-c", cp, NULL); > > Not sure if that works, but the current code in `grave' > is certain > to screw things up when used with vfork. The grave function is a big mystery to me. I have done some tracing to learn which parts of it are vfork-safe but then I gave up. The hack using "execlp" works in most cases but is *slow*. The drawback is that any variables in the backticks aren't expanded; therefore this won't work, DIR="/home" RESULT=`ls -l $DIR` > So, I'd recommend: > > Apply my earlier patch to use _exit > > Think about doing something with the `grave' function > similar to what > Stefan did. -Stefan -- Stefan Soucek, ssoucek at coactive.com Coactive Networks, 28 Liberty Ship Way, Sausalito, CA 94965 Ph: (415) 289 7840, Fax: (415) 289 1320 This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From JLiu at PacesetterElectronics.com Tue Mar 5 15:26:29 2002 From: JLiu at PacesetterElectronics.com (JINGSONG LIU) Date: Tue, 5 Mar 2002 14:26:29 -0600 Subject: [uClinux-dev] RE: [BusyBox] Re: BusyBox `msh' doesn't work co rrectly with vfork / uClibc Message-ID: Thank you very much ! I'm new to Linux and uclinux would you please give me some more information about how to set up uclinux under the readhat 7.2 ? Thank you in advance JSL -----Original Message----- From: Stefan Soucek [mailto:ssoucek at coactive.com] Sent: Tuesday, March 05, 2002 12:18 PM To: 'Miles Bader'; andersen at codepoet.org Cc: uclibc at uclibc.org; uclinux-dev at uclinux.org; busybox at busybox.net Subject: [uClinux-dev] RE: [BusyBox] Re: BusyBox `msh' doesn't work correctly with vfork / uClibc Thanks for looking through. This has been sleeping for a while and I might put some work into it to synch up our msh version. > I looked at that patch, and found that most of it had already been > applied in the version of busybox/msh.c that I have (which is from > uClinux-dist-20020220, and claims to be version 0.60.0). Those parts > seems sane. > > However, Stefan's patch doesn't address the problem with `exit' that I > reported, so my patch (that changes things to use `_exit') or > something > similar is still necessary. I obviously missed those parts. I don't know why our shell doesn't crash on them but maybe they never get activated. > The parts of Stefan's patch which _aren't_ in the current sources are: > > (1) Some ifdefs to allow either fork or vfork to be used; > the current > sources just always use vfork, which seems fine to me. The ifdefs are rather there to switch back to fork() if the vfork() patch breaks msh. > (2) A change to the `grave' function to avoid problems with vfork; > since that function appears to do pretty complicated things > within the child, Stefan's patch just does (where cp is the > command): > > execlp("/bin/sh", "/bin/sh", "-c", cp, NULL); > > Not sure if that works, but the current code in `grave' > is certain > to screw things up when used with vfork. The grave function is a big mystery to me. I have done some tracing to learn which parts of it are vfork-safe but then I gave up. The hack using "execlp" works in most cases but is *slow*. The drawback is that any variables in the backticks aren't expanded; therefore this won't work, DIR="/home" RESULT=`ls -l $DIR` > So, I'd recommend: > > Apply my earlier patch to use _exit > > Think about doing something with the `grave' function > similar to what > Stefan did. -Stefan -- Stefan Soucek, ssoucek at coactive.com Coactive Networks, 28 Liberty Ship Way, Sausalito, CA 94965 Ph: (415) 289 7840, Fax: (415) 289 1320 This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From JLiu at PacesetterElectronics.com Tue Mar 5 16:25:13 2002 From: JLiu at PacesetterElectronics.com (JINGSONG LIU) Date: Tue, 5 Mar 2002 15:25:13 -0600 Subject: [uClinux-dev] RE: [BusyBox] Re: BusyBox `msh' doesn't work co rrectly with vfork / uClibc Message-ID: Dear sir I'm a new to uclinux and linux. My Company want to use uclinux as a OS of our product. Target CPU is M5272 and I had M5272C3 in hand for starting development. I download sources and binaries from http://www.uclinux.org/ports/coldfire/ The source from http://www.uclinux.org/ports/coldfire/source.html Full Source Distribution : uClinux-dist-20020214.tar and a too chain : m68k-elf-tools-20011219.tar. My computer had two OS one is winnt and another is linux ( redhat 7.2 ) One logic disk Hold linux with 2G space. Another logic disk with 2G space hold uClinux-dist-20020214.tar. I use < mount -t vfat /dev/hda1 /mnt/vfat > and then tar xvzf uClinux-dist-20020214.tar.gz This is files and directory in the first level. Directory of C:\UCLINUX\uClinux-dist 03/04/02 03:09p . 03/04/02 03:09p .. 03/04/02 03:09p bin 03/04/02 04:32p config 02/13/02 08:38p 18,007 COPYING 03/05/02 02:12p Documentation 03/04/02 03:09p freeswan 03/04/02 03:10p images 03/04/02 03:10p lib 03/04/02 04:28p linux 03/05/02 02:07p linux-2.4.x 02/13/02 08:38p 3,228 MAINTAINERS 02/13/02 08:38p 7,004 Makefile 02/13/02 08:38p 5,334 README 02/13/02 08:38p 1,397 SOURCE 03/04/02 03:21p tools 03/04/02 03:21p uClibc 03/04/02 03:22p user 03/04/02 03:29p vendors 03/04/02 04:31p 2,354 config.in 03/04/02 04:32p 1,990 .config 03/05/02 02:52p 0 1 03/04/02 04:32p 88 .oldconfig 03/04/02 04:31p 1,992 .config.old 03/04/02 04:31p 1,898 .config.save 03/04/02 03:59p 1,767 .tmpconfig.h 03/05/02 12:36p bak 27 File(s) 45,059 bytes 407,863,296 bytes free With linux login as root cd / tar xvzf m68k-elf-tools-20011219.tar.gz After that cd uClinux-dist, make xconfig, make dep ane make I got many error message !!! [root at localhost uclinux-dist]# ls -l total 640 -rwxr-xr-x 1 root root 1884 Mar 5 15:03 1 drwxr-xr-x 2 root root 32768 Mar 5 12:36 bak drwxr-xr-x 2 root root 32768 Mar 4 15:09 bin drwxr-xr-x 3 root root 32768 Mar 4 16:32 config -rwxr-xr-x 1 root root 2354 Mar 4 16:31 config.in -rwxr-xr-x 1 root root 18007 Feb 13 20:38 copying drwxr-xr-x 2 root root 32768 Mar 5 14:12 Documentation drwxr-xr-x 9 root root 32768 Mar 4 15:09 freeswan drwxr-xr-x 2 root root 32768 Mar 4 15:10 images drwxr-xr-x 12 root root 32768 Mar 4 15:10 lib drwxr-xr-x 15 root root 32768 Mar 4 16:28 linux drwxr-xr-x 15 root root 32768 Mar 5 14:07 linux-2.4.x -rwxr-xr-x 1 root root 3228 Feb 13 20:38 MAINTAINERS -rwxr-xr-x 1 root root 7004 Feb 13 20:38 Makefile -rwxr-xr-x 1 root root 5334 Feb 13 20:38 readme -rwxr-xr-x 1 root root 1397 Feb 13 20:38 source drwxr-xr-x 3 root root 32768 Mar 4 15:21 tools drwxr-xr-x 15 root root 32768 Mar 4 15:21 uClibc drwxr-xr-x 2 root root 32768 Mar 4 15:22 user drwxr-xr-x 19 root root 32768 Mar 4 15:29 vendors [root at localhost uclinux-dist]# make make -C linux-2.4.x || exit 1 make[1]: Entering directory `/mnt/vfat/uClinux/uclinux-dist/linux-2.4.x' gcc -D__KERNEL__ -I/mnt/vfat/uClinux/uclinux-dist/linux-2.4.x/include -Wall -Wstrict-prototypes -Wno-trigraphs -O1 -g -fno-strict-aliasing -pipe -mpreferred-stack-boundary=2 -c -o init/main.o init/main.c In file included from /mnt/vfat/uClinux/uclinux-dist/linux-2.4.x/include/linux/highmem.h:5, from /mnt/vfat/uClinux/uclinux-dist/linux-2.4.x/include/linux/pagemap.h:16, from /mnt/vfat/uClinux/uclinux-dist/linux-2.4.x/include/linux/locks.h:8, from /mnt/vfat/uClinux/uclinux-dist/linux-2.4.x/include/linux/devfs_fs_kernel.h:6 , from init/main.c:16: /usr/include/asm/pgalloc.h: In function `get_pgd_slow': /usr/include/asm/pgalloc.h:62: `PAGE_OFFSET_RAW' undeclared (first use in this function) /usr/include/asm/pgalloc.h:62: (Each undeclared identifier is reported only once /usr/include/asm/pgalloc.h:62: for each function it appears in.) In file included from /usr/include/asm/dma.h:13, from /mnt/vfat/uClinux/uclinux-dist/linux-2.4.x/include/linux/bootmem.h:8, from init/main.c:28: /usr/include/asm/io.h: In function `virt_to_phys': /usr/include/asm/io.h:132: `PAGE_OFFSET_RAW' undeclared (first use in this function) /usr/include/asm/io.h:133: warning: control reaches end of non-void function /usr/include/asm/io.h: In function `phys_to_virt': /usr/include/asm/io.h:137: `PAGE_OFFSET_RAW' undeclared (first use in this function) /usr/include/asm/io.h:138: warning: control reaches end of non-void function /usr/include/asm/io.h: In function `isa_check_signature': /usr/include/asm/io.h:239: `PAGE_OFFSET_RAW' undeclared (first use in this function) init/main.c: In function `start_kernel': init/main.c:583: `CONFIG_X86_L1_CACHE_SHIFT' undeclared (first use in this function) init/main.c:583: `PAGE_OFFSET_RAW' undeclared (first use in this function) init/main.c:599: warning: implicit declaration of function `pgtable_cache_init' make[1]: *** [init/main.o] Error 1 make[1]: Leaving directory `/mnt/vfat/uClinux/uclinux-dist/ I don't how to make thing forward . Would you please tell me how deal with this thing. I'm new to uclinux. Thank you very much ! Contact info : Pacesetter Electronics Inc 3851 Washington Ave St.Louis MO 63108 Phone 314-531-6363E123 Jason Liu 3/5/2002 This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From SMerrifield at tecom.com.au Tue Mar 5 17:28:48 2002 From: SMerrifield at tecom.com.au (Steven Merrifield) Date: Tue, 5 Mar 2002 23:28:48 +0100 Subject: [uClinux-dev] CGI and uClinux in ARM7-EB40LS ? Message-ID: <7FF1185EA0D3D511BC0500B0D0D0C24A1B5845@melexc01.tecom.com.au> Hi Stefan, I am doing CGI using the boa web server (included in the user directory). I have a /htdocs directory that contains the html pages, and a /htdocs/cgi-bin directory that contains the CGI binaries. All CGI programs are written in C. For development work, I NFS mount the /htdocs directory on a PC, to test the CGIs before making a new image. Once I am happy the CGI works properly, I copy the binary into the romfs/ and run genromfs to make a new image. steve > -----Original Message----- > From: Stefan Jonsson [mailto:stefan.jonsson at faltcom.se] > Sent: Tuesday, 5 March 2002 11:38 > To: uClinux-dev mailing list > Subject: [uClinux-dev] CGI and uClinux in ARM7-EB40LS ? > > > Hello list, > > 1) I am in need of running CGI-scripts on a webserver on > EB40LS. That it > will not be I Perl but I for instance C, I have understood. > But what do I > need to make this possible? What web-server (if it is > relevant) and do I > need a /home/httpd/cgi-bin ... or something? And where it > might be possible > to find this (in that case) needed things. > > 2) I always make an image.bin and download that to target. Is > it possible to > just download an application and get that to work as > expected? I mean, how > do I compile this application? How do I safely download that > to my card to > get it to work? > > I am most greatful for any answers to my problem. > > Regards, > Stefan Jonsson, > University of Umea, Sweden. > > This message resent by the uclinux-dev at uclinux.org list > server http://www.uClinux.org/ > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From davidm at snapgear.com Tue Mar 5 18:12:36 2002 From: davidm at snapgear.com (David McCullough) Date: Wed, 6 Mar 2002 09:12:36 +1000 Subject: [uClinux-dev] problems building elf tool In-Reply-To: ; from mirco_cremasco@hotmail.com on Tue, Mar 05, 2002 at 05:29:03PM +0100 References: Message-ID: <20020306091236.A10549@beast.internal.moreton.com.au> Jivin mirco cremasco lays it down ... > I've successfully rebuild-ed gcc-2.95.3, binutils-2.10, genromfs-0.5.1 and > now I'm able to compile uClinux-dist-20020220 for a m5272c3 board. > But there is still a problem: the generated bin image is too big!! > images/romfs.bin is 3.9M !! (using m68k-els-tools-20020218 it is 1M) > looking into romfs/bin directory I've found that telnet (as example) is 326K > in place of 70K compiled with m68k-els-tools-20020218. > I didn't changed any compiler option uClinux-dist makefiles. > There are any option for compile gcc?? > I used > ./configure --target=m68k-elf --prefix=/opt/uClinux \ > --with-headers=/opt/uClinux/uClinux-dist/lib/libc/include > > make > make install Use the build-uclinux-tools.sh to build the tools, it does things to make the install work properly. I would guess that you have installed "ld" and not replaced it with ld-elf2flt as required, thus you have a romfs/bin full of ELF files and not flat files, Cheers, Davidm -- David McCullough: Ph: +61 7 3435 2815 http://www.SnapGear.com davidm at snapgear.com Fx: +61 7 3891 3630 825 Stanley St., W'gabba QLD 4102, Oz This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From davidm at snapgear.com Tue Mar 5 18:18:32 2002 From: davidm at snapgear.com (David McCullough) Date: Wed, 6 Mar 2002 09:18:32 +1000 Subject: [uClinux-dev] Unable to build kernel 2.4 for m68k (Dragonball VZ - uCdimm) In-Reply-To: <1015349240.11145.11.camel@id13.bingdata.es>; from ferrin@bingdata.es on Tue, Mar 05, 2002 at 06:27:19PM +0100 References: <1015349240.11145.11.camel@id13.bingdata.es> Message-ID: <20020306091832.B10549@beast.internal.moreton.com.au> Jivin Agustin Ferrin Pozuelo lays it down ... > I am afraid I am missing some patch in order to build kernel 2.4 for the > M68VZ328/uCdim. > > I am using "uClinux-dist-20020220" and the m68k-elf toolchain from > Bernhard Kuhn. > > I get this when it goes to compile 'a.out' execution: > -- > binfmt_aout.c: In function `load_aout_binary': > binfmt_aout.c:302: `SEGMENT_SIZE' undeclared (first use in this > function) > binfmt_aout.c:302: (Each undeclared identifier is reported only once > binfmt_aout.c:302: for each function it appears in.) > binfmt_aout.c:307: structure has no member named `mmap' > -- > What am I doing wrong? Somehow your kernel configuration is way wrong. You should not be including a.out support in the kernel. > ...I just got fed up with the uClinux-2.0.38 that Lineo sent me along > with the M68VZ328/uCdimm development kit. Second uart support is wrong > and I have not the time to mend it. > So I am trying to migrate to kernel version 2.4, but I don't even manage > to build the kernel. There should be a valid uCdimm on 2.4 target in the uClinux distro, it is in vendors/Lineo/uCdimm/*-2.4.x. Just run "make xconfig" and choose uCdimm/2.4 and select "default all settings" and you should be away. Cheers, Davidm -- David McCullough: Ph: +61 7 3435 2815 http://www.SnapGear.com davidm at snapgear.com Fx: +61 7 3891 3630 825 Stanley St., W'gabba QLD 4102, Oz This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From justin.wojdacki at analog.com Tue Mar 5 20:18:49 2002 From: justin.wojdacki at analog.com (Justin Wojdacki) Date: Tue, 05 Mar 2002 17:18:49 -0800 Subject: [uClinux-dev] Porting Proprietary TCP/IP stack to uclinux References: Message-ID: <3C856E79.E0F52C83@analog.com> Christian Pellegrin wrote: > > So also Analog is going to release Linux support for something like a VoIP > processor like TI and Aplio :-p. Unfortunatly they didn't understand the > GNU manifesto that it's 100% interest of hardware manufactures to do Open > Source :-(((((. > Not true. Contrary to what Suresh has said, ADI has no plan to remove the Linux IP stack and replace it with a proprietary one. Once we get the port finished, you will be able to run Linux on the hardware in question without requiring any binary-only modules. I should know, I've been doing all the kernel work myself. Please be patient, and we'll release the code as soon as it's stable (and we've caught up with the rest of the uclinux dev tree ;) ). -- ------------------------------------------------- Justin Wojdacki justin.wojdacki at analog.com (408) 350-5032 Communications Processors Group -- Analog Devices This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From superwen at 263.net Tue Mar 5 20:30:27 2002 From: superwen at 263.net (kz) Date: Wed, 6 Mar 2002 9:30:27 +0800 Subject: [uClinux-dev] Please help me out of this. Thank you Message-ID: <20020306012319.1D0261E1EB551@smtp.263.net> Dear All, I'm a newbie with uClinux,and I got some interest in it. I got a M68VZ328ADS board,so I want to port uClinux into it.I've read the maillist and I found that many people did such a work a couple of months ago. Here is what I've done following the instructions in those messages. I used linux-2.0.38.tar.gz with patch uClinus-2.0.38.1pre7.diff.gz I modified the rom.ld to the memory map of my ADS board,and did the change in the is_in_rom function (memory.c).Then I compiled the kernel as below, "make menuconfig" Choose MC68EZ328ADS. (There is no "VZ" one) "make dep" "make FLASH_LOAD_ADDR=0x10000 flash.b", Since the flash0.b I'm using is copy the RAM space start with 0x10000 to the flash. I use this flash_load_addr Is the above right? I got a bbugv+.exe for VZ328ADS, and I use it at my NT machine.Here is what I was doing. >bbugv+ #choose com1 and uart1 Bootstrap:>init Bootstrap:>in flash0.b //The flash copy program Bootstrap:>in flash.b //Load my kernel. It took me almost 4 hours to load it!God,I can't believe it! I use my COM port with baud rate 115200 to load it.Why so slowly? Bootstrap:>go 2000 //Start flash0.b to copy the RAM to flash After print out a bunch of "WWWWVVVVVV", it told me "PASS". Switch back to normal mode, then press reset.Then nothing happen, the heart beat LED doesn't blink any more. And I can't get any info in HyperTerminal. So, I think I was failed on this..... Does anybody did the same work to fit the uclinux-2.0.28 to Mot's VZ328ADS? I'm looking forward to your reply. Best Regards Kevin Zhu ------------------------ ? _..--"\ /`'^`) ? ? ? ( .~\ \W//^~-' ? ? \_/ ( @) ? ( =-oOo-) ? I'm thinking about it... ? _.`^^- "o0/ ? ??(oO\ ###/ \`)???????????????? This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From kj_lin at accton.com.tw Tue Mar 5 21:50:09 2002 From: kj_lin at accton.com.tw (kj_lin at accton.com.tw) Date: Wed, 6 Mar 2002 10:50:09 +0800 Subject: [uClinux-dev] Why arm-elf instead of arm-linux? Message-ID: Hi All, I am curious that why we always configure the crossdev for uclinux with " --target=arm-elf" instead of "--target=arm-linux". Is it due to the issue of elf2flt tool? KJ This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From andersen at codepoet.org Tue Mar 5 22:11:06 2002 From: andersen at codepoet.org (Erik Andersen) Date: Tue, 5 Mar 2002 20:11:06 -0700 Subject: [uClinux-dev] Why arm-elf instead of arm-linux? In-Reply-To: References: Message-ID: <20020306031106.GA17034@codepoet.org> On Wed Mar 06, 2002 at 10:50:09AM +0800, kj_lin at accton.com.tw wrote: > I am curious that why we always configure the crossdev for uclinux with " > > --target=arm-elf" instead of "--target=arm-linux". > > Is it due to the issue of elf2flt tool? Primarily since it makes the compiler build much easier. I have been working on an arm-linux toolchain which is nearly ready. Presumably Greg and I will be able to sync up once I have my stuff working, -Erik -- Erik B. Andersen http://codepoet-consulting.com/ --This message was written using 73% post-consumer electrons-- This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From sheeliang at lit.org.sg Tue Mar 5 22:42:58 2002 From: sheeliang at lit.org.sg (sheeliang at lit.org.sg) Date: Wed, 6 Mar 2002 11:42:58 +0800 Subject: [uClinux-dev] How does mm work on uClinux? Message-ID: <20020306114258.A1393@localhost.localdomain> This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From gerg at snapgear.com Tue Mar 5 23:15:46 2002 From: gerg at snapgear.com (Greg Ungerer) Date: Wed, 06 Mar 2002 14:15:46 +1000 Subject: [uClinux-dev] Why arm-elf instead of arm-linux? References: <20020306031106.GA17034@codepoet.org> Message-ID: <3C8597F2.3090701@snapgear.com> Erik Andersen wrote: > On Wed Mar 06, 2002 at 10:50:09AM +0800, kj_lin at accton.com.tw wrote: > >>I am curious that why we always configure the crossdev for uclinux with " >> >>--target=arm-elf" instead of "--target=arm-linux". >> >>Is it due to the issue of elf2flt tool? >> > > Primarily since it makes the compiler build much easier. > I have been working on an arm-linux toolchain which is > nearly ready. Presumably Greg and I will be able to > sync up once I have my stuff working. Yep. The compiler build is often easier with a *-elf target. Some uClinux targets are not general linux targets (for example i960). Regards Greg ------------------------------------------------------------------------ Greg Ungerer -- Chief Software Wizard EMAIL: gerg at snapgear.com SnapGear PHONE: +61 7 3435 2888 825 Stanley St, FAX: +61 7 3891 3630 Woolloongabba, QLD, 4102, Australia WEB: www.snapgear.com This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From shiyas at globaledgesoft.com Wed Mar 6 00:47:41 2002 From: shiyas at globaledgesoft.com (shiaz) Date: Wed, 6 Mar 2002 11:17:41 +0530 Subject: [uClinux-dev] uclinux-dev:assembly language syntax for ARM7TDMI In-Reply-To: <010701c1ae95$728a2220$7d01a8c0@tekdigitel.com> References: <010701c1ae95$728a2220$7d01a8c0@tekdigitel.com> Message-ID: <02030611174101.00949@shiaz> Hi stella and list I've been blocked with building the kernel downloaded from uclinux.org for ARM7TDMI.I was able to build the gcc toolchain. Are u able to build the uclinux kernel using Lineo development kit? Any help on this is well appreciated.. Thankz in advance... Shiaz On Wednesday 06 February 2002 04:05 am, you wrote: > Hi, All > > I am new to uclinux. And I would like to port uclinux to ARM7DTMI . > I got my distribution CD from Lieno. > It seems that the assembly code for ARM7 looks a little different from what > I used before. So, Can somebody give me some Info/Clue on where I can get > the assembly language syntax for ARM7TDMI ? > > Thanks a lot. > > Stella Zhang ---------------------------------------- Content-Type: text/html; charset="iso-8859-1"; name="Attachment: 1" Content-Transfer-Encoding: quoted-printable Content-Description: ---------------------------------------- This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From sgvagabond at yahoo.com.cn Wed Mar 6 01:47:46 2002 From: sgvagabond at yahoo.com.cn (=?gb2312?q?=E1=BF=C3=F7=20=C9=CF=B9=D9?=) Date: Wed, 6 Mar 2002 14:47:46 +0800 (CST) Subject: [uClinux-dev] How does mm work on uClinux? In-Reply-To: <20020306114258.A1393@localhost.localdomain> Message-ID: <20020306064746.702.qmail@web15005.mail.bjs.yahoo.com> what is mm? --- sheeliang at lit.org.sg ????> This message resent by the uclinux-dev at uclinux.org > list server http://www.uClinux.org/ _________________________________________________________ Do You Yahoo!? ???? Yahoo! ?? http://survey.yahoo.com/cgi-bin/Express/survey.cgi This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From sheeliang at lit.org.sg Wed Mar 6 01:54:25 2002 From: sheeliang at lit.org.sg (sheeliang at lit.org.sg) Date: Wed, 6 Mar 2002 14:54:25 +0800 Subject: [uClinux-dev] This is a test Message-ID: <20020306145425.A1484@localhost.localdomain> This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From kj_lin at accton.com.tw Wed Mar 6 01:47:41 2002 From: kj_lin at accton.com.tw (kj_lin at accton.com.tw) Date: Wed, 6 Mar 2002 14:47:41 +0800 Subject: [uClinux-dev] Arm toolchain on C++ Message-ID: Hi, Does the toolchain of arm-elf-20011219 really work well on C++? I can not have the "cin" and "cout" work on my platform. But except "cin" & "cout", the C++ code seem to work fine. Does any user of arm-elf-20011219 has the same problem? Or anybody can give me some hints? KJ This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From stefan.jonsson at faltcom.se Wed Mar 6 03:12:32 2002 From: stefan.jonsson at faltcom.se (Stefan Jonsson) Date: Wed, 6 Mar 2002 09:12:32 +0100 Subject: SV: [uClinux-dev] CGI and uClinux in ARM7-EB40LS ? In-Reply-To: <7FF1185EA0D3D511BC0500B0D0D0C24A1B5845@melexc01.tecom.com.au> Message-ID: Hello Steve, Thank you for your help. I am new to uClinux and not so familiar with everything. You say: "I am doing CGI using the boa web server (included in the user directory)". Is it in a user directury somewhere in "userland"? Or in some download on uclinux.org? I have seen some boa directory in the CVS's but do not know what is needed and how to use that :-( Do u know what httpd is included on Lineos official uClinux disc (that makes the official testpage)? Maybe it can handle CGI aswell? (Would save me some trouble maybe, but it would b good to learn about how to install boa also). / Stefan -----Ursprungligt meddelande----- Fr?n: owner-uclinux-dev at uclinux.org [mailto:owner-uclinux-dev at uclinux.org]F?r Steven Merrifield Skickat: den 5 mars 2002 23:29 Till: uclinux-dev at uclinux.org ?mne: RE: [uClinux-dev] CGI and uClinux in ARM7-EB40LS ? Hi Stefan, I am doing CGI using the boa web server (included in the user directory). I have a /htdocs directory that contains the html pages, and a /htdocs/cgi-bin directory that contains the CGI binaries. All CGI programs are written in C. For development work, I NFS mount the /htdocs directory on a PC, to test the CGIs before making a new image. Once I am happy the CGI works properly, I copy the binary into the romfs/ and run genromfs to make a new image. steve > -----Original Message----- > From: Stefan Jonsson [mailto:stefan.jonsson at faltcom.se] > Sent: Tuesday, 5 March 2002 11:38 > To: uClinux-dev mailing list > Subject: [uClinux-dev] CGI and uClinux in ARM7-EB40LS ? > > > Hello list, > > 1) I am in need of running CGI-scripts on a webserver on > EB40LS. That it > will not be I Perl but I for instance C, I have understood. > But what do I > need to make this possible? What web-server (if it is > relevant) and do I > need a /home/httpd/cgi-bin ... or something? And where it > might be possible > to find this (in that case) needed things. > > 2) I always make an image.bin and download that to target. Is > it possible to > just download an application and get that to work as > expected? I mean, how > do I compile this application? How do I safely download that > to my card to > get it to work? > > I am most greatful for any answers to my problem. > > Regards, > Stefan Jonsson, > University of Umea, Sweden. > > This message resent by the uclinux-dev at uclinux.org list > server http://www.uClinux.org/ > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From chri at infis.univ.trieste.it Wed Mar 6 03:29:39 2002 From: chri at infis.univ.trieste.it (Christian Pellegrin) Date: Wed, 6 Mar 2002 09:29:39 +0100 (CET) Subject: [uClinux-dev] Porting Proprietary TCP/IP stack to uclinux In-Reply-To: <3C856E79.E0F52C83@analog.com> Message-ID: On Tue, 5 Mar 2002, Justin Wojdacki wrote: > Christian Pellegrin wrote: > > Not true. Contrary to what Suresh has said, ADI has no plan to remove > the Linux IP stack and replace it with a proprietary one. Once we get > the port finished, you will be able to run Linux on the hardware in > question without requiring any binary-only modules. I should know, > I've been doing all the kernel work myself. Please be patient, and > we'll release the code as soon as it's stable (and we've caught up > with the rest of the uclinux dev tree ;) ). > OK, sorry, I apologize. I'm just angry with analog since they made me booting windows since they didn't release sim2181 and asm2181 for Linux :-)))). Jokes apart I'm happy Linux will rule the scene :-) Sorry again, Bye! This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From baiwei99 at yahoo.com.cn Wed Mar 6 04:56:25 2002 From: baiwei99 at yahoo.com.cn (=?gb2312?q?bai=20wei?=) Date: Wed, 6 Mar 2002 17:56:25 +0800 (CST) Subject: [uClinux-dev] A question about ls>/dev/ttys0! Message-ID: <20020306095625.82010.qmail@web13208.mail.yahoo.com> hello! When I was using 'ls>/dev/ttyS0' on my kernel,a error appeared: ls>/dev/ttyS0: Bad command or file name I using sash and the /dev/ttyS0 was working all right. And the other functions fo 'ls' are OK.Is there any problem on the Sash or my kernel(linux-2.4)? Think you! _________________________________________________________ Do You Yahoo!? ???? Yahoo! ?? http://survey.yahoo.com/cgi-bin/Express/survey.cgi This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From mac at young.url.com.tw Wed Mar 6 04:53:52 2002 From: mac at young.url.com.tw (Mac Wang) Date: Wed, 6 Mar 2002 17:53:52 +0800 Subject: [uClinux-dev] Samsung 4510 support... Message-ID: <005401c1c4f4$d2fc29d0$eccb718c@zythum> Hi... I have ported the uClinux 2.4.17 to this Samsung 4510 chip. It can enter sash shell successfully, and do something like get IP by dhcpcd then telnet to other host. But the serial/ethernet drivers are not fully tested. Anyone interested it can visit my 4510 homepage below. Any suggestion is appericated. http://mac.os.nctu.edu.tw -- Mac Wang National Chiao Tung University Department of Computer & Information Science Operating System Lab. E-mail: mac at os.nctu.edu.tw This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From mac at young.url.com.tw Wed Mar 6 05:13:03 2002 From: mac at young.url.com.tw (Mac Wang) Date: Wed, 6 Mar 2002 18:13:03 +0800 Subject: [uClinux-dev] Samsung 4510 support Message-ID: <007001c1c4f7$8160b750$eccb718c@zythum> Hi... I have ported the uClinux 2.4.17 to this Samsung 4510 chip. It can enter sash shell successfully, and do something like get IP by dhcpcd then telnet to other host. But the serial/ethernet drivers are not fully tested. Anyone interested it can visit my 4510 homepage below. This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From baiwei99 at yahoo.com.cn Wed Mar 6 05:17:43 2002 From: baiwei99 at yahoo.com.cn (=?gb2312?q?bai=20wei?=) Date: Wed, 6 Mar 2002 18:17:43 +0800 (CST) Subject: [uClinux-dev] About max3100 on VZ board with SPI1 Message-ID: <20020306101743.63967.qmail@web13202.mail.yahoo.com> hi: I just write a dirver for the UART by SPI1 of 68VZ328 ,and I was using max3100 . First I writed a program of asm. Just to config the max3100 and read the config back. I write a 0xcc0a to the max3100,but when I write a 0x4000 to read the configuration ,also a 0x4000 backed. Before I write data to the max3100 I have writing a 0 to the cs,and after the transfer writing a 1. Why I can not read the configuration? think you! _________________________________________________________ Do You Yahoo!? ???? Yahoo! ?? http://survey.yahoo.com/cgi-bin/Express/survey.cgi This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From olivierp at actimage.fr Wed Mar 6 05:24:31 2002 From: olivierp at actimage.fr (Olivier Pierrat) Date: Wed, 06 Mar 2002 11:24:31 +0100 Subject: [uClinux-dev] [uClinux] memory region, at an absolute address Message-ID: <3C85EE5F.7000509@actimage.fr> Hi, I try to define a memory region starting at 0x00300000, and with a lenght of 0x8000. The start address must be an absolute one. I want to read an write within this region. My script file is (myscript.ld), which is a copy of elf2flt.ld to which I add my memory region : myscript.ld ============= MEMORY { flatmem : ORIGIN = 0x0, LENGTH = 0x100000 common_data : ORIGIN = 0x00300000, LENGTH = 0x8000 } SECTIONS { .text 0 : { . = ALIGN(0x4) ; _stext = . ; _start = . ; *(.text) *(.text.*) *(.gnu.warning) *(.stub) *(.gnu.linkonce.t*) *(.init) . = ALIGN(0x10) ; _etext = . ; } > flatmem .data : { . = ALIGN(0x4) ; _sdata = . ; __data_start = . ; data_start = . ; *(.rela.got) *(.rel.got) *(.got.plt) *(.got) LONG(-1) *(.rodata) *(.rodata.*) *(.gnu.linkonce.r*) *(.rodata1) *(.data) *(.data.*) *(.gnu.linkonce.d*) *(.data1) *(.eh_frame) *(.gcc_except_table) *(.sdata) *(.sdata.*) *(.gnu.linkonce.s.*) . = ALIGN(4) ; *(.ctors.*) *(.ctors) LONG(0) *(.dtors.*) *(.dtors) LONG(0) . = ALIGN(0x10) ; _edata = . ; } > flatmem .bss : { . = ALIGN(0x4) ; _sbss = ALIGN(0x4) ; __bss_start = . ; *(.dynsbss) *(.sbss) *(.sbss.*) *(.scommon) *(.dynbss) *(.bss) *(.bss.*) *(COMMON) . = ALIGN(0x4) ; _ebss = . ; _end = . ; end = . ; } > flatmem .common.data : { *(.common.data) ; } > common_data } My user application to test this, is test.c, compile with the command : m68k-elf-gcc --m5200 -msep-data -Wl,-elf2flt -Wl,-T ./myscript.ld -o test test.c -lc test.c ======== #include #include #include #include #include #define __commondata __attribute__ ((__section__ (".common.data"))) int datatmp __commondata = 0 ; /*****************************************************************************/ int main(int argc, char *argv[]) { int tmp = 1; printf("datatmp adress = %08x \n", (int)&datatmp); // printf("_datatmp value = %08d \n", (int)datatmp); fflush(stdout); // Mandatory, otherwise printout is buffered exit(0); } The adress of datatmp is 0x04E6EF8 ??? I think it's the relative address + 0x00300000, whereas I want that 0x00300000 was an absolute address. What is wrong in my method ? Can you give an explanation or a clue ? Thanks. Olivier. This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From startec at ms11.hinet.net Wed Mar 6 05:50:27 2002 From: startec at ms11.hinet.net (Steve Tsai) Date: Wed, 6 Mar 2002 18:50:27 +0800 Subject: [uClinux-dev] Samsung 4510 support... References: <005401c1c4f4$d2fc29d0$eccb718c@zythum> Message-ID: <004d01c1c4fc$bab5bdc0$5501a8c0@synso.com.tw> Hi Mac, I had a 4510 board using another ARM linux distrubution without MMU. Maybe I can test your codes on my board, and I have some question. Which toolchain do you use? Do you use littleendian mode or bigendian mode? What's your bootstrap code? Steve Tsai ----- Original Message ----- From: "Mac Wang" To: "uclinux-dev" Sent: Wednesday, March 06, 2002 5:53 PM Subject: [uClinux-dev] Samsung 4510 support... > Hi... > I have ported the uClinux 2.4.17 to this Samsung 4510 chip. It can enter > sash shell successfully, and do something like get IP by dhcpcd then telnet > to other host. But the serial/ethernet drivers are not fully tested. Anyone > interested it can visit my 4510 homepage below. Any suggestion is > appericated. > > http://mac.os.nctu.edu.tw > -- > Mac Wang > National Chiao Tung University > Department of Computer & Information Science > Operating System Lab. > E-mail: mac at os.nctu.edu.tw > > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From mac at young.url.com.tw Wed Mar 6 06:15:08 2002 From: mac at young.url.com.tw (Mac Wang) Date: Wed, 6 Mar 2002 19:15:08 +0800 Subject: [uClinux-dev] A question about ls>/dev/ttys0! References: <20020306095625.82010.qmail@web13208.mail.yahoo.com> Message-ID: <001801c1c500$2d506580$eccb718c@zythum> sash doesn't support pipe... so you can't use '>' ----- Original Message ----- From: "bai wei" To: Sent: Wednesday, March 06, 2002 5:56 PM Subject: [uClinux-dev] A question about ls>/dev/ttys0! > > hello! > When I was using 'ls>/dev/ttyS0' on my kernel,a > error appeared: > > ls>/dev/ttyS0: Bad command or file name > > I using sash and the /dev/ttyS0 was working all > right. And the other functions fo 'ls' are OK.Is there > any problem on the Sash or my kernel(linux-2.4)? > > Think you! > > > _________________________________________________________ > Do You Yahoo!? > ???? Yahoo! ?? > http://survey.yahoo.com/cgi-bin/Express/survey.cgi > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From mac at young.url.com.tw Wed Mar 6 07:08:58 2002 From: mac at young.url.com.tw (Mac Wang) Date: Wed, 6 Mar 2002 20:08:58 +0800 Subject: [uClinux-dev] Samsung 4510 support... References: <005401c1c4f4$d2fc29d0$eccb718c@zythum> <004d01c1c4fc$bab5bdc0$5501a8c0@synso.com.tw> Message-ID: <001201c1c507$b3699a90$eccb718c@zythum> Hi Steve, 1. The toolchain I put on my page already, it's from Lineo's uClinux distribution CD... 2. I use little endian on my board... 3. My bootstrap code is ~linux-2.4.x/arch/armnommu/boot/compressed/head.S... take a look on it... ----- Original Message ----- From: "Steve Tsai" To: Sent: Wednesday, March 06, 2002 6:50 PM Subject: Re: [uClinux-dev] Samsung 4510 support... > Hi Mac, > > I had a 4510 board using another ARM linux distrubution without MMU. Maybe > I can test your codes on my board, and I have some question. Which toolchain > do you use? Do you use littleendian mode or bigendian mode? What's your > bootstrap code? > > Steve Tsai > > > ----- Original Message ----- > From: "Mac Wang" > To: "uclinux-dev" > Sent: Wednesday, March 06, 2002 5:53 PM > Subject: [uClinux-dev] Samsung 4510 support... > > > > Hi... > > I have ported the uClinux 2.4.17 to this Samsung 4510 chip. It can > enter > > sash shell successfully, and do something like get IP by dhcpcd then > telnet > > to other host. But the serial/ethernet drivers are not fully tested. > Anyone > > interested it can visit my 4510 homepage below. Any suggestion is > > appericated. > > > > http://mac.os.nctu.edu.tw > > -- > > Mac Wang > > National Chiao Tung University > > Department of Computer & Information Science > > Operating System Lab. > > E-mail: mac at os.nctu.edu.tw > > > > This message resent by the uclinux-dev at uclinux.org list server > http://www.uClinux.org/ > > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From gmenie at akamai.com Wed Mar 6 08:59:29 2002 From: gmenie at akamai.com (Menie, Georges) Date: Wed, 6 Mar 2002 08:59:29 -0500 Subject: [uClinux-dev] Please help me out of this. Thank you Message-ID: >from my own experience, I would not use the rom build first, but the himem build. And may be it would be better to use the latest uClinux distribution from http://www.uclinux.org/pub/uClinux/dist/index.html and build the 2.4 kernel. Choose one of the setting for the Lineo/ucdimm or Exys/de2 boards which are 68VZ328 based, modify the himem.ld for your board, and compile like you did. Then download the image.bin file onto your board RAM and run it (remember: the start address is the download address+0x400) You may have to modify other files to have a working image, then you will be able to build a rom image. In the make menuconfig step, set everything by default first (yes for "Default all settings") later on, when your image will be ok, you might customize it. Regards, Georges > -----Original Message----- > From: kz [mailto:superwen at 263.net] > Sent: mercredi 6 mars 2002 02:30 > To: uclinux > Subject: [uClinux-dev] Please help me out of this. Thank you > > > Dear All, > > I'm a newbie with uClinux,and I got some interest in it. > > I got a M68VZ328ADS board,so I want to port uClinux into > it.I've read the maillist and I found that many people did > such a work a couple of months ago. Here is what I've done > following the instructions in those messages. > > I used linux-2.0.38.tar.gz with patch uClinus-2.0.38.1pre7.diff.gz > > I modified the rom.ld to the memory map of my ADS board,and > did the change in the is_in_rom function (memory.c).Then I > compiled the kernel as below, > "make menuconfig" Choose MC68EZ328ADS. (There is no "VZ" one) > "make dep" > "make FLASH_LOAD_ADDR=0x10000 flash.b", Since the flash0.b > I'm using is copy the RAM space start with 0x10000 to the > flash. I use this flash_load_addr > > Is the above right? > > I got a bbugv+.exe for VZ328ADS, and I use it at my NT > machine.Here is what I was doing. > > >bbugv+ > #choose com1 and uart1 > Bootstrap:>init > Bootstrap:>in flash0.b //The flash copy program > Bootstrap:>in flash.b //Load my kernel. > > It took me almost 4 hours to load it!God,I can't believe it! > I use my COM port with baud rate 115200 to load it.Why so slowly? > > Bootstrap:>go 2000 //Start flash0.b to copy the RAM to flash > > After print out a bunch of "WWWWVVVVVV", it told me "PASS". > > Switch back to normal mode, then press reset.Then nothing > happen, the heart beat LED doesn't blink any more. And I > can't get any info in HyperTerminal. So, I think I was failed > on this..... > > Does anybody did the same work to fit the uclinux-2.0.28 to > Mot's VZ328ADS? > I'm looking forward to your reply. > Best Regards > Kevin Zhu > ------------------------ > ? _..--"\ /`'^`) ? ? ? > ( .~\ \W//^~-' ? ? > \_/ ( @) ? > ( =-oOo-) ? I'm thinking about it... ? > _.`^^- "o0/ ? > ??(oO\ ###/ \`)???????????????? > > This message resent by the uclinux-dev at uclinux.org list > server http://www.uClinux.org/ > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From pgccoop at hotmail.com Wed Mar 6 13:25:30 2002 From: pgccoop at hotmail.com (Jonathan Schmidt) Date: Wed, 06 Mar 2002 18:25:30 +0000 Subject: [uClinux-dev] About max3100 on VZ board with SPI1 Message-ID: I too have a MAX3100 on the SPI2 interface on my VZ board. I have not started developing drivers for it yet, but I do need to start within the next week or so. Would it be possible to see what you have written for a driver so far? That would be really helpful to me! Thanks. Jonathan >From: bai wei >Reply-To: uclinux-dev at uclinux.org >To: uclinux-dev at uclinux.org >Subject: [uClinux-dev] About max3100 on VZ board with SPI1 >Date: Wed, 6 Mar 2002 18:17:43 +0800 (CST) >MIME-Version: 1.0 >Received: from [216.57.57.93] by hotmail.com (3.2) with ESMTP id >MHotMailBE4F3C8C00A84004325ED839395DCA5E0; Wed, 06 Mar 2002 02:38:22 -0800 >Received: (from mdom at localhost)by uclinux-dev-listclinux.org >(8.11.3/8.11.3/SuSE Linux 8.11.1-0.5) id g26AHpj21830for uclinux-dev-list; >Wed, 6 Mar 2002 05:17:51 -0500 >From owner-uclinux-dev at uclinux.org Wed, 06 Mar 2002 02:39:02 -0800 >Message-ID: <20020306101743.63967.qmail at web13202.mail.yahoo.com> >Sender: owner-uclinux-dev at uclinux.org >Precedence: bulk > >hi: > I just write a dirver for the UART by SPI1 of >68VZ328 ,and I was using max3100 . First I writed a >program of asm. Just to config the max3100 and read >the config back. I write a 0xcc0a to the max3100,but >when I write a 0x4000 to read the configuration ,also >a 0x4000 backed. Before I write data to the max3100 I >have writing a 0 to the cs,and after the transfer >writing a 1. > Why I can not read the configuration? > >think you! > >_________________________________________________________ >Do You Yahoo!? >???????? Yahoo! ???? >http://survey.yahoo.com/cgi-bin/Express/survey.cgi >This message resent by the uclinux-dev at uclinux.org list server >http://www.uClinux.org/ _________________________________________________________________ Chat with friends online, try MSN Messenger: http://messenger.msn.com This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From skibria at win.trlabs.ca Wed Mar 6 14:09:27 2002 From: skibria at win.trlabs.ca (Sami Kibria) Date: 06 Mar 2002 13:09:27 -0600 Subject: [uClinux-dev] re: flash runtime Message-ID: <1015441767.6614.0.camel@feedtheworms.win.trlabs.ca> hey guys...i just have a really quick question... can you allocate a certain area in flash for writing to, lets say we want to keep some data logging...can we do that, or like in the, tini (ds80 dual can high speed micro), that you can't write to flash while executing an application, can you write to flash with that atmel at91, or is not allowed? thanks... ps...just in case i am not inquiring about code shadowing... :) -- -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- Sami Kibria Researcher email: skibria at win.trlabs.ca TRLabs - Winnipeg, MB, CANADA http://www.win.trlabs.ca/~skibria -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- __ _ / / (_)__ __ ____ __ / /__/ / _ \/ // /\ \/ / . . . t h e c h o i c e o f a /____/_/_//_/\_,_/ /_/\_\ G N U g e n e r a t i o n -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From szhang at tekdigitel.com Wed Mar 6 14:15:12 2002 From: szhang at tekdigitel.com (Stella) Date: Wed, 6 Mar 2002 14:15:12 -0500 Subject: [uClinux-dev] uclinux-dev:assembly language syntax for ARM7TDMI References: <010701c1ae95$728a2220$7d01a8c0@tekdigitel.com> <02030611174101.00949@shiaz> Message-ID: <002f01c1c543$3d8e4820$7d01a8c0@tekdigitel.com> I successfully built the kernel under linux platform but have not tested yet. But I failed to build the kernel under cygwin. There are some errors. I am still working on this. Stella ----- Original Message ----- From: "shiaz" To: Sent: Wednesday, March 06, 2002 12:47 AM Subject: Re: [uClinux-dev] uclinux-dev:assembly language syntax for ARM7TDMI > Hi stella and list > I've been blocked with building the kernel downloaded from uclinux.org for > ARM7TDMI.I was able to build the gcc toolchain. > Are u able to build the uclinux kernel using Lineo development kit? > Any help on this is well appreciated.. > Thankz in advance... > > Shiaz > > > > > On Wednesday 06 February 2002 04:05 am, you wrote: > > Hi, All > > > > I am new to uclinux. And I would like to port uclinux to ARM7DTMI . > > I got my distribution CD from Lieno. > > It seems that the assembly code for ARM7 looks a little different from what > > I used before. So, Can somebody give me some Info/Clue on where I can get > > the assembly language syntax for ARM7TDMI ? > > > > Thanks a lot. > > > > Stella Zhang > > ---------------------------------------- > Content-Type: text/html; charset="iso-8859-1"; name="Attachment: 1" > Content-Transfer-Encoding: quoted-printable > Content-Description: > ---------------------------------------- > > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ > > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From hamilton at SEDSystems.ca Wed Mar 6 14:43:43 2002 From: hamilton at SEDSystems.ca (Kendrick Hamilton) Date: Wed, 6 Mar 2002 13:43:43 -0600 (CST) Subject: [uClinux-dev] re: flash runtime In-Reply-To: <1015441767.6614.0.camel@feedtheworms.win.trlabs.ca> Message-ID: Sami, I depends on if your code is running from RAM or from flash. If your code is running from RAM (kernel and applications) then it should not be a problem. Just make sure that nothing (including interrupt vectors) are being pulled from flash. What processor are you using? On 6 Mar 2002, Sami Kibria wrote: > hey guys...i just have a really quick question... > > can you allocate a certain area in flash for writing to, lets say we > want to keep some data logging...can we do that, or like in the, tini > (ds80 dual can high speed micro), that you can't write to flash while > executing an application, can you write to flash with that atmel at91, > or is not allowed? > > thanks... > > ps...just in case i am not inquiring about code shadowing... > > :) > > > -- Kendrick Hamilton E.I.T. SED Systems, a division of Calian Ltd. 18 Innovation Blvd. PO Box 1464 Saskatoon, Saskatchewan Canada S7N 3R1 Hamilton at sedsystems.ca Tel: (306) 933-1453 Fax: (306) 933-1486 This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From skibria at win.trlabs.ca Wed Mar 6 15:21:47 2002 From: skibria at win.trlabs.ca (Sami Kibria) Date: 06 Mar 2002 14:21:47 -0600 Subject: [uClinux-dev] re: runtime flash at91 Message-ID: <1015446111.6614.2.camel@feedtheworms.win.trlabs.ca> hey guys...i just have a really quick question... can you allocate a certain area in flash for writing to, lets say we want to keep some data logging...can we do that, or like in the, tini (ds80 dual can high speed micro), that you can't write to flash while executing an application, can you write to flash with that atmel at91, or is not allowed? thanks... ps...just in case i am not inquiring about code shadowing... :) -- -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- Sami Kibria Researcher email: skibria at win.trlabs.ca TRLabs - Winnipeg, MB, CANADA http://www.win.trlabs.ca/~skibria -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- __ _ / / (_)__ __ ____ __ / /__/ / _ \/ // /\ \/ / . . . t h e c h o i c e o f a /____/_/_//_/\_,_/ /_/\_\ G N U g e n e r a t i o n -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From jjef at inforamp.net Wed Mar 6 14:52:17 2002 From: jjef at inforamp.net (John Jeffers) Date: Wed, 06 Mar 2002 14:52:17 -0500 Subject: [uClinux-dev] Hi anyone have a good theory on why flash will work from a monitor and not from an application? Message-ID: <5.1.0.14.2.20020306143958.032e9d10@209.89.75.25> Hi I have used Kiwi for my flash monitor. It loads itself beautifully from itself and uClinux too. I go through trap 2 and am using the same code. I have turned all interrupts off except serial port and relocated the stack back into monitor space 10000H-10H The first flash sector erase stops it and on the EZ328 you cannot see where it got to. Actually it got to the sector erase line in the AMD code that worked previously in monitor only. So I have the stack moved so no one can step on me, I have all interrupts off. I am executing from 0 to 4000H and stacked FFE0h down. from my point of view the conditions are exactly the same as the monitor in which everything worked well and it's the same code running. One additional point is if I put puts() all over that code it works really really slow but it works. Any idea's? The flash.c is just about identical to the published Kiwi on sourceforge and the application taking it to trap2 is flashloader.c from uClinux june222001. I'll send it all to anyone who wants to have a go. Cheers John This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From davidm at snapgear.com Wed Mar 6 18:43:43 2002 From: davidm at snapgear.com (David McCullough) Date: Thu, 7 Mar 2002 09:43:43 +1000 Subject: [uClinux-dev] [uClinux] memory region, at an absolute address In-Reply-To: <3C85EE5F.7000509@actimage.fr>; from olivierp@actimage.fr on Wed, Mar 06, 2002 at 11:24:31AM +0100 References: <3C85EE5F.7000509@actimage.fr> Message-ID: <20020307094343.A8005@beast.internal.moreton.com.au> Jivin Olivier Pierrat lays it down ... > The adress of datatmp is 0x04E6EF8 ??? I think it's the relative address > + 0x00300000, whereas I want that 0x00300000 was an absolute address. > What is wrong in my method ? Can you give an explanation or a clue ? Flat format relocates everything relative to where the program is loaded, thus the problem, your fixed locations have been relocated. I think it would be fairly hard to sto pthis unless you can convince the linker/compiler not to generate relocation entries for those accesses. Checkout the linker script DOC, you might find a way. Depending on what you are using that region for, you could just set a pointer to 0x00300000, ie., struct fancy_stuff *sp = (struct fancy_stuff *) 0x00300000; sp->field = value; ... Cheers, Davidm -- David McCullough: Ph: +61 7 3435 2815 http://www.SnapGear.com davidm at snapgear.com Fx: +61 7 3891 3630 825 Stanley St., W'gabba QLD 4102, Oz This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From baiwei99 at yahoo.com.cn Wed Mar 6 20:32:34 2002 From: baiwei99 at yahoo.com.cn (=?gb2312?q?bai=20wei?=) Date: Thu, 7 Mar 2002 09:32:34 +0800 (CST) Subject: [uClinux-dev] About max3100 on VZ board with SPI1 In-Reply-To: Message-ID: <20020307013234.61086.qmail@web13207.mail.yahoo.com> My program is very easy and I want to write my driver referrence the max311Xserial.c. You can get this program at here: http://www.bdt.com/david/ucsimm/max311Xserial-0.20-R1.tar.gz. Send me some message too after your success! --- Jonathan Schmidt ????> I too have a MAX3100 on the SPI2 interface on my VZ > board. I have not > started developing drivers for it yet, but I do need > to start within the > next week or so. Would it be possible to see what > you have written for a > driver so far? That would be really helpful to me! > Thanks. > > Jonathan > > > >From: bai wei > >Reply-To: uclinux-dev at uclinux.org > >To: uclinux-dev at uclinux.org > >Subject: [uClinux-dev] About max3100 on VZ board > with SPI1 > >Date: Wed, 6 Mar 2002 18:17:43 +0800 (CST) > >MIME-Version: 1.0 > >Received: from [216.57.57.93] by hotmail.com (3.2) > with ESMTP id > >MHotMailBE4F3C8C00A84004325ED839395DCA5E0; Wed, 06 > Mar 2002 02:38:22 -0800 > >Received: (from mdom at localhost)by > uclinux-dev-listclinux.org > >(8.11.3/8.11.3/SuSE Linux 8.11.1-0.5) id > g26AHpj21830for uclinux-dev-list; > >Wed, 6 Mar 2002 05:17:51 -0500 > >From owner-uclinux-dev at uclinux.org Wed, 06 Mar 2002 > 02:39:02 -0800 > >Message-ID: > <20020306101743.63967.qmail at web13202.mail.yahoo.com> > >Sender: owner-uclinux-dev at uclinux.org > >Precedence: bulk > > > >hi: > > I just write a dirver for the UART by SPI1 of > >68VZ328 ,and I was using max3100 . First I writed a > >program of asm. Just to config the max3100 and read > >the config back. I write a 0xcc0a to the > max3100,but > >when I write a 0x4000 to read the configuration > ,also > >a 0x4000 backed. Before I write data to the max3100 > I > >have writing a 0 to the cs,and after the transfer > >writing a 1. > > Why I can not read the configuration? > > > >think you! > > > >_________________________________________________________ > >Do You Yahoo!? > >???? Yahoo! ?? > >http://survey.yahoo.com/cgi-bin/Express/survey.cgi > >This message resent by the uclinux-dev at uclinux.org > list server > >http://www.uClinux.org/ > > > > > _________________________________________________________________ > Chat with friends online, try MSN Messenger: > http://messenger.msn.com > > This message resent by the uclinux-dev at uclinux.org > list server http://www.uClinux.org/ _________________________________________________________ Do You Yahoo!? ???? Yahoo! ?? http://survey.yahoo.com/cgi-bin/CNITsolution/survey.cgi This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From gerg at snapgear.com Wed Mar 6 21:49:45 2002 From: gerg at snapgear.com (Greg Ungerer) Date: Thu, 07 Mar 2002 12:49:45 +1000 Subject: [uClinux-dev] Samsung 4510 support... References: <005401c1c4f4$d2fc29d0$eccb718c@zythum> Message-ID: <3C86D549.4000606@snapgear.com> Hi Mac, I have integrated these patches. Some did not apply clean, you should check the CVS and see if everything is there. Just a suggestion :-) If you do a "make distclean" in linux-2.4.x then your patches won't contain any of the generated files. Makes it easier for us to integrate patches. Regards Greg Mac Wang wrote: > Hi... > I have ported the uClinux 2.4.17 to this Samsung 4510 chip. It can enter > sash shell successfully, and do something like get IP by dhcpcd then telnet > to other host. But the serial/ethernet drivers are not fully tested. Anyone > interested it can visit my 4510 homepage below. Any suggestion is > appericated. > > http://mac.os.nctu.edu.tw > -- > Mac Wang > National Chiao Tung University > Department of Computer & Information Science > Operating System Lab. > E-mail: mac at os.nctu.edu.tw > > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ > > -- ------------------------------------------------------------------------ Greg Ungerer -- Chief Software Wizard EMAIL: gerg at snapgear.com SnapGear PHONE: +61 7 3435 2888 825 Stanley St, FAX: +61 7 3891 3630 Woolloongabba, QLD, 4102, Australia WEB: www.snapgear.com This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From baiwei99 at yahoo.com.cn Wed Mar 6 21:50:21 2002 From: baiwei99 at yahoo.com.cn (=?gb2312?q?bai=20wei?=) Date: Thu, 7 Mar 2002 10:50:21 +0800 (CST) Subject: [uClinux-dev] About framebuffer for LCD on 2.4 kernel Message-ID: <20020307025021.711.qmail@web13208.mail.yahoo.com> hi: I want to write something to my LCD by framebuffer.I has modified /arch/m68knmmu/console/68328fb.c for my 320x240 LCD.And I also wrote a application to test it.My test program is like this: static int screen_fd=0; /* Framebuffer file handle. */ int screen_width,screen_height; unsigned char *screen_ptr; struct fb_var_screeninfo screeninfo; int ret = 0; unsigned char screen_data[100]={1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1}; /** test if screen have been initilized . **/ if ( screen_fd >= 0 ) return 0; screen_fd = open(SCR_DEV, O_RDWR); if (screen_fd == -1) { perror("Unable to open frame buffer device /dev/fb0"); return -1; } /*if ( (gb2312fonts_fd = open(GB2312FONTS_FILE,O_RDONLY)) == -1 ) { ret = -1; goto init_error; }*/ if (ioctl(screen_fd, FBIOGET_VSCREENINFO, &screeninfo)==-1) { perror("Unable to retrieve framebuffer information"); ret = -1; goto init_error; } screen_width = screeninfo.xres_virtual; screen_height = screeninfo.yres_virtual; screen_ptr = mmap(0, screen_height * screen_width / 8, PROT_READ|PROT_WRITE, 0, screen_fd, 0); if (screen_ptr==-1) { perror("Unable to mmap frame buffer"); ret = -1; goto init_error; } memset(screen_ptr,0,screen_height * screen_width / 8); memcpy(screen_ptr,screen_data,100); return 0; But after running this program,nothing was happen on my LCD. I need some help and I want to know how dose the 68328fb.c connect to /dev/fb0 (29,0)? think you! _________________________________________________________ Do You Yahoo!? ???? Yahoo! ?? http://survey.yahoo.com/cgi-bin/CNITsolution/survey.cgi This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From superwen at 263.net Thu Mar 7 03:05:18 2002 From: superwen at 263.net (Kevin Zhu) Date: Thu, 7 Mar 2002 16:5:18 +0800 Subject: [uClinux-dev] Why the download process is so slow? Message-ID: <20020307080251.6D4A11D005F3B@smtp.263.net> Hi, All, The uClinux kernel file(B-Record) I compiled is about 2MB. I use the Mot's bbugv+ to download it into my MC68VZ328ADS board. With the baud rate 115200, it took me more than 5 hours to load it into RAM? Is this normal? Does anybody get some experience on this? (My machine is NT Workstation 4.0.) Thank you and best Regards Kevin Zhu ------------------------ ? _..--"\ /`'^`) ? ? ? ( .~\ \W//^~-' ? ? \_/ ( @) ? ( =-oOo-) ? I'm thinking about it... ? _.`^^- "o0/ ? ??(oO\ ###/ \`)???????????????? This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From chri at infis.univ.trieste.it Thu Mar 7 03:05:41 2002 From: chri at infis.univ.trieste.it (Christian Pellegrin) Date: Thu, 7 Mar 2002 09:05:41 +0100 (CET) Subject: [uClinux-dev] re: runtime flash at91 In-Reply-To: <1015446111.6614.2.camel@feedtheworms.win.trlabs.ca> Message-ID: On 6 Mar 2002, Sami Kibria wrote: > hey guys...i just have a really quick question... > > executing an application, can you write to flash with that atmel at91, Hi, it depends on which flash do you have. I ported the stuff from the EB55 example software to uclinux (the flash supported is at49bxxxx and at29bxxxx). I'm just cleaning out the crap before releasing it, if you want it early for testing just drop me a line. I'm not sure if Linux (the MTD layer or whatever) supports that flash directly, please check in the latest distro or in the CVS. Bye! This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From ducuy at tlnet.com.vn Thu Mar 7 03:40:04 2002 From: ducuy at tlnet.com.vn (VAN DUC UY) Date: Thu, 7 Mar 2002 15:40:04 +0700 Subject: [uClinux-dev] Samsung 4510 support... References: <005401c1c4f4$d2fc29d0$eccb718c@zythum> Message-ID: <003e01c1c5b3$adf77c60$9c0110ac@ducuy> Hi Mac, I downloaded all pakages from your site and i'm starting to make linux image. My board is used 4510 CPU and similar with SNDS100 (it can run vLinux 2.2.14 http://www.vitals.co.kr well). but I have some trouble when i make. The error output like following : /opt/uClinux/arm-elf/bin/gld: cannot open romfs.o: No such file or directory make[1]: *** [linux] Error 1 make[1]: Leaving directory `/home/share/kernel/uClinux/nctu/uClinux-samsung/linux-2.4.x' make: *** [linux] Error 1 May be i'm missing something when install tools. could you give me some hints for makeing kernel. Thanks in advance. DUC UY ----- Original Message ----- From: "Mac Wang" To: "uclinux-dev" Sent: Wednesday, March 06, 2002 4:53 PM Subject: [uClinux-dev] Samsung 4510 support... > Hi... > I have ported the uClinux 2.4.17 to this Samsung 4510 chip. It can enter > sash shell successfully, and do something like get IP by dhcpcd then telnet > to other host. But the serial/ethernet drivers are not fully tested. Anyone > interested it can visit my 4510 homepage below. Any suggestion is > appericated. > > http://mac.os.nctu.edu.tw > -- > Mac Wang > National Chiao Tung University > Department of Computer & Information Science > Operating System Lab. > E-mail: mac at os.nctu.edu.tw > > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From zl at imail.com.cn Thu Mar 7 04:00:02 2002 From: zl at imail.com.cn (zhanglei) Date: Thu, 7 Mar 2002 17:0:2 +0800 Subject: [uClinux-dev] Why the download process is so slow? Message-ID: <20020307090522.3589EAB5D5@e3500-2.email.com.cn> Please Make use of bbugv on Win98. Hi, All, The uClinux kernel file(B-Record) I compiled is about 2MB. I use the Mot's bbugv+ to download it into my MC68VZ328ADS board. With the baud rate 115200, it took me more than 5 hours to load it into RAM? Is this normal? Does anybody get some experience on this? (My machine is NT Workstation 4.0.) Thank you and best Regards Kevin Zhu ------------------------ ? _..--"\ /`'^`) ? ? ? ( .~\ \W//^~-' ? ? \_/ ( @) ? ( =-oOo-) ? I'm thinking about it... ? _.`^^- "o0/ ? ??(oO\ ###/ \`)???????????????? This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: LittleBoy.GIF Type: image/gif Size: 1021 bytes Desc: not available URL: From fly.chen at msa.hinet.net Thu Mar 7 05:12:36 2002 From: fly.chen at msa.hinet.net (Anderson Chen(HiNET) (E-mail)) Date: Thu, 7 Mar 2002 18:12:36 +0800 Subject: [uClinux-dev] About PPPoE In-Reply-To: <1015446111.6614.2.camel@feedtheworms.win.trlabs.ca> Message-ID: <000901c1c5c0$a34c64d0$1501a8c0@anderson> Hi, I am a new member. May anyone teach me how to setup uClinux with 2.4.17 kernel to run RP-PPPOE. When I download image file to M5272C3 board, I can not find any scripts. If I run "pppoe" directly, it shows some strange characters and then exits. Is any modification needed or not ?? This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From stefan.jonsson at faltcom.se Thu Mar 7 05:24:32 2002 From: stefan.jonsson at faltcom.se (Stefan Jonsson) Date: Thu, 7 Mar 2002 11:24:32 +0100 Subject: [uClinux-dev] simple compile for uClinux on ARM EB40LS Message-ID: Hello, I have recently began with this and have only made some changes to my resulting image.bin. What I wonder (that do not succeed for me) is how to compile a simple "hello world" (for instance) -program? I simply typed: "gcc -o hello.cgi hello.c" ... that seemed fine but the program did not work in either my Red Hat 7.2 environment or on EB40LS uClinux. So what can i do to compile own programs and fit them into my rom-filesystem? No, I'm not so comfortable with Linux yet, due to lack of experience in the subject, therefore a basic, descriptive and detailed answer would be most appriciated ;-) (any help on CGI with C on uClinux and what's neede 4 that would also b great) Thank you! Reagards, Stefan Jonsson Student @ University of Umea, Sweden. This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From dheeruch at cmcltd.com Thu Mar 7 05:52:44 2002 From: dheeruch at cmcltd.com (dheeru) Date: Thu, 7 Mar 2002 16:22:44 +0530 Subject: [uClinux-dev] boot error Message-ID: <002501c1c5c6$37c09910$6a0c10ac@DHEERU> Hello, I'm using uClinux-dist-20020220 distribution for M5307C3 board.I have compiled the code as per the documentation and downloaded the 'image.bin' file into the ram.While booting, the bootup sequence is stopped by giving message as: warning: unable to open an initial console. The binary image from net 'image-cadre3-5307-big.bin' is booting perfectly.Plz help me. The complete boot snapshot is: dBUG>go 20000 uClinux/COLDFIRE(m5307) COLDFIRE port done by Greg Ungerer, gerg at snapgear.com Modified for M5307 by Dave Miller, dmiller at intellistor.com Flat model support (C) 1998,1999 Kenneth Albanowski, D. Jeff Dionne On node 0 totalpages: 2048 zone(0): 0 pages. zone(1): 2048 pages. zone(2): 0 pages. Kernel command line: Calibrating delay loop... 59.80 BogoMIPS Memory available: 5704k/8192k RAM, 0k/0k ROM (662k kernel code, 196k data) kmem_create: Forcing size word alignment - vm_area_struct kmem_create: Forcing size word alignment - mm_struct kmem_create: Forcing size word alignment - filp Dentry-cache hash table entries: 1024 (order: 1, 8192 bytes) Inode-cache hash table entries: 512 (order: 0, 4096 bytes) kmem_create: Forcing size word alignment - inode_cache Mount-cache hash table entries: 512 (order: 0, 4096 bytes) kmem_create: Forcing size word alignment - bdev_cache kmem_create: Forcing size word alignment - cdev_cache Buffer-cache hash table entries: 1024 (order: 0, 4096 bytes) Page-cache hash table entries: 2048 (order: 1, 8192 bytes) POSIX conformance testing by UNIFIX Linux NET4.0 for Linux 2.4 Based upon Swansea University Computer Society NET3.039 kmem_create: Forcing size word alignment - sock Initializing RT netlink socket Starting kswapd kmem_create: Forcing size word alignment - file lock cache ColdFire internal UART serial driver version 1.00 ttyS0 at 0x100001c0 (irq = 73) is a builtin ColdFire UART ttyS1 at 0x10000200 (irq = 74) is a builtin ColdFire UART kmem_create: Forcing size word alignment - blkdev_requests block: 64 slots per queue, batch=16 RAMDISK driver initialized: 16 RAM disks of 4096K size 1024 blocksize ne.c:v1.10 9/23/94 Donald Becker (becker at scyld.com) Last modified Nov 1, 2000 by Paul Gortmaker NE*000 ethercard probe at 0x40000300: 00 50 ba d8 f4 1f eth0: NE2000 found at 0x40000300, using IRQ 27. SLIP: version 0.8.4-NET3.019-NEWTTY (dynamic channels, max=256). CSLIP: code copyright 1989 Regents of the University of California. Blkmem copyright 1998,1999 D. Jeff Dionne Blkmem copyright 1998 Kenneth Albanowski Blkmem 7 disk images: 0: F6CD4-249CD3 [VIRTUAL F6CD4-249CD3] (RO) 1: FFE00000-FFE1FFFF [VIRTUAL FFE00000-FFE1FFFF] (RW) 2: FFE20000-FFEDFFFF [VIRTUAL FFE20000-FFEDFFFF] (RW) 3: FFEE0000-FFEEFFFF [VIRTUAL FFEE0000-FFEEFFFF] (RW) 4: FFEF0000-FFEF3FFF [VIRTUAL FFEF0000-FFEF3FFF] (RW) 5: FFEF4000-FFEF7FFF [VIRTUAL FFEF4000-FFEF7FFF] (RW) 6: FFEF8000-FFEFFFFF [VIRTUAL FFEF8000-FFEFFFFF] (RW) PPP generic driver version 2.4.1 PPP MPPE compression module registered NET4: Linux TCP/IP 1.0 for NET4.0 IP Protocols: ICMP, UDP, TCP kmem_create: Forcing size word alignment - ip_dst_cache IP: routing cache hash table of 512 buckets, 4Kbytes TCP: Hash tables configured (established 512 bind 512) NET4: Unix domain sockets 1.0/SMP for Linux NET4.0. VFS: Mounted root (romfs filesystem) readonly. Freeing unused kernel memory: 24k freed (0xd0000 - 0xd5000) Warning: unable to open an initial console. Thanks & Rgs Dheeru. -------------- next part -------------- An HTML attachment was scrubbed... URL: From mac at young.url.com.tw Thu Mar 7 05:51:46 2002 From: mac at young.url.com.tw (Mac Wang) Date: Thu, 7 Mar 2002 18:51:46 +0800 Subject: [uClinux-dev] Samsung 4510 support... References: <005401c1c4f4$d2fc29d0$eccb718c@zythum> <003e01c1c5b3$adf77c60$9c0110ac@ducuy> Message-ID: <003201c1c5c6$13e425c0$eccb718c@zythum> Hi... I put romfs image as one section in kernel image, when you make kernel without romfs image, this error will occur... (I don't know how to use arm-elf-objcopy to link these two file together) Please follow these steps 1. make lib_only 2. make user_only 3. make romfs 4. make image you will get a error message here, just ignore it... 5. make then you will have two images in ~/image, image.rom and image.ram you can download image.ram to you board by ICE directly or program the image.rom to your flash. !!!notice!!! the console is on com0 19200Bps 8n1 ----- Original Message ----- From: "VAN DUC UY" To: Sent: Thursday, March 07, 2002 4:40 PM Subject: Re: [uClinux-dev] Samsung 4510 support... > Hi Mac, > I downloaded all pakages from your site and i'm starting to make linux > image. > My board is used 4510 CPU and similar with SNDS100 (it can run vLinux 2.2.14 > http://www.vitals.co.kr well). > but I have some trouble when i make. > The error output like following : > > /opt/uClinux/arm-elf/bin/gld: cannot open romfs.o: No such file or directory > make[1]: *** [linux] Error 1 > make[1]: Leaving directory > `/home/share/kernel/uClinux/nctu/uClinux-samsung/linux-2.4.x' > make: *** [linux] Error 1 > > > May be i'm missing something when install tools. could you give me some > hints for makeing kernel. > > Thanks in advance. > > DUC UY > > ----- Original Message ----- > From: "Mac Wang" > To: "uclinux-dev" > Sent: Wednesday, March 06, 2002 4:53 PM > Subject: [uClinux-dev] Samsung 4510 support... > > > > Hi... > > I have ported the uClinux 2.4.17 to this Samsung 4510 chip. It can > enter > > sash shell successfully, and do something like get IP by dhcpcd then > telnet > > to other host. But the serial/ethernet drivers are not fully tested. > Anyone > > interested it can visit my 4510 homepage below. Any suggestion is > > appericated. > > > > http://mac.os.nctu.edu.tw > > -- > > Mac Wang > > National Chiao Tung University > > Department of Computer & Information Science > > Operating System Lab. > > E-mail: mac at os.nctu.edu.tw > > > > This message resent by the uclinux-dev at uclinux.org list server > http://www.uClinux.org/ > > > > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From gerg at snapgear.com Thu Mar 7 07:41:09 2002 From: gerg at snapgear.com (gerg) Date: Thu, 07 Mar 2002 22:41:09 +1000 Subject: [uClinux-dev] Re: boot error References: <002501c1c5c6$37c09910$6a0c10ac@DHEERU> Message-ID: <3C875FE5.B7411FCA@snapgear.com> Hi Dheeru, dheeru wrote: > > Hello, > I'm using uClinux-dist-20020220 distribution for M5307C3 board.I have > compiled the code as per the documentation and downloaded the > 'image.bin' file into the ram.While booting, the bootup sequence is > stopped by giving message as: > warning: unable to open an initial console. The most common problem is that you are using an old genromfs. Check which genromfs ("type genromfs" will tell you where your genromfs is). Almost certainly there is an old one installed on your system, and it is in your path before the one that he m68k-elf-tools package contains (which is in /usr/local/bin). Find the old one and move or remove it. Regards Greg > The binary image from net 'image-cadre3-5307-big.bin' is booting > perfectly.Plz help me. > > The complete boot snapshot is: > > dBUG>go 20000 > uClinux/COLDFIRE(m5307) > COLDFIRE port done by Greg Ungerer, gerg at snapgear.com > Modified for M5307 by Dave Miller, dmiller at intellistor.com > Flat model support (C) 1998,1999 Kenneth Albanowski, D. Jeff Dionne > On node 0 totalpages: 2048 > zone(0): 0 pages. > zone(1): 2048 pages. > zone(2): 0 pages. > Kernel command line: > Calibrating delay loop... 59.80 BogoMIPS > Memory available: 5704k/8192k RAM, 0k/0k ROM (662k kernel code, 196k > data) > kmem_create: Forcing size word alignment - vm_area_struct > kmem_create: Forcing size word alignment - mm_struct > kmem_create: Forcing size word alignment - filp > Dentry-cache hash table entries: 1024 (order: 1, 8192 bytes) > Inode-cache hash table entries: 512 (order: 0, 4096 bytes) > kmem_create: Forcing size word alignment - inode_cache > Mount-cache hash table entries: 512 (order: 0, 4096 bytes) > kmem_create: Forcing size word alignment - bdev_cache > kmem_create: Forcing size word alignment - cdev_cache > Buffer-cache hash table entries: 1024 (order: 0, 4096 bytes) > Page-cache hash table entries: 2048 (order: 1, 8192 bytes) > POSIX conformance testing by UNIFIX > Linux NET4.0 for Linux 2.4 > Based upon Swansea University Computer Society NET3.039 > kmem_create: Forcing size word alignment - sock > Initializing RT netlink socket > Starting kswapd > kmem_create: Forcing size word alignment - file lock cache > ColdFire internal UART serial driver version 1.00 > ttyS0 at 0x100001c0 (irq = 73) is a builtin ColdFire UART > ttyS1 at 0x10000200 (irq = 74) is a builtin ColdFire UART > kmem_create: Forcing size word alignment - blkdev_requests > block: 64 slots per queue, batch=16 > RAMDISK driver initialized: 16 RAM disks of 4096K size 1024 blocksize > ne.c:v1.10 9/23/94 Donald Becker (becker at scyld.com) > Last modified Nov 1, 2000 by Paul Gortmaker > NE*000 ethercard probe at 0x40000300: 00 50 ba d8 f4 1f > eth0: NE2000 found at 0x40000300, using IRQ 27. > SLIP: version 0.8.4-NET3.019-NEWTTY (dynamic channels, max=256). > CSLIP: code copyright 1989 Regents of the University of California. > Blkmem copyright 1998,1999 D. Jeff Dionne > Blkmem copyright 1998 Kenneth Albanowski > Blkmem 7 disk images: > 0: F6CD4-249CD3 [VIRTUAL F6CD4-249CD3] (RO) > 1: FFE00000-FFE1FFFF [VIRTUAL FFE00000-FFE1FFFF] (RW) > 2: FFE20000-FFEDFFFF [VIRTUAL FFE20000-FFEDFFFF] (RW) > 3: FFEE0000-FFEEFFFF [VIRTUAL FFEE0000-FFEEFFFF] (RW) > 4: FFEF0000-FFEF3FFF [VIRTUAL FFEF0000-FFEF3FFF] (RW) > 5: FFEF4000-FFEF7FFF [VIRTUAL FFEF4000-FFEF7FFF] (RW) > 6: FFEF8000-FFEFFFFF [VIRTUAL FFEF8000-FFEFFFFF] (RW) > PPP generic driver version 2.4.1 > PPP MPPE compression module registered > NET4: Linux TCP/IP 1.0 for NET4.0 > IP Protocols: ICMP, UDP, TCP > kmem_create: Forcing size word alignment - ip_dst_cache > IP: routing cache hash table of 512 buckets, 4Kbytes > TCP: Hash tables configured (established 512 bind 512) > NET4: Unix domain sockets 1.0/SMP for Linux NET4.0. > VFS: Mounted root (romfs filesystem) readonly. > Freeing unused kernel memory: 24k freed (0xd0000 - 0xd5000) > Warning: unable to open an initial console. > > > Thanks & Rgs > Dheeru. -- ------------------------------------------------------------------------ Greg Ungerer -- Chief Software Wizard EMAIL: gerg at snapgear.com Snapgear PHONE: +61 7 3279 1822 825 Stanley St, FAX: +61 7 3279 1820 Woolloongabba, QLD, 4102, Australia WEB: www.snapgear.com This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From jody.pearson at cern.ch Thu Mar 7 09:55:13 2002 From: jody.pearson at cern.ch (Jody Pearson) Date: Thu, 07 Mar 2002 15:55:13 +0100 Subject: [uClinux-dev] Problems with running init ? References: <20020306214508.72425.qmail@web14309.mail.yahoo.com> Message-ID: <3C877F51.4772ECA3@cern.ch> Hi all, Well, I'm getting closer (teasingly closer) to getting to a shell on xcopilot with my compiled rom.. I am using ; VERSION = 2 PATCHLEVEL = 0 SUBLEVEL = 40 UCRELEASE = 0 and m68k-coff-gcc toolchain. Here are the current achievements; Basically - i believe that the kernel is trying to run /etc/init. I can put debug printk's in ${TOPDIR}/init/main.c, and they appear. The offending code is; if (!execute_command) { execve("/etc/init",argv_init,envp_init); execve("/bin/init",argv_init,envp_init); execve("/sbin/init",argv_init,envp_init); /* if this fails, fall through to original stuff */ pid = kernel_thread(do_rc, "/etc/rc", SIGCHLD); if (pid>0) while (pid != wait(&i)) /* nothing */; } while (1) { pid = kernel_thread(do_shell, execute_command ? execute_command : "/bin/sh", SIGCHLD); if (pid < 0) { printf("Fork failed in init\n\r"); continue; } while (1) if (pid == wait(&i)) // printk goes here.... break; printf("\n\rchild %d died with code %04x\n\r",pid,i); sync(); } return -1; Obviously from init/main.c If I insert a printk where the comment is I appears (in an endless loop). Has anyone seen this before ? ( I have attached the .config file as well) Thanks. Jody Screen output follows....... if [ -f romfs.img ]; then\ cat linux.text linux.data romfs.img > linux.bin;\ else\ cat linux.text linux.data > linux.bin;\ fi perl arch/m68knommu/platform/PalmVx/fixup.pl pilot /root/.xcopilot/ load /root/.xcopilot//pilot.rom memory.c,444 - 0x40620008 load /root/.xcopilot//pilot.rom memory.c,444 - 0x40620008 Attached to process. Start debugger I - CPU Started Fatal(custom.c,1419) Fatal(custom.c,1628) Fatal(custom.c,1628) 68EZ328 DragonBallEZ support (C) 2001 Empower Technologies, Inc Flat model support (C) 1998,1999 Kenneth Albanowski, D. Jeff Dionne start_mem is 0x2ce50 virtual_end is 0x100000 before free_area_init free_area_init -> start_mem is 0x2fe50 virtual_end is 0x100000 Console driver (without boot console): mono mc68328 40x26, 1 virtual console (max 63) Calibrating delay loop.. ok - 0.00 BogoMIPS memory start:7677c,memory_end:100000 Mem_init: start=7677c, end=100000 Memory available: 548k/1023k RAM, 696k/1024k ROM (476k kernel data, 328k code) Swansea University Computer Society NET3.035 for Linux 2.0 NET3: Unix domain sockets 0.13 for Linux NET3.035. LinuxDA ver 2.0.40.0 (root at pcitpdp20) (gcc version 2.7.2.3) #53 Thu Mar 7 15:45:51 CET 2002 MC68328 serial driver version 1.00 ttyS0 at 0xfffff900 (irq = 64) is a builtin MC68328 UART Ramdisk driver initialized : 16 ramdisks of 4096K size block write addr:10c27474,rd_request:10c3eec4, device name :Blkmem Blkmem 1 disk images:0: 10C60160-10CB315F (RO) VFS: Mounted root (romfs filesystem) readonly. Illegal instruction: cc7e ILL at 0x1001e -------------- next part -------------- # # Automatically generated make config: don't edit # CONFIG_LINUXDA=y # # Code maturity level options # # CONFIG_EXPERIMENTAL is not set # # Loadable module support # # CONFIG_MODULES is not set # # Flash memory chipset # # CONFIG_FLASH_SHARP is not set # CONFIG_FLASH_INTEL is not set # # Platform dependant setup # # CONFIG_M68000 is not set # CONFIG_M68328 is not set CONFIG_MC68EZ328=y # CONFIG_M68332 is not set # CONFIG_M68360 is not set # # Platform # CONFIG_PILOT=y # CONFIG_PALM3X is not set CONFIG_PALMVX=y # # General setup # CONFIG_NET=y CONFIG_SYSVIPC=y CONFIG_REDUCED_MEMORY=y CONFIG_BINFMT_FLAT=y # CONFIG_PILOT_MEMORY_DISPLAY is not set CONFIG_CONSOLE=y CONFIG_FRAMEBUFFER=y # CONFIG_PILOT_CONSOLE is not set # # Floppy, IDE, and other block devices # CONFIG_BLK_DEV_BLKMEM=y # # Additional Block Devices # # CONFIG_BLK_DEV_LOOP is not set # CONFIG_BLK_DEV_MD is not set CONFIG_BLK_DEV_RAM=y CONFIG_RD_RELEASE_BLOCKS=y # # Networking options # # CONFIG_FIREWALL is not set # CONFIG_NET_ALIAS is not set # CONFIG_INET is not set # # # # CONFIG_IPX is not set # CONFIG_ATALK is not set # CONFIG_AX25 is not set # CONFIG_NETLINK is not set # # SCSI support # # CONFIG_SCSI is not set # # Network device support # # CONFIG_NETDEVICES is not set # CONFIG_UCCS8900 is not set # # Filesystems # # CONFIG_QUOTA is not set # CONFIG_MINIX_FS is not set # CONFIG_EXT_FS is not set # CONFIG_EXT2_FS is not set # CONFIG_XIA_FS is not set # CONFIG_NLS is not set CONFIG_PROC_FS=y # CONFIG_HPFS_FS is not set # CONFIG_SYSV_FS is not set CONFIG_ROMFS_FS=y # CONFIG_UFS_FS is not set # # Character devices # # CONFIG_UMISC is not set CONFIG_68328_SERIAL=y CONFIG_68328_DIGI=y # CONFIG_USERIAL is not set # CONFIG_WATCHDOG is not set # # Kernel hacking # CONFIG_PROFILE=y CONFIG_PROFILE_SHIFT=2 From Frank_Woods at datamaxcorp.com Thu Mar 7 10:44:09 2002 From: Frank_Woods at datamaxcorp.com (Frank Woods) Date: Thu, 7 Mar 2002 10:44:09 -0500 Subject: [uClinux-dev] PCMCIA Support Message-ID: Does uCLinux currently have drivers for PC Cards? If so, where in the source tree are the files? Regards, Frank Woods -------------- next part -------------- An HTML attachment was scrubbed... URL: From skibria at win.trlabs.ca Thu Mar 7 11:15:22 2002 From: skibria at win.trlabs.ca (Sami Kibria) Date: 07 Mar 2002 10:15:22 -0600 Subject: [uClinux-dev] re: flash runtime In-Reply-To: References: Message-ID: <1015517724.14979.0.camel@feedtheworms.win.trlabs.ca> hi, thanks for you email...actually i was just starting to look into this type of processor...let me give you an example of an application i was looking at, lets say i had some device attached to the serial port, and the device keep sending us data, so what i would want to do is keep a data log of that data coming in from the serial port, can i save that data log to flash and continuously update that data using the at91 1mb flash...or do i need some type of mtd layer (ie memory technology device layer - subsystem used by linux)...does that make any sense??? thanks :) On Wed, 2002-03-06 at 13:43, Kendrick Hamilton wrote: > Sami, > I depends on if your code is running from RAM or from flash. If your > code is running from RAM (kernel and applications) then it should not be a > problem. Just make sure that nothing (including interrupt vectors) are > being pulled from flash. What processor are you using? > > > On 6 Mar 2002, Sami Kibria wrote: > > > hey guys...i just have a really quick question... > > > > can you allocate a certain area in flash for writing to, lets say we > > want to keep some data logging...can we do that, or like in the, tini > > (ds80 dual can high speed micro), that you can't write to flash while > > executing an application, can you write to flash with that atmel at91, > > or is not allowed? > > > > thanks... > > > > ps...just in case i am not inquiring about code shadowing... > > > > :) > > > > > > > > -- > Kendrick Hamilton E.I.T. > SED Systems, a division of Calian Ltd. > 18 Innovation Blvd. > PO Box 1464 > Saskatoon, Saskatchewan > Canada > S7N 3R1 > > Hamilton at sedsystems.ca > Tel: (306) 933-1453 > Fax: (306) 933-1486 > > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ > -- -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- Sami Kibria Researcher email: skibria at win.trlabs.ca TRLabs - Winnipeg, MB, CANADA http://www.win.trlabs.ca/~skibria -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- __ _ / / (_)__ __ ____ __ / /__/ / _ \/ // /\ \/ / . . . t h e c h o i c e o f a /____/_/_//_/\_,_/ /_/\_\ G N U g e n e r a t i o n -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From mirco_cremasco at hotmail.com Thu Mar 7 12:55:43 2002 From: mirco_cremasco at hotmail.com (mirco cremasco) Date: Thu, 07 Mar 2002 18:55:43 +0100 Subject: [uClinux-dev] problems building elf tool Message-ID: >From: David McCullough >Reply-To: uclinux-dev at uclinux.org >To: uclinux-dev at uclinux.org >Subject: Re: [uClinux-dev] problems building elf tool >Date: Wed, 6 Mar 2002 09:12:36 +1000 > > >Jivin mirco cremasco lays it down ... > > I've successfully rebuild-ed gcc-2.95.3, binutils-2.10, genromfs-0.5.1 >and > > now I'm able to compile uClinux-dist-20020220 for a m5272c3 board. > > But there is still a problem: the generated bin image is too big!! > > images/romfs.bin is 3.9M !! (using m68k-els-tools-20020218 it is 1M) > > looking into romfs/bin directory I've found that telnet (as example) is >326K > > in place of 70K compiled with m68k-els-tools-20020218. > > I didn't changed any compiler option uClinux-dist makefiles. > > There are any option for compile gcc?? > > I used > > ./configure --target=m68k-elf --prefix=/opt/uClinux \ > > --with-headers=/opt/uClinux/uClinux-dist/lib/libc/include > > > > make > > make install > > >Use the build-uclinux-tools.sh to build the tools, it does things to make >the install work properly. > >I would guess that you have installed "ld" and not replaced it with >ld-elf2flt as required, thus you have a romfs/bin full of ELF files and >not flat files, > >Cheers, >Davidm > >-- >David McCullough: Ph: +61 7 3435 2815 http://www.SnapGear.com >davidm at snapgear.com Fx: +61 7 3891 3630 825 Stanley St., W'gabba QLD >4102, Oz >This message resent by the uclinux-dev at uclinux.org list server >http://www.uClinux.org/ I've downloaded al CVS and tar.gz files as suggeted inside build-uclinux-tools-sh script. Everything is in one directory and I'm root but the script stop with this error: STAGE 2 - complete -------------------------------------------------------- STAGE 3 - needs building /bin/sh: m68k-elf-gcc: command not found /bin/sh: include/bits/uClibc_config.h: No such file or directory Why? Tnks Mirco Cremasco _________________________________________________________________ MSN Photos ? il metodo pi? semplice per condividere e stampare le tue foto: http://photos.msn.it/Support/WorldWide.aspx This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From Dave_Pfaltzgraff at patapsco.com Thu Mar 7 16:03:04 2002 From: Dave_Pfaltzgraff at patapsco.com (Dave_Pfaltzgraff at patapsco.com) Date: Thu, 07 Mar 2002 16:03:04 -0500 Subject: [uClinux-dev] Questions on the IP stack Message-ID: <85256B75.0073A39F.00@patapsco.com> Dave Pfaltzgraff at PATAPSCO 03/07/2002 04:03 PM We have a situation where uCLinux is passing data up to a central server using a TCP connection. Because the uCLinux application is 'mission critical', we've enabled the watchdog and use the SIGALRM signal to generate a 10 msec 'interrupt' in the user program. The user level interrupt program then addresses the watchdog. This all seems to work. However, every once in a long while (3 or 4 days), the central server gets too busy to service the TCP traffic. I would have thought that Linux would let the TCP buffers back up and when they are full and the user application attempts to write would return an error condition. However, what seems to be happening is that somewhere in Linux the application does not get processor time for a long enough period that the watchdog times out. (Unfortunately, it's an embedded system and it reboots so there's no trace of what was going on just prior to the reset!) We have increased the number of buffers in fec.c for the original 16 to 32 and later to 64. The first change made a slight improvement, but extending it from 32 to 64 didn't seem to make any further improvement. We also increased the watchdog time to several seconds but that had no apparent effect. Anyone with IP stack experience willing to comment on what may be going on? Thanks, Dave This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From andyk at poliac.com Thu Mar 7 17:38:14 2002 From: andyk at poliac.com (Andrew P. Karels) Date: Thu, 7 Mar 2002 16:38:14 -0600 Subject: [uClinux-dev] Questions on the IP stack References: <85256B75.0073A39F.00@patapsco.com> Message-ID: <001801c1c628$cac7ae80$4301a8c0@andyk> Dave, We ran into what may be a similar situation. We tried putting out watchdog tickle in a user app also, but under very heavy network traffic (like for example flood pinging with large packets) the user app never got processing time. It turned out that the scheduler first looks for bottom halves of interrupts that need processing before anything else, and since processing packets off the ethernet device is done in the bottom half of an interrupt, all the processor time was taken up handling ethernet packets. What we ended up doing, at least for now, is tickling the watchdog in the scheduler and going into a infinite loop (to prevent the scheduler from running) when an exception is encountered. Andy K. ----- Original Message ----- From: To: Sent: Thursday, March 07, 2002 3:03 PM Subject: [uClinux-dev] Questions on the IP stack > > > > > Dave Pfaltzgraff at PATAPSCO > 03/07/2002 04:03 PM > > We have a situation where uCLinux is passing data up to a central server using a > TCP connection. Because the uCLinux application is 'mission critical', we've > enabled the watchdog and use the SIGALRM signal to generate a 10 msec > 'interrupt' in the user program. The user level interrupt program then addresses > the watchdog. This all seems to work. > > However, every once in a long while (3 or 4 days), the central server gets too > busy to service the TCP traffic. I would have thought that Linux would let the > TCP buffers back up and when they are full and the user application attempts to > write would return an error condition. However, what seems to be happening is > that somewhere in Linux the application does not get processor time for a long > enough period that the watchdog times out. (Unfortunately, it's an embedded > system and it reboots so there's no trace of what was going on just prior to the > reset!) > > We have increased the number of buffers in fec.c for the original 16 to 32 and > later to 64. The first change made a slight improvement, but extending it from > 32 to 64 didn't seem to make any further improvement. We also increased the > watchdog time to several seconds but that had no apparent effect. > > Anyone with IP stack experience willing to comment on what may be going on? > > Thanks, > Dave > > > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From davidm at snapgear.com Thu Mar 7 18:52:23 2002 From: davidm at snapgear.com (David McCullough) Date: Fri, 8 Mar 2002 09:52:23 +1000 Subject: [uClinux-dev] Problems with running init ? In-Reply-To: <3C877F51.4772ECA3@cern.ch>; from jody.pearson@cern.ch on Thu, Mar 07, 2002 at 03:55:13PM +0100 References: <20020306214508.72425.qmail@web14309.mail.yahoo.com> <3C877F51.4772ECA3@cern.ch> Message-ID: <20020308095223.A30299@beast.internal.moreton.com.au> Jivin Jody Pearson lays it down ... > Hi all, > > Well, I'm getting closer (teasingly closer) to getting to a shell on xcopilot > with my compiled rom.. The uClinux distribution has an xcopilot target which will build and run in xcopilot if you want to save yourself some pain ;-) You will need the m68k-elf-tools though, Cheers, Davidm -- David McCullough: Ph: +61 7 3435 2815 http://www.SnapGear.com davidm at snapgear.com Fx: +61 7 3891 3630 825 Stanley St., W'gabba QLD 4102, Oz This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From davidm at snapgear.com Thu Mar 7 18:57:11 2002 From: davidm at snapgear.com (David McCullough) Date: Fri, 8 Mar 2002 09:57:11 +1000 Subject: [uClinux-dev] problems building elf tool In-Reply-To: ; from mirco_cremasco@hotmail.com on Thu, Mar 07, 2002 at 06:55:43PM +0100 References: Message-ID: <20020308095711.B30299@beast.internal.moreton.com.au> Jivin mirco cremasco lays it down ... ... > I've downloaded al CVS and tar.gz files as suggeted inside > build-uclinux-tools-sh script. > Everything is in one directory and I'm root > but the script stop with this error: > > STAGE 2 - complete > -------------------------------------------------------- > STAGE 3 - needs building > /bin/sh: m68k-elf-gcc: command not found > /bin/sh: include/bits/uClibc_config.h: No such file or directory You need to have /usr/local/bin (or PREFIX/bin if you change it) in your path. The next version of the script will do this for you ;-) Cheers, Davidm -- David McCullough: Ph: +61 7 3435 2815 http://www.SnapGear.com davidm at snapgear.com Fx: +61 7 3891 3630 825 Stanley St., W'gabba QLD 4102, Oz This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From superwen at 263.net Thu Mar 7 19:44:12 2002 From: superwen at 263.net (kz) Date: Fri, 8 Mar 2002 8:44:12 +0800 Subject: [uClinux-dev] Why the download process is so slow? Message-ID: <20020308004113.566FB1D005EDB@smtp.263.net> Thank you, Mr.Zhang, But where can I get one? When I run the bbugv+ I had in Win98 machine(MSDOS prompt), it acts the same with or without the ADS board connected. And if I use "md" command to display the content in memory, it just displayed the same thing like "30 30 30 30", which is absolutely not the right thing I downloaded. Please advises. Thanks a lot. Kevin Zhu >Please Make use of bbugv on Win98. > > > > >Hi, All, > >The uClinux kernel file(B-Record) I compiled is about 2MB. I use the Mot's bbugv+ to download it into my MC68VZ328ADS board. With the baud rate 115200, it took me more than 5 hours to load it into RAM? Is this normal? > >Does anybody get some experience on this? (My machine is NT Workstation 4.0.) > >Thank you and best Regards >Kevin Zhu > > > >------------------------ >? _..--"\ /`'^`) ? ? ? > ( .~\ \W//^~-' ? ? > \_/ ( @) ? > ( =-oOo-) ? I'm thinking about it... ? > _.`^^- "o0/ ? >??(oO\ ###/ \`)???????????????? > >This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ ? ?? kz superwen at 263.net This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From gerg at snapgear.com Thu Mar 7 19:23:50 2002 From: gerg at snapgear.com (Greg Ungerer) Date: Fri, 08 Mar 2002 10:23:50 +1000 Subject: [uClinux-dev] Questions on the IP stack References: <85256B75.0073A39F.00@patapsco.com> <001801c1c628$cac7ae80$4301a8c0@andyk> Message-ID: <3C880496.8030504@snapgear.com> Hi Dave, Andrew, Tickling the watchdog should really be done in a driver. It is relatively strait forward to implement a watchdog driver with all the required logic. You can use the kernel timer with a clean API (use add_timer), so nasty hacks are required. Have a look at ~/linux/drivers/char/mcfwatchdog.c as an example. Regards Greg Andrew P. Karels wrote: > Dave, > > We ran into what may be a similar situation. We tried putting out watchdog > tickle in a user app also, but under very heavy network traffic (like for > example flood pinging with large packets) the user app never got processing > time. It turned out that the scheduler first looks for bottom halves of > interrupts that need processing before anything else, and since processing > packets off the ethernet device is done in the bottom half of an interrupt, > all the processor time was taken up handling ethernet packets. What we ended > up doing, at least for now, is tickling the watchdog in the scheduler and > going into a infinite loop (to prevent the scheduler from running) when an > exception is encountered. > > Andy K. > > ----- Original Message ----- > From: > To: > Sent: Thursday, March 07, 2002 3:03 PM > Subject: [uClinux-dev] Questions on the IP stack > > > >> >> >> >>Dave Pfaltzgraff at PATAPSCO >>03/07/2002 04:03 PM >> >>We have a situation where uCLinux is passing data up to a central server >> > using a > >>TCP connection. Because the uCLinux application is 'mission critical', >> > we've > >>enabled the watchdog and use the SIGALRM signal to generate a 10 msec >>'interrupt' in the user program. The user level interrupt program then >> > addresses > >>the watchdog. This all seems to work. >> >>However, every once in a long while (3 or 4 days), the central server gets >> > too > >>busy to service the TCP traffic. I would have thought that Linux would let >> > the > >>TCP buffers back up and when they are full and the user application >> > attempts to > >>write would return an error condition. However, what seems to be happening >> > is > >>that somewhere in Linux the application does not get processor time for a >> > long > >>enough period that the watchdog times out. (Unfortunately, it's an >> > embedded > >>system and it reboots so there's no trace of what was going on just prior >> > to the > >>reset!) >> >>We have increased the number of buffers in fec.c for the original 16 to 32 >> > and > >>later to 64. The first change made a slight improvement, but extending it >> > from > >>32 to 64 didn't seem to make any further improvement. We also increased >> > the > >>watchdog time to several seconds but that had no apparent effect. >> >>Anyone with IP stack experience willing to comment on what may be going >> > on? > >>Thanks, >>Dave >> >> >>This message resent by the uclinux-dev at uclinux.org list server >> > http://www.uClinux.org/ > > > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ > > -- ------------------------------------------------------------------------ Greg Ungerer -- Chief Software Wizard EMAIL: gerg at snapgear.com SnapGear PHONE: +61 7 3435 2888 825 Stanley St, FAX: +61 7 3891 3630 Woolloongabba, QLD, 4102, Australia WEB: www.snapgear.com This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From Sanjeev at coronanetworks.com Thu Mar 7 19:53:21 2002 From: Sanjeev at coronanetworks.com (Sanjeev Chakravarty) Date: Thu, 7 Mar 2002 16:53:21 -0800 Subject: [uClinux-dev] Questions on the IP stack Message-ID: Hi Andrew, My understanding is that the 2.4.x kernel TCP/IP stack implementation doesn't use bottom half approach instead uses software interrupt. Sanjeev -----Original Message----- From: Andrew P. Karels [mailto:andyk at poliac.com] Sent: Thursday, March 07, 2002 2:38 PM To: uclinux-dev at uclinux.org Subject: Re: [uClinux-dev] Questions on the IP stack Dave, We ran into what may be a similar situation. We tried putting out watchdog tickle in a user app also, but under very heavy network traffic (like for example flood pinging with large packets) the user app never got processing time. It turned out that the scheduler first looks for bottom halves of interrupts that need processing before anything else, and since processing packets off the ethernet device is done in the bottom half of an interrupt, all the processor time was taken up handling ethernet packets. What we ended up doing, at least for now, is tickling the watchdog in the scheduler and going into a infinite loop (to prevent the scheduler from running) when an exception is encountered. Andy K. ----- Original Message ----- From: To: Sent: Thursday, March 07, 2002 3:03 PM Subject: [uClinux-dev] Questions on the IP stack > > > > > Dave Pfaltzgraff at PATAPSCO > 03/07/2002 04:03 PM > > We have a situation where uCLinux is passing data up to a central server using a > TCP connection. Because the uCLinux application is 'mission critical', we've > enabled the watchdog and use the SIGALRM signal to generate a 10 msec > 'interrupt' in the user program. The user level interrupt program then addresses > the watchdog. This all seems to work. > > However, every once in a long while (3 or 4 days), the central server gets too > busy to service the TCP traffic. I would have thought that Linux would let the > TCP buffers back up and when they are full and the user application attempts to > write would return an error condition. However, what seems to be happening is > that somewhere in Linux the application does not get processor time for a long > enough period that the watchdog times out. (Unfortunately, it's an embedded > system and it reboots so there's no trace of what was going on just prior to the > reset!) > > We have increased the number of buffers in fec.c for the original 16 to 32 and > later to 64. The first change made a slight improvement, but extending it from > 32 to 64 didn't seem to make any further improvement. We also increased the > watchdog time to several seconds but that had no apparent effect. > > Anyone with IP stack experience willing to comment on what may be going on? > > Thanks, > Dave > > > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From robertd at vantagecontrols.com Thu Mar 7 19:52:31 2002 From: robertd at vantagecontrols.com (Robert Daniels) Date: Thu, 7 Mar 2002 17:52:31 -0700 Subject: [uClinux-dev] Shared Memory... at last??? Message-ID: <002e01c1c63b$8863da60$01000001@intranet.transera.com> I recently discovered that having system V ipc shared memory would be a nice thing, so after taking a look at what it's problems were, I whipped up a quick hack that gives me some basic shm support. Since this has been a question that pops up every now and then, I'm including a patch for shm.c(2.4.x kernel), you'll also have to make sure your sys_ipc call is correct too (I had to un-ifdef the section in sys_m68k.c for shm calls). I admit that I'm not a linux kernel guru, so if someone sees something in this hack that is bad, please let me know! Thanks! Robert Daniels Vantage Controls Inc. robertd at vantagecontrols.inc -------------- next part -------------- A non-text attachment was scrubbed... Name: shm.patch Type: application/octet-stream Size: 9646 bytes Desc: not available URL: From superwen at 263.net Thu Mar 7 23:02:26 2002 From: superwen at 263.net (Kevin Zhu) Date: Fri, 8 Mar 2002 12:2:26 +0800 Subject: [uClinux-dev] About uClinux ctr0_fixed.S Message-ID: <20020308035959.477C41D005E4B@smtp.263.net> Hi, All, I want port the uClinux 2.4 to my VZ board, and I got a question in Memory map part. If I want to build a flash-based kernel with my memory map, I think maybe I need to do some changes in rom.ld & ctr0_rom.S, which is actually ctro_fixed.S. But I can't find the "chip select" settings in either files. Doesn't that mean what I've done in rom.ld is actually not applied in the real board? Please advises.Thank you. Have a nice day. Kevin Zhu ------------------------ ? _..--"\ /`'^`) ? ? ? ( .~\ \W//^~-' ? ? \_/ ( @) ? ( =-oOo-) ? I'm thinking about it... ? _.`^^- "o0/ ? ??(oO\ ###/ \`)???????????????? This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From gerg at snapgear.com Thu Mar 7 22:58:30 2002 From: gerg at snapgear.com (Greg Ungerer) Date: Fri, 08 Mar 2002 13:58:30 +1000 Subject: [uClinux-dev] Shared Memory... at last??? References: <002e01c1c63b$8863da60$01000001@intranet.transera.com> Message-ID: <3C8836E6.4010603@snapgear.com> Hi Robert, Robert Daniels wrote: > I recently discovered that having system V ipc shared memory would be a nice > thing, so after taking a look at what it's problems were, I whipped up a > quick hack that gives me some basic shm support. Since this has been a > question that pops up every now and then, I'm including a patch for > shm.c(2.4.x kernel), you'll also have to make sure your sys_ipc call is > correct too (I had to un-ifdef the section in sys_m68k.c for shm calls). I > admit that I'm not a linux kernel guru, so if someone sees something in this > hack that is bad, please let me know! Thanks! Looks ok at a quick look. I have CVS'ed it. No harm really, won't hurt anyone who doesn't try to use it. Regards Greg ------------------------------------------------------------------------ Greg Ungerer -- Chief Software Wizard EMAIL: gerg at snapgear.com SnapGear PHONE: +61 7 3435 2888 825 Stanley St, FAX: +61 7 3891 3630 Woolloongabba, QLD, 4102, Australia WEB: www.snapgear.com This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From kj_lin at accton.com.tw Fri Mar 8 00:10:31 2002 From: kj_lin at accton.com.tw (kj_lin at accton.com.tw) Date: Fri, 8 Mar 2002 13:10:31 +0800 Subject: [uClinux-dev] The cflags of the arm-elf Message-ID: Hi all, The arm-elf toolchain has multilib version of uClibc and STDC++ library and can be switched by the cflags. The cflags seems can be "-fpic", "-mapcs-26", "-mbig-endian" and " -msingle-pic-base". I had tested them by simple C code and C++ code on my platform. All of them were compiled OK. But when executed, some worked, some failed. Testing by simple hello.c(printf("Hello World C!\n");): arm-elf-gcc -Wl,-elf2flt -o hello hello.c ---> executed ok. arm-elf-gcc -fpic -Wl,-elf2flt -o hello hello.c --> executed fail. arm-elf-gcc -mapcs-26 -Wl,-elf2flt -o hello hello.c --> executed ok. arm-elf-gcc -msingle-pic-base -Wl,-elf2flt -o hello hello.c --> executed ok . Testing by simple hello.cxx(cout<<"Hello World C++!\n";): arm-elf-g++ -Wl,-elf2flt -o hello hello.cxx ---> executed fail. arm-elf-g++ -fpic -Wl,-elf2flt -o hello hello.cxx --> executed fail. arm-elf-g++ -mapcs-26 -Wl,-elf2flt -o hello hello.cxx --> executed ok. arm-elf-g++ -msingle-pic-base -Wl,-elf2flt -o hello hello.cxx --> executed ok. I am curious that what is the difference between them and what is the most suitable option for my platform? My board's CPU is ARM940T and is running uclinux-2.4.10-u2 kernel. Thanks, KJ This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From asky at syncom.com.tw Fri Mar 8 01:18:11 2002 From: asky at syncom.com.tw (Jiun-Shian Ho) Date: Fri, 8 Mar 2002 14:18:11 +0800 Subject: [uClinux-dev] Does anyone have experence of using visionProbe-II from WindRiver Message-ID: <001501c1c669$07deeb40$6701a8c0@syncom.com.tw> Hi, all I am new to use uClinux. My Board is Samsung S3C4510 ARM7TDMI. I use pSOS before. I'm glad that Mac Wang release distribution for this chip By following the steps, I have compiled success. However, my tool, visionProbe-II from WindRiver, can not recognize file format of image.rom I tried to transfer linux-2.4.x/linux to flat-binary format by using convert tool from visionProbe It transfers but can not work/boot-up. And then I try to debug it, but can not find source code of libgcc1.S Does anyone have experience of using visionProbe-II from WindRiver on uClinux ?! Could be kind to give me some opinions or suggestions ?! Best regrds This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From ramkumar at intotoinc.com Fri Mar 8 02:36:41 2002 From: ramkumar at intotoinc.com (Ram Kumar K V) Date: Thu, 7 Mar 2002 23:36:41 -0800 (PST) Subject: [uClinux-dev] uClinux - MIPS Message-ID: <200203080736.g287af519193@intotoinc.com> Hi All, I am getting the Below Error, when i tried to read a file on the Host through NFS. >RPC: sendmsg returned error 128 >nfs: RPC call returned error 128 My target is MIPS board and i am using uClinux 2.4.14. Interesting issue is i did not get any problem, when i tried to write to a file on the Host through NFS. Please coment on this. thanks, Ramkumar. This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From eauth at softsys.co.at Fri Mar 8 05:10:46 2002 From: eauth at softsys.co.at (Erwin Authried) Date: Fri, 8 Mar 2002 11:10:46 +0100 Subject: [uClinux-dev] PCMCIA Support Message-ID: <01C1C691.E62F4660@smithwicks.softsys.co.at> Frank Woods[SMTP:Frank_Woods at datamaxcorp.com] wrote: > > Does uCLinux currently have drivers for PC Cards? If so, where in the > source tree are the files? > > Regards, > Frank Woods > > The drivers are in the standard linux distribution, in drivers/pcmcia, drivers/net, drivers/net/wireless, and they can be used with uClinux without problems, even without module support. The only part that you have to write from scratch are the socket services, depending on the hardware that you use to interface to your cards. If you don't need hot plugging, you can use my modified version of the cardctl utility to bind a driver to your card in your startup script: http://uclinux.home.at/midori/apps/pcmcia-cs-3.1.29_2.mlz -Erwin -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/ms-tnef Size: 1724 bytes Desc: not available URL: From davidm at snapgear.com Fri Mar 8 08:18:06 2002 From: davidm at snapgear.com (David McCullough) Date: Fri, 8 Mar 2002 23:18:06 +1000 Subject: [uClinux-dev] The cflags of the arm-elf In-Reply-To: ; from kj_lin@accton.com.tw on Fri, Mar 08, 2002 at 01:10:31PM +0800 References: Message-ID: <20020308231806.A14423@beast.internal.moreton.com.au> Jivin kj_lin at accton.com.tw lays it down ... > Hi all, > > > The arm-elf toolchain has multilib version of uClibc and STDC++ library and > > can be switched by the cflags. > > The cflags seems can be "-fpic", "-mapcs-26", "-mbig-endian" and " > > -msingle-pic-base". > > I had tested them by simple C code and C++ code on my platform. > > All of them were compiled OK. > > But when executed, some worked, some failed. > > > > Testing by simple hello.c(printf("Hello World C!\n");): > > arm-elf-gcc -Wl,-elf2flt -o hello hello.c ---> executed ok. > > arm-elf-gcc -fpic -Wl,-elf2flt -o hello hello.c --> executed fail. > > arm-elf-gcc -mapcs-26 -Wl,-elf2flt -o hello hello.c --> executed ok. > > arm-elf-gcc -msingle-pic-base -Wl,-elf2flt -o hello hello.c --> executed ok > > . > > > > Testing by simple hello.cxx(cout<<"Hello World C++!\n";): > > arm-elf-g++ -Wl,-elf2flt -o hello hello.cxx ---> executed fail. > > arm-elf-g++ -fpic -Wl,-elf2flt -o hello hello.cxx --> executed fail. > > arm-elf-g++ -mapcs-26 -Wl,-elf2flt -o hello hello.cxx --> executed ok. > > arm-elf-g++ -msingle-pic-base -Wl,-elf2flt -o hello hello.cxx --> executed > > ok. > > > > I am curious that what is the difference between them and what is the most > > suitable option for my platform? > > My board's CPU is ARM940T and is running uclinux-2.4.10-u2 kernel. -fpic is not suitable for uClinux. It is just PIC code but it doesn't know how to manage it data area properly under uClinux. The 1st and 3rd above are fully relocated at runtime. They work fine but both require more RAM as each time you run the same program it makes a copy of the text area and relocates it. I guess that you process can run both mapcs-26 and whatever the default target is. -msingle-pic-base is suitable if you want to do execute in place (ie., from rom) which saves a lot of memory as the text segment is shared between all instances of a program and doesn't need any relocation. Cheers, Davidm -- David McCullough: Ph: +61 7 3435 2815 http://www.SnapGear.com davidm at snapgear.com Fx: +61 7 3891 3630 825 Stanley St., W'gabba QLD 4102, Oz This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From andyk at poliac.com Fri Mar 8 10:34:28 2002 From: andyk at poliac.com (Andrew P. Karels) Date: Fri, 8 Mar 2002 09:34:28 -0600 Subject: [uClinux-dev] Questions on the IP stack References: <85256B75.0073A39F.00@patapsco.com> <001801c1c628$cac7ae80$4301a8c0@andyk> <3C880496.8030504@snapgear.com> Message-ID: <001601c1c6b6$c1f0dd20$4301a8c0@andyk> Actually, we tried to do the tickle in a timer also (in the kernel, not in a seperate driver) and still would have occational resets, though much less frequently than when the tickle was in a user app. (Bottom halves are also done before timers in the scheduler.) Incidently, I should have mentioned we where using version 2.0.x of the kernel. Andy K. ----- Original Message ----- From: "Greg Ungerer" To: Sent: Thursday, March 07, 2002 6:23 PM Subject: Re: [uClinux-dev] Questions on the IP stack > > Hi Dave, Andrew, > > Tickling the watchdog should really be done in a driver. > It is relatively strait forward to implement a watchdog > driver with all the required logic. You can use the > kernel timer with a clean API (use add_timer), so nasty > hacks are required. > > Have a look at ~/linux/drivers/char/mcfwatchdog.c > as an example. > > Regards > Greg > > > > Andrew P. Karels wrote: > > > Dave, > > > > We ran into what may be a similar situation. We tried putting out watchdog > > tickle in a user app also, but under very heavy network traffic (like for > > example flood pinging with large packets) the user app never got processing > > time. It turned out that the scheduler first looks for bottom halves of > > interrupts that need processing before anything else, and since processing > > packets off the ethernet device is done in the bottom half of an interrupt, > > all the processor time was taken up handling ethernet packets. What we ended > > up doing, at least for now, is tickling the watchdog in the scheduler and > > going into a infinite loop (to prevent the scheduler from running) when an > > exception is encountered. > > > > Andy K. > > > > ----- Original Message ----- > > From: > > To: > > Sent: Thursday, March 07, 2002 3:03 PM > > Subject: [uClinux-dev] Questions on the IP stack > > > > > > > >> > >> > >> > >>Dave Pfaltzgraff at PATAPSCO > >>03/07/2002 04:03 PM > >> > >>We have a situation where uCLinux is passing data up to a central server > >> > > using a > > > >>TCP connection. Because the uCLinux application is 'mission critical', > >> > > we've > > > >>enabled the watchdog and use the SIGALRM signal to generate a 10 msec > >>'interrupt' in the user program. The user level interrupt program then > >> > > addresses > > > >>the watchdog. This all seems to work. > >> > >>However, every once in a long while (3 or 4 days), the central server gets > >> > > too > > > >>busy to service the TCP traffic. I would have thought that Linux would let > >> > > the > > > >>TCP buffers back up and when they are full and the user application > >> > > attempts to > > > >>write would return an error condition. However, what seems to be happening > >> > > is > > > >>that somewhere in Linux the application does not get processor time for a > >> > > long > > > >>enough period that the watchdog times out. (Unfortunately, it's an > >> > > embedded > > > >>system and it reboots so there's no trace of what was going on just prior > >> > > to the > > > >>reset!) > >> > >>We have increased the number of buffers in fec.c for the original 16 to 32 > >> > > and > > > >>later to 64. The first change made a slight improvement, but extending it > >> > > from > > > >>32 to 64 didn't seem to make any further improvement. We also increased > >> > > the > > > >>watchdog time to several seconds but that had no apparent effect. > >> > >>Anyone with IP stack experience willing to comment on what may be going > >> > > on? > > > >>Thanks, > >>Dave > >> > >> > >>This message resent by the uclinux-dev at uclinux.org list server > >> > > http://www.uClinux.org/ > > > > > > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ > > > > > > > -- > ------------------------------------------------------------------------ > Greg Ungerer -- Chief Software Wizard EMAIL: gerg at snapgear.com > SnapGear PHONE: +61 7 3435 2888 > 825 Stanley St, FAX: +61 7 3891 3630 > Woolloongabba, QLD, 4102, Australia WEB: www.snapgear.com > > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From mirco_cremasco at hotmail.com Fri Mar 8 11:38:59 2002 From: mirco_cremasco at hotmail.com (mirco cremasco) Date: Fri, 08 Mar 2002 17:38:59 +0100 Subject: [uClinux-dev] problems building elf tool Message-ID: >From: David McCullough >Reply-To: uclinux-dev at uclinux.org >To: uclinux-dev at uclinux.org >Subject: Re: [uClinux-dev] problems building elf tool >Date: Fri, 8 Mar 2002 09:57:11 +1000 > > >Jivin mirco cremasco lays it down ... >... > > I've downloaded al CVS and tar.gz files as suggeted inside > > build-uclinux-tools-sh script. > > Everything is in one directory and I'm root > > but the script stop with this error: > > > > STAGE 2 - complete > > -------------------------------------------------------- > > STAGE 3 - needs building > > /bin/sh: m68k-elf-gcc: command not found > > /bin/sh: include/bits/uClibc_config.h: No such file or directory > >You need to have /usr/local/bin (or PREFIX/bin if you change it) in >your path. > >The next version of the script will do this for you ;-) > >Cheers, >Davidm > >-- >David McCullough: Ph: +61 7 3435 2815 http://www.SnapGear.com >davidm at snapgear.com Fx: +61 7 3891 3630 825 Stanley St., W'gabba QLD >4102, Oz >This message resent by the uclinux-dev at uclinux.org list server >http://www.uClinux.org/ Sorry if I'm boring you. I've done this: 1) I changed host, now on a Debian Linux. 2) modified in build-uclinux-tools.sh: -> MAKE = make (gmake unrecognized) -> PREFIX = /opt/uClinux 3) put /opt/uClinux/bin in $PATH exec ./build-uclinux-tools.sh build 2>$1 it stop width: STAGE 3 - needs building /bin/sh: include/bits/uClibc_config.h: No such file or directory make: *** [uClibc_config] Error 1 so: mkdir uClibc/include/bits and rerun build-uclinux-tools.sh width continue command it stop again, so I added 'cp include/bits/uClibc_config.h .' to make it work ./build-uclinux-tools.sh continue 2>$1 it stop with this output (long): ... make GCC_FOR_TARGET="/opt/uClinux/m68k-elf-gcc/gcc/xgcc -B/opt/uClinux/m68k-elf-gcc/gcc/ -B/opt/uClinux/m68k-elf/bin/ -I/opt/uClinux/m68k-elf/include" \ AR_FOR_TARGET="m68k-elf-ar" \ AR_FLAGS_FOR_TARGET="rc" \ CC="gcc" CFLAGS="-g -O2" \ HOST_PREFIX="" HOST_PREFIX_1="loser-" \ LANGUAGES="c proto gcov " \ LIBGCC2_CFLAGS="-O2 -DCROSS_COMPILE -DIN_GCC -g -O2 -I./include -g1 -DIN_LIBGCC2 -D__GCC_FLOAT_NOT_NEEDED " libgcc2.a make[3]: Entering directory `/opt/uClinux/m68k-elf-gcc/gcc' rm -f tmplibgcc2.a for name in _muldi3 _divdi3 _moddi3 _udivdi3 _umoddi3 _negdi2 _lshrdi3 _ashldi3 _ashrdi3 _ffsdi2 _udiv_w_sdiv _udivmoddi4 _cmpdi2 _ucmpdi2 _floatdidf _floatdisf _fixunsdfsi _fixunssfsi _fixunsdfdi _fixdfdi _fixunssfdi _fixsfdi _fixxfdi _fixunsxfdi _floatdixf _fixunsxfsi _fixtfdi _fixunstfdi _floatditf __gcc_bcmp _varargs __dummy _eprintf _bb _shtab _clear_cache _trampoline __main _exit _ctors _pure; \ do \ echo ${name}; \ /opt/uClinux/m68k-elf-gcc/gcc/xgcc -B/opt/uClinux/m68k-elf-gcc/gcc/ -B/opt/uClinux/m68k-elf/bin/ -I/opt/uClinux/m68k-elf/include -O2 -DCROSS_COMPILE -DIN_GCC -g -O2 -I./include -g1 -DIN_LIBGCC2 -D__GCC_FLOAT_NOT_NEEDED -I. -I../../gcc-2.95.3/gcc -I../../gcc-2.95.3/gcc/config -I../../gcc-2.95.3/gcc/../include -c -DL${name} \ ../../gcc-2.95.3/gcc/libgcc2.c -o ${name}.o; \ if [ $? -eq 0 ] ; then true; else exit 1; fi; \ m68k-elf-ar rc tmplibgcc2.a ${name}.o; \ rm -f ${name}.o; \ done _muldi3 In file included from /opt/uClinux/lib/gcc-lib/m68k-elf/2.95.3/../../../../m68k-elf/sys-include/stdlib.h:394, from ../../gcc-2.95.3/gcc/libgcc2.c:41: /opt/uClinux/lib/gcc-lib/m68k-elf/2.95.3/../../../../m68k-elf/sys-include/sys/types.h:30: bits/types.h: No such file or directory In file included from /opt/uClinux/lib/gcc-lib/m68k-elf/2.95.3/../../../../m68k-elf/sys-include/sys/types.h:132, from /opt/uClinux/lib/gcc-lib/m68k-elf/2.95.3/../../../../m68k-elf/sys-include/stdlib.h:394, from ../../gcc-2.95.3/gcc/libgcc2.c:41: /opt/uClinux/lib/gcc-lib/m68k-elf/2.95.3/../../../../m68k-elf/sys-include/time.h:67: bits/types.h: No such file or directory /opt/uClinux/lib/gcc-lib/m68k-elf/2.95.3/../../../../m68k-elf/sys-include/time.h:79: bits/types.h: No such file or directory /opt/uClinux/lib/gcc-lib/m68k-elf/2.95.3/../../../../m68k-elf/sys-include/time.h:91: bits/types.h: No such file or directory In file included from /opt/uClinux/lib/gcc-lib/m68k-elf/2.95.3/../../../../m68k-elf/sys-include/sys/types.h:215, from /opt/uClinux/lib/gcc-lib/m68k-elf/2.95.3/../../../../m68k-elf/sys-include/stdlib.h:394, from ../../gcc-2.95.3/gcc/libgcc2.c:41: /opt/uClinux/lib/gcc-lib/m68k-elf/2.95.3/../../../../m68k-elf/sys-include/sys/select.h:28: bits/types.h: No such file or directory /opt/uClinux/lib/gcc-lib/m68k-elf/2.95.3/../../../../m68k-elf/sys-include/sys/select.h:31: bits/select.h: No such file or directory /opt/uClinux/lib/gcc-lib/m68k-elf/2.95.3/../../../../m68k-elf/sys-include/sys/select.h:34: bits/sigset.h: No such file or directory In file included from /opt/uClinux/lib/gcc-lib/m68k-elf/2.95.3/../../../../m68k-elf/sys-include/sys/types.h:215, from /opt/uClinux/lib/gcc-lib/m68k-elf/2.95.3/../../../../m68k-elf/sys-include/stdlib.h:394, from ../../gcc-2.95.3/gcc/libgcc2.c:41: /opt/uClinux/lib/gcc-lib/m68k-elf/2.95.3/../../../../m68k-elf/sys-include/sys/select.h:45: bits/time.h: No such file or directory In file included from ../../gcc-2.95.3/gcc/libgcc2.c:42: /opt/uClinux/lib/gcc-lib/m68k-elf/2.95.3/../../../../m68k-elf/sys-include/unistd.h:175: bits/posix_opt.h: No such file or directory /opt/uClinux/lib/gcc-lib/m68k-elf/2.95.3/../../../../m68k-elf/sys-include/unistd.h:179: bits/environments.h: No such file or directory /opt/uClinux/lib/gcc-lib/m68k-elf/2.95.3/../../../../m68k-elf/sys-include/unistd.h:190: bits/types.h: No such file or directory In file included from ../../gcc-2.95.3/gcc/libgcc2.c:42: /opt/uClinux/lib/gcc-lib/m68k-elf/2.95.3/../../../../m68k-elf/sys-include/unistd.h:503: bits/confname.h: No such file or directory make[3]: *** [libgcc2.a] Error 1 make[3]: Leaving directory `/opt/uClinux/m68k-elf-gcc/gcc' make[2]: *** [stmp-multilib-sub] Error 2 make[2]: Leaving directory `/opt/uClinux/m68k-elf-gcc/gcc' make[1]: *** [stmp-multilib] Error 1 make[1]: Leaving directory `/opt/uClinux/m68k-elf-gcc/gcc' make: *** [all-gcc] Error 2 Any suggestion?? Thanks Mirco Cremasco _________________________________________________________________ MSN Photos ? il metodo pi? semplice per condividere e stampare le tue foto: http://photos.msn.it/Support/WorldWide.aspx This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From olivierp at actimage.fr Fri Mar 8 12:16:10 2002 From: olivierp at actimage.fr (Olivier Pierrat) Date: Fri, 08 Mar 2002 18:16:10 +0100 Subject: [uClinux-dev] uClinux and a RAM accessed in 16 bits. Message-ID: <3C88F1DA.50603@actimage.fr> Hi, A month ago I start to work with uClinux on a coldfire M5272C3 board, in order to see if I can use it in a final device. But the evaluation board has a 4Mbytes RAM, and the coldfire access to the RAM in 32 bits. It has a 4Mbytes flash, accessed in 16 bits. Do you think, there is big problems to run uClinux and an application, if the RAM access is in 16 bits, instead of 32 bits ? Does uClinux works fine with this configuration (RAM access in 16 bits) ? What is the influence on the accuracy of uClinux ? I make some measures of the time the kernel takes to detect an interruption ( TMR2 interruption ar 1ms). The maximum latency I notice, with an application consumming all the cpu time, is 50 micro seconds. Is there a great influence with an 16 bits RAM access ? Is it possible to make some tests on the evaluation board, with the coldfire configure in 16 bits external access ? Thanks, for your help. Olivier. This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From hamilton at SEDSystems.ca Fri Mar 8 12:35:37 2002 From: hamilton at SEDSystems.ca (Kendrick Hamilton) Date: Fri, 8 Mar 2002 11:35:37 -0600 (Canada Central Standard Time) Subject: [uClinux-dev] Re: watchdog timers In-Reply-To: Message-ID: What is the time out on the coldfire watchdog timer? TIA, Kendrick This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From justin.wojdacki at analog.com Fri Mar 8 13:57:09 2002 From: justin.wojdacki at analog.com (Justin Wojdacki) Date: Fri, 08 Mar 2002 10:57:09 -0800 Subject: [uClinux-dev] uClinux and a RAM accessed in 16 bits. References: <3C88F1DA.50603@actimage.fr> Message-ID: <3C890985.E35A387@analog.com> Olivier Pierrat wrote: > > But the evaluation board has a 4Mbytes RAM, and the coldfire access to > the RAM in 32 bits. It has a 4Mbytes flash, accessed in 16 bits. > So accesses to RAM are 32-bits wide and accesses to flash are 16-bits wide? That's not that unusual. > > Do you think, there is big problems to run uClinux and an application, > if the RAM access is in 16 bits, instead of 32 bits ? Does uClinux works > fine with this configuration (RAM access in 16 bits) ? What is the > influence on the accuracy of uClinux ? > The difference will be that 32-bit RAM accesses will take longer. Exactly how much longer depends on the memory device and the memory controller used. So you'll lose some performance, but it should still run. If you can access RAM 32-bits at a time, why change it to 16? (Or did I misunderstand something here?). > > I make some measures of the time the kernel takes to detect an > interruption ( TMR2 interruption ar 1ms). The maximum latency I notice, > with an application consumming all the cpu time, is 50 micro seconds. Is > there a great influence with an 16 bits RAM access ? > Instruction fetches will take longer. So you'll probably see an increase in latency. > > Is it possible to make some tests on the evaluation board, with the > coldfire configure in 16 bits external access ? > You'll need to identify how to reprogram the memory controller, and add that code to the initialization sequence for the controller. Sorry I can't help you more than that, I'm not familiar with the details of the Coldfire boards. -- ------------------------------------------------- Justin Wojdacki justin.wojdacki at analog.com (408) 350-5032 Communications Processors Group -- Analog Devices This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From Dave_Pfaltzgraff at patapsco.com Fri Mar 8 14:16:07 2002 From: Dave_Pfaltzgraff at patapsco.com (Dave_Pfaltzgraff at patapsco.com) Date: Fri, 08 Mar 2002 14:16:07 -0500 Subject: [uClinux-dev] Re: watchdog timers Message-ID: <85256B76.0069D8EE.00@patapsco.com> Dave Pfaltzgraff at PATAPSCO 03/08/2002 02:16 PM Kendrick Hamilton asks: > What is the time out on the coldfire watchdog timer? As with all things programmable, that depends... You need to look at the section on the "Software Watchdog Timer" in the approriate manual. It's section 6.2.8 in the "MCF5272 Coldfire Integrated Microprocessor User's Manual" Our procedure is to: During initialization set up the control register, where START_WDG is defined as 0x00000001 to enable the watchdog. (*WRRR) = 1611 | START_WDG; // Start the watchdog with 0.8 sec timeout // Each count is 0.4965 msec Then within the program address it periodically. In our case: (*WCR) = 2; // And feed the dog. This allows for // 2 timeouts or 1.6 seconds Hope this helps... Dave This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From hamilton at SEDSystems.ca Fri Mar 8 16:11:25 2002 From: hamilton at SEDSystems.ca (Kendrick Hamilton) Date: Fri, 8 Mar 2002 15:11:25 -0600 (Canada Central Standard Time) Subject: [uClinux-dev] Re: watchdog timers In-Reply-To: <85256B76.0069D8EE.00@patapsco.com> Message-ID: I was wondering about the timeout under uClinux. Thank You. On Fri, 8 Mar 2002 Dave_Pfaltzgraff at patapsco.com wrote: > > > > > Dave Pfaltzgraff at PATAPSCO > 03/08/2002 02:16 PM > > Kendrick Hamilton asks: > > What is the time out on the coldfire watchdog timer? > > As with all things programmable, that depends... You need > to look at the section on the "Software Watchdog Timer" in > the approriate manual. It's section 6.2.8 in the "MCF5272 > Coldfire Integrated Microprocessor User's Manual" > > Our procedure is to: > > During initialization set up the control register, where START_WDG > is defined as 0x00000001 to enable the watchdog. > (*WRRR) = 1611 | START_WDG; // Start the watchdog with 0.8 sec timeout > // Each count is 0.4965 msec > > Then within the program address it periodically. In our case: > (*WCR) = 2; // And feed the dog. This allows for > // 2 timeouts or 1.6 seconds > > Hope this helps... > Dave > > > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From jjef at inforamp.net Fri Mar 8 15:17:39 2002 From: jjef at inforamp.net (John Jeffers) Date: Fri, 08 Mar 2002 15:17:39 -0500 Subject: [uClinux-dev] Communications to monitor Trap Message-ID: <5.1.0.14.2.20020308151539.028ad190@209.89.75.25> Hello Is there any good reason the M68K cannot use PIC or ELF location on an application drop through to a trap passing parameters on registers and using coff in a monitor servicing the Trap. Regards John This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From superwen at 263.net Sat Mar 9 00:25:51 2002 From: superwen at 263.net (Kevin Zhu) Date: Sat, 9 Mar 2002 13:25:51 +0800 Subject: [uClinux-dev] Is there a bbug32? Message-ID: <20020309051844.8CC6E1DE8310B@smtp.263.net> Hi, All, Is there a bbugv32 or something for windows I can use to download, my bbugv+ can't run properly in Windows. Thank you. Best Regards Kevin Zhu ----------------------------- This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From superwen at 263.net Sat Mar 9 00:24:36 2002 From: superwen at 263.net (Kevin Zhu) Date: Sat, 9 Mar 2002 13:24:36 +0800 Subject: [uClinux-dev] Is there a bbug32? Message-ID: <20020309051825.28FC51DE831CB@smtp.263.net> Hi, All, Is there a bbugv32 or something for windows I can use to download, my bbugv+ can't run properly in Windows. Thank you. Best Regards Kevin Zhu ----------------------------- This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From zl at imail.com.cn Sat Mar 9 00:50:21 2002 From: zl at imail.com.cn (zhanglei) Date: Sat, 9 Mar 2002 13:50:21 +0800 Subject: [uClinux-dev] Is there a bbug32? Message-ID: <20020309055207.22EA01033D6@smtp227.email.com.cn> Try this one! http://www.apspg.com/products/dragonballvz/vz.html Hi, All, Is there a bbugv32 or something for windows I can use to download, my bbugv+ can't run properly in Windows. Thank you. Best Regards Kevin Zhu ----------------------------- This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: LittleBoy.GIF Type: image/gif Size: 1021 bytes Desc: not available URL: From superwen at 263.net Sat Mar 9 05:23:50 2002 From: superwen at 263.net (Kevin Zhu) Date: Sat, 9 Mar 2002 18:23:50 +0800 Subject: [uClinux-dev] A question about printk Message-ID: <20020309102121.5FA8E1CC42492@smtp.263.net> Hi, All, I've download the uClinux 2.4 kernel into my VZ328ADS board. Since the "HEART BEAT" LED is blinking, I think my kernel is runing. But I can't get any info output at my HyperTerminal window(Win98), is this correct? where does the printk statements write strings to? I'm new on this, so the questions I asked maybe stupid, but I really learn from it. Thank you all. Best Regards Kevin Zhu ------------------------ ? _..--"\ /`'^`) ? ? ? ( .~\ \W//^~-' ? ? \_/ ( @) ? ( =-oOo-) ? If you smell what Kevin is cooking ... ? _.`^^- "o0/ ? ??(oO\ ###/ \`)???????????????? This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From davidm at snapgear.com Sat Mar 9 07:26:31 2002 From: davidm at snapgear.com (David McCullough) Date: Sat, 9 Mar 2002 22:26:31 +1000 Subject: [uClinux-dev] problems building elf tool In-Reply-To: ; from mirco_cremasco@hotmail.com on Fri, Mar 08, 2002 at 05:38:59PM +0100 References: Message-ID: <20020309222631.A24325@beast.internal.moreton.com.au> Jivin mirco cremasco lays it down ... > >From: David McCullough > >Reply-To: uclinux-dev at uclinux.org > >To: uclinux-dev at uclinux.org > >Subject: Re: [uClinux-dev] problems building elf tool > >Date: Fri, 8 Mar 2002 09:57:11 +1000 > > > > > >Jivin mirco cremasco lays it down ... > >... > > > I've downloaded al CVS and tar.gz files as suggeted inside > > > build-uclinux-tools-sh script. > > > Everything is in one directory and I'm root > > > but the script stop with this error: > > > > > > STAGE 2 - complete > > > -------------------------------------------------------- > > > STAGE 3 - needs building > > > /bin/sh: m68k-elf-gcc: command not found > > > /bin/sh: include/bits/uClibc_config.h: No such file or directory > > > >You need to have /usr/local/bin (or PREFIX/bin if you change it) in > >your path. > > > >The next version of the script will do this for you ;-) > > > >Cheers, > >Davidm > > > >-- > >David McCullough: Ph: +61 7 3435 2815 http://www.SnapGear.com > >davidm at snapgear.com Fx: +61 7 3891 3630 825 Stanley St., W'gabba QLD > >4102, Oz > >This message resent by the uclinux-dev at uclinux.org list server > >http://www.uClinux.org/ > > > Sorry if I'm boring you. > > I've done this: > 1) I changed host, now on a Debian Linux. > 2) modified in build-uclinux-tools.sh: > -> MAKE = make (gmake unrecognized) > -> PREFIX = /opt/uClinux > 3) put /opt/uClinux/bin in $PATH > > exec ./build-uclinux-tools.sh build 2>$1 > > it stop width: > STAGE 3 - needs building > /bin/sh: include/bits/uClibc_config.h: No such file or directory > make: *** [uClibc_config] Error 1 Are you using the uClibc.tar.gz that is in the source directory with the tools or are you using the latest from CVS/snapshot off uClinux.org ? If you aren't already, use: http://www.uclinux.org/pub/uClinux/m68k-elf-tools/tools-20020218/uClibc.tar.gz > so: mkdir uClibc/include/bits and rerun build-uclinux-tools.sh width > continue command You shouldn't have to do this. It seems like the uClibc headers did not get copied into place correctly. Check in the /opt/uClinux/m68k-elf/include directory (I think, I never build the tools this way). [...] > Any suggestion?? Other than finding out what uClibc you are using, make sure you have PATH'd it correctly in the top of build-uclinux-tools.sh. Cheers, Davidm -- David McCullough: Ph: +61 7 3435 2815 http://www.SnapGear.com davidm at snapgear.com Fx: +61 7 3891 3630 825 Stanley St., W'gabba QLD 4102, Oz This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From gerg at snapgear.com Sat Mar 9 07:42:08 2002 From: gerg at snapgear.com (gerg) Date: Sat, 09 Mar 2002 22:42:08 +1000 Subject: [uClinux-dev] uClinux and a RAM accessed in 16 bits. References: <3C88F1DA.50603@actimage.fr> Message-ID: <3C8A0320.77647F82@snapgear.com> Hi Olivier, Olivier Pierrat wrote: > Do you think, there is big problems to run uClinux and an application, > if the RAM access is in 16 bits, instead of 32 bits ? No, no problems at all. It works fine. > Does uClinux works > fine with this configuration (RAM access in 16 bits) ? Yes. > What is the > influence on the accuracy of uClinux ? It still runs just as accurate :-) It is a little slower, the CPU cache reduces the penalty of having slower memory accesses. My measuements show that is of the order of 10% udner uClinux. > I make some measures of the time the kernel takes to detect an > interruption ( TMR2 interruption ar 1ms). The maximum latency I notice, > with an application consumming all the cpu time, is 50 micro seconds. Is > there a great influence with an 16 bits RAM access ? I expect it will that this latency will be longer. > Is it possible to make some tests on the evaluation board, with the > coldfire configure in 16 bits external access ? I don't know. I did it on our own 5272 board. Regards Greg ------------------------------------------------------------------------ Greg Ungerer -- Chief Software Wizard EMAIL: gerg at snapgear.com Snapgear PHONE: +61 7 3279 1822 825 Stanley St, FAX: +61 7 3279 1820 Woolloongabba, QLD, 4102, Australia WEB: www.snapgear.com This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From gerg at snapgear.com Sat Mar 9 07:46:05 2002 From: gerg at snapgear.com (gerg) Date: Sat, 09 Mar 2002 22:46:05 +1000 Subject: [uClinux-dev] Re: watchdog timers References: Message-ID: <3C8A040D.9BDCE1B1@snapgear.com> Hi Kendrick, Kendrick Hamilton wrote: > What is the time out on the coldfire watchdog timer? The mcfwatchdog.c driver sets it to the maximum value, which if IIRC is a little over 1 second (at least on a 90MHz 5307). It is trivial to set it to any valid value. Regards Greg ------------------------------------------------------------------------ Greg Ungerer -- Chief Software Wizard EMAIL: gerg at snapgear.com Snapgear PHONE: +61 7 3279 1822 825 Stanley St, FAX: +61 7 3279 1820 Woolloongabba, QLD, 4102, Australia WEB: www.snapgear.com This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From gerg at snapgear.com Sat Mar 9 07:49:15 2002 From: gerg at snapgear.com (gerg) Date: Sat, 09 Mar 2002 22:49:15 +1000 Subject: [uClinux-dev] uClinux and a RAM accessed in 16 bits. References: <3C88F1DA.50603@actimage.fr> <3C890985.E35A387@analog.com> Message-ID: <3C8A04CB.6F49D67B@snapgear.com> Hi Justin, Justin Wojdacki wrote: > If you can access RAM 32-bits at a time, why change it to 16? (Or did > I misunderstand something here?). Cost. It is usually cheaper to use a single 16bit RAM chip then 2 16bit devices. And this is a common trade off. Regards Greg ------------------------------------------------------------------------ Greg Ungerer -- Chief Software Wizard EMAIL: gerg at snapgear.com Snapgear PHONE: +61 7 3279 1822 825 Stanley St, FAX: +61 7 3279 1820 Woolloongabba, QLD, 4102, Australia WEB: www.snapgear.com This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From MGroden at hxi.com Sat Mar 9 09:40:21 2002 From: MGroden at hxi.com (Mike Groden) Date: Sat, 9 Mar 2002 09:40:21 -0500 Subject: [uClinux-dev] uClinux and a RAM accessed in 16 bits. Message-ID: <879EA880A0FED411996B00B0D0B082EC10080D@hxi_exch01.hxi.com> And maybe size. Also the PCB will be a little less complex with a 16-bit bus. Mike -----Original Message----- From: gerg [mailto:gerg at snapgear.com] Sent: Saturday, March 09, 2002 7:49 AM To: uclinux-dev at uclinux.org Subject: Re: [uClinux-dev] uClinux and a RAM accessed in 16 bits. Hi Justin, Justin Wojdacki wrote: > If you can access RAM 32-bits at a time, why change it to 16? (Or did > I misunderstand something here?). Cost. It is usually cheaper to use a single 16bit RAM chip then 2 16bit devices. And this is a common trade off. Regards Greg ------------------------------------------------------------------------ Greg Ungerer -- Chief Software Wizard EMAIL: gerg at snapgear.com Snapgear PHONE: +61 7 3279 1822 825 Stanley St, FAX: +61 7 3279 1820 Woolloongabba, QLD, 4102, Australia WEB: www.snapgear.com This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From hamilton at SEDSystems.ca Sat Mar 9 15:11:12 2002 From: hamilton at SEDSystems.ca (Kendrick Hamilton) Date: Sat, 9 Mar 2002 14:11:12 -0600 (CST) Subject: [uClinux-dev] Re: watchdog timers In-Reply-To: <3C8A040D.9BDCE1B1@snapgear.com> Message-ID: I was curious because it is timing out for people. I am also using a watchdog and don't want it timing out unless something is not working properly. My watchdog is set to 16 seconds (the next step lower was 4 seconds and it is not long enough for auto-negotiate). Kendrick Hamilton E.I.T. SED Systems, a division of Calian Ltd. 18 Innovation Blvd. PO Box 1464 Saskatoon, Saskatchewan Canada S7N 3R1 Hamilton at sedsystems.ca Tel: (306) 933-1453 Fax: (306) 933-1486 On Sat, 9 Mar 2002, gerg wrote: > > Hi Kendrick, > > Kendrick Hamilton wrote: > > What is the time out on the coldfire watchdog timer? > > The mcfwatchdog.c driver sets it to the maximum value, > which if IIRC is a little over 1 second (at least on a > 90MHz 5307). It is trivial to set it to any valid value. > > Regards > Greg > > > ------------------------------------------------------------------------ > Greg Ungerer -- Chief Software Wizard EMAIL: gerg at snapgear.com > Snapgear PHONE: +61 7 3279 1822 > 825 Stanley St, FAX: +61 7 3279 1820 > Woolloongabba, QLD, 4102, Australia WEB: www.snapgear.com > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From andersen at codepoet.org Sat Mar 9 15:42:46 2002 From: andersen at codepoet.org (Erik Andersen) Date: Sat, 9 Mar 2002 13:42:46 -0700 Subject: [uClinux-dev] problems building elf tool In-Reply-To: <20020309222631.A24325@beast.internal.moreton.com.au> References: <20020309222631.A24325@beast.internal.moreton.com.au> Message-ID: <20020309204246.GA5303@codepoet.org> On Sat Mar 09, 2002 at 10:26:31PM +1000, David McCullough wrote: > > so: mkdir uClibc/include/bits and rerun build-uclinux-tools.sh width > > continue command > > > You shouldn't have to do this. It seems like the uClibc headers did not > get copied into place correctly. > > Check in the /opt/uClinux/m68k-elf/include directory (I think, I never > build the tools this way). I suspect he is using the latest and greatest CVS version of uClibc. This past week I did a major update of all the include/bits directories (eliminating all the redundant files by grouping them together). That probably is causing the toolchin build some grief (the part where it hand-copies all the files) since everything in include/bits is now a relative symlink, so a 'cp -a' now does Bad Things(tm), -Erik -- Erik B. Andersen http://codepoet-consulting.com/ --This message was written using 73% post-consumer electrons-- This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From tzengym at ms8.hinet.net Sat Mar 9 19:48:49 2002 From: tzengym at ms8.hinet.net (Tonny Tzeng) Date: Sun, 10 Mar 2002 08:48:49 +0800 Subject: [uClinux-dev] aplio toolchain References: <02022812264802.04902@shiaz> <02022821095000.23146@shiaz> <003001c1c0c0$06e7e880$1700a8c0@wiscore.com> <02030510092500.00976@shiaz> Message-ID: <002f01c1c7cd$753debe0$80e7a8c0@quantatw> You should use uClibc or other C libraries which implements the stdlib in terms of Linux system calls. As I know, the newlibc can be compiled to implement stdlib functions by using Angel or Daemon calling convensions, neither can be work in Linux. You could link the sample programs with uClibc for a trial if you have installed the uClibc. Somthing like the following: arm-elf-gcc -c hello.c arm-elf-ld -L /lib/crt0.o hello.o -lc -lgcc -lc Best regards, Tonny > I've downloaded the gcc-3.0.1 and binutils and newlibc. > I've built the toolchain and tried compiling some sample programs > trying some system calls, but to my surprise when using vfork this error it > gives. > > /tmp/ccRU5l0N.o: In function `main': > /tmp/ccRU5l0N.o(.text+0x38): undefined reference to `vfork' > collect2: ld returned 1 exit status > > Please comment on this. > > Thankz in advance.. > regards > shiaz > > > On Friday 01 March 2002 06:54 am, you wrote: > > Can you compile the "Hello world" sample program by typing the > > following command? > > arm-elf-gcc -o hello hello.c > > If the C runtime has been installed properly, the ld should be invoked > > by the gcc to create the final executable. Otherwise, you have to link > > the object files with runtime explicitly. As I know, the "C compiler > > cannot create executables." error was caused by the above test > > command. > > > > Best regards, > > Tonny > > > > > I've given make install in gcc dir... > > > In my install directory crt?.o is installed . After this i tried to make > > > > gcc > > > > > in build dir it's givinme this error.. > > > > > > On Thursday 28 February 2002 01:39 pm, you wrote: > > > > The installation of GCC will try to build a dummy executable by using > > > > $(CC) directly, but it is impossible until the crt?.o have been > > > > installed. > > > > > > So you could jump into the gcc directory and 'make install', then > > > > compile > > > > > > and install the C library, then re-compile the gcc toolchain should > > > > pass the error. > > > > > > > > Best regards, > > > > Tonny > > > > > > > > > Can anyone tell me the cause of the error while building the > > > > > toolchain for arm uclinux using the aplio Armtools(got from > > > > > ftp://ftp.wiscore.com/ARMTools/Aplio/arm-tools-src.tgz.) > > > > > Configuring in arm-elf/libf2c > > > > > loading cache ./config.cache > > > > > checking if compiler f771 has been built... yes > > > > > checking for gcc... /home/shiaz/Aplio/apliotools/build-egcs/gcc/xgcc > > > > > -B/home/shiaz/Aplio/apliotools/build-egcs/gcc/ > > > > > -B/home/shiaz/Aplio/apliotools/build-egcs/arm-elf/newlib/ > > > > > -B/home/shiaz/Aplio/apliotools/instdir/arm-elf/bin/ -idirafter > > > > > /home/shiaz/Aplio/apliotools/build-egcs/arm-elf/newlib/targ-include > > > > > -idirafter /home/shiaz/Aplio/apliotools/egcs/newlib/libc/include > > > > > -nostdinc checking whether the C compiler > > > > > (/home/shiaz/Aplio/apliotools/build-egcs/gcc/xgcc > > > > > -B/home/shiaz/Aplio/apliotools/build-egcs/gcc/ > > > > > -B/home/shiaz/Aplio/apliotools/build-egcs/arm-elf/newlib/ > > > > > -B/home/shiaz/Aplio/apliotools/instdir/arm-elf/bin/ -idirafter > > > > > /home/shiaz/Aplio/apliotools/build-egcs/arm-elf/newlib/targ-include > > > > > -idirafter > > > > > > > > /home/shiaz/Aplio/apliotools/egcs/newlib/libc/include -nostdinc -g > > > > > > > > > -O2 ) works... no > > > > > > > > > > configure: error: installation or configuration problem: C compiler > > > > > cannot create executables. > > > > > > > > This message resent by the uclinux-dev at uclinux.org list server > > > > http://www.uClinux.org/ > > > > > > This message resent by the uclinux-dev at uclinux.org list server > > > > http://www.uClinux.org/ > > > > > > This message resent by the uclinux-dev at uclinux.org list server > > http://www.uClinux.org/ > > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From davidm at snapgear.com Sun Mar 10 07:59:38 2002 From: davidm at snapgear.com (David McCullough) Date: Sun, 10 Mar 2002 22:59:38 +1000 Subject: [uClinux-dev] new version of xcopilot for uClinux Message-ID: <20020310225937.A2842@beast.internal.moreton.com.au> Hi all, For anyone wanting to play with xcopilot/uClinux I have put a new version up on uClinux.org: http://www.uclinux.org/pub/uClinux/utilities/xcopilot-0.6.6-uc0.tar.gz It is also in CVS. Check the README.uClinux for details on building and running using the uClinux distribution. Many thanks to Mark McChrystal who has helped a lot with getting this done, Cheers, Davidm -- David McCullough: Ph: +61 7 3435 2815 http://www.SnapGear.com davidm at snapgear.com Fx: +61 7 3891 3630 825 Stanley St., W'gabba QLD 4102, Oz This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From carlos at embedded.cl Sun Mar 10 21:18:08 2002 From: carlos at embedded.cl (Carlos Manuel Duclos Vergara) Date: Sun, 10 Mar 2002 22:18:08 -0400 Subject: [uClinux-dev] About framebuffer for LCD on 2.4 kernel In-Reply-To: <20020307025021.711.qmail@web13208.mail.yahoo.com> References: <20020307025021.711.qmail@web13208.mail.yahoo.com> Message-ID: <20020310221808.6e7563e0.carlos@embedded.cl> > hi: hi! > I want to write something to my LCD by framebuffer.I > has modified /arch/m68knmmu/console/68328fb.c for my > 320x240 LCD.And I also wrote a application to test > it.My test program is like this: first, can you see the boot process????? (if you compile this and everything is correctly plugged in you _should_ see something in your lcd before you run your program) second, i'm not sure about the framebuffer support in uclinux. I've been (well, last week i did nothing because i took a little vacations on beach) working with that and i found some little problems... and i still doesn't fix them. I'm working with a 5206e Coldfire not with a dragonball, but i think that they must be close enough. Anyway, first try to see the bott process(uClinux 2.4......... booting....... blablabla) bye -- "Solo me arrepiento de unos * de menos y unos ++ de sobra" Carlos Manuel Duclos Vergara This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From gmenie at akamai.com Mon Mar 11 04:17:53 2002 From: gmenie at akamai.com (Menie, Georges) Date: Mon, 11 Mar 2002 04:17:53 -0500 Subject: [uClinux-dev] Debugging on the 68VZ328 using 2.4.10 and elf Message-ID: I made some significant progress to allow on target debugging with gdb/gdbserver. The trace traps were missing in the 68328, 68EZ328 and 68VZ328 architectures. The problem also was that the 68000 core does not put the trap vector on the stack so we need to simulate this. Once the traps are added, I did recompile the kernel with the Full source debugging option set (it's not clear to me if this is needed to debug user apps...). Then gdb/gdbserver almost works ! I can set breakpoints and run a couple of s/n commands and then the debugged program receive a SIGILL signal. Of course the program under debug is working just fine, so I know something is still missing. Any hints would be much appreciated... Regards, Georges > -----Original Message----- > From: Daniel Haensse [mailto:daniel.haensse at fhk.usz.ch] > Sent: mardi 19 f?vrier 2002 12:40 > To: uclinux-dev at uclinux.org > Subject: Re: [uClinux-dev] Debugging on the 68VZ328 using > 2.4.10 and elf > > > Is there a description about the mechanism that is needed for > the debugger? > Maybe somebody of the coldfire guys knows about it. I assume > that it will be > similair for the vz. Can anybody point the way? > > Dani > > Am Dienstag, 19. Februar 2002 11.36 schrieben Sie: > > I did give it a quick try a couple of weeks ago when porting uClinux > > on a new 68VZ328 board, but had no success. gdbserver is running > > but may need some traps which I did not implement. It is something > > I want to fix too... > > > > Regards, > > Georges > > > > > -----Original Message----- > > > From: Daniel Haensse [mailto:daniel.haensse at fhk.usz.ch] > > > Sent: lundi 18 f?vrier 2002 15:35 > > > To: uclinux-dev at uclinux.org > > > Subject: [uClinux-dev] Debugging on the 68VZ328 using > 2.4.10 and elf > > > > > > > > > Hi folks, > > > > > > is anybody using GDB on the 68VZ328? Is any patch needed or > > > is it just > > > working? > > > > > > Thanks > > > > > > Dani > > > This message resent by the uclinux-dev at uclinux.org list > > > server http://www.uClinux.org/ > > > > This message resent by the uclinux-dev at uclinux.org list server > > http://www.uClinux.org/ > This message resent by the uclinux-dev at uclinux.org list > server http://www.uClinux.org/ > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From sujz at cad.csie.ncku.edu.tw Mon Mar 11 04:31:31 2002 From: sujz at cad.csie.ncku.edu.tw (SuCC) Date: Mon, 11 Mar 2002 17:31:31 +0800 Subject: [uClinux-dev] About page allocate Message-ID: <001101c1c8df$8e83a960$1af7748c@SCC> Hi, all I run pptp client on uclinux-dist on ARM9, and first, it work very well!! But when I ping the other side(ex. 192.168.1.1) and take a while, system will crash, and dump these messages kernel Bug at page-alloc.c : 86 kernel Bug at page-alloc.c : 88 kernel Bug at page-alloc.c : 98 kernel Bug at page-alloc.c : 100 kernel Bug at page-alloc.c : 225 (not ordered in real) And, when I run "ps" many times. The system crash, too!! and another message is BINFMT_FLAT: reloc outside program ........... thks ------------------------------------------------------ Sincerely, Chien-Chung Su Intelligent System/Media Processing Lab. Department of CSIE, NCKU http://ismp.csie.ncku.edu.tw -------------- next part -------------- An HTML attachment was scrubbed... URL: From Dave_Pfaltzgraff at patapsco.com Mon Mar 11 08:02:42 2002 From: Dave_Pfaltzgraff at patapsco.com (Dave_Pfaltzgraff at patapsco.com) Date: Mon, 11 Mar 2002 08:02:42 -0500 Subject: [uClinux-dev] Re: watchdog timers Message-ID: <85256B79.0047A957.00@patapsco.com> Dave Pfaltzgraff at PATAPSCO 03/11/2002 08:02 AM Kendrick Hamilton said: > I was wondering about the timeout under uClinux. Thank You. As far as I know, there is none. This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From davidm at snapgear.com Mon Mar 11 08:24:34 2002 From: davidm at snapgear.com (David McCullough) Date: Mon, 11 Mar 2002 23:24:34 +1000 Subject: [uClinux-dev] Debugging on the 68VZ328 using 2.4.10 and elf In-Reply-To: ; from gmenie@akamai.com on Mon, Mar 11, 2002 at 04:17:53AM -0500 References: Message-ID: <20020311232434.C10577@beast.internal.moreton.com.au> Jivin Menie, Georges lays it down ... > I made some significant progress to allow on target debugging > with gdb/gdbserver. The trace traps were missing in the 68328, > 68EZ328 and 68VZ328 architectures. The problem also was > that the 68000 core does not put the trap vector on the stack > so we need to simulate this. Once the traps are added, > I did recompile the kernel with the Full source debugging > option set (it's not clear to me if this is needed to debug > user apps...). Then gdb/gdbserver almost works ! Only needed to debug the kernel. If you are using the uClinux-distro, you can build the libs/apps with debug info under the "vendor/apps" config menu in the Debug section. > I can set breakpoints and run a couple of s/n commands > and then the debugged program receive a SIGILL signal. > Of course the program under debug is working just fine, > so I know something is still missing. > > Any hints would be much appreciated... Does it break on the breakpoints ? It could be that it is writing the wrong code back or writing to not quite the right place ? Set the breakpoint then diasassemble/dump the location to see what it was changed to. If you know where the SIGILL appears, check the code that is there with what is supposed to be there, it might offer some clues. Be careful when debugging a program that is execute in place. Firstly, if it is in ROM you cannot write breakpoints back to it, if its in ram, a breakpoint will affect all instances of the running program and not just the one you are debugging. Probably best to add "FLTFLAGS += -r" to the makefile for the app you are debugging if any of the above is true. Cheers, Davidm -- David McCullough: Ph: +61 7 3435 2815 http://www.SnapGear.com davidm at snapgear.com Fx: +61 7 3891 3630 825 Stanley St., W'gabba QLD 4102, Oz This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From davidm at snapgear.com Mon Mar 11 08:28:44 2002 From: davidm at snapgear.com (David McCullough) Date: Mon, 11 Mar 2002 23:28:44 +1000 Subject: [uClinux-dev] About page allocate In-Reply-To: <001101c1c8df$8e83a960$1af7748c@SCC>; from sujz@cad.csie.ncku.edu.tw on Mon, Mar 11, 2002 at 05:31:31PM +0800 References: <001101c1c8df$8e83a960$1af7748c@SCC> Message-ID: <20020311232844.D10577@beast.internal.moreton.com.au> Jivin SuCC lays it down ... > Hi, all > I run pptp client on uclinux-dist on ARM9, and first, it work very well!! > But when I ping the other side(ex. 192.168.1.1) and take a while, > system will crash, and dump these messages > > kernel Bug at page-alloc.c : 86 > kernel Bug at page-alloc.c : 88 > kernel Bug at page-alloc.c : 98 > kernel Bug at page-alloc.c : 100 > kernel Bug at page-alloc.c : 225 (not ordered in real) > > And, when I run "ps" many times. The system crash, too!! > and another message is > BINFMT_FLAT: reloc outside program ........... Sounds like you are getting memory corruption. Double check that all the correct memory is being reserved at startup (including your filesystem). If that is ok then look carefully at the apps you are using and the areas of memory being affected. You may be able to spot strings or patterns in the corruption that will help you find the cause. Turn on DEBUG in binfmt_flat.c so you can see the memory being allocated to applications, that way you will see if they get located somewhere bad. Also check the BUG's above in mmnommu/page_alloc.c lines 86/88/98 ... to see why it is complaining. Cheers, Davidm -- David McCullough: Ph: +61 7 3435 2815 http://www.SnapGear.com davidm at snapgear.com Fx: +61 7 3891 3630 825 Stanley St., W'gabba QLD 4102, Oz This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From Dave_Pfaltzgraff at patapsco.com Mon Mar 11 08:20:19 2002 From: Dave_Pfaltzgraff at patapsco.com (Dave_Pfaltzgraff at patapsco.com) Date: Mon, 11 Mar 2002 08:20:19 -0500 Subject: [uClinux-dev] Re: watchdog timers Message-ID: <85256B79.004945E6.00@patapsco.com> Dave Pfaltzgraff at PATAPSCO 03/11/2002 08:20 AM Oops, I spoke too soon when I responded to Kendrick Hamilton. Thanks, Greg for the following input: > The mcfwatchdog.c driver sets it to the maximum value, > which if IIRC is a little over 1 second (at least on a > 90MHz 5307). It is trivial to set it to any valid value. After looking at the source code, I was wondering if you'd be willing to share a snippet of code showing its intended use? I was not aware of this module and it may have saved me some time. Thanks, Dave This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From theresa.beizer at aei.com Mon Mar 11 08:32:51 2002 From: theresa.beizer at aei.com (Beizer, Theresa) Date: Mon, 11 Mar 2002 08:32:51 -0500 Subject: [uClinux-dev] Comparison between eCos and uClinux Message-ID: <5CD700D456B5D511B0C50000D11AD2321B78A2@MDMSERVER> Has anyone yet written a comparison of the two? -- Theresa Beizer, Sr. Staff Engineer AE Voorhees Division VOX: (856) 627-6100 x2422 FAX: (856) 627-1311 mailto:theresa.beizer at aei.com _______________________________________________________ This message, including any attachments, may contain information that is confidential and proprietary information of Advanced Energy Industries, Inc. The dissemination, distribution, use or copying of this message or any of its attachments is strictly prohibited without the express written consent of Advanced Energy Industries, Inc. This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From stefan.jonsson at faltcom.se Mon Mar 11 10:55:53 2002 From: stefan.jonsson at faltcom.se (Stefan Jonsson) Date: Mon, 11 Mar 2002 16:55:53 +0100 Subject: [uClinux-dev] building a "userland" for ARM Message-ID: Hello list, I have both the Lineo kit with the official CD and EB40LC with uCnet and a plain EB40 with memory module AT91MEC. I want to build a working environment for my other board without using the CD. Patching the kernel seems to work fine but then the problems starts. I cannot find "the proper way" of doing this. All readme:s seems to be for 68k. 1) Is there an explicit way of doing this for a EB40 standard board and is it described somewhere. (because it cannot be the same files for ARM as for 68k, can it?) 2) Is it possible to use "the official CD installation of "useland" on the other card (with different hardware) and in that case how can I modify things to fit for that purpose? Regards, Stefan Jonsson, Student, University of Umea, Sweden. This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From justin.wojdacki at analog.com Mon Mar 11 13:03:28 2002 From: justin.wojdacki at analog.com (Justin Wojdacki) Date: Mon, 11 Mar 2002 10:03:28 -0800 Subject: [uClinux-dev] uClinux and a RAM accessed in 16 bits. References: <3C88F1DA.50603@actimage.fr> <3C890985.E35A387@analog.com> <3C8A04CB.6F49D67B@snapgear.com> Message-ID: <3C8CF170.5B4C7A4B@analog.com> gerg wrote: > > Hi Justin, > > Justin Wojdacki wrote: > > If you can access RAM 32-bits at a time, why change it to 16? (Or did > > I misunderstand something here?). > > Cost. It is usually cheaper to use a single 16bit RAM chip then 2 > 16bit devices. And this is a common trade off. > Well aware of this, if you're designing a new board. I got the impression from Olivier's message that there already was a 32-bit bus on the board in question. -- ------------------------------------------------- Justin Wojdacki justin.wojdacki at analog.com (408) 350-5032 Communications Processors Group -- Analog Devices This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From rwehrli at azpower.com Mon Mar 11 14:36:10 2002 From: rwehrli at azpower.com (Rob Wehrli) Date: Mon, 11 Mar 2002 12:36:10 -0700 Subject: [uClinux-dev] Comparison between eCos and uClinux References: <5CD700D456B5D511B0C50000D11AD2321B78A2@MDMSERVER> Message-ID: <3C8D072A.A52FD5C@azpower.com> "Beizer, Theresa" wrote: > > Has anyone yet written a comparison of the two? You might want to take a look at: http://www.linuxdevices.com/articles/AT3393503683.html > > -- > Theresa Beizer, Sr. Staff Engineer > AE Voorhees Division > VOX: (856) 627-6100 x2422 FAX: (856) 627-1311 > mailto:theresa.beizer at aei.com > Take Care. Rob! This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From rwehrli at azpower.com Mon Mar 11 14:57:30 2002 From: rwehrli at azpower.com (Rob Wehrli) Date: Mon, 11 Mar 2002 12:57:30 -0700 Subject: [uClinux-dev] Showing uClinux port at ESC SF... Message-ID: <3C8D0C2A.612DAE07@azpower.com> BTW...uClinux developers..I'd just like to invite anyone in the area to come down to Moscone Center downtown SF and take a look at the Hitachi H8S-uClinux port 2148-based demo platform... I'll have some free "uClinux penguin" t-shirts to give-away and some cool little stuffed "embedded" penguins, too :) See: http://www.linuxdevices.com/news/NS2254127984.html ...for the announcement. Take Care. Rob! This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From Fabrice_Gautier at sdesigns.com Mon Mar 11 15:27:26 2002 From: Fabrice_Gautier at sdesigns.com (Fabrice Gautier) Date: Mon, 11 Mar 2002 12:27:26 -0800 Subject: [uClinux-dev] Comparison between eCos and uClinux Message-ID: indeed: http://www.linuxdevices.com/articles/AT3393503683.html -- Fabrice Gautier, Fabrice_Gautier at sdesigns.com > -----Original Message----- > From: Beizer, Theresa [mailto:theresa.beizer at aei.com] > Sent: Monday, March 11, 2002 5:33 AM > To: uclinux-dev at uclinux.org > Subject: [uClinux-dev] Comparison between eCos and uClinux > > > Has anyone yet written a comparison of the two? > > -- > Theresa Beizer, Sr. Staff Engineer > AE Voorhees Division > VOX: (856) 627-6100 x2422 FAX: (856) 627-1311 > mailto:theresa.beizer at aei.com > > > > _______________________________________________________ > > This message, including any attachments, may contain > information that is > confidential and proprietary information of Advanced Energy > Industries, Inc. > The dissemination, distribution, use or copying of this > message or any of > its attachments is strictly prohibited without the express > written consent > of Advanced Energy Industries, Inc. > This message resent by the uclinux-dev at uclinux.org list > server http://www.uClinux.org/ > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From danielp at cse.unsw.edu.au Mon Mar 11 18:16:16 2002 From: danielp at cse.unsw.edu.au (Daniel Potts) Date: Tue, 12 Mar 2002 10:16:16 +1100 (EST) Subject: [uClinux-dev] patch - MC68328.h & MC68VZ328.h ICR definition Message-ID: ICR definition was 0xffff302 instead of 0xfffff302 (missing an f). It appears to be correct in MC68EZ328.h. Cheers, Daniel -------------- next part -------------- Common subdirectories: uClinux-2.4.x.orig/include/asm-m68knommu/CVS and linux-2.4.x/include/asm-m68knommu/CVS diff -Nau uClinux-2.4.x.orig/include/asm-m68knommu/MC68328.h linux-2.4.x/include/asm-m68knommu/MC68328.h --- uClinux-2.4.x.orig/include/asm-m68knommu/MC68328.h Tue Mar 12 10:04:47 2002 +++ linux-2.4.x/include/asm-m68knommu/MC68328.h Tue Mar 12 10:09:06 2002 @@ -240,7 +240,7 @@ /* * Interrupt control Register */ -#define ICR_ADRR 0xffff302 +#define ICR_ADRR 0xfffff302 #define ICR WORD_REF(ICR_ADDR) #define ICR_ET6 0x0100 /* Edge Trigger Select for IRQ6 */ diff -Nau uClinux-2.4.x.orig/include/asm-m68knommu/MC68VZ328.h linux-2.4.x/include/asm-m68knommu/MC68VZ328.h --- uClinux-2.4.x.orig/include/asm-m68knommu/MC68VZ328.h Tue Mar 12 10:04:47 2002 +++ linux-2.4.x/include/asm-m68knommu/MC68VZ328.h Tue Mar 12 10:09:24 2002 @@ -205,7 +205,7 @@ /* * Interrupt control Register */ -#define ICR_ADDR 0xffff302 +#define ICR_ADDR 0xfffff302 #define ICR WORD_REF(ICR_ADDR) #define ICR_POL5 0x0080 /* Polarity Control for IRQ5 */ From gerg at snapgear.com Mon Mar 11 19:02:39 2002 From: gerg at snapgear.com (Greg Ungerer) Date: Tue, 12 Mar 2002 10:02:39 +1000 Subject: [uClinux-dev] Re: watchdog timers References: <85256B79.004945E6.00@patapsco.com> Message-ID: <3C8D459F.3020706@snapgear.com> Hi Dave, Dave_Pfaltzgraff at patapsco.com wrote: > Oops, I spoke too soon when I responded to Kendrick > Hamilton. :-) > Thanks, Greg for the following input: > > >>The mcfwatchdog.c driver sets it to the maximum value, >>which if IIRC is a little over 1 second (at least on a >>90MHz 5307). It is trivial to set it to any valid value. >> > > After looking at the source code, I was wondering if > you'd be willing to share a snippet of code showing its > intended use? I was not aware of this module and it > may have saved me some time. I use the driver "as is", no supporting user apps. I mostly use it for debugging. So when I have a problem with that "hangs" the system I enable to watchdog. 9 times out of 10 it is an access to an invalid memory region, and this is trapped by the watchdog. I have coded the watchdog trap routine in this driver to dump regs, code, stack, etc, and then send a kill signal to the current user process. More often than not the system can keep running after this! Regards Greg ------------------------------------------------------------------------ Greg Ungerer -- Chief Software Wizard EMAIL: gerg at snapgear.com SnapGear PHONE: +61 7 3435 2888 825 Stanley St, FAX: +61 7 3891 3630 Woolloongabba, QLD, 4102, Australia WEB: www.snapgear.com This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From stefan.jonsson at faltcom.se Tue Mar 12 02:40:33 2002 From: stefan.jonsson at faltcom.se (Stefan Jonsson) Date: Tue, 12 Mar 2002 08:40:33 +0100 Subject: SV: [uClinux-dev] building a "userland" for ARM In-Reply-To: Message-ID: OOPS ! EB40LS of course ... "slip of the finger" :-) / Stefan -----Ursprungligt meddelande----- Fr?n: owner-uclinux-dev at uclinux.org [mailto:owner-uclinux-dev at uclinux.org]F?r Stefan Jonsson Skickat: den 11 mars 2002 16:56 Till: uClinux-dev mailing list ?mne: [uClinux-dev] building a "userland" for ARM Hello list, I have both the Lineo kit with the official CD and EB40LC with uCnet and a plain EB40 with memory module AT91MEC. I want to build a working environment for my other board without using the CD. Patching the kernel seems to work fine but then the problems starts. I cannot find "the proper way" of doing this. All readme:s seems to be for 68k. 1) Is there an explicit way of doing this for a EB40 standard board and is it described somewhere. (because it cannot be the same files for ARM as for 68k, can it?) 2) Is it possible to use "the official CD installation of "useland" on the other card (with different hardware) and in that case how can I modify things to fit for that purpose? Regards, Stefan Jonsson, Student, University of Umea, Sweden. This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From joe_buczek at yahoo.com Tue Mar 12 03:22:03 2002 From: joe_buczek at yahoo.com (Joe Buczek) Date: Tue, 12 Mar 2002 00:22:03 -0800 (PST) Subject: [uClinux-dev] Comparison between eCos and uClinux In-Reply-To: Message-ID: <20020312082203.92049.qmail@web20306.mail.yahoo.com> Nice article, but it failed to mention one design aspect of eCos that I found disturbing: they use a patched GCC to allow them to utilize C++ exception handling in the eCos core. The designers seemed careful enough to use only the exception handling of C++, but still, is a lot of risk and extra maintenence to get this feature. It didn't seem justified to me. The rest of eCos seemed pretty reasonable, but the configuration tool was buggy when I used it a year ago: if you changed a setting, you had better save the config, quit the tool and restart it rather than saving and building. Other than these two complaints, which weren't mentioned in the article mentioned below, I found eCos to be a very nicely thought out embedded OS. And, certainly, if minimalism is key, eCos is probably a better place to start looking. JMHO, Cheers, --Joe --- Fabrice Gautier wrote: > indeed: > > http://www.linuxdevices.com/articles/AT3393503683.html > > -- > Fabrice Gautier, > Fabrice_Gautier at sdesigns.com > > > > -----Original Message----- > > From: Beizer, Theresa [mailto:theresa.beizer at aei.com] > > Sent: Monday, March 11, 2002 5:33 AM > > To: uclinux-dev at uclinux.org > > Subject: [uClinux-dev] Comparison between eCos and uClinux > > > > > > Has anyone yet written a comparison of the two? > > > > -- > > Theresa Beizer, Sr. Staff Engineer > > AE Voorhees Division > > VOX: (856) 627-6100 x2422 FAX: (856) 627-1311 > > mailto:theresa.beizer at aei.com > > > > > > > > _______________________________________________________ > > > > This message, including any attachments, may contain > > information that is > > confidential and proprietary information of Advanced Energy > > Industries, Inc. > > The dissemination, distribution, use or copying of this > > message or any of > > its attachments is strictly prohibited without the express > > written consent > > of Advanced Energy Industries, Inc. > > This message resent by the uclinux-dev at uclinux.org list > > server http://www.uClinux.org/ > > > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ __________________________________________________ Do You Yahoo!? Try FREE Yahoo! Mail - the world's greatest free email! http://mail.yahoo.com/ This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From jody.pearson at cern.ch Tue Mar 12 04:13:45 2002 From: jody.pearson at cern.ch (Jody Pearson) Date: Tue, 12 Mar 2002 10:13:45 +0100 Subject: [uClinux-dev] Problems with running init ? References: <20020306214508.72425.qmail@web14309.mail.yahoo.com> <3C877F51.4772ECA3@cern.ch> <20020308095223.A30299@beast.internal.moreton.com.au> Message-ID: <3C8DC6C8.44AA80A6@cern.ch> Hi David, I got the latest source tar from uclinux.org. Installed m68k-elf toolchain, you're right ! This is a lot easier. How much success have you had with the userland ? I seem to be getting a LOT of trouble with microwindows, and bash. Also can't seem to get framebuffer support under a 2.4.x kernel. Have you sorted that one ? Thanks Jody David McCullough wrote: > Jivin Jody Pearson lays it down ... > > Hi all, > > > > Well, I'm getting closer (teasingly closer) to getting to a shell on xcopilot > > with my compiled rom.. > > The uClinux distribution has an xcopilot target which will build and run in > xcopilot if you want to save yourself some pain ;-) > > You will need the m68k-elf-tools though, > > Cheers, > Davidm > > -- > David McCullough: Ph: +61 7 3435 2815 http://www.SnapGear.com > davidm at snapgear.com Fx: +61 7 3891 3630 825 Stanley St., W'gabba QLD 4102, Oz > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From gmenie at akamai.com Tue Mar 12 04:36:14 2002 From: gmenie at akamai.com (Menie, Georges) Date: Tue, 12 Mar 2002 04:36:14 -0500 Subject: [uClinux-dev] Debugging on the 68VZ328 using 2.4.10 and elf Message-ID: > -----Original Message----- > From: David McCullough [mailto:davidm at snapgear.com] > Sent: lundi 11 mars 2002 14:25 > To: uclinux-dev at uclinux.org > Subject: Re: [uClinux-dev] Debugging on the 68VZ328 using > 2.4.10 and elf > > Jivin Menie, Georges lays it down ... > > I made some significant progress to allow on target debugging > > with gdb/gdbserver. The trace traps were missing in the 68328, > > 68EZ328 and 68VZ328 architectures. The problem also was > > that the 68000 core does not put the trap vector on the stack > > so we need to simulate this. Once the traps are added, > > I did recompile the kernel with the Full source debugging > > option set (it's not clear to me if this is needed to debug > > user apps...). Then gdb/gdbserver almost works ! > > Only needed to debug the kernel. If you are using the uClinux-distro, > you can build the libs/apps with debug info under the "vendor/apps" > config menu in the Debug section. ok > > I can set breakpoints and run a couple of s/n commands > > and then the debugged program receive a SIGILL signal. > > Of course the program under debug is working just fine, > > so I know something is still missing. > > > > Any hints would be much appreciated... > > Does it break on the breakpoints ? It could be that it is writing > the wrong code back or writing to not quite the right place ? > > Set the breakpoint then diasassemble/dump the location to see > what it was > changed to. If you know where the SIGILL appears, check the > code that is > there with what is supposed to be there, it might offer some clues. > > Be careful when debugging a program that is execute in place. > Firstly, if > it is in ROM you cannot write breakpoints back to it, if its > in ram, a > breakpoint will affect all instances of the running program > and not just > the one you are debugging. Probably best to add "FLTFLAGS += > -r" to the > makefile for the app you are debugging if any of the above is true. Thanks for your inputs, I am debuging in ram, with only one instance of the prog running. I did not check yet where the bp was set, but it was because if I don't use bp, I can run the debugged prog completely without error from gdb. If I set a bp, I sometimes got a SIGILL after the break so I assume you are right : the bp is not set correctly. Also, the print function of gdb was not working right, as if the stack were corrupted. I think I did half the job and I need to better simulate the format/vector info on the stack. What I did was: add a entry for most of the traps and exception: ENTRY(trap1) SAVE_ALL_INT GET_CURRENT(%d0) movel %sp,%sp at - /* stack frame pointer argument*/ movel #132,%sp at - /* pseudo-vector number */ bsrw SYMBOL_NAME(trap_cv) addql #8,%sp jra SYMBOL_NAME(ret_from_exception) adding the pseudo vector as the first parameter to trap_cv the trap_cv is a wrapping function to trap_c: asmlinkage void trap_cv(int vector, struct frame *fp) { asmlinkage void trap_c(struct frame *fp); struct frame fakefp; memcpy(&fakefp, fp, sizeof fakefp); fakefp.ptregs.format = 0; fakefp.ptregs.vector = vector; trap_c(&fakefp); } I did this because I though that it would be enough, the trap_c function 'consume' the format/vector info and do not pass it to other functions. All the exception/trap have a modified entry, but not the system_call and buserr exceptions and I think the problem is there : system_call is still missing the pseudo format/vector word and this why the stack appears to be corrupted from gdb. So I am going to really add the format/vector to all the traps and exceptions by modifying the stack immediately upon entry and by patching back the stack before the "rte". If it works, I will add a CONFIG_ option so that we can build kernel with user app debug possibilities but somewhat slower in the exception processing, and 'production' kernel without. Thanks again for the hints, Cheers, Georges This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From davidm at snapgear.com Tue Mar 12 06:47:05 2002 From: davidm at snapgear.com (David McCullough) Date: Tue, 12 Mar 2002 21:47:05 +1000 Subject: [uClinux-dev] Problems with running init ? In-Reply-To: <3C8DC6C8.44AA80A6@cern.ch>; from jody.pearson@cern.ch on Tue, Mar 12, 2002 at 10:13:45AM +0100 References: <20020306214508.72425.qmail@web14309.mail.yahoo.com> <3C877F51.4772ECA3@cern.ch> <20020308095223.A30299@beast.internal.moreton.com.au> <3C8DC6C8.44AA80A6@cern.ch> Message-ID: <20020312214705.D3628@beast.internal.moreton.com.au> Jivin Jody Pearson lays it down ... > Hi David, > > I got the latest source tar from uclinux.org. > > Installed m68k-elf toolchain, you're right ! This is a lot easier. > > How much success have you had with the userland ? > > I seem to be getting a LOT of trouble with microwindows, and bash. Microwindows used to work on Xcopilot, thatys where I first ran it :-) Perhaps with all the kernel changes and what not it needs a little maintenance. Forget bash, it will never run on uClinux without a LOT of work. Use "sash" or the Minis shell (user/sh). > Also can't seem to get framebuffer support under a 2.4.x kernel. I don't think the driver is there yet. I use 2.0 for all my serious pilot stuff as the kernel is quite a bit smaller and allows you to get some real apps in there. Have you tried Microwindows with the 2.0 kernel, it should have all the required headers etc. > Have you sorted that one ? No, although I think someone mentioned it on the list just recently. Check the latest in CVS as it may have been added there since the last distribution cut. Cheers, Davidm -- David McCullough: Ph: +61 7 3435 2815 http://www.SnapGear.com davidm at snapgear.com Fx: +61 7 3891 3630 825 Stanley St., W'gabba QLD 4102, Oz This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From securez at teleline.es Tue Mar 12 06:44:40 2002 From: securez at teleline.es (Securez) Date: Tue, 12 Mar 2002 12:44:40 +0100 Subject: [uClinux-dev] MTD on uClinux Message-ID: <5.1.0.14.2.20020312122948.01ee17b0@pop3.iqc-services.com> I try to compile the kernel with support for MTD. I use the Dragonix board (http://www.openhardware.net/Dragonix), 68VZ328 and the FLASH are two banks of 1Mx16 (29LV160BE). The start adress for the FLASH are 0x2000000, i put this in the start adress in MTD, and 0x400000 for size. I get aproblem compiling the kernel, in the asm/io.h the macros: #define __raw_readb readb #define __raw_readw readw #define __raw_readl readl #define __raw_writeb writeb #define __raw_writew writew #define __raw_writel writel I don't know if this place is the better, or if it's correct. When i boot the kernel the message of MTD are show: physmap flash device:40000 at 2000000 but after this the kernel reboots, when i try this the first time, i use bbug to upload the image, and when the kernel startup, after the MTD message the loader program starts again and begin to erase the flash for coping the image again. Are the macro definitios correct? Are the asm/io.h the best palce? This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From Dave_Pfaltzgraff at patapsco.com Tue Mar 12 08:01:09 2002 From: Dave_Pfaltzgraff at patapsco.com (Dave_Pfaltzgraff at patapsco.com) Date: Tue, 12 Mar 2002 08:01:09 -0500 Subject: [uClinux-dev] Re: watchdog timers Message-ID: <85256B7A.00478502.00@patapsco.com> Dave Pfaltzgraff at PATAPSCO 03/12/2002 08:01 AM Greg wrote: > I use the driver "as is", no supporting user apps. I mostly > use it for debugging. So when I have a problem with that > "hangs" the system I enable to watchdog. 9 times out of 10 > ... What I meant by my request was: In order to use it, do you enable it (e.g: call watchdog_init()) from within a user application? Is there a corresponding 'h' file to define the interface? I'm not 100% sure, but it doesn't look like this is being included in our kernel builds. Is there a directive that controls this? Just as further input, the use of the watchdog as you've implemented it is good for debuging or for an evironment where multitasking is important. In our environment, there is only one task and if that fails we want to restart the whole system. This puts a slightly different slant on how you use the watchdog. Thanks once again, Dave This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From hamilton at SEDSystems.ca Tue Mar 12 09:36:11 2002 From: hamilton at SEDSystems.ca (Kendrick Hamilton) Date: Tue, 12 Mar 2002 08:36:11 -0600 (CST) Subject: [uClinux-dev] Re: watchdog timers In-Reply-To: <85256B7A.00478502.00@patapsco.com> Message-ID: We are using the watchdog for the same purpose. It is built into the 2.0.x kernel. See files driver/char/misc.c and arch/m68knommu/platform/68360/quicc/quicc_wdt.c -- Kendrick Hamilton E.I.T. SED Systems, a division of Calian Ltd. 18 Innovation Blvd. PO Box 1464 Saskatoon, Saskatchewan Canada S7N 3R1 Hamilton at sedsystems.ca Tel: (306) 933-1453 Fax: (306) 933-1486 On Tue, 12 Mar 2002 Dave_Pfaltzgraff at patapsco.com wrote: > > > > > Dave Pfaltzgraff at PATAPSCO > 03/12/2002 08:01 AM > > Greg wrote: > > I use the driver "as is", no supporting user apps. I mostly > > use it for debugging. So when I have a problem with that > > "hangs" the system I enable to watchdog. 9 times out of 10 > > ... > > What I meant by my request was: In order to use it, do you > enable it (e.g: call watchdog_init()) from within a user > application? Is there a corresponding 'h' file to define > the interface? > > I'm not 100% sure, but it doesn't look like this is being > included in our kernel builds. Is there a directive that > controls this? > > Just as further input, the use of the watchdog as you've > implemented it is good for debuging or for an evironment > where multitasking is important. In our environment, there > is only one task and if that fails we want to restart the > whole system. This puts a slightly different slant on how > you use the watchdog. > > Thanks once again, > Dave > > > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From Fabio.Ricciato at TILAB.COM Tue Mar 12 10:02:59 2002 From: Fabio.Ricciato at TILAB.COM (Ricciato Fabio) Date: Tue, 12 Mar 2002 16:02:59 +0100 Subject: [uClinux-dev] boot uClinux on ARM7TDMI and Integrator AP Message-ID: <6ECEC1E214F2E342814ABB1ED10E79550875F0@EXC2K01B.cselt.it> I have some questions about booting...probably dummy questions....:-< I've compiled my uclinux 2.4.17. Now I'd like to run it on my Integrator AP/ARM7TDMI using the serial port console to communicate with the kernel. I'd like to know: 1) which kind of boot loader do I have to use to load my uclinux binary and eventually the ramdisk ?- currently I've in the flash bootPROM from arm; 2) how can I have the console on the serial ? - I've included ARM serial support in the compiled kernel 3) how can I launch the login process ? Thanks a lot Fabio This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From stefan.jonsson at faltcom.se Tue Mar 12 11:13:46 2002 From: stefan.jonsson at faltcom.se (Stefan Jonsson) Date: Tue, 12 Mar 2002 17:13:46 +0100 Subject: [uClinux-dev] A few "simple" questions. Message-ID: Hello list, I am (still) new to uClinux (and linux) as well as to CGI (I have basic knoledge to make things happen though). Problem: I want to run CGI-script (and SSI) in C on my EB40LS board. My HTML document calls a cgi-script but the browser just wants to present the file on the screen instead of executing it, wether it is a plain shell-script or a compiled c program. Questions: Is the functionality of CGI dependent of the httpd? In that case, is the preinstalled httpd capable of CGI/SSI? Is it supposed to be some kind of config-file in /htdocs/cgi-bin/ where I keep my test.cgi -programs and what is supposed to be in that in that case? (have'nt found anything about that). Simply, what is needed? Any help or tips on this would be great, thank you. Regards, Stefan Jonsson, Student, University of Ume?, Sweden. This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From yvon at tempemicroelectronics.com Tue Mar 12 11:36:18 2002 From: yvon at tempemicroelectronics.com (yvon) Date: Tue, 12 Mar 2002 09:36:18 -0700 Subject: [uClinux-dev] Lineo uCEvolution board and uCDimm Message-ID: <005001c1c9e4$09c66d10$0200a8c0@CX2040412A> Hello, I have a uCEvolution board and 2 uCDimm that I bought from Lineo for a project that I intended to do with uCLinux. Unfortunately my customer read too much of Linux not suitable for 'real time' that I was forced to switch to a 68332 and uCOSII. If any of you is interested I will sell the package for $600.00 + shipment. Yvon PS: I am still convinced that uCLinux was the right choice for that application! -------------- next part -------------- An HTML attachment was scrubbed... URL: From hamilton at SEDSystems.ca Tue Mar 12 12:33:45 2002 From: hamilton at SEDSystems.ca (Kendrick Hamilton) Date: Tue, 12 Mar 2002 11:33:45 -0600 (CST) Subject: [uClinux-dev] memory/malloc problem Message-ID: Hi, I am using the uClinux 2.0.39.uc2 release of uClinux on a 68360 based system and am running into a malloc problem. Out application code does the following: StatusElement_t *newnode = new StatusElement_t; EAH_DEBUG("newnode = %8p", newnode); It returns that newnode is 0x01337098 Almost immediately after a bus error occurs with the KERNEL-STACK = 0132d000. The startup code initializes the kernel stack to 013ffffc. Has the kernel stack been moved or is this malloc returning a pointer in the middle of the stack? Note: chipselect 0 is devactived after all code has been relocated out of it. The interrupt vectors are also relocated. I did this so de-referencing a NULL pointer causes a bus error. Here is the complete dump: traps.c(73): BUSS ERROR TRAP CURRENT PROCESS: COMM=sios PID=51 TEXT=010c0058-010e3548 DATA=010e3548-010f93a8 BSS=010f93a8-010f9da4 USER-STACK=010fff30 KERNEL-STACK=0132d000 PC: 010d5a1a SR: 00000000 SP: 0132dfc0 d0: ffffffff d1: ffffffff d2: 010f1914 d3: 010fec1c d4: 010eb3b1 d5: 01319d58 a0: 00011b4a a1: 010f5f2c USP: 010fe9dc TRAPFRAME: 0132dfc0 CODE: 010d59fa: 296a0034 00342f0a 41f90000 899e4ebb 010d5a0a: 88002045 53a80014 588f2047 20aefdfc 010d5a1a: 42a741f9 00004e88 4ebb8800 20402f10 010d5a2a: 4878001b 2f2d0858 2f2d085c 486ef9f4 KERNEL STACK: 0132df80: 00f877e2 00fdff2e 00fdff2e 010f1914 0132df90: 0132dfc0 01337098 00000000 010e3548 0132dfa0: 00f81460 0132dfc0 00fdf839 00fdf831 0132dfb0: 00000049 01337018 00f8634c 0132dfc0 0132dfc0: ffffffff 010f1914 010fec1c 010eb3b1 0132dfd0: 01319d58 00011b4a 010f5f2c ffffffff 0132dfe0: ffffffff 00000000 0000010d 5a1ac008 0132dff0: 00011b4a 0000000c 010d5a1a 00090201 0132e000: 64657669 63650046 696c6520 746f6f20 0132e010: 6c617267 65005465 78742066 696c6520 0132e020: 62757379 004e6f74 20612074 79706577 0132e030: 72697465 7200546f 6f206d61 6e79206f USER STACK: 010fe9cc: 010f1914 010ff024 010d5a0c 01337018 010fe9dc: 010ff45c 010e55a2 010ffa9c 01319d98 010fe9ec: 01331678 00000007 010ff89c 01319dd8 010fe9fc: 01319dd8 2e630000 00013300 00000006 010fea0c: 00000001 0000003b 00000002 703a3131 010fea1c: 300a0001 010fea44 010db3ac 010feabf 010fea2c: 00000000 0000000a 00000000 00000000 010fea3c: 010feabf 010feb76 00000000 00000001 --- Kendrick Hamilton E.I.T. SED Systems, a division of Calian Ltd. 18 Innovation Blvd. PO Box 1464 Saskatoon, Saskatchewan Canada S7N 3R1 Hamilton at sedsystems.ca Tel: (306) 933-1453 Fax: (306) 933-1486 This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From ssoucek at coactive.com Tue Mar 12 12:44:48 2002 From: ssoucek at coactive.com (Stefan Soucek) Date: Tue, 12 Mar 2002 09:44:48 -0800 Subject: [uClinux-dev] building a "userland" for ARM Message-ID: I am not totally sure what your question "what the proper way" is; and frankly I don't know the official Lineo CD. You can, however, build all the stuff you need to get a working ARM system from source off the net. There are plenty of docs on www.uclinux.org, www.uclibc.org, www.busybox.net. A search on google for "uClinux" will also be useful to start with. This is a very rough overview of the things you need: - a cross-compilation tool chain (i.e. arm-elf tool chain) - boot code for your board - a uClinux kernel (for non-MMU) or a stock kernel (for MMU) - a console driver - uClibc - applications (init, sh, tinylogin, busybox) If you are able to build these you end up with a system you can log on to. Then there is a variety of networking apps out there (for which you may also need to write an Ethernet driver if not yet supplied with your board). Hope you can use those hints, -Stefan > -----Original Message----- > From: Stefan Jonsson [mailto:stefan.jonsson at faltcom.se] > Sent: Monday, March 11, 2002 11:41 PM > To: uclinux-dev at uclinux.org > Subject: SV: [uClinux-dev] building a "userland" for ARM > > > OOPS ! > > EB40LS of course ... "slip of the finger" :-) > > / Stefan > > -----Ursprungligt meddelande----- > Fr?n: owner-uclinux-dev at uclinux.org > [mailto:owner-uclinux-dev at uclinux.org]F?r Stefan Jonsson > Skickat: den 11 mars 2002 16:56 > Till: uClinux-dev mailing list > ?mne: [uClinux-dev] building a "userland" for ARM > > > Hello list, > > I have both the Lineo kit with the official CD and EB40LC > with uCnet and a > plain EB40 with memory module AT91MEC. I want to build a > working environment > for my other board without using the CD. Patching the kernel > seems to work > fine but then the problems starts. I cannot find "the proper > way" of doing > this. All readme:s seems to be for 68k. > > 1) Is there an explicit way of doing this for a EB40 standard > board and is > it described somewhere. (because it cannot be the same files > for ARM as for > 68k, can it?) > > 2) Is it possible to use "the official CD installation of > "useland" on the > other card (with different hardware) and in that case how can I modify > things to fit for that purpose? > > Regards, > Stefan Jonsson, > Student, University of Umea, Sweden. > > This message resent by the uclinux-dev at uclinux.org list server > http://www.uClinux.org/ > > This message resent by the uclinux-dev at uclinux.org list > server http://www.uClinux.org/ > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From eauth at softsys.co.at Tue Mar 12 14:30:06 2002 From: eauth at softsys.co.at (Erwin Authried) Date: Tue, 12 Mar 2002 20:30:06 +0100 Subject: [uClinux-dev] building a "userland" for ARM Message-ID: <01C1CA04.B33BB880@smithwicks.softsys.co.at> Stefan Jonsson[SMTP:stefan.jonsson at faltcom.se] wrote: > Hello list, > > I have both the Lineo kit with the official CD and EB40LC with uCnet and a > plain EB40 with memory module AT91MEC. I want to build a working environment > for my other board without using the CD. Patching the kernel seems to work > fine but then the problems starts. I cannot find "the proper way" of doing > this. All readme:s seems to be for 68k. > > 1) Is there an explicit way of doing this for a EB40 standard board and is > it described somewhere. (because it cannot be the same files for ARM as for > 68k, can it?) > > 2) Is it possible to use "the official CD installation of "useland" on the > other card (with different hardware) and in that case how can I modify > things to fit for that purpose? > > Regards, > Stefan Jonsson, > Student, University of Umea, Sweden. > > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ > > You could try my Midori uClinux distribution at http://uclinux.home.at/midori/ It contains the complete toolchain including elf2flt and the tools to create a rom image, a web based configuration tool, uClibc and the most important applications. -Erwin -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/ms-tnef Size: 2167 bytes Desc: not available URL: From ken at reasonability.net Tue Mar 12 19:42:03 2002 From: ken at reasonability.net (Ken Treis) Date: Tue, 12 Mar 2002 16:42:03 -0800 Subject: [uClinux-dev] XIP Message-ID: <3C8EA05B.4080201@reasonability.net> How can I tell if I'm doing XIP? I've compiled my code with -msep-data, but it's running out of JFFS. Is it even possible to do XIP in this situation? Any help greatly appreciated. Along similar lines, I've been asked if there are any memory profilers that will run on uClinux. I have no experience in this area, but I know a mailing list full of people who might know. :) Ken Treis ken at reasonability.net This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From falls at huayuan.org Tue Mar 12 23:07:10 2002 From: falls at huayuan.org (falls) Date: Wed, 13 Mar 2002 04:07:10 GMT Subject: [uClinux-dev] uClinux crash when I copy large file to directory shared by samba Message-ID: <20020313040710.30779.qmail@huayuan.org> Hello all! 1. I have successfully mounted my IDE disk(4G,ext2) on my uClinux(distribution 20020220). 2. Then I rewrited smb.conf to share /mnt/disk1. 3. Reboot uClinux 4. Start nmbd; 5. I can see a pc named "uClinux-coldfire" in M$ windows2000 Network Neighbor. 6. Open "uClinux-coldfire" in Network Neighbor and see the disk1(shared by samba). 7. I copy a file(size=500k) to disk1 and delete it,all is ok. 8. I am very happy. 9. I tried to copy a large file(size=3M) to disk1 from windows2000; 10. The uClinux crashed when the large file is transfered about 60%-70%. I rebooted uClinux many many times. But every time when I copied large file to disk1 ,the uClinux crashed. Who can help me ? Thx!!!!! This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From miles at lsi.nec.co.jp Wed Mar 13 02:29:27 2002 From: miles at lsi.nec.co.jp (Miles Bader) Date: 13 Mar 2002 16:29:27 +0900 Subject: [uClinux-dev] buglet in `tinylogin' with new uClibc stdio Message-ID: When I telnet to my board, which uses apps compiled with the new uClibc stdio, I can login succesfully, but /etc/motd (which is output by tinylogin before execing the shell) is truncated: (images) telnet mcspx1 Trying 10.30.114.220... Connected to mcspx1.ucom.lsi.nec.co.jp. Escape character is '^]'. mcspx1.ucom.lsi.nec.co.jp login: miles No directory, logging in with HOME=/ Welcome to ____ _ _ _ ___ ____ ___ _ _ / ___| (_)_ __ _ _ _ _ / >_ _( _ ) ___| / _ \ | | | | | | | | '_ \| | | < \/ > / /< \ / > _ \___ \| | | | | |_| | |_ BusyBox v0.60.0 (2002.03.13-02:50+0000) Built-in shell (msh) Enter 'help' for a list of built-in commands. $ Looking at the code (in `user/tinylogin/login.c' from the uClinux-dist-20020214 distribution), it outputs motd using a loop, in the `motd' function: if ((fp = fopen("/etc/motd", "r")) != NULL) { while ((c = getc(fp)) != EOF) { putchar(c); } fclose(fp); } and then the caller of `motd' (`login_main') goes ahead and execs the shell. So it seems that the problem is that stdio isn't flushed. Here's a patch that fixes the problem (for the uClinux-dev people): diff -ruP -X../diff.xcl ../orig/uClinux-dist/user/tinylogin/login.c ./user/tinylogin/login.c --- ../orig/uClinux-dist/user/tinylogin/login.c Thu Feb 14 11:44:34 2002 +++ user/tinylogin/login.c Wed Mar 13 16:15:42 2002 @@ -405,6 +405,7 @@ while ((c = getc(fp)) != EOF) { putchar(c); } + fflush (stdout); fclose(fp); } } However, I wonder why this didn't happen with the old stdio.... Was it not buffered, or did it default to line buffering? Is this a potential compatibility problem? -Miles -- Love is a snowmobile racing across the tundra. Suddenly it flips over, pinning you underneath. At night the ice weasels come. --Nietzsche This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From andersen at codepoet.org Wed Mar 13 02:54:53 2002 From: andersen at codepoet.org (Erik Andersen) Date: Wed, 13 Mar 2002 00:54:53 -0700 Subject: [uClinux-dev] Re: [uClibc]buglet in `tinylogin' with new uClibc stdio In-Reply-To: References: Message-ID: <20020313075453.GA7425@codepoet.org> On Wed Mar 13, 2002 at 04:29:27PM +0900, Miles Bader wrote: > When I telnet to my board, which uses apps compiled with the new uClibc > stdio, I can login succesfully, but /etc/motd (which is output by > tinylogin before execing the shell) is truncated: [---------snip----------] > So it seems that the problem is that stdio isn't flushed. Here's a > patch that fixes the problem (for the uClinux-dev people): Patch is obviously correct. Applied (to tinylogin). > However, I wonder why this didn't happen with the old stdio.... Guess you just got lucky. > Was it not buffered, or did it default to line buffering? Is this a > potential compatibility problem? The old stdio was also buffering. Regardless, the new behavior is correct and seems to do a better job at showing bugs... I'll take obviously broken over silently broken any day. -Erik -- Erik B. Andersen http://codepoet-consulting.com/ --This message was written using 73% post-consumer electrons-- This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From christian.opel at brooks.com Wed Mar 13 03:14:39 2002 From: christian.opel at brooks.com (Opel, Christian) Date: Wed, 13 Mar 2002 03:14:39 -0500 Subject: [uClinux-dev] problems with libio and libiostream Message-ID: <316532F8F79ED41184540008C7282A54396B44@mail-europe3.brooks.com> We`re trying to build some oo c++ code for the motorola m5272 coldfire processor running uClinux Distr 2.0.38. The host dev. system is Suse 7.3. Compiler, Linker, libs etc. are installed successfully and even if i try to compile and link an empty Project i get executable binaries. The nec. c++ headers are also installed. But if i try to use a functionality of iostream or stdio stderr etc., I get linker errors like this one. /usr/local/lib/gcc-lib/m68k-elf/2.95.2/libgcc.a(_eh.o): in function `__default_terminate`: /home/davidm/play/m68k-elf-tools/m68k-elf-gcc/gcc/../../gcc-2.95.2/gcc/libgc c2.c(.text=0x6): undefined reference to `abort` and many other undefined references ( _stdout _stderr ). It seems I am using the wrong libraries. So my question is : Are there any cpu - specific libraries i have to use. Does anyone can give me a short explanation of successfully porting c++ code. I use libc.a and libm.a builded with the distribution of uClinux. C code is compiled and linked without errors but c++ ... Any help greatly appreciated. This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From seemanta at recjai.ac.in Wed Mar 13 03:46:14 2002 From: seemanta at recjai.ac.in (Seemanta Dutta) Date: Wed, 13 Mar 2002 14:16:14 +0530 (IST) Subject: [uClinux-dev] A few "simple" questions. In-Reply-To: Message-ID: > Hello list, > > I am (still) new to uClinux (and linux) as well as to CGI (I have basic > knoledge to make things happen though). > > Problem: I want to run CGI-script (and SSI) in C on my EB40LS board. My HTML > document calls a cgi-script but the browser just wants to present the file > on the screen instead of executing it, wether it is a plain shell-script or > a compiled c program. > > Questions: > Is the functionality of CGI dependent of the httpd? > > In that case, is the preinstalled httpd capable of CGI/SSI? > > Is it supposed to be some kind of config-file in /htdocs/cgi-bin/ where I > keep my test.cgi -programs and what is supposed to be in that in that case? > (have'nt found anything about that). > > Simply, what is needed? > > Any help or tips on this would be great, thank you. > > Regards, > Stefan Jonsson, > Student, University of Ume?, Sweden. > > ******************************************** hello stefan/list happy to hear that the list contains newbies too....well i am a complete newbie to the world of embedded linux though not one in the world of embeded systems.... i am thinking of doing my summer project on uClinux this year..but till now i have not been able to get a good start due to two reasoons: 1. lack of good documentation for *beginners* 2. someone to guide. well if you have any idea from where to start or from where to get any downloadable turorials please help me out of this... ******************************************** now ur problem... answer 1. yes cgi sripts are affected by the httpd. answer 2. yes and no.yes because in most cases htpd comes configured...no because otherwise u would have to reconfigure it ureself.... thatis simple....just go to /etc/httpd/config and then edit httpd.conf file..syntax of this file can be obtained from the manuals itself.. i think after that u sohuld be able to solve ur problem... it is the conf file that determines where ur scripots wouyld be kept and whether they can run or not..please note that the above mentioned path is for my 7.1 linux box only wherein i have sucessfully configured and ran cgi/ssi scripts.... i hope i am able to help u....btw a few diagnostic messages or error messages would have been more helpful for me to provide a good answer... ******************************************** please reply soon with details of uClinux soon... yours sincerely seemanta dutta -- 2nd Year B.E. Electronics and Communication Engineering Hostel No:6,Room No:05 Malaviya Regional Engineering College Jaipur,Rajasthan PIN-302017 INDIA P.S. While replying please send a copy to seemanta_18 at yahoo.com. Thank You. This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From chenypg at hotmail.com Wed Mar 13 06:59:25 2002 From: chenypg at hotmail.com (chen yp) Date: Wed, 13 Mar 2002 11:59:25 +0000 Subject: [uClinux-dev] Ide Compile problem! Message-ID: Hi,all When I try to add ide disk support to the 2.4.x kernel using the latest distribution uClinux-dist-20020306.tar.gz issued by snapgear,I encounted the following problem. {standard input}: Assembler messages: {standard input}:11392: Error: operands mismatch -- statement `move.l 4(%a0,%d0. l),-106(%a6)' ignored What happen to kernel source. Thanks all for help. _________________________________________________________________ ???????? Web ?????? ?? MSN Hotmail? http://www.hotmail.com/cn This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From mirco_cremasco at hotmail.com Wed Mar 13 08:02:12 2002 From: mirco_cremasco at hotmail.com (mirco cremasco) Date: Wed, 13 Mar 2002 14:02:12 +0100 Subject: [uClinux-dev] Ide Compile problem! Message-ID: >From: "chen yp" >Reply-To: uclinux-dev at uclinux.org >To: uclinux-dev at uclinux.org >Subject: [uClinux-dev] Ide Compile problem! >Date: Wed, 13 Mar 2002 11:59:25 +0000 > >Hi,all > >When I try to add ide disk support to the 2.4.x kernel using the latest >distribution uClinux-dist-20020306.tar.gz issued by snapgear,I encounted >the following problem. > >{standard input}: Assembler messages: >{standard input}:11392: Error: operands mismatch -- statement `move.l >4(%a0,%d0. >l),-106(%a6)' ignored > >What happen to kernel source. Thanks all for help. > I've found the same problem, try to modify include/linux/genhd.h file on line 76 deleting the "unsigned" in the struct members definition: struct hd_struct { unsigned long start_sect; ^^^^^^^^ unsigned long nr_sects; ^^^^^^^ devfs_handle_t de; /* primary (master) devfs entry */ int number; /* stupid old code wastes space */ } A question: which board are you using. I'm developing a board that will support ide disk in near future. Let's stay in touch. Bye Mirco Cremasco _________________________________________________________________ Prova MSN Messenger per conversare in linea con i tuoi amici: http://messenger.msn.it This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From christian.opel at brooks.com Wed Mar 13 08:19:35 2002 From: christian.opel at brooks.com (Christian Opel) Date: 13 Mar 2002 14:19:35 +0100 Subject: [uClinux-dev] A few "simple" questions. In-Reply-To: References: Message-ID: <1016025578.1796.13.camel@christian> Am Mit, 2002-03-13 um 09.46 schrieb Seemanta Dutta: > > > Hello list, > > > > I am (still) new to uClinux (and linux) as well as to CGI (I have basic > > knoledge to make things happen though). > > > > Problem: I want to run CGI-script (and SSI) in C on my EB40LS board. My HTML > > document calls a cgi-script but the browser just wants to present the file > > on the screen instead of executing it, wether it is a plain shell-script or > > a compiled c program. > > Make shure your web server is running and/or you provide a cgi interface Check settings in /etc/httpd.conf like your path to cgi-bin Regards Christian This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From sujz at cad.csie.ncku.edu.tw Wed Mar 13 09:12:57 2002 From: sujz at cad.csie.ncku.edu.tw (SuCC) Date: Wed, 13 Mar 2002 22:12:57 +0800 Subject: [uClinux-dev] Problem of page allocation Message-ID: <000001c1ca99$33a6eff0$1af7748c@SCC> Hi, I have a problem about uClinux page allocation. When I run a program, the system will allocate some pages to this program. But when this program is terminated, I cat /proc/meminfo and find that these pages are not released. This will cause the memory decreasing and finally empty. I call show_mem() in arch/armnommu/mm/init.c and there are some information show that the value of "pages shared" increase each time when I run a program Is this a normal case or this is my system bugs, thks!! -------------- next part -------------- An HTML attachment was scrubbed... URL: From hamilton at SEDSystems.ca Wed Mar 13 09:46:43 2002 From: hamilton at SEDSystems.ca (Kendrick Hamilton) Date: Wed, 13 Mar 2002 08:46:43 -0600 (CST) Subject: [uClinux-dev] Problem of page allocation In-Reply-To: <000001c1ca99$33a6eff0$1af7748c@SCC> Message-ID: Which kernel are you using? Kendrick Hamilton E.I.T. SED Systems, a division of Calian Ltd. 18 Innovation Blvd. PO Box 1464 Saskatoon, Saskatchewan Canada S7N 3R1 Hamilton at sedsystems.ca Tel: (306) 933-1453 Fax: (306) 933-1486 On Wed, 13 Mar 2002, SuCC wrote: > Hi, > I have a problem about uClinux page allocation. > When I run a program, the system will allocate some pages to this > program. > But when this program is terminated, I cat /proc/meminfo and find that > these pages are not released. > This will cause the memory decreasing and finally empty. > > I call show_mem() in arch/armnommu/mm/init.c and there are some > information show that > the value of "pages shared" increase each time when I run a program > > Is this a normal case or this is my system bugs, thks!! > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From sujz at cad.csie.ncku.edu.tw Wed Mar 13 11:16:30 2002 From: sujz at cad.csie.ncku.edu.tw (SuCC) Date: Thu, 14 Mar 2002 00:16:30 +0800 Subject: [uClinux-dev] Problem of page allocation In-Reply-To: Message-ID: <000e01c1caaa$6f69cba0$1af7748c@SCC> The kernel version is 2.4.10-uc2. -----Original Message----- From: owner-uclinux-dev at uclinux.org [mailto:owner-uclinux-dev at uclinux.org] On Behalf Of Kendrick Hamilton Sent: Wednesday, March 13, 2002 10:47 PM To: uclinux-dev at uclinux.org Subject: Re: [uClinux-dev] Problem of page allocation Which kernel are you using? Kendrick Hamilton E.I.T. SED Systems, a division of Calian Ltd. 18 Innovation Blvd. PO Box 1464 Saskatoon, Saskatchewan Canada S7N 3R1 Hamilton at sedsystems.ca Tel: (306) 933-1453 Fax: (306) 933-1486 On Wed, 13 Mar 2002, SuCC wrote: > Hi, > I have a problem about uClinux page allocation. > When I run a program, the system will allocate some pages to this > program. > But when this program is terminated, I cat /proc/meminfo and find that > these pages are not released. > This will cause the memory decreasing and finally empty. > > I call show_mem() in arch/armnommu/mm/init.c and there are some > information show that > the value of "pages shared" increase each time when I run a program > > Is this a normal case or this is my system bugs, thks!! > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From hamilton at SEDSystems.ca Wed Mar 13 11:57:14 2002 From: hamilton at SEDSystems.ca (Kendrick Hamilton) Date: Wed, 13 Mar 2002 10:57:14 -0600 (CST) Subject: [uClinux-dev] Which GDB Message-ID: Which version of gdb are people using with gdb server to debug application code? Where did people download gdb server from? TIA, Kendrick Hamilton E.I.T. SED Systems, a division of Calian Ltd. 18 Innovation Blvd. PO Box 1464 Saskatoon, Saskatchewan Canada S7N 3R1 Hamilton at sedsystems.ca Tel: (306) 933-1453 Fax: (306) 933-1486 This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From hamilton at SEDSystems.ca Wed Mar 13 11:44:34 2002 From: hamilton at SEDSystems.ca (Kendrick Hamilton) Date: Wed, 13 Mar 2002 10:44:34 -0600 (CST) Subject: [uClinux-dev] EN360 parts (fwd) Message-ID: ---------- Forwarded message ---------- Date: Wed, 13 Mar 2002 10:04:43 -0500 From: John Fabrizio To: Kendrick Hamilton Subject: EN360 parts Hi Kendrick, I have been lurking on the uclinux dev list for a couple of years now and have noted that you have done work on the EN360. We (Rt Control/Lineo Canada) have as well. We have a stock of MC68EN360EM25VL that we would like to move if you are interested. Of course the price would be much more favourable than current market. Let me know if you are at all interested. Thanks, John Fabrizio Lineo Canada Corp. This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From ken at reasonability.net Wed Mar 13 13:23:57 2002 From: ken at reasonability.net (Ken Treis) Date: Wed, 13 Mar 2002 10:23:57 -0800 Subject: [uClinux-dev] Which GDB References: Message-ID: <3C8F993D.9060608@reasonability.net> Kendrick Hamilton wrote: > Which version of gdb are people using with gdb server to debug > application > code? Where did people download gdb server from? > I use the m68k-bdm-elf-gdb that I built for BDM debugging (gdb 5.0 + patches). For gdbserver, I use the one in uClinux-dist-20020220/user/gdbserver. HTH, -- Ken Treis / Software Engineer Key Technology, Inc. ktreis at keyww.com ---- -------------- next part -------------- An HTML attachment was scrubbed... URL: From hamilton at SEDSystems.ca Wed Mar 13 14:38:50 2002 From: hamilton at SEDSystems.ca (Kendrick Hamilton) Date: Wed, 13 Mar 2002 13:38:50 -0600 (CST) Subject: [uClinux-dev] Which GDB In-Reply-To: <3C8F993D.9060608@reasonability.net> Message-ID: Have you had this combination of gdb and gdb server working together? Kendrick Hamilton E.I.T. SED Systems, a division of Calian Ltd. 18 Innovation Blvd. PO Box 1464 Saskatoon, Saskatchewan Canada S7N 3R1 Hamilton at sedsystems.ca Tel: (306) 933-1453 Fax: (306) 933-1486 On Wed, 13 Mar 2002, Ken Treis wrote: > Kendrick Hamilton wrote: > > > Which version of gdb are people using with gdb server to debug > > application > > code? Where did people download gdb server from? > > > I use the m68k-bdm-elf-gdb that I built for BDM debugging (gdb 5.0 + > patches). For gdbserver, I use the one in > uClinux-dist-20020220/user/gdbserver. > > HTH, > > -- > Ken Treis / Software Engineer > Key Technology, Inc. > ktreis at keyww.com > ---- > > > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From ken at reasonability.net Wed Mar 13 16:23:22 2002 From: ken at reasonability.net (Ken Treis) Date: Wed, 13 Mar 2002 13:23:22 -0800 Subject: [uClinux-dev] Which GDB References: Message-ID: <3C8FC34A.7090409@reasonability.net> Kendrick Hamilton wrote: > Have you had this combination of gdb and gdb server working together? > At one point, I seem to remember having these working together. I used remote debugging over a network connection. But now, it's not working for me. I can even manage to panic the uClinux (2.4.17) kernel (all it says is "BUG!"). Sorry for the red herring. -- Ken Treis ken at reasonability.net -------------- next part -------------- An HTML attachment was scrubbed... URL: From pgccoop at hotmail.com Wed Mar 13 16:44:52 2002 From: pgccoop at hotmail.com (Jonathan Schmidt) Date: Wed, 13 Mar 2002 21:44:52 +0000 Subject: [uClinux-dev] "Calibrating delay loop..." crashes Message-ID: I have been building kernels from Daniel Haenesse's "unified-2.4.6" distribution for my DragonBall VZ board, and they have been working just fine. I tried yesterday to build a kernel straight from the newest uClinux CVS (2.4.17 mostly, as I understand). When the image is programmed to the flash (using a modified "2flash" to support the TE28F320B3 Intel flash), and I reboot, the kernel hangs at "Calibrating delay loop...". I checked in the .c file that is responsible for this part of the bootup sequence, and there doesn't seem to be anything that could cause a crash. I was just wondering if anyone has had this type of problem with a VZ board running a 2.4 kernel before, and if so, how did you resolve it. Thanks in advance for any help you can provide. _________________________________________________________________ Join the world?s largest e-mail service with MSN Hotmail. http://www.hotmail.com This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From gerg at snapgear.com Wed Mar 13 20:45:27 2002 From: gerg at snapgear.com (Greg Ungerer) Date: Thu, 14 Mar 2002 11:45:27 +1000 Subject: [uClinux-dev] Re: watchdog timers References: <85256B7A.00478502.00@patapsco.com> Message-ID: <3C9000B7.4090005@snapgear.com> Hi Dave, Dave_Pfaltzgraff at patapsco.com wrote: > What I meant by my request was: In order to use it, do you > enable it (e.g: call watchdog_init()) from within a user > application? Is there a corresponding 'h' file to define > the interface? Tp use it enable the kernel config option "Watchdog Timer Support" in the character devices section. > I'm not 100% sure, but it doesn't look like this is being > included in our kernel builds. Is there a directive that > controls this? > > Just as further input, the use of the watchdog as you've > implemented it is good for debuging or for an evironment > where multitasking is important. In our environment, there > is only one task and if that fails we want to restart the > whole system. This puts a slightly different slant on how > you use the watchdog. It does, and the current code doesn't deal with this mode of operation. But it would be very simple to add, and making it a kernel config option it can co-exist with the current functionality. Regards Greg ------------------------------------------------------------------------ Greg Ungerer -- Chief Software Wizard EMAIL: gerg at snapgear.com SnapGear PHONE: +61 7 3435 2888 825 Stanley St, FAX: +61 7 3891 3630 Woolloongabba, QLD, 4102, Australia WEB: www.snapgear.com This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From gerg at snapgear.com Wed Mar 13 21:42:46 2002 From: gerg at snapgear.com (Greg Ungerer) Date: Thu, 14 Mar 2002 12:42:46 +1000 Subject: [uClinux-dev] patch - MC68328.h & MC68VZ328.h ICR definition References: Message-ID: <3C900E26.40207@snapgear.com> Hi Daniel, Daniel Potts wrote: > ICR definition was 0xffff302 instead of 0xfffff302 (missing an f). > > It appears to be correct in MC68EZ328.h. Applied. Thanks Greg ------------------------------------------------------------------------ Greg Ungerer -- Chief Software Wizard EMAIL: gerg at snapgear.com SnapGear PHONE: +61 7 3435 2888 825 Stanley St, FAX: +61 7 3891 3630 Woolloongabba, QLD, 4102, Australia WEB: www.snapgear.com This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From chenypg at hotmail.com Wed Mar 13 22:03:11 2002 From: chenypg at hotmail.com (chen yp) Date: Thu, 14 Mar 2002 03:03:11 +0000 Subject: [uClinux-dev] Ide disk write data problem Message-ID: Thanks for your help. I have solved this problem. But I have another problem which looks like that some interrupt happened when system was executing schedul process when I try write some data to ide disk. Where should I mask ide interrupt in the kernel. The following is the erro massage.Any suggestion is appreciating! enter ide_intr enter ide_intr bad frame format: 00000000 PC: [<00104664>] SR: 2009 SP: 0022bf90 a2: 003b0000 d0: f43fd806 d1: 21539780 d2: 001aa000 d3: 0025a310 d4: 00000000 d5: 001a35a0 a0: 2f002f00 a1: 15936f87 Process ksoftirqd_CPU0 (pid: 3, stackpage=0022b000) Frame format=4 Stack from 0022bfcc: 00000008 001a35a0 10000000 0001b074 00010f00 00106a1a 001a35c8 0022bff4 00108fc0 00235fa8 00235fb4 001003b2 00000000 Call Trace: [<00106a1a>] [<00108fc0>] [<001003b2>] Code: 2228 0020 e281 7014 90a8 0024 e480 2241 43f1 0801 bad frame format: 00000000 PC: [<00107796>] SR: 2000 SP: 0022bf00 a2: 2f002f00 Any suggestion is appreciating! >From: "mirco cremasco" >Reply-To: uclinux-dev at uclinux.org >To: uclinux-dev at uclinux.org >Subject: Re: [uClinux-dev] Ide Compile problem! >Date: Wed, 13 Mar 2002 14:02:12 +0100 > > > > >>From: "chen yp" >>Reply-To: uclinux-dev at uclinux.org >>To: uclinux-dev at uclinux.org >>Subject: [uClinux-dev] Ide Compile problem! >>Date: Wed, 13 Mar 2002 11:59:25 +0000 >> >>Hi,all >> >>When I try to add ide disk support to the 2.4.x kernel using the >>latest >>distribution uClinux-dist-20020306.tar.gz issued by snapgear,I >>encounted >>the following problem. >> >>{standard input}: Assembler messages: >>{standard input}:11392: Error: operands mismatch -- statement >>`move.l >>4(%a0,%d0. >>l),-106(%a6)' ignored >> >>What happen to kernel source. Thanks all for help. >> >I've found the same problem, >try to modify include/linux/genhd.h file on line 76 >deleting the "unsigned" in the struct members definition: > struct hd_struct { > unsigned long start_sect; > ^^^^^^^^ > unsigned long nr_sects; > ^^^^^^^ > devfs_handle_t de; /* primary (master) devfs entry */ > int number; /* stupid old code wastes space */ > } > >A question: which board are you using. I'm developing a board that >will >support ide disk in near future. Let's stay in touch. > >Bye Mirco Cremasco > >_________________________________________________________________ >Prova MSN Messenger per conversare in linea con i tuoi amici: >http://messenger.msn.it > >This message resent by the uclinux-dev at uclinux.org list server >http://www.uClinux.org/ _________________________________________________________________ ???????? Web ?????? ?? MSN Hotmail? http://www.hotmail.com/cn This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From superwen at 263.net Wed Mar 13 22:46:20 2002 From: superwen at 263.net (Kevin Zhu) Date: Thu, 14 Mar 2002 11:46:20 +0800 Subject: [uClinux-dev] how to do with stack pointer? Message-ID: <20020314034340.7D4BE1C865831@smtp.263.net> Hi,all, I'm trying to port uClinux 2.4 on my MC68VZ328ADS board. But failed. It starts up correctly, and can print out 'ABCDEFGH' in my terminal, but failed right at the "jsr start_kernel". I have traced the startup sequence and found out that the kernel is failed in printk(), it can't print out the message. I'm wondering if I did something wrong at the stack pointer? I actually have no idea what value should be passed to it?and stack size?... Can anybody help me on this? Thanks & Best Regards Kevin ------------------------------------- moveal #_sbss, %a0 moveal #_ebss, %a1 /*Fill .bss with ZEROs until %a0 == %a1 */ L1: movel #0,%a0 at + cmpal %a0, %a1 bhi L1 /* Copy data segment from ROM to RAM */ moveal #__data_rom_start, %a0 moveal #_sdata, %a1 moveal #_edata, %a2 /* Copy %a0 to %a1 until %a1 == %a2 */ L2: movel %a0 at +, %a1 at + cmpal %a1, %a2 bhi L2 /*************Copy it from uCdimm ctr0_fixed.S**********/ movel #_sdata, %d0 movel %d0, _rambase movel #_ebss, %d0 movel %d0, _ramstart movel #__ramend, %d0 movel %d0, _ramend movel #__ramvec, %d0 movel %d0, _ramvec #ifdef CONFIG_68328_SERIAL_UART2 /* Enable RXD TXD port bits to enable UART2 */ moveb #0xcf, 0xfffff43b #endif /* * load the current task pointer and stack */ moveq #71,%d7 /*'G'*/ moveb %d7,0xfffff907 lea init_task_union,%a0 movel %a0, _current_task lea 0x2000(%a0),%sp moveq #72,%d7 /*'H' can be print out correctly, but it is stuck right after that*/ moveb %d7,0xfffff907 lp: jsr start_kernel jmp lp This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From sheeliang at lit.org.sg Thu Mar 14 01:58:14 2002 From: sheeliang at lit.org.sg (Chia Shee Liang) Date: Thu, 14 Mar 2002 14:58:14 +0800 Subject: [uClinux-dev] Memory corruption? Message-ID: <20020314145814.A1695@localhost.localdomain> Does anybody have any advice on debugging memory corruption? My kernel has this problem, looks as if the free lists is not maintained properly, resulting in double usage. Running uClinux 2.4.17-uc0 on ARM7TDMI on Integrator AP. This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From gmenie at akamai.com Thu Mar 14 03:59:42 2002 From: gmenie at akamai.com (Menie, Georges) Date: Thu, 14 Mar 2002 03:59:42 -0500 Subject: [uClinux-dev] Which GDB Message-ID: the latest toolchain and distrib from : http://www.uclinux.org/pub/uClinux/dist/index.html contains a working m68k-elf-gdb and gdbserver the problem might be the kernel, I know for the 68328/EZ/VZ that the traps are *not* set in a way to allow both tools to work together, I am trying to make it work but I don't have much time right now. I will send update to the list as soon as possible probably next week. Georges -----Original Message----- From: Ken Treis [mailto:ken at reasonability.net] Sent: mercredi 13 mars 2002 19:24 To: uclinux-dev at uclinux.org Subject: Re: [uClinux-dev] Which GDB Kendrick Hamilton wrote: Which version of gdb are people using with gdb server to debug application code? Where did people download gdb server from? I use the m68k-bdm-elf-gdb that I built for BDM debugging (gdb 5.0 + patches). For gdbserver, I use the one in uClinux-dist-20020220/user/gdbserver. HTH, -- Ken Treis / Software Engineer Key Technology, Inc. ktreis at keyww.com ---- -------------- next part -------------- An HTML attachment was scrubbed... URL: From christian.opel at brooks.com Thu Mar 14 07:29:41 2002 From: christian.opel at brooks.com (Christian Opel) Date: 14 Mar 2002 13:29:41 +0100 Subject: [uClinux-dev] Message-ID: <1016108984.1283.3.camel@christian> Is there already a pthread library for uClinux. I am trying to port JTC for Motorola coldfire. This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From dchasse at sierracom.com Thu Mar 14 05:09:08 2002 From: dchasse at sierracom.com (Denise Chasse) Date: Thu, 14 Mar 2002 10:09:08 -0000 Subject: [uClinux-dev] Applications, processes, and threads... Message-ID: <01C1CB40.485EC380.dchasse@sierracom.com> After mulling through the FAQs and email forum, I am still very confused. With such confusion I would not even label myself as a newbie, that will be my next promotion. So with that said, here goes: The goal is to port an application that I wrote for PSOS to uClinux. My platform is the MCF5272. I'm trying to understand how applications, processes, and threads all share resources. Without an MMU, I have certain limitations creating processes, ie vfork instead of fork. Apparently, I need to follow vfork() with a call to exec() before the parent runs again. After I call exec(), is the child terminated? Can I get back to the child without recreating it with another vfork()? This is my interpretation from the archived information but it doesn't seem to make sense. If this is truly the case, how is this effectively any different than just calling a function from within the parent and just skip to vfork all together? If I can't have a process running in the background, can I have multiple applications. In addition to my specific application, eventually I would like to support SNMP, FTP, DHCP and others. Can these all be done? Any caveats? Thanks, Newbie-in-training This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From stefan.jonsson at faltcom.se Thu Mar 14 10:48:12 2002 From: stefan.jonsson at faltcom.se (Stefan Jonsson) Date: Thu, 14 Mar 2002 16:48:12 +0100 Subject: [uClinux-dev] CGI and httpd again, explicit questions. Message-ID: Hello list, Yes, hold on now, because here comes some real "newbie" questions ;-) It is regarding uClinux on EB40LS (from Lineo) and CGI again. Simply, I don't get it to work. I did get a suggestion like this from the uClinux-mail-list: "Make shure your web server is running and/or you provide a cgi interface Check settings in /etc/httpd.conf like your path to cgi-bin" *** Explicit-question-1 *** But how do I make sure it "provides a CGI interface"? I made a httpd.conf (because there were none) like this in the suggested location. # Trying to get httpd to exec my CGI- # -scripts located in /htdocs/cgi-bin/ # 2002-03-14 Stefan Jonsson ScriptAlias /cgi-bin/ "/htdocs/cgi-bin/" Options ExecCGI # CGI exec permitted Options Includes # SSI includes allowed Since I really don't know, I wonder if this is "the way". I have found really nothing about this on uClinux, unfortunately ... and therefore I bother u ;-) *** Explicit-question-2 *** Does the provided httpd work with cgi and provide that interface or is the problem there? My call for the cgi-script (in C) looks like this:
*** Explicit-question-3 *** Is it right so far? (the problem is that it does not execute, just shows the text "bFLT" (binary flat format?) on the screen ... or text if I try to execute a shellscript as cgi). *** Explicit-question-4 and on ... *** 4.1 Shall the httpd.conf be located in /etc/? (or shall the config be named something else and be located elsewhere?) 4.2 Should it be a httpd.conf at all in uClinux shipped installation from Lineo (since there is none such to begin with), or should it maybe be configurated elsewhere (as in inetd.conf, 'services' or something)? 4.3 In services it says: http stream tcp nowait root /sbin/httpd -i What does the '-i' stand for? Is it possible to define something here? 4.4 Should something special be defined somewhere in the HTML-"code" to get a browser to understand that there will be CGI? So, now my questions are finally to an end ... for this time. I thank you for already have been helpful in many ways, but this problem remains for me. I would be happy for any suggestions and tips/tricks about this matter. Thanks! Regards, Stefan Jonsson, Student, University of Ume?, Sweden. This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From david.coldfire at tiscali.co.uk Thu Mar 14 11:14:10 2002 From: david.coldfire at tiscali.co.uk (Mr Anderson) Date: Thu, 14 Mar 2002 16:14:10 -0000 Subject: [uClinux-dev] MTD with uClinux-2.4... ioctl error Invalid ioctl Message-ID: <000801c1cb73$474275e0$d7a7403e@coldfiretiscali.co.uk> Hello folks, Does anyone know the cause of Invalid ioctl 5401 ( MEMGETINFO = 0x8........ ) with the mtdchar driver. I'm using MCF5272C3 motorola development platform, with the tarifa patch for 2.0.38 kernel fixed for 2.4.. (I think). I'm making the mtd devices O.K. and they appear allocated into the correct boundaries. When I use debug to write ABCDEFGH string to the user flash area, I expect to be able to cat /dev/mtdx and find the string in one of the mtd devices. The cat command performs mtd_open mtd_ioctl Invalid ioctl ( details above ) mtd_close The mtd dumped memory actually appears to access RAM and not FLASH, which makes sense if MEMGETINFO did not work ( MTD would have no knowledge of the base address of flash ). Someone has been here before ! Regards David Anderson mailto:djanderson at iee.org ============================================== FF := 2B | (~2B) ! ============================================== -------------- next part -------------- An HTML attachment was scrubbed... URL: From philip at engarts.com Thu Mar 14 12:31:10 2002 From: philip at engarts.com (Philip Nye) Date: Thu, 14 Mar 2002 17:31:10 -0000 Subject: [uClinux-dev] Applications, processes, and threads... References: <01C1CB40.485EC380.dchasse@sierracom.com> Message-ID: <01cf01c1cb7e$084f0a00$5801a8c0@localdomain> Denise, I am not right up to speed on this, but with the vfork() restriction, the parent can call vfork() then exec() multiple times to run multiple copies of another program. It just can't divide itself directly. Yes exec() terminates the calling program and replaces it with the called one. But if you called vfork() first the parent then restarts and you end with the parent and child programs running. uClinux is very much a multi-process system. Philip Nye ----- Original Message ----- From: "Denise Chasse" ... > Apparently, I need to follow vfork() with a call to exec() before the > parent runs again. After I call exec(), is the child terminated? Can I > get back to the child without recreating it with another vfork()? This > is my interpretation from the archived information but it doesn't seem to > make sense. If this is truly the case, how is this effectively any > different than just calling a function from within the parent and just skip > to vfork all together? > > If I can't have a process running in the background, can I have multiple > applications. In addition to my specific application, eventually I would > like to support SNMP, FTP, > DHCP and others. Can these all be done? Any caveats? > > Thanks, > > Newbie-in-training > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From justin.wojdacki at analog.com Thu Mar 14 13:42:15 2002 From: justin.wojdacki at analog.com (Justin Wojdacki) Date: Thu, 14 Mar 2002 10:42:15 -0800 Subject: [uClinux-dev] Applications, processes, and threads... References: <01C1CB40.485EC380.dchasse@sierracom.com> Message-ID: <3C90EF07.883D5B41@analog.com> Denise Chasse wrote: > > Apparently, I need to follow vfork() with a call to exec() before the > parent runs again. After I call exec(), is the child terminated? Can I > get back to the child without recreating it with another vfork()? This > is my interpretation from the archived information but it doesn't seem to > make sense. If this is truly the case, how is this effectively any > different than just calling a function from within the parent and just skip > to vfork all together? > When you vfork() and exec(), you get another schedulable "thing" (task or process depending on your nomenclature). Which one is active at any given moment is up to the scheduler, but on a higher level, you effectively do have multiple processes running at the same time. So when you vfork() and exec() in your parent process, you end up with a child process running in the background. Does this help clear anything up? And what do you intend to mean by "get back to the child"? If you're wondering if the task will continue to run, yes, it will, as long as it doesn't exit(). > > If I can't have a process running in the background, can I have multiple > applications. In addition to my specific application, eventually I would > like to support SNMP, FTP, > DHCP and others. Can these all be done? Any caveats? > Can and already are. :) You just need to build the appropriate applications in the user directory tree to get these kinds of applications. -- ------------------------------------------------- Justin Wojdacki justin.wojdacki at analog.com (408) 350-5032 Communications Processors Group -- Analog Devices This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From siddons at bnl.gov Thu Mar 14 13:33:50 2002 From: siddons at bnl.gov (D. Peter Siddons) Date: Thu, 14 Mar 2002 13:33:50 -0500 Subject: [uClinux-dev] CGI and httpd again, explicit questions. References: Message-ID: <3C90ED0E.1F23503C@bnl.gov> Hi Stefan, I don't know about your board, but for the 68Vz328 the server which I was able to use with CGI was boa. The simple servers do not do so. For me, boa worked fine. Pete. Stefan Jonsson wrote: > > Hello list, > > Yes, hold on now, because here comes some real "newbie" questions ;-) > > It is regarding uClinux on EB40LS (from Lineo) and CGI again. Simply, I > don't get it to work. I did get a suggestion like this from the > uClinux-mail-list: > > "Make shure your web server is running and/or you provide a cgi interface > Check settings in /etc/httpd.conf like your path to cgi-bin" > > *** Explicit-question-1 *** > But how do I make sure it "provides a CGI interface"? > > I made a httpd.conf (because there were none) like this in the suggested > location. > > # Trying to get httpd to exec my CGI- > # -scripts located in /htdocs/cgi-bin/ > # 2002-03-14 Stefan Jonsson > ScriptAlias /cgi-bin/ "/htdocs/cgi-bin/" > > Options ExecCGI # CGI exec permitted > Options Includes # SSI includes allowed > > > Since I really don't know, I wonder if this is "the way". I have found > really nothing about this on uClinux, unfortunately ... and therefore I > bother u ;-) > > *** Explicit-question-2 *** > Does the provided httpd work with cgi and provide that interface or is the > problem there? > > My call for the cgi-script (in C) looks like this: > >
method="GET"> > >
> > *** Explicit-question-3 *** > Is it right so far? (the problem is that it does not execute, just shows the > text "bFLT" (binary flat format?) on the screen ... or text if I try to > execute a shellscript as cgi). > > *** Explicit-question-4 and on ... *** > 4.1 Shall the httpd.conf be located in /etc/? (or shall the config be named > something else and be located elsewhere?) > > 4.2 Should it be a httpd.conf at all in uClinux shipped installation from > Lineo (since there is none such to begin with), or should it maybe be > configurated elsewhere (as in inetd.conf, 'services' or something)? > > 4.3 In services it says: > http stream tcp nowait root /sbin/httpd -i > What does the '-i' stand for? Is it possible to define something here? > > 4.4 Should something special be defined somewhere in the HTML-"code" to get > a browser to understand that there will be CGI? > > So, now my questions are finally to an end ... for this time. > > I thank you for already have been helpful in many ways, but this problem > remains for me. I would be happy for any suggestions and tips/tricks about > this matter. > > Thanks! > > Regards, > Stefan Jonsson, > Student, University of Ume?, Sweden. > > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ -- ----------------- D. Peter Siddons National Synchrotron Light Source, Bldg. 725D Brookhaven National Laboratory Upton New York 11973 Email: siddons at bnl.gov This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From dbyron at coactive.com Thu Mar 14 14:59:00 2002 From: dbyron at coactive.com (David Byron) Date: Thu, 14 Mar 2002 11:59:00 -0800 Subject: [uClinux-dev] elf2flt question Message-ID: <7118259C3044D311942700508B2CA5BB0112D066@balance.coactive.com> I think I've found a problem with elf2flt and how it handles big-endian ARM code. I don't think endianness is relevant, but you never know. I'm trying to get the uClibc setjmp test to run, but the first call from main to setjmp (which a macro turns into _setjmp) ends up going into an infinite loop. Running arm-linux-objdump -d on the gdb file that elf2flt produces gives me: 00001890 <__sigsetjmp>: 1890:?e8806ff0 ?stmia?r0, {r4, r5, r6, r7, r8, r9, sl, fp, sp, lr} 1894:?e2400030 ?sub?r0, r0, #48?; 0x30 1898:?ea000000 ?b?18a0 <__sigjmp_save> 189c:?e1a00000 ?nop???(mov r0,r0) But, the flat file shows e8806ff0 e2400030 eb000000 e1a00000 Note the branch instruction has changed from ea to eb. I believe this turns the b into a bl which is what's causing me grief. When a return statement finally comes, it returns to the wrong place, instead of back to the place that called _setjmp since the link register gets changed with the bl instruction. I do see that some b instructions don't get changed to bl, so I'm not sure what elf2flt is doing here. The following is what I think is relevant from using -v with elf2flt: PLT32 vma=0x1a90, value=0x18a0, address=0x1898 sym_addr=0x18a0 rs=0x2, opcode=0xfeffffea RELOC[40]: offset=1898 symbol=__sigjmp_save+0xffffe762 section=.text size=4 fixup=2 (reloc=0x1898) Other b instructions just seem to have the RELOC entry, but not the PLT32 entry. As an aside, I think the display of the opcode is wrong. Not sure if changing %x to %lx is enough at line 494. I'm using binutils 2.11.2 with no patches and gcc-2_95-branch with Phil Blundell's patch. The patches I've seen from www.codepoet.org/toolchain and http://www.uclinux.org/pub/uClinux/m68k-elf-tools/tools-20020218/ don't seem relevant here. I have another toolchain based binutils 2.11 (but a different elf2flt) that works. I've tried 2.95.3 as well and configured both as arm-elf and still no luck. I'm pretty much in toolchain hell but if any knows why this might be happening, it would be a big help. Thanks much. -DB --- David Byron dbyron at coactive.com Coactive Networks, Inc. http://www.coactive.com 28 Liberty Ship Way voice:(415)289-7800 Sausalito, CA 94965 fax:(415)289-1320 This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From dchasse at sierracom.com Thu Mar 14 10:37:11 2002 From: dchasse at sierracom.com (Denise Chasse) Date: Thu, 14 Mar 2002 15:37:11 -0000 Subject: [uClinux-dev] Applications, processes, and threads... Message-ID: <01C1CB6E.1CA37500.dchasse@sierracom.com> >> Does this help clear anything up? Yes, thanks very much. I think we can put this one to rest. Denise -----Original Message----- From: Justin Wojdacki [SMTP:justin.wojdacki at analog.com] Sent: Thursday, March 14, 2002 6:42 PM To: uclinux-dev at uclinux.org Subject: Re: [uClinux-dev] Applications, processes, and threads... Denise Chasse wrote: > > Apparently, I need to follow vfork() with a call to exec() before the > parent runs again. After I call exec(), is the child terminated? Can I > get back to the child without recreating it with another vfork()? This > is my interpretation from the archived information but it doesn't seem to > make sense. If this is truly the case, how is this effectively any > different than just calling a function from within the parent and just skip > to vfork all together? > When you vfork() and exec(), you get another schedulable "thing" (task or process depending on your nomenclature). Which one is active at any given moment is up to the scheduler, but on a higher level, you effectively do have multiple processes running at the same time. So when you vfork() and exec() in your parent process, you end up with a child process running in the background. Does this help clear anything up? And what do you intend to mean by "get back to the child"? If you're wondering if the task will continue to run, yes, it will, as long as it doesn't exit(). > > If I can't have a process running in the background, can I have multiple > applications. In addition to my specific application, eventually I would > like to support SNMP, FTP, > DHCP and others. Can these all be done? Any caveats? > Can and already are. :) You just need to build the appropriate applications in the user directory tree to get these kinds of applications. -- ------------------------------------------------- Justin Wojdacki justin.wojdacki at analog.com (408) 350-5032 Communications Processors Group -- Analog Devices This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From SMerrifield at tecom.com.au Thu Mar 14 18:07:35 2002 From: SMerrifield at tecom.com.au (Steven Merrifield) Date: Fri, 15 Mar 2002 00:07:35 +0100 Subject: [uClinux-dev] CGI and httpd again, explicit questions. Message-ID: <7FF1185EA0D3D511BC0500B0D0D0C24A1B5898@melexc01.tecom.com.au> Hi Stefan, Confirm that the server you are using does actually support CGI requests. There are several web servers in the uClinux-distribution, not all of them support CGI. > Is it right so far? (the problem is that it does not execute, > just shows the > text "bFLT" (binary flat format?) on the screen ... or text > if I try to > execute a shellscript as cgi). If you execute it from the command line, it should work, ie if your CGI program sends "Hello World" to the web browser, then running ./hellotml.cgi from a shell should do the same thing, but send the output to the console (stdout). In it's simplest form, a CGI executable is no different to any other executable. You could, for example, cp /bin/date /htdocs/cgi-bin/date.cgi then call
to display the current date. You may need to add a mime-type entry to tell the server to execute the CGI program rather than echo it. Something like this perhaps: application/x-httpd-cgi cgi Hope this helps, steve This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From SMerrifield at tecom.com.au Thu Mar 14 18:12:43 2002 From: SMerrifield at tecom.com.au (Steven Merrifield) Date: Fri, 15 Mar 2002 00:12:43 +0100 Subject: [uClinux-dev] Message-ID: <7FF1185EA0D3D511BC0500B0D0D0C24A1B5899@melexc01.tecom.com.au> Hi, pthreads are now supported in uClibc. http://www.uclibc.org/cgi-bin/cvsweb/uClibc/libpthread/ steve > -----Original Message----- > From: Christian Opel [mailto:christian.opel at brooks.com] > Sent: Thursday, 14 March 2002 10:30 > To: Uclinux email Forum > Subject: [uClinux-dev] > > > > Is there already a pthread library for uClinux. > I am trying to port JTC for Motorola coldfire. > > > This message resent by the uclinux-dev at uclinux.org list > server http://www.uClinux.org/ > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From bruce at tele-ip.com Thu Mar 14 18:18:57 2002 From: bruce at tele-ip.com (Bruce Paterson) Date: Fri, 15 Mar 2002 10:18:57 +1100 Subject: [uClinux-dev] Applications, processes, and threads... References: <01C1CB40.485EC380.dchasse@sierracom.com> Message-ID: <3C912FE1.A4495BBB@tele-ip.com> Denise Chasse wrote: > > The goal is to port an application that I wrote for PSOS to uClinux. My > platform is the MCF5272. I'm trying to understand how applications, > processes, and threads all share resources. Without an MMU, I have certain > limitations creating processes, ie vfork instead of fork. I am doing similar. pSOS is of course multi-threaded rather than multi-tasking (static variables are shared). I figure the easiet way to port pSOS code is to actually put most of it in 1 uClinux process, but with many threads representing the original pSOS tasks. In some cases completely seperate processes can be made where there is no need for the sharing of statics. This means threading has to be supported, which means a recent kernel and a re-entrant version of uClibc that is just around the corner. > If I can't have a process running in the background, can I have multiple > applications. In addition to my specific application, eventually I would > like to support SNMP, FTP, > DHCP and others. Can these all be done? Any caveats? There appear to be at least 2 SNMP stacks in the uClinux user area. FTP and DHCP are also fully supported. In my case the "kernel" porting is taking more time than I would have wished (board support) so I haven't got around to porting pSOS application code as yet. [Maily problems with MC68360 ethernet support under the 2.4.x kernels.] -- Cheers, Bruce ------------------------------------------------------------------- This email and any files transmitted with it are confidential and intended solely for the use of the individual or entity to whom they are addressed. If you have received this email in error please notify the system manager. /\\\/\\\/\\\ / / Bruce Paterson / \\\ \\\ \\\ / / Senior Design Engineer / /\\\/\\\/\\\/ / 87 Peters Ave, Mulgrave, Vic, 3170 / / \\\ \\\ \\\ / PO Box 4112, Mulgrave, Vic, 3170, Australia / / \\\/\\\ \\\/ Ph: +61 3 8561 4232 Fax: +61 3 9560 9055 Tele-IP Ltd. Email: bruce at tele-ip.com Icq: #32015991 WWW: http://www.tele-ip.com VK3TJN ------------------------------------------------------------------- This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From superwen at 263.net Thu Mar 14 21:18:08 2002 From: superwen at 263.net (Kevin) Date: Fri, 15 Mar 2002 10:18:8 +0800 Subject: [uClinux-dev] About console Message-ID: <20020315021055.D2A621DE906C2@smtp.263.net> Hi, there, I got a question in porting uClinux 2.4 to MC68VZ328ADS. I can use function console_68328_print() to print out messages in my terminal, but failed with printk(), how could this happen? I might do something wrong in my startup codes, but I don't know where. Best Regards Kevin This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From gerg at snapgear.com Fri Mar 15 02:22:11 2002 From: gerg at snapgear.com (Greg Ungerer) Date: Fri, 15 Mar 2002 17:22:11 +1000 Subject: [uClinux-dev] elf2flt question References: <7118259C3044D311942700508B2CA5BB0112D066@balance.coactive.com> Message-ID: <3C91A123.9040301@snapgear.com> Hi David, Which version of elf2flt are you using? Can you do the same using the latest elf2flt sources from cvs.uclinux.org? Regards Greg David Byron wrote: > I think I've found a problem with elf2flt and how it handles big-endian ARM code. > I don't think endianness is relevant, but you never know. I'm trying to get the > uClibc setjmp test to run, but the first call from main to setjmp (which a macro > turns into _setjmp) ends up going into an infinite loop. > > Running arm-linux-objdump -d on the gdb file that elf2flt produces gives me: > > 00001890 <__sigsetjmp>: > 1890: e8806ff0 stmia r0, {r4, r5, r6, r7, r8, r9, sl, fp, sp, lr} > 1894: e2400030 sub r0, r0, #48 ; 0x30 > 1898: ea000000 b 18a0 <__sigjmp_save> > 189c: e1a00000 nop (mov r0,r0) > > But, the flat file shows > > e8806ff0 > e2400030 > eb000000 > e1a00000 > > Note the branch instruction has changed from ea to eb. I believe this > turns the b into a bl which is what's causing me grief. When a return > statement finally comes, it returns to the wrong place, instead of back > to the place that called _setjmp since the link register gets changed > with the bl instruction. > > I do see that some b instructions don't get changed to bl, so I'm not > sure what elf2flt is doing here. The following is what I think is > relevant from using -v with elf2flt: > > PLT32 vma=0x1a90, value=0x18a0, address=0x1898 sym_addr=0x18a0 rs=0x2, opcode=0xfeffffea > RELOC[40]: offset=1898 symbol=__sigjmp_save+0xffffe762 section=.text size=4 fixup=2 (reloc=0x1898) > > Other b instructions just seem to have the RELOC entry, but not the PLT32 entry. > As an aside, I think the display of the opcode is wrong. Not sure if changing > %x to %lx is enough at line 494. > > I'm using binutils 2.11.2 with no patches and gcc-2_95-branch with Phil Blundell's > patch. The patches I've seen from www.codepoet.org/toolchain and > http://www.uclinux.org/pub/uClinux/m68k-elf-tools/tools-20020218/ don't seem relevant > here. I have another toolchain based binutils 2.11 (but a different elf2flt) that > works. I've tried 2.95.3 as well and configured both as arm-elf and still no luck. > > I'm pretty much in toolchain hell but if any knows why this might be happening, it > would be a big help. > > Thanks much. > > -DB > --- > David Byron dbyron at coactive.com > Coactive Networks, Inc. http://www.coactive.com > 28 Liberty Ship Way voice:(415)289-7800 > Sausalito, CA 94965 fax:(415)289-1320 > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ > > -- ------------------------------------------------------------------------ Greg Ungerer -- Chief Software Wizard EMAIL: gerg at snapgear.com SnapGear PHONE: +61 7 3435 2888 825 Stanley St, FAX: +61 7 3891 3630 Woolloongabba, QLD, 4102, Australia WEB: www.snapgear.com This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From julienb at actimage.fr Fri Mar 15 03:57:15 2002 From: julienb at actimage.fr (Julien Boibessot) Date: Fri, 15 Mar 2002 09:57:15 +0100 Subject: [uClinux-dev] Gdb+bdm+5272C3 References: <7FF1185EA0D3D511BC0500B0D0D0C24A1B5899@melexc01.tecom.com.au> Message-ID: <05a401c1cbff$6784a5d0$8947ec95@bruker.fr> Hi everybody ! I'm trying to use GDB thrue BDM (P&E Adaptator) with the Motorola 5272C3. I have compile and insert BDM driver ( gdb-bdm-20010901 from cybertec) in my host linux and it works (chk /dev/bdmcf0 is running and all tests are ok). I've got GDB 5.1.1 sources, patche it for m68k/coldfire BDM debugging and recompile it to get a running "m68k-bdm-elf-gdb". Then, I launch gdb with: "m68k-bdm-elf-gdb --command=5272.gdb image.elf". 5272.gdb is my init script for GDB and image.elf is my linux image. gdb is running and I can do whatever I want. So I launch the "load" command, take a coffee to wait the end of the download, and "set $pc=0x20000". But then when I try to execute the program by typing command "c", gdb crashes (I get a "segmentation fault"). My board is reseted. Does anybody knows what is happening ??? Thanks ! Julien This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From tomasheran at yahoo.co.uk Fri Mar 15 04:47:46 2002 From: tomasheran at yahoo.co.uk (Tomas Heran) Date: Fri, 15 Mar 2002 10:47:46 +0100 Subject: [uClinux-dev] "Calibrating delay loop..." crashes In-Reply-To: References: Message-ID: <20020315094746.GA7988@obelia.praha.cz.ibm.com> On Wed, Mar 13, 2002 at 09:44:52PM +0000, Jonathan Schmidt wrote: [snip snip] > and I reboot, the kernel hangs at "Calibrating delay loop...". I checked > in the .c file that is responsible for this part of the bootup sequence, > and there doesn't seem to be anything that could cause a crash. The one possible reason for this is that timer is not working properly. Can you verify that? For example by checking whether you're getting timer interrupts? [snip snip] Regards, Tomas This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From okan.vural at eliar.com.tr Fri Mar 15 05:07:04 2002 From: okan.vural at eliar.com.tr (Okan) Date: Fri, 15 Mar 2002 12:07:04 +0200 Subject: [uClinux-dev] Timer interrupt Message-ID: <02031512070400.01737@okan> Hi all, I've a function and I want it to run perodically and asynchronouns my main...Can anyone suggest me an example code about how I can use timer interrupt in user mode ? If I'm not wrong init_timer(),add_timer(),functions dont run in user mode... Thanks anyway.... This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From kadionik at enseirb.fr Fri Mar 15 06:14:42 2002 From: kadionik at enseirb.fr (Patrice Kadionik) Date: Fri, 15 Mar 2002 12:14:42 +0100 Subject: [uClinux-dev] CGI and httpd again, explicit questions. References: <3C90ED0E.1F23503C@bnl.gov> Message-ID: <3C91D7A2.D050B61A@enseirb.fr> hi Stefan, In the uClinux distri, just the boa and thttp server support CGI interface. The http package does'nt support it !!! Pat "D. Peter Siddons" wrote: > Hi Stefan, > I don't know about your board, but for the 68Vz328 the server which I > was able to use with CGI was boa. The simple servers do not do so. For > me, > boa worked fine. > Pete. > > Stefan Jonsson wrote: > > > > Hello list, > > > > Yes, hold on now, because here comes some real "newbie" questions ;-) > > > > It is regarding uClinux on EB40LS (from Lineo) and CGI again. Simply, I > > don't get it to work. I did get a suggestion like this from the > > uClinux-mail-list: > > > > "Make shure your web server is running and/or you provide a cgi interface > > Check settings in /etc/httpd.conf like your path to cgi-bin" > > > > *** Explicit-question-1 *** > > But how do I make sure it "provides a CGI interface"? > > > > I made a httpd.conf (because there were none) like this in the suggested > > location. > > > > # Trying to get httpd to exec my CGI- > > # -scripts located in /htdocs/cgi-bin/ > > # 2002-03-14 Stefan Jonsson > > ScriptAlias /cgi-bin/ "/htdocs/cgi-bin/" > > > > Options ExecCGI # CGI exec permitted > > Options Includes # SSI includes allowed > > > > > > Since I really don't know, I wonder if this is "the way". I have found > > really nothing about this on uClinux, unfortunately ... and therefore I > > bother u ;-) > > > > *** Explicit-question-2 *** > > Does the provided httpd work with cgi and provide that interface or is the > > problem there? > > > > My call for the cgi-script (in C) looks like this: > > > > > method="GET"> > > > > > > > > *** Explicit-question-3 *** > > Is it right so far? (the problem is that it does not execute, just shows the > > text "bFLT" (binary flat format?) on the screen ... or text if I try to > > execute a shellscript as cgi). > > > > *** Explicit-question-4 and on ... *** > > 4.1 Shall the httpd.conf be located in /etc/? (or shall the config be named > > something else and be located elsewhere?) > > > > 4.2 Should it be a httpd.conf at all in uClinux shipped installation from > > Lineo (since there is none such to begin with), or should it maybe be > > configurated elsewhere (as in inetd.conf, 'services' or something)? > > > > 4.3 In services it says: > > http stream tcp nowait root /sbin/httpd -i > > What does the '-i' stand for? Is it possible to define something here? > > > > 4.4 Should something special be defined somewhere in the HTML-"code" to get > > a browser to understand that there will be CGI? > > > > So, now my questions are finally to an end ... for this time. > > > > I thank you for already have been helpful in many ways, but this problem > > remains for me. I would be happy for any suggestions and tips/tricks about > > this matter. > > > > Thanks! > > > > Regards, > > Stefan Jonsson, > > Student, University of Ume?, Sweden. > > > > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ > > -- > ----------------- > D. Peter Siddons > National Synchrotron Light Source, Bldg. 725D > Brookhaven National Laboratory > Upton > New York 11973 > Email: siddons at bnl.gov > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ -- Patrice Kadionik. F6KQH / F4CUQ ----------- +---------------------------------------------------------------------+ +"Tout doit etre aussi simple que possible, pas seulement plus simple"+ +---------------------------------------------------------------------+ + Patrice Kadionik http://www.enseirb.fr/~kadionik + + ENSEIRB fax : +33 5.56.37.20.23 + + PO BOX 99 voice : +33 5.56.84.23.47 + + 33402 TALENCE Cedex http://www.enseirb.fr + + FRANCE mailto:kadionik at enseirb.fr + +---------------------------------------------------------------------+ This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From chri at infis.univ.trieste.it Fri Mar 15 07:11:31 2002 From: chri at infis.univ.trieste.it (Christian Pellegrin) Date: Fri, 15 Mar 2002 13:11:31 +0100 (CET) Subject: [uClinux-dev] Timer interrupt In-Reply-To: <02031512070400.01737@okan> Message-ID: On Fri, 15 Mar 2002, Okan wrote: > Hi all, > I've a function and I want it to run perodically and > asynchronouns my main...Can anyone suggest me an example > code about how I can use timer interrupt in user mode ? If > I'm not wrong init_timer(),add_timer(),functions dont run > in user mode... Hi, it depends on the timer resolution you want. If it's not high try man alarm. Bye! This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From okan.vural at eliar.com.tr Fri Mar 15 07:39:29 2002 From: okan.vural at eliar.com.tr (Okan) Date: Fri, 15 Mar 2002 14:39:29 +0200 Subject: [uClinux-dev] Timer interrupt In-Reply-To: References: Message-ID: <02031514392901.01737@okan> Well,I work on miliseconds time base (alarm works on seconds as I see),and I dont want to use SIGALARM or any other alarm signals (no signal),actually I want a real, time interrupt.Is it possible ??? On Friday 15 March 2002 14:11, Christian Pellegrin wrote: > On Fri, 15 Mar 2002, Okan wrote: > > Hi all, > > I've a function and I want it to run perodically and > > asynchronouns my main...Can anyone suggest me an > > example code about how I can use timer interrupt in > > user mode ? If I'm not wrong > > init_timer(),add_timer(),functions dont run in user > > mode... > > Hi, it depends on the timer resolution you want. If it's > not high try man alarm. > > Bye! This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From chri at infis.univ.trieste.it Fri Mar 15 08:39:42 2002 From: chri at infis.univ.trieste.it (Christian Pellegrin) Date: Fri, 15 Mar 2002 14:39:42 +0100 (CET) Subject: [uClinux-dev] Timer interrupt In-Reply-To: <02031514392901.01737@okan> Message-ID: On Fri, 15 Mar 2002, Okan wrote: > Well,I work on miliseconds time base (alarm works on > seconds as I see),and I dont want to use SIGALARM or any > other alarm signals (no signal),actually I want a real, > time interrupt.Is it possible ??? > I guess your best bet is write a driver for the timer on your board and pass the events to user level via some standard unix mean (signals, select). But usually such short intervals are handled by device drivers or real-time extensions. Bye! This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From gmarselis at patras.atmel.com Fri Mar 15 08:46:57 2002 From: gmarselis at patras.atmel.com (George Marselis) Date: Fri, 15 Mar 2002 15:46:57 +0200 Subject: [uClinux-dev] CGI and httpd again, explicit questions. References: Message-ID: <002d01c1cc27$f1e5e1d0$f5feaa0a@IronLung> [GetYourLearningOnRegister 0x01] Spell Ingredients: * http://www.apache.org/docs Everything you wanted to read about the best httpd server and you were too afraid to ask (for the eyes) * http://cgi-spec.golux.com/ The Common Gateway Interface - RFC Project Page (for the mind) * Marry J Blige F. Method Man "You're all I need to survive" (for the soul ) * Wierd "Al" Yankovitch - "It's all about the Pentiums" ( for the ears ) * http://www.oreilly.com/catalog/apache2/ - Apache The definite Guide - By O'Reilly and associates (for the feel) * 1 glass of strong Greek coffee (for the smell) // Greek coffee is consummed in glasses, not cups. * 1 ergonomic keyboard (for the body) [STARTMEUP, OH, BABY, STARTMEUP] > Hello list, Hey Stephan - use the Force; read those docs. They will give you a better understanding of what CGI is/is not. [snip] > Its regarding uClinux on EB40LS (from Lineo) and CGI again. Simply, I > don't get it to work. I did get a suggestion like this from the > uClinux-mail-list: > "Make shure your web server is running and/or you provide a cgi interface > Check settings in /etc/httpd.conf like your path to cgi-bin" that's a standard unix answer[tm] alright. > *** Explicit-question-1 *** > But how do I make sure it "provides a CGI interface"? Here are simple suggestions. 1) read around the web site you downloaded the webserver from. It should have a page with all the bug^H^H^H^H^H erh, features the server you are looking at. 2) read the README file in the dir you unpacked the tarball or in the ./Documentation dir. 3) read the source and see if there's a CGI header. if you do find one, chances are, there's a CGI implementation >I made a httpd.conf (because there were none) like this in the suggested >location. (gramatical error - was not were) ># Trying to get httpd to exec my CGI- ># -scripts located in /htdocs/cgi-bin/ Bad tactic #1: Never, never, *ever* put the cgi-bin dir under htdocs, unless you want bad things to happen like root access and rm -rf / on your system. cgi-bin should reside under the ServerRoot i.e. if ServerRoot is /web/httpd, then cgi-bin should be /web/httpd/cgi-bin [snip] >Since I really don't know, I wonder if this is "the way". I have found >really nothing about this on uClinux, unfortunately ... and therefore I >bother u ;-) Well, all fine and dandy, but do try to read up/ask around first. It's the UNIX way :) >*** Explicit-question-2 *** >Does the provided httpd work with cgi and provide that interface or is the >problem there? Stephan, what webserver are you using? tiny? boa? simple? please be more specific. boa is a full-featured web server, so, it has a CGI 1.1 implementation. tiny has only a POST CGI capability. so, in essence, you can write some CGI programs, but you can't do everything apache/boa does. simple httpd has no CGI implementation >My call for the cgi-script (in C) looks like this: > >
method="GET"> > >
Uhm. could you please be more specific? that's html, not C . Do you mean, you coded the hellotml.cgi program in C? Is that snippet inside a #include #include int main( int argc, char *argv[]) { printf(" your code here\n"); exit(0);}? Or is it how you called the cgi from your html? >*** Explicit-question-3 *** >Is it right so far? (the problem is that it does not execute, just shows the >text "bFLT" (binary flat format?) on the screen ... or text if I try to >execute a shellscript as cgi). didja remember to chmod 755 the cgi program? the CGI 1.1 standard mentions that httpd has to exec() (or any of its derivatives) to start the >*** Explicit-question-4 and on ... *** >4.1 Shall the httpd.conf be located in /etc/? (or shall the config be named >something else and be located elsewhere?) You should be able to place it anywhere. Look at the makefile/config script. you should be able to manually configure it/pass it as a param to config >4.2 Should it be a httpd.conf at all in uClinux shipped installation from >Lineo (since there is none such to begin with), or should it maybe be >configurated elsewhere (as in inetd.conf, 'services' or something)? That's up the vendor to decide if it will include a configure script. I'm pretty sure they have. why don't you try to use the command find to locate it? If there isn't one, try looking up a manpage/docs the vendor included. they should have some examples included on how to configure the server. >4.3 In services it says: >http stream tcp nowait root /sbin/httpd -i >What does the '-i' stand for? Is it possible to define something here? I can't tell you what -i stands for, since I don't know what server you are running. And, no, /etc/services is not where you define httpd directives. >4.4 Should something special be defined somewhere in the HTML-"code" to get >a browser to understand that there will be CGI? nope CGI is server-side; all the client gets is HTML see http://www.oreilly.com/catalog/cgi/ "CGI prorgaming with C & Perl" from O'Reilly & Associates >So, now my questions are finally to an end ... for this time. Cool, glad to be of help. But do try to read up next time. All this information could be really easilly answered by just reading the CGI standard alone... >Regards, >Stefan Jonsson, >Student, University of Ume?, Sweden. -><- George Marselis WuTang Soldier Spunky Misunderstood Genious http://www.hack.gr/users/isobiths/ This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From dchasse at sierracom.com Fri Mar 15 03:43:49 2002 From: dchasse at sierracom.com (Denise Chasse) Date: Fri, 15 Mar 2002 08:43:49 -0000 Subject: [uClinux-dev] Applications, processes, and threads... Message-ID: <01C1CBFD.87D915C0.dchasse@sierracom.com> I've convinced myself, please correct me if I'm wrong, that the processes created with a vfork closely represent threads. Since the child processes created within a MMU-less environment has access to its parent's variables, I can leverage on this to simulate the behavior of a thread. Thanks, Denise -----Original Message----- From: Bruce Paterson [SMTP:bruce at tele-ip.com] Sent: Thursday, March 14, 2002 11:19 PM To: uclinux-dev at uclinux.org Subject: Re: [uClinux-dev] Applications, processes, and threads... Denise Chasse wrote: > > The goal is to port an application that I wrote for PSOS to uClinux. My > platform is the MCF5272. I'm trying to understand how applications, > processes, and threads all share resources. Without an MMU, I have certain > limitations creating processes, ie vfork instead of fork. I am doing similar. pSOS is of course multi-threaded rather than multi-tasking (static variables are shared). I figure the easiet way to port pSOS code is to actually put most of it in 1 uClinux process, but with many threads representing the original pSOS tasks. In some cases completely seperate processes can be made where there is no need for the sharing of statics. This means threading has to be supported, which means a recent kernel and a re-entrant version of uClibc that is just around the corner. > If I can't have a process running in the background, can I have multiple > applications. In addition to my specific application, eventually I would > like to support SNMP, FTP, > DHCP and others. Can these all be done? Any caveats? There appear to be at least 2 SNMP stacks in the uClinux user area. FTP and DHCP are also fully supported. In my case the "kernel" porting is taking more time than I would have wished (board support) so I haven't got around to porting pSOS application code as yet. [Maily problems with MC68360 ethernet support under the 2.4.x kernels.] -- Cheers, Bruce ------------------------------------------------------------------- This email and any files transmitted with it are confidential and intended solely for the use of the individual or entity to whom they are addressed. If you have received this email in error please notify the system manager. /\\\/\\\/\\\ / / Bruce Paterson / \\\ \\\ \\\ / / Senior Design Engineer / /\\\/\\\/\\\/ / 87 Peters Ave, Mulgrave, Vic, 3170 / / \\\ \\\ \\\ / PO Box 4112, Mulgrave, Vic, 3170, Australia / / \\\/\\\ \\\/ Ph: +61 3 8561 4232 Fax: +61 3 9560 9055 Tele-IP Ltd. Email: bruce at tele-ip.com Icq: #32015991 WWW: http://www.tele-ip.com VK3TJN ------------------------------------------------------------------- This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From JLiu at PacesetterElectronics.com Fri Mar 15 09:22:40 2002 From: JLiu at PacesetterElectronics.com (JINGSONG LIU) Date: Fri, 15 Mar 2002 08:22:40 -0600 Subject: [uClinux-dev] Others cannot access tftp server. Message-ID: Dear sir : Thank you for your help for installing and configing tftpd server in my linux 7.2. After tftpd server is running in oder , I tested tftpd server by issuing a tftp client in the same machine as root user, so I can get and put file from or to tftpd server. but question is that I cannot get or put file from another computer which had a tftp client. I checked IP address of the computer which had a tftpd server and it is correct. Would you please tell me how to deal with this thing ? Thank you very much ! Jason Liu -------------- next part -------------- An HTML attachment was scrubbed... URL: From julienb at actimage.fr Fri Mar 15 09:04:10 2002 From: julienb at actimage.fr (Julien Boibessot) Date: Fri, 15 Mar 2002 15:04:10 +0100 Subject: [uClinux-dev] Timer interrupt References: <02031514392901.01737@okan> Message-ID: <064f01c1cc2a$47dfda30$8947ec95@bruker.fr> Hi ! I was already faced to this problem and that is the solution I've found (if someone can tell me if it is the optimal one.......): You will have to write an interupt handler in the kernel which will catch a custom timer interrupt you have defined, and inside this handler you will have to send a signal (eg SIGUSR1) to your application (but it may take several ms to arrive depending on system load - generally it's fast (hundred of ?s)). But if you don't want to send signals, you can define a variable common to kernel and application (by pointer; (uClinux doesn't handle mmu)). The kernel will modify this variable and application will test it regularly. It's not very clean but....:-)) Hope it will help you.... Julien ----- Original Message ----- From: "Okan" To: "Christian Pellegrin" Cc: Sent: Friday, March 15, 2002 1:39 PM Subject: Re: [uClinux-dev] Timer interrupt > Well,I work on miliseconds time base (alarm works on > seconds as I see),and I dont want to use SIGALARM or any > other alarm signals (no signal),actually I want a real, > time interrupt.Is it possible ??? > > On Friday 15 March 2002 14:11, Christian Pellegrin wrote: > > On Fri, 15 Mar 2002, Okan wrote: > > > Hi all, > > > I've a function and I want it to run perodically and > > > asynchronouns my main...Can anyone suggest me an > > > example code about how I can use timer interrupt in > > > user mode ? If I'm not wrong > > > init_timer(),add_timer(),functions dont run in user > > > mode... > > > > Hi, it depends on the timer resolution you want. If it's > > not high try man alarm. > > > > Bye! > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From chenypg at hotmail.com Fri Mar 15 09:42:57 2002 From: chenypg at hotmail.com (chen yp) Date: Fri, 15 Mar 2002 14:42:57 +0000 Subject: [uClinux-dev] Memory bug of uClinux 2.4.x kernel ? Message-ID: Hi,all Today,I write one program to test the memory managment of uClinux 2.4.x kernel.There is 16M memory in my board which is composed of MC5272 microprocessor. Following is main section of my program: FILE *f = fopen("/proc/meminfo", "r"); while(1){ n++; buff = (unsigned char*)malloc(0x400); memset(buff,0xff,0x400); if((n % 15) == 0){ fseek(f,0,SEEK_SET); for(i=0;i<3;i++) { fgets(buf, 250, f); fputs(buf, stdout); } } } I malloc memory for buff and printf the memory usage info. system crash ,when memory is left 8.4M. Previously I have tested the memory and ensured memory no error. Whether or not it means the kernel has bug somewhere. Any suggestion is appreciated.Thanks a lot. _________________________________________________________________ ???????? Web ?????? ?? MSN Hotmail? http://www.hotmail.com/cn This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From dbyron at coactive.com Fri Mar 15 10:44:18 2002 From: dbyron at coactive.com (David Byron) Date: Fri, 15 Mar 2002 07:44:18 -0800 Subject: [uClinux-dev] elf2flt question Message-ID: <7118259C3044D311942700508B2CA5BB0112D06B@balance.coactive.com> > -----Original Message----- > From: Greg Ungerer [mailto:gerg at snapgear.com] > Sent: Thursday, March 14, 2002 11:22 PM > > Which version of elf2flt are you using? > Can you do the same using the latest elf2flt sources from > cvs.uclinux.org? I'm using the latest sources from cvs.uclinux.org. The only mod I've made is to configure an explicit path to look for the linker script. -DB --- David Byron dbyron at coactive.com Coactive Networks, Inc. http://www.coactive.com 28 Liberty Ship Way voice:(415)289-7800 Sausalito, CA 94965 fax:(415)289-1320 This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From chri at infis.univ.trieste.it Fri Mar 15 11:07:57 2002 From: chri at infis.univ.trieste.it (Christian Pellegrin) Date: Fri, 15 Mar 2002 17:07:57 +0100 (CET) Subject: [uClinux-dev] Timer interrupt In-Reply-To: <064f01c1cc2a$47dfda30$8947ec95@bruker.fr> Message-ID: On Fri, 15 Mar 2002, Julien Boibessot wrote: > But if you don't want to send signals, you can define a variable common to > kernel and application (by pointer; (uClinux doesn't handle mmu)). The > kernel will modify this variable and application will test it regularly. > It's not very clean but....:-)) > Well, if you want to play even dirtier you can pass a pointer to a function to the kernel and a pointer to some data you need to modify from you "user level interrupt routine". Then the int handler calls the function with the pointer to the data as argument. Just don't call library functions and don't access data except via the argument passed. This should work since there is not MMU. I tried on an Atmel AT91M55800 with < 50 microsecond worse-case latency by fine-tuning the AIC. But anyway a RT extensions will do the work in a much much more elegant way :-)). Bye! This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From dbyron at coactive.com Fri Mar 15 13:01:48 2002 From: dbyron at coactive.com (David Byron) Date: Fri, 15 Mar 2002 10:01:48 -0800 Subject: [uClinux-dev] elf2flt question Message-ID: <7118259C3044D311942700508B2CA5BB0112D06C@balance.coactive.com> There's no break in this case of a switch statement in elf2flt.c. Is that intentional? ????case R_ARM_PLT32: ?????if (verbose) ??????fprintf(stderr, ???????"%s vma=0x%x, value=0x%x, address=0x%x " ???????"sym_addr=0x%x rs=0x%x, opcode=0x%lx\n", ???????"PLT32", ???????sym_vma, (*(q->sym_ptr_ptr))->value, ???????q->address, sym_addr, ???????(*p)->howto->rightshift, ???????*((unsigned long *) (sectionp + q->address))); /* NO BREAK HERE */ ????case R_ARM_PC24: ?????sym_vma = 0; ?????sym_addr = (sym_addr-q->address)>>(*p)->howto->rightshift; ?????break; I'll dive into elf2flt to see where/why the branch instruction is changing to a branch w/ link. -DB > -----Original Message----- > From: David Byron > Sent: Friday, March 15, 2002 7:44 AM > To: 'uclinux-dev at uclinux.org' > Subject: RE: [uClinux-dev] elf2flt question > > > -----Original Message----- > > From: Greg Ungerer [mailto:gerg at snapgear.com] > > Sent: Thursday, March 14, 2002 11:22 PM > > > > Which version of elf2flt are you using? > > Can you do the same using the latest elf2flt sources from > > cvs.uclinux.org? > > I'm using the latest sources from cvs.uclinux.org. The only mod > I've made is to configure an explicit path to look for the linker > script. This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From tgriggs at keyww.com Fri Mar 15 13:21:12 2002 From: tgriggs at keyww.com (Travis Griggs) Date: Fri, 15 Mar 2002 10:21:12 -0800 Subject: [uClinux-dev] C++ space issues Message-ID: <3C923B98.4040401@keyww.com> We're doing uClinux coldfire development. We've got a lot of good help from people in this group. We have been trying to develop our little control apps that run on the coldfire with C++. We use the STL so that we can get at things like vectors and strings and such. What we are finding though is that we run out of size *very* quickly. Doing things the "C++ way" just seems to be chewing up tons of memory, both in program file size, as well as executable size. A simple "tell me a number and I'll print it back to you program in C" takes 47072 bytes process space. Doing the same with cin, cout, << and >>, uses 169952. Almost 4X. It's not taking much of a program before we run out of space to even load and start the program. Is anyone else writing C++ apps of any size for uClinux/coldfire? And are you having success. We're hoping there's some switch one can flick and the compiler dramatically reduces the code usage. BTW: when one does a cat /proc//status, how exactly does one interpret the "slack" entry? -- Travis Griggs Key Technology 10 2 letter words: "If it is to be it is up to me" This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From hamilton at SEDSystems.ca Fri Mar 15 14:02:42 2002 From: hamilton at SEDSystems.ca (Kendrick Hamilton) Date: Fri, 15 Mar 2002 13:02:42 -0600 (CST) Subject: [uClinux-dev] C++ space issues In-Reply-To: <3C923B98.4040401@keyww.com> Message-ID: Travis, We are using uClinux and C++ on a 68360. We don't use 'STL'. We also don't use cin and cout. We avoided both of them because of size. We also avoid STL so we have better control of memory allocation. On Fri, 15 Mar 2002, Travis Griggs wrote: > We're doing uClinux coldfire development. We've got a lot of good help > from people in this group. > > We have been trying to develop our little control apps that run on the > coldfire with C++. We use the STL so that we can get at things like > vectors and strings and such. > > What we are finding though is that we run out of size *very* quickly. > Doing things the "C++ way" just seems to be chewing up tons of memory, > both in program file size, as well as executable size. A simple "tell me > a number and I'll print it back to you program in C" takes 47072 bytes > process space. Doing the same with cin, cout, << and >>, uses 169952. > Almost 4X. It's not taking much of a program before we run out of space > to even load and start the program. > > Is anyone else writing C++ apps of any size for uClinux/coldfire? And > are you having success. We're hoping there's some switch one can flick > and the compiler dramatically reduces the code usage. > > BTW: when one does a cat /proc//status, how exactly does one > interpret the "slack" entry? > > -- > Travis Griggs > Key Technology > 10 2 letter words: "If it is to be it is up to me" > > > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From dbyron at coactive.com Fri Mar 15 14:46:18 2002 From: dbyron at coactive.com (David Byron) Date: Fri, 15 Mar 2002 11:46:18 -0800 Subject: [uClinux-dev] elf2flt question Message-ID: <7118259C3044D311942700508B2CA5BB0112D071@balance.coactive.com> I can't say I understand why, but the following patch seems to fix things for me. Hopefully someone who knows more about ARM relocations can say if this makes sense. At least the setjmp test passes now. I'll let you know if the rest of everything still works -- boot code, kernel, init, shell, busybox, etc. -DB Index: elf2flt.c =================================================================== RCS file: /prj/salem/cvs/elf2flt/elf2flt.c,v retrieving revision 1.1.1.2 diff -u -b -B -w -p -u -r1.1.1.2 elf2flt.c --- elf2flt.c?2002/02/19 19:00:31?1.1.1.2 +++ elf2flt.c?2002/03/15 19:45:25 @@ -598,7 +598,8 @@ dump_symbols(symbols, number_of_symbols) ????tmp.l = *((unsigned long *) (sectionp + q->address)); ????hl = tmp.c[i0] | (tmp.c[i1] << 8) | (tmp.c[i2] << 16); -????if ((*p)->howto->type != R_ARM_PC24) +????if (((*p)->howto->type != R_ARM_PC24) && +???? ((*p)->howto->type != R_ARM_PLT32)) ?????hl |= (tmp.c[i3] << 24); ????else if (tmp.c[i2] & 0x80) ?????hl |= 0xff000000; /* sign extend */ @@ -606,7 +607,8 @@ dump_symbols(symbols, number_of_symbols) ????tmp.c[i0] = hl & 0xff; ????tmp.c[i1] = (hl >> 8) & 0xff; ????tmp.c[i2] = (hl >> 16) & 0xff; -????if ((*p)->howto->type != R_ARM_PC24) +????if (((*p)->howto->type != R_ARM_PC24) && +???? ((*p)->howto->type != R_ARM_PLT32)) ?????tmp.c[i3] = (hl >> 24) & 0xff; ????if ((*p)->howto->type == R_ARM_ABS32) ?????*((unsigned long *) (sectionp + q->address)) = htonl(hl); > -----Original Message----- > From: David Byron > Sent: Thursday, March 14, 2002 11:59 AM > To: 'uclinux-dev at uclinux.org' > Subject: [uClinux-dev] elf2flt question > > I think I've found a problem with elf2flt and how it handles > big-endian ARM code. I don't think endianness is relevant, > but you never know. I'm trying to get the uClibc setjmp > test to run, but the first call from main to setjmp (which a > macro turns into _setjmp) ends up going into an infinite > loop. > > Running arm-linux-objdump -d on the gdb file that elf2flt > produces gives me: > > 00001890 <__sigsetjmp>: > 1890:?e8806ff0 ?stmia?r0, {r4, r5, r6, r7, r8, r9, sl, fp, sp, lr} > 1894:?e2400030 ?sub?r0, r0, #48?; 0x30 > 1898:?ea000000 ?b?18a0 <__sigjmp_save> > 189c:?e1a00000 ?nop???(mov r0,r0) > > But, the flat file shows > > e8806ff0 > e2400030 > eb000000 > e1a00000 > > Note the branch instruction has changed from ea to eb. I > believe this turns the b into a bl which is what's causing > me grief. When a return statement finally comes, it returns > to the wrong place, instead of back to the place that called > _setjmp since the link register gets changed with the bl > instruction. > > I do see that some b instructions don't get changed to bl, so I'm not > sure what elf2flt is doing here. The following is what I think is > relevant from using -v with elf2flt: > > PLT32 vma=0x1a90, value=0x18a0, address=0x1898 > sym_addr=0x18a0 rs=0x2, opcode=0xfeffffea > RELOC[40]: offset=1898 symbol=__sigjmp_save+0xffffe762 > section=.text size=4 fixup=2 (reloc=0x1898) > > Other b instructions just seem to have the RELOC entry, but > not the PLT32 entry. As an aside, I think the display of > the opcode is wrong. Not sure if changing %x to %lx is > enough at line 494. > > I'm using binutils 2.11.2 with no patches and gcc-2_95-branch > with Phil Blundell's patch. The patches I've seen from > www.codepoet.org/toolchain and > http://www.uclinux.org/pub/uClinux/m68k-elf-tools/tools-20020218/ > don't seem relevant here. I have another toolchain based binutils 2.11 > (but a different elf2flt) that works. I've tried 2.95.3 as well and > configured both as arm-elf and still no luck. This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From gmenie at akamai.com Sun Mar 17 05:46:09 2002 From: gmenie at akamai.com (Menie, Georges) Date: Sun, 17 Mar 2002 05:46:09 -0500 Subject: [uClinux-dev] MC68VZ328.h again Message-ID: here is a patch for the linux-2.4.x/include/asm-m68knommu/MC68VZ328.h file (I did not check if it can be applied to linux/include..., but it should). it will add the Port K and Port J definitions. Georges -------------- next part -------------- An embedded and charset-unspecified text was scrubbed... Name: patch-vz328.txt URL: From gmenie at akamai.com Sun Mar 17 05:40:15 2002 From: gmenie at akamai.com (Menie, Georges) Date: Sun, 17 Mar 2002 05:40:15 -0500 Subject: [uClinux-dev] Debugging on the 68VZ328 using 2.4.10 and elf Message-ID: Here is a patch for the 68328/68EZ328/68VZ328 arch which will allow gdb and gdbserver to run together. It adds the vector number on the stack like does a 68010 (format=0). It slow down a little bit the exception handling, so I did add a config switch to enable/disable it : "Simulate vectorized exceptions (debugging user apps on 68000)". I did test it on the DragonEngine port, but the following port should be checked : 68328 68328/pilot 68EZ328 68VZ328/ucdimm The patch allow for a normal gdb/gdbserver session with breakpoints but displaying variables shows sometimes wrong data, I have to check this but I don't think it related to the exception processing... The patch is against the latest (2002-03-16) CVS tree for linux-2.4.x Cheers, Georges -------------- next part -------------- An embedded and charset-unspecified text was scrubbed... Name: patch-gdb.txt URL: From juergen.mueller at muellerhydraulik.de Sun Mar 17 13:27:51 2002 From: juergen.mueller at muellerhydraulik.de (Juergen Mueller) Date: Sun, 17 Mar 2002 19:27:51 +0100 Subject: [uClinux-dev] Flash Write In-Reply-To: Message-ID: Hi!! Is it possible to write into the flash at runtime. Is there something like a flash filesystem which allows to write into the flash. I'm using uCdimm MC68Vz328 and Kernel 2.4.x. Any suggestion is appreciated. Thanks! Juergen This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From danielp at cse.unsw.edu.au Sun Mar 17 16:50:28 2002 From: danielp at cse.unsw.edu.au (Daniel Potts) Date: Mon, 18 Mar 2002 08:50:28 +1100 (EST) Subject: [uClinux-dev] 68*328 power management In-Reply-To: Message-ID: Hi, I've been trying to implement some power management on the ucdimm, but I'm not having much luck when I turn the clock off (PLLCR |= PLLCR_DISPLL). When I do this, it appears that memory gets corrupted when you leave it off for more than 10 seconds.. if you leave it off for over 30 seconds it won't come back at all. I'm guessing this has something do to with the refreshing of the SDRAM that isn't quite right. I've tried a whole heap of different settings for the DRAM but it hasn't changed the result at all. Perhaps it is some other peripheral device causing the problem, I'm not sure. Doze mode works without any issues, as does sleep if you only sleep for a few seconds :) Main code is attached. I've also modified the serial driver, cs89x00 driver and so on to support the pm routines. Ideas? Thanks, Daniel -------------- next part -------------- /* * 68328 Power Management Routines * * Copyright (C) 2002 Daniel Potts * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA */ /* * Debug macros */ #define DEBUG 1 #ifdef DEBUG # define DPRINTK(fmt, args...) printk("%s: " fmt, __FUNCTION__ , ## args) #else # define DPRINTK(fmt, args...) #endif #include #include #include #include #include #include /* WARNING: included for CTL_ACPI and ACPI_S1_SLP_TYP */ #include #include /* for interrupt stuff (that shouldn't be here)*/ #include unsigned long sleep_irq_mask = 0L; /* irqs to enable while in sleep mode */ #ifdef CONFIG_68328_SERIAL void shutdown_console(void); void startup_console(void); #endif void set_sleep_mask(unsigned long mask) { sleep_irq_mask = mask; } int cpu_suspend(void) { unsigned long flags; unsigned long imr_flags; /* ZZZZzzzz */ save_flags(flags); cli(); imr_flags = IMR; #ifdef CONFIG_68328_SERIAL /* Console is shutdown here as a special device, * to ensure that the kernel will not hang trying to write to console. */ shutdown_console(); #endif /* write out sleep enabled interrupts */ IMR = ~(sleep_irq_mask); #if 1 /* Stop clock - DOZE mode */ PCTRL = PCTRL_PCEN; #else PLLCR |= PLLCR_DISPLL; #endif __asm__("stop #0x2000" : : : "cc"); #ifdef CONFIG_68328_SERIAL startup_console(); #endif IMR = imr_flags; restore_flags(flags); return 0; } static int sysctl_pm_do_suspend(ctl_table *table, int write, struct file *filp, void *buffer, size_t *lenp) { int r = 0; *lenp = 0; DPRINTK("Suspending..\n"); r = pm_send_all(PM_SUSPEND, (void *)2); if (r) { return r; } r = cpu_suspend(); pm_send_all(PM_RESUME, (void *)0); DPRINTK("Resumed..\n"); return r; } static struct ctl_table pm_table[] = { {ACPI_S1_SLP_TYP, "suspend", NULL, 0, 0600, NULL, (proc_handler *)&sysctl_pm_do_suspend}, {0} }; static struct ctl_table pm_dir_table[] = { {CTL_ACPI, "pm", NULL, 0, 0555, pm_table}, {0} }; void pm_irq(int irq, void *dev_id, struct pt_regs *regs) { { unsigned long flags; save_flags(flags); cli(); ISR |= ISR_IRQ3; /* ack edge interrupt */ PDSEL |= 0x01; PDDIR |= 0x01; PDDATA ^= 0x01; restore_flags(flags); } } /* * Initialize power interface */ static int __init pm_init(void) { printk("Power management driver for 68328, Daniel Potts\n"); /* register a "power button" that we wait on when sleeping */ power_button(); register_sysctl_table(pm_dir_table, 1); return 0; } __initcall(pm_init); /* power_button(): BELONGS IN ANOTHER FILE: */ void power_button() { int err; unsigned long flags; /* init interrupt section */ PDDIR &= ~PD_IRQ3; PDSEL &= ~PD_IRQ3; ICR |= ( ICR_ET3); /* edge sensitive */ ISR |= ISR_IRQ3; /* clear it */ set_sleep_mask(IMR_MIRQ3); err = request_irq(IRQ3_IRQ_NUM, pm_irq, IRQ_FLG_STD, "power button", NULL); if(err) printk("power irq failed\n"); } From mark_mcchrystal at yahoo.com Sun Mar 17 18:40:58 2002 From: mark_mcchrystal at yahoo.com (Mark McChrystal) Date: Sun, 17 Mar 2002 15:40:58 -0800 (PST) Subject: [uClinux-dev] 68*328 power management In-Reply-To: Message-ID: <20020317234058.40291.qmail@web14303.mail.yahoo.com> Can you send a patch for everything? It sounds like the SDRAM isn't being refreshed. Is the SDRAM self refresh? If you have a scope look at the RAS and CAS lines while it's in sleep. If you don't have a scope, I'll try your code on a ucsimm board and let you know what I see. Mark __________________________________________________ Do You Yahoo!? Yahoo! Sports - live college hoops coverage http://sports.yahoo.com/ This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From gerg at snapgear.com Sun Mar 17 18:49:20 2002 From: gerg at snapgear.com (Greg Ungerer) Date: Mon, 18 Mar 2002 09:49:20 +1000 Subject: [uClinux-dev] elf2flt question References: <7118259C3044D311942700508B2CA5BB0112D071@balance.coactive.com> Message-ID: <3C952B80.2040601@snapgear.com> Hi David, If it looks good, I will commit this change. Regards Greg David Byron wrote: > I can't say I understand why, but the following patch seems to fix things for me. > Hopefully someone who knows more about ARM relocations can say if this makes sense. > > At least the setjmp test passes now. I'll let you know if the rest of everything > still works -- boot code, kernel, init, shell, busybox, etc. > > -DB > > Index: elf2flt.c > =================================================================== > RCS file: /prj/salem/cvs/elf2flt/elf2flt.c,v > retrieving revision 1.1.1.2 > diff -u -b -B -w -p -u -r1.1.1.2 elf2flt.c > --- elf2flt.c 2002/02/19 19:00:31 1.1.1.2 > +++ elf2flt.c 2002/03/15 19:45:25 > @@ -598,7 +598,8 @@ dump_symbols(symbols, number_of_symbols) > > tmp.l = *((unsigned long *) (sectionp + q->address)); > hl = tmp.c[i0] | (tmp.c[i1] << 8) | (tmp.c[i2] << 16); > - if ((*p)->howto->type != R_ARM_PC24) > + if (((*p)->howto->type != R_ARM_PC24) && > + ((*p)->howto->type != R_ARM_PLT32)) > hl |= (tmp.c[i3] << 24); > else if (tmp.c[i2] & 0x80) > hl |= 0xff000000; /* sign extend */ > @@ -606,7 +607,8 @@ dump_symbols(symbols, number_of_symbols) > tmp.c[i0] = hl & 0xff; > tmp.c[i1] = (hl >> 8) & 0xff; > tmp.c[i2] = (hl >> 16) & 0xff; > - if ((*p)->howto->type != R_ARM_PC24) > + if (((*p)->howto->type != R_ARM_PC24) && > + ((*p)->howto->type != R_ARM_PLT32)) > tmp.c[i3] = (hl >> 24) & 0xff; > if ((*p)->howto->type == R_ARM_ABS32) > *((unsigned long *) (sectionp + q->address)) = htonl(hl); > ------------------------------------------------------------------------ Greg Ungerer -- Chief Software Wizard EMAIL: gerg at snapgear.com SnapGear PHONE: +61 7 3435 2888 825 Stanley St, FAX: +61 7 3891 3630 Woolloongabba, QLD, 4102, Australia WEB: www.snapgear.com This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From gerg at snapgear.com Sun Mar 17 18:53:06 2002 From: gerg at snapgear.com (Greg Ungerer) Date: Mon, 18 Mar 2002 09:53:06 +1000 Subject: [uClinux-dev] Gdb+bdm+5272C3 References: <7FF1185EA0D3D511BC0500B0D0D0C24A1B5899@melexc01.tecom.com.au> <05a401c1cbff$6784a5d0$8947ec95@bruker.fr> Message-ID: <3C952C62.2080506@snapgear.com> Hi Julien, Julien Boibessot wrote: > I'm trying to use GDB thrue BDM (P&E Adaptator) with the Motorola 5272C3. > > I have compile and insert BDM driver ( gdb-bdm-20010901 from cybertec) in my > host linux and it works (chk /dev/bdmcf0 is running and all tests are ok). > I've got GDB 5.1.1 sources, patche it for m68k/coldfire BDM debugging and > recompile it to get a running "m68k-bdm-elf-gdb". > > Then, I launch gdb with: "m68k-bdm-elf-gdb --command=5272.gdb image.elf". > 5272.gdb is my init script for GDB and image.elf is my linux image. > > gdb is running and I can do whatever I want. So I launch the "load" command, > take a coffee to wait the end of the download, and "set $pc=0x20000". But > then when I try to execute the program by typing command "c", gdb crashes (I > get a "segmentation fault"). My board is reseted. > > Does anybody knows what is happening ??? Does you gdbinit script initialize RAM/SDRAM? If it doesn't then your load cannot possibly work, since it will be trying to load the binary into RAM. Regards Greg ------------------------------------------------------------------------ Greg Ungerer -- Chief Software Wizard EMAIL: gerg at snapgear.com SnapGear PHONE: +61 7 3435 2888 825 Stanley St, FAX: +61 7 3891 3630 Woolloongabba, QLD, 4102, Australia WEB: www.snapgear.com This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From gerg at snapgear.com Sun Mar 17 19:01:00 2002 From: gerg at snapgear.com (Greg Ungerer) Date: Mon, 18 Mar 2002 10:01:00 +1000 Subject: [uClinux-dev] Applications, processes, and threads... References: <01C1CBFD.87D915C0.dchasse@sierracom.com> Message-ID: <3C952E3C.9050201@snapgear.com> Hi Denise, Denise Chasse wrote: > I've convinced myself, please correct me if I'm wrong, that the processes > created with a vfork closely represent threads. Since the child processes > created within a MMU-less environment has access to its parent's variables, > I can leverage on this to simulate the behavior of a thread. Couple of problems: 1. both parent and child are using the same stack 2. the parent is put to sleep (after the vfork) until the child either exec()s or exit()s. I expect you could work around number 1 (by say mallocing some memory and pointing your child stack into it - dirty, but I think it could be made to work). Number 2 is a problem if you want both threads to be runing :-) Regards Greg ------------------------------------------------------------------------ Greg Ungerer -- Chief Software Wizard EMAIL: gerg at snapgear.com SnapGear PHONE: +61 7 3435 2888 825 Stanley St, FAX: +61 7 3891 3630 Woolloongabba, QLD, 4102, Australia WEB: www.snapgear.com This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From parksik at nuri.keti.re.kr Sun Mar 17 19:39:35 2002 From: parksik at nuri.keti.re.kr (=?ks_c_5601-1987?B?udq6zr3E?=) Date: Mon, 18 Mar 2002 09:39:35 +0900 Subject: [uClinux-dev] [Q] First Image... Message-ID: <001c01c1ce15$60da7850$7980fdcb@BOO> Hi~ :) I'm a freshman in the field of embedded linux... ^^* Help me~~! ^^; I bought "Lineo uClinux ARM7TDMI Board Development Kit". Using the rpm files in the CD, I installed the packages such as gcc cross compiler and so on. And I made my work directory, did 'buildenv', 'make' and 'sh deftemplate.sh'. I found the 'image.bin' file. I modified the 'romdisk/etc/rc' file so that I set my network address and I did 'make' again. I found the new 'image.bin' file too. I downloaded new image file into the ARM7TDMI Borad using 'minicom' and push the reset button but nothing except 'goram!' or 'go!' message appear in debugging console. How the image.bin file do like the Linux doing 'Login', 'response to ping request', etc..~? Pu-Sik Park System IC Research Center Korea Electronics Technology Institute(KETI) 451-865 MaSan-Ri, JinWi-Myon, PyungTaek-Si, KyungGi-Do 451-865, Korea Tel: +82-31-6104-394 M.P: 017-317-9493 Fax: +82-31-6104-048 E-mail: parksik at nuri.keti.re.kr This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From wenfeng_liu at ccermail.net Sun Mar 17 20:59:17 2002 From: wenfeng_liu at ccermail.net (Wenfeng Liu) Date: Mon, 18 Mar 2002 9:59:17 +0800 Subject: [uClinux-dev] Ask for help of uCGardener Board Message-ID: <200203180157.g2I1vAk31260@clinux.org> Hi, all I have got a uCGardener board. But how can I debug it after all of the components are sealed together. And could anyone tell me how to put a LCD on this board? Does anyone has some useful links or documents? Thank you very much for help me. Kent W. Liu wenfeng_liu@ This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From gerg at snapgear.com Sun Mar 17 22:33:52 2002 From: gerg at snapgear.com (Greg Ungerer) Date: Mon, 18 Mar 2002 13:33:52 +1000 Subject: [uClinux-dev] linux-2.5.6uc0 Message-ID: <3C956020.2060908@snapgear.com> Hey All, I have put a linux-2.5.6 uClinux patch set up on www.uclinux.org. http://www.uclinux.org/pub/uClinux/uClinux-2.5.6/linux-2.5.6uc0.patch.gz There is still a memory de-allocation problem I need to track down, but it basically works... Regards Greg ------------------------------------------------------------------------ Greg Ungerer -- Chief Software Wizard EMAIL: gerg at snapgear.com SnapGear PHONE: +61 7 3435 2888 825 Stanley St, FAX: +61 7 3891 3630 Woolloongabba, QLD, 4102, Australia WEB: www.snapgear.com This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From danielp at cse.unsw.edu.au Sun Mar 17 23:20:10 2002 From: danielp at cse.unsw.edu.au (Daniel Potts) Date: Mon, 18 Mar 2002 15:20:10 +1100 (EST) Subject: [uClinux-dev] 68*328 power management In-Reply-To: <20020317234058.40291.qmail@web14303.mail.yahoo.com> Message-ID: I believe this is a patch for everything that I've changed. Sorry for the messy state of the code.. it will be cleaned up when I'm happy with it. You need to enable SYSCTL and PM options when building the kernel. Also, if you compile in cs89x0 support, please double check where the "/sleep" wire is connected to, and which port it is. I believe the current cvs code is wrong (at least for ucdimm it is) and I've fixed it up in this patch. The SDRAM part is a micron.com part. I don't remember off the top of my head exactly what the part number is, however I believe it supports self-refresh. The earliest I'll be able to use a scope is next weekend, so I'd appreciate it if you can have a look for me. This current code does not enable the DRAM self-refresh bit or any DRAM timing, etc changes.. I took out all this code when it didn't appear to improve the situation. Cheers, Daniel On Sun, 17 Mar 2002, Mark McChrystal wrote: > Can you send a patch for everything? It sounds like > the SDRAM isn't being refreshed. Is the SDRAM self > refresh? If you have a scope look at the RAS and CAS > lines while it's in sleep. If you don't have a scope, > I'll try your code on a ucsimm board and let you know > what I see. -------------- next part -------------- diff --exclude=.* --exclude=*~ -Naur uClinux-2.4.x.orig/arch/m68knommu/platform/68VZ328/Makefile linux-2.4.x/arch/m68knommu/platform/68VZ328/Makefile --- uClinux-2.4.x.orig/arch/m68knommu/platform/68VZ328/Makefile Sun Mar 3 12:33:33 2002 +++ linux-2.4.x/arch/m68knommu/platform/68VZ328/Makefile Thu Mar 7 00:31:56 2002 @@ -17,6 +17,8 @@ O_TARGET := platform.o obj-y := entry.o config.o signal.o traps.o ints.o +obj-$(CONFIG_PM) += pm.o + $(BOARD)/crt0_$(MODEL).o: $(BOARD)/crt0_$(MODEL).S $(BOARD)/crt0_fixed.S $(BOARD)/bootlogo.rh entry.o: entry.S m68k_defs.h diff --exclude=.* --exclude=*~ -Naur uClinux-2.4.x.orig/arch/m68knommu/platform/68VZ328/pm.c linux-2.4.x/arch/m68knommu/platform/68VZ328/pm.c --- uClinux-2.4.x.orig/arch/m68knommu/platform/68VZ328/pm.c Thu Jan 1 10:00:00 1970 +++ linux-2.4.x/arch/m68knommu/platform/68VZ328/pm.c Mon Mar 18 08:37:07 2002 @@ -0,0 +1,191 @@ +/* + * 68328 Power Management Routines + * + * Copyright (C) 2002 Daniel Potts + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + */ + +/* + * Debug macros + */ +#define DEBUG 1 +#ifdef DEBUG +# define DPRINTK(fmt, args...) printk("%s: " fmt, __FUNCTION__ , ## args) +#else +# define DPRINTK(fmt, args...) +#endif + + +#include +#include +#include +#include +#include + +#include /* WARNING: included for CTL_ACPI and ACPI_S1_SLP_TYP */ + +#include + +#include /* for interrupt stuff (that shouldn't be here)*/ +#include + +unsigned long sleep_irq_mask = 0L; /* irqs to enable while in sleep mode */ + +#ifdef CONFIG_68328_SERIAL +void shutdown_console(void); +void startup_console(void); +#endif + +void set_sleep_mask(unsigned long mask) +{ + sleep_irq_mask = mask; +} + +int cpu_suspend(void) +{ + unsigned long flags; + unsigned long imr_flags; + + /* ZZZZzzzz */ + save_flags(flags); + cli(); + imr_flags = IMR; + +#ifdef CONFIG_68328_SERIAL + /* Console is shutdown here as a special device, + * to ensure that the kernel will not hang trying to write to console. + */ + shutdown_console(); +#endif + + /* write out sleep enabled interrupts */ + IMR = ~(sleep_irq_mask); + +#if 1 + /* Stop clock - DOZE mode */ + PCTRL = PCTRL_PCEN; +#else + PLLCR |= PLLCR_DISPLL; +#endif + + __asm__("stop #0x2000" : : : "cc"); + +#ifdef CONFIG_68328_SERIAL + startup_console(); +#endif + + IMR = imr_flags; + restore_flags(flags); + + return 0; +} + +static int sysctl_pm_do_suspend(ctl_table *table, int write, struct file *filp, + void *buffer, size_t *lenp) +{ + int r = 0; + + *lenp = 0; + + DPRINTK("Suspending..\n"); + + r = pm_send_all(PM_SUSPEND, (void *)2); + if (r) { + return r; + } + + r = cpu_suspend(); + + pm_send_all(PM_RESUME, (void *)0); + + DPRINTK("Resumed..\n"); + + return r; +} + + +static struct ctl_table pm_table[] = +{ + {ACPI_S1_SLP_TYP, "suspend", NULL, 0, 0600, NULL, (proc_handler *)&sysctl_pm_do_suspend}, + {0} +}; + +static struct ctl_table pm_dir_table[] = +{ + {CTL_ACPI, "pm", NULL, 0, 0555, pm_table}, + {0} +}; + +void pm_irq(int irq, void *dev_id, struct pt_regs *regs) +{ + + + { + unsigned long flags; + + save_flags(flags); + cli(); + + ISR |= ISR_IRQ3; /* ack edge interrupt */ + + PDSEL |= 0x01; + PDDIR |= 0x01; + PDDATA ^= 0x01; + + restore_flags(flags); + } +} + +/* + * Initialize power interface + */ +static int __init pm_init(void) +{ + + printk("Power management driver for 68328, Daniel Potts\n"); + + /* register a "power button" that we wait on when sleeping */ + power_button(); + + register_sysctl_table(pm_dir_table, 1); + return 0; +} + +__initcall(pm_init); + + +/* power_button(): BELONGS IN ANOTHER FILE: */ +void power_button() +{ + int err; + unsigned long flags; + + + /* init interrupt section */ + PDDIR &= ~PD_IRQ3; + PDSEL &= ~PD_IRQ3; + + ICR |= ( ICR_ET3); /* edge sensitive */ + ISR |= ISR_IRQ3; /* clear it */ + + set_sleep_mask(IMR_MIRQ3); + + err = request_irq(IRQ3_IRQ_NUM, pm_irq, + IRQ_FLG_STD, "power button", NULL); + + if(err) + printk("power irq failed\n"); +} diff --exclude=.* --exclude=*~ -Naur uClinux-2.4.x.orig/arch/m68knommu/platform/68VZ328/ucdimm/ints.c linux-2.4.x/arch/m68knommu/platform/68VZ328/ucdimm/ints.c --- uClinux-2.4.x.orig/arch/m68knommu/platform/68VZ328/ucdimm/ints.c Thu Dec 13 12:00:11 2001 +++ linux-2.4.x/arch/m68knommu/platform/68VZ328/ucdimm/ints.c Thu Mar 7 01:49:09 2002 @@ -367,3 +367,11 @@ mach_get_irq_list = M68VZ328_get_irq_list; mach_process_int = M68VZ328_do_irq; } + + +void init_irq_proc(void); +void init_irq_proc(void) +{ + /* Insert /proc/irq driver here */ +} + -------------- next part -------------- diff --exclude=*modem* --exclude=*.in --exclude=*digi* --exclude=.* --exclude=*~ -Naur uClinux-2.4.x.orig/drivers/char/68328serial.c linux-2.4.x/drivers/char/68328serial.c --- uClinux-2.4.x.orig/drivers/char/68328serial.c Sun Mar 3 12:34:03 2002 +++ linux-2.4.x/drivers/char/68328serial.c Tue Mar 12 21:06:26 2002 @@ -28,6 +28,7 @@ #include #include #include +#include #include #include @@ -1444,7 +1445,51 @@ printk("MC68328 serial driver version 1.00\n"); } -volatile int test_done; +#ifdef CONFIG_PM +/* Serial Power management + * The console (currently fixed at line 0) is a special case for power + * management because the kernel is so chatty. The console will be + * explicitly disabled my our power manager as the last minute, so we won't + * mess with it here. + */ +static struct pm_dev *serial_pm[NR_PORTS]; + +static int serial_pm_callback(struct pm_dev *dev, pm_request_t request, void *data) +{ + struct m68k_serial *info = (struct m68k_serial *)dev->data; + + if(info == NULL) + return -1; + + /* special case for line 0 - pm restores it */ + if(info->line == 0) + return 0; + + switch (request) { + case PM_SUSPEND: + shutdown(info); + break; + + case PM_RESUME: + startup(info); + break; + } + return 0; +} + +void shutdown_console(void) +{ + struct m68k_serial *info = &m68k_soft[0]; + shutdown(info); +} + +void startup_console(void) +{ + struct m68k_serial *info = &m68k_soft[0]; + startup(info); +} +#endif + /* rs_init inits the driver */ static int __init @@ -1545,10 +1590,19 @@ IRQ_FLG_STD, "M68328_UART", NULL)) panic("Unable to attach 68328 serial interrupt\n"); +#ifdef CONFIG_PM + serial_pm[i] = pm_register(PM_SYS_DEV, PM_SYS_COM, serial_pm_callback); + if (serial_pm[i]) { + printk("info ptr 0x%p\n", info); + serial_pm[i]->data = info; + } +#endif } restore_flags(flags); return 0; } + + /* * register_serial and unregister_serial allows for serial ports to be diff --exclude=*modem* --exclude=*.in --exclude=*digi* --exclude=.* --exclude=*~ -Naur uClinux-2.4.x.orig/drivers/net/cs89x0.c linux-2.4.x/drivers/net/cs89x0.c --- uClinux-2.4.x.orig/drivers/net/cs89x0.c Sun Mar 3 12:34:35 2002 +++ linux-2.4.x/drivers/net/cs89x0.c Sat Mar 16 14:49:27 2002 @@ -425,6 +425,7 @@ printk("cs89x0: Setting up uCcs8900 Chip Select & IRQ ioaddr = 0x%X\n",ioaddr); #endif +#if 0 /* set up the chip select */ *(volatile unsigned char *)0xfffff42b |= 0x01; /* output /sleep */ *(volatile unsigned short *)0xfffff428 |= 0x0101; /* not sleeping */ @@ -434,6 +435,29 @@ *(volatile unsigned short *)0xfffff102 = 0x8000; /* 0x04000000 */ *(volatile unsigned short *)0xfffff112 = 0x01e1; /* 128k, 2ws, FLASH, en */ +#else + /* set up the chip select */ +// *(volatile unsigned char *)0xfffff42b |= 0x01; /* output /sleep */ +// *(volatile unsigned short *)0xfffff428 |= 0x0101; /* not sleeping */ + + *(volatile unsigned char *)0xfffff430 |= 0x08; +// *(volatile unsigned char *)0xfffff432 &= ~(0x08); + *(volatile unsigned char *)0xfffff433 |= 0x08; + *(volatile unsigned char *)0xfffff431 |= (0x08); /* currently high - drive low for + * hware sleep + * indicator */ + + + *(volatile unsigned char *)0xfffff42b &= ~0x02; /* input irq5 */ + *(volatile unsigned short *)0xfffff428 &= ~0x0202; /* irq5 fcn on */ + + *(volatile unsigned short *)0xfffff102 = 0x8000; /* 0x04000000 */ + *(volatile unsigned short *)0xfffff112 = 0x01e1; /* 128k, 2ws, FLASH, + * en */ + + +#endif + #endif /* Initialize the device structure. */ @@ -1379,6 +1403,8 @@ #endif ); netif_start_queue(dev); + + if (net_debug > 1) printk("cs89x0: net_open() succeeded\n"); return 0; @@ -1617,6 +1643,8 @@ { struct net_local *lp = (struct net_local *)dev->priv; + printk("net_close()\n"); + netif_stop_queue(dev); writereg(dev, PP_RxCFG, 0); @@ -1633,6 +1661,14 @@ } #endif +#if 1 + *(volatile unsigned char *)0xfffff431 &= ~(0x08); + printk("put it to sleep\n"); + writereg(dev, PP_SelfCTL, readreg(dev, PP_SelfCTL) | SLEEP_ON); + +#endif + + /* Update the statistics here. */ return 0; } @@ -1725,7 +1761,25 @@ } #endif +#if 0 +static stuct pm_dev *cs89x0_pm; + +static int cs89x0_pm_callback(struct pm_dev *dev, pm_request_t request, void *data) +{ + + switch(request) { + case PM_SUSPEND: + break; + case PM_RESUME: + break; + } + + return 0; + +} + +#endif #ifdef MODULE From danielp at cse.unsw.edu.au Sun Mar 17 23:23:26 2002 From: danielp at cse.unsw.edu.au (Daniel Potts) Date: Mon, 18 Mar 2002 15:23:26 +1100 (EST) Subject: [uClinux-dev] 68*328 power management In-Reply-To: <20020317234058.40291.qmail@web14303.mail.yahoo.com> Message-ID: btw, To enable power management with the previous patch I sent, do this: echo >/proc/sys/pm/suspend To get out of sleep/doze hit the button connected to IRQ3. On Sun, 17 Mar 2002, Mark McChrystal wrote: > Can you send a patch for everything? It sounds like > the SDRAM isn't being refreshed. Is the SDRAM self > refresh? If you have a scope look at the RAS and CAS > lines while it's in sleep. If you don't have a scope, > I'll try your code on a ucsimm board and let you know > what I see. > > > > Mark > > > > > __________________________________________________ > Do You Yahoo!? > Yahoo! Sports - live college hoops coverage > http://sports.yahoo.com/ > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ > -- http://www.cse.unsw.edu.au/~danielp This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From daniel.haensse at fhk.usz.ch Mon Mar 18 01:58:28 2002 From: daniel.haensse at fhk.usz.ch (Daniel Haensse) Date: Mon, 18 Mar 2002 07:58:28 +0100 Subject: [uClinux-dev] "Calibrating delay loop..." crashes In-Reply-To: References: Message-ID: <200203180702.IAA00812@ds-00122.usz.ch> Hi Jonathan, Am Mittwoch, 13. M?rz 2002 22.44 schrieben Sie: > I have been building kernels from Daniel Haenesse's "unified-2.4.6" > distribution for my DragonBall VZ board, and they have been working just > fine. I tried yesterday to build a kernel straight from the newest uClinux > CVS (2.4.17 mostly, as I understand). When the image is programmed to the > flash (using a modified "2flash" to support the TE28F320B3 Intel flash), > and I reboot, the kernel hangs at "Calibrating delay loop...". I checked > in the .c file that is responsible for this part of the bootup sequence, > and there doesn't seem to be anything that could cause a crash. > > I was just wondering if anyone has had this type of problem with a VZ board > running a 2.4 kernel before, and if so, how did you resolve it. Thanks in > advance for any help you can provide. I had the same bug with 2.0.38. There is something wrong with your initialisation of the start and end of the sdram or flash. Dani ( Haensse :-) ) > > > _________________________________________________________________ > Join the world?s largest e-mail service with MSN Hotmail. > http://www.hotmail.com > > This message resent by the uclinux-dev at uclinux.org list server > http://www.uClinux.org/ This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From christian.opel at brooks.com Mon Mar 18 02:09:03 2002 From: christian.opel at brooks.com (Christian Opel) Date: 18 Mar 2002 08:09:03 +0100 Subject: [uClinux-dev] CGI and httpd again, explicit questions. In-Reply-To: <002d01c1cc27$f1e5e1d0$f5feaa0a@IronLung> References: <002d01c1cc27$f1e5e1d0$f5feaa0a@IronLung> Message-ID: <1016435346.1283.4.camel@christian> I don`t know which web server you are running, but even if it supports cgi you`ll need the nec. interpreters e.g. Perl to execute a script.... Am Fre, 2002-03-15 um 14.46 schrieb George Marselis: > > [GetYourLearningOnRegister 0x01] > > Spell Ingredients: > * http://www.apache.org/docs Everything you wanted to read about the best > httpd server and you were too afraid to ask (for the eyes) > * http://cgi-spec.golux.com/ The Common Gateway Interface - RFC Project > Page (for the mind) > * Marry J Blige F. Method Man "You're all I need to survive" (for the soul ) > * Wierd "Al" Yankovitch - "It's all about the Pentiums" ( for the ears ) > * http://www.oreilly.com/catalog/apache2/ - Apache The definite Guide - By > O'Reilly and associates (for the feel) > * 1 glass of strong Greek coffee (for the smell) // Greek coffee is > consummed in glasses, not cups. > * 1 ergonomic keyboard (for the body) > > [STARTMEUP, OH, BABY, STARTMEUP] > > > Hello list, > Hey Stephan - use the Force; read those docs. They will give you a > better understanding of > what CGI is/is not. > > [snip] > > > Its regarding uClinux on EB40LS (from Lineo) and CGI again. Simply, I > > don't get it to work. I did get a suggestion like this from the > > uClinux-mail-list: > > "Make shure your web server is running and/or you provide a cgi interface > > Check settings in /etc/httpd.conf like your path to cgi-bin" > that's a standard unix answer[tm] alright. > > > *** Explicit-question-1 *** > > But how do I make sure it "provides a CGI interface"? > Here are simple suggestions. > 1) read around the web site you downloaded the webserver from. It should > have a page with all the bug^H^H^H^H^H erh, features the server you are > looking at. > 2) read the README file in the dir you unpacked the tarball or in the > ./Documentation dir. > 3) read the source and see if there's a CGI header. if you do find one, > chances are, there's a CGI implementation > > >I made a httpd.conf (because there were none) like this in the suggested > >location. > (gramatical error - was not were) > > ># Trying to get httpd to exec my CGI- > ># -scripts located in /htdocs/cgi-bin/ > Bad tactic #1: Never, never, *ever* put the cgi-bin dir under htdocs, > unless you want bad things to happen like root access and rm -rf / on your > system. cgi-bin should reside under the ServerRoot i.e. if ServerRoot is > /web/httpd, then cgi-bin should be /web/httpd/cgi-bin > [snip] > > >Since I really don't know, I wonder if this is "the way". I have found > >really nothing about this on uClinux, unfortunately ... and therefore I > >bother u ;-) > Well, all fine and dandy, but do try to read up/ask around first. It's > the UNIX way :) > > >*** Explicit-question-2 *** > >Does the provided httpd work with cgi and provide that interface or is the > >problem there? > Stephan, what webserver are you using? tiny? boa? simple? please be more > specific. > boa is a full-featured web server, so, it has a CGI 1.1 implementation. > tiny has only a POST CGI capability. so, in essence, you can write some CGI > programs, but you can't do everything apache/boa does. > simple httpd has no CGI implementation > > >My call for the cgi-script (in C) looks like this: > > > >
> method="GET"> > > > >
> Uhm. could you please be more specific? that's html, not C . Do you > mean, you coded the hellotml.cgi program in C? > Is that snippet inside a #include #include int > main( int argc, char *argv[]) { printf(" your code here\n"); exit(0);}? Or > is it how you called the cgi from your html? > > >*** Explicit-question-3 *** > >Is it right so far? (the problem is that it does not execute, just shows > the > >text "bFLT" (binary flat format?) on the screen ... or text if I try to > >execute a shellscript as cgi). > didja remember to chmod 755 the cgi program? the CGI 1.1 standard mentions > that httpd has to exec() (or any of its derivatives) to start the > > >*** Explicit-question-4 and on ... *** > >4.1 Shall the httpd.conf be located in /etc/? (or shall the config be named > >something else and be located elsewhere?) > You should be able to place it anywhere. Look at the makefile/config > script. you should be able to manually configure it/pass it as a param to > config > > >4.2 Should it be a httpd.conf at all in uClinux shipped installation from > >Lineo (since there is none such to begin with), or should it maybe be > >configurated elsewhere (as in inetd.conf, 'services' or something)? > That's up the vendor to decide if it will include a configure script. > I'm pretty sure they have. why don't you try to use the command find to > locate it? If there isn't one, try looking up a manpage/docs the vendor > included. they should have some examples included on how to configure the > server. > > >4.3 In services it says: > >http stream tcp nowait root /sbin/httpd -i > >What does the '-i' stand for? Is it possible to define something here? > I can't tell you what -i stands for, since I don't know what server you are > running. And, no, /etc/services is not where you define httpd directives. > > >4.4 Should something special be defined somewhere in the HTML-"code" to get > >a browser to understand that there will be CGI? > nope CGI is server-side; all the client gets is HTML > see http://www.oreilly.com/catalog/cgi/ "CGI prorgaming with C & Perl" from > O'Reilly & Associates > > >So, now my questions are finally to an end ... for this time. > Cool, glad to be of help. But do try to read up next time. All this > information could be really easilly answered by just reading the CGI > standard alone... > > >Regards, > >Stefan Jonsson, > >Student, University of Ume?, Sweden. > > -><- > George Marselis > WuTang Soldier Spunky Misunderstood Genious > http://www.hack.gr/users/isobiths/ > > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From stefan.jonsson at faltcom.se Mon Mar 18 02:46:41 2002 From: stefan.jonsson at faltcom.se (Stefan Jonsson) Date: Mon, 18 Mar 2002 08:46:41 +0100 Subject: [uClinux-dev] "Userland" ARM Message-ID: Hello list, I have both Lineos EB40LS and an EB40+MEC01. I am doing a project in my study. I have been asked to make a new "userland" not from the CD and put uClinux into the standard EB40 with memory extension card. Is it possible to have two paraller "userlands" or will the compilation or depandencies or other things be disturbed by this? (strange question eh?) I just wonder if someone knows anything about this. I am trying to do this now. Regards, Stefan Jonsson, Student, University of Umea, Sweden. This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From julienb at actimage.fr Mon Mar 18 03:04:09 2002 From: julienb at actimage.fr (Julien Boibessot) Date: Mon, 18 Mar 2002 09:04:09 +0100 Subject: [uClinux-dev] Gdb+bdm+5272C3 References: <7FF1185EA0D3D511BC0500B0D0D0C24A1B5899@melexc01.tecom.com.au> <05a401c1cbff$6784a5d0$8947ec95@bruker.fr> <3C952C62.2080506@snapgear.com> Message-ID: <009701c1ce53$7b969c80$8947ec95@bruker.fr> Hi Greg ! Well, in fact my gdb init script is the one you posted on this list some times ago.....:-) And it initializes DRAM controller...... Thanks Julien ----- Original Message ----- From: "Greg Ungerer" To: Sent: Monday, March 18, 2002 12:53 AM Subject: Re: [uClinux-dev] Gdb+bdm+5272C3 > > Hi Julien, > > Julien Boibessot wrote: > > > I'm trying to use GDB thrue BDM (P&E Adaptator) with the Motorola 5272C3. > > > > I have compile and insert BDM driver ( gdb-bdm-20010901 from cybertec) in my > > host linux and it works (chk /dev/bdmcf0 is running and all tests are ok). > > I've got GDB 5.1.1 sources, patche it for m68k/coldfire BDM debugging and > > recompile it to get a running "m68k-bdm-elf-gdb". > > > > Then, I launch gdb with: "m68k-bdm-elf-gdb --command=5272.gdb image.elf". > > 5272.gdb is my init script for GDB and image.elf is my linux image. > > > > gdb is running and I can do whatever I want. So I launch the "load" command, > > take a coffee to wait the end of the download, and "set $pc=0x20000". But > > then when I try to execute the program by typing command "c", gdb crashes (I > > get a "segmentation fault"). My board is reseted. > > > > Does anybody knows what is happening ??? > > > Does you gdbinit script initialize RAM/SDRAM? > > If it doesn't then your load cannot possibly work, since > it will be trying to load the binary into RAM. > > Regards > Greg > > > ------------------------------------------------------------------------ > Greg Ungerer -- Chief Software Wizard EMAIL: gerg at snapgear.com > SnapGear PHONE: +61 7 3435 2888 > 825 Stanley St, FAX: +61 7 3891 3630 > Woolloongabba, QLD, 4102, Australia WEB: www.snapgear.com > > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From stefan.jonsson at faltcom.se Mon Mar 18 04:12:39 2002 From: stefan.jonsson at faltcom.se (Stefan Jonsson) Date: Mon, 18 Mar 2002 10:12:39 +0100 Subject: SV: [uClinux-dev] CGI and httpd again, explicit questions. In-Reply-To: <1016435346.1283.4.camel@christian> Message-ID: Hello, Unfortunately I do not know which httpd is shipped on board from Lineo on EB40LS+uCnet. I was hoping that someone knows. Therefore I don't know either, and no documentation about it is to be found. As it seems though (from an answer by mail to this questions), not every version of the shipped httpd works with cgi. What is a "nec"? I am scripting in C, that should need nothing extra to work if I am right? Thanks for your help. Regards, Stefan Jonsson -----Ursprungligt meddelande----- Fr?n: owner-uclinux-dev at uclinux.org [mailto:owner-uclinux-dev at uclinux.org]F?r Christian Opel Skickat: den 18 mars 2002 08:09 Till: Uclinux email Forum ?mne: Re: [uClinux-dev] CGI and httpd again, explicit questions. I don`t know which web server you are running, but even if it supports cgi you`ll need the nec. interpreters e.g. Perl to execute a script.... Am Fre, 2002-03-15 um 14.46 schrieb George Marselis: > > [GetYourLearningOnRegister 0x01] > > Spell Ingredients: > * http://www.apache.org/docs Everything you wanted to read about the best > httpd server and you were too afraid to ask (for the eyes) > * http://cgi-spec.golux.com/ The Common Gateway Interface - RFC Project > Page (for the mind) > * Marry J Blige F. Method Man "You're all I need to survive" (for the soul ) > * Wierd "Al" Yankovitch - "It's all about the Pentiums" ( for the ears ) > * http://www.oreilly.com/catalog/apache2/ - Apache The definite Guide - By > O'Reilly and associates (for the feel) > * 1 glass of strong Greek coffee (for the smell) // Greek coffee is > consummed in glasses, not cups. > * 1 ergonomic keyboard (for the body) > > [STARTMEUP, OH, BABY, STARTMEUP] > > > Hello list, > Hey Stephan - use the Force; read those docs. They will give you a > better understanding of > what CGI is/is not. > > [snip] > > > Its regarding uClinux on EB40LS (from Lineo) and CGI again. Simply, I > > don't get it to work. I did get a suggestion like this from the > > uClinux-mail-list: > > "Make shure your web server is running and/or you provide a cgi interface > > Check settings in /etc/httpd.conf like your path to cgi-bin" > that's a standard unix answer[tm] alright. > > > *** Explicit-question-1 *** > > But how do I make sure it "provides a CGI interface"? > Here are simple suggestions. > 1) read around the web site you downloaded the webserver from. It should > have a page with all the bug^H^H^H^H^H erh, features the server you are > looking at. > 2) read the README file in the dir you unpacked the tarball or in the > ./Documentation dir. > 3) read the source and see if there's a CGI header. if you do find one, > chances are, there's a CGI implementation > > >I made a httpd.conf (because there were none) like this in the suggested > >location. > (gramatical error - was not were) > > ># Trying to get httpd to exec my CGI- > ># -scripts located in /htdocs/cgi-bin/ > Bad tactic #1: Never, never, *ever* put the cgi-bin dir under htdocs, > unless you want bad things to happen like root access and rm -rf / on your > system. cgi-bin should reside under the ServerRoot i.e. if ServerRoot is > /web/httpd, then cgi-bin should be /web/httpd/cgi-bin > [snip] > > >Since I really don't know, I wonder if this is "the way". I have found > >really nothing about this on uClinux, unfortunately ... and therefore I > >bother u ;-) > Well, all fine and dandy, but do try to read up/ask around first. It's > the UNIX way :) > > >*** Explicit-question-2 *** > >Does the provided httpd work with cgi and provide that interface or is the > >problem there? > Stephan, what webserver are you using? tiny? boa? simple? please be more > specific. > boa is a full-featured web server, so, it has a CGI 1.1 implementation. > tiny has only a POST CGI capability. so, in essence, you can write some CGI > programs, but you can't do everything apache/boa does. > simple httpd has no CGI implementation > > >My call for the cgi-script (in C) looks like this: > > > >
> method="GET"> > > > >
> Uhm. could you please be more specific? that's html, not C . Do you > mean, you coded the hellotml.cgi program in C? > Is that snippet inside a #include #include int > main( int argc, char *argv[]) { printf(" your code here\n"); exit(0);}? Or > is it how you called the cgi from your html? > > >*** Explicit-question-3 *** > >Is it right so far? (the problem is that it does not execute, just shows > the > >text "bFLT" (binary flat format?) on the screen ... or text if I try to > >execute a shellscript as cgi). > didja remember to chmod 755 the cgi program? the CGI 1.1 standard mentions > that httpd has to exec() (or any of its derivatives) to start the > > >*** Explicit-question-4 and on ... *** > >4.1 Shall the httpd.conf be located in /etc/? (or shall the config be named > >something else and be located elsewhere?) > You should be able to place it anywhere. Look at the makefile/config > script. you should be able to manually configure it/pass it as a param to > config > > >4.2 Should it be a httpd.conf at all in uClinux shipped installation from > >Lineo (since there is none such to begin with), or should it maybe be > >configurated elsewhere (as in inetd.conf, 'services' or something)? > That's up the vendor to decide if it will include a configure script. > I'm pretty sure they have. why don't you try to use the command find to > locate it? If there isn't one, try looking up a manpage/docs the vendor > included. they should have some examples included on how to configure the > server. > > >4.3 In services it says: > >http stream tcp nowait root /sbin/httpd -i > >What does the '-i' stand for? Is it possible to define something here? > I can't tell you what -i stands for, since I don't know what server you are > running. And, no, /etc/services is not where you define httpd directives. > > >4.4 Should something special be defined somewhere in the HTML-"code" to get > >a browser to understand that there will be CGI? > nope CGI is server-side; all the client gets is HTML > see http://www.oreilly.com/catalog/cgi/ "CGI prorgaming with C & Perl" from > O'Reilly & Associates > > >So, now my questions are finally to an end ... for this time. > Cool, glad to be of help. But do try to read up next time. All this > information could be really easilly answered by just reading the CGI > standard alone... > > >Regards, > >Stefan Jonsson, > >Student, University of Ume?, Sweden. > > -><- > George Marselis > WuTang Soldier Spunky Misunderstood Genious > http://www.hack.gr/users/isobiths/ > > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From philippe.ney at smartdata.ch Mon Mar 18 04:54:26 2002 From: philippe.ney at smartdata.ch (Philippe Ney) Date: Mon, 18 Mar 2002 10:54:26 +0100 Subject: [uClinux-dev] 68*328 power management In-Reply-To: References: Message-ID: <20020318105426.0aa9f31c.philippe.ney@smartdata.ch> > Hi, > > I've been trying to implement some power management on the ucdimm, but I'm > not having much luck when I turn the clock off (PLLCR |= PLLCR_DISPLL). > ... Hi, Did you ensure that low-power refresh mode is set for your RAM? /* RAM w/o self-refresh mode * -> ensure that low-power refresh mode is set */ DRAMC |= DRAMC_LPR; /* DRAMC:LPR to high */ BTW, I also try using the self-refresh mode settings with apropriate RAM but without success... /* we have RAM whit self-refresh mode */ DRAMC |= DRAMC_RM; /* DRAMC:RM to self-refresh mode */ And I had to set my RAM as w/o self-refresh-mode Maybe this help, -philippe This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From superwen at 263.net Mon Mar 18 05:42:33 2002 From: superwen at 263.net (Kevin) Date: Mon, 18 Mar 2002 18:42:33 +0800 Subject: [uClinux-dev] About console Message-ID: <20020318103522.55F8C1DECF461@smtp.263.net> Hi, There, I still can't use the printk(), so I just use the console_68328_print() to replace it. then looks like I can go a little bit further, but I'm not sure if it's OK. Can anybody give me a hand on this? My kernel finally failed with "unable to open an initial console". So I'm wondering if this is caused by the above replacing? Best Regards Kevin ---------- here I attached the message I got from uart. KERNEL -> TEXT=0x1000400-0x104e308 DATA=0x020000-0x03c000 BSS=0x03c000-0x04f198 CAT ROMARRAY! KERNEL -> ROMFS=0x106a308-0x1125f78 MEM=0x050000-0x800000 STACK=0x800000-0x80000 0 On node 0 totalpages: 2048 zone(0): 0 pages. zone(1): 2048 pages. zone(2): 0 pages. Done setup_arch init/main.c:line 561: Setup_Arch Done Kernel command line: -------This looks not good. Calibrating delay loop... 1.40 BogoMIPS init/main.c:line 586: Calibrate_delay.Done Memory available: 7728k/8064k RAM, 0k/0k ROM (311k kernel code, 188k data) Dentry-cache hash table entries: 1024 (order: 1, 8192 bytes) Inode-cache hash table entries: 512 (order: 0, 4096 bytes) Mount-cache hash table entries: 512 (order: 0, 4096 bytes) Buffer-cache hash table entries: 1024 (order: 0, 4096 bytes) Page-cache hash table entries: 2048 (order: 1, 8192 bytes) POSIX conformance testing by UNIFIX <6>Linux NET4.0 for Linux 2.4 <6>Based upon Swansea University Computer Society NET3.039 Starting kswapd pty: 256 Unix98 ptys configured MC68328 serial driver version 1.00 ttyS0 at 0xfffff900 (irq = 64) is a builtin MC68328 UART block: 64 slots per queue, batch=16 RAMDISK driver initialized: 16 RAM disks of 4096K size 1024 blocksize Blkmem copyright 1998,1999 D. Jeff Dionne Blkmem copyright 1998 Kenneth Albanowski Blkmem 1 disk images: 0: 106A308-1126307 [VIRTUAL 106A308-1126307] (RO) VFS: Mounted root (romfs filesystem) readonly. Warning: unable to open an initial console. This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From kovacs_kernig at hotmail.com Mon Mar 18 06:02:05 2002 From: kovacs_kernig at hotmail.com (kernig kovacs) Date: Mon, 18 Mar 2002 19:02:05 +0800 Subject: [uClinux-dev] How about samba in ARM7 NOMMU platform? References: Message-ID: To kind all I want to compile smbd and nmbd of ARM7 NOMMU version. The sources are from cvs of uclinux. I failed with messages below. I also solve an erron by modifying elf2flt.ld from LENGTH = 0x100000 to LENGTH = 0x200000. MEMORY { flatmem : ORIGIN = 0x0, LENGTH = 0x200000 } I think this may cause problems. Can someone tell me how to solve this problem? Regards, Kovacs. MESSAGE: [root at localhost source]# make Using FLAGS = -D_LIBC -D__linux__ -DEMBED -I../../../library/include -Iinclude -I./inc lude -I./ubiqx -I./smb wrapper -DSMBLOGFILE="/usr/local/samba/var/log.smb" -DNMBLOGFILE="/usr/loca l/samba/var/log.nmb" -DCONFIGFILE= "/usr/local/samba/lib/smb.conf" -DLMHOSTSFILE="/usr/local/samba/lib/lmhosts" -DSWATDIR="/usr/local/samba/swa t" -DSBINDIR="/usr/local/samba/bin" -DLOCKDIR="/usr/local/samba/var/locks" - DSMBRUN="/usr/local/samba/bin/smbr un" -DCODEPAGEDIR="/usr/local/samba/lib/codepages" -DDRIVERFILE="/usr/local/ samba/lib/printers.def" -DBINDIR=" /usr/local/samba/bin" -DHAVE_INCLUDES_H -DPASSWD_PROGRAM="/bin/passwd" -DSMB _PASSWD_FILE="/usr/local/samba/pri vate/smbpasswd" Using FLAGS32 = -D_LIBC -D__linux__ -DEMBED -I../../../library/include -Iinclude -I./inc lude -I./ubiqx -I./s mbwrapper -DSMBLOGFILE="/usr/local/samba/var/log.smb" -DNMBLOGFILE="/usr/lo cal/samba/var/log.nmb" -DCONFIGFIL E="/usr/local/samba/lib/smb.conf" -DLMHOSTSFILE="/usr/local/samba/lib/lmhost s" -DSWATDIR="/usr/local/samba/s wat" -DSBINDIR="/usr/local/samba/bin" -DLOCKDIR="/usr/local/samba/var/locks" -DSMBRUN="/usr/local/samba/bin/sm brun" -DCODEPAGEDIR="/usr/local/samba/lib/codepages" -DDRIVERFILE="/usr/loca l/samba/lib/printers.def" -DBINDIR ="/usr/local/samba/bin" -DHAVE_INCLUDES_H -DPASSWD_PROGRAM="/bin/passwd" -DS MB_PASSWD_FILE="/usr/local/samba/p rivate/smbpasswd" Using LIBS = Linking bin/smbd arm-elf-gcc -D_LIBC -D__linux__ -DEMBED -I../../../library/include -Iinclu de -I./include -I./ubiqx -I./smbwr apper -DSMBLOGFILE=\"/usr/local/samba/var/log.smb\" -DNMBLOGFILE=\"/usr/loc al/samba/var/log.nmb\" -DCONFIGFIL E=\"/usr/local/samba/lib/smb.conf\" -DLMHOSTSFILE=\"/usr/local/samba/lib/lmh osts\" -DSWATDIR=\"/usr/local/sa mba/swat\" -DSBINDIR=\"/usr/local/samba/bin\" -DLOCKDIR=\"/usr/local/samba/v ar/locks\" -DSMBRUN=\"/usr/local/s amba/bin/smbrun\" -DCODEPAGEDIR=\"/usr/local/samba/lib/codepages\" -DDRIVERF ILE=\"/usr/local/samba/lib/printer /print_cups.o profile/profile.o lib/charcnv.o lib/charset.o lib/debug.o lib/fault.o lib/getsmbpass.o lib/interface.o lib/kanji.o lib/md4.o lib/interfaces.o lib/pidfile.o lib/replace.o lib/signal.o lib/slprintf.o lib/system.o lib/doscalls.o lib/time.o lib/ufc.o lib/genrand.o lib/username.o lib/access.o lib/smbrun.o lib/bitmap.o lib/crc32.o lib/snprintf.o lib/util_str.o lib/util_sid.o lib/util_unistr.o lib/util_file.o lib/util.o lib/util_sock.o lib/util_sec.o smbd/ssl.o lib/fnmatch.o lib/talloc.o lib/hash.o -d -fno-rtti -nostartfiles -T ./elf2flt.ld -Ur -L../../../library/lib -lc -lgcc -lcrypt collect2: ld terminated with signal 11 [Segmentation fault], core dumped /usr/local/gnu/arm-elf/bin/ld: warning: no memory region specified for section `.rel.plt' /usr/local/gnu/arm-elf/bin/ld: warning: no memory region specified for section `.rel.bss' /usr/local/gnu/arm-elf/bin/ld: warning: no memory region specified for section `.plt' /usr/local/gnu/arm-elf/bin/ld: warning: no memory region specified for section `.hash' /usr/local/gnu/arm-elf/bin/ld: warning: no memory region specified for section `.dynstr' /usr/local/gnu/arm-elf/bin/ld: warning: no memory region specified for section `.dynsym' /usr/local/gnu/arm-elf/bin/ld: warning: no memory region specified for section `.gnu.version_r' /usr/local/gnu/arm-elf/bin/ld: warning: no memory region specified for section `.gnu.version' /usr/local/gnu/arm-elf/bin/ld: warning: no memory region specified for section `.gnu.version_d' /usr/local/gnu/arm-elf/bin/ld: warning: no memory region specified for section `.interp' /usr/local/gnu/arm-elf/bin/ld: warning: no memory region specified for section `.dynamic' /usr/local/gnu/arm-elf/bin/ld: section .rel.plt [00000000 -> 000005d7] overlaps section .text [00000000 -> 001064ff] /usr/local/gnu/arm-elf/bin/ld: section .rel.bss [000005d8 -> 000005ff] overlaps section .text [00000000 -> 001064ff] /usr/local/gnu/arm-elf/bin/ld: section .plt [00000600 -> 000011bf] overlaps section .text [00000000 -> 001064ff] /usr/local/gnu/arm-elf/bin/ld: section .hash [000011c0 -> 000017ff] overlaps section .text [00000000 -> 001064ff] /usr/local/gnu/arm-elf/bin/ld: section .dynstr [00001800 -> 00001e3c] overlaps section .text [00000000 -> 001064ff] /usr/local/gnu/arm-elf/bin/ld: section .dynsym [00001e40 -> 00002acf] overlaps section .text [00000000 -> 001064ff] /usr/local/gnu/arm-elf/bin/ld: section .interp [00002ad0 -> 00002ae0] overlaps section .text [00000000 -> 001064ff] /usr/local/gnu/arm-elf/bin/ld: section .dynamic [00002ae4 -> 00002b63] overlaps section .text [00000000 -> 001064ff] /usr/local/gnu/arm-elf/bin/ld: bfd assertion fail ../../bfd/elf32-arm.h:3127 make: *** [bin/smbd] Error 1 ELF2FLT.LD : MEMORY { flatmem : ORIGIN = 0x0, LENGTH = 0x200000 } SECTIONS { .text 0 : { . = ALIGN(0x4) ; _stext = . ; _start = . ; *(.text) *(.text.*) *(.gnu.warning) *(.stub) *(.gnu.linkonce.t*) *(.glue_7t) *(.glue_7) *(.init) . = ALIGN(0x10) ; _etext = . ; } > flatmem .data : { . = ALIGN(0x4) ; _sdata = . ; __data_start = . ; data_start = . ; *(.rela.got) *(.rel.got) *(.got.plt) *(.got) LONG(-1) *(.rodata) *(.rodata.*) *(.gnu.linkonce.r*) *(.rodata1) *(.data) *(.data.*) *(.gnu.linkonce.d*) *(.data1) *(.eh_frame) *(.gcc_except_table) *(.sdata) *(.sdata.*) *(.gnu.linkonce.s.*) . = ALIGN(4) ; *(.ctors.*) *(.ctors) LONG(0) *(.dtors.*) *(.dtors) LONG(0) . = ALIGN(0x10) ; _edata = . ; } > flatmem .bss : { . = ALIGN(0x4) ; _sbss = ALIGN(0x4) ; __bss_start = . ; *(.dynsbss) *(.sbss) *(.sbss.*) *(.scommon) *(.dynbss) *(.bss) *(.bss.*) *(COMMON) . = ALIGN(0x4) ; _ebss = . ; _end = . ; end = . ; } > flatmem } This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From gerg at snapgear.com Mon Mar 18 07:38:01 2002 From: gerg at snapgear.com (gerg) Date: Mon, 18 Mar 2002 22:38:01 +1000 Subject: [uClinux-dev] Gdb+bdm+5272C3 References: <7FF1185EA0D3D511BC0500B0D0D0C24A1B5899@melexc01.tecom.com.au> <05a401c1cbff$6784a5d0$8947ec95@bruker.fr> <3C952C62.2080506@snapgear.com> <009701c1ce53$7b969c80$8947ec95@bruker.fr> Message-ID: <3C95DFA9.F505952D@snapgear.com> Hi Julien, Julien Boibessot wrote: > Well, in fact my gdb init script is the one you posted on this list some > times ago.....:-) > And it initializes DRAM controller...... OK :-) Check after you do the "load" that the code looks right in memory. Dump it and check. If it looks good then single step through some code to verify that you can execute code. Regards Greg ------------------------------------------------------------------------ Greg Ungerer -- Chief Software Wizard EMAIL: gerg at snapgear.com Snapgear PHONE: +61 7 3279 1822 825 Stanley St, FAX: +61 7 3279 1820 Woolloongabba, QLD, 4102, Australia WEB: www.snapgear.com This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From stefan.jonsson at faltcom.se Mon Mar 18 07:51:27 2002 From: stefan.jonsson at faltcom.se (Stefan Jonsson) Date: Mon, 18 Mar 2002 13:51:27 +0100 Subject: [uClinux-dev] Download and boot on EB40 Message-ID: Hello list, I have an EB40+MEC01 memory extension card. Atmels tools though is for windows. I am using linux and want to put uClinux on the EB40+MEC. I wonder about: 1) I have seen something about modifying the arch/armnommu/kernel/head-arm-atmel.S (that mail was from june 2001, is this still valid?). 2) how to download an image to target and how to program it into flash (on MEC01)? I get some kind of contact with angel debug monitor but don't know how to use it for my purpose. 3) how to boot from the downloaded image (from MEC)? (Or just how to execute it at least?) Regards, Stefan Jonsson, Student, University of Umea, Sweden. This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From gmenie at akamai.com Mon Mar 18 08:37:50 2002 From: gmenie at akamai.com (Menie, Georges) Date: Mon, 18 Mar 2002 08:37:50 -0500 Subject: [uClinux-dev] About console Message-ID: there is several reason for this to happen, go to http://home.at/uclinux/ click on the "Searchable MARC of ?Clinux-dev list" link and enter your error message in the search box. You will see all the related emails, check your settings according to the replies. Sorry, I don't know which one is the answer for your case, but I'm pretty sure the answer is there. Regards, Georges > -----Original Message----- > From: Kevin [mailto:superwen at 263.net] > Sent: lundi 18 mars 2002 11:43 > To: uclinux-dev at uclinux.org > Subject: Re: [uClinux-dev] About console > > > Hi, There, > > I still can't use the printk(), so I just use the > console_68328_print() to replace it. then looks like I can go > a little bit further, but I'm not sure if it's OK. Can > anybody give me a hand on this? > > My kernel finally failed with "unable to open an initial > console". So I'm wondering if this is caused by the above replacing? > > Best Regards > Kevin > > ---------- > here I attached the message I got from uart. > > KERNEL -> TEXT=0x1000400-0x104e308 DATA=0x020000-0x03c000 > BSS=0x03c000-0x04f198 > CAT ROMARRAY! > KERNEL -> ROMFS=0x106a308-0x1125f78 MEM=0x050000-0x800000 > STACK=0x800000-0x80000 > 0 > On node 0 totalpages: 2048 > zone(0): 0 pages. > zone(1): 2048 pages. > zone(2): 0 pages. > Done setup_arch > init/main.c:line 561: Setup_Arch Done > Kernel command line: -------This looks > not good. > Calibrating delay loop... 1.40 BogoMIPS > init/main.c:line 586: Calibrate_delay.Done > Memory available: 7728k/8064k RAM, 0k/0k ROM (311k kernel > code, 188k data) > Dentry-cache hash table entries: 1024 (order: 1, 8192 bytes) > Inode-cache hash table entries: 512 (order: 0, 4096 bytes) > Mount-cache hash table entries: 512 (order: 0, 4096 bytes) > Buffer-cache hash table entries: 1024 (order: 0, 4096 bytes) > Page-cache hash table entries: 2048 (order: 1, 8192 bytes) > POSIX conformance testing by UNIFIX > <6>Linux NET4.0 for Linux 2.4 > <6>Based upon Swansea University Computer Society NET3.039 > Starting kswapd > pty: 256 Unix98 ptys configured > MC68328 serial driver version 1.00 > ttyS0 at 0xfffff900 (irq = 64) is a builtin MC68328 UART > block: 64 slots per queue, batch=16 > RAMDISK driver initialized: 16 RAM disks of 4096K size 1024 blocksize > Blkmem copyright 1998,1999 D. Jeff Dionne > Blkmem copyright 1998 Kenneth Albanowski > Blkmem 1 disk images: > 0: 106A308-1126307 [VIRTUAL 106A308-1126307] (RO) > VFS: Mounted root (romfs filesystem) readonly. > Warning: unable to open an initial console. > > > > This message resent by the uclinux-dev at uclinux.org list > server http://www.uClinux.org/ > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From eauth at softsys.co.at Mon Mar 18 09:36:18 2002 From: eauth at softsys.co.at (Erwin Authried) Date: Mon, 18 Mar 2002 15:36:18 +0100 Subject: [uClinux-dev] Download and boot on EB40 Message-ID: <01C1CE92.A64E9740@smithwicks.softsys.co.at> Stefan Jonsson[SMTP:stefan.jonsson at faltcom.se] wrote: > Hello list, > > I have an EB40+MEC01 memory extension card. Atmels tools though is for > windows. I am using linux and want to put uClinux on the EB40+MEC. I wonder > about: > > 1) I have seen something about modifying the > arch/armnommu/kernel/head-arm-atmel.S (that mail was from june 2001, is this > still valid?). That's the code that configures the EBI for the additional memory. I think it shouldn't go into the kernel, it's the bootloader's task to configure the EBI and remap the memory. > > 2) how to download an image to target and how to program it into flash (on > MEC01)? > I get some kind of contact with angel debug monitor but don't know how to > use it for my purpose. > > 3) how to boot from the downloaded image (from MEC)? (Or just how to execute > it at least?) > Have a look at my arm-boot package. It contains the armtool monitor that allows you to initialize your board, download binary images, execute them, and program the flash. http://cvs.home.at/ http://uclinux.home.at/midori/apps/armtool-0.1_3.mlz -Erwin -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/ms-tnef Size: 2043 bytes Desc: not available URL: From hamilton at SEDSystems.ca Mon Mar 18 09:41:52 2002 From: hamilton at SEDSystems.ca (Kendrick Hamilton) Date: Mon, 18 Mar 2002 08:41:52 -0600 (CST) Subject: [uClinux-dev] Flash Write In-Reply-To: Message-ID: Have you looked at JFFS Kendrick Hamilton E.I.T. SED Systems, a division of Calian Ltd. 18 Innovation Blvd. PO Box 1464 Saskatoon, Saskatchewan Canada S7N 3R1 Hamilton at sedsystems.ca Tel: (306) 933-1453 Fax: (306) 933-1486 On Sun, 17 Mar 2002, Juergen Mueller wrote: > Hi!! > > Is it possible to write into the flash at runtime. Is there something like a > flash filesystem which allows to write into the flash. > I'm using uCdimm MC68Vz328 and Kernel 2.4.x. > > Any suggestion is appreciated. > > Thanks! > > Juergen > > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From tahir at snom.de Mon Mar 18 12:23:22 2002 From: tahir at snom.de (Muhammad Usman Tahir) Date: Mon, 18 Mar 2002 18:23:22 +0100 Subject: [uClinux-dev] m68k-elf-g++: compiling c++ files. Message-ID: <013101c1cea1$9b347f50$2200a8c0@flamenco> Hi there, I'm using uClinux 2.0.38 with m68k tools version 2.95.3 from DavidM for an MCF 5206 (m68k-elf-tools-20020218.tar.gz). The kernel and built-in user applications were compiled successfully and are running fine. I now want to add an application of my own. Incidentally it is in C++ (I want to reuse an existing app). I have followed all the conventional steps to add an application in uClinux-coldfire/user and editing the corresponding makefiles etc. When I try to compile the source files through this, the following errors come up: cpp: myapp.cxx: C++ compiler not installed on this system cpp: -lang-c++: linker input file unused since linking not done m68k-elf-g++: installation problem, cannot exec `cc1plus': No such file or directory make: *** [obj--r/myapp.o] Error 1 >from the look of things it seems to be a linker error i.e. m68k-elf-ld. I even tried to compile a simple c++ sample source file independently without any luck (something like HelloWorld.cxx). I may add here that the application added doesn't have any known problems and compiles fine with simple gcc. I tried the cross compiler technique recently suggested by Ilguiz Latypov for colilo (new README), in hopes of getting m68k-gcc working properly. But still the problem remains the same. Any ideas regarding the issue and ways to get c++ sources compiled for MCF 52XX by m68k tools? Do I have to compile the tools myself? Thanks, Usman. This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From chris at cjx.com Mon Mar 18 12:27:56 2002 From: chris at cjx.com (Chris Allen) Date: Mon, 18 Mar 2002 17:27:56 +0000 Subject: [uClinux-dev] Looking for LCD module source in the UK Message-ID: <20020318172756.A16455@cjx.com> I am looking for a source for a reasonably cheap LCD module for use with ucLinux on the ucSimm and similar. Ideally with a built-in touch screen, or an easily attachable touch screen.. Does anybody have a source for this in the UK? - preferably somwhere that can ship out small quantities from stock in return for a credit card number. I have looked in RS, Rapid, Arrow and Maplin but none of those has a very good selection. .. and one of Sharp's distributors quoted me over GBP200 for a Sharp module - for this price I could buy a couple of Psion Revos and pull the screens out of them!! Many thanks, Chris Allen chris at cjx.com This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From JSpark104 at aol.com Mon Mar 18 13:10:37 2002 From: JSpark104 at aol.com (JSpark104 at aol.com) Date: Mon, 18 Mar 2002 13:10:37 EST Subject: [uClinux-dev] Looking for LCD module source in the UK Message-ID: Hey Chris, I know that they are not in the UK, but www.eio.com has a decent selection of surplus LCD's, in small quantities. They accept paypal, and I beleive they now allow you to coordinate the order over e-mail. Their stock rotates around, so I'm not sure if they have any touch screens, but if they do, they will be much less expensive than 200 pounds. -Josh This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From castet at firstlinux.net Mon Mar 18 13:48:49 2002 From: castet at firstlinux.net (laurent castet) Date: Mon, 18 Mar 2002 10:48:49 -0800 (PST) Subject: [uClinux-dev] ks8737 Message-ID: <20020318184849.A8E6E2756@sitemail.everyone.net> An embedded and charset-unspecified text was scrubbed... Name: not available URL: From yvon at tempemicroelectronics.com Mon Mar 18 13:54:34 2002 From: yvon at tempemicroelectronics.com (yvon) Date: Mon, 18 Mar 2002 11:54:34 -0700 Subject: [uClinux-dev] Looking for LCD module source in the UK References: <20020318172756.A16455@cjx.com> Message-ID: <00d501c1ceae$58f6df90$0200a8c0@CX2040412A> Earth LCD is a good source.Their web site is www.earthlcd.com. The EG9013-nz-1 is 640-480, they also have a touch screen kit for $199.00. Another solution is a monochrome LCD with touch screen for PDA sold for $99.00 (Samsung UG-24UD1) Yvon ----- Original Message ----- From: "Chris Allen" To: Sent: Monday, March 18, 2002 10:27 AM Subject: [uClinux-dev] Looking for LCD module source in the UK > I am looking for a source for a reasonably cheap LCD > module for use with ucLinux on the ucSimm and similar. > Ideally with a built-in touch screen, or an easily > attachable touch screen.. > > Does anybody have a source for this in the UK? - preferably > somwhere that can ship out small quantities from stock in return for > a credit card number. I have looked in RS, Rapid, Arrow and > Maplin but none of those has a very good selection. > > > .. and one of Sharp's distributors quoted me over GBP200 for > a Sharp module - for this price I could buy a couple of > Psion Revos and pull the screens out of them!! > > > Many thanks, > > > Chris Allen > chris at cjx.com > > > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From tgriggs at keyww.com Mon Mar 18 16:13:44 2002 From: tgriggs at keyww.com (Travis Griggs) Date: Mon, 18 Mar 2002 13:13:44 -0800 Subject: [uClinux-dev] reboot on the coldfire... Message-ID: <3C965888.9040606@keyww.com> Is this known to not work? I tried reboot, both the sash builtin and the busy box one. They both blow up similiarily. Some sort of kernel trap I think. Here's what Minicom spits back when I try it: /> reboot Nov 30 00:02:05 dhcpcd[17]: terminating on signal Restarting system. bad frame format: 00000000 PC: [<374e2010>] SR: 2700 SP: 00293e60 a2: 001695d8 d0: 00000016 d1: 00000c75 d2: 28121969 d3: 0002d720 d4: 001cffd5 d5: 003f6000 a0: 374e2010 a1: 000cfc20 Process reboot (pid: 26, stackpage=00293000) Frame format=4 Stack from 00293e9c: 00293fc4 0002d7ee 00000000 000bd897 000e3d78 00000001 00000000 00000058 001695d8 00169528 00000000 003f6000 001cffac 001cfdc4 000373bc 000372de 00169510 003f6000 000372de 00169510 003f6000 00398e10 00398750 003985d0 00398630 00398570 00398810 00293f20 000382aa 00398810 00000000 00000000 00398810 00293f48 00037372 00398810 00000000 00000000 00000001 00292000 001aa000 00292000 000cf990 003f6000 00293f78 00024762 00293f8c 00003214 Call Trace: [<0002d7ee>] [<000bd897>] [<000373bc>] [<000372de>] [<000372de>] [<000382aa>] [<00037372>] [<00024762>] [<000244ee>] [<00024412>] [<0002bd74>] [<0002bc6c>] [<00022286>] Code: bad frame format: 00000000 PC: [<000209e4>] SR: 2714 SP: 00293dd4 a2: 00294000 d0: 00000000 d1: 00000000 d2: 00000000 d3: 0016959c d4: 00294000 d5: 003f6000 a0: 374e2010 a1: 000cfc20 Process reboot (pid: 26, stackpage=00293000) Frame format=4 Stack from 00293e10: 001cffd5 00026ac6 00293e60 00293e40 00020dac 00293e60 000bc757 0029223c 0000001a 00293000 001695d8 fee1dead 00293e54 00020736 000bc49e 00293e60 00000000 00293e94 000221c0 00293e60 00000c75 28121969 0002d720 001cffd5 003f6000 374e2010 000cfc20 001695d8 00000016 ffffffff 00000000 44082700 374e2010 00293e9c 000202b2 00293fc4 0002d7ee 00000000 000bd897 000e3d78 00000001 00000000 00000058 001695d8 00169528 00000000 003f6000 001cffac Call Trace: [<00026ac6>] [<00020dac>] [<000bc757>] [<00020736>] [<000bc49e>] [<000221c0>] [<0002d720>] [<000202b2>] [<0002d7ee>] [<000bd897>] [<000373bc>] [<000372de>] [<000372de>] [<000382aa>] [<00037372>] [<00024762>] [<000244ee>] [<00024412>] [<0002bd74>] [<0002bc6c>] [<00022286>] Code: 3030 2a00 2f00 4879 000b c638 4eb9 0002 6ac6 508f Any hints, tips, slams, welcome. Travis Griggs This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From ilatypov at superbt.com Mon Mar 18 16:20:25 2002 From: ilatypov at superbt.com (Ilguiz Latypov) Date: Mon, 18 Mar 2002 16:20:25 -0500 (EST) Subject: [uClinux-dev] start offset in FLAT image and 20020218 m68-elf-tools Message-ID: Hello, When I switched to the newer m68k-elf-tools cross-compiler from uclinux.org of 20020218, I found that the compiled applications wouldn't start. I am getting the "Illegal instruction" error code. Is that because the uClinux kernel 2.4.6 might not honour the _start offset in the FLAT image, jumping immediately to offset 0 of the .text section? That would explain the above difference. The new linker script produces a 4 byte zeroed space in the beginning of .text. My uClinux and uClibc code is few months old. I attached a piece of disassembly from the "hello, world" C application and a piece of relevant elf2flt.ld. Ilguiz ==================================================================== Disassembly of section .text: 00000000 <_stext-0x4>: 0: 0000 0000 orib #0,%d0 00000004 <_stext>: 4: 4e71 4e71 NqNq 00000008 <_start>: 8: 2a45 moveal %d5,%a5 a: 41f9 0000 40d6 lea 40d6 ,%a0 10: 4ebb 8800 jsr %pc@(12 ,%a0:l) ==================================================================== Linker script from uclinux.org's m68k-elf-tools of 20020218 ENTRY (_start) MEMORY { flatmem : ORIGIN = 0x0, LENGTH = 0x100000 junk : ORIGIN = 0x0, LENGTH = 0x100000 } SECTIONS { .text 0 : { . = . + 4; . = ALIGN(0x4) ; _stext = . ; *(.text) ... ==================================================================== This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From jadb at redhat.com Mon Mar 18 19:23:18 2002 From: jadb at redhat.com (Joe deBlaquiere) Date: Mon, 18 Mar 2002 18:23:18 -0600 Subject: [uClinux-dev] Re: need your help about port uclinux 2.4.17 to KS32C50100 References: Message-ID: <3C9684F6.2010409@redhat.com> Well, yes, trying to access a non-existent coprocessor would cause Bad Things (tm) to happen. proc-arm7tdmi.S is in the 2.0.X tree in CVS on uclinux.org you can find it at: http://cvs.uclinux.org/cgi-bin/cvsweb/uClinux-2.0.x/arch/armnommu/mm/ I really thought the code in fault-armv.c was supposed to handle alignments in a sans-CP15 case. It's generally a little more readable and surely more tested. I don't know if anybody has adapted proc-arm7tdmi.S to the 2.4 kernel, but it can't be that difficult... Heck, it's the best-documented ARM asm code I've ever written (not saying _that_ much). Of course it's pretty convoluted too... -- Joe ? ?? wrote: > Dear Joe, > I traced the kernel with ICE again, here is the output from windows > HyperTerminal: > .... > Backtrace: no frame pointer > Code: e59fc0a4 e59cc000 (ee01cf10) e3a0c013 e121f00c > ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ <--------Notice > here. > Kernel panic: Attempted to kill the idle task! > In idle task - not syncing > Internal error: Oops - undefined instruction: 10e00 > CPU: 0 > pc : [<00012dfc>] lr : [<000140bc>] Not tainted > ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ <-----Notice . > sp : 0016fe44 ip : 0016fe70 fp : 0016fe50 > r10: 00101014 r9 : ef900078 r8 : 0016ffb8 > r7 : 60000013 r6 : 00176430 r5 : 001921f8 r4 : 00000800 > r3 : 00012230 r2 : 00010e00 r1 : 00000000 r0 : 00010f00 > Flags: nZCv IRQs on FIQs on Mode SVC_32 Segment kernel > Control: 0 > Process swapper (pid: 0, stackpage=0016f000) > > It shows when pc=00012dfc the error happens. I had a look at the context > of the address in SDRAM and listed it below: > > Address Context Assemble code > ... > 0x012df4 0xe59cc0a4 ldr r12,0x00012ea0 > 0x012df8 0xe59cc000 ldr r12,[r12,#0] > 0x012dfc 0xee01cf10 mcr p15,0x0,r12,c1,c0,0 > 0x012e00 0xe3a0c013 mov r12,#0x13 > 0x012e04 0xe121f00c msr cpsr_c,r12 > ... > ( I attatched the gif file that showed the context and assemble code in > attachment.) > > It seems no undefined instructions except the "mcr p15,0x0,r12,c1,c0,0". > instructions because KS32C50100-- the CPU we are using is based on > ARM7TDMI, it has no MMU unit and coprocessor at all. > > Is the problem resides here? Thanks for your guide. > > best regards. > > ziming > > >> From: Joe deBlaquiere >> To: ? ?? >> Subject: Re: need your help about port uclinux 2.4.17 to KS32C50100 >> Date: Mon, 18 Mar 2002 09:56:56 -0600 >> >> The unaligned data access handler is present in the 2.4.17 kernel, so >> you shouldn't be having problems with unaligned data access. Regardless, >> if your program is properly linked, you should _NEVER_ be doing an >> instruction fetch from an unaligned address. You've either got a linker >> command file problem causing code to be statically linked to an >> unaligned address, or you're dereferencing a pointer to a function that >> contains garbage data. You should be able to stepi with the ICE and find >> the offending instruction. >> >> -- >> Joe >> > > > _________________________________________________________________ > ???? MSN Explorer?http://explorer.msn.com/lccn/intl.asp? > > ------------------------------------------------------------------------ > > > ------------------------------------------------------------------------ > -- Joe deBlaquiere Senior Architect, Embedded Linux Red Hat, Inc. voice : 256-217-0123 mobile: 256-527-5633 fax : 256-837-3839 jadb at redhat.com This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From artu at 263.net Mon Mar 18 19:42:58 2002 From: artu at 263.net (ziming) Date: Tue, 19 Mar 2002 8:42:58 +0800 Subject: [uClinux-dev] kernel OOPS-alignment problem? mcr instruction problem? Message-ID: <20020319002214.BBFAB1D005F53@smtp.263.net> Dear All, We are porting uclinux 2.4.17 to KS32C50100- a CPU based on ARM7TDMI and have no MMU unit. when boot it the kernel OOPS, below is the output from windows HyperTerminal: ... Backtrace: no frame pointer Code: e59fc0a4 e59cc000 (ee01cf10) e3a0c013 e121f00c ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ <--------Notice here. Kernel panic: Attempted to kill the idle task! In idle task - not syncing Internal error: Oops - undefined instruction: 10e00 CPU: 0 pc : [<00012dfc>] lr : [<000140bc>] Not tainted ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ <-----Notice . sp : 0016fe44 ip : 0016fe70 fp : 0016fe50 r10: 00101014 r9 : ef900078 r8 : 0016ffb8 r7 : 60000013 r6 : 00176430 r5 : 001921f8 r4 : 00000800 r3 : 00012230 r2 : 00010e00 r1 : 00000000 r0 : 00010f00 Flags: nZCv IRQs on FIQs on Mode SVC_32 Segment kernel Control: 0 Process swapper (pid: 0, stackpage=0016f000) It shows when pc=00012dfc the error happens. I had a look at the context of the address in SDRAM and listed it below: Address Context Assemble code .. 0x012df4 0xe59cc0a4 ldr r12,0x00012ea0 0x012df8 0xe59cc000 ldr r12,[r12,#0] 0x012dfc 0xee01cf10 mcr p15,0x0,r12,c1,c0,0 0x012e00 0xe3a0c013 mov r12,#0x13 0x012e04 0xe121f00c msr cpsr_c,r12 .. It seems no undefined instructions except the "mcr p15,0x0,r12,c1,c0,0". instructions because KS32C50100-- the CPU we are using is based on ARM7TDMI, it has no MMU unit and coprocessor at all. Is the problem resides here? Thanks for any suggestions. best regards. ziming artu at 263.net This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From david at bdt.com Mon Mar 18 20:48:59 2002 From: david at bdt.com (David Beckemeyer) Date: Mon, 18 Mar 2002 17:48:59 -0800 Subject: [uClinux-dev] About max3100 on VZ board with SPI1 In-Reply-To: <20020307013234.61086.qmail@web13207.mail.yahoo.com>; from baiwei99@yahoo.com.cn on Thu, Mar 07, 2002 at 09:32:34AM +0800 References: <20020307013234.61086.qmail@web13207.mail.yahoo.com> Message-ID: <20020318174859.B1173@rhea.toyz.org> Please let me know if you get this ported to your platform and also if you get interrupt driven output working. Thanks. On Thu, Mar 07, 2002 at 09:32:34AM +0800, bai wei wrote: > My program is very easy and I want to write my driver > referrence the max311Xserial.c. You can get this > program at here: > > http://www.bdt.com/david/ucsimm/max311Xserial-0.20-R1.tar.gz. > Send me some message too after your success! > > > --- Jonathan Schmidt ????????> > I too have a MAX3100 on the SPI2 interface on my VZ > > board. I have not > > started developing drivers for it yet, but I do need > > to start within the > > next week or so. Would it be possible to see what > > you have written for a > > driver so far? That would be really helpful to me! > > Thanks. > > > > Jonathan > > > > > > >From: bai wei > > >Reply-To: uclinux-dev at uclinux.org > > >To: uclinux-dev at uclinux.org > > >Subject: [uClinux-dev] About max3100 on VZ board > > with SPI1 > > >Date: Wed, 6 Mar 2002 18:17:43 +0800 (CST) > > >MIME-Version: 1.0 > > >Received: from [216.57.57.93] by hotmail.com (3.2) > > with ESMTP id > > >MHotMailBE4F3C8C00A84004325ED839395DCA5E0; Wed, 06 > > Mar 2002 02:38:22 -0800 > > >Received: (from mdom at localhost)by > > uclinux-dev-listclinux.org > > >(8.11.3/8.11.3/SuSE Linux 8.11.1-0.5) id > > g26AHpj21830for uclinux-dev-list; > > >Wed, 6 Mar 2002 05:17:51 -0500 > > >From owner-uclinux-dev at uclinux.org Wed, 06 Mar 2002 > > 02:39:02 -0800 > > >Message-ID: > > <20020306101743.63967.qmail at web13202.mail.yahoo.com> > > >Sender: owner-uclinux-dev at uclinux.org > > >Precedence: bulk > > > > > >hi: > > > I just write a dirver for the UART by SPI1 of > > >68VZ328 ,and I was using max3100 . First I writed a > > >program of asm. Just to config the max3100 and read > > >the config back. I write a 0xcc0a to the > > max3100,but > > >when I write a 0x4000 to read the configuration > > ,also > > >a 0x4000 backed. Before I write data to the max3100 > > I > > >have writing a 0 to the cs,and after the transfer > > >writing a 1. > > > Why I can not read the configuration? > > > > > >think you! > > > > > > >_________________________________________________________ > > >Do You Yahoo!? > > >???????? Yahoo! ???? > > >http://survey.yahoo.com/cgi-bin/Express/survey.cgi > > >This message resent by the uclinux-dev at uclinux.org > > list server > > >http://www.uClinux.org/ > > > > > > > > > > > _________________________________________________________________ > > Chat with friends online, try MSN Messenger: > > http://messenger.msn.com > > > > This message resent by the uclinux-dev at uclinux.org > > list server http://www.uClinux.org/ > > _________________________________________________________ > Do You Yahoo!? > ???????? Yahoo! ???? > http://survey.yahoo.com/cgi-bin/CNITsolution/survey.cgi > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From gerg at snapgear.com Mon Mar 18 21:44:26 2002 From: gerg at snapgear.com (Greg Ungerer) Date: Tue, 19 Mar 2002 12:44:26 +1000 Subject: [uClinux-dev] reboot on the coldfire... References: <3C965888.9040606@keyww.com> Message-ID: <3C96A60A.5080204@snapgear.com> Hi Travis, Travis Griggs wrote: > Is this known to not work? It works on some platforms... If it doesn't work it will almost certainly be the code in linux/include/asm-m68knommu/system.h broken for your platform. Regards Greg > I tried reboot, both the sash builtin and the busy box one. They both > blow up similiarily. Some sort of kernel trap I think. Here's what > Minicom spits back when I try it: > > /> reboot > Nov 30 00:02:05 dhcpcd[17]: terminating on signal Restarting system. > bad frame format: 00000000 > PC: [<374e2010>] > SR: 2700 SP: 00293e60 a2: 001695d8 > d0: 00000016 d1: 00000c75 d2: 28121969 d3: 0002d720 > d4: 001cffd5 d5: 003f6000 a0: 374e2010 a1: 000cfc20 > Process reboot (pid: 26, stackpage=00293000) > Frame format=4 Stack from 00293e9c: > 00293fc4 0002d7ee 00000000 000bd897 000e3d78 00000001 00000000 > 00000058 > 001695d8 00169528 00000000 003f6000 001cffac 001cfdc4 000373bc > 000372de > 00169510 003f6000 000372de 00169510 003f6000 00398e10 00398750 > 003985d0 > 00398630 00398570 00398810 00293f20 000382aa 00398810 00000000 > 00000000 > 00398810 00293f48 00037372 00398810 00000000 00000000 00000001 > 00292000 > 001aa000 00292000 000cf990 003f6000 00293f78 00024762 00293f8c > 00003214 > Call Trace: > [<0002d7ee>] [<000bd897>] [<000373bc>] [<000372de>] > [<000372de>] [<000382aa>] [<00037372>] [<00024762>] > [<000244ee>] [<00024412>] [<0002bd74>] [<0002bc6c>] > [<00022286>] > Code: bad frame format: 00000000 > PC: [<000209e4>] > SR: 2714 SP: 00293dd4 a2: 00294000 > d0: 00000000 d1: 00000000 d2: 00000000 d3: 0016959c > d4: 00294000 d5: 003f6000 a0: 374e2010 a1: 000cfc20 > Process reboot (pid: 26, stackpage=00293000) > Frame format=4 Stack from 00293e10: > 001cffd5 00026ac6 00293e60 00293e40 00020dac 00293e60 000bc757 > 0029223c > 0000001a 00293000 001695d8 fee1dead 00293e54 00020736 000bc49e > 00293e60 > 00000000 00293e94 000221c0 00293e60 00000c75 28121969 0002d720 > 001cffd5 > 003f6000 374e2010 000cfc20 001695d8 00000016 ffffffff 00000000 > 44082700 > 374e2010 00293e9c 000202b2 00293fc4 0002d7ee 00000000 000bd897 > 000e3d78 > 00000001 00000000 00000058 001695d8 00169528 00000000 003f6000 > 001cffac > Call Trace: > [<00026ac6>] [<00020dac>] [<000bc757>] [<00020736>] > [<000bc49e>] [<000221c0>] [<0002d720>] [<000202b2>] > [<0002d7ee>] [<000bd897>] [<000373bc>] [<000372de>] > [<000372de>] [<000382aa>] [<00037372>] [<00024762>] > [<000244ee>] [<00024412>] [<0002bd74>] [<0002bc6c>] > [<00022286>] > Code: 3030 2a00 2f00 4879 000b c638 4eb9 0002 6ac6 508f > > Any hints, tips, slams, welcome. > > Travis Griggs > > This message resent by the uclinux-dev at uclinux.org list server > http://www.uClinux.org/ > -- ------------------------------------------------------------------------ Greg Ungerer -- Chief Software Wizard EMAIL: gerg at snapgear.com SnapGear PHONE: +61 7 3435 2888 825 Stanley St, FAX: +61 7 3891 3630 Woolloongabba, QLD, 4102, Australia WEB: www.snapgear.com This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From gerg at snapgear.com Mon Mar 18 21:46:11 2002 From: gerg at snapgear.com (Greg Ungerer) Date: Tue, 19 Mar 2002 12:46:11 +1000 Subject: [uClinux-dev] start offset in FLAT image and 20020218 m68-elf-tools References: Message-ID: <3C96A673.9050902@snapgear.com> Hi Ilguiz, Ilguiz Latypov wrote: > When I switched to the newer m68k-elf-tools cross-compiler from > uclinux.org of 20020218, I found that the compiled applications wouldn't > start. I am getting the "Illegal instruction" error code. > > Is that because the uClinux kernel 2.4.6 might not honour the _start > offset in the FLAT image, jumping immediately to offset 0 of the .text > section? That would explain the above difference. The new linker script > produces a 4 byte zeroed space in the beginning of .text. Certainly sounds right. Try getting the latest binfmt_flat.c from CVS and plugging that into your source tree. Regards Greg > My uClinux and uClibc code is few months old. I attached a piece of > disassembly from the "hello, world" C application and a piece of relevant > elf2flt.ld. > > Ilguiz > > ==================================================================== > Disassembly of section .text: > > 00000000 <_stext-0x4>: > 0: 0000 0000 orib #0,%d0 > > 00000004 <_stext>: > 4: 4e71 4e71 NqNq > > 00000008 <_start>: > 8: 2a45 moveal %d5,%a5 > a: 41f9 0000 40d6 lea 40d6 ,%a0 > 10: 4ebb 8800 jsr %pc@(12 ,%a0:l) > > ==================================================================== > Linker script from uclinux.org's m68k-elf-tools of 20020218 > > ENTRY (_start) > > MEMORY { > flatmem : ORIGIN = 0x0, LENGTH = 0x100000 > junk : ORIGIN = 0x0, LENGTH = 0x100000 > } > > SECTIONS { > .text 0 : { > . = . + 4; > . = ALIGN(0x4) ; > _stext = . ; > *(.text) > ... > ==================================================================== > > > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ > > -- ------------------------------------------------------------------------ Greg Ungerer -- Chief Software Wizard EMAIL: gerg at snapgear.com SnapGear PHONE: +61 7 3435 2888 825 Stanley St, FAX: +61 7 3891 3630 Woolloongabba, QLD, 4102, Australia WEB: www.snapgear.com This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From christian.opel at brooks.com Tue Mar 19 02:21:33 2002 From: christian.opel at brooks.com (Christian Opel) Date: 19 Mar 2002 08:21:33 +0100 Subject: [uClinux-dev] m68k-elf-g++: compiling c++ files. In-Reply-To: <013101c1cea1$9b347f50$2200a8c0@flamenco> References: <013101c1cea1$9b347f50$2200a8c0@flamenco> Message-ID: <1016522495.1310.10.camel@christian> It is not necessary to compile the toolchain yourself! Try out a newer toolchain! I`ve had the same problems with rel 20010228. After installing the m68k-elf-toolchain 20010716 compiling and linking worked fine both. You said it`s a linker error! probably ut would work if you install the c++ libraries and headers in the correct directories and retry compiling. This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From mathias.fritzson at mecel.se Tue Mar 19 02:58:42 2002 From: mathias.fritzson at mecel.se (mathias.fritzson at mecel.se) Date: Tue, 19 Mar 2002 08:58:42 +0100 Subject: [uClinux-dev] Compiling a romfs Message-ID: <200203190758.g2J7wRk09907@clinux.org> Hi again We got ourselves a problem here, using Cygwin bash we can't process the romfs-inst.sh Do we got anyone that has successfully fixed this problem... Or any scriptcunning person that can analyze what's gone wrong. TIA /Mathias Output from compilation, the romfs dir is created but no files is moved into the tree -------------------------------8<--------------- [ -d /opt/uClinux-dist/romfs/$i ] || mkdir -p /opt/uClinux-dist/romfs for i in bin dev etc home lib mnt proc usr var home/httpd home/httpd/cgi-bin; do \ [ -d /opt/uClinux-dist/romfs/$i ] || mkdir -p /opt/uClinux-dist/romfs/$i ; \ done for i in tty,c,5,0 console,c,5,1 cua0,c,5,64 cua1,c,5,65 mem,c,1, 1 kmem,c,1,2 null,c,1,3 ram0,b,1,0 ram1,b,1,1 ptyp0,c,2,0 ptyp1,c,2,1 ptyp2,c,2,2 ptyp3,c,2,3 ptyp4,c,2,4 ptyp5,c,2,5 ptyp6,c,2,6 ptyp7,c,2,7 ptyp8,c,2,8 ptyp9,c,2,9 ptypa,c,2,10 ptypb,c,2,11 ptypc,c,2,12 ptypd,c,2,13 ptype,c,2,14 ptypf,c,2,15 r om0,b,31,0 rom1,b,31,1 rom2,b,31,2 rom3,b,31,3 rom4,b,31,4 rom 5,b,31,5 rom6,b,31,6 rom7,b,31,7 rom8,b,31,8 rom9,b,31,9 tty0,c,4 ,0 tty1,c,4,1 tty2,c,4,2 tty3,c,4,3 ttyS0,c,4,64 ttyS1,c,4,6 5 ttyp0,c,3,0 ttyp1,c,3,1 ttyp2,c,3,2 ttyp3,c,3,3 ttyp4,c,3,4 ttyp5,c,3,5 ttyp6,c,3,6 ttyp7,c,3,7 ttyp8,c,3,8 ttyp9,c,3,9 ttypa,c,3,10 ttypb,c,3,11 ttypc,c,3,12 ttypd,c,3,13 ttype,c,3,14 ttypf,c,3,15 hda,b,3,0 hdb,b,3,64 zero,c,1,5 random,c,1,8 uran dom,c,1,9 modem,c,24,64 cuam,c,25,64 ipsec,c,36,10 ledman,c,126,0; do \ touch /opt/uClinux-dist/romfs/dev/@$i; \ done for i in ../../Generic/cgi; do make -C $i romfs || exit ; done make[3]: Entering directory `/opt/uClinux-dist/vendors/Generic/cgi' romfs-inst.sh ../httpd /home/httpd expr: syntax error rm: `/opt/uClinux-dist/romfs/home/httpd' is a directory cp: cannot stat `httpd': No such file or directory romfs-inst.sh /home/httpd/cgi-bin/cgi_demo expr: syntax error /opt/uClinux-dist/tools/romfs-inst.sh: [options] [src] dst -v : output actions performed. -e env-var : only take action if env-var is set to "y". -o option : only take action if option is set to "y". -p perms : chmod style permissions for dst. -a text : append text to dst. -l link : dst is a link to 'link'. -s sym-link : dst is a sym-link to 'sym-link'. if "src" is not provided, basename is run on dst to determine the source in the current directory. multiple -e and -o options are ANDed together. To achieve an OR affect use a single -e/-o with 1 or more y/n/"" chars in the condition. if src is a directory, everything in it is copied recursively to dst with special files removed (currently CVS dirs). make[3]: *** [romfs] Error 1 make[3]: Leaving directory `/opt/uClinux-dist/vendors/Generic/cgi' --------------------------------------->8---------------------------- This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From stefan.jonsson at faltcom.se Tue Mar 19 03:29:31 2002 From: stefan.jonsson at faltcom.se (Stefan Jonsson) Date: Tue, 19 Mar 2002 09:29:31 +0100 Subject: SV: [uClinux-dev] Download and boot on EB40 In-Reply-To: <01C1CE92.A64E9740@smithwicks.softsys.co.at> Message-ID: Thank you Erwin, I will look at those things right now :-). http://cvs.home.at/ This seems to be made for EB01 and/or EB63, will it work for EB40? Thanks again ... Reagards, Stefan Jonsson. -----Ursprungligt meddelande----- Fr?n: owner-uclinux-dev at uclinux.org [mailto:owner-uclinux-dev at uclinux.org]F?r Erwin Authried Skickat: den 18 mars 2002 15:36 Till: 'uclinux-dev at uclinux.org' ?mne: Re: [uClinux-dev] Download and boot on EB40 Stefan Jonsson[SMTP:stefan.jonsson at faltcom.se] wrote: > Hello list, > > I have an EB40+MEC01 memory extension card. Atmels tools though is for > windows. I am using linux and want to put uClinux on the EB40+MEC. I wonder > about: > > 1) I have seen something about modifying the > arch/armnommu/kernel/head-arm-atmel.S (that mail was from june 2001, is this > still valid?). That's the code that configures the EBI for the additional memory. I think it shouldn't go into the kernel, it's the bootloader's task to configure the EBI and remap the memory. > > 2) how to download an image to target and how to program it into flash (on > MEC01)? > I get some kind of contact with angel debug monitor but don't know how to > use it for my purpose. > > 3) how to boot from the downloaded image (from MEC)? (Or just how to execute > it at least?) > Have a look at my arm-boot package. It contains the armtool monitor that allows you to initialize your board, download binary images, execute them, and program the flash. http://cvs.home.at/ http://uclinux.home.at/midori/apps/armtool-0.1_3.mlz -Erwin This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From stefan.jonsson at faltcom.se Tue Mar 19 03:50:49 2002 From: stefan.jonsson at faltcom.se (Stefan Jonsson) Date: Tue, 19 Mar 2002 09:50:49 +0100 Subject: SV: [uClinux-dev] Download and boot on EB40 In-Reply-To: <01C1CE92.A64E9740@smithwicks.softsys.co.at> Message-ID: Hello again, I have some additional questions to this ... Do I have to have all of Midori along with this armtool to run/(use) it? This seems to be for JTAG use. I do not have such cable. I use serial A (on EB40). Is it possible to use this (or anything else that would serve my purposes) with the serial connection? If not, how do i make and connect the JTAG cable? Regards, Stefan Jonsson. -----Ursprungligt meddelande----- Fr?n: owner-uclinux-dev at uclinux.org [mailto:owner-uclinux-dev at uclinux.org]F?r Erwin Authried Skickat: den 18 mars 2002 15:36 Till: 'uclinux-dev at uclinux.org' ?mne: Re: [uClinux-dev] Download and boot on EB40 Stefan Jonsson[SMTP:stefan.jonsson at faltcom.se] wrote: > Hello list, > > I have an EB40+MEC01 memory extension card. Atmels tools though is for > windows. I am using linux and want to put uClinux on the EB40+MEC. I wonder > about: > > 1) I have seen something about modifying the > arch/armnommu/kernel/head-arm-atmel.S (that mail was from june 2001, is this > still valid?). That's the code that configures the EBI for the additional memory. I think it shouldn't go into the kernel, it's the bootloader's task to configure the EBI and remap the memory. > > 2) how to download an image to target and how to program it into flash (on > MEC01)? > I get some kind of contact with angel debug monitor but don't know how to > use it for my purpose. > > 3) how to boot from the downloaded image (from MEC)? (Or just how to execute > it at least?) > Have a look at my arm-boot package. It contains the armtool monitor that allows you to initialize your board, download binary images, execute them, and program the flash. http://cvs.home.at/ http://uclinux.home.at/midori/apps/armtool-0.1_3.mlz -Erwin This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From mathias.fritzson at mecel.se Tue Mar 19 04:03:32 2002 From: mathias.fritzson at mecel.se (mathias.fritzson at mecel.se) Date: Tue, 19 Mar 2002 10:03:32 +0100 Subject: [uClinux-dev] Compiling a romfs Message-ID: <200203190903.g2J93Hk10513@clinux.org> I should add that $OPTIND seems to contain nothing.. /Mathias Mathias Fritzson wrote: >Hi again > >We got ourselves a problem here, using Cygwin bash we can't process the >romfs-inst.sh > >Do we got anyone that has successfully fixed this problem... Or any >scriptcunning person that can analyze what's gone wrong. > >TIA > >/Mathias > >Output from compilation, the romfs dir is created but no files is moved >into the tree >-------------------------------8<--------------- >[ -d /opt/uClinux-dist/romfs/$i ] || mkdir -p /opt/uClinux-dist/romfs >for i in bin dev etc home lib mnt proc usr var home/httpd home/httpd/cgi-bin; do > \ > [ -d /opt/uClinux-dist/romfs/$i ] || mkdir -p /opt/uClinux-dist/romfs/$i >; \ >done >for i in tty,c,5,0 console,c,5,1 cua0,c,5,64 cua1,c,5,65 mem,c,1, >1 kmem,c,1,2 null,c,1,3 ram0,b,1,0 ram1,b,1,1 ptyp0,c,2,0 >ptyp1,c,2,1 ptyp2,c,2,2 ptyp3,c,2,3 ptyp4,c,2,4 ptyp5,c,2,5 > ptyp6,c,2,6 ptyp7,c,2,7 ptyp8,c,2,8 ptyp9,c,2,9 ptypa,c,2,10 > ptypb,c,2,11 ptypc,c,2,12 ptypd,c,2,13 ptype,c,2,14 ptypf,c,2,15 r >om0,b,31,0 rom1,b,31,1 rom2,b,31,2 rom3,b,31,3 rom4,b,31,4 rom >5,b,31,5 rom6,b,31,6 rom7,b,31,7 rom8,b,31,8 rom9,b,31,9 tty0,c,4 >,0 tty1,c,4,1 tty2,c,4,2 tty3,c,4,3 ttyS0,c,4,64 ttyS1,c,4,6 >5 ttyp0,c,3,0 ttyp1,c,3,1 ttyp2,c,3,2 ttyp3,c,3,3 ttyp4,c,3,4 >ttyp5,c,3,5 ttyp6,c,3,6 ttyp7,c,3,7 ttyp8,c,3,8 ttyp9,c,3,9 > ttypa,c,3,10 ttypb,c,3,11 ttypc,c,3,12 ttypd,c,3,13 ttype,c,3,14 > ttypf,c,3,15 hda,b,3,0 hdb,b,3,64 zero,c,1,5 random,c,1,8 uran >dom,c,1,9 modem,c,24,64 cuam,c,25,64 ipsec,c,36,10 ledman,c,126,0; do \ > touch /opt/uClinux-dist/romfs/dev/@$i; \ >done >for i in ../../Generic/cgi; do make -C $i romfs || exit ; done >make[3]: Entering directory `/opt/uClinux-dist/vendors/Generic/cgi' >romfs-inst.sh ../httpd /home/httpd > >expr: syntax error >rm: `/opt/uClinux-dist/romfs/home/httpd' is a directory >cp: cannot stat `httpd': No such file or directory >romfs-inst.sh /home/httpd/cgi-bin/cgi_demo > >expr: syntax error >/opt/uClinux-dist/tools/romfs-inst.sh: [options] [src] dst > -v : output actions performed. > -e env-var : only take action if env-var is set to "y". > -o option : only take action if option is set to "y". > -p perms : chmod style permissions for dst. > -a text : append text to dst. > -l link : dst is a link to 'link'. > -s sym-link : dst is a sym-link to 'sym-link'. > > if "src" is not provided, basename is run on dst to determine the > source in the current directory. > > multiple -e and -o options are ANDed together. To achieve an OR affect > use a single -e/-o with 1 or more y/n/"" chars in the condition. > > if src is a directory, everything in it is copied recursively to dst > with special files removed (currently CVS dirs). >make[3]: *** [romfs] Error 1 >make[3]: Leaving directory `/opt/uClinux-dist/vendors/Generic/cgi' >--------------------------------------->8---------------------------- > >This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From sheeliang at lit.org.sg Tue Mar 19 04:14:15 2002 From: sheeliang at lit.org.sg (sheeliang at lit.org.sg) Date: Tue, 19 Mar 2002 17:14:15 +0800 Subject: [uClinux-dev] Network cards on Integrator AP Message-ID: <20020319171343.A9118@localhost.localdomain> Wondering if anybody has any experience getting network cards to work on Integrator AP, which uses the PCI V3 chipset? I have tried pcnet32 and 3com cards without avail. Initialization runs ok, ie, I can see the correct ethernet addresses, IRQ assigned, ioport, etc on startup, but actual pinging on both sides fail. The core is a ARM7TDMI running uClinux-2.4.17. Help, anyone? This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From eauth at softsys.co.at Tue Mar 19 04:44:20 2002 From: eauth at softsys.co.at (Erwin Authried) Date: Tue, 19 Mar 2002 10:44:20 +0100 Subject: [uClinux-dev] Download and boot on EB40 In-Reply-To: Message-ID: > -----Urspr?ngliche Nachricht----- > Von: owner-uclinux-dev at uclinux.org > [mailto:owner-uclinux-dev at uclinux.org]Im Auftrag von Stefan Jonsson > Gesendet: Dienstag, 19. M?rz 2002 09:30 > An: uclinux-dev at uclinux.org > Betreff: SV: [uClinux-dev] Download and boot on EB40 > > > Thank you Erwin, > > I will look at those things right now :-). > > http://cvs.home.at/ > This seems to be made for EB01 and/or EB63, will it work for EB40? Of course, The EB01 used the AT91M40400, and the EB40 is the same board, with the newer AT91M40800. The CPUs are almost equivalent the AT91M40800 has more on-chip SRAM. Actually, the monitor will probably work with all CPUs from the AT91 series. Depending on your target hardware, you will have to modify the flash programming routines. -Erwin This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From eauth at softsys.co.at Tue Mar 19 04:57:05 2002 From: eauth at softsys.co.at (Erwin Authried) Date: Tue, 19 Mar 2002 10:57:05 +0100 Subject: [uClinux-dev] Download and boot on EB40 In-Reply-To: Message-ID: > -----Urspr?ngliche Nachricht----- > Von: owner-uclinux-dev at uclinux.org > [mailto:owner-uclinux-dev at uclinux.org]Im Auftrag von Stefan Jonsson > Gesendet: Dienstag, 19. M?rz 2002 09:51 > An: uclinux-dev at uclinux.org > Betreff: SV: [uClinux-dev] Download and boot on EB40 > > > Hello again, > > I have some additional questions to this ... > > Do I have to have all of Midori along with this armtool to run/(use) it? > No, that's not necessary. The .mlz files are ordinary gzipped tar archives, with a build script and some extra files that are required by the Midori build system. In this case, just go into the src/ directory and type "make" to build armtool. > This seems to be for JTAG use. I do not have such cable. I use > serial A (on > EB40). Is it possible to use this (or anything else that would serve my > purposes) with the serial connection? > > If not, how do i make and connect the JTAG cable? > There are many different schematics available. They connect to the PC's parallel port. Look into the JTAG folder at http://uclinux.home.at. There's a cable with schematics in "Welcome to Jtag-Arm9", or take the schematics from "Altera Byteblaster Cable". Don't forget to change the pin definitions in jtag_io.c. The advantage over serial communication is that you don't require any boot/monitor software on your board. Happy hacking! -Erwin This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From chris at cjx.com Tue Mar 19 05:42:35 2002 From: chris at cjx.com (Chris Allen) Date: Tue, 19 Mar 2002 10:42:35 +0000 Subject: [uClinux-dev] Looking for LCD module source in the UK In-Reply-To: <20020318172756.A16455@cjx.com>; from chris@cjx.com on Mon, Mar 18, 2002 at 05:27:56PM +0000 References: <20020318172756.A16455@cjx.com> Message-ID: <20020319104235.A21090@cjx.com> On Mon, Mar 18, 2002 at 05:27:56PM +0000, Chris Allen wrote: > I am looking for a source for a reasonably cheap LCD > module for use with ucLinux on the ucSimm and similar. > Ideally with a built-in touch screen, or an easily > attachable touch screen.. > Many thanks to those who have replied. I have ordered a couple of the Samsung LCDs from EarthLCD. This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From kovacs_kernig at hotmail.com Tue Mar 19 07:22:54 2002 From: kovacs_kernig at hotmail.com (kernig kovacs) Date: Tue, 19 Mar 2002 20:22:54 +0800 Subject: [uClinux-dev] Does linux mtd subsystem support INTEL TE28F160 flash? References: <20020318172756.A16455@cjx.com> <20020319104235.A21090@cjx.com> Message-ID: Hi all Does linux mtd subsystem support INTEL TE28F160 flash? Regards, Kovacs This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From superwen at 263.net Tue Mar 19 07:04:36 2002 From: superwen at 263.net (Kevin) Date: Tue, 19 Mar 2002 20:4:36 +0800 Subject: [uClinux-dev] kernel command line is NULL Message-ID: <20020319120151.55CE21D89B6C1@smtp.263.net> Hi, All, The kernel command line I got is NULL(please see below), I'm wondering if I can add it manually? Or where should I check it? Please advise, thank you very much -------------------------------- here I attached the message I got from uart. KERNEL -> TEXT=0x1000400-0x104e308 DATA=0x020000-0x03c000 BSS=0x03c000-0x04f198 CAT ROMARRAY! KERNEL -> ROMFS=0x106a308-0x1125f78 MEM=0x050000-0x800000 STACK=0x800000-0x80000 0 On node 0 totalpages: 2048 zone(0): 0 pages. zone(1): 2048 pages. zone(2): 0 pages. Done setup_arch init/main.c:line 561: Setup_Arch Done Kernel command line: -------This looks not good. Calibrating delay loop... 1.40 BogoMIPS init/main.c:line 586: Calibrate_delay.Done Memory available: 7728k/8064k RAM, 0k/0k ROM (311k kernel code, 188k data) Dentry-cache hash table entries: 1024 (order: 1, 8192 bytes) Inode-cache hash table entries: 512 (order: 0, 4096 bytes) Mount-cache hash table entries: 512 (order: 0, 4096 bytes) Buffer-cache hash table entries: 1024 (order: 0, 4096 bytes) Page-cache hash table entries: 2048 (order: 1, 8192 bytes) POSIX conformance testing by UNIFIX <6>Linux NET4.0 for Linux 2.4 <6>Based upon Swansea University Computer Society NET3.039 Starting kswapd pty: 256 Unix98 ptys configured MC68328 serial driver version 1.00 ttyS0 at 0xfffff900 (irq = 64) is a builtin MC68328 UART block: 64 slots per queue, batch=16 RAMDISK driver initialized: 16 RAM disks of 4096K size 1024 blocksize Blkmem copyright 1998,1999 D. Jeff Dionne Blkmem copyright 1998 Kenneth Albanowski Blkmem 1 disk images: 0: 106A308-1126307 [VIRTUAL 106A308-1126307] (RO) VFS: Mounted root (romfs filesystem) readonly. Warning: unable to open an initial console. This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From gmenie at akamai.com Tue Mar 19 08:20:28 2002 From: gmenie at akamai.com (Menie, Georges) Date: Tue, 19 Mar 2002 08:20:28 -0500 Subject: [uClinux-dev] kernel command line is NULL Message-ID: It is not a problem, the command line is a way to setup extra parameters for the kernel, You may not need it. The code which setup this command line is in the config_BSP() function (see the config.c file for your platform). An empty command line is ok. Georges > -----Original Message----- > From: Kevin [mailto:superwen at 263.net] > Sent: lundi 18 mars 2002 19:01 > To: uclinux > Subject: [uClinux-dev] kernel command line is NULL > > > Hi, All, > The kernel command line I got is NULL(please see below), I'm > wondering if I can add it manually? Or where should I check it? > Please advise, thank you very much > > -------------------------------- > here I attached the message I got from uart. > > KERNEL -> TEXT=0x1000400-0x104e308 DATA=0x020000-0x03c000 > BSS=0x03c000-0x04f198 > CAT ROMARRAY! > KERNEL -> ROMFS=0x106a308-0x1125f78 MEM=0x050000-0x800000 > STACK=0x800000-0x80000 > 0 > On node 0 totalpages: 2048 > zone(0): 0 pages. > zone(1): 2048 pages. > zone(2): 0 pages. > Done setup_arch > init/main.c:line 561: Setup_Arch Done > Kernel command line: -------This looks > not good. > Calibrating delay loop... 1.40 BogoMIPS > init/main.c:line 586: Calibrate_delay.Done > Memory available: 7728k/8064k RAM, 0k/0k ROM (311k kernel > code, 188k data) > Dentry-cache hash table entries: 1024 (order: 1, 8192 bytes) > Inode-cache hash table entries: 512 (order: 0, 4096 bytes) > Mount-cache hash table entries: 512 (order: 0, 4096 bytes) > Buffer-cache hash table entries: 1024 (order: 0, 4096 bytes) > Page-cache hash table entries: 2048 (order: 1, 8192 bytes) > POSIX conformance testing by UNIFIX > <6>Linux NET4.0 for Linux 2.4 > <6>Based upon Swansea University Computer Society NET3.039 > Starting kswapd > pty: 256 Unix98 ptys configured > MC68328 serial driver version 1.00 > ttyS0 at 0xfffff900 (irq = 64) is a builtin MC68328 UART > block: 64 slots per queue, batch=16 > RAMDISK driver initialized: 16 RAM disks of 4096K size 1024 blocksize > Blkmem copyright 1998,1999 D. Jeff Dionne > Blkmem copyright 1998 Kenneth Albanowski > Blkmem 1 disk images: > 0: 106A308-1126307 [VIRTUAL 106A308-1126307] (RO) > VFS: Mounted root (romfs filesystem) readonly. > Warning: unable to open an initial console. > > > > This message resent by the uclinux-dev at uclinux.org list > server http://www.uClinux.org/ > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From gmenie at akamai.com Tue Mar 19 08:44:20 2002 From: gmenie at akamai.com (Menie, Georges) Date: Tue, 19 Mar 2002 08:44:20 -0500 Subject: [uClinux-dev] Debugging on the 68VZ328 using 2.4.10 and elf Message-ID: for the record: the patch is breaking (at least) the trap #2 processing on the ucdimm. I'm going to send a better patch in a couple of days... sorry for that, Georges > -----Original Message----- > From: Menie, Georges [mailto:gmenie at akamai.com] > Sent: dimanche 17 mars 2002 11:40 > To: 'uclinux-dev at uclinux.org' > Subject: RE: [uClinux-dev] Debugging on the 68VZ328 using > 2.4.10 and elf > > > Here is a patch for the 68328/68EZ328/68VZ328 arch which will > allow gdb and > gdbserver to run together. It adds the vector number on the > stack like does > a 68010 (format=0). It slow down a little bit the exception > handling, so I > did > add a config switch to enable/disable it : "Simulate > vectorized exceptions > (debugging user apps on 68000)". I did test it on the > DragonEngine port, > but the following port should be checked : > 68328 > 68328/pilot > 68EZ328 > 68VZ328/ucdimm > The patch allow for a normal gdb/gdbserver session with > breakpoints but > displaying variables shows sometimes wrong data, I have to check this > but I don't think it related to the exception processing... > The patch is against the latest (2002-03-16) CVS tree for linux-2.4.x > > Cheers, > Georges > > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From siddons at bnl.gov Tue Mar 19 09:42:35 2002 From: siddons at bnl.gov (D. Peter Siddons) Date: Tue, 19 Mar 2002 09:42:35 -0500 Subject: [uClinux-dev] goram References: <3C965888.9040606@keyww.com> Message-ID: <3C974E5B.72A13B2B@bnl.gov> I am trying to understand the Lineo bootloader which is supplied with the uCdimm and uCsimm, but there is not much said in the manual about it. In particular, I am unable to execute a program using the 'goram' command. If I understand correctly, this command executes code starting at 0x20400. I tried to load a program into this location and execute it, without success. A memory dump of this area verifies that the instruction codes I intended are indeed loaded properly. THe program is essentially just the first few lines of crt0_ram.S, just as a test to get the loader right: movew #0x2700, %sr /* Exceptions off! */ moveq #70, %d7 /* 'F' */ moveb %d7,0xfffff907 /* No absolute addresses */ pclp1: movew 0xfffff906, %d7 andw #0x2000, %d7 beq pclp1 ...... which I would expect to issue a 'F' character from the serial port. Dumping memory shows: B$ md 20400 00020400: 46 fc 27 00 7e 46 11 c7 f9 07 3e 38 f9 06 02 47 F.'.~F....>8...G which looks OK. Executing 'goram gives: B$ goram go from RAM! illegal instruction PC[10c00400] SSP[fffffffa] USP[c4ff0718] BootstrapSP[0001fffe] D0 D1 D2 D3 D4 D5 D6 D7 00000000 0000000d 00000001 ffffff81 00000000 00000000 eeefa819 e5c44184 A0 A1 A2 A3 A4 A5 A6 00000000 00001e94 0000453c 00004534 007fff18 007fff98 007fff98 So what am I doing wrong here? As far as I can see, the ram-based kernel starts in the same way. If I build a ram kernel and run it, it eventually crashes, but certainly gets further than this! (why the ram kernel crashes is another question....). Since the serial port is already configured by the bootloader (or I couldn't do the above) I shouldn't need to redo that, and the ram control and memory map is already configured by the same argument, so what is wrong? Pete. -- ----------------- D. Peter Siddons National Synchrotron Light Source, Bldg. 725D Brookhaven National Laboratory Upton New York 11973 Email: siddons at bnl.gov This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From juergen.mueller at muellerhydraulik.de Tue Mar 19 10:05:18 2002 From: juergen.mueller at muellerhydraulik.de (Juergen Mueller) Date: Tue, 19 Mar 2002 16:05:18 +0100 Subject: AW: [uClinux-dev] Flash Write In-Reply-To: Message-ID: Yes I've done. But it's not quite clear how to use jffs with mtd and blkmem. I think I have to part my flash Can somone give me advise. -----Ursprungliche Nachricht----- Von: owner-uclinux-dev at uclinux.org [mailto:owner-uclinux-dev at uclinux.org]Im Auftrag von Kendrick Hamilton Gesendet: Montag, 18. Marz 2002 15:42 An: uclinux-dev at uclinux.org Betreff: Re: [uClinux-dev] Flash Write Have you looked at JFFS Kendrick Hamilton E.I.T. SED Systems, a division of Calian Ltd. 18 Innovation Blvd. PO Box 1464 Saskatoon, Saskatchewan Canada S7N 3R1 Hamilton at sedsystems.ca Tel: (306) 933-1453 Fax: (306) 933-1486 On Sun, 17 Mar 2002, Juergen Mueller wrote: > Hi!! > > Is it possible to write into the flash at runtime. Is there something like a > flash filesystem which allows to write into the flash. > I'm using uCdimm MC68Vz328 and Kernel 2.4.x. > > Any suggestion is appreciated. > > Thanks! > > Juergen > > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From hamilton at SEDSystems.ca Tue Mar 19 11:03:11 2002 From: hamilton at SEDSystems.ca (Kendrick Hamilton) Date: Tue, 19 Mar 2002 10:03:11 -0600 (CST) Subject: AW: [uClinux-dev] Flash Write In-Reply-To: Message-ID: Sorry, I cannot help you with that. I use JFFS with uClinux 2.0.x. It does not have the MTD layer. Instead it has flash.c to interface to the flash device. In that, you would add support for your flash device (if it is not already there). You specify what address flash starts at and it probes for flash at that address. It autodetects the manufacturer and model of flash. You just have to add the flash sector sizes if they are not already there. I suggest looking into mtd code to figure out how it does it. Kendrick Hamilton E.I.T. SED Systems, a division of Calian Ltd. 18 Innovation Blvd. PO Box 1464 Saskatoon, Saskatchewan Canada S7N 3R1 Hamilton at sedsystems.ca Tel: (306) 933-1453 Fax: (306) 933-1486 On Tue, 19 Mar 2002, Juergen Mueller wrote: > Yes I've done. > But it's not quite clear how to use jffs with mtd and blkmem. > I think I have to part my flash > > Can somone give me advise. > > > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From tgriggs at keyww.com Tue Mar 19 12:07:23 2002 From: tgriggs at keyww.com (Travis Griggs) Date: Tue, 19 Mar 2002 09:07:23 -0800 Subject: [uClinux-dev] reboot on the coldfire... References: <3C965888.9040606@keyww.com> <3C96A60A.5080204@snapgear.com> Message-ID: <3C97704B.7020500@keyww.com> Greg Ungerer wrote: > > Hi Travis, > > Travis Griggs wrote: > > > Is this known to not work? > > > It works on some platforms... > > If it doesn't work it will almost certainly be the code in > linux/include/asm-m68knommu/system.h broken for your platform. > OK, I'll look in there. I tested on both our own coldfire board as well as the Motorolla C3. No luck in either case. Travis Griggs This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From pgccoop at hotmail.com Tue Mar 19 15:03:39 2002 From: pgccoop at hotmail.com (Jonathan Schmidt) Date: Tue, 19 Mar 2002 20:03:39 +0000 Subject: [uClinux-dev] "Calibrating delay loop..." crashes Message-ID: I am now using the fixed.ld file from the de2 (Dragen2 board), as an include from the rom.ld file, as follows: MEMORY { romvec : ORIGIN = 0x10c00000, LENGTH = 0x0000400 flash : ORIGIN = 0x10c00400, LENGTH = 0x0400000 - 0x0000400 eflash : ORIGIN = 0x11000000, LENGTH = 0 ramvec : ORIGIN = 0x00000000, LENGTH = 0x0000400 ram : ORIGIN = 0x00020000, LENGTH = 0x1000000 - 0x0020000 eram : ORIGIN = 0x01000000, LENGTH = 0 } INCLUDE arch/m68knommu/platform/68VZ328/pgc_vz/fixed.ld A copy of fixed.ld is not included here because it is available from the CVS in arch/m68knommu/platform/68VZ328/de2/fixed.ld. I am getting the same error (crash at "Calibrating delay loop...") as I was before. I have 16MB of SDRAM (single chip) and 4MB of Intel Flash (single chip). Both are 16bit. It would seem to me that 0x0400000 is in fact 4MB, and 0x1000000 is 16MB. I don't really see a problem here. Perhaps there is something else wrong? This .ld file works just fine with "unified-2.4.6" of Daniel's. I am really confused! Please let me know if anyone has come across this before. Thanks! Jonathan Schmidt >From: Daniel Haensse >Reply-To: uclinux-dev at uclinux.org >To: uclinux-dev at uclinux.org >Subject: Re: [uClinux-dev] "Calibrating delay loop..." crashes >Date: Mon, 18 Mar 2002 07:58:28 +0100 >MIME-Version: 1.0 >Received: from [216.57.57.93] by hotmail.com (3.2) with ESMTP id >MHotMailBE5EDF2700C64004324BD839395D8AEB0; Sun, 17 Mar 2002 23:15:59 -0800 >Received: (from mdom at localhost)by uclinux-dev-listclinux.org >(8.11.3/8.11.3/SuSE Linux 8.11.1-0.5) id g2I72Qq00654for uclinux-dev-list; >Mon, 18 Mar 2002 02:02:26 -0500 >From owner-uclinux-dev at uclinux.org Sun, 17 Mar 2002 23:17:38 -0800 >Message-Id: <200203180702.IAA00812 at ds-00122.usz.ch> >X-Mailer: KMail [version 1.3] >References: >In-Reply-To: >Sender: owner-uclinux-dev at uclinux.org >Precedence: bulk > >Hi Jonathan, > >Am Mittwoch, 13. M?rz 2002 22.44 schrieben Sie: > > I have been building kernels from Daniel Haenesse's "unified-2.4.6" > > distribution for my DragonBall VZ board, and they have been working just > > fine. I tried yesterday to build a kernel straight from the newest >uClinux > > CVS (2.4.17 mostly, as I understand). When the image is programmed to >the > > flash (using a modified "2flash" to support the TE28F320B3 Intel flash), > > and I reboot, the kernel hangs at "Calibrating delay loop...". I >checked > > in the .c file that is responsible for this part of the bootup sequence, > > and there doesn't seem to be anything that could cause a crash. > > > > I was just wondering if anyone has had this type of problem with a VZ >board > > running a 2.4 kernel before, and if so, how did you resolve it. Thanks >in > > advance for any help you can provide. >I had the same bug with 2.0.38. There is something wrong with your >initialisation of the start and end of the sdram or flash. > >Dani ( Haensse :-) ) > > > > > > > _________________________________________________________________ > > Join the world?s largest e-mail service with MSN Hotmail. > > http://www.hotmail.com > > > > This message resent by the uclinux-dev at uclinux.org list server > > http://www.uClinux.org/ >This message resent by the uclinux-dev at uclinux.org list server >http://www.uClinux.org/ _________________________________________________________________ Join the world?s largest e-mail service with MSN Hotmail. http://www.hotmail.com This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From tgriggs at keyww.com Tue Mar 19 15:51:27 2002 From: tgriggs at keyww.com (Travis Griggs) Date: Tue, 19 Mar 2002 12:51:27 -0800 Subject: [uClinux-dev] memory usage question Message-ID: <3C97A4CF.5030101@keyww.com> I have a program running which ps tells me is using 120K for a memory footprint. What it doesn't tell me was what the largest request for contiguous memory it made (or was initialized with) was. Is there anyway to figure this out? I'm particularly interested actually with what the initial program start memory requirements were. Because this is where we ran into problems before. Travis Griggs This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From gmenie at akamai.com Tue Mar 19 17:08:09 2002 From: gmenie at akamai.com (Menie, Georges) Date: Tue, 19 Mar 2002 17:08:09 -0500 Subject: [uClinux-dev] "Calibrating delay loop..." crashes Message-ID: Could it be a timer interrupt problem ? One of the first thing which is setup is the timer irq, just before the calibrate_delay() function... Regards, Georges > -----Original Message----- > From: Jonathan Schmidt [mailto:pgccoop at hotmail.com] > Sent: mardi 19 mars 2002 21:04 > To: uclinux-dev at uclinux.org > Subject: Re: [uClinux-dev] "Calibrating delay loop..." crashes > > > I am now using the fixed.ld file from the de2 (Dragen2 > board), as an include > from the rom.ld file, as follows: > > MEMORY > { > romvec : ORIGIN = 0x10c00000, LENGTH = 0x0000400 > flash : ORIGIN = 0x10c00400, LENGTH = 0x0400000 - 0x0000400 > eflash : ORIGIN = 0x11000000, LENGTH = 0 > ramvec : ORIGIN = 0x00000000, LENGTH = 0x0000400 > ram : ORIGIN = 0x00020000, LENGTH = 0x1000000 - 0x0020000 > eram : ORIGIN = 0x01000000, LENGTH = 0 > } > > INCLUDE arch/m68knommu/platform/68VZ328/pgc_vz/fixed.ld > > A copy of fixed.ld is not included here because it is > available from the CVS > in arch/m68knommu/platform/68VZ328/de2/fixed.ld. I am > getting the same > error (crash at "Calibrating delay loop...") as I was before. > I have 16MB > of SDRAM (single chip) and 4MB of Intel Flash (single chip). > Both are > 16bit. It would seem to me that 0x0400000 is in fact 4MB, > and 0x1000000 is > 16MB. I don't really see a problem here. Perhaps there is > something else > wrong? This .ld file works just fine with "unified-2.4.6" of > Daniel's. > > I am really confused! Please let me know if anyone has come > across this > before. Thanks! > > Jonathan Schmidt > > > >From: Daniel Haensse > >Reply-To: uclinux-dev at uclinux.org > >To: uclinux-dev at uclinux.org > >Subject: Re: [uClinux-dev] "Calibrating delay loop..." crashes > >Date: Mon, 18 Mar 2002 07:58:28 +0100 > >MIME-Version: 1.0 > >Received: from [216.57.57.93] by hotmail.com (3.2) with ESMTP id > >MHotMailBE5EDF2700C64004324BD839395D8AEB0; Sun, 17 Mar 2002 > 23:15:59 -0800 > >Received: (from mdom at localhost)by uclinux-dev-listclinux.org > >(8.11.3/8.11.3/SuSE Linux 8.11.1-0.5) id g2I72Qq00654for > uclinux-dev-list; > >Mon, 18 Mar 2002 02:02:26 -0500 > >From owner-uclinux-dev at uclinux.org Sun, 17 Mar 2002 23:17:38 -0800 > >Message-Id: <200203180702.IAA00812 at ds-00122.usz.ch> > >X-Mailer: KMail [version 1.3] > >References: > >In-Reply-To: > >Sender: owner-uclinux-dev at uclinux.org > >Precedence: bulk > > > >Hi Jonathan, > > > >Am Mittwoch, 13. M?rz 2002 22.44 schrieben Sie: > > > I have been building kernels from Daniel Haenesse's > "unified-2.4.6" > > > distribution for my DragonBall VZ board, and they have > been working just > > > fine. I tried yesterday to build a kernel straight from > the newest > >uClinux > > > CVS (2.4.17 mostly, as I understand). When the image is > programmed to > >the > > > flash (using a modified "2flash" to support the > TE28F320B3 Intel flash), > > > and I reboot, the kernel hangs at "Calibrating delay loop...". I > >checked > > > in the .c file that is responsible for this part of the > bootup sequence, > > > and there doesn't seem to be anything that could cause a crash. > > > > > > I was just wondering if anyone has had this type of > problem with a VZ > >board > > > running a 2.4 kernel before, and if so, how did you > resolve it. Thanks > >in > > > advance for any help you can provide. > >I had the same bug with 2.0.38. There is something wrong with your > >initialisation of the start and end of the sdram or flash. > > > >Dani ( Haensse :-) ) > > > > > > > > > > > _________________________________________________________________ > > > Join the world?s largest e-mail service with MSN Hotmail. > > > http://www.hotmail.com > > > > > > This message resent by the uclinux-dev at uclinux.org list server > > > http://www.uClinux.org/ > >This message resent by the uclinux-dev at uclinux.org list server > >http://www.uClinux.org/ > > > > > _________________________________________________________________ > Join the world's largest e-mail service with MSN Hotmail. > http://www.hotmail.com > > This message resent by the uclinux-dev at uclinux.org list > server http://www.uClinux.org/ > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From gmenie at akamai.com Tue Mar 19 17:19:31 2002 From: gmenie at akamai.com (Menie, Georges) Date: Tue, 19 Mar 2002 17:19:31 -0500 Subject: [uClinux-dev] Debugging on the 68VZ328 using 2.4.10 and elf Message-ID: here is the latest patch to allow debugging user apps on a 68328/68EZ328/68VZ328. I did remove the trap#2-trap#14 entries overwrite, so that boards using this traps will not break. These traps are not used by the kernel anyway. the patch is for the linux-2.4.x module in CVS. Regards, Georges -------------- next part -------------- An embedded and charset-unspecified text was scrubbed... Name: patch-gdb.txt URL: From artu at 263.net Tue Mar 19 19:47:15 2002 From: artu at 263.net (ziming) Date: Wed, 20 Mar 2002 8:47:15 +0800 Subject: [uClinux-dev] "Calibrating delay loop..." crashes Message-ID: <20020320002625.DF6F81CF84953@smtp.263.net> Calibrating delay loop.. crash means the timer interrupt is not directed into the kernel to update value of jiffes. You should check the timer interrupt reponse is correctly set or not. >I am now using the fixed.ld file from the de2 (Dragen2 board), as an include >from the rom.ld file, as follows: > >MEMORY > { > romvec : ORIGIN = 0x10c00000, LENGTH = 0x0000400 > flash : ORIGIN = 0x10c00400, LENGTH = 0x0400000 - 0x0000400 > eflash : ORIGIN = 0x11000000, LENGTH = 0 > ramvec : ORIGIN = 0x00000000, LENGTH = 0x0000400 > ram : ORIGIN = 0x00020000, LENGTH = 0x1000000 - 0x0020000 > eram : ORIGIN = 0x01000000, LENGTH = 0 > } > >INCLUDE arch/m68knommu/platform/68VZ328/pgc_vz/fixed.ld > >A copy of fixed.ld is not included here because it is available from the CVS >in arch/m68knommu/platform/68VZ328/de2/fixed.ld. I am getting the same >error (crash at "Calibrating delay loop...") as I was before. I have 16MB >of SDRAM (single chip) and 4MB of Intel Flash (single chip). Both are >16bit. It would seem to me that 0x0400000 is in fact 4MB, and 0x1000000 is >16MB. I don't really see a problem here. Perhaps there is something else >wrong? This .ld file works just fine with "unified-2.4.6" of Daniel's. > >I am really confused! Please let me know if anyone has come across this >before. Thanks! > >Jonathan Schmidt > > >>From: Daniel Haensse >>Reply-To: uclinux-dev at uclinux.org >>To: uclinux-dev at uclinux.org >>Subject: Re: [uClinux-dev] "Calibrating delay loop..." crashes >>Date: Mon, 18 Mar 2002 07:58:28 +0100 >>MIME-Version: 1.0 >>Received: from [216.57.57.93] by hotmail.com (3.2) with ESMTP id >>MHotMailBE5EDF2700C64004324BD839395D8AEB0; Sun, 17 Mar 2002 23:15:59 -0800 >>Received: (from mdom at localhost)by uclinux-dev-listclinux.org >>(8.11.3/8.11.3/SuSE Linux 8.11.1-0.5) id g2I72Qq00654for uclinux-dev-list; >>Mon, 18 Mar 2002 02:02:26 -0500 >>From owner-uclinux-dev at uclinux.org Sun, 17 Mar 2002 23:17:38 -0800 >>Message-Id: <200203180702.IAA00812 at ds-00122.usz.ch> >>X-Mailer: KMail [version 1.3] >>References: >>In-Reply-To: >>Sender: owner-uclinux-dev at uclinux.org >>Precedence: bulk >> >>Hi Jonathan, >> >>Am Mittwoch, 13. M?rz 2002 22.44 schrieben Sie: >> > I have been building kernels from Daniel Haenesse's "unified-2.4.6" >> > distribution for my DragonBall VZ board, and they have been working just >> > fine. I tried yesterday to build a kernel straight from the newest >>uClinux >> > CVS (2.4.17 mostly, as I understand). When the image is programmed to >>the >> > flash (using a modified "2flash" to support the TE28F320B3 Intel flash), >> > and I reboot, the kernel hangs at "Calibrating delay loop...". I >>checked >> > in the .c file that is responsible for this part of the bootup sequence, >> > and there doesn't seem to be anything that could cause a crash. >> > >> > I was just wondering if anyone has had this type of problem with a VZ >>board >> > running a 2.4 kernel before, and if so, how did you resolve it. Thanks >>in >> > advance for any help you can provide. >>I had the same bug with 2.0.38. There is something wrong with your >>initialisation of the start and end of the sdram or flash. >> >>Dani ( Haensse :-) ) >> >> > >> > >> > _________________________________________________________________ >> > Join the world?s largest e-mail service with MSN Hotmail. >> > http://www.hotmail.com >> > >> > This message resent by the uclinux-dev at uclinux.org list server >> > http://www.uClinux.org/ >>This message resent by the uclinux-dev at uclinux.org list server >>http://www.uClinux.org/ > > > > >_________________________________________________________________ >Join the world?s largest e-mail service with MSN Hotmail. >http://www.hotmail.com > >This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From superwen at 263.net Tue Mar 19 19:48:51 2002 From: superwen at 263.net (Kevin) Date: Wed, 20 Mar 2002 8:48:51 +0800 Subject: [uClinux-dev] kernel command line is NULL Message-ID: <20020320004141.7EF6E1DB9861B@smtp.263.net> Thanks a lot, Georges, 1,I've look into the config_BSP function,since I'm using a mot's M68VZ328ADS board, which doesn't have cs8900, so I comment those lines. And the kernel just stucked at p=getbenv("APPEND"),so I comment that also. Then it goes a little further,but finally I still failed with "Unable to open an initial console". I'm wondering if it's caused by the comment I did. void config_BSP(char *command, int len) { unsigned char *p; printk("\n68VZ328 DragonBallVZ support (c) 2001 Lineo, Inc.\n"); /****** ----------I comment this printk("uCdimm serial string [%s]\n",getserialnum()); p = cs8900a_hwaddr = gethwaddr(0); printk("uCdimm hwaddr %.2x:%.2x:%.2x:%.2x:%.2x:%.2x\n", p[0], p[1], p[2], p[3], p[4], p[5]); p = getbenv("APPEND"); if (p) strcpy(p,command); else command[0] = 0; ******/ mach_sched_init = BSP_sched_init; mach_tick = BSP_tick; mach_gettimeoffset = BSP_gettimeoffset; mach_gettod = BSP_gettod; mach_hwclk = NULL; mach_set_clock_mmss = NULL; // mach_mksound = NULL; mach_reset = BSP_reset; // mach_debug_init = NULL; config_M68VZ328_irq(); } 2,"unable to open an initial console", I searched the maillist, some other people got the same problem, but most of them are because of genromfs, I only got one genromfs in my machine, which is in /usr/local/bin/genromfs. I used command "genromfs -h" and at top is "genromfs 0.5.1",I'm not sure if it's the right one. 3,The kernel could get through the warning message "unable to open..", but it stucked at the following init thread sbin/init -----looks like it links to /bin/init. I'm wondering if the mount romfs file system could cause this? but it mount successfully. >> VFS: Mounted root (romfs filesystem) readonly. I traced down the warning message, it's caused by the return value of the system call ---open().I mount the romfs.img in PC, and the /dev/console is there, however, what I want is /dev/ttyS0, it's also there, ----------------------- if (open("/dev/console", O_RDWR, 0) < 0) ---I modified it to "open("/dev/ttyS0", O_RDWR, 0)", but got the same warning. printk("Warning: unable to open an initial console.\n"); (void) dup(0); (void) dup(0); if (execute_command) execve(execute_command,argv_init,envp_init); execve("/sbin/init",argv_init,envp_init); execve("/etc/init",argv_init,envp_init); execve("/bin/init",argv_init,envp_init); execve("/bin/sh",argv_init,envp_init); panic("No init found. Try passing init= option to kernel."); }------the end of init-------- Too many questions, forgive me, I've stucked here for a couple of days. Thanks again Best Regards Kevin ------------------ -----Original Message-------- >RE: [uClinux-dev] kernel command line is NULL >It is not a problem, the command line is a way to setup extra parameters for >the kernel, >You may not need it. The code which setup this command line is in the >config_BSP() >function (see the config.c file for your platform). An empty command line is >ok. > >Georges > >> -----Original Message----- >> From: Kevin [mailto:superwen at 263.net] >> Sent: lundi 18 mars 2002 19:01 >> To: uclinux >> Subject: [uClinux-dev] kernel command line is NULL >> >> >> Hi, All, >> The kernel command line I got is NULL(please see below), I'm >> wondering if I can add it manually? Or where should I check it? >> Please advise, thank you very much >> >> -------------------------------- >> here I attached the message I got from uart. >> >> KERNEL -> TEXT=0x1000400-0x104e308 DATA=0x020000-0x03c000 >> BSS=0x03c000-0x04f198 >> CAT ROMARRAY! >> KERNEL -> ROMFS=0x106a308-0x1125f78 MEM=0x050000-0x800000 >> STACK=0x800000-0x80000 >> 0 >> On node 0 totalpages: 2048 >> zone(0): 0 pages. >> zone(1): 2048 pages. >> zone(2): 0 pages. >> Done setup_arch >> init/main.c:line 561: Setup_Arch Done >> Kernel command line: -------This looks >> not good. >> Calibrating delay loop... 1.40 BogoMIPS >> init/main.c:line 586: Calibrate_delay.Done >> Memory available: 7728k/8064k RAM, 0k/0k ROM (311k kernel >> code, 188k data) >> Dentry-cache hash table entries: 1024 (order: 1, 8192 bytes) >> Inode-cache hash table entries: 512 (order: 0, 4096 bytes) >> Mount-cache hash table entries: 512 (order: 0, 4096 bytes) >> Buffer-cache hash table entries: 1024 (order: 0, 4096 bytes) >> Page-cache hash table entries: 2048 (order: 1, 8192 bytes) >> POSIX conformance testing by UNIFIX >> <6>Linux NET4.0 for Linux 2.4 >> <6>Based upon Swansea University Computer Society NET3.039 >> Starting kswapd >> pty: 256 Unix98 ptys configured >> MC68328 serial driver version 1.00 >> ttyS0 at 0xfffff900 (irq = 64) is a builtin MC68328 UART >> block: 64 slots per queue, batch=16 >> RAMDISK driver initialized: 16 RAM disks of 4096K size 1024 blocksize >> Blkmem copyright 1998,1999 D. Jeff Dionne >> Blkmem copyright 1998 Kenneth Albanowski >> Blkmem 1 disk images: >> 0: 106A308-1126307 [VIRTUAL 106A308-1126307] (RO) >> VFS: Mounted root (romfs filesystem) readonly. >> Warning: unable to open an initial console. >> >> >> >> This message resent by the uclinux-dev at uclinux.org list >> server http://www.uClinux.org/ >> >This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ -------------------------------- This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From michael_chen at ipmental.com.tw Tue Mar 19 20:13:42 2002 From: michael_chen at ipmental.com.tw (Michael Chen) Date: Wed, 20 Mar 2002 09:13:42 +0800 Subject: AW: [uClinux-dev] Flash Write References: Message-ID: <000d01c1cfac$8b0b9080$c600a8c0@Domain> Hi. Juergen, Try http://www.rcs.ei.tum.de/~kuhn/uclinux/coldfire/tarifa/20011206/README It may help you. Michael Chen IPmental Inc. 8F-9, No. 35 Hsin-Tai Road, Chupei City, Hsincuh Hsien, Taiwan Tel: +886-3-5529868 ext 225 Fax: +886-3-5551296 E-mail: michael_chen at ipmental.com.tw ----- Original Message ----- From: Kendrick Hamilton To: Sent: Wednesday, March 20, 2002 12:03 AM Subject: Re: AW: [uClinux-dev] Flash Write > Sorry, > I cannot help you with that. I use JFFS with uClinux 2.0.x. It > does not have the MTD layer. Instead it has flash.c to interface to the > flash device. In that, you would add support for your flash device (if it > is not already there). You specify what address flash starts at and it > probes for flash at that address. It autodetects the manufacturer and > model of flash. You just have to add the flash sector sizes if they are > not already there. > I suggest looking into mtd code to figure out how it does it. > > Kendrick Hamilton E.I.T. > SED Systems, a division of Calian Ltd. > 18 Innovation Blvd. > PO Box 1464 > Saskatoon, Saskatchewan > Canada > S7N 3R1 > > Hamilton at sedsystems.ca > Tel: (306) 933-1453 > Fax: (306) 933-1486 > > On Tue, 19 Mar 2002, Juergen Mueller wrote: > > > Yes I've done. > > But it's not quite clear how to use jffs with mtd and blkmem. > > I think I have to part my flash > > > > Can somone give me advise. > > > > > > > > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From gerg at snapgear.com Tue Mar 19 22:43:27 2002 From: gerg at snapgear.com (Greg Ungerer) Date: Wed, 20 Mar 2002 13:43:27 +1000 Subject: [uClinux-dev] reboot on the coldfire... References: <3C965888.9040606@keyww.com> <3C96A60A.5080204@snapgear.com> <3C97704B.7020500@keyww.com> Message-ID: <3C98055F.9010307@snapgear.com> Hi Travis, Travis Griggs wrote: > Greg Ungerer wrote: >> Travis Griggs wrote: >> >> > Is this known to not work? >> >> >> It works on some platforms... >> >> If it doesn't work it will almost certainly be the code in >> linux/include/asm-m68knommu/system.h broken for your platform. >> > OK, I'll look in there. > > I tested on both our own coldfire board as well as the Motorolla C3. No > luck in either case. It works on the SnapGear/NETtel boards. Regards Greg ------------------------------------------------------------------------ Greg Ungerer -- Chief Software Wizard EMAIL: gerg at snapgear.com SnapGear PHONE: +61 7 3435 2888 825 Stanley St, FAX: +61 7 3891 3630 Woolloongabba, QLD, 4102, Australia WEB: www.snapgear.com This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From uclinux at softhome.net Wed Mar 20 00:34:39 2002 From: uclinux at softhome.net (Kevin) Date: Wed, 20 Mar 2002 13:34:39 +0800 Subject: [uClinux-dev] romfs offset Message-ID: Hi,There, I got a question about my romdisk image.-I'm playing with mot's MC68VZ328ADS. The message I got from UART is like below, it failed with "unable to open an initial console". KERNEL -> TEXT=0x1000400-0x104e308 DATA=0x020000-0x03c000 BSS=0x03c000-0x04f198 KERNEL -> ROMFS=0x106a308-0x1125f78 MEM=0x050000-0x800000 STACK=0x800000-0x800000 But when I used some hex editor to check the image.bin I got and the B-record file. I found the romfs is actually start from 0x1062138. NOT 0x106a308 as above!!. Why there is an offset? Please advise. Thank you very much. Best Regards Beckhag This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From daniel.haensse at fhk.usz.ch Wed Mar 20 02:39:58 2002 From: daniel.haensse at fhk.usz.ch (Daniel Haensse) Date: Wed, 20 Mar 2002 08:39:58 +0100 Subject: [uClinux-dev] "Calibrating delay loop..." crashes In-Reply-To: References: Message-ID: <200203200743.IAA01659@ds-00122.usz.ch> When I came across the problem I spend a lot of time searching in the calibrate and timer irq functions and finally found the memory range problem. It looks like the calibration_delay function is the first function that "rocks" with C code and because of the bad settings the system crashes. There is also a bad definition in BSP_sched_init in config.c that makes the tick timer slightly slower (2seconds on 1 minute), but this gives some extra boost of bogomips on the bootmessage, which is not true. I corrected the function for the dragonix board. from my crt0rom.S as an example, somebody introduced this mechanism earlier .global _start .global _stext .global _rambase .global _ramvec .global _ramstart .global _ramend and later in the file: movel #__ramend, %d0 movel %d0, _ramend while __ramend is defined in my rom.ld, later in the kernel in setup.c memory_start = PAGE_ALIGN(_ramstart); memory_end = _ramend; /* by now the stack is part of the init task */ the fine difference in the var name is the _ also check for the other vars to be correctly set. Maybe it is really an other problem with the same symptoms, but I would check this first. Dani Am Dienstag, 19. M?rz 2002 23.08 schrieben Sie: > Could it be a timer interrupt problem ? > One of the first thing which is setup is the timer irq, just > before the calibrate_delay() function... > > Regards, > Georges > > > -----Original Message----- > > From: Jonathan Schmidt [mailto:pgccoop at hotmail.com] > > Sent: mardi 19 mars 2002 21:04 > > To: uclinux-dev at uclinux.org > > Subject: Re: [uClinux-dev] "Calibrating delay loop..." crashes > > > > > > I am now using the fixed.ld file from the de2 (Dragen2 > > board), as an include > > from the rom.ld file, as follows: > > > > MEMORY > > { > > romvec : ORIGIN = 0x10c00000, LENGTH = 0x0000400 > > flash : ORIGIN = 0x10c00400, LENGTH = 0x0400000 - 0x0000400 > > eflash : ORIGIN = 0x11000000, LENGTH = 0 > > ramvec : ORIGIN = 0x00000000, LENGTH = 0x0000400 > > ram : ORIGIN = 0x00020000, LENGTH = 0x1000000 - 0x0020000 > > eram : ORIGIN = 0x01000000, LENGTH = 0 > > } > > > > INCLUDE arch/m68knommu/platform/68VZ328/pgc_vz/fixed.ld > > > > A copy of fixed.ld is not included here because it is > > available from the CVS > > in arch/m68knommu/platform/68VZ328/de2/fixed.ld. I am > > getting the same > > error (crash at "Calibrating delay loop...") as I was before. > > I have 16MB > > of SDRAM (single chip) and 4MB of Intel Flash (single chip). > > Both are > > 16bit. It would seem to me that 0x0400000 is in fact 4MB, > > and 0x1000000 is > > 16MB. I don't really see a problem here. Perhaps there is > > something else > > wrong? This .ld file works just fine with "unified-2.4.6" of > > Daniel's. > > > > I am really confused! Please let me know if anyone has come > > across this > > before. Thanks! > > > > Jonathan Schmidt > > > > >From: Daniel Haensse > > >Reply-To: uclinux-dev at uclinux.org > > >To: uclinux-dev at uclinux.org > > >Subject: Re: [uClinux-dev] "Calibrating delay loop..." crashes > > >Date: Mon, 18 Mar 2002 07:58:28 +0100 > > >MIME-Version: 1.0 > > >Received: from [216.57.57.93] by hotmail.com (3.2) with ESMTP id > > >MHotMailBE5EDF2700C64004324BD839395D8AEB0; Sun, 17 Mar 2002 > > > > 23:15:59 -0800 > > > > >Received: (from mdom at localhost)by uclinux-dev-listclinux.org > > >(8.11.3/8.11.3/SuSE Linux 8.11.1-0.5) id g2I72Qq00654for > > > > uclinux-dev-list; > > > > >Mon, 18 Mar 2002 02:02:26 -0500 > > >From owner-uclinux-dev at uclinux.org Sun, 17 Mar 2002 23:17:38 -0800 > > >Message-Id: <200203180702.IAA00812 at ds-00122.usz.ch> > > >X-Mailer: KMail [version 1.3] > > >References: > > >In-Reply-To: > > >Sender: owner-uclinux-dev at uclinux.org > > >Precedence: bulk > > > > > >Hi Jonathan, > > > > > >Am Mittwoch, 13. M?rz 2002 22.44 schrieben Sie: > > > > I have been building kernels from Daniel Haenesse's > > > > "unified-2.4.6" > > > > > > distribution for my DragonBall VZ board, and they have > > > > been working just > > > > > > fine. I tried yesterday to build a kernel straight from > > > > the newest > > > > >uClinux > > > > > > > CVS (2.4.17 mostly, as I understand). When the image is > > > > programmed to > > > > >the > > > > > > > flash (using a modified "2flash" to support the > > > > TE28F320B3 Intel flash), > > > > > > and I reboot, the kernel hangs at "Calibrating delay loop...". I > > > > > >checked > > > > > > > in the .c file that is responsible for this part of the > > > > bootup sequence, > > > > > > and there doesn't seem to be anything that could cause a crash. > > > > > > > > I was just wondering if anyone has had this type of > > > > problem with a VZ > > > > >board > > > > > > > running a 2.4 kernel before, and if so, how did you > > > > resolve it. Thanks > > > > >in > > > > > > > advance for any help you can provide. > > > > > >I had the same bug with 2.0.38. There is something wrong with your > > >initialisation of the start and end of the sdram or flash. > > > > > >Dani ( Haensse :-) ) > > > > > > > _________________________________________________________________ > > > > Join the world?s largest e-mail service with MSN Hotmail. > > > > http://www.hotmail.com > > > > > > > > This message resent by the uclinux-dev at uclinux.org list server > > > > http://www.uClinux.org/ > > > > > >This message resent by the uclinux-dev at uclinux.org list server > > >http://www.uClinux.org/ > > > > _________________________________________________________________ > > Join the world's largest e-mail service with MSN Hotmail. > > http://www.hotmail.com > > > > This message resent by the uclinux-dev at uclinux.org list > > server http://www.uClinux.org/ > > This message resent by the uclinux-dev at uclinux.org list server > http://www.uClinux.org/ This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From sheeliang at lit.org.sg Wed Mar 20 02:53:02 2002 From: sheeliang at lit.org.sg (sheeliang at lit.org.sg) Date: Wed, 20 Mar 2002 15:53:02 +0800 Subject: [uClinux-dev] Network cards on Integrator AP In-Reply-To: <20020319171343.A9118@localhost.localdomain>; from sheeliang@lit.org.sg on Tue, Mar 19, 2002 at 05:14:15PM +0800 References: <20020319171343.A9118@localhost.localdomain> Message-ID: <20020320155302.A1154@localhost.localdomain> Okok, seems like nobody has much idea. I have moved on a bit though. Now the problem is that I can receive packets, but I cannot send. I can receive arp-whohas and ICMP echo request, but it seems like I can only send arp-replies. This is pretty strange, being able to send out arp replies but not ICMP echo replies. Loopback works fine though. The card is a AMD pcnet32. Going to try 3Com Vortex next. On Tue, Mar 19, 2002 at 05:14:15PM +0800, sheeliang at lit.org.sg wrote: > Wondering if anybody has any experience getting network cards to work on > Integrator AP, which uses the PCI V3 chipset? I have tried pcnet32 and 3com > cards without avail. Initialization runs ok, ie, I can see the correct ethernet > addresses, IRQ assigned, ioport, etc on startup, but actual pinging on both > sides fail. The core is a ARM7TDMI running uClinux-2.4.17. Help, anyone? > > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From martijn.schipper at intersil.com Wed Mar 20 03:14:34 2002 From: martijn.schipper at intersil.com (Schipper, Martijn (Contractor)) Date: Wed, 20 Mar 2002 03:14:34 -0500 Subject: [uClinux-dev] Network cards on Integrator AP Message-ID: <716A7F62A194D41195AB001083FC3AE4BF61F2@inlmx1.inl.intersil.com> Hello Shee(?), It looks to me as a routing problem. Can you give some more information on the IP settings on your machine? Best regards, Martijn Schipper Software Engineer Intersil Netherlands BV. Rembrandtlaan 1a 3723 BG Bilthoven The Netherlands Tel: +31 (0)30 2296076 Fax: +31 (0)30 2296061 E-mail: Martijn.Schipper at Intersil.com Internet: www.intersil.com -----Original Message----- From: sheeliang at lit.org.sg [mailto:sheeliang at lit.org.sg] Sent: woensdag 20 maart 2002 8:53 To: uclinux-dev at uclinux.org Subject: Re: [uClinux-dev] Network cards on Integrator AP Okok, seems like nobody has much idea. I have moved on a bit though. Now the problem is that I can receive packets, but I cannot send. I can receive arp-whohas and ICMP echo request, but it seems like I can only send arp-replies. This is pretty strange, being able to send out arp replies but not ICMP echo replies. Loopback works fine though. The card is a AMD pcnet32. Going to try 3Com Vortex next. On Tue, Mar 19, 2002 at 05:14:15PM +0800, sheeliang at lit.org.sg wrote: > Wondering if anybody has any experience getting network cards to work on > Integrator AP, which uses the PCI V3 chipset? I have tried pcnet32 and 3com > cards without avail. Initialization runs ok, ie, I can see the correct ethernet > addresses, IRQ assigned, ioport, etc on startup, but actual pinging on both > sides fail. The core is a ARM7TDMI running uClinux-2.4.17. Help, anyone? > > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From michiel.thuys at intersil.com Wed Mar 20 03:53:39 2002 From: michiel.thuys at intersil.com (Thuys, Michiel) Date: Wed, 20 Mar 2002 03:53:39 -0500 Subject: [uClinux-dev] Executing from NFS mounted directory on 2.4 kernel crash Message-ID: <716A7F62A194D41195AB001083FC3AE4B5E7A6@inlmx1.inl.intersil.com> Hi all, I have a problem with executing files in an NFS mounted directory. When I try to execute a flat binary in an NFS mounted directory, i get a segmentation fault. When I copy the file to the ram disk and then execute it, it works fine. I run uclinux 2.4.17 on an ARM940. Does this work for anynone or do you have similar problems? Michiel This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From eauth at softsys.co.at Wed Mar 20 04:32:52 2002 From: eauth at softsys.co.at (Erwin Authried) Date: Wed, 20 Mar 2002 10:32:52 +0100 Subject: [uClinux-dev] Network cards on Integrator AP In-Reply-To: <20020320155302.A1154@localhost.localdomain> Message-ID: > -----Ursprungliche Nachricht----- > Von: owner-uclinux-dev at uclinux.org > [mailto:owner-uclinux-dev at uclinux.org]Im Auftrag von > sheeliang at lit.org.sg > Gesendet: Mittwoch, 20. Marz 2002 08:53 > An: uclinux-dev at uclinux.org > Betreff: Re: [uClinux-dev] Network cards on Integrator AP > > > Okok, seems like nobody has much idea. I have moved on a bit though. > Now the problem is that I can receive packets, but I cannot send. > I can receive arp-whohas and ICMP echo request, but it seems like > I can only > send arp-replies. This is pretty strange, being able to send out > arp replies > but not ICMP echo replies. Loopback works fine though. The card is a AMD > pcnet32. > Sounds pretty much like a packet alignment problem somewhere in the driver. Maybe a missing "packed" attribute that doesn't cause problems with other architectures. I had similar problems with the Orinoco driver on the Atmel AT91 cpus. -Erwin This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From daniel.haensse at fhk.usz.ch Wed Mar 20 04:39:13 2002 From: daniel.haensse at fhk.usz.ch (Daniel Haensse) Date: Wed, 20 Mar 2002 10:39:13 +0100 Subject: [uClinux-dev] Dragonix VZ mainboards Message-ID: <200203200943.KAA03366@ds-00122.usz.ch> Hi folks, Some people asked me about assembled boards in the last couple of months. I have updated the openhardware homepage ( http://www.dragonix.net it's a forward ) . Now kits (assembled and tested boards) for 600 Euro including B/W display are available, also including latest distro cd and serial cable to run it out of the box. I have a couple of boards on stock, next shipping is this friday) thanks Dani This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From gmenie at akamai.com Wed Mar 20 04:45:30 2002 From: gmenie at akamai.com (Menie, Georges) Date: Wed, 20 Mar 2002 04:45:30 -0500 Subject: [uClinux-dev] kernel command line is NULL Message-ID: Hi Kevin, you did the right thing, the code you commented out is uCdimm specific. just keep : command[0] = 0; before the line: mach_sched_init = BSP_sched_init; it's just signify there is no specific boot option... genromfs 0.5.1 seems to be the good version, could you check that the device is created in your romfs/dev subdir ? Regards, Georges > -----Original Message----- > From: Kevin [mailto:superwen at 263.net] > Sent: mercredi 20 mars 2002 01:49 > To: uclinux-dev at uclinux.org > Subject: Re: RE: [uClinux-dev] kernel command line is NULL > > > Thanks a lot, Georges, > 1,I've look into the config_BSP function,since I'm using a > mot's M68VZ328ADS board, which doesn't have cs8900, so I > comment those lines. And the kernel just stucked at > p=getbenv("APPEND"),so I comment that also. Then it goes a > little further,but finally I still failed with "Unable to > open an initial console". I'm wondering if it's caused by the > comment I did. > void config_BSP(char *command, int len) > { > unsigned char *p; > > printk("\n68VZ328 DragonBallVZ support (c) 2001 Lineo, Inc.\n"); > /****** ----------I comment > this > printk("uCdimm serial string [%s]\n",getserialnum()); > p = cs8900a_hwaddr = gethwaddr(0); > printk("uCdimm hwaddr %.2x:%.2x:%.2x:%.2x:%.2x:%.2x\n", > p[0], p[1], p[2], p[3], p[4], p[5]); > p = getbenv("APPEND"); > if (p) strcpy(p,command); > else command[0] = 0; > ******/ > mach_sched_init = BSP_sched_init; > mach_tick = BSP_tick; > mach_gettimeoffset = BSP_gettimeoffset; > mach_gettod = BSP_gettod; > mach_hwclk = NULL; > mach_set_clock_mmss = NULL; > // mach_mksound = NULL; > mach_reset = BSP_reset; > // mach_debug_init = NULL; > > config_M68VZ328_irq(); > } > > 2,"unable to open an initial console", I searched the > maillist, some other people got the same problem, but most of > them are because of genromfs, I only got one genromfs in my > machine, which is in /usr/local/bin/genromfs. I used command > "genromfs -h" and at top is "genromfs 0.5.1",I'm not sure if > it's the right one. > 3,The kernel could get through the warning message "unable to > open..", but it stucked at the following init thread > sbin/init -----looks like it links to /bin/init. I'm > wondering if the mount romfs file system could cause this? > but it mount successfully. > >> VFS: Mounted root (romfs filesystem) readonly. > I traced down the warning message, it's caused by the return > value of the system call ---open().I mount the romfs.img in > PC, and the /dev/console is there, however, what I want is > /dev/ttyS0, it's also there, > ----------------------- > if (open("/dev/console", O_RDWR, 0) < 0) ---I modified it > to "open("/dev/ttyS0", O_RDWR, 0)", but got the same warning. > printk("Warning: unable to open an initial console.\n"); > > (void) dup(0); > (void) dup(0); > > if (execute_command) > execve(execute_command,argv_init,envp_init); > execve("/sbin/init",argv_init,envp_init); > execve("/etc/init",argv_init,envp_init); > execve("/bin/init",argv_init,envp_init); > execve("/bin/sh",argv_init,envp_init); > panic("No init found. Try passing init= option to kernel."); > }------the end of init-------- > > Too many questions, forgive me, I've stucked here for a > couple of days. > Thanks again > Best Regards > Kevin > ------------------ > > > -----Original Message-------- > >RE: [uClinux-dev] kernel command line is NULL > >It is not a problem, the command line is a way to setup > extra parameters for > >the kernel, > >You may not need it. The code which setup this command line is in the > >config_BSP() > >function (see the config.c file for your platform). An empty > command line is > >ok. > > > >Georges > > > >> -----Original Message----- > >> From: Kevin [mailto:superwen at 263.net] > >> Sent: lundi 18 mars 2002 19:01 > >> To: uclinux > >> Subject: [uClinux-dev] kernel command line is NULL > >> > >> > >> Hi, All, > >> The kernel command line I got is NULL(please see below), I'm > >> wondering if I can add it manually? Or where should I check it? > >> Please advise, thank you very much > >> > >> -------------------------------- > >> here I attached the message I got from uart. > >> > >> KERNEL -> TEXT=0x1000400-0x104e308 DATA=0x020000-0x03c000 > >> BSS=0x03c000-0x04f198 > >> CAT ROMARRAY! > >> KERNEL -> ROMFS=0x106a308-0x1125f78 MEM=0x050000-0x800000 > >> STACK=0x800000-0x80000 > >> 0 > >> On node 0 totalpages: 2048 > >> zone(0): 0 pages. > >> zone(1): 2048 pages. > >> zone(2): 0 pages. > >> Done setup_arch > >> init/main.c:line 561: Setup_Arch Done > >> Kernel command line: -------This looks > >> not good. > >> Calibrating delay loop... 1.40 BogoMIPS > >> init/main.c:line 586: Calibrate_delay.Done > >> Memory available: 7728k/8064k RAM, 0k/0k ROM (311k kernel > >> code, 188k data) > >> Dentry-cache hash table entries: 1024 (order: 1, 8192 bytes) > >> Inode-cache hash table entries: 512 (order: 0, 4096 bytes) > >> Mount-cache hash table entries: 512 (order: 0, 4096 bytes) > >> Buffer-cache hash table entries: 1024 (order: 0, 4096 bytes) > >> Page-cache hash table entries: 2048 (order: 1, 8192 bytes) > >> POSIX conformance testing by UNIFIX > >> <6>Linux NET4.0 for Linux 2.4 > >> <6>Based upon Swansea University Computer Society NET3.039 > >> Starting kswapd > >> pty: 256 Unix98 ptys configured > >> MC68328 serial driver version 1.00 > >> ttyS0 at 0xfffff900 (irq = 64) is a builtin MC68328 UART > >> block: 64 slots per queue, batch=16 > >> RAMDISK driver initialized: 16 RAM disks of 4096K size > 1024 blocksize > >> Blkmem copyright 1998,1999 D. Jeff Dionne > >> Blkmem copyright 1998 Kenneth Albanowski > >> Blkmem 1 disk images: > >> 0: 106A308-1126307 [VIRTUAL 106A308-1126307] (RO) > >> VFS: Mounted root (romfs filesystem) readonly. > >> Warning: unable to open an initial console. > >> > >> > >> > >> This message resent by the uclinux-dev at uclinux.org list > >> server http://www.uClinux.org/ > >> > >This message resent by the uclinux-dev at uclinux.org list > server http://www.uClinux.org/ > > > > -------------------------------- > > > > > This message resent by the uclinux-dev at uclinux.org list > server http://www.uClinux.org/ > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From sheeliang at lit.org.sg Wed Mar 20 04:49:20 2002 From: sheeliang at lit.org.sg (sheeliang at lit.org.sg) Date: Wed, 20 Mar 2002 17:49:20 +0800 Subject: [uClinux-dev] Network cards on Integrator AP In-Reply-To: <716A7F62A194D41195AB001083FC3AE4BF61F2@inlmx1.inl.intersil.com>; from martijn.schipper@intersil.com on Wed, Mar 20, 2002 at 03:14:34AM -0500 References: <716A7F62A194D41195AB001083FC3AE4BF61F2@inlmx1.inl.intersil.com> Message-ID: <20020320174920.A1671@localhost.localdomain> Oh yeah, forgot to mention that. It's the simplest setup possible. One network card on the Integrator and another on my desktop linked by a network cross cable. After turning on the debug options in the drivers, whenenver I send packets, I have messages like 'Too much work in interrupt'(for the 3c59x) or 'Tx FIFO error'(for pcnet32). On Wed, Mar 20, 2002 at 03:14:34AM -0500, Schipper, Martijn (Contractor) wrote: > Hello Shee(?), > > It looks to me as a routing problem. Can you give some more information > on the IP settings on your machine? > > Best regards, > > Martijn Schipper > Software Engineer > Intersil Netherlands BV. > Rembrandtlaan 1a > 3723 BG Bilthoven > The Netherlands > Tel: +31 (0)30 2296076 > Fax: +31 (0)30 2296061 > E-mail: Martijn.Schipper at Intersil.com > Internet: www.intersil.com > > > -----Original Message----- > From: sheeliang at lit.org.sg [mailto:sheeliang at lit.org.sg] > Sent: woensdag 20 maart 2002 8:53 > To: uclinux-dev at uclinux.org > Subject: Re: [uClinux-dev] Network cards on Integrator AP > > Okok, seems like nobody has much idea. I have moved on a bit though. > Now the problem is that I can receive packets, but I cannot send. > I can receive arp-whohas and ICMP echo request, but it seems like I can > only > send arp-replies. This is pretty strange, being able to send out arp > replies > but not ICMP echo replies. Loopback works fine though. The card is a AMD > pcnet32. > > Going to try 3Com Vortex next. > > On Tue, Mar 19, 2002 at 05:14:15PM +0800, sheeliang at lit.org.sg wrote: > > Wondering if anybody has any experience getting network cards to work > on > > Integrator AP, which uses the PCI V3 chipset? I have tried pcnet32 and > 3com > > cards without avail. Initialization runs ok, ie, I can see the correct > ethernet > > addresses, IRQ assigned, ioport, etc on startup, but actual pinging on > both > > sides fail. The core is a ARM7TDMI running uClinux-2.4.17. Help, > anyone? > > > > This message resent by the uclinux-dev at uclinux.org list server > http://www.uClinux.org/ > This message resent by the uclinux-dev at uclinux.org list server > http://www.uClinux.org/ > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From gmenie at akamai.com Wed Mar 20 04:59:50 2002 From: gmenie at akamai.com (Menie, Georges) Date: Wed, 20 Mar 2002 04:59:50 -0500 Subject: [uClinux-dev] 68*328 timer was: "Calibrating delay loop..." crashes Message-ID: That's interesting, i did notice the wrong timing and I wanted to look at it, could you share the modifications you made ? do you have a reasonable accuracy now ? Regards, Georges > -----Original Message----- > From: Daniel Haensse [mailto:daniel.haensse at fhk.usz.ch] > Sent: mercredi 20 mars 2002 08:40 > To: uclinux-dev at uclinux.org > Subject: Re: [uClinux-dev] "Calibrating delay loop..." crashes > > > When I came across the problem I spend a lot of time > searching in the calibrate and timer irq functions and > finally found the > memory range problem. It looks like the calibration_delay > function is the > first function that "rocks" with C code and because of the > bad settings the > system crashes. There is also a bad definition in > BSP_sched_init in config.c > that makes the tick timer slightly slower (2seconds on 1 > minute), but this > gives some extra boost of bogomips on the bootmessage, which > is not true. > I corrected the function for the dragonix board. > > > from my crt0rom.S as an example, somebody introduced this > mechanism earlier > > .global _start > .global _stext > > .global _rambase > .global _ramvec > .global _ramstart > .global _ramend > > and later in the file: > > movel #__ramend, %d0 > movel %d0, _ramend > > while __ramend is defined in my rom.ld, later in the kernel in setup.c > memory_start = PAGE_ALIGN(_ramstart); > memory_end = _ramend; /* by now the stack is part of the init task */ > > the fine difference in the var name is the _ > also check for the other vars to be correctly set. > > Maybe it is really an other problem with the same symptoms, > but I would check > this first. > > Dani > > > > > > Am Dienstag, 19. M?rz 2002 23.08 schrieben Sie: > > Could it be a timer interrupt problem ? > > One of the first thing which is setup is the timer irq, just > > before the calibrate_delay() function... > > > > Regards, > > Georges > > > > > -----Original Message----- > > > From: Jonathan Schmidt [mailto:pgccoop at hotmail.com] > > > Sent: mardi 19 mars 2002 21:04 > > > To: uclinux-dev at uclinux.org > > > Subject: Re: [uClinux-dev] "Calibrating delay loop..." crashes > > > > > > > > > I am now using the fixed.ld file from the de2 (Dragen2 > > > board), as an include > > > from the rom.ld file, as follows: > > > > > > MEMORY > > > { > > > romvec : ORIGIN = 0x10c00000, LENGTH = 0x0000400 > > > flash : ORIGIN = 0x10c00400, LENGTH = 0x0400000 > - 0x0000400 > > > eflash : ORIGIN = 0x11000000, LENGTH = 0 > > > ramvec : ORIGIN = 0x00000000, LENGTH = 0x0000400 > > > ram : ORIGIN = 0x00020000, LENGTH = 0x1000000 > - 0x0020000 > > > eram : ORIGIN = 0x01000000, LENGTH = 0 > > > } > > > > > > INCLUDE arch/m68knommu/platform/68VZ328/pgc_vz/fixed.ld > > > > > > A copy of fixed.ld is not included here because it is > > > available from the CVS > > > in arch/m68knommu/platform/68VZ328/de2/fixed.ld. I am > > > getting the same > > > error (crash at "Calibrating delay loop...") as I was before. > > > I have 16MB > > > of SDRAM (single chip) and 4MB of Intel Flash (single chip). > > > Both are > > > 16bit. It would seem to me that 0x0400000 is in fact 4MB, > > > and 0x1000000 is > > > 16MB. I don't really see a problem here. Perhaps there is > > > something else > > > wrong? This .ld file works just fine with "unified-2.4.6" of > > > Daniel's. > > > > > > I am really confused! Please let me know if anyone has come > > > across this > > > before. Thanks! > > > > > > Jonathan Schmidt > > > > > > >From: Daniel Haensse > > > >Reply-To: uclinux-dev at uclinux.org > > > >To: uclinux-dev at uclinux.org > > > >Subject: Re: [uClinux-dev] "Calibrating delay loop..." crashes > > > >Date: Mon, 18 Mar 2002 07:58:28 +0100 > > > >MIME-Version: 1.0 > > > >Received: from [216.57.57.93] by hotmail.com (3.2) with ESMTP id > > > >MHotMailBE5EDF2700C64004324BD839395D8AEB0; Sun, 17 Mar 2002 > > > > > > 23:15:59 -0800 > > > > > > >Received: (from mdom at localhost)by uclinux-dev-listclinux.org > > > >(8.11.3/8.11.3/SuSE Linux 8.11.1-0.5) id g2I72Qq00654for > > > > > > uclinux-dev-list; > > > > > > >Mon, 18 Mar 2002 02:02:26 -0500 > > > >From owner-uclinux-dev at uclinux.org Sun, 17 Mar 2002 > 23:17:38 -0800 > > > >Message-Id: <200203180702.IAA00812 at ds-00122.usz.ch> > > > >X-Mailer: KMail [version 1.3] > > > >References: > > > >In-Reply-To: > > > >Sender: owner-uclinux-dev at uclinux.org > > > >Precedence: bulk > > > > > > > >Hi Jonathan, > > > > > > > >Am Mittwoch, 13. M?rz 2002 22.44 schrieben Sie: > > > > > I have been building kernels from Daniel Haenesse's > > > > > > "unified-2.4.6" > > > > > > > > distribution for my DragonBall VZ board, and they have > > > > > > been working just > > > > > > > > fine. I tried yesterday to build a kernel straight from > > > > > > the newest > > > > > > >uClinux > > > > > > > > > CVS (2.4.17 mostly, as I understand). When the image is > > > > > > programmed to > > > > > > >the > > > > > > > > > flash (using a modified "2flash" to support the > > > > > > TE28F320B3 Intel flash), > > > > > > > > and I reboot, the kernel hangs at "Calibrating delay > loop...". I > > > > > > > >checked > > > > > > > > > in the .c file that is responsible for this part of the > > > > > > bootup sequence, > > > > > > > > and there doesn't seem to be anything that could > cause a crash. > > > > > > > > > > I was just wondering if anyone has had this type of > > > > > > problem with a VZ > > > > > > >board > > > > > > > > > running a 2.4 kernel before, and if so, how did you > > > > > > resolve it. Thanks > > > > > > >in > > > > > > > > > advance for any help you can provide. > > > > > > > >I had the same bug with 2.0.38. There is something wrong > with your > > > >initialisation of the start and end of the sdram or flash. > > > > > > > >Dani ( Haensse :-) ) > > > > > > > > > > _________________________________________________________________ > > > > > Join the world?s largest e-mail service with MSN Hotmail. > > > > > http://www.hotmail.com > > > > > > > > > > This message resent by the uclinux-dev at uclinux.org list server > > > > > http://www.uClinux.org/ > > > > > > > >This message resent by the uclinux-dev at uclinux.org list server > > > >http://www.uClinux.org/ > > > > > > _________________________________________________________________ > > > Join the world's largest e-mail service with MSN Hotmail. > > > http://www.hotmail.com > > > > > > This message resent by the uclinux-dev at uclinux.org list > > > server http://www.uClinux.org/ > > > > This message resent by the uclinux-dev at uclinux.org list server > > http://www.uClinux.org/ > This message resent by the uclinux-dev at uclinux.org list > server http://www.uClinux.org/ > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From julienb at actimage.fr Wed Mar 20 04:55:56 2002 From: julienb at actimage.fr (Julien Boibessot) Date: Wed, 20 Mar 2002 10:55:56 +0100 Subject: [uClinux-dev] Gdb+bdm+5272C3 References: <7FF1185EA0D3D511BC0500B0D0D0C24A1B5899@melexc01.tecom.com.au> <05a401c1cbff$6784a5d0$8947ec95@bruker.fr> <3C952C62.2080506@snapgear.com> <009701c1ce53$7b969c80$8947ec95@bruker.fr> <3C95DFA9.F505952D@snapgear.com> Message-ID: <05a801c1cff5$6f6a9e40$8947ec95@bruker.fr> Hi Greg, thank you for your help !!! I've found my problem: I was using GDB 5.1.1 and the patch from Chris Johns seems to work only with gdb 5.0 and lower..... So I've downloaded a 5.0 source archive, patche it and recompile it and now "continue" command works and I can debug my linux kernel..... I know there are no bugs in kernel, I'm just playing with gdb-bdm....:-)))) Julien ----- Original Message ----- From: "gerg" To: Sent: Monday, March 18, 2002 1:38 PM Subject: Re: [uClinux-dev] Gdb+bdm+5272C3 > > Hi Julien, > > Julien Boibessot wrote: > > Well, in fact my gdb init script is the one you posted on this list some > > times ago.....:-) > > And it initializes DRAM controller...... > > OK :-) > > Check after you do the "load" that the code looks right in memory. > Dump it and check. If it looks good then single step through > some code to verify that you can execute code. > > Regards > Greg > > > ------------------------------------------------------------------------ > Greg Ungerer -- Chief Software Wizard EMAIL: gerg at snapgear.com > Snapgear PHONE: +61 7 3279 1822 > 825 Stanley St, FAX: +61 7 3279 1820 > Woolloongabba, QLD, 4102, Australia WEB: www.snapgear.com > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From stefan.jonsson at faltcom.se Wed Mar 20 05:20:22 2002 From: stefan.jonsson at faltcom.se (Stefan Jonsson) Date: Wed, 20 Mar 2002 11:20:22 +0100 Subject: [uClinux-dev] EB40 boot and download, again. Message-ID: Hello list, With my EB40LS+uCnet from Lineo, everything is of course easy since there is a very "easy-to-use" -bootloader preinstalled and the "userland" that installs from the Lineo CD is already configured for that target. However I have my EB40 (standard) + MEC01 on which I want to run uClinux. So far so good. I have no possibility to make a JTAG cable for the moment so I want to use the serial !!! I manage to download (to EB40 flash) prebuilt testapplications with the "bincom" (in ms-windows) that comes with the card from Atmel. I want to download via the serial cable from linux since that is where I have my "userland" and so on. I don't manage to "speak" with Angel (but I have only tried with minicom. Is it possible to use minicom for download in this case?). What program shall I use for downloading to my EB40+MEC01? Is there a bootloader available for just booting up uClinux on my EB40? (uClinux in that case has to be stored on my MEC since EB40 can not fit it in the tiny flash). Basicly there is two problems for me. 1) Downloading (to specific address) to EB40+MEC. 2) Boot and initialisation of the MEC. (... and I will have to use the serial cable) Anyone have done this? ... or know how to do? ... or have program/bootloader already ready? I am not too familiar with the world of linux yet, so I could need some help (descriptive and easy). Any help is greatly appreciated. Thank you! Regards, Stefan Jonsson, Student, University of Umea, Sweden. This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From daniel.haensse at fhk.usz.ch Wed Mar 20 05:21:15 2002 From: daniel.haensse at fhk.usz.ch (Daniel Haensse) Date: Wed, 20 Mar 2002 11:21:15 +0100 Subject: [uClinux-dev] 68*328 timer was: "Calibrating delay loop..." crashes In-Reply-To: References: Message-ID: <200203201025.LAA04051@ds-00122.usz.ch> Hi Georges, kitkat is going to submit a patch to the uclinux kernel anyway (this will be included). If you guys like it, we could remove the ifdef. Dani void BSP_sched_init(void (*timer_routine)(int, void *, struct pt_regs *)) { #ifdef CONFIG_DRAGONIXVZ /* Restart mode, Enable int, 32KHz, Enable timer */ TCTL = TCTL_OM | TCTL_IRQEN | TCTL_CLKSOURCE_32KHZ | TCTL_TEN; /* Set prescaler (Divide 32.768KHz by 1)*/ TPRER = 0; /* Set compare register 32.768Khz / 32 / 10 = 100 */ TCMP = 327; #elif /* Restart mode, Enable int, 32KHz, Enable timer */ TCTL = TCTL_OM | TCTL_IRQEN | TCTL_CLKSOURCE_32KHZ | TCTL_TEN; /* Set prescaler (Divide 32KHz by 32)*/ TPRER = 31; /* Set compare register 32Khz / 32 / 10 = 100 */ TCMP = 10; #endif request_irq(TMR_IRQ_NUM, timer_routine, IRQ_FLG_LOCK, "timer", NULL); } Am Mittwoch, 20. M?rz 2002 10.59 schrieben Sie: > That's interesting, i did notice the wrong timing and I wanted > to look at it, could you share the modifications you made ? > do you have a reasonable accuracy now ? > > Regards, > Georges > > > -----Original Message----- > > From: Daniel Haensse [mailto:daniel.haensse at fhk.usz.ch] > > Sent: mercredi 20 mars 2002 08:40 > > To: uclinux-dev at uclinux.org > > Subject: Re: [uClinux-dev] "Calibrating delay loop..." crashes > > > > > > When I came across the problem I spend a lot of time > > searching in the calibrate and timer irq functions and > > finally found the > > memory range problem. It looks like the calibration_delay > > function is the > > first function that "rocks" with C code and because of the > > bad settings the > > system crashes. There is also a bad definition in > > BSP_sched_init in config.c > > that makes the tick timer slightly slower (2seconds on 1 > > minute), but this > > gives some extra boost of bogomips on the bootmessage, which > > is not true. > > I corrected the function for the dragonix board. > > > > > > from my crt0rom.S as an example, somebody introduced this > > mechanism earlier > > > > .global _start > > .global _stext > > > > .global _rambase > > .global _ramvec > > .global _ramstart > > .global _ramend > > > > and later in the file: > > > > movel #__ramend, %d0 > > movel %d0, _ramend > > > > while __ramend is defined in my rom.ld, later in the kernel in setup.c > > memory_start = PAGE_ALIGN(_ramstart); > > memory_end = _ramend; /* by now the stack is part of the init task */ > > > > the fine difference in the var name is the _ > > also check for the other vars to be correctly set. > > > > Maybe it is really an other problem with the same symptoms, > > but I would check > > this first. > > > > Dani > > > > Am Dienstag, 19. M?rz 2002 23.08 schrieben Sie: > > > Could it be a timer interrupt problem ? > > > One of the first thing which is setup is the timer irq, just > > > before the calibrate_delay() function... > > > > > > Regards, > > > Georges > > > > > > > -----Original Message----- > > > > From: Jonathan Schmidt [mailto:pgccoop at hotmail.com] > > > > Sent: mardi 19 mars 2002 21:04 > > > > To: uclinux-dev at uclinux.org > > > > Subject: Re: [uClinux-dev] "Calibrating delay loop..." crashes > > > > > > > > > > > > I am now using the fixed.ld file from the de2 (Dragen2 > > > > board), as an include > > > > from the rom.ld file, as follows: > > > > > > > > MEMORY > > > > { > > > > romvec : ORIGIN = 0x10c00000, LENGTH = 0x0000400 > > > > flash : ORIGIN = 0x10c00400, LENGTH = 0x0400000 > > > > - 0x0000400 > > > > > > eflash : ORIGIN = 0x11000000, LENGTH = 0 > > > > ramvec : ORIGIN = 0x00000000, LENGTH = 0x0000400 > > > > ram : ORIGIN = 0x00020000, LENGTH = 0x1000000 > > > > - 0x0020000 > > > > > > eram : ORIGIN = 0x01000000, LENGTH = 0 > > > > } > > > > > > > > INCLUDE arch/m68knommu/platform/68VZ328/pgc_vz/fixed.ld > > > > > > > > A copy of fixed.ld is not included here because it is > > > > available from the CVS > > > > in arch/m68knommu/platform/68VZ328/de2/fixed.ld. I am > > > > getting the same > > > > error (crash at "Calibrating delay loop...") as I was before. > > > > I have 16MB > > > > of SDRAM (single chip) and 4MB of Intel Flash (single chip). > > > > Both are > > > > 16bit. It would seem to me that 0x0400000 is in fact 4MB, > > > > and 0x1000000 is > > > > 16MB. I don't really see a problem here. Perhaps there is > > > > something else > > > > wrong? This .ld file works just fine with "unified-2.4.6" of > > > > Daniel's. > > > > > > > > I am really confused! Please let me know if anyone has come > > > > across this > > > > before. Thanks! > > > > > > > > Jonathan Schmidt > > > > > > > > >From: Daniel Haensse > > > > >Reply-To: uclinux-dev at uclinux.org > > > > >To: uclinux-dev at uclinux.org > > > > >Subject: Re: [uClinux-dev] "Calibrating delay loop..." crashes > > > > >Date: Mon, 18 Mar 2002 07:58:28 +0100 > > > > >MIME-Version: 1.0 > > > > >Received: from [216.57.57.93] by hotmail.com (3.2) with ESMTP id > > > > >MHotMailBE5EDF2700C64004324BD839395D8AEB0; Sun, 17 Mar 2002 > > > > > > > > 23:15:59 -0800 > > > > > > > > >Received: (from mdom at localhost)by uclinux-dev-listclinux.org > > > > >(8.11.3/8.11.3/SuSE Linux 8.11.1-0.5) id g2I72Qq00654for > > > > > > > > uclinux-dev-list; > > > > > > > > >Mon, 18 Mar 2002 02:02:26 -0500 > > > > >From owner-uclinux-dev at uclinux.org Sun, 17 Mar 2002 > > > > 23:17:38 -0800 > > > > > > >Message-Id: <200203180702.IAA00812 at ds-00122.usz.ch> > > > > >X-Mailer: KMail [version 1.3] > > > > >References: > > > > >In-Reply-To: > > > > >Sender: owner-uclinux-dev at uclinux.org > > > > >Precedence: bulk > > > > > > > > > >Hi Jonathan, > > > > > > > > > >Am Mittwoch, 13. M?rz 2002 22.44 schrieben Sie: > > > > > > I have been building kernels from Daniel Haenesse's > > > > > > > > "unified-2.4.6" > > > > > > > > > > distribution for my DragonBall VZ board, and they have > > > > > > > > been working just > > > > > > > > > > fine. I tried yesterday to build a kernel straight from > > > > > > > > the newest > > > > > > > > >uClinux > > > > > > > > > > > CVS (2.4.17 mostly, as I understand). When the image is > > > > > > > > programmed to > > > > > > > > >the > > > > > > > > > > > flash (using a modified "2flash" to support the > > > > > > > > TE28F320B3 Intel flash), > > > > > > > > > > and I reboot, the kernel hangs at "Calibrating delay > > > > loop...". I > > > > > > >checked > > > > > > > > > > > in the .c file that is responsible for this part of the > > > > > > > > bootup sequence, > > > > > > > > > > and there doesn't seem to be anything that could > > > > cause a crash. > > > > > > > > I was just wondering if anyone has had this type of > > > > > > > > problem with a VZ > > > > > > > > >board > > > > > > > > > > > running a 2.4 kernel before, and if so, how did you > > > > > > > > resolve it. Thanks > > > > > > > > >in > > > > > > > > > > > advance for any help you can provide. > > > > > > > > > >I had the same bug with 2.0.38. There is something wrong > > > > with your > > > > > > >initialisation of the start and end of the sdram or flash. > > > > > > > > > >Dani ( Haensse :-) ) > > > > _________________________________________________________________ > > > > > > > > Join the world?s largest e-mail service with MSN Hotmail. > > > > > > http://www.hotmail.com > > > > > > > > > > > > This message resent by the uclinux-dev at uclinux.org list server > > > > > > http://www.uClinux.org/ > > > > > > > > > >This message resent by the uclinux-dev at uclinux.org list server > > > > >http://www.uClinux.org/ > > > > > > > > _________________________________________________________________ > > > > Join the world's largest e-mail service with MSN Hotmail. > > > > http://www.hotmail.com > > > > > > > > This message resent by the uclinux-dev at uclinux.org list > > > > server http://www.uClinux.org/ > > > > > > This message resent by the uclinux-dev at uclinux.org list server > > > http://www.uClinux.org/ > > > > This message resent by the uclinux-dev at uclinux.org list > > server http://www.uClinux.org/ > > This message resent by the uclinux-dev at uclinux.org list server > http://www.uClinux.org/ This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From castet at firstlinux.net Wed Mar 20 07:16:32 2002 From: castet at firstlinux.net (laurent castet) Date: Wed, 20 Mar 2002 04:16:32 -0800 (PST) Subject: [uClinux-dev] usbflash Message-ID: <20020320121632.1D41D36F9@sitemail.everyone.net> An embedded and charset-unspecified text was scrubbed... Name: not available URL: From mathias.fritzson at mecel.se Wed Mar 20 07:33:52 2002 From: mathias.fritzson at mecel.se (mathias.fritzson at mecel.se) Date: Wed, 20 Mar 2002 13:33:52 +0100 Subject: [uClinux-dev] 68*328 timer was: "Calibrating delay loop..." crashes Message-ID: <200203201233.g2KCXOk21731@clinux.org> This is a nice fix, but what is the reason behind it ? I rewrote this part for the '336/'376 a few weeks ago and thought about this problem, since we're not running any time critical stuff yet I left it as it was. Do we lag those two secs due to the latency in the timer interrupt or what's the reason, if this is the proble, as I think it might be there might be a nicer way to fix it.. We have a "reset" in this function void BSP_tick(void) { /* Reset Timer1 */ MCSM2CNT = 65536-2500; MCSM2SIC = MCSM2SIC & ~MCSM2SIC_INT_FLAG; /*TSTAT &= 0;*/ } The "TSTAT &= 0;" line that is in use by the '328 should be the fixed one imho. Since the interrupt delay in the one that should be compensated for.. I'm not sure if changing the frequency could create any other errors.... (In our case we should lower the 2500 part..) Regards /Mathias Fritzson >Hi Georges, > >kitkat is going to submit a patch to the uclinux kernel anyway (this will be >included). If you guys like it, we could remove the >ifdef. > >Dani > >void BSP_sched_init(void (*timer_routine)(int, void *, struct pt_regs *)) >{ >#ifdef CONFIG_DRAGONIXVZ > /* Restart mode, Enable int, 32KHz, Enable timer */ > TCTL = TCTL_OM | TCTL_IRQEN | TCTL_CLKSOURCE_32KHZ | TCTL_TEN; > /* Set prescaler (Divide 32.768KHz by 1)*/ > TPRER = 0; > /* Set compare register 32.768Khz / 32 / 10 = 100 */ > TCMP = 327; >#elif > /* Restart mode, Enable int, 32KHz, Enable timer */ > TCTL = TCTL_OM | TCTL_IRQEN | TCTL_CLKSOURCE_32KHZ | TCTL_TEN; > /* Set prescaler (Divide 32KHz by 32)*/ > TPRER = 31; > /* Set compare register 32Khz / 32 / 10 = 100 */ > TCMP = 10; >#endif > > request_irq(TMR_IRQ_NUM, timer_routine, IRQ_FLG_LOCK, "timer", NULL); >} This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From andersen at codepoet.org Wed Mar 20 07:41:55 2002 From: andersen at codepoet.org (Erik Andersen) Date: Wed, 20 Mar 2002 05:41:55 -0700 Subject: [uClinux-dev] Re: BusyBox `msh' doesn't work correctly with vfork / uClibc In-Reply-To: References: <20020301080856.GA26459@codepoet.org> Message-ID: <20020320124155.GA14782@codepoet.org> On Tue Mar 05, 2002 at 04:57:35PM +0900, Miles Bader wrote: > So, I'd recommend: > > Apply my earlier patch to use _exit > > Think about doing something with the `grave' function similar to what > Stefan did. Could you take a look at the latest and greatest busybox-stable in CVS and let me know if it now works for you? -Erik -- Erik B. Andersen http://codepoet-consulting.com/ --This message was written using 73% post-consumer electrons-- This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From superwen at 263.net Wed Mar 20 07:49:09 2002 From: superwen at 263.net (Kevin) Date: Wed, 20 Mar 2002 20:49:9 +0800 Subject: [uClinux-dev] kernel command line is NULL Message-ID: <20020320124155.77D121DBFBCF0@smtp.263.net> Thanks a lot, Georges, I've checked the device in romfs/dev [root at linux3 dev]# ls @console,c,5,1 @ptyp3,c,2,3 @ptype,c,2,14 @rom6,b,31,6 @ttyp2,c,3,2 @ttypd,c,3,13 @cua0,c,5,64 @ptyp4,c,2,4 @ptypf,c,2,15 @rom7,b,31,7 @ttyp3,c,3,3 @ttype,c,3,14 @cua1,c,5,65 @ptyp5,c,2,5 @ram0,b,1,0 @rom8,b,31,8 @ttyp4,c,3,4 @ttypf,c,3,15 @ipsec,c,36,10 @ptyp6,c,2,6 @ram1,b,1,1 @rom9,b,31,9 @ttyp5,c,3,5 @ttyS0,c,4,64 @kmem,c,1,2 @ptyp7,c,2,7 @random,c,1,8 @tty0,c,4,0 @ttyp6,c,3,6 @ttyS1,c,4,65 @ledman,c,126,0 @ptyp8,c,2,8 @rom0,b,31,0 @tty1,c,4,1 @ttyp7,c,3,7 @urandom,c,1,9 @mem,c,1,1 @ptyp9,c,2,9 @rom1,b,31,1 @tty2,c,4,2 @ttyp8,c,3,8 @zero,c,1,5 @null,c,1,3 @ptypa,c,2,10 @rom2,b,31,2 @tty3,c,4,3 @ttyp9,c,3,9 @ptyp0,c,2,0 @ptypb,c,2,11 @rom3,b,31,3 @tty,c,5,0 @ttypa,c,3,10 @ptyp1,c,2,1 @ptypc,c,2,12 @rom4,b,31,4 @ttyp0,c,3,0 @ttypb,c,3,11 @ptyp2,c,2,2 @ptypd,c,2,13 @rom5,b,31,5 @ttyp1,c,3,1 @ttypc,c,3,12 looks like the console is there--the first one, and major & minor number looks right. I'm really confused. The warning message comes from --------------in init/main.c--- if (open("/dev/console", O_RDWR, 0) < 0) printk("Warning: unable to open an initial console.\n"); ------------------------------- Should I do some modification with it? I'm wondering besides the genromfs problem, what else could cause such a warning message? I actually got confused by it,I mean the console_init is successfull,the rom disk image mount correctly.I also did some memory dump after I download the image.bin into flash. The whole image.bin are in the right place; and I check the romfs.img by mount it on PC.The console&ttyS0 are there......what else should I check? Thanks again and best regards Kevin -----Original Message-------- >RE: RE: [uClinux-dev] kernel command line is NULL >Hi Kevin, > >you did the right thing, the code you commented out is uCdimm specific. >just keep : >command[0] = 0; >before the line: >mach_sched_init = BSP_sched_init; >it's just signify there is no specific boot option... > >genromfs 0.5.1 seems to be the good version, could you check that the device >is created >in your romfs/dev subdir ? > >Regards, >Georges >> -----Original Message----- >> From: Kevin [mailto:superwen at 263.net] >> Sent: mercredi 20 mars 2002 01:49 >> To: uclinux-dev at uclinux.org >> Subject: Re: RE: [uClinux-dev] kernel command line is NULL >> >> >> Thanks a lot, Georges, >> 1,I've look into the config_BSP function,since I'm using a >> mot's M68VZ328ADS board, which doesn't have cs8900, so I >> comment those lines. And the kernel just stucked at >> p=getbenv("APPEND"),so I comment that also. Then it goes a >> little further,but finally I still failed with "Unable to >> open an initial console". I'm wondering if it's caused by the >> comment I did. >> void config_BSP(char *command, int len) >> { >> unsigned char *p; >> >> printk("\n68VZ328 DragonBallVZ support (c) 2001 Lineo, Inc.\n"); >> /****** ----------I comment >> this >> printk("uCdimm serial string [%s]\n",getserialnum()); >> p = cs8900a_hwaddr = gethwaddr(0); >> printk("uCdimm hwaddr %.2x:%.2x:%.2x:%.2x:%.2x:%.2x\n", >> p[0], p[1], p[2], p[3], p[4], p[5]); >> p = getbenv("APPEND"); >> if (p) strcpy(p,command); >> else command[0] = 0; >> ******/ >> mach_sched_init = BSP_sched_init; >> mach_tick = BSP_tick; >> mach_gettimeoffset = BSP_gettimeoffset; >> mach_gettod = BSP_gettod; >> mach_hwclk = NULL; >> mach_set_clock_mmss = NULL; >> // mach_mksound = NULL; >> mach_reset = BSP_reset; >> // mach_debug_init = NULL; >> >> config_M68VZ328_irq(); >> } >> >> 2,"unable to open an initial console", I searched the >> maillist, some other people got the same problem, but most of >> them are because of genromfs, I only got one genromfs in my >> machine, which is in /usr/local/bin/genromfs. I used command >> "genromfs -h" and at top is "genromfs 0.5.1",I'm not sure if >> it's the right one. >> 3,The kernel could get through the warning message "unable to >> open..", but it stucked at the following init thread >> sbin/init -----looks like it links to /bin/init. I'm >> wondering if the mount romfs file system could cause this? >> but it mount successfully. >> >> VFS: Mounted root (romfs filesystem) readonly. >> I traced down the warning message, it's caused by the return >> value of the system call ---open().I mount the romfs.img in >> PC, and the /dev/console is there, however, what I want is >> /dev/ttyS0, it's also there, >> ----------------------- >> if (open("/dev/console", O_RDWR, 0) < 0) ---I modified it >> to "open("/dev/ttyS0", O_RDWR, 0)", but got the same warning. >> printk("Warning: unable to open an initial console.\n"); >> >> (void) dup(0); >> (void) dup(0); >> >> if (execute_command) >> execve(execute_command,argv_init,envp_init); >> execve("/sbin/init",argv_init,envp_init); >> execve("/etc/init",argv_init,envp_init); >> execve("/bin/init",argv_init,envp_init); >> execve("/bin/sh",argv_init,envp_init); >> panic("No init found. Try passing init= option to kernel."); >> }------the end of init-------- >> >> Too many questions, forgive me, I've stucked here for a >> couple of days. >> Thanks again >> Best Regards >> Kevin >> ------------------ >> >> >> -----Original Message-------- >> >RE: [uClinux-dev] kernel command line is NULL >> >It is not a problem, the command line is a way to setup >> extra parameters for >> >the kernel, >> >You may not need it. The code which setup this command line is in the >> >config_BSP() >> >function (see the config.c file for your platform). An empty >> command line is >> >ok. >> > >> >Georges >> > >> >> -----Original Message----- >> >> From: Kevin [mailto:superwen at 263.net] >> >> Sent: lundi 18 mars 2002 19:01 >> >> To: uclinux >> >> Subject: [uClinux-dev] kernel command line is NULL >> >> >> >> >> >> Hi, All, >> >> The kernel command line I got is NULL(please see below), I'm >> >> wondering if I can add it manually? Or where should I check it? >> >> Please advise, thank you very much >> >> >> >> -------------------------------- >> >> here I attached the message I got from uart. >> >> >> >> KERNEL -> TEXT=0x1000400-0x104e308 DATA=0x020000-0x03c000 >> >> BSS=0x03c000-0x04f198 >> >> CAT ROMARRAY! >> >> KERNEL -> ROMFS=0x106a308-0x1125f78 MEM=0x050000-0x800000 >> >> STACK=0x800000-0x80000 >> >> 0 >> >> On node 0 totalpages: 2048 >> >> zone(0): 0 pages. >> >> zone(1): 2048 pages. >> >> zone(2): 0 pages. >> >> Done setup_arch >> >> init/main.c:line 561: Setup_Arch Done >> >> Kernel command line: -------This looks >> >> not good. >> >> Calibrating delay loop... 1.40 BogoMIPS >> >> init/main.c:line 586: Calibrate_delay.Done >> >> Memory available: 7728k/8064k RAM, 0k/0k ROM (311k kernel >> >> code, 188k data) >> >> Dentry-cache hash table entries: 1024 (order: 1, 8192 bytes) >> >> Inode-cache hash table entries: 512 (order: 0, 4096 bytes) >> >> Mount-cache hash table entries: 512 (order: 0, 4096 bytes) >> >> Buffer-cache hash table entries: 1024 (order: 0, 4096 bytes) >> >> Page-cache hash table entries: 2048 (order: 1, 8192 bytes) >> >> POSIX conformance testing by UNIFIX >> >> <6>Linux NET4.0 for Linux 2.4 >> >> <6>Based upon Swansea University Computer Society NET3.039 >> >> Starting kswapd >> >> pty: 256 Unix98 ptys configured >> >> MC68328 serial driver version 1.00 >> >> ttyS0 at 0xfffff900 (irq = 64) is a builtin MC68328 UART >> >> block: 64 slots per queue, batch=16 >> >> RAMDISK driver initialized: 16 RAM disks of 4096K size >> 1024 blocksize >> >> Blkmem copyright 1998,1999 D. Jeff Dionne >> >> Blkmem copyright 1998 Kenneth Albanowski >> >> Blkmem 1 disk images: >> >> 0: 106A308-1126307 [VIRTUAL 106A308-1126307] (RO) >> >> VFS: Mounted root (romfs filesystem) readonly. >> >> Warning: unable to open an initial console. >> >> >> >> >> >> >> >> This message resent by the uclinux-dev at uclinux.org list >> >> server http://www.uClinux.org/ >> >> >> >This message resent by the uclinux-dev at uclinux.org list >> server http://www.uClinux.org/ >> >> >> >> -------------------------------- >> >> >> >> >> This message resent by the uclinux-dev at uclinux.org list >> server http://www.uClinux.org/ >> >This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ -------------------------------- This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From daniel.haensse at fhk.usz.ch Wed Mar 20 07:49:40 2002 From: daniel.haensse at fhk.usz.ch (Daniel Haensse) Date: Wed, 20 Mar 2002 13:49:40 +0100 Subject: [uClinux-dev] 68*328 timer was: "Calibrating delay loop..." crashes In-Reply-To: References: <200203201233.g2KCXOk21731@clinux.org> Message-ID: <200203201253.NAA06045@ds-00122.usz.ch> Hi Mathias, > This is a nice fix, but what is the reason behind it ? >>do you have a reasonable accuracy now ? Well, it depends on the definition of reasonable. An American President had a similair problem some time ago with a different word. :-) For me 2 seconds every minute was not reasonable accuracy and I wanted to be sure that the problem did not come from the hardware itself, because we use an battery backup RTC on the dragonix mainboard that feeds the clk32 signal of the 68vz328. Dani This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From vinayak at ionicmicro.com Wed Mar 20 12:16:54 2002 From: vinayak at ionicmicro.com (Vinayak Kore) Date: Wed, 20 Mar 2002 22:46:54 +0530 Subject: [uClinux-dev] Problem in build Message-ID: <01C1D061.24DC3160@VINAYAKA_98> Hi, I am installing ucLinux source on linux 7.2 machine. I have downloaded uClinux-dist-20020306.tar.gz & unzipped it. As given in instructions for compiling, I could do make xconfig & make dep. But while doing make it gives the following error message - make[1]:m68k-elf-gcc : command not found. During make xconfig, I have selected 5272 processor & kernal 2.0. Please guide me. Regards, Vinayak -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/ms-tnef Size: 1948 bytes Desc: not available URL: From hamilton at SEDSystems.ca Wed Mar 20 12:44:20 2002 From: hamilton at SEDSystems.ca (Kendrick Hamilton) Date: Wed, 20 Mar 2002 11:44:20 -0600 (CST) Subject: [uClinux-dev] Problem in build In-Reply-To: <01C1D061.24DC3160@VINAYAKA_98> Message-ID: On Wed, 20 Mar 2002, Vinayak Kore wrote: > Hi, > > I am installing ucLinux source on linux 7.2 machine. Linux 7.2??? (Redhat Linux 7.2, Suse Linux 7.2) I have downloaded uClinux-dist-20020306.tar.gz & unzipped it. As given in instructions for compiling, I could do make xconfig & make dep. But while doing make it gives the following error message - > > make[1]:m68k-elf-gcc : command not found. Have you installed the m68k-elf tools. They should be in /usr/local/bin (assuming default install path). If you have not, you need to download them from the uClinux web site. http://www.uclinux.org/pub/uClinux/m68k-elf-tools/ You can download source and build them or download the pre-compiled executables. > > During make xconfig, I have selected 5272 processor & kernal 2.0. Please guide me. > > Regards, > Vinayak This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From JLiu at PacesetterElectronics.com Wed Mar 20 16:57:39 2002 From: JLiu at PacesetterElectronics.com (JINGSONG LIU) Date: Wed, 20 Mar 2002 15:57:39 -0600 Subject: [uClinux-dev] need help for mount from linux to my uclinux Message-ID: All : I'm new to uclinux and I got a problem for that mount a dir from a host redhat 7.2 to uclinux of M5272C3. I modify /etc/exports and /etc/hosts in redhat 7.2 and restart service nfs and it seems every is fine in the host site. I can mount some dir from local computer. but when in the command line I issue from uclinux I got "/>mount failed:No such device" uClinux and redhat I had can ping each other and telnet each other. This is the contents of rc from uClinux /etc> cat rc hostname uClinux /bin/expand /etc/ramfs.img /dev/ram0 mount -t proc proc /proc mount -t ext2 /dev/ram0 /var mkdir /var/tmp mkdir /var/log mkdir /var/run mkdir /var/lock ifconfig lo 127.0.0.1 route add -net 127.0.0.0 netmask 255.0.0.0 lo dhcpcd -p -a eth0 & cat /etc/motd and /etc> ifconfig eth0 Link encap:Ethernet HWaddr 00:CF:52:72:C3:01 inet addr:10.0.0.18 Bcast:10.0.0.255 Mask:255.255.255.0 UP BROADCAST NOTRAILERS RUNNING MTU:1500 Metric:1 RX packets:3058 errors:0 dropped:0 overruns:0 frame:0 TX packets:664 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:100 Base address:0x840 lo Link encap:Local Loopback inet addr:127.0.0.1 Mask:255.0.0.0 UP LOOPBACK RUNNING MTU:16436 Metric:1 RX packets:0 errors:0 dropped:0 overruns:0 frame:0 TX packets:0 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:0 and using uClinux-dist-20020214.tar.gz. Woul you please tell how to deal with this thing ? Thank you very much !!!!! Jason Liu This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From ken at reasonability.net Wed Mar 20 18:49:12 2002 From: ken at reasonability.net (Ken Treis) Date: Wed, 20 Mar 2002 15:49:12 -0800 Subject: [uClinux-dev] need help for mount from linux to my uclinux References: Message-ID: <3C991FF8.6080004@reasonability.net> Make sure you have NFS enabled in your uClinux kernel; it is not enabled by default. You should also make sure that you have the portmapper (portmap) running on your uClinux system. HTH, -- Ken Treis ken at reasonability.net JINGSONG LIU wrote: > All : > > I'm new to uclinux and I got a problem for that mount a dir from a host > redhat 7.2 to > uclinux of M5272C3. I modify /etc/exports and /etc/hosts in redhat 7.2 > and > restart > service nfs and it seems every is fine in the host site. I can mount some > dir from local > computer. but when in the command line I issue /mnt> from uclinux > I got "/>mount failed:No such device" uClinux and redhat I had can > ping each > other and > telnet each other. > > This is the contents of rc from uClinux > /etc> cat rc > hostname uClinux > /bin/expand /etc/ramfs.img /dev/ram0 > mount -t proc proc /proc > mount -t ext2 /dev/ram0 /var > mkdir /var/tmp > mkdir /var/log > mkdir /var/run > mkdir /var/lock > ifconfig lo 127.0.0.1 > route add -net 127.0.0.0 netmask 255.0.0.0 lo > dhcpcd -p -a eth0 & > cat /etc/motd > > and > > /etc> ifconfig > eth0 Link encap:Ethernet HWaddr 00:CF:52:72:C3:01 > inet addr:10.0.0.18 Bcast:10.0.0.255 Mask:255.255.255.0 > UP BROADCAST NOTRAILERS RUNNING MTU:1500 Metric:1 > RX packets:3058 errors:0 dropped:0 overruns:0 frame:0 > TX packets:664 errors:0 dropped:0 overruns:0 carrier:0 > collisions:0 txqueuelen:100 > Base address:0x840 > > lo Link encap:Local Loopback > inet addr:127.0.0.1 Mask:255.0.0.0 > UP LOOPBACK RUNNING MTU:16436 Metric:1 > RX packets:0 errors:0 dropped:0 overruns:0 frame:0 > TX packets:0 errors:0 dropped:0 overruns:0 carrier:0 > collisions:0 txqueuelen:0 > > > and using uClinux-dist-20020214.tar.gz. > > Woul you please tell how to deal with this thing ? > > Thank you very much !!!!! > > Jason Liu > > This message resent by the uclinux-dev at uclinux.org list server > http://www.uClinux.org/ > -------------- next part -------------- An HTML attachment was scrubbed... URL: From superwen at 263.net Wed Mar 20 18:55:09 2002 From: superwen at 263.net (Kevin) Date: Thu, 21 Mar 2002 7:55:9 +0800 Subject: [uClinux-dev] Problem in build Message-ID: <20020320234755.E79071CCDDADA@smtp.263.net> Hi, Did you install the m68k-elf- tools? Best Regards Kevin -----Original Message-------- >[uClinux-dev] Problem in build >Hi, > >I am installing ucLinux source on linux 7.2 machine. I have downloaded uClinux-dist-20020306.tar.gz & unzipped it. As given in instructions for compiling, I could do make xconfig & make dep. But while doing make it gives the following error message - > >make[1]:m68k-elf-gcc : command not found. > >During make xconfig, I have selected 5272 processor & kernal 2.0. Please guide me. > >Regards, >Vinayak >x?>";? -------------------------------- This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From gerg at snapgear.com Wed Mar 20 20:15:32 2002 From: gerg at snapgear.com (Greg Ungerer) Date: Thu, 21 Mar 2002 11:15:32 +1000 Subject: [uClinux-dev] need help for mount from linux to my uclinux References: Message-ID: <3C993434.3030105@snapgear.com> Hi Jason, Did you enable NFS support in the linux kernel config? Regards Greg JINGSONG LIU wrote: > All : > > I'm new to uclinux and I got a problem for that mount a dir from a host > redhat 7.2 to > uclinux of M5272C3. I modify /etc/exports and /etc/hosts in redhat 7.2 and > restart > service nfs and it seems every is fine in the host site. I can mount some > dir from local > computer. but when in the command line I issue /mnt> from uclinux > I got "/>mount failed:No such device" uClinux and redhat I had can ping each > other and > telnet each other. > > This is the contents of rc from uClinux > /etc> cat rc > hostname uClinux > /bin/expand /etc/ramfs.img /dev/ram0 > mount -t proc proc /proc > mount -t ext2 /dev/ram0 /var > mkdir /var/tmp > mkdir /var/log > mkdir /var/run > mkdir /var/lock > ifconfig lo 127.0.0.1 > route add -net 127.0.0.0 netmask 255.0.0.0 lo > dhcpcd -p -a eth0 & > cat /etc/motd > > and > > /etc> ifconfig > eth0 Link encap:Ethernet HWaddr 00:CF:52:72:C3:01 > inet addr:10.0.0.18 Bcast:10.0.0.255 Mask:255.255.255.0 > UP BROADCAST NOTRAILERS RUNNING MTU:1500 Metric:1 > RX packets:3058 errors:0 dropped:0 overruns:0 frame:0 > TX packets:664 errors:0 dropped:0 overruns:0 carrier:0 > collisions:0 txqueuelen:100 > Base address:0x840 > > lo Link encap:Local Loopback > inet addr:127.0.0.1 Mask:255.0.0.0 > UP LOOPBACK RUNNING MTU:16436 Metric:1 > RX packets:0 errors:0 dropped:0 overruns:0 frame:0 > TX packets:0 errors:0 dropped:0 overruns:0 carrier:0 > collisions:0 txqueuelen:0 > > > and using uClinux-dist-20020214.tar.gz. > > Woul you please tell how to deal with this thing ? > > Thank you very much !!!!! > > Jason Liu > > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ > > -- ------------------------------------------------------------------------ Greg Ungerer -- Chief Software Wizard EMAIL: gerg at snapgear.com SnapGear PHONE: +61 7 3435 2888 825 Stanley St, FAX: +61 7 3891 3630 Woolloongabba, QLD, 4102, Australia WEB: www.snapgear.com This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From sheeliang at lit.org.sg Wed Mar 20 21:09:58 2002 From: sheeliang at lit.org.sg (sheeliang at lit.org.sg) Date: Thu, 21 Mar 2002 10:09:58 +0800 Subject: [uClinux-dev] Pathetic BogoMIPS Message-ID: <20020321100957.A936@localhost.localdomain> What's the nominal BogoMIPS for an ARM7TDMI? The core is running at 50 MHz, the memory bus at 20 MHz, system bus at 20 MHz and PCI bus at 33 MHz, all default values. The BogoMIPS I get is in a pathetic range of 1.5 BogoMIPS to 4.8 BogoMIPS(after tweaking values). Is this normal? This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From venkateshwarlut at myw.ltindia.com Wed Mar 20 21:49:30 2002 From: venkateshwarlut at myw.ltindia.com (Venkateshwarlu Torlakonda) Date: Thu, 21 Mar 2002 08:19:30 +0530 Subject: [uClinux-dev] Question on USB port Test on MCF5272 evaluation board Message-ID: Hi , Iam using uClinux on MCF5272 Evaluation board. I heard that MCF5272 board contains USB port which is client. I want to connect this port to PC USB port and to see whether PC is detecting USB port.For that Which USB drivers should i include in the uClinux image.Basically my aim is to test USB port of MCF5272 board. Any suggestions will be appreciated. Thanks Venkat.T This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From yu_sea at 263.net Wed Mar 20 21:53:51 2002 From: yu_sea at 263.net (yu) Date: Thu, 21 Mar 2002 10:53:51 +0800 Subject: [uClinux-dev] How to config jffs2 for m5272c3(motorola)? References: <3C991FF8.6080004@reasonability.net> Message-ID: <000f01c1d083$a77461c0$2300a8c0@LocalHost> Could anyone give some statement? Thanks. -------------- next part -------------- An HTML attachment was scrubbed... URL: From jeff at arcturusnetworks.com Wed Mar 20 22:58:12 2002 From: jeff at arcturusnetworks.com (D. Jeff Dionne) Date: Thu, 21 Mar 2002 03:58:12 -0000 Subject: [uClinux-dev] Pathetic BogoMIPS Message-ID: <200203210358.g2L3wCF15547@arcturusnetworks.com> sheeliang at lit.org.sg said: > What's the nominal BogoMIPS for an ARM7TDMI? The core is running at 50 MHz, the > memory bus at 20 MHz, system bus at 20 MHz and PCI bus at 33 MHz, all default > values. The BogoMIPS I get is in a pathetic range of 1.5 BogoMIPS to 4.8 > BogoMIPS(after tweaking values). Is this normal? Yes, without cache. With an 8k cache and running 50MHz in comparison, we get about 44 BogoMIPS (on our S3C4530 based reference platform) insted of 4 BogoMIPS. -- D. Jeff Dionne Jeff at ArcturusNetworks.com -------------- next part -------------- A non-text attachment was scrubbed... Name: Type: application/octet-stream Size: 2 bytes Desc: not available URL: From shankarr at connexus.com.sg Wed Mar 20 23:58:13 2002 From: shankarr at connexus.com.sg (Shankar) Date: Thu, 21 Mar 2002 12:58:13 +0800 Subject: [uClinux-dev] Memory Mapping for ARM7TDMI Message-ID: <000c01c1d095$01fdcf80$1a000a0a@cxs8> Dear All, I have a ARM7TDMI development board and i am trying to port my uclinux kernel version 2.4.4. and I want to run the kernel from the FLASH memory. I am using LH28F640BFHE-PBTL90 Flash from Sharp.It has eight 4K -word parameter blocks and 127-32K Main blocks.But I dont know how to map the different memories i have. I have given the memory mapping for your kind reference. The memory map is: |----------|- 0x43FF FFFF | | | 32MB (SDRAM 1) protected by MPU segement 5 | | |----------| 0x4200 0000 |----------|- 0x41FF FFFF | | | 32MB (SDRAM 0) protected by MPU segement 4 | | |----------| 0x4000 0000 |----------|- 0x85FF FFFF | | ___ | 16MB ( for Peripheral 2 memory interface) --- CS3 | | |----------| 0x8500 0000 |----------|- 0x84FF FFFF | | ___ | 16MB ( for Peripheral 1 memory interface) --- CS2 | | |----------| 0x8400 0000 |----------|- 0x80FF FFFF | | ___ | 8MB (FLASH 1) --- CS1 | | |----------| 0x8080 0000 |----------|- 0x807F 0000 | | ___ | 8MB (FLASH 0) --- CS0 | | |----------|- 0x8000 0000 I too eager to know what are the files that i have to do the changes..and where to do the changes in the files ? It is necessary to download the kernel first in SDRAM and then move to Flash part..and how to do this? Is it necessary to know the assembly language of ARM7 processor? Do i want to do any changes regarding my memory map in makefile and Config files? .......it will be great help for me if i get the feedback to kick the process... Thanks in advance, Shankar. This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From yglee at issserver.kaist.ac.kr Thu Mar 21 00:33:18 2002 From: yglee at issserver.kaist.ac.kr (Yun-Gu Lee) Date: Thu, 21 Mar 2002 14:33:18 +0900 Subject: [uClinux-dev] I have linker problems... Message-ID: I'm work with uClinux kernel 2.4.0 I installed cross compiler (binutiles and egcs) in http://www.uclinux.org/pub/ports/arm7dmi and uClibc-0.99 in http://www.uclibc.org But when I compile application, some problems are occurred. Below are example. . ------------------ test1.c --------------------------------- #include "test.h" void good2(); int main() { void (*tests)(); void (*test2)(); tests = test_program; tests(); test2 = good2; test2(); } void good2() { printf("good2!\n"); } ------------------ test2.c --------------------------------- #include "test.h" void test_program() { printf("test\n"); } --------------------- test.h ------------------------------------ #ifndef __TEST_H_ #define __TEST_H_ #include void test_program(); #endif ----------------------------------------------------------------- I complied this code as follows arm-uclinux-gcc -mdisable-got -mcpu=arm7tdmi -c test1.c -o test1.o arm-uclinux-gcc -mdisable-got -mcpu=arm7tdmi -c test2.c -o test2.o arm-uclinux-gcc -nostartfiles -Wl,-m -Wl,armelf_uclinux /usr/local/arm/arm-uclinux/lib/crt0.o test1.o test2.o -o test.elf -mcpu=arm7tdmi I look into the test.elf using arm-uclinux-objdump arm-uclinux-objdump -d test.elf ....................... 00000098
: 98: e1a0c00d mov r12, sp 9c: e92dd800 stmdb sp!, {r11, r12, lr, pc} a0: e24cb004 sub r11, r12, #4 ; 0x4 a4: e24dd008 sub sp, sp, #8 ; 0x8 a8: e59f3030 ldr r3, [pc, #30] ; e0 ac: e08f3003 add r3, pc, r3 b0: e50b3010 str r3, [r11, -#16] b4: e51b3010 ldr r3, [r11, -#16] b8: e1a0e00f mov lr, pc bc: e1a0f003 mov pc, r3 c0: e59f301c ldr r3, [pc, #1c] ; e4 c4: e08f3003 add r3, pc, r3 c8: e50b3014 str r3, [r11, -#20] cc: e51b3014 ldr r3, [r11, -#20] d0: e1a0e00f mov lr, pc d4: e1a0f003 mov pc, r3 d8: e3a00001 mov r0, #1 ; 0x1 dc: e91ba800 ldmdb r11, {r11, sp, pc} e0: 000000a0 andeq r0, r0, r0, lsr #1 e4: 0000001c andeq r0, r0, r12, lsl r0 000000e8 : e8: e1a0c00d mov r12, sp ec: e92dd800 stmdb sp!, {r11, r12, lr, pc} f0: e24cb004 sub r11, r12, #4 ; 0x4 f4: e59f300c ldr r3, [pc, #c] ; 108 f8: e0893003 add r3, r9, r3 fc: e1a00003 mov r0, r3 100: eb0000d3 bl 454 104: e91ba800 ldmdb r11, {r11, sp, pc} 108: 0000205c andeq r2, r0, r12, asr r0 0000010c : 10c: e1a0c00d mov r12, sp 110: e92dd800 stmdb sp!, {r11, r12, lr, pc} 114: e24cb004 sub r11, r12, #4 ; 0x4 118: e59f300c ldr r3, [pc, #c] ; 12c 11c: e0893003 add r3, r9, r3 120: e1a00003 mov r0, r3 124: eb0000ca bl 454 128: e91ba800 ldmdb r11, {r11, sp, pc} 12c: 00002064 andeq r2, r0, r4, rrx ......................................... As you know, the register r3 of address:bc is wrong. a8: e59f3030 ldr r3, [pc, #30] ; e0 ac: e08f3003 add r3, pc, r3 The value of r3 maybe is 0xac + 0xa0 + 0x8. But it must be 0x10c. But, the register r3 of addres:d4 is right. c0: e59f301c ldr r3, [pc, #1c] ; e4 c4: e08f3003 add r3, pc, r3 The value of r3 maybe is 0x1c + 0xc4 + 0x8 = 0xe8 Yes. it is the address of good2() function. What's the problem? Is liker option wrong? thank you. This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From sheeliang at lit.org.sg Thu Mar 21 01:38:32 2002 From: sheeliang at lit.org.sg (sheeliang at lit.org.sg) Date: Thu, 21 Mar 2002 14:38:32 +0800 Subject: [uClinux-dev] Network cards on Integrator AP In-Reply-To: ; from eauth@softsys.co.at on Wed, Mar 20, 2002 at 10:32:52AM +0100 References: <20020320155302.A1154@localhost.localdomain> Message-ID: <20020321143832.A1202@localhost.localdomain> I think you're right, the values of skb->data is aligned on a 2-byte boundary instead of 4 or 16. So, um, how do I fix it? I can't just decrement it by 2, can I? On Wed, Mar 20, 2002 at 10:32:52AM +0100, Erwin Authried wrote: > Sounds pretty much like a packet alignment problem somewhere in the driver. > Maybe a missing "packed" attribute that doesn't cause problems with other > architectures. I had similar problems with the Orinoco driver on the Atmel > AT91 cpus. > > -Erwin This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From jim at stardot-tech.com Thu Mar 21 02:00:56 2002 From: jim at stardot-tech.com (Jim Treadway) Date: Wed, 20 Mar 2002 23:00:56 -0800 (PST) Subject: [uClinux-dev] Re: BusyBox `msh' doesn't work correctly with vfork / uClibc In-Reply-To: <20020320124155.GA14782@codepoet.org> Message-ID: On Wed, 20 Mar 2002, Erik Andersen wrote: > On Tue Mar 05, 2002 at 04:57:35PM +0900, Miles Bader wrote: > > So, I'd recommend: > > > > Apply my earlier patch to use _exit > > > > Think about doing something with the `grave' function similar to what > > Stefan did. > > Could you take a look at the latest and greatest busybox-stable > in CVS and let me know if it now works for you? The new version works much better on uClinux-coldfire (it no longer exits the shell) but it doesn't evaluate variables in the subshell, so something like "A = 1; B = `eval $A`; echo $B" fails. The following patch fixes this (it's actually a partial merge from some code elsewhere in the uClinux-coldfire distribution). It also moves the check for __UCLIBC__ to after "busybox.h" is included (which is needed for the uClinux-coldfire build). --- busybox/msh.c.orig Wed Mar 20 18:36:30 2002 +++ busybox/msh.c Wed Mar 20 22:51:42 2002 @@ -23,13 +23,6 @@ * Original copyright notice is retained at the end of this file. */ -/* define this to use fork on MMU-systems instead of vfork */ -#if !defined(__UCLIBC__) || defined(__UCLIBC_HAS_MMU__) -#define USE_FORK -#else -#undef USE_FORK -#endif - /* -------- sh.h -------- */ #include @@ -52,6 +45,14 @@ #include "cmdedit.h" #include "busybox.h" + +/* define this to use fork on MMU-systems instead of vfork */ +#if !defined(__UCLIBC__) || defined(__UCLIBC_HAS_MMU__) +#define USE_FORK +#else +#undef USE_FORK +#endif + #ifdef test //#undef BB_FEATURE_COMMAND_EDITING const char *applet_name = "msh"; @@ -248,7 +249,7 @@ static int rlookup(char *n); static struct wdblock *glob(char *cp, struct wdblock *wb); static int subgetc(int ec, int quoted); -static char **makenv(void); +static char **makenv(int all); static char **eval(char **ap, int f); static int setstatus(int s); static int waitfor(int lastpid, int canintr); @@ -2497,7 +2498,7 @@ _exit(execute(t->left, NOPIPE, NOPIPE, FEXEC)); if (wp[0] == NULL) _exit(0); - cp = rexecve(wp[0], wp, makenv()); + cp = rexecve(wp[0], wp, makenv(0)); prs(wp[0]); prs(": "); warn(cp); @@ -2898,7 +2899,7 @@ signal(SIGINT, SIG_DFL); signal(SIGQUIT, SIG_DFL); } - cp = rexecve(t->words[0], t->words, makenv()); + cp = rexecve(t->words[0], t->words, makenv(0)); prs(t->words[0]); prs(": "); err(cp); @@ -3371,14 +3372,14 @@ * names in the dictionary. Keyword assignments * will already have been done. */ -char **makenv() +char **makenv(int all) { register struct wdblock *wb; register struct var *vp; wb = NULL; for (vp = vlist; vp; vp = vp->next) - if (vp->status & EXPORT) + if (all || vp->status & EXPORT) wb = addword(vp->name, wb); wb = addword((char *) 0, wb); return (getwords(wb)); @@ -3659,6 +3660,10 @@ register char *cp, *s; register int i, c; int pf[2]; +#ifndef USE_FORK + char *av[4]; + char **env; +#endif c = readc(); s = e.linep; @@ -3677,6 +3682,9 @@ } if (openpipe(pf) < 0) return (0); +#ifndef USE_FORK + env = makenv(1); +#endif if ((i = XFORK()) == -1) { closepipe(pf); err("try again"); @@ -3714,7 +3722,11 @@ /* NOTE: When we are vfork()'ing we've got to exec here. * Therefore we start the command using /bin/sh -c. */ - execlp("/bin/sh", "/bin/sh", "-c", cp, NULL); + av[0] = "sh"; + av[1] = "-c"; + av[2] = cp; + av[3] = NULL; + execve("/bin/sh", av, env); err("grave: execlp failed\n"); _exit(1); #endif This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From andersen at codepoet.org Thu Mar 21 02:11:40 2002 From: andersen at codepoet.org (Erik Andersen) Date: Thu, 21 Mar 2002 00:11:40 -0700 Subject: [uClinux-dev] Re: BusyBox `msh' doesn't work correctly with vfork / uClibc In-Reply-To: References: <20020320124155.GA14782@codepoet.org> Message-ID: <20020321071139.GA8071@codepoet.org> On Wed Mar 20, 2002 at 11:00:56PM -0800, Jim Treadway wrote: > > Could you take a look at the latest and greatest busybox-stable > > in CVS and let me know if it now works for you? > > The new version works much better on uClinux-coldfire (it no longer exits > the shell) but it doesn't evaluate variables in the subshell, so something > like "A = 1; B = `eval $A`; echo $B" fails. > > The following patch fixes this (it's actually a partial merge from some > code elsewhere in the uClinux-coldfire distribution). Patch looks quite sane. Applied, thanks! > It also moves the check for __UCLIBC__ to after "busybox.h" is included > (which is needed for the uClinux-coldfire build). Oops. That was awfully dumb of me wasn't it. -Erik -- Erik B. Andersen http://codepoet-consulting.com/ --This message was written using 73% post-consumer electrons-- This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From mathias.fritzson at mecel.se Thu Mar 21 02:43:53 2002 From: mathias.fritzson at mecel.se (mathias.fritzson at mecel.se) Date: Thu, 21 Mar 2002 08:43:53 +0100 Subject: [uClinux-dev] 68*328 timer was: "Calibrating delay loop..." crashes Message-ID: <200203210743.g2L7hKk28712@clinux.org> Hi Dani >Hi Mathias, > >> This is a nice fix, but what is the reason behind it ? > >>>do you have a reasonable accuracy now ? >Well, it depends on the definition of reasonable. An American President had a >similair problem some time ago with a different word. :-) > Ok, bad word =) I guess I wanted to know how exact your board is now.. >For me 2 seconds every minute was not reasonable accuracy and I wanted to be >sure that the problem did not come from the hardware itself, because we use >an battery backup RTC on the dragonix mainboard that feeds the clk32 signal >of the 68vz328. > Thats a good reason for you to fix it, I guess I wished to track down the problem that caused the delay... =) /Mathias This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From martijn.schipper at intersil.com Thu Mar 21 02:52:34 2002 From: martijn.schipper at intersil.com (Schipper, Martijn (Contractor)) Date: Thu, 21 Mar 2002 02:52:34 -0500 Subject: [uClinux-dev] Network cards on Integrator AP Message-ID: <716A7F62A194D41195AB001083FC3AE4BF61F5@inlmx1.inl.intersil.com> Hello Shee, You can reserve 2 bytes at the beginning of the skbuff with: skb_reserve (skb, 2); /* word align IP header */ Hope this helps, Martijn Schipper Software Engineer Intersil Netherlands BV. Rembrandtlaan 1a 3723 BG Bilthoven The Netherlands Tel: +31 (0)30 2296076 Fax: +31 (0)30 2296061 E-mail: Martijn.Schipper at Intersil.com Internet: www.intersil.com -----Original Message----- From: sheeliang at lit.org.sg [mailto:sheeliang at lit.org.sg] Sent: donderdag 21 maart 2002 7:39 To: uclinux-dev at uclinux.org Subject: Re: [uClinux-dev] Network cards on Integrator AP I think you're right, the values of skb->data is aligned on a 2-byte boundary instead of 4 or 16. So, um, how do I fix it? I can't just decrement it by 2, can I? On Wed, Mar 20, 2002 at 10:32:52AM +0100, Erwin Authried wrote: > Sounds pretty much like a packet alignment problem somewhere in the driver. > Maybe a missing "packed" attribute that doesn't cause problems with other > architectures. I had similar problems with the Orinoco driver on the Atmel > AT91 cpus. > > -Erwin This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From stefan.jonsson at faltcom.se Thu Mar 21 03:17:36 2002 From: stefan.jonsson at faltcom.se (Stefan Jonsson) Date: Thu, 21 Mar 2002 09:17:36 +0100 Subject: [uClinux-dev] compiling from prebuilt toolchain (ARM) Message-ID: Hello list, I have one userland for Lineos EB40LS+uCnet but I want to have a userland built for my other EB40+MEC also. (I think it is too much to modify otherwise maybe? ... and I also want to keep on doing things on my EB40LS, so I need both). I have "installed" the prebuilt arm-elf tool chain. It creates /usr/local/ ... and so on. Where shall I put the source I want to compile into a userland? (no, I'm not so familiar with the world of linux ... yet) Will compiler/linker or such be disturbed by the precence of "the other userland"? Any suggestions? How should I arrange this? Any help would (as usual) be greatly appreciated. Thank you, danke sch?n, tack s? mycket, merci beaucoup, grazias, sukria ... Regards, Stefan Jonsson, Student, University of Umea, Sweden. This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From gmenie at akamai.com Thu Mar 21 03:43:43 2002 From: gmenie at akamai.com (Menie, Georges) Date: Thu, 21 Mar 2002 03:43:43 -0500 Subject: [uClinux-dev] 68*328 timer was: "Calibrating delay loop..." c rashes Message-ID: Thanks Dani, > #ifdef CONFIG_DRAGONIXVZ > /* Restart mode, Enable int, 32KHz, Enable timer */ > TCTL = TCTL_OM | TCTL_IRQEN | TCTL_CLKSOURCE_32KHZ | TCTL_TEN; > /* Set prescaler (Divide 32.768KHz by 1)*/ > TPRER = 0; > /* Set compare register 32.768Khz / 32 / 10 = 100 */ > TCMP = 327; > #elif > /* Restart mode, Enable int, 32KHz, Enable timer */ > TCTL = TCTL_OM | TCTL_IRQEN | TCTL_CLKSOURCE_32KHZ | TCTL_TEN; > /* Set prescaler (Divide 32KHz by 32)*/ > TPRER = 31; > /* Set compare register 32Khz / 32 / 10 = 100 */ > TCMP = 10; > #endif 32,768/32/10 -> 102.4 Hz (actual 68*328 setting) 32,768/1/327 -> 100.2 Hz (DragonixVZ) The timer tick is defined at 100 Hz, that's definitely a patch to be applied on the 68*328 arch. Anyone tried to use a 64 Hz interrupt and change the HZ constant to 64 ? What would be the drawback having the timer running lower than 100 Hz ? it will give more CPU to the running apps, the scheduler will run every 16 ms instead of 10 ms, is this a problem ?. I did a quick test, everything seems fine, and the clock should be even more accurate. Regards, Georges This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From daniel.haensse at fhk.usz.ch Thu Mar 21 03:46:05 2002 From: daniel.haensse at fhk.usz.ch (Daniel Haensse) Date: Thu, 21 Mar 2002 09:46:05 +0100 Subject: [uClinux-dev] 68*328 timer was: "Calibrating delay loop..." crashes In-Reply-To: References: <200203210743.g2L7hKk28712@clinux.org> Message-ID: <200203210850.JAA02682@ds-00122.usz.ch> Hi Mathias, > Ok, bad word =) I guess I wanted to know how exact your board is now.. I think it counts up to the CMP value and needs one cycle to reset to 0, so a value of 1 needs 2 cycles for one period. The prescalor value of 0 means division by 1. f1 = 100Hz (as it should be and defined with HZ)=> 24hours = 8640000 ticks 32768Hz/32/11 = 93.09 Hz = f2 32768Hz/1/328 = 99.9 Hz = f3 8640000 * (1/f2-1/f1) = 6412.5 s => per day the clock has an offset of 107 minutes per day (unfixed) 8640000 * (1/f3-1/f1) = 84.4 s => per day the clock has 1.4 minutes offset per day (fixed) Anyway in either case, I don't want this as my watch. Does anybody have a deeper knowledge of how cron and other stuff works? Would it work to get the time from the RTC every hour to correct this misbehaviour? I think it is a demand, that the tick value is continously increased, isn't it? Dani This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From daniel.haensse at fhk.usz.ch Thu Mar 21 04:13:54 2002 From: daniel.haensse at fhk.usz.ch (Daniel Haensse) Date: Thu, 21 Mar 2002 10:13:54 +0100 Subject: [uClinux-dev] 68*328 timer was: "Calibrating delay loop..." c rashes In-Reply-To: References: Message-ID: <200203210917.KAA03068@ds-00122.usz.ch> George, > Anyone tried to use a 64 Hz interrupt and change the HZ > constant to 64 ? What would be the drawback having the timer > running lower than 100 Hz ? it will give more CPU to the > running apps, the scheduler will run every 16 ms instead > of 10 ms, is this a problem ?. I did a quick test, everything > seems fine, and the clock should be even more accurate. Isn't it possible to use the 33MHz clock for the timers clock? We don't use this, because it makes the coding more complex due to the different processor clocks we use, but I like the 100Hz. As I know, we use 16,20 and 33MHz versions so far (and 66 for the sz in the future) Dani This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From eric.bosch at fourtress.nl Thu Mar 21 04:39:41 2002 From: eric.bosch at fourtress.nl (Eric Bosch (Fourtress)) Date: Thu, 21 Mar 2002 10:39:41 +0100 Subject: [uClinux-dev] Problems with the kernel compilation and uploading user programs to uCsimm Message-ID: Hi guys, I?m a newbie in Linux programming, so I need some help with my problems with the ?uClinux Development Kit for the Motorola Dragonball 68EZ328?. The operating system on my computer is Linux Mandrake version 8.1. I?m involved with developing software with the Lineo target. Unfortunatly, I have some trouble with building the kernel, downloading my ?Hello world? test-program to the target (via the serial cable) and try to download some .bin files with minicom. The following problems arises: 1) When I try to 'make dep' with the sources and all, the following error messages appear: /opt/uClinux/linux/scripts/mkdep *.[chS] > .depend; fi 68302serial.h doesn't need config 68332serial.c doesn't need config make[2]: *** [fastdep] Error 135 make[2]: Leaving directory `/opt/uClinux/linux/drivers/char' make[1]: *** [fastdep] Error 2 make[1]: Leaving directory `/opt/uClinux/linux/drivers' make: *** [dep-files] Error 2 So I don't succeed in building the kernel!! 2) If I compile and build my own ?Hello World? program, and download this program with minicom to RAM, I can not execute it!. I tried several bootloader options like program, go and goram but regrettably with a unsatisfied result. I don?t see ?Hello World? on my screen! Looking at the RAM Addresses 0x00200000 etc, nothing is changed so I wonder if the image is placed correctly in RAM. What is wrong? Thanks, Eric mailto: eric.bosch at fourtress.nl -------------- next part -------------- An HTML attachment was scrubbed... URL: From andersen at codepoet.org Thu Mar 21 06:51:49 2002 From: andersen at codepoet.org (Erik Andersen) Date: Thu, 21 Mar 2002 04:51:49 -0700 Subject: [uClinux-dev] uClibc 0.9.10 released Message-ID: <20020321115149.GA19878@codepoet.org> CodePoet Consulting is pleased to announce the immediate availability of uClibc 0.9.10. This release adds pthreads support (including pthreads support for mmu-less systems!). Additionally, thanks to Manuel Novoa III, we now have a completely new stdio library, which is small, standards compliant, supports pthreads, wide/narrow streams, large files, and can even operate in a low-memory unbuffered mode. Many, many bugs have been fixed and a number of additional applications now compile and run perfectly. Even with all these changes, the uClibc continues to be very small. On x86, a default build of the uClibc C library is still just 168k. To make things more interesting, the release also adds support for C++ constructors and destructors. To make it easy to use uClibc when developing C++ applications, this release also provides a wrapper for the GNU C++ compiler. Of course, for more complex C++ applications, such as those using iostreams, a standard C++ library (libstdc++) is required. A native GNU toolchain (binutils/gcc) that provides libstdc++ linked with uClibc 0.9.10 will be released in the next couple of days, so stay tuned. The uClibc 0.9.10 release can be obtained from: http://www.uclibc.org/downloads/uClibc-0.9.10.tar.bz2 The Changelog for this release is here: http://www.uclibc.org/downloads/Changelog http://www.uclibc.org/downloads/Changelog.full The uClibc web site can be found at: http://www.uclibc.org/ -Erik -- Erik B. Andersen http://codepoet-consulting.com/ --This message was written using 73% post-consumer electrons-- This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From yglee at issserver.kaist.ac.kr Thu Mar 21 07:39:40 2002 From: yglee at issserver.kaist.ac.kr (Yun-Gu Lee) Date: Thu, 21 Mar 2002 21:39:40 +0900 Subject: [uClinux-dev] I have linker problems... In-Reply-To: Message-ID: sorry... mistake... in assembly code, good --> test_program ------------------------------------------------------------------------------------------ I'm work with uClinux kernel 2.4.0 I installed cross compiler (binutiles and egcs) in http://www.uclinux.org/pub/ports/arm7dmi and uClibc-0.99 in http://www.uclibc.org But when I compile application, some problems are occurred. Below are example. . ------------------ test1.c --------------------------------- #include "test.h" void good2(); int main() { void (*tests)(); void (*test2)(); tests = test_program; tests(); test2 = good2; test2(); } void good2() { printf("good2!\n"); } ------------------ test2.c --------------------------------- #include "test.h" void test_program() { printf("test\n"); } --------------------- test.h ------------------------------------ #ifndef __TEST_H_ #define __TEST_H_ #include void test_program(); #endif ----------------------------------------------------------------- I complied this code as follows arm-uclinux-gcc -mdisable-got -mcpu=arm7tdmi -c test1.c -o test1.o arm-uclinux-gcc -mdisable-got -mcpu=arm7tdmi -c test2.c -o test2.o arm-uclinux-gcc -nostartfiles -Wl,-m -Wl,armelf_uclinux /usr/local/arm/arm-uclinux/lib/crt0.o test1.o test2.o -o test.elf -mcpu=arm7tdmi I look into the test.elf using arm-uclinux-objdump arm-uclinux-objdump -d test.elf ....................... 00000098
: 98: e1a0c00d mov r12, sp 9c: e92dd800 stmdb sp!, {r11, r12, lr, pc} a0: e24cb004 sub r11, r12, #4 ; 0x4 a4: e24dd008 sub sp, sp, #8 ; 0x8 a8: e59f3030 ldr r3, [pc, #30] ; e0 ac: e08f3003 add r3, pc, r3 b0: e50b3010 str r3, [r11, -#16] b4: e51b3010 ldr r3, [r11, -#16] b8: e1a0e00f mov lr, pc bc: e1a0f003 mov pc, r3 c0: e59f301c ldr r3, [pc, #1c] ; e4 c4: e08f3003 add r3, pc, r3 c8: e50b3014 str r3, [r11, -#20] cc: e51b3014 ldr r3, [r11, -#20] d0: e1a0e00f mov lr, pc d4: e1a0f003 mov pc, r3 d8: e3a00001 mov r0, #1 ; 0x1 dc: e91ba800 ldmdb r11, {r11, sp, pc} e0: 000000a0 andeq r0, r0, r0, lsr #1 e4: 0000001c andeq r0, r0, r12, lsl r0 000000e8 : e8: e1a0c00d mov r12, sp ec: e92dd800 stmdb sp!, {r11, r12, lr, pc} f0: e24cb004 sub r11, r12, #4 ; 0x4 f4: e59f300c ldr r3, [pc, #c] ; 108 f8: e0893003 add r3, r9, r3 fc: e1a00003 mov r0, r3 100: eb0000d3 bl 454 104: e91ba800 ldmdb r11, {r11, sp, pc} 108: 0000205c andeq r2, r0, r12, asr r0 0000010c : 10c: e1a0c00d mov r12, sp 110: e92dd800 stmdb sp!, {r11, r12, lr, pc} 114: e24cb004 sub r11, r12, #4 ; 0x4 118: e59f300c ldr r3, [pc, #c] ; 12c 11c: e0893003 add r3, r9, r3 120: e1a00003 mov r0, r3 124: eb0000ca bl 454 128: e91ba800 ldmdb r11, {r11, sp, pc} 12c: 00002064 andeq r2, r0, r4, rrx ......................................... As you know, the register r3 of address:bc is wrong. a8: e59f3030 ldr r3, [pc, #30] ; e0 ac: e08f3003 add r3, pc, r3 The value of r3 maybe is 0xac + 0xa0 + 0x8. But it must be 0x10c. But, the register r3 of addres:d4 is right. c0: e59f301c ldr r3, [pc, #1c] ; e4 c4: e08f3003 add r3, pc, r3 The value of r3 maybe is 0x1c + 0xc4 + 0x8 = 0xe8 Yes. it is the address of good2() function. What's the problem? Is liker option wrong? thank you. -----Original Message----- From: owner-uclinux-dev at uclinux.org [mailto:owner-uclinux- dev at uclinux.org]On Behalf Of Yun-Gu Lee Sent: Thursday, March 21, 2002 2:33 PM To: uclinux-dev at uclinux.org Subject: [uClinux-dev] I have linker problems... I'm work with uClinux kernel 2.4.0 I installed cross compiler (binutiles and egcs) in http://www.uclinux.org/pub/ports/arm7dmi and uClibc-0.99 in http://www.uclibc.org But when I compile application, some problems are occurred. Below are example. . ------------------ test1.c --------------------------------- #include "test.h" void good2(); int main() { void (*tests)(); void (*test2)(); tests = test_program; tests(); test2 = good2; test2(); } void good2() { printf("good2!\n"); } ------------------ test2.c --------------------------------- #include "test.h" void test_program() { printf("test\n"); } --------------------- test.h ------------------------------------ #ifndef __TEST_H_ #define __TEST_H_ #include void test_program(); #endif ----------------------------------------------------------------- I complied this code as follows arm-uclinux-gcc -mdisable-got -mcpu=arm7tdmi -c test1.c -o test1.o arm-uclinux-gcc -mdisable-got -mcpu=arm7tdmi -c test2.c -o test2.o arm-uclinux-gcc -nostartfiles -Wl,-m -Wl,armelf_uclinux /usr/local/arm/arm-uclinux/lib/crt0.o test1.o test2.o -o test.elf -mcpu=arm7tdmi I look into the test.elf using arm-uclinux-objdump arm-uclinux-objdump -d test.elf ....................... 00000098
: 98: e1a0c00d mov r12, sp 9c: e92dd800 stmdb sp!, {r11, r12, lr, pc} a0: e24cb004 sub r11, r12, #4 ; 0x4 a4: e24dd008 sub sp, sp, #8 ; 0x8 a8: e59f3030 ldr r3, [pc, #30] ; e0 ac: e08f3003 add r3, pc, r3 b0: e50b3010 str r3, [r11, -#16] b4: e51b3010 ldr r3, [r11, -#16] b8: e1a0e00f mov lr, pc bc: e1a0f003 mov pc, r3 c0: e59f301c ldr r3, [pc, #1c] ; e4 c4: e08f3003 add r3, pc, r3 c8: e50b3014 str r3, [r11, -#20] cc: e51b3014 ldr r3, [r11, -#20] d0: e1a0e00f mov lr, pc d4: e1a0f003 mov pc, r3 d8: e3a00001 mov r0, #1 ; 0x1 dc: e91ba800 ldmdb r11, {r11, sp, pc} e0: 000000a0 andeq r0, r0, r0, lsr #1 e4: 0000001c andeq r0, r0, r12, lsl r0 000000e8 : e8: e1a0c00d mov r12, sp ec: e92dd800 stmdb sp!, {r11, r12, lr, pc} f0: e24cb004 sub r11, r12, #4 ; 0x4 f4: e59f300c ldr r3, [pc, #c] ; 108 f8: e0893003 add r3, r9, r3 fc: e1a00003 mov r0, r3 100: eb0000d3 bl 454 104: e91ba800 ldmdb r11, {r11, sp, pc} 108: 0000205c andeq r2, r0, r12, asr r0 0000010c : 10c: e1a0c00d mov r12, sp 110: e92dd800 stmdb sp!, {r11, r12, lr, pc} 114: e24cb004 sub r11, r12, #4 ; 0x4 118: e59f300c ldr r3, [pc, #c] ; 12c 11c: e0893003 add r3, r9, r3 120: e1a00003 mov r0, r3 124: eb0000ca bl 454 128: e91ba800 ldmdb r11, {r11, sp, pc} 12c: 00002064 andeq r2, r0, r4, rrx ......................................... As you know, the register r3 of address:bc is wrong. a8: e59f3030 ldr r3, [pc, #30] ; e0 ac: e08f3003 add r3, pc, r3 The value of r3 maybe is 0xac + 0xa0 + 0x8. But it must be 0x10c. But, the register r3 of addres:d4 is right. c0: e59f301c ldr r3, [pc, #1c] ; e4 c4: e08f3003 add r3, pc, r3 The value of r3 maybe is 0x1c + 0xc4 + 0x8 = 0xe8 Yes. it is the address of good2() function. What's the problem? Is liker option wrong? thank you. This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From hanel.jm at t-online.de Thu Mar 21 06:56:00 2002 From: hanel.jm at t-online.de (=?ISO-8859-1?b?Ikr8Z2VuIE1hZGVyIg==?=) Date: 21 Mar 2002 11:56 GMT Subject: [uClinux-dev] Java and uClinux Message-ID: <16o1B3-1TaezAC@fwd06.sul.t-online.com> Hi, we want to run a java program on an coldfire MCF5272 with uClinux. Does anybody know if there exists a runtime environment for uClinx or a tool with which we can compile our code for the coldfire? J?rgen This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From chyrag at yahoo.com Thu Mar 21 08:40:43 2002 From: chyrag at yahoo.com (Chirag Kantharia) Date: Thu, 21 Mar 2002 05:40:43 -0800 (PST) Subject: [uClinux-dev] Java and uClinux In-Reply-To: <16o1B3-1TaezAC@fwd06.sul.t-online.com> Message-ID: <20020321134043.1133.qmail@web14310.mail.yahoo.com> try waba.sourceforge.net ;) chyrag. --- "J?gen Mader" wrote: > we want to run a java program on an coldfire MCF5272 with uClinux. > Does anybody know if there exists a runtime environment for uClinx or > a tool with which we can compile our code for the coldfire? ===== Chirag Kantharia __________________________________________________ Do You Yahoo!? Yahoo! Movies - coverage of the 74th Academy Awards? http://movies.yahoo.com/ This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From siddons at bnl.gov Thu Mar 21 08:54:57 2002 From: siddons at bnl.gov (D. Peter Siddons) Date: Thu, 21 Mar 2002 08:54:57 -0500 Subject: [uClinux-dev] 68*328 timer was: "Calibrating delay loop..." crashes References: Message-ID: <3C99E631.AA26CFE3@bnl.gov> Hi Georges, I use the RTC sample timer at 128Hz for the tick without obvious problems, but I don't really care about the absolute time and have never checked. If you like I can set it running and check how well it keeps time. Pete. "Menie, Georges" wrote: > > Anyone tried to use a 64 Hz interrupt and change the HZ > constant to 64 ? What would be the drawback having the timer > running lower than 100 Hz ? it will give more CPU to the > running apps, the scheduler will run every 16 ms instead > of 10 ms, is this a problem ?. I did a quick test, everything > seems fine, and the clock should be even more accurate. > > Regards, > Georges > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ -- ----------------- D. Peter Siddons National Synchrotron Light Source, Bldg. 725D Brookhaven National Laboratory Upton New York 11973 Email: siddons at bnl.gov This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From JLiu at PacesetterElectronics.com Thu Mar 21 10:11:53 2002 From: JLiu at PacesetterElectronics.com (JINGSONG LIU) Date: Thu, 21 Mar 2002 09:11:53 -0600 Subject: [uClinux-dev] need help for mount from linux to my uclinux Message-ID: Greg : I think I did by service nfs start after updating /etc/exports and /etc/hosts files. Another point I used build in mount command in the uClnux to try to mount dir from linux because I couldn't find mount in /bin. Does mount a dir form host have to be use mount in /bin in other words when make config there is a special setting for add mount in /bin instead of build in mount ? your question ? Did you enable NFS support in the linux kernel config? you mean uClinux, I didn't do anything about the uClinux kernel config for NFS support. Actually I don't know how to do that. Thank you very much !!! Jason -----Original Message----- From: Greg Ungerer [mailto:gerg at snapgear.com] Sent: Wednesday, March 20, 2002 7:16 PM To: uclinux-dev at uclinux.org Subject: Re: [uClinux-dev] need help for mount from linux to my uclinux Hi Jason, Did you enable NFS support in the linux kernel config? Regards Greg JINGSONG LIU wrote: > All : > > I'm new to uclinux and I got a problem for that mount a dir from a host > redhat 7.2 to > uclinux of M5272C3. I modify /etc/exports and /etc/hosts in redhat 7.2 and > restart > service nfs and it seems every is fine in the host site. I can mount some > dir from local > computer. but when in the command line I issue /mnt> from uclinux > I got "/>mount failed:No such device" uClinux and redhat I had can ping each > other and > telnet each other. > > This is the contents of rc from uClinux > /etc> cat rc > hostname uClinux > /bin/expand /etc/ramfs.img /dev/ram0 > mount -t proc proc /proc > mount -t ext2 /dev/ram0 /var > mkdir /var/tmp > mkdir /var/log > mkdir /var/run > mkdir /var/lock > ifconfig lo 127.0.0.1 > route add -net 127.0.0.0 netmask 255.0.0.0 lo > dhcpcd -p -a eth0 & > cat /etc/motd > > and > > /etc> ifconfig > eth0 Link encap:Ethernet HWaddr 00:CF:52:72:C3:01 > inet addr:10.0.0.18 Bcast:10.0.0.255 Mask:255.255.255.0 > UP BROADCAST NOTRAILERS RUNNING MTU:1500 Metric:1 > RX packets:3058 errors:0 dropped:0 overruns:0 frame:0 > TX packets:664 errors:0 dropped:0 overruns:0 carrier:0 > collisions:0 txqueuelen:100 > Base address:0x840 > > lo Link encap:Local Loopback > inet addr:127.0.0.1 Mask:255.0.0.0 > UP LOOPBACK RUNNING MTU:16436 Metric:1 > RX packets:0 errors:0 dropped:0 overruns:0 frame:0 > TX packets:0 errors:0 dropped:0 overruns:0 carrier:0 > collisions:0 txqueuelen:0 > > > and using uClinux-dist-20020214.tar.gz. > > Woul you please tell how to deal with this thing ? > > Thank you very much !!!!! > > Jason Liu > > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ > > -- ------------------------------------------------------------------------ Greg Ungerer -- Chief Software Wizard EMAIL: gerg at snapgear.com SnapGear PHONE: +61 7 3435 2888 825 Stanley St, FAX: +61 7 3891 3630 Woolloongabba, QLD, 4102, Australia WEB: www.snapgear.com This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From eric.bosch at fourtress.nl Thu Mar 21 10:50:39 2002 From: eric.bosch at fourtress.nl (Eric Bosch (Fourtress)) Date: Thu, 21 Mar 2002 16:50:39 +0100 Subject: [uClinux-dev] How can I execute my own user program from the uCsimm (HELP!!!) Message-ID: Hi guys, I upload my own "hello world" program to the Lineo target (uCsimm), with minicom - RX - program, and write it in flash ROM . When I try to execute the program with the command go , I only see on my terminal screen: Go! ABCD... But I except "Hello World" on my screen! What is wrong! Can you help me? Thank you guys! Eric mailto: eric.bosch at fourtress.nl -------------- next part -------------- An HTML attachment was scrubbed... URL: From bernd.pfaff at dspsolutions.de Thu Mar 21 10:58:53 2002 From: bernd.pfaff at dspsolutions.de (Bernd Pfaff) Date: Thu, 21 Mar 2002 16:58:53 +0100 Subject: [uClinux-dev] Enableing module support Message-ID: <200203211558.QAA18072@caleb.dspsolutions> Hi I am using uCLinux for the Coldfire M5272C3 board. Latest distribution and latest toolchain. If I enable the module support within the kernel setup the kernel will boot and handle modules, however it will fail to mount the /proc and /tmp file system. See boot messages below. Any ideas what's going wrong? ------------------------------------------------------------------------------ Page-cache hash table entries: 1024 (order: 0, 4096 bytes) POSIX conformance testing by UNIFIX Linux NET4.0 for Linux 2.4 Based upon Swansea University Computer Society NET3.039 kmem_create: Forcing size word alignment - sock Initializing RT netlink socket Starting kswapd kmem_create: Forcing size word alignment - file lock cache ColdFire internal UART serial driver version 1.00 ttyS0 at 0x10000100 (irq = 73) is a builtin ColdFire UART ttyS1 at 0x10000140 (irq = 74) is a builtin ColdFire UART kmem_create: Forcing size word alignment - blkdev_requests block: 64 slots per queue, batch=16 RAMDISK driver initialized: 16 RAM disks of 4096K size 1024 blocksize eth0: FEC ENET Version 0.2, 00:cf:52:72:c3:01 fec: Phy @ 0x0, type 0x001378e2 SLIP: version 0.8.4-NET3.019-NEWTTY (dynamic channels, max=256). CSLIP: code copyright 1989 Regents of the University of California. Blkmem copyright 1998,1999 D. Jeff Dionne Blkmem copyright 1998 Kenneth Albanowski Blkmem 1 disk images: 0: 107DA4-15B9A3 [VIRTUAL 107DA4-15B9A3] (RO) PPP generic driver version 2.4.1 NET4: Linux TCP/IP 1.0 for NET4.0 IP Protocols: ICMP, UDP, TCP kmem_create: Forcing size word alignment - ip_dst_cache IP: routing cache hash table of 512 buckets, 4Kbytes TCP: Hash tables configured (established 512 bind 512) NET4: Unix domain sockets 1.0/SMP for Linux NET4.0. VFS: Mounted root (romfs filesystem) readonly. Freeing unused kernel memory: 20k freed (0xe0000 - 0xe4000) Shell invoked to run file: /etc/rc Command: hostname uClinux Command: /bin/expand /etc/ramfs.img /dev/ram0 Command: mount -t proc proc /proc bad frame format: 00000000 PC: [<000bef2c>] SR: 2000 SP: 001b5f40 a2: 0018b000 d0: 00001000 d1: 000003ef d2: 0018b000 d3: 003fffb9 d4: c0ed0000 d5: 00175580 a0: 00400001 a1: 0018b048 Process mount (pid: 9, stackpage=001b5000) Frame format=4 Stack from 001b5f7c: 000497e0 0018b000 003fffb9 00001000 00000015 0004995c 003f3c04 001113c8 001b5fc4 00049976 003fffb9 001b5fc0 00000015 0004995c 003f3c04 00000021 00000000 00000000 003ff938 0002228e 001f1800 001f1c00 003fffb9 c0ed0000 00175580 003f3820 003fffba 003f3c04 ffffffda 00000015 00000000 40800000 00115502 Call Trace: [<000497e0>] [<0004995c>] [<00049976>] [<0004995c>] [<0002228e>] Code: 22d8 5381 66fa 0800 0001 6702 32d8 0800 0000 6702 pid 9: failed 11 Command: mount -t ext2 /dev/ram0 /var bad frame format: 00000000 PC: [<000bef2c>] SR: 2000 SP: 001b5f40 a2: 00181000 d0: 00001000 d1: 000003ee d2: 00181000 d3: 003fffb5 d4: c0ed0000 d5: 00175580 a0: 00400001 a1: 0018104c Process mount (pid: 10, stackpage=001b5000) Frame format=4 Stack from 001b5f7c: 000497e0 00181000 003fffb5 00001000 00000015 0004995c 003f3c04 001113c8 001b5fc4 00049976 003fffb5 001b5fc0 00000015 0004995c 003f3c04 00000021 00000000 00000000 003ff934 0002228e 001f1800 001f1c00 003fffb5 c0ed0000 00175580 003f3820 003fffb6 003f3c04 ffffffda 00000015 00000000 40800000 00115502 Call Trace: [<000497e0>] [<0004995c>] [<00049976>] [<0004995c>] [<0002228e>] Code: 22d8 5381 66fa 0800 0001 6702 32d8 0800 0000 6702 pid 10: failed 11 Command: mkdir /var/tmp /var/tmp: Read-only file system Command: mkdir /var/log /var/log: Read-only file system Command: mkdir /var/erun /var/run: Read-only tfile system Command: mkhdir /var/lock /var/lock:0 Read-only file system :Command: ifconfig lo 127. 0.0.1 Command: route addc -net 127.0.0.0 netmask 2o55.0.0.0 lo Command: ifcnonfig eth0 192.168.0.2 fig: auto-negotiation on, 100FDX, 100HDX, 10FDX, 10HDX. Command: portmap& [14] Command: mount -tnfs 192.168.0.1:/uclinux /mnt eth0: status: link up, 100MBit Full Duplex, auto-negotiation complete. mount: /proc/mounts: Invalid argument mount: /proc/mounts: Invalid argument Command: cat /etc/motd Welcome to ____ _ _ / __| ||_| _ _| | | | _ ____ _ _ _ _ | | | | | | || | _ \| | | |\ \/ / | |_| | |__| || | | | | |_| |/ \ | ___\____|_||_|_| |_|\____|\_/\_/ | | |_| For further information check: http://www.uclinux.org/ Execution Finished, Exiting Sash command shell (version 1.1.1) This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From justin.wojdacki at analog.com Thu Mar 21 13:17:28 2002 From: justin.wojdacki at analog.com (Justin Wojdacki) Date: Thu, 21 Mar 2002 10:17:28 -0800 Subject: [uClinux-dev] Question on USB port Test on MCF5272 evaluation board References: Message-ID: <3C9A23B8.49E0D652@analog.com> Venkateshwarlu Torlakonda wrote: > > Hi , > Iam using uClinux on MCF5272 Evaluation board. I heard that MCF5272 board > contains USB port which is client. I want to connect this port to PC USB > port and to see whether PC is detecting USB port.For that Which USB > drivers should i include in the uClinux image.Basically my aim is to test > USB port of MCF5272 board. > Any suggestions will be appreciated. > As this USB interface is a device interface, I don't think you'll have much luck with the built-in Linux USB support. Linux's default USB support is for host interfaces, which behave differently to device interfaces. (Note: I could be wrong, somebody could have added support for USB device interfaces). -- ------------------------------------------------- Justin Wojdacki justin.wojdacki at analog.com (408) 350-5032 Communications Processors Group -- Analog Devices This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From Fabrice_Gautier at sdesigns.com Thu Mar 21 14:22:38 2002 From: Fabrice_Gautier at sdesigns.com (Fabrice Gautier) Date: Thu, 21 Mar 2002 11:22:38 -0800 Subject: [uClinux-dev] RE: [uClibc]uClibc 0.9.10 released Message-ID: > -----Original Message----- > From: Erik Andersen [mailto:andersen at codepoet.org] > Sent: Thursday, March 21, 2002 3:52 AM > To: uClibc > Cc: uclinux-dev at uclinux.org > Subject: [uClibc]uClibc 0.9.10 released > > availability of uClibc 0.9.10. This release adds pthreads support > (including pthreads support for mmu-less systems!). What is you level of confidence in the stability of the pthread code, and the thread-safeness of uClibc , especially in a non-mmu target ? Also will gdb correctly debug multi-threaded applications linked with uClibc? (In a word, If i'm going to use threads, do I have to expect headaches debugging ?) Thanks -- Fabrice Gautier, Fabrice_Gautier at sdesigns.com This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From ktreis at keyww.com Thu Mar 21 15:47:10 2002 From: ktreis at keyww.com (Ken Treis) Date: Thu, 21 Mar 2002 12:47:10 -0800 Subject: [uClinux-dev] 5272 C++ patch Message-ID: <3C9A46CE.4010407@keyww.com> Attached is a minor patch to the 5272's config.arch for user-space C++. It removes the -fname-mangling-version-0 flag (which is only necessary for backwards compatibility with compilers older than gcc 2.8) and adds cpu-specific stuff (i.e.-m5307) to $(CXX). -- Ken Treis / Software Engineer Key Technology, Inc. ktreis at keyww.com ---- -------------- next part -------------- An embedded and charset-unspecified text was scrubbed... Name: cxx.patch URL: From siddons at bnl.gov Thu Mar 21 17:12:31 2002 From: siddons at bnl.gov (D. Peter Siddons) Date: Thu, 21 Mar 2002 17:12:31 -0500 Subject: [uClinux-dev] 68*328 timer was: "Calibrating delay loop..." crashes References: Message-ID: <3C9A5ACF.C727E077@bnl.gov> So I set the clock at 12:00.00 today, and just checked it at 17:00.00, using my wristwatch as a calibration (!) and it seems within 1 second, which is probably my twitching finger's problem. This is using uCdimm, and the 128Hz sample timer from the on-chip RTC as the tick. Pete. "Menie, Georges" wrote: > > Anyone tried to use a 64 Hz interrupt and change the HZ > constant to 64 ? What would be the drawback having the timer > running lower than 100 Hz ? it will give more CPU to the > running apps, the scheduler will run every 16 ms instead > of 10 ms, is this a problem ?. I did a quick test, everything > seems fine, and the clock should be even more accurate. > > Regards, > Georges > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ -- ----------------- D. Peter Siddons National Synchrotron Light Source, Bldg. 725D Brookhaven National Laboratory Upton New York 11973 Email: siddons at bnl.gov This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From szhang at tekdigitel.com Thu Mar 21 18:23:50 2002 From: szhang at tekdigitel.com (Stella) Date: Thu, 21 Mar 2002 18:23:50 -0500 Subject: [uClinux-dev] kernel make dep problem Message-ID: <00c001c1d12f$75a99280$7d01a8c0@tekdigitel.com> Hi, all I am porting uClinux to ARM7TDMI platform. When I "make dep", following message are shown on the screen. mkdep: mmap: Permission denied Since I am not familiar with linux kernel programming, can somebody tell me why does this happen? How can I solve this problem? I appreciate any help and information. Regards, Stella -------------- next part -------------- An HTML attachment was scrubbed... URL: From ktreis at keyww.com Thu Mar 21 18:22:22 2002 From: ktreis at keyww.com (Ken Treis) Date: Thu, 21 Mar 2002 15:22:22 -0800 Subject: [uClinux-dev] Board Testing Message-ID: <3C9A6B2E.1020001@keyww.com> One of the challenges that we face when bringing up new boards is that we never know if they're wired correctly. It's one thing to develop on a well-tested eval board, but it's another thing entirely to try to deploy on an alpha board. Right now, we're trying to debug an alpha board that periodically locks up. We can still connect to it via telnet, but we never get a prompt, and our application is dead. We've also seen some kprintfs that say "ksize on unknown page type (index=14)!" (or 41, or 7 3, or...), so I'm starting to wonder if the board might have memory problems. What sort of board tests are people doing? -- Ken Treis / Software Engineer Key Technology, Inc. ktreis at keyww.com ---- This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From bkuhn at lineo.com Thu Mar 21 20:17:03 2002 From: bkuhn at lineo.com (Bernhard Kuhn) Date: Fri, 22 Mar 2002 02:17:03 +0100 Subject: [uClinux-dev] Enableing module support References: <200203211558.QAA18072@caleb.dspsolutions> Message-ID: <3C9A860F.EA68558E@lineo.com> Bernd Pfaff schrieb: > > Any ideas what's going wrong? Yes, this problem (and its solution) was discussed (two months ago?) on the mailing list. Just go to http://www.uclinux.org/pub/uClinux/archive/index.html And search for all posts with "kernel crash with bb-mount" as subject. You are probably especialy interessted in: http://www.uclinux.org/pub/uClinux/archive/6130.html best regards Bernhard -- Bernhard Kuhn, Software Engineer, Lineo Inc. (Where Open Meets Smart) This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From gerg at snapgear.com Thu Mar 21 20:21:27 2002 From: gerg at snapgear.com (Greg Ungerer) Date: Fri, 22 Mar 2002 11:21:27 +1000 Subject: [uClinux-dev] 5272 C++ patch References: <3C9A46CE.4010407@keyww.com> Message-ID: <3C9A8717.5030201@snapgear.com> Hi Ken, Applied. Thanks Greg Ken Treis wrote: > Attached is a minor patch to the 5272's config.arch for user-space C++. > It removes the -fname-mangling-version-0 flag (which is only necessary > for backwards compatibility with compilers older than gcc 2.8) and adds > cpu-specific stuff (i.e.-m5307) to $(CXX). > ------------------------------------------------------------------------ Greg Ungerer -- Chief Software Wizard EMAIL: gerg at snapgear.com SnapGear PHONE: +61 7 3435 2888 825 Stanley St, FAX: +61 7 3891 3630 Woolloongabba, QLD, 4102, Australia WEB: www.snapgear.com This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From bkuhn at lineo.com Thu Mar 21 20:36:49 2002 From: bkuhn at lineo.com (Bernhard Kuhn) Date: Fri, 22 Mar 2002 02:36:49 +0100 Subject: [uClinux-dev] RE: [uClibc]uClibc 0.9.10 released References: Message-ID: <3C9A8AB1.44D7AFEE@lineo.com> Fabrice Gautier schrieb: > (In a word, If i'm going to use threads, do I have to expect headaches > debugging ?) You will have to expect headaches while debugging multithreaded appliactions - regardless if the debugger works properly or not :-) Sorry, but couldn't resist :-) Bernhard -- Bernhard Kuhn, Software Engineer, Lineo Inc. (Where Open Meets Smart) This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From andersen at codepoet.org Thu Mar 21 21:55:17 2002 From: andersen at codepoet.org (Erik Andersen) Date: Thu, 21 Mar 2002 19:55:17 -0700 Subject: [uClinux-dev] Re: [uClibc]uClibc 0.9.10 released In-Reply-To: References: Message-ID: <20020322025517.GA30703@codepoet.org> On Thu Mar 21, 2002 at 11:22:38AM -0800, Fabrice Gautier wrote: > > -----Original Message----- > > From: Erik Andersen [mailto:andersen at codepoet.org] > > Sent: Thursday, March 21, 2002 3:52 AM > > To: uClibc > > Cc: uclinux-dev at uclinux.org > > Subject: [uClibc]uClibc 0.9.10 released > > > > availability of uClibc 0.9.10. This release adds pthreads support > > (including pthreads support for mmu-less systems!). > > What is you level of confidence in the stability of the pthread code, and > the thread-safeness of uClibc , especially in a non-mmu target ? It works. Not all thread safe interfaces are implemented yet. Most interfaces requiring thread safety are, but I have not yet done a through thread safety audit. > Also will gdb correctly debug multi-threaded applications linked with > uClibc? gdb does not yet correctly debug applications using uClibc + threads, so I have no cause to expect it to work properly for mmu-less yet. > (In a word, If i'm going to use threads, do I have to expect headaches > debugging ?) So far, yes. You will really want to debug things on your host system as much as possible before deploying a threaded app on an mmu-less system. -Erik -- Erik B. Andersen http://codepoet-consulting.com/ --This message was written using 73% post-consumer electrons-- This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From gerg at snapgear.com Thu Mar 21 23:15:41 2002 From: gerg at snapgear.com (Greg Ungerer) Date: Fri, 22 Mar 2002 14:15:41 +1000 Subject: [uClinux-dev] need help for mount from linux to my uclinux References: Message-ID: <3C9AAFED.8040204@snapgear.com> Hi Jason, JINGSONG LIU wrote: > I think I did by service nfs start after updating /etc/exports and > /etc/hosts files. Easy way to check the host setup is to try to NFS mount on the host. So if you are exporting /home on your RedHat host then try to mount it there: mount A.B.C.D:/home /mnt If this doesn't work then you have some server setup to figure out first. > Another point I used build in mount command in the uClnux to try to mount > dir from linux > because I couldn't find mount in /bin. Does mount a dir form host have to be > use mount in > /bin in other words when make config there is a special setting for add > mount in /bin > instead of build in mount ? You will need a mount capable of NFS mounting. If I recall correctly the sash shell builint mount is not capable of NFS mounting. > your question ? > Did you enable NFS support in the linux kernel config? > you mean uClinux, I didn't do anything about the uClinux kernel config for > NFS support. > Actually I don't know how to do that. OK, I assume you are not compiling uClinux yourself then? Most likely cause of your problem is simply that the uCLinux kernel you are running isn't configured with NFS support. You probably will need to get the sources and compiler and compile a system for yourself with NFS in uClinux kernel enabled. Regards Greg ------------------------------------------------------------------------ Greg Ungerer -- Chief Software Wizard EMAIL: gerg at snapgear.com SnapGear PHONE: +61 7 3435 2888 825 Stanley St, FAX: +61 7 3891 3630 Woolloongabba, QLD, 4102, Australia WEB: www.snapgear.com This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From vinayak at ionicmicro.com Thu Mar 21 23:30:54 2002 From: vinayak at ionicmicro.com (Vinayak Kore) Date: Fri, 22 Mar 2002 10:00:54 +0530 Subject: [uClinux-dev] Problem in build Message-ID: <01C1D188.756CB320@VINAYAKA_98> Thanks Kendrick!! It worked. Vinayak -----Original Message----- From: Kendrick Hamilton [SMTP:hamilton at SEDSystems.ca] Sent: Wednesday, March 20, 2002 11:14 PM To: 'uclinux-dev at uclinux.org' Subject: Re: [uClinux-dev] Problem in build On Wed, 20 Mar 2002, Vinayak Kore wrote: > Hi, > > I am installing ucLinux source on linux 7.2 machine. Linux 7.2??? (Redhat Linux 7.2, Suse Linux 7.2) I have downloaded uClinux-dist-20020306.tar.gz & unzipped it. As given in instructions for compiling, I could do make xconfig & make dep. But while doing make it gives the following error message - > > make[1]:m68k-elf-gcc : command not found. Have you installed the m68k-elf tools. They should be in /usr/local/bin (assuming default install path). If you have not, you need to download them from the uClinux web site. http://www.uclinux.org/pub/uClinux/m68k-elf-tools/ You can download source and build them or download the pre-compiled executables. > > During make xconfig, I have selected 5272 processor & kernal 2.0. Please guide me. > > Regards, > Vinayak This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/ms-tnef Size: 2606 bytes Desc: not available URL: From sheeliang at lit.org.sg Fri Mar 22 00:27:15 2002 From: sheeliang at lit.org.sg (sheeliang at lit.org.sg) Date: Fri, 22 Mar 2002 13:27:15 +0800 Subject: [uClinux-dev] Network cards on Integrator AP In-Reply-To: <716A7F62A194D41195AB001083FC3AE4BF61F5@inlmx1.inl.intersil.com>; from martijn.schipper@intersil.com on Thu, Mar 21, 2002 at 02:52:34AM -0500 References: <716A7F62A194D41195AB001083FC3AE4BF61F5@inlmx1.inl.intersil.com> Message-ID: <20020322132715.A2860@localhost.localdomain> Looks like alignment of the packets is not the problem after all. ARP sends fine aligned on a 2-byte. Problem seems to be the length of the packets. Transmission succeeds with skb->len <= 74, means ping -s 32 succeeds. Can anyone tell me how to solve this? On Thu, Mar 21, 2002 at 02:52:34AM -0500, Schipper, Martijn (Contractor) wrote: > Hello Shee, > > You can reserve 2 bytes at the beginning of the skbuff with: > > skb_reserve (skb, 2); /* word align IP header */ > > Hope this helps, > > Martijn Schipper This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From uclinux at softhome.net Fri Mar 22 03:17:09 2002 From: uclinux at softhome.net (k.beckhag) Date: Fri, 22 Mar 2002 16:17:9 +0800 Subject: [uClinux-dev] Difference between reset & power off and on. Message-ID: Hi,There, I got a very strange problem, I've made a ROM-version image.bin, and download it successfully, when I just push the reset button of my board(MC68VZ328ADS),it works fairly well, the command prompt comes out,and everything is just fine. But when I power the board off and then power it back, the program then crash, however, it can still output the "ABCDEF", which is the codes in my ctr0_rom.S. In my understanding, the "reset" just reset the dragonball, but power off & on is gonna reset all the stuff. So, I'm wondering if some necessary delay would cause this, not sure.Could somebody give me a hint on this? Thanks a lot. Best Regards Beckhag ------------ This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From venkateshwarlut at myw.ltindia.com Fri Mar 22 05:33:21 2002 From: venkateshwarlut at myw.ltindia.com (Venkateshwarlu Torlakonda) Date: Fri, 22 Mar 2002 16:03:21 +0530 Subject: [uClinux-dev] Compact Flash support under uClinux Message-ID: Hi all, Iam using uClinux dist on MCF5272 board.I interfaced Sandisk CompactFlash (64MB)in TrueIDE mode. These are my hardware details. Sandisk Signal ColdfireSignal A3-10, ATASEL,CSEL,VS2,VS1,GND,GND1 GND A0,A1,A2 A0, A1, A2 D0-D15 D16 - D31 CS1,CS2 CS3,CS4, INTRA INT IORD,IOWR GPIO signals REG,WE,VCC1 VCC What are the drivers to be included to detect my Compact Flash on uClinux kernel image .Iam generating IORD,IOWR throug GPIO signals and chip selects in colilo stuff.I have to generate IORD,IOWR manually before read/write operation to Sandisk In uclinux kernel under BlockDevices XTharddisk support(this is ISA 8 bit bus) is there and also under ATA/IDE/MFM/RLL support is avilable (that has to options old hard disk and IDE/ATA-2). If i include ATA/IDE(old hard disk) and if iam correct what are the necessary files to be changed. I dont know much about uclinux IDE drievrs.Any help on this issue will be highly appreciated.Iam usins uClinux linux-2.4.x kernel. Thanks and Regards Venkat.T This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From vinayak at ionicmicro.com Fri Mar 22 08:06:14 2002 From: vinayak at ionicmicro.com (Vinayak Kore) Date: Fri, 22 Mar 2002 18:36:14 +0530 Subject: [uClinux-dev] How to make a function run? Message-ID: <01C1D1D0.731A7100@VINAYAKA_98> Hi, I have written a simple test_code.c.It has only one function test_func( ), which prints a message continueously. I have made appropriate changes in /user/Makefile , configure.help & config.in as explained in Documentation/Adding-User-Apps-HOWTO. I can even see my code built from the messages displayed duting make. When I download the image.bin to a M5272c3 eval board, the kernal comes up. It gives the prompt as usual. But I don't see the prints as given in my test code. What should be done to make a call to the test_func( )? Is there anything such as task spawn in vxWorks? Am I doing any other mistake? How to activate a perticular function? Please guide. Vinayak Kore -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/ms-tnef Size: 2215 bytes Desc: not available URL: From JLiu at PacesetterElectronics.com Fri Mar 22 08:44:15 2002 From: JLiu at PacesetterElectronics.com (JINGSONG LIU) Date: Fri, 22 Mar 2002 07:44:15 -0600 Subject: [uClinux-dev] need help for mount from linux to my uclinux Message-ID: Hi Greg: Thank you for your help. I can mount on the host without problem. I worked with uClinux-dist-20020214.tar, and did make xconfig, make dep and make. finally I got image.bin. and using dn command from my M5272C3 to download into M5272C3 then go 20000. Is something I did wrong ? I couldn't find setting for enable NFS or something relate to in xconfig interface . Many thanks! Jason -----Original Message----- From: Greg Ungerer [mailto:gerg at snapgear.com] Sent: Thursday, March 21, 2002 10:16 PM To: uclinux-dev at uclinux.org Subject: Re: [uClinux-dev] need help for mount from linux to my uclinux Hi Jason, JINGSONG LIU wrote: > I think I did by service nfs start after updating /etc/exports and > /etc/hosts files. Easy way to check the host setup is to try to NFS mount on the host. So if you are exporting /home on your RedHat host then try to mount it there: mount A.B.C.D:/home /mnt If this doesn't work then you have some server setup to figure out first. > Another point I used build in mount command in the uClnux to try to mount > dir from linux > because I couldn't find mount in /bin. Does mount a dir form host have to be > use mount in > /bin in other words when make config there is a special setting for add > mount in /bin > instead of build in mount ? You will need a mount capable of NFS mounting. If I recall correctly the sash shell builint mount is not capable of NFS mounting. > your question ? > Did you enable NFS support in the linux kernel config? > you mean uClinux, I didn't do anything about the uClinux kernel config for > NFS support. > Actually I don't know how to do that. OK, I assume you are not compiling uClinux yourself then? Most likely cause of your problem is simply that the uCLinux kernel you are running isn't configured with NFS support. You probably will need to get the sources and compiler and compile a system for yourself with NFS in uClinux kernel enabled. Regards Greg ------------------------------------------------------------------------ Greg Ungerer -- Chief Software Wizard EMAIL: gerg at snapgear.com SnapGear PHONE: +61 7 3435 2888 825 Stanley St, FAX: +61 7 3891 3630 Woolloongabba, QLD, 4102, Australia WEB: www.snapgear.com This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From siddons at bnl.gov Fri Mar 22 09:11:29 2002 From: siddons at bnl.gov (D. Peter Siddons) Date: Fri, 22 Mar 2002 09:11:29 -0500 Subject: [uClinux-dev] Re: How to make a function run? References: <01C1D1D0.731A7100@VINAYAKA_98> Message-ID: <3C9B3B91.3E0D83E6@bnl.gov> Hi Vinayak, After you log in to the system, you should find your program in the /bin directory. Just type the name of your program and it should execute. i.e. cd /bin ls ...your program should be included in the listing.... test_code ....your program's output here..... Pete. Vinayak Kore wrote: > > Hi, > I have written a simple test_code.c.It has only one function test_func( > ), which prints a message continueously. I have made appropriate changes > in /user/Makefile , configure.help & config.in as explained in > Documentation/Adding-User-Apps-HOWTO. I can even see my code built from > the messages displayed duting make. > > When I download the image.bin to a M5272c3 eval board, the kernal comes > up. It gives the prompt as usual. But I don't see the prints as given in > my test code. > > What should be done to make a call to the test_func( )? > > Is there anything such as task spawn in vxWorks? > Am I doing any other mistake? > How to activate a perticular function? > > Please guide. > > Vinayak Kore -- ----------------- D. Peter Siddons National Synchrotron Light Source, Bldg. 725D Brookhaven National Laboratory Upton New York 11973 Email: siddons at bnl.gov This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From siddons at bnl.gov Fri Mar 22 09:12:41 2002 From: siddons at bnl.gov (D. Peter Siddons) Date: Fri, 22 Mar 2002 09:12:41 -0500 Subject: [uClinux-dev] Re: How to make a function run? References: <01C1D1D0.731A7100@VINAYAKA_98> Message-ID: <3C9B3BD9.77877741@bnl.gov> By the way, your program does have a main() which calls test_func(), right? Pete. Vinayak Kore wrote: > > Hi, > I have written a simple test_code.c.It has only one function test_func( > ), which prints a message continueously. I have made appropriate changes > in /user/Makefile , configure.help & config.in as explained in > Documentation/Adding-User-Apps-HOWTO. I can even see my code built from > the messages displayed duting make. > > When I download the image.bin to a M5272c3 eval board, the kernal comes > up. It gives the prompt as usual. But I don't see the prints as given in > my test code. > > What should be done to make a call to the test_func( )? > > Is there anything such as task spawn in vxWorks? > Am I doing any other mistake? > How to activate a perticular function? > > Please guide. > > Vinayak Kore -- ----------------- D. Peter Siddons National Synchrotron Light Source, Bldg. 725D Brookhaven National Laboratory Upton New York 11973 Email: siddons at bnl.gov This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From gerg at snapgear.com Fri Mar 22 09:24:36 2002 From: gerg at snapgear.com (gerg) Date: Sat, 23 Mar 2002 00:24:36 +1000 Subject: [uClinux-dev] need help for mount from linux to my uclinux References: Message-ID: <3C9B3EA4.5A179B65@snapgear.com> Hi Jason, JINGSONG LIU wrote: > Thank you for your help. > > I can mount on the host without problem. I worked with > uClinux-dist-20020214.tar, and did make xconfig, make dep and make. finally > I got > image.bin. and using dn command from my M5272C3 to download into > M5272C3 then go 20000. Is something I did wrong ? I couldn't find setting > for enable NFS > or something relate to in xconfig interface . You need to select the option to configure the kernel (same screen as the target board selection menu). Then save and exit. Next a standard Linux kernel configuration screen will come up. You need to enable the NFS support options under the "FileSystems" tab. Once done you should do a "make clean; make dep; make" again. Regards Greg ------------------------------------------------------------------------ Greg Ungerer -- Chief Software Wizard EMAIL: gerg at snapgear.com Snapgear PHONE: +61 7 3279 1822 825 Stanley St, FAX: +61 7 3279 1820 Woolloongabba, QLD, 4102, Australia WEB: www.snapgear.com This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From vinayak at ionicmicro.com Fri Mar 22 11:06:08 2002 From: vinayak at ionicmicro.com (Vinayak Kore) Date: Fri, 22 Mar 2002 21:36:08 +0530 Subject: [uClinux-dev] Re: How to make a function run? Message-ID: <01C1D1E9.94F73A60@VINAYAKA_98> Thanks Pete. But I can't see the file name appearing in the /bin directory. Another thing is, there is nothing like logging in. It directly gives the prompt. Is it wrong? Vinayak -----Original Message----- From: D. Peter Siddons [SMTP:siddons at bnl.gov] Sent: Friday, March 22, 2002 7:41 PM To: uclinux-dev at uclinux.org Subject: [uClinux-dev] Re: How to make a function run? Hi Vinayak, After you log in to the system, you should find your program in the /bin directory. Just type the name of your program and it should execute. i.e. cd /bin ls ..your program should be included in the listing.... test_code ...your program's output here..... Pete. Vinayak Kore wrote: > > Hi, > I have written a simple test_code.c.It has only one function test_func( > ), which prints a message continueously. I have made appropriate changes > in /user/Makefile , configure.help & config.in as explained in > Documentation/Adding-User-Apps-HOWTO. I can even see my code built from > the messages displayed duting make. > > When I download the image.bin to a M5272c3 eval board, the kernal comes > up. It gives the prompt as usual. But I don't see the prints as given in > my test code. > > What should be done to make a call to the test_func( )? > > Is there anything such as task spawn in vxWorks? > Am I doing any other mistake? > How to activate a perticular function? > > Please guide. > > Vinayak Kore -- ----------------- D. Peter Siddons National Synchrotron Light Source, Bldg. 725D Brookhaven National Laboratory Upton New York 11973 Email: siddons at bnl.gov This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/ms-tnef Size: 2973 bytes Desc: not available URL: From dchasse at sierracom.com Fri Mar 22 06:05:45 2002 From: dchasse at sierracom.com (Denise Chasse) Date: Fri, 22 Mar 2002 11:05:45 -0000 Subject: [uClinux-dev] snmpd, cannot find mib module Message-ID: <01C1D191.84659F00.dchasse@sierracom.com> I am trying to extend ucdsnmp with a private mib. When I invoke snmpd, I get the following message: Cannot find module (ABC-MIB): at line 0 in (none). The header and source file for ABC-MIB was generated using the mib2c tool. The file config.h which is included in abc.c defines the path and mibs that are included. I verified that ABC-MIB.txt resides in the defined path. Any idea why I would get the message? Thanks, Denise This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From Anuragb at ami.com Fri Mar 22 12:02:12 2002 From: Anuragb at ami.com (Anurag Bhatia) Date: Fri, 22 Mar 2002 12:02:12 -0500 Subject: [uClinux-dev] out of the box v3 support in ucdsnmp ?? Message-ID: <8CCBDD5583C50E4196F012E79439B45C364465@atl-ms1.megatrends.com> Hi, As known from http://net-snmp.sourceforge.net/snmpv3.html, The ucd-snmp-4.0 (or later) code supports, out of the box, only authNoPriv with MD5 authentication checksums. So, I'm trying to test the same thing. I connect to the agent by running manager on a different machine. It doesn't allow me to connect and gives me usmStatsWrongDigests in response packet. But, if I try the same operation locally (the same machine as agent), it works. I'm using ver 4.1.2 on uClinux. the configuration files on my agent look like as: snmp.conf: defVersion 3 defCommunity public defSecurityLevel authNoPriv defAuthType MD5 snmpd.conf: com2sec AnuragB 155.229.81.221(manager's IP) public group AMIRWGroup usm AnuragB view all included .1 80 access AMIRWGroup "" any noauth exact all all none Have someone got this working?? I want to use authNoPriv (MD5, default in ucdsnmp), create a user and connect to the agent running on my board using a remote manager. Thanks in advance, Anurag This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From vinayak at ionicmicro.com Fri Mar 22 12:09:33 2002 From: vinayak at ionicmicro.com (Vinayak Kore) Date: Fri, 22 Mar 2002 22:39:33 +0530 Subject: [uClinux-dev] Re: How to make a function run? Message-ID: <01C1D1F2.7110C5E0@VINAYAKA_98> Hey Pete, It worked. It was a bug in my makefile. In romfs field of makefile, I had used a tab in place of space after $(ROMFSINST). I corrected it. Then the test_code appeared in /bin. Thanks a lot for your help. Happy weekend ! Vinayak -----Original Message----- From: D. Peter Siddons [SMTP:siddons at bnl.gov] Sent: Friday, March 22, 2002 7:41 PM To: uclinux-dev at uclinux.org Subject: [uClinux-dev] Re: How to make a function run? Hi Vinayak, After you log in to the system, you should find your program in the /bin directory. Just type the name of your program and it should execute. i.e. cd /bin ls ..your program should be included in the listing.... test_code ...your program's output here..... Pete. Vinayak Kore wrote: > > Hi, > I have written a simple test_code.c.It has only one function test_func( > ), which prints a message continueously. I have made appropriate changes > in /user/Makefile , configure.help & config.in as explained in > Documentation/Adding-User-Apps-HOWTO. I can even see my code built from > the messages displayed duting make. > > When I download the image.bin to a M5272c3 eval board, the kernal comes > up. It gives the prompt as usual. But I don't see the prints as given in > my test code. > > What should be done to make a call to the test_func( )? > > Is there anything such as task spawn in vxWorks? > Am I doing any other mistake? > How to activate a perticular function? > > Please guide. > > Vinayak Kore -- ----------------- D. Peter Siddons National Synchrotron Light Source, Bldg. 725D Brookhaven National Laboratory Upton New York 11973 Email: siddons at bnl.gov This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/ms-tnef Size: 3037 bytes Desc: not available URL: From Abhishek.Thakur at motorola.com Fri Mar 22 17:37:42 2002 From: Abhishek.Thakur at motorola.com (Thakur Abhishek-MGIA0740) Date: Fri, 22 Mar 2002 15:37:42 -0700 Subject: [uClinux-dev] snmpd, cannot find mib module Message-ID: <5725AF6EE9FAD51195DC009027E37702C70EFA@az33exm35.corp.mot.com> though this might not be relevant to uClinux mailing group, have you ensured that MIBS and MIBDIRS environment variables are set OK? are you attempting to use snmpd itself or one of the other tools like snmpget/snmpset etc? for snmpd, I did not have to put any MIB files on target. Abhishek > -----Original Message----- > From: Denise Chasse [mailto:dchasse at sierracom.com] > Sent: Friday, March 22, 2002 3:06 AM > To: uclinux-dev at uclinux.org > Subject: [uClinux-dev] snmpd, cannot find mib module > > > > I am trying to extend ucdsnmp with a private mib. When I > invoke snmpd, I > get the following message: > > Cannot find module (ABC-MIB): at line 0 in (none). > > The header and source file for ABC-MIB was generated using > the mib2c tool. > The file config.h which is included in abc.c defines the > path and mibs > that are included. > I verified that ABC-MIB.txt resides in the defined path. Any > idea why I > would get the message? > > Thanks, > > Denise > > > This message resent by the uclinux-dev at uclinux.org list > server http://www.uClinux.org/ > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From shankarr at connexus.com.sg Fri Mar 22 23:45:30 2002 From: shankarr at connexus.com.sg (Shankar) Date: Sat, 23 Mar 2002 12:45:30 +0800 Subject: [uClinux-dev] linker problem... In-Reply-To: Message-ID: <000b01c1d225$900d53e0$1a000a0a@cxs8> Hi, I got errors when i trying to execute the following command in RH linux env... what could be the problem.. > arm-uclinux-gcc -o cjpeg cjpeg.o rdppm.o the errors are.. /usr/local/lib/gcc-lib/arm-uclinux/2.96/ ../../../../arm-uclinux/bin/ld: cannot open crt1.o: No such file or directory.. with regards, Siva This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From vinayak at ionicmicro.com Sat Mar 23 06:55:35 2002 From: vinayak at ionicmicro.com (Vinayak Kore) Date: Sat, 23 Mar 2002 17:25:35 +0530 Subject: [uClinux-dev] problem regarding __THROW Message-ID: <01C1D28F.BF24C9A0@VINAYAKA_98> Hi, While compiling my C code which includes sys-include/pthread.h, I am getting following error /usr/local/lib/gcc-lib/m68k-elf/2.95.3/../../../../m68-elf/sys-include/pthread.h :673:parse error before '__THROW' warning:data defination has no prototype or storage class Simillar error I am getting for bits/sigthread.h I am using RH linux 7.2 machine, gcc 2.95.3 compiler for motorola M5272C3 eval board. Please guide. Regards, Vinayak Kore -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/ms-tnef Size: 2041 bytes Desc: not available URL: From bkuhn at lineo.com Sat Mar 23 17:29:16 2002 From: bkuhn at lineo.com (Bernhard Kuhn) Date: Sat, 23 Mar 2002 23:29:16 +0100 Subject: [uClinux-dev] [Announcement] Software archive closed down Message-ID: <3C9D01BC.E6618CC7@lineo.com> Hi Everybody! My former employer, the Institute of Real-Time Computer Systems, thankfully provided me with write access to it's web server for nearly three years since i quit. But now, the Institute has to restrict the usage of it's resources for former employees. So i removed my Real Time Linux and uClinux software archives from the server. I will try to find a new home for the archives and supply the alternative URL. For those interessted in hosting the software archives in the mean while may download them from http://www.rcs.ei.tum.de/~kuhn/realtime+uclinux.tar.bz2 Hopefully the tarball will be still available 'til mid of the next week before it gets deleted :-) best regards Bernhard -- Bernhard Kuhn, Software Engineer, Lineo Inc. (Where Open Meets Smart) This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From Anuragb at ami.com Sat Mar 23 17:49:10 2002 From: Anuragb at ami.com (Anurag Bhatia) Date: Sat, 23 Mar 2002 17:49:10 -0500 Subject: [uClinux-dev] latest version of ucd-snmp on uclinux?? Message-ID: <8CCBDD5583C50E4196F012E79439B45C36448A@atl-ms1.megatrends.com> Hi, What's the latest version of ucd-snmp on uClinux ? I'm using ucd-snmp 4.1.2 agent on my MCF5272 board and not able to connect using MD5 with authNoPriv. The agent returns an error to the manager (running on a different machine) saying "WrongDigest". Is it a known problem ? I don't want to include openssl and want to work without privacy but with authentiation. Thanks, Anurag This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From Anuragb at ami.com Sat Mar 23 18:00:52 2002 From: Anuragb at ami.com (Anurag Bhatia) Date: Sat, 23 Mar 2002 18:00:52 -0500 Subject: [uClinux-dev] Re: SNMP agent support on uClinux Message-ID: <8CCBDD5583C50E4196F012E79439B45C36448C@atl-ms1.megatrends.com> > Hi, > > I saw the following on uclinux-dev list... Date: Thu Dec 14 2000 - 09:53:29 EST * Previous message: davidm at lineo.com: "Re: [uClinux-dev] Re: SNMP agent support on uClinux" <0600.html> > * In reply to: davidm at lineo.com: "Re: [uClinux-dev] Re: SNMP agent > support on uClinux" <0600.html> > * Next in thread: Juan Jos? : "Re: [uClinux-dev] Re: SNMP agent > support on uClinux" <0604.html> > * Reply: Juan Jos? : "Re: [uClinux-dev] Re: SNMP agent support on > uClinux" <0604.html> > * Messages sorted by: [ date ] [ thread ] [ > subject ] [ author ] > > And now for the patch :-) > -- > David McCullough: Ph: +61 7 3279 1822 Lineo (The Embedded Linux Company) > * application/octet-stream attachment: ucd-snmp.patch.gz > > This message resent by the "uclinux-dev at uclinux.org > 0agent%20support%20on%20uClinux&In-Reply-To=<20001215005329.B823 at moby.more > ton.com.au> list server > > Could someone tell me what's this patch for?? > > Thanks in advance, > Anurag > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From Anuragb at ami.com Sat Mar 23 18:22:17 2002 From: Anuragb at ami.com (Anurag Bhatia) Date: Sat, 23 Mar 2002 18:22:17 -0500 Subject: [uClinux-dev] problems in ucd-snmp 4.1.2?? Message-ID: <8CCBDD5583C50E4196F012E79439B45C36448D@atl-ms1.megatrends.com> Are there any problems in ucd-snmp 4.1.2 under latest uClinux port for MCF5272??? Thanks in advance, Anurag This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From uclinux at softhome.net Wed Mar 20 00:34:39 2002 From: uclinux at softhome.net (Kevin) Date: Wed, 20 Mar 2002 13:34:39 +0800 Subject: [uClinux-dev] romfs offset Message-ID: Hi,There, I got a question about my romdisk image.-I'm playing with mot's MC68VZ328ADS. The message I got from UART is like below, it failed with "unable to open an initial console". KERNEL -> TEXT=0x1000400-0x104e308 DATA=0x020000-0x03c000 BSS=0x03c000-0x04f198 KERNEL -> ROMFS=0x106a308-0x1125f78 MEM=0x050000-0x800000 STACK=0x800000-0x800000 But when I used some hex editor to check the image.bin I got and the B-record file. I found the romfs is actually start from 0x1062138. NOT 0x106a308 as above!!. Why there is an offset? Please advise. Thank you very much. Best Regards Beckhag This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From fly.chen at msa.hinet.net Sun Mar 24 03:40:32 2002 From: fly.chen at msa.hinet.net (Anderson Chen(HiNET) (E-mail)) Date: Sun, 24 Mar 2002 16:40:32 +0800 Subject: [uClinux-dev] Setting ADSL Dial up In-Reply-To: Message-ID: <000101c1d30f$96b7e990$1501a8c0@anderson> Hi, all Can anyone give some hints to setup ADSL connection for Motorola M5272C3 Board I include the rp-pppoe to be compiled. But the operation is not same to rp-pppoe in RedHat. Regards Anderson Chen in Taiwan Mail : meifeng at mail.nihs.tp.edu.tw Mail : fly.chen at msa.hinet.net This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From vinayak at ionicmicro.com Sun Mar 24 06:14:26 2002 From: vinayak at ionicmicro.com (Vinayak Kore) Date: Sun, 24 Mar 2002 16:44:26 +0530 Subject: [uClinux-dev] Multiple definations of '__errno_location' Message-ID: <01C1D353.29C04FA0@VINAYAKA_98> Hi all, I have written code to create two threads. In the code I have included pthread.h & sched.h. In the makefile, I have included LIBPTHREAD (libpthread.a) & also used -D_REENTRANT compiler option. I am getting the following error /uclinux/uClinux-dist/lib/libc/./libc.a(errno_loc.o): In function '__errno_location' /uclinux/uClinux-dist/lib/libc/sysdeps/errno_loc.c:8:multiple definations of '__errno_location' /usr/local/m68k-elf/bin/ld:line 99:12366 Segmentation fault $LINKER -r -o $OFILE.elf2flt $ARG1 Collect2:ld returned 139 exit status make[2]: ***[tasks] Error 1 make[1]: ***[all] Error 2 make: *** [subdirs] Error 1 Please guide !! I am using the following makefile in my directory ------------------------------------------------------------------------------------------- # Makefile for test directory # # CC = gcc LIBPTHREAD = /usr/lib/libpthread.a CFLAGS += -g -O -Wall -I.. -D_REENTRANT EXECS = tasks OBJS = test.o osatask.o all: $(EXECS) $(EXECS): $(OBJS) $(CC) $(LDFLAGS) -o $@ $(OBJS) $(LIBPTHREAD) $(LDLIBS) romfs: $(ROMFSINST) /bin/$(EXECS) clean: rm -f $(EXECS) *.elf *.gdb *.o ---------------------------------------------------------------------------------------- Am I doing any mistake? Do I need to change anything related to the compilor? Please guide. Thanks, Vinayak Kore -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/ms-tnef Size: 2590 bytes Desc: not available URL: From mdurrant at uclinux.org Sun Mar 24 11:04:44 2002 From: mdurrant at uclinux.org (Michael Durrant) Date: Sun, 24 Mar 2002 11:04:44 -0500 Subject: [uClinux-dev] [Announcement] Software archive closed down References: <3C9D01BC.E6618CC7@lineo.com> Message-ID: <004301c1d34d$9d5c50e0$b782fea9@sympatico.ca> Bernhard et al. We should have the archive mirrored on www.uclinux.org in the next few days as www.uclinux.org/bkuhn -- Michael mdurrant at uclinux.org ----- Original Message ----- From: "Bernhard Kuhn" To: Sent: Saturday, March 23, 2002 5:29 PM Subject: [uClinux-dev] [Announcement] Software archive closed down > > Hi Everybody! > > My former employer, the Institute of Real-Time Computer Systems, > thankfully provided me with write access to it's web server for > nearly three years since i quit. But now, the Institute has to > restrict the usage of it's resources for former employees. > > So i removed my Real Time Linux and uClinux software archives > from the server. > > I will try to find a new home for the archives and supply the > alternative URL. For those interessted in hosting the software > archives in the mean while may download them from > > http://www.rcs.ei.tum.de/~kuhn/realtime+uclinux.tar.bz2 > > Hopefully the tarball will be still available 'til mid > of the next week before it gets deleted :-) > > best regards > > Bernhard > > > > -- > Bernhard Kuhn, Software Engineer, Lineo Inc. (Where Open Meets Smart) > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From miles at lsi.nec.co.jp Mon Mar 25 00:08:51 2002 From: miles at lsi.nec.co.jp (Miles Bader) Date: 25 Mar 2002 14:08:51 +0900 Subject: [uClinux-dev] out-of-line BUG Message-ID: Those of you who read the linux-kernel mailing list may have noticed Andrew Morton's patch to shrink the kernel by changing calls to the `BUG' macro in inline functions in header files to call a normal function `out_of_line_bug'. Here's Andrew's original message (Message-ID: <3C982824.60091B4A at zip.com.au>): This is the result of a search-and-destroy mission against oft-repeated strings in the kernel text. These come about due to the toolchain's inability to common up strings between compilation units. In particular, if there is a printk inside an inline function then the string gets expanded into every object file which included that inline function. So the patch goes through and removes the offending strings in various ways. Kernel size is reduced by another 110 kbytes in my build. I haven't sent this to Linus. gcc-3.0.x does the right thing, and as 3.x will likely be the preferred compiler when 2.6 is released, this change isn't needed for Linus kernels. But it's needed for 2.4. [patch deleted] I tried his patch with uclinux, and it applied (with a bit of fuzziness) and works fine. I fixed up my arch tree to use `out_of_line_bug' too, and fixed one missing instance in the kernel generic source (in include/linux/highmem.h). In all, it saved about 40K. So whad'ya think? It's kind of nice to save 40K without any effort, but I guess this will make it into uclinux eventually as long as Marcelo picks it up, and you guys keep tracking 2.4.x (so far only Alan Cox seems to use it). -Miles Here's my patch against uclinux; this is Andrew's patch + similar changes to the v850 arch + one more call in include/linux/highmem.h: -------------- next part -------------- A non-text attachment was scrubbed... Name: oolb.patch Type: text/x-patch Size: 16650 bytes Desc: uClinux version of Andrew Morton's `incredible shrinking kernel patch' URL: -------------- next part -------------- -- .Numeric stability is probably not all that important when you're guessing. From Manigandanv at myw.ltindia.com Mon Mar 25 04:05:25 2002 From: Manigandanv at myw.ltindia.com (Manigandan.V) Date: Mon, 25 Mar 2002 14:35:25 +0530 Subject: [uClinux-dev] Regarding gdb bdm of m5272 References: <3C9D01BC.E6618CC7@lineo.com> <004301c1d34d$9d5c50e0$b782fea9@sympatico.ca> Message-ID: <001001c1d3dc$33ac99e0$4806a8c0@emsysdom> Hi, I am using gdb version as 4.18 --host=i686-pc-linux-gnu --target=m68k-bdm-elf .. But when I try to debug kernel using m68k-elf-gdb --command=01-5272.gdb i am getting error as follows 01-5272.gdb:151:Error in source command file (target bdm /dev/bdmcf0) BDM error: Invalid argument If i do them manually in gdb prompt .. (gdb)target bdm /dev/bdmcf0 Invalid argument (But target is resetted) The ./chk /dev/bdmcf0 is working properly See Attachment 01-5272.gdb Now I have 2 questions 1.What is wrong in my setup either gdb or patch for bdm or 01-5272.gdb script ? 2.I would like to upgrade but where do i get latest gdb, bdm patch for M5272 board ..I tried (ftp://www.cybertec.com.au/pub/bdm/) but I couldnot download, any other download location. 3.My aim is to download colilo into flash of my m5272 board using gdm-bdm .How to do it (any document's for reference) Regards Manigandan -------------- next part -------------- A non-text attachment was scrubbed... Name: 01-5272.gdb Type: application/octet-stream Size: 3562 bytes Desc: not available URL: From stefan.jonsson at faltcom.se Mon Mar 25 05:18:31 2002 From: stefan.jonsson at faltcom.se (Stefan Jonsson) Date: Mon, 25 Mar 2002 11:18:31 +0100 Subject: [uClinux-dev] About make for midori ... Message-ID: Hello list, ... and hello Erwin :-) I have an EB40+AT91MEC extra memory card. I was recommended, on this list, to try midori. I have downloaded midori now and made 'webconfig' and tried to 'make'. I get some errors though. I have put midori in: /hdc/midori/ and the kernel (2.4.6) in: /usr/src/linux/ I have attached the "make messages" with errors in a file. So if anyone wants to have a look at that and help me out, I would be most happy. Reagards, Stefan Jonsson, Student, University of Umea, Sweden. -------------- next part -------------- An embedded and charset-unspecified text was scrubbed... Name: midori_log.txt URL: From julienb at actimage.fr Mon Mar 25 05:40:01 2002 From: julienb at actimage.fr (Julien Boibessot) Date: Mon, 25 Mar 2002 11:40:01 +0100 Subject: [uClinux-dev] Regarding gdb bdm of m5272 References: <3C9D01BC.E6618CC7@lineo.com> <004301c1d34d$9d5c50e0$b782fea9@sympatico.ca> <001001c1d3dc$33ac99e0$4806a8c0@emsysdom> Message-ID: <011501c1d3e9$6cbcae70$8947ec95@bruker.fr> Hi !! for newer gdb look at : http://sources.redhat.com/gdb/ (you should download 5.0 version because patch from cybertec don't work (by me) with 5.1.x) your adress for the bdm patch is ok but the ftp server seems to be down today....:-(( fo your problem I would say it's probably your gdb version, try 5.0, it works fine...... hope it will help you Julien ----- Original Message ----- From: "Manigandan.V" To: Sent: Monday, March 25, 2002 10:05 AM Subject: [uClinux-dev] Regarding gdb bdm of m5272 > Hi, > I am using > gdb version as 4.18 --host=i686-pc-linux-gnu --target=m68k-bdm-elf .. > But when I try to debug kernel using > m68k-elf-gdb --command=01-5272.gdb > i am getting error as follows > 01-5272.gdb:151:Error in source command file (target bdm /dev/bdmcf0) > BDM error: Invalid argument > If i do them manually in gdb prompt .. > (gdb)target bdm /dev/bdmcf0 > Invalid argument (But target is resetted) > The ./chk /dev/bdmcf0 is working properly > See Attachment 01-5272.gdb > > Now I have 2 questions > 1.What is wrong in my setup either gdb or patch for bdm or 01-5272.gdb > script ? > 2.I would like to upgrade but where do i get latest gdb, bdm patch for M5272 > board ..I tried (ftp://www.cybertec.com.au/pub/bdm/) but I couldnot > download, any other download location. > 3.My aim is to download colilo into flash of my m5272 board using gdm-bdm > .How to do it (any document's for reference) > > Regards > Manigandan > > > > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From jody.pearson at cern.ch Mon Mar 25 06:27:22 2002 From: jody.pearson at cern.ch (Jody Pearson) Date: Mon, 25 Mar 2002 12:27:22 +0100 Subject: [uClinux-dev] Picogui... Message-ID: <3C9F099A.F16A1A90@cern.ch> Hi, Has anyone had any success compiling (or should I say linking) picogui under uClinux running on a dragonball ? I am having a terrible time trying to link the thing, after it has apparently compiled successfully ; Does anybody have some patches to the picogui source for exapmle, which enables it to compile and link correctly ? My output is; make[2]: Entering directory `/asis.local/src/uclinux/pgui-dev20020311/pgserver' m68k-elf-gcc -m68000 -O2 -fomit-frame-pointer -I/opt/src/uclinux/uClinux-dist/lib/libc/include /opt/src/uclinux/uClinux-dist/lib/libc/libc.a -fomit-frame-pointer -o pgserve r pgmain.o gcore/libgcore.a widget/libwidget.a appmgr/libappmgr.a net/libnet.a theme/libtheme.a debug/ libdebug.a video/libvideo.a vidwrap/libvidwrap.a font/libfont.a vidbase/libvidbase.a input/libinput.a formats/libformats.a pgmain.o: In function `run_config_process': pgmain.o(.text+0xc): undefined reference to `getpid' pgmain.o(.text+0x34): undefined reference to `vfork' pgmain.o(.text+0x68): undefined reference to `execve' pgmain.o(.text+0x80): undefined reference to `kill' pgmain.o: In function `main': pgmain.o(.text+0xca): undefined reference to `getenv' pgmain.o(.text+0xe2): undefined reference to `strlen' pgmain.o(.text+0x10e): undefined reference to `strcpy' pgmain.o(.text+0x11e): undefined reference to `strcat' pgmain.o(.text+0x13a): undefined reference to `getopt' pgmain.o(.text+0x1fc): undefined reference to `optarg' pgmain.o(.text+0x202): undefined reference to `strchr' pgmain.o(.text+0x216): undefined reference to `optarg' Obviously all of the references exist in uclibc, so i tried using; m68k-elf-gcc -o pgserver pgmain.o gcore/libgcore.a widget/libwidget.a appmgr/libappmgr.a net/libnet.a theme/libtheme.a debug/libdebug.a video/libvideo.a vidwrap/libvidwrap.a font/libfont.a vidbase/libvidbase.a input/libinput.a formats/libformats.a /usr/lib/libjpeg.a /usr/lib/libpng.a /opt/src/uclinux/uClinux-dist/lib/libsmalloc/libsmalloc.a \ /opt/src/uclinux/uClinux-dist/lib/libc/libc.a Which gives; pgmain.o: In function `main': pgmain.o(.text+0xaa): undefined reference to `__main' gcore/libgcore.a(div.o): In function `divnode_split': div.o(.text+0x34a): undefined reference to `__divsi3' div.o(.text+0x434): undefined reference to `__divsi3' gcore/libgcore.a(timer.o): In function `getticks': timer.o(.text+0xf8): undefined reference to `__divsi3' gcore/libgcore.a(timer.o): In function `inactivity_check': timer.o(.text+0x1d8): undefined reference to `__udivsi3' gcore/libgcore.a(hotspot.o): In function `hotspot_closest': hotspot.o(.text+0x27a): undefined reference to `__mulsi3' gcore/libgcore.a(render.o): In function `gropnode_scaletranslate': render.o(.text+0x852): undefined reference to `__mulsi3' render.o(.text+0x864): undefined reference to `__divsi3' gcore/libgcore.a(render.o): In function `gropnode_map': Any ideas ? Thanks Jody This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From madhusudhan at knowsys.net Mon Mar 25 07:15:09 2002 From: madhusudhan at knowsys.net (Madhusudhan) Date: Mon, 25 Mar 2002 17:45:09 +0530 Subject: [uClinux-dev] SWI Relocation Error Message-ID: <009201c1d3f6$bd6df4c0$5401a8c0@knowsys.net> Hi I am trying to compile uClibc for uClinux which is to be run on arm processor. I have my uclinux kernel source in usr/src/siap-uclinux/linux directoryonm my local machine. I also have 2.4 kernel in my host machine.I have mentioned the kernel source to be of usr/src/siap-uclinux/linux while compiling uclibc. When i try to do this,.. i am getting a error "Cannot represent SWI relocation in this object file" siap-uClinux-gcc -Wall -Os -falign-functions=1 -fstrict-aliasing -fno-builtin -nostdinc -pipe -nostdinc -I../../../../include -I/usr/local/lib/gcc-lib/arm-uclinux/2.96/include -I. -D_LIBC -DNDEBUG -D_LIBC_REENTRANT -DL___libc_pread syscalls.c -c -o __libc_pread.o{standard input}: Assembler messages: {standard input}:20: Error: Can not represent SWI relocation in this object file format (0) make[4]: *** [__libc_pread.o] Error 1 make[4]: Leaving directory `/usr/src/siap-uClinux-2.2.5/uClibc/libc/sysdeps/linux/common' make[3]: *** [_dir_common] Error 2 make[3]: Leaving directory `/usr/src/siap-uClinux-2.2.5/uClibc/libc/sysdeps/linux' make[2]: *** [_dir_linux] Error 2 make[2]: Leaving directory `/usr/src/siap-uClinux-2.2.5/uClibc/libc/sysdeps' make[1]: *** [_dir_sysdeps] Error 2 make[1]: Leaving directory `/usr/src/siap-uClinux-2.2.5/uClibc/libc' make: *** [_dir_libc] Error 2 I think this is because my uclibc is getting the wrong unistd.h file. But i have properly made a soft link in the kernel source linux/include/asm to arm related files. But the Rules.mak file which is used to compile the uclibc does "uname -m" which gives i686 which is of my host PC. I think that is why i am getting this error. Somebody please suggest me a way out of this, Regards Madhusudhan M Software Engineer Knowledge Sytems Pvt Ltd #470, East End Main Road, 9th Block Jayanagar, Bangalore 560069. Off 6545252/53 Res 6587680 -------------- next part -------------- An HTML attachment was scrubbed... URL: From eauth at softsys.co.at Mon Mar 25 07:36:04 2002 From: eauth at softsys.co.at (Erwin Authried) Date: Mon, 25 Mar 2002 13:36:04 +0100 Subject: [uClinux-dev] About make for midori ... In-Reply-To: Message-ID: The kernel must be configured before you do the midori build, the "kernel" package that automates this process is not yet ready. Otherwise the kernel header files cannot be found. Try "make config" in your kernel source directory first. -Erwin > -----Urspr?ngliche Nachricht----- > Von: owner-uclinux-dev at uclinux.org > [mailto:owner-uclinux-dev at uclinux.org]Im Auftrag von Stefan Jonsson > Gesendet: Montag, 25. M?rz 2002 11:19 > An: uClinux-dev mailing list > Betreff: [uClinux-dev] About make for midori ... > > > Hello list, > > ... and hello Erwin :-) > I have an EB40+AT91MEC extra memory card. I was recommended, on this list, > to try midori. > I have downloaded midori now and made 'webconfig' and tried to > 'make'. I get > some errors though. > > I have put midori in: /hdc/midori/ > and the kernel (2.4.6) in: /usr/src/linux/ > > I have attached the "make messages" with errors in a file. So if anyone > wants to have a look at that and help me out, I would be most happy. > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From gmenie at akamai.com Mon Mar 25 08:04:46 2002 From: gmenie at akamai.com (Menie, Georges) Date: Mon, 25 Mar 2002 08:04:46 -0500 Subject: [uClinux-dev] 68*328 timer was: "Calibrating delay loop..." c rashes Message-ID: thanks for the try, it seems that 128Hz give enough accuracy, I did try it also, and I had to add a mach_gettimeoffset() function in order to have an acceptable accuracy. I'm going to send a new patch for the DragenEngine board, which includes the timer changes and other de2 specific changes. If there is some interest, I can generate a timer patch for the 68328 arch. Cheers, Georges > -----Original Message----- > From: D. Peter Siddons [mailto:siddons at bnl.gov] > Sent: jeudi 21 mars 2002 23:13 > To: uclinux-dev at uclinux.org > Subject: Re: [uClinux-dev] 68*328 timer was: "Calibrating > delay loop..." > crashes > > > So I set the clock at 12:00.00 today, and just checked it > at 17:00.00, using my wristwatch as a calibration (!) and > it seems within 1 second, which is probably my twitching > finger's problem. This is using uCdimm, and the 128Hz > sample timer from the on-chip RTC as the tick. > Pete. > > > "Menie, Georges" wrote: > > > > Anyone tried to use a 64 Hz interrupt and change the HZ > > constant to 64 ? What would be the drawback having the timer > > running lower than 100 Hz ? it will give more CPU to the > > running apps, the scheduler will run every 16 ms instead > > of 10 ms, is this a problem ?. I did a quick test, everything > > seems fine, and the clock should be even more accurate. > > > > Regards, > > Georges > > This message resent by the uclinux-dev at uclinux.org list > server http://www.uClinux.org/ > > -- > ----------------- > D. Peter Siddons > National Synchrotron Light Source, Bldg. 725D > Brookhaven National Laboratory > Upton > New York 11973 > Email: siddons at bnl.gov > This message resent by the uclinux-dev at uclinux.org list > server http://www.uClinux.org/ > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From daniel.haensse at fhk.usz.ch Mon Mar 25 08:28:22 2002 From: daniel.haensse at fhk.usz.ch (Daniel Haensse) Date: Mon, 25 Mar 2002 14:28:22 +0100 Subject: [uClinux-dev] 68*328 timer was: "Calibrating delay loop..." c rashes In-Reply-To: References: Message-ID: <200203251332.OAA07602@ds-00122.usz.ch> Am Montag, 25. M?rz 2002 14.04 schrieben Sie: > thanks for the try, it seems that 128Hz give enough accuracy, > I did try it also, and I had to add a mach_gettimeoffset() > function in order to have an acceptable accuracy. > > I'm going to send a new patch for the DragenEngine board, > which includes the timer changes and other de2 specific changes. > If there is some interest, I can generate a timer patch > for the 68328 arch. Yes, please, we're interested. > > Cheers, > Georges > > > -----Original Message----- > > From: D. Peter Siddons [mailto:siddons at bnl.gov] > > Sent: jeudi 21 mars 2002 23:13 > > To: uclinux-dev at uclinux.org > > Subject: Re: [uClinux-dev] 68*328 timer was: "Calibrating > > delay loop..." > > crashes > > > > > > So I set the clock at 12:00.00 today, and just checked it > > at 17:00.00, using my wristwatch as a calibration (!) and > > it seems within 1 second, which is probably my twitching > > finger's problem. This is using uCdimm, and the 128Hz > > sample timer from the on-chip RTC as the tick. > > Pete. > > > > "Menie, Georges" wrote: > > > Anyone tried to use a 64 Hz interrupt and change the HZ > > > constant to 64 ? What would be the drawback having the timer > > > running lower than 100 Hz ? it will give more CPU to the > > > running apps, the scheduler will run every 16 ms instead > > > of 10 ms, is this a problem ?. I did a quick test, everything > > > seems fine, and the clock should be even more accurate. > > > > > > Regards, > > > Georges > > > This message resent by the uclinux-dev at uclinux.org list > > > > server http://www.uClinux.org/ > > > > -- > > ----------------- > > D. Peter Siddons > > National Synchrotron Light Source, Bldg. 725D > > Brookhaven National Laboratory > > Upton > > New York 11973 > > Email: siddons at bnl.gov > > This message resent by the uclinux-dev at uclinux.org list > > server http://www.uClinux.org/ > > This message resent by the uclinux-dev at uclinux.org list server > http://www.uClinux.org/ This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From gmenie at akamai.com Mon Mar 25 08:34:42 2002 From: gmenie at akamai.com (Menie, Georges) Date: Mon, 25 Mar 2002 08:34:42 -0500 Subject: [uClinux-dev] new de2 support files Message-ID: Here is the latest change on the Exys/de2 board. This patch is for the CVS repository. The changes are: * better timer support * removed EEPROM read at startup (not needed as the IEEE hardware address is stored on flash) * rework the cs89x0.c changes so that there less #ifdef DRAGEN2 * less dependencies on other 68328 files, to allow the code to be changed without breaking other 68*328 arch * UART2 enabled Cheers, Georges -------------- next part -------------- An embedded and charset-unspecified text was scrubbed... Name: patch-de2.txt URL: From siddons at bnl.gov Mon Mar 25 09:12:12 2002 From: siddons at bnl.gov (D. Peter Siddons) Date: Mon, 25 Mar 2002 09:12:12 -0500 Subject: [uClinux-dev] 68*328 timer was: "Calibrating delay loop..." crashes References: Message-ID: <3C9F303C.69E7EFC7@bnl.gov> Hi Georges, I think that using the RTC sample timer is a superior way to using the timer1, because it is a truly periodic interrupt and doesn't require software to re-initialize it. I think the residual error I see is just a matter of crystal accuracy, and could certainly be tuned out. I adapted a patch from Bernhard Kuhn which switches the tick interrupt and modifies the service routine apropriately. I think it has disappeared from the web, so I attach his original patch, and the reference to his notes. I have been trying to find time to generate a patch to reflect my modifications and have the option included in the distribution, but haven't got there yet. All I did was generalize this one to include the VZ part and the uCdimm. Bernhard's notes are at http://www.uclinux.org/pub/uClinux/archive/2785.html and the patch he points to is attached. Cheers, Pete. "Menie, Georges" wrote: > > thanks for the try, it seems that 128Hz give enough accuracy, > I did try it also, and I had to add a mach_gettimeoffset() > function in order to have an acceptable accuracy. > > I'm going to send a new patch for the DragenEngine board, > which includes the timer changes and other de2 specific changes. > If there is some interest, I can generate a timer patch > for the 68328 arch. > > Cheers, > Georges > > -----Original Message----- > > From: D. Peter Siddons [mailto:siddons at bnl.gov] > > Sent: jeudi 21 mars 2002 23:13 > > To: uclinux-dev at uclinux.org > > Subject: Re: [uClinux-dev] 68*328 timer was: "Calibrating > > delay loop..." > > crashes > > > > > > So I set the clock at 12:00.00 today, and just checked it > > at 17:00.00, using my wristwatch as a calibration (!) and > > it seems within 1 second, which is probably my twitching > > finger's problem. This is using uCdimm, and the 128Hz > > sample timer from the on-chip RTC as the tick. > > Pete. > > > > > > "Menie, Georges" wrote: > > > > > > Anyone tried to use a 64 Hz interrupt and change the HZ > > > constant to 64 ? What would be the drawback having the timer > > > running lower than 100 Hz ? it will give more CPU to the > > > running apps, the scheduler will run every 16 ms instead > > > of 10 ms, is this a problem ?. I did a quick test, everything > > > seems fine, and the clock should be even more accurate. > > > > > > Regards, > > > Georges > > > This message resent by the uclinux-dev at uclinux.org list > > server http://www.uClinux.org/ > > > > -- > > ----------------- > > D. Peter Siddons > > National Synchrotron Light Source, Bldg. 725D > > Brookhaven National Laboratory > > Upton > > New York 11973 > > Email: siddons at bnl.gov > > This message resent by the uclinux-dev at uclinux.org list > > server http://www.uClinux.org/ > > > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ -- ----------------- D. Peter Siddons National Synchrotron Light Source, Bldg. 725D Brookhaven National Laboratory Upton New York 11973 Email: siddons at bnl.gov -------------- next part -------------- diff -urN uClinux-2.4.x-CVS-20010618/arch/m68knommu/config.in uClinux-2.4.x-CVS-20010618-uCsimm-20010619-rtc/arch/m68knommu/config.in --- uClinux-2.4.x-CVS-20010618/arch/m68knommu/config.in Mon Jun 18 16:52:04 2001 +++ uClinux-2.4.x-CVS-20010618-uCsimm-20010619-rtc/arch/m68knommu/config.in Wed Jun 20 01:43:06 2001 @@ -69,6 +69,7 @@ bool 'ALMA Electronics board support' CONFIG_ALMA_ANS bool 'uCsimm module support' CONFIG_UCSIMM bool 'PalmV support' CONFIG_PILOT + bool 'Use RTC as system timer' CONFIG_M68EZ328_USE_RTC fi if [ "$CONFIG_M68EN302" = "y" ]; then diff -urN uClinux-2.4.x-CVS-20010618/arch/m68knommu/defconfig uClinux-2.4.x-CVS-20010618-uCsimm-20010619-rtc/arch/m68knommu/defconfig --- uClinux-2.4.x-CVS-20010618/arch/m68knommu/defconfig Thu May 24 15:18:41 2001 +++ uClinux-2.4.x-CVS-20010618-uCsimm-20010619-rtc/arch/m68knommu/defconfig Wed Jun 20 01:56:51 2001 @@ -47,6 +47,7 @@ # CONFIG_RAM8MB is not set # CONFIG_RAM16MB is not set # CONFIG_RAM32MB is not set +# CONFIG_M68EZ328_USE_RTC is not set CONFIG_RAMKERNEL=y # CONFIG_ROMKERNEL is not set # CONFIG_HIMEMKERNEL is not set diff -urN uClinux-2.4.x-CVS-20010618/arch/m68knommu/platform/68EZ328/config.c uClinux-2.4.x-CVS-20010618-uCsimm-20010619-rtc/arch/m68knommu/platform/68EZ328/config.c --- uClinux-2.4.x-CVS-20010618/arch/m68knommu/platform/68EZ328/config.c Thu May 24 14:49:51 2001 +++ uClinux-2.4.x-CVS-20010618-uCsimm-20010619-rtc/arch/m68knommu/platform/68EZ328/config.c Wed Jun 20 01:46:43 2001 @@ -32,6 +32,15 @@ void BSP_sched_init(void (*timer_routine)(int, void *, struct pt_regs *)) { +#ifdef CONFIG_M68EZ328_USE_RTC + /* Enable RTC */ + RTCCTL |= RTCCTL_EN; + + /* Enable 128 Hz interrupt line */ + RTCIENR |= RTCIENR_SAM5; + + request_irq(SAM_IRQ_NUM, timer_routine, IRQ_FLG_LOCK, "timer", NULL); +#else /* Restart mode, Enable int, 32KHz, Enable timer */ TCTL = TCTL_OM | TCTL_IRQEN | TCTL_CLKSOURCE_32KHZ | TCTL_TEN; /* Set prescaler (Divide 32KHz by 32)*/ @@ -39,13 +48,18 @@ /* Set compare register 32Khz / 32 / 10 = 100 */ TCMP = 10; - request_irq(TMR_IRQ_NUM, timer_routine, IRQ_FLG_LOCK, "timer", NULL); + request_irq(TMR_IRQ_NUM, timer_routine, IRQ_FLG_LOCK, "timer", NULL); +#endif } void BSP_tick(void) { /* Reset Timer1 */ - TSTAT &= 0; +#ifdef CONFIG_M68EZ328_USE_RTC + RTCISR |= RTCISR_SAM5; +#else + TSTAT &= 0; +#endif } unsigned long BSP_gettimeoffset (void) diff -urN uClinux-2.4.x-CVS-20010618/include/asm-m68knommu/param.h uClinux-2.4.x-CVS-20010618-uCsimm-20010619-rtc/include/asm-m68knommu/param.h --- uClinux-2.4.x-CVS-20010618/include/asm-m68knommu/param.h Wed Jun 13 00:15:35 2001 +++ uClinux-2.4.x-CVS-20010618-uCsimm-20010619-rtc/include/asm-m68knommu/param.h Wed Jun 20 01:57:32 2001 @@ -14,10 +14,11 @@ #define HZ 100 #endif #ifdef CONFIG_M68EZ328 +#ifdef CONFIG_M68EZ328_USE_RTC +#define HZ 128 +#else #define HZ 100 #endif -#ifdef CONFIG_UCSIMM -#define HZ 100 #endif #ifdef CONFIG_SHGLCORE #define HZ 50 From mark_mcchrystal at yahoo.com Mon Mar 25 11:20:44 2002 From: mark_mcchrystal at yahoo.com (Mark McChrystal) Date: Mon, 25 Mar 2002 08:20:44 -0800 (PST) Subject: [uClinux-dev] Picogui... In-Reply-To: <3C9F099A.F16A1A90@cern.ch> Message-ID: <20020325162044.64300.qmail@web14306.mail.yahoo.com> try: LIBS="-lc" CLAGS="-m68000 -W1,-elf2flt" ./configure --host=m68k-elf --with-prefix Mark __________________________________________________ Do You Yahoo!? Yahoo! Movies - coverage of the 74th Academy Awards? http://movies.yahoo.com/ This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From jeff at arcturusnetworks.com Mon Mar 25 13:16:51 2002 From: jeff at arcturusnetworks.com (D. Jeff Dionne) Date: Mon, 25 Mar 2002 18:16:51 -0000 Subject: [uClinux-dev] Picogui... Message-ID: <200203251816.g2PIGpF19439@arcturusnetworks.com> Jody Pearson said: > Hi, .. You do need libc (second from) the end, and libgcc at the very end. It looks like a link order problem. You should be getting libc and libgcc (in the right order) from the specs file installed with the toolchain. I would look into that first. J > > make[2]: Entering directory > `/asis.local/src/uclinux/pgui-dev20020311/pgserver' > m68k-elf-gcc -m68000 -O2 -fomit-frame-pointer > -I/opt/src/uclinux/uClinux-dist/lib/libc/include > /opt/src/uclinux/uClinux-dist/lib/libc/libc.a -fomit-frame-pointer -o > pgserve > r pgmain.o gcore/libgcore.a widget/libwidget.a > appmgr/libappmgr.a net/libnet.a > theme/libtheme.a debug/ > libdebug.a video/libvideo.a > vidwrap/libvidwrap.a font/libfont.a > vidbase/libvidbase.a input/libinput.a > formats/libformats.a > pgmain.o: In function `run_config_process': > pgmain.o(.text+0xc): undefined reference to `getpid' > pgmain.o(.text+0x34): undefined reference to `vfork' > pgmain.o(.text+0x68): undefined reference to `execve' > pgmain.o(.text+0x80): undefined reference to `kill' > pgmain.o: In function `main': > pgmain.o(.text+0xca): undefined reference to `getenv' > pgmain.o(.text+0xe2): undefined reference to `strlen' > pgmain.o(.text+0x10e): undefined reference to `strcpy' > pgmain.o(.text+0x11e): undefined reference to `strcat' > pgmain.o(.text+0x13a): undefined reference to `getopt' > pgmain.o(.text+0x1fc): undefined reference to `optarg' > pgmain.o(.text+0x202): undefined reference to `strchr' > pgmain.o(.text+0x216): undefined reference to `optarg' > > Obviously all of the references exist in uclibc, so i tried using; > > > m68k-elf-gcc -o pgserver pgmain.o gcore/libgcore.a widget/libwidget.a > appmgr/libappmgr.a net/libnet.a theme/libtheme.a debug/libdebug.a > video/libvideo.a vidwrap/libvidwrap.a font/libfont.a > vidbase/libvidbase.a input/libinput.a formats/libformats.a > /usr/lib/libjpeg.a /usr/lib/libpng.a > /opt/src/uclinux/uClinux-dist/lib/libsmalloc/libsmalloc.a \ > /opt/src/uclinux/uClinux-dist/lib/libc/libc.a > > Which gives; > > pgmain.o: In function `main': > pgmain.o(.text+0xaa): undefined reference to `__main' > gcore/libgcore.a(div.o): In function `divnode_split': > div.o(.text+0x34a): undefined reference to `__divsi3' > div.o(.text+0x434): undefined reference to `__divsi3' > gcore/libgcore.a(timer.o): In function `getticks': > timer.o(.text+0xf8): undefined reference to `__divsi3' > gcore/libgcore.a(timer.o): In function `inactivity_check': > timer.o(.text+0x1d8): undefined reference to `__udivsi3' > gcore/libgcore.a(hotspot.o): In function `hotspot_closest': > hotspot.o(.text+0x27a): undefined reference to `__mulsi3' > gcore/libgcore.a(render.o): In function `gropnode_scaletranslate': > render.o(.text+0x852): undefined reference to `__mulsi3' > render.o(.text+0x864): undefined reference to `__divsi3' > gcore/libgcore.a(render.o): In function `gropnode_map': > > Any ideas ? > > Thanks > > Jody > > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ > -- D. Jeff Dionne Jeff at ArcturusNetworks.com -------------- next part -------------- A non-text attachment was scrubbed... Name: Type: application/octet-stream Size: 2 bytes Desc: not available URL: From gerg at snapgear.com Mon Mar 25 20:33:23 2002 From: gerg at snapgear.com (Greg Ungerer) Date: Tue, 26 Mar 2002 11:33:23 +1000 Subject: [uClinux-dev] Re: new de2 support files References: Message-ID: <3C9FCFE3.5070002@snapgear.com> Hi Menie, Applied. Thanks Greg Menie, Georges wrote: > Here is the latest change on the Exys/de2 board. > This patch is for the CVS repository. > > The changes are: > * better timer support > * removed EEPROM read at startup (not needed as the IEEE > hardware address is stored on flash) > * rework the cs89x0.c changes so that there less #ifdef DRAGEN2 > * less dependencies on other 68328 files, to allow the code > to be changed without breaking other 68*328 arch > * UART2 enabled > > Cheers, > Georges > ------------------------------------------------------------------------ Greg Ungerer -- Chief Software Wizard EMAIL: gerg at snapgear.com SnapGear PHONE: +61 7 3435 2888 825 Stanley St, FAX: +61 7 3891 3630 Woolloongabba, QLD, 4102, Australia WEB: www.snapgear.com This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From gerg at snapgear.com Tue Mar 26 00:10:22 2002 From: gerg at snapgear.com (Greg Ungerer) Date: Tue, 26 Mar 2002 15:10:22 +1000 Subject: [uClinux-dev] 68*328 timer was: "Calibrating delay loop..." crashes References: <3C9F303C.69E7EFC7@bnl.gov> Message-ID: <3CA002BE.2020901@snapgear.com> Hi Peter, This looks usefull, and the patch is pretty clean. So I have applied it to the 2.4.x CVS. Regards Greg D. Peter Siddons wrote: > Hi Georges, > I think that using the RTC sample timer is a superior way > to using the timer1, because it is a truly periodic interrupt > and doesn't require software to re-initialize it. I think the > residual error I see is just a matter of crystal accuracy, and > could certainly be tuned out. > I adapted a patch from Bernhard Kuhn which switches the tick > interrupt and modifies the service routine apropriately. I > think it has disappeared from the web, so I attach his original > patch, and the reference to his notes. I have been trying to > find time to generate a patch to reflect my modifications and > have the option included in the distribution, but haven't got > there yet. All I did was generalize this one to include the VZ > part and the uCdimm. Bernhard's notes are at > http://www.uclinux.org/pub/uClinux/archive/2785.html > and the patch he points to is attached. > Cheers, > Pete. > > > "Menie, Georges" wrote: > >>thanks for the try, it seems that 128Hz give enough accuracy, >>I did try it also, and I had to add a mach_gettimeoffset() >>function in order to have an acceptable accuracy. >> >>I'm going to send a new patch for the DragenEngine board, >>which includes the timer changes and other de2 specific changes. >>If there is some interest, I can generate a timer patch >>for the 68328 arch. >> >>Cheers, >>Georges >> >>>-----Original Message----- >>>From: D. Peter Siddons [mailto:siddons at bnl.gov] >>>Sent: jeudi 21 mars 2002 23:13 >>>To: uclinux-dev at uclinux.org >>>Subject: Re: [uClinux-dev] 68*328 timer was: "Calibrating >>>delay loop..." >>>crashes >>> >>> >>>So I set the clock at 12:00.00 today, and just checked it >>>at 17:00.00, using my wristwatch as a calibration (!) and >>>it seems within 1 second, which is probably my twitching >>>finger's problem. This is using uCdimm, and the 128Hz >>>sample timer from the on-chip RTC as the tick. >>>Pete. >>> >>> >>>"Menie, Georges" wrote: >>> >>>>Anyone tried to use a 64 Hz interrupt and change the HZ >>>>constant to 64 ? What would be the drawback having the timer >>>>running lower than 100 Hz ? it will give more CPU to the >>>>running apps, the scheduler will run every 16 ms instead >>>>of 10 ms, is this a problem ?. I did a quick test, everything >>>>seems fine, and the clock should be even more accurate. >>>> >>>>Regards, >>>>Georges >>>>This message resent by the uclinux-dev at uclinux.org list >>>> >>>server http://www.uClinux.org/ >>> >>>-- >>>----------------- >>>D. Peter Siddons >>>National Synchrotron Light Source, Bldg. 725D >>>Brookhaven National Laboratory >>>Upton >>>New York 11973 >>>Email: siddons at bnl.gov >>>This message resent by the uclinux-dev at uclinux.org list >>>server http://www.uClinux.org/ >>> >>> >>This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ >> > > > ------------------------------------------------------------------------ > > diff -urN uClinux-2.4.x-CVS-20010618/arch/m68knommu/config.in uClinux-2.4.x-CVS-20010618-uCsimm-20010619-rtc/arch/m68knommu/config.in > --- uClinux-2.4.x-CVS-20010618/arch/m68knommu/config.in Mon Jun 18 16:52:04 2001 > +++ uClinux-2.4.x-CVS-20010618-uCsimm-20010619-rtc/arch/m68knommu/config.in Wed Jun 20 01:43:06 2001 > @@ -69,6 +69,7 @@ > bool 'ALMA Electronics board support' CONFIG_ALMA_ANS > bool 'uCsimm module support' CONFIG_UCSIMM > bool 'PalmV support' CONFIG_PILOT > + bool 'Use RTC as system timer' CONFIG_M68EZ328_USE_RTC > fi > > if [ "$CONFIG_M68EN302" = "y" ]; then > diff -urN uClinux-2.4.x-CVS-20010618/arch/m68knommu/defconfig uClinux-2.4.x-CVS-20010618-uCsimm-20010619-rtc/arch/m68knommu/defconfig > --- uClinux-2.4.x-CVS-20010618/arch/m68knommu/defconfig Thu May 24 15:18:41 2001 > +++ uClinux-2.4.x-CVS-20010618-uCsimm-20010619-rtc/arch/m68knommu/defconfig Wed Jun 20 01:56:51 2001 > @@ -47,6 +47,7 @@ > # CONFIG_RAM8MB is not set > # CONFIG_RAM16MB is not set > # CONFIG_RAM32MB is not set > +# CONFIG_M68EZ328_USE_RTC is not set > CONFIG_RAMKERNEL=y > # CONFIG_ROMKERNEL is not set > # CONFIG_HIMEMKERNEL is not set > diff -urN uClinux-2.4.x-CVS-20010618/arch/m68knommu/platform/68EZ328/config.c uClinux-2.4.x-CVS-20010618-uCsimm-20010619-rtc/arch/m68knommu/platform/68EZ328/config.c > --- uClinux-2.4.x-CVS-20010618/arch/m68knommu/platform/68EZ328/config.c Thu May 24 14:49:51 2001 > +++ uClinux-2.4.x-CVS-20010618-uCsimm-20010619-rtc/arch/m68knommu/platform/68EZ328/config.c Wed Jun 20 01:46:43 2001 > @@ -32,6 +32,15 @@ > > void BSP_sched_init(void (*timer_routine)(int, void *, struct pt_regs *)) > { > +#ifdef CONFIG_M68EZ328_USE_RTC > + /* Enable RTC */ > + RTCCTL |= RTCCTL_EN; > + > + /* Enable 128 Hz interrupt line */ > + RTCIENR |= RTCIENR_SAM5; > + > + request_irq(SAM_IRQ_NUM, timer_routine, IRQ_FLG_LOCK, "timer", NULL); > +#else > /* Restart mode, Enable int, 32KHz, Enable timer */ > TCTL = TCTL_OM | TCTL_IRQEN | TCTL_CLKSOURCE_32KHZ | TCTL_TEN; > /* Set prescaler (Divide 32KHz by 32)*/ > @@ -39,13 +48,18 @@ > /* Set compare register 32Khz / 32 / 10 = 100 */ > TCMP = 10; > > - request_irq(TMR_IRQ_NUM, timer_routine, IRQ_FLG_LOCK, "timer", NULL); > + request_irq(TMR_IRQ_NUM, timer_routine, IRQ_FLG_LOCK, "timer", NULL); > +#endif > } > > void BSP_tick(void) > { > /* Reset Timer1 */ > - TSTAT &= 0; > +#ifdef CONFIG_M68EZ328_USE_RTC > + RTCISR |= RTCISR_SAM5; > +#else > + TSTAT &= 0; > +#endif > } > > unsigned long BSP_gettimeoffset (void) > diff -urN uClinux-2.4.x-CVS-20010618/include/asm-m68knommu/param.h uClinux-2.4.x-CVS-20010618-uCsimm-20010619-rtc/include/asm-m68knommu/param.h > --- uClinux-2.4.x-CVS-20010618/include/asm-m68knommu/param.h Wed Jun 13 00:15:35 2001 > +++ uClinux-2.4.x-CVS-20010618-uCsimm-20010619-rtc/include/asm-m68knommu/param.h Wed Jun 20 01:57:32 2001 > @@ -14,10 +14,11 @@ > #define HZ 100 > #endif > #ifdef CONFIG_M68EZ328 > +#ifdef CONFIG_M68EZ328_USE_RTC > +#define HZ 128 > +#else > #define HZ 100 > #endif > -#ifdef CONFIG_UCSIMM > -#define HZ 100 > #endif > #ifdef CONFIG_SHGLCORE > #define HZ 50 > > patch-uClinux-2.4.x-CVS-20010618-uCsimm-20010619-rtc > > Content-Type: > > text/plain > Content-Encoding: > > 7bit > > -- ------------------------------------------------------------------------ Greg Ungerer -- Chief Software Wizard EMAIL: gerg at snapgear.com SnapGear PHONE: +61 7 3435 2888 825 Stanley St, FAX: +61 7 3891 3630 Woolloongabba, QLD, 4102, Australia WEB: www.snapgear.com This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From miles at lsi.nec.co.jp Tue Mar 26 00:41:11 2002 From: miles at lsi.nec.co.jp (Miles Bader) Date: 26 Mar 2002 14:41:11 +0900 Subject: [uClinux-dev] v850 update patch 20020326 Message-ID: Please apply: -------------- next part -------------- A non-text attachment was scrubbed... Name: uclinux-2.4-v850-20020326-dist.patch Type: text/x-patch Size: 15355 bytes Desc: v850 update patch 20020326 URL: -------------- next part -------------- Thanks, -Miles -- Fast, small, soon; pick any 2. From miles at lsi.nec.co.jp Tue Mar 26 01:05:33 2002 From: miles at lsi.nec.co.jp (Miles Bader) Date: 26 Mar 2002 15:05:33 +0900 Subject: [uClinux-dev] patch to add v850 support to busybox insmod Message-ID: Hi, This patch adds support for the v850e processor to busybox's insmod program. The patch is against busybox 0.60.0 from uClinux-dist-20020214. Could you apply it? Thanks, -Miles Patch: -------------- next part -------------- A non-text attachment was scrubbed... Name: busybox-insmod-v850.patch Type: text/x-patch Size: 8368 bytes Desc: patch to add v850 support to busybox insmod URL: -------------- next part -------------- -- I'm beginning to think that life is just one long Yoko Ono album; no rhyme or reason, just a lot of incoherent shrieks and then it's over. --Ian Wolff From gerg at snapgear.com Tue Mar 26 01:49:20 2002 From: gerg at snapgear.com (Greg Ungerer) Date: Tue, 26 Mar 2002 16:49:20 +1000 Subject: [uClinux-dev] v850 update patch 20020326 References: Message-ID: <3CA019F0.2050402@snapgear.com> Hi Miles, Applied. Regards Greg Miles Bader wrote: > Please apply: > > > > ------------------------------------------------------------------------ [snip] > ------------------------------------------------------------------------ > > > > Thanks, > > -Miles > > uclinux-2.4-v850-20020326-dist.patch > > Content-Description: > > v850 update patch 20020326 > Content-Type: > > text/x-patch > > > ------------------------------------------------------------------------ > Part 1.3 > > Content-Type: > > text/plain > > -- ------------------------------------------------------------------------ Greg Ungerer -- Chief Software Wizard EMAIL: gerg at snapgear.com SnapGear PHONE: +61 7 3435 2888 825 Stanley St, FAX: +61 7 3891 3630 Woolloongabba, QLD, 4102, Australia WEB: www.snapgear.com This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From gmenie at akamai.com Tue Mar 26 03:40:12 2002 From: gmenie at akamai.com (Menie, Georges) Date: Tue, 26 Mar 2002 03:40:12 -0500 Subject: [uClinux-dev] 68*328 timer was: "Calibrating delay loop..." c rashes Message-ID: You are right, the RTC needs less CPU power, however it does not give the current ticks count which is used by gettimeoffset() to adjust the time. Anyway, I'm going to merge the patch so the choice will be in the user hand. Cheers, Georges > -----Original Message----- > From: D. Peter Siddons [mailto:siddons at bnl.gov] > Sent: lundi 25 mars 2002 15:12 > To: uclinux-dev at uclinux.org > Subject: Re: [uClinux-dev] 68*328 timer was: "Calibrating > delay loop..." > crashes > > > Hi Georges, > I think that using the RTC sample timer is a superior way > to using the timer1, because it is a truly periodic interrupt > and doesn't require software to re-initialize it. I think the > residual error I see is just a matter of crystal accuracy, and > could certainly be tuned out. > I adapted a patch from Bernhard Kuhn which switches the tick > interrupt and modifies the service routine apropriately. I > think it has disappeared from the web, so I attach his original > patch, and the reference to his notes. I have been trying to > find time to generate a patch to reflect my modifications and > have the option included in the distribution, but haven't got > there yet. All I did was generalize this one to include the VZ > part and the uCdimm. Bernhard's notes are at > http://www.uclinux.org/pub/uClinux/archive/2785.html > and the patch he points to is attached. > Cheers, > Pete. > > > "Menie, Georges" wrote: > > > > thanks for the try, it seems that 128Hz give enough accuracy, > > I did try it also, and I had to add a mach_gettimeoffset() > > function in order to have an acceptable accuracy. > > > > I'm going to send a new patch for the DragenEngine board, > > which includes the timer changes and other de2 specific changes. > > If there is some interest, I can generate a timer patch > > for the 68328 arch. > > > > Cheers, > > Georges > > > -----Original Message----- > > > From: D. Peter Siddons [mailto:siddons at bnl.gov] > > > Sent: jeudi 21 mars 2002 23:13 > > > To: uclinux-dev at uclinux.org > > > Subject: Re: [uClinux-dev] 68*328 timer was: "Calibrating > > > delay loop..." > > > crashes > > > > > > > > > So I set the clock at 12:00.00 today, and just checked it > > > at 17:00.00, using my wristwatch as a calibration (!) and > > > it seems within 1 second, which is probably my twitching > > > finger's problem. This is using uCdimm, and the 128Hz > > > sample timer from the on-chip RTC as the tick. > > > Pete. > > > > > > > > > "Menie, Georges" wrote: > > > > > > > > Anyone tried to use a 64 Hz interrupt and change the HZ > > > > constant to 64 ? What would be the drawback having the timer > > > > running lower than 100 Hz ? it will give more CPU to the > > > > running apps, the scheduler will run every 16 ms instead > > > > of 10 ms, is this a problem ?. I did a quick test, everything > > > > seems fine, and the clock should be even more accurate. > > > > > > > > Regards, > > > > Georges > > > > This message resent by the uclinux-dev at uclinux.org list > > > server http://www.uClinux.org/ > > > > > > -- > > > ----------------- > > > D. Peter Siddons > > > National Synchrotron Light Source, Bldg. 725D > > > Brookhaven National Laboratory > > > Upton > > > New York 11973 > > > Email: siddons at bnl.gov > > > This message resent by the uclinux-dev at uclinux.org list > > > server http://www.uClinux.org/ > > > > > This message resent by the uclinux-dev at uclinux.org list > server http://www.uClinux.org/ > > -- > ----------------- > D. Peter Siddons > National Synchrotron Light Source, Bldg. 725D > Brookhaven National Laboratory > Upton > New York 11973 > Email: siddons at bnl.gov > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From gmenie at akamai.com Tue Mar 26 03:40:12 2002 From: gmenie at akamai.com (Menie, Georges) Date: Tue, 26 Mar 2002 03:40:12 -0500 Subject: [uClinux-dev] 68*328 timer was: "Calibrating delay loop..." c rashes Message-ID: Attached is the patch for the ucdimm; could you try it, then if everything is ok, I will generate a patch for the whole 68*328 arch, merging the RTC patch of Bernard Kuhn so everyone will have the choice to use timer 1 or the rtc to generate ticks. After applying this patch, you will have to edit your linux-2.4.x/include/m68knommu/param.h to change the HZ constant to 128 to get a better accuracy. Cheers, Georges > -----Original Message----- > From: Daniel Haensse [mailto:daniel.haensse at fhk.usz.ch] > Sent: lundi 25 mars 2002 14:28 > To: uclinux-dev at uclinux.org > Subject: Re: [uClinux-dev] 68*328 timer was: "Calibrating > delay loop..." > c rashes > > > Am Montag, 25. M?rz 2002 14.04 schrieben Sie: > > thanks for the try, it seems that 128Hz give enough accuracy, > > I did try it also, and I had to add a mach_gettimeoffset() > > function in order to have an acceptable accuracy. > > > > I'm going to send a new patch for the DragenEngine board, > > which includes the timer changes and other de2 specific changes. > > If there is some interest, I can generate a timer patch > > for the 68328 arch. > Yes, please, we're interested. > > > > Cheers, > > Georges -------------- next part -------------- An embedded and charset-unspecified text was scrubbed... Name: patch-timer.txt URL: From miles at lsi.nec.co.jp Tue Mar 26 03:47:12 2002 From: miles at lsi.nec.co.jp (Miles Bader) Date: 26 Mar 2002 17:47:12 +0900 Subject: [uClinux-dev] Re: v850 update patch 20020326 In-Reply-To: <3CA019F0.2050402@snapgear.com> References: <3CA019F0.2050402@snapgear.com> Message-ID: Thanks I forgot one file, though; could you apply this patch too? Patch: Index: include/asm-v850/dma.h =================================================================== RCS file: /var/cvs/uClinux-2.4.x/include/asm-v850/dma.h,v retrieving revision 1.1 diff -u -r1.1 dma.h --- include/asm-v850/dma.h 2001/10/22 13:39:02 1.1 +++ include/asm-v850/dma.h 2002/03/26 08:09:31 @@ -9,6 +9,10 @@ /* release it again */ extern void free_dma (unsigned int dmanr); +#ifdef CONFIG_PCI +extern int isa_dma_bridge_buggy; +#else #define isa_dma_bridge_buggy (0) +#endif #endif /* __V850_DMA_H__ */ -Miles -- The secret to creativity is knowing how to hide your sources. --Albert Einstein This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From sheeliang at lit.org.sg Tue Mar 26 04:10:11 2002 From: sheeliang at lit.org.sg (Chia Shee Liang) Date: Tue, 26 Mar 2002 17:10:11 +0800 Subject: [uClinux-dev] General qn on task_struct Message-ID: <20020326171011.A5750@localhost.localdomain> This is a general newbie question on struct task_struct. I read somewhere that it is not a good idea to change the members of this struct as the offsets are often hardcoded in assembly language routines. So is it safe to add members towards the end of the struct, thereby increasing the size, but not touching anything else? This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From jody.pearson at cern.ch Tue Mar 26 11:36:07 2002 From: jody.pearson at cern.ch (Jody Pearson) Date: Tue, 26 Mar 2002 17:36:07 +0100 Subject: [uClinux-dev] Picogui - compiled and linked.... Message-ID: <3CA0A377.CC036AF5@cern.ch> Hi All, Well I have made serious changes to the normal linking process. The situation is; 1) Get the snapshot. 2) Run the autogen script 3) Hack the configure script to include; CFLAGS="-m68000 -g -O2 -fomit-frame-pointer /opt/src/uclinux/uClinux-dist/lib/libc/libc.a" (I know - it's a crap hack) 4) Run ./configure --host=m68k-elf --with-perfix=m68k-elf 5) type make config, 6) type make 7) Linking still fails, so i hacked the following; #!/bin/sh CMDLIBS="\ /usr/local/m68k-elf/lib/m68000/libc.a /usr/local/lib/gcc-lib/m68k-elf/2.95.3/m68000/libgcc.a \ /opt/src/uclinux/uClinux-dist/lib/libc/libc.a \ /opt/src/uclinux/uClinux-dist/lib/libm/libm.a \ " m68k-elf-gcc -m68000 -O2 -fomit-frame-pointer -Wl,-elf2flt -o pgserver pgmain.o gcore/libgcore.a widget/libwidget.a appmgr/libappmgr.a net/libnet.a theme/libtheme.a debug/libdebug.a video/libvideo.a vidwrap/libvidwrap.a font/libfont.a vidbase/libvidbase.a input/libinput.a formats/libformats.a $CMDLIBS Again, it's crap, but it works. SO i get a binary.. -rwxr--r-- 1 root 5764 299808 Mar 26 17:24 pgserver # file pgserver* pgserver: data pgserver.gdb: ELF 32-bit MSB executable, Motorola 68000, version 1, statically linked, not stripped # Then i copy that into me romfs, and remake the image. SO when I try to run the bin, i get; Sash command shell (version 1.1.1) /> pgserver MAP_SHARED not supported (cannot write mappings to disk) /> cat /etc/pgserver.conf # Sample config file [pgserver] xres = 160 yres = 160 fullscreen [video-fbdev] caption = Welcome to PicoGUI! [video-fbdev] vt = auto /> Any suggestions please ? Thanks Jody This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From stefan.keller-tuberg at alcatel.com Tue Mar 26 12:08:45 2002 From: stefan.keller-tuberg at alcatel.com (Stefan Keller-Tuberg) Date: Tue, 26 Mar 2002 12:08:45 -0500 Subject: [uClinux-dev] uClibc for uCsimm Message-ID: <3CA0AB1D.5F83B20B@alcatel.com> I've been trying to build uClibc for a uCsimm. The previous postings to this group don't clarify how to actually do it and the procedure I've followed ends in grief. It would be wonderful if somebody could clarify this. Step by step, this is what I've done. Under "make menuconfig", I have selected the following target platform options: Lineo/uCsimm, linux-2.4.x, uClibc, default all settings. The kernel builds OK and then the library compile bombs out with the following error... >make[2]: Entering directory `/opt/uClinux-dist/uClibc' >Rules.mak:25: Config: No such file or directory >You didn't read the README, did you... =) >Choose a configuration file in extras/Config/ and then run > ln -s ./extra/Configs/Config. ./Config The solution to this seems obvious. I "cp ./vendors/Lineo/uCdimm/config.uClibc ./vendors/Lineo/uCsimm" and make again with the same settings as above. The library compile bombs out a little further along... >m68k-elf-gcc -Wall -fno-builtin -nostdinc -m68000 -nostdinc > -I../../../../include > -I/usr/local/lib/gcc-lib/m68k-elf/2.95.3/include -I. > -D_LIBC -m68000 -O2 -g -fomit-frame-pointer -DCONFIG_LINEO > -fno-builtin -msep-data -DEMBED > -I/opt/uClinux-dist/lib/uClibc/include -I/opt/uClinux-dist > -Dlinux -D__linux__ -D__uClinux__ -Dunix > -I/opt/uClinux-dist/linux-2.4.x/include -DNDEBUG > -DL_ioperm syscalls.c -c -o ioperm.o >syscalls.c:767: sys/io.h: No such file or directory >make[6]: *** [ioperm.o] Error 1 A recent posting suggests a solution. Rather than copy individual missing files, I untared an older uClibc-snapshot into "./uClibc" and then overcopied with the uClibc-0.9.10 tree. The make goes further still and stops with... >m68k-elf-gcc -Wall -fno-builtin -c crti.S -o crti.o >cp crti.o ../../../../lib/ >cp: cannot create regular file > `../../../../lib/crti.o': No such file or directory >make[6]: *** [crti.o] Error 1 Now I'm lost. I made a wild guess which is probably wrong and created a directory called ./uClibc/lib and recompiled... >make[6]: Entering directory > `/opt/uClinux-dist/uClibc/libc/sysdeps/linux/m68k' >m68k-elf-gcc -Wall -fno-builtin -nostdinc -m68000 > -nostdinc -I../../../../include > -I/usr/local/lib/gcc-lib/m68k-elf/2.95.3/include -I. > -D_LIBC -m68000 -O2 -g -fomit-frame-pointer -DCONFIG_LINEO > -fno-builtin -msep-data -DEMBED > -I/opt/uClinux-dist/lib/uClibc/include -I/opt/uClinux-dist > -Dlinux -D__linux__ -D__uClinux__ -Dunix > -I/opt/uClinux-dist/linux-2.4.x/include -DNDEBUG -c __longjmp.S > -o __longjmp.o >../../../../include/bits/setjmp.h: Assembler messages: >../../../../include/bits/setjmp.h:25: Error: > Unknown operator -- statement `typedef struct' ignored etc My host system is RH7.2 and I've installed m68k-elf-tools-20020218.tar.gz under "/" and uClinux-dist-20020306.tar.gz under "/opt". Any suggestions? Thanks. Stefan This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From JLiu at PacesetterElectronics.com Tue Mar 26 12:17:10 2002 From: JLiu at PacesetterElectronics.com (JINGSONG LIU) Date: Tue, 26 Mar 2002 11:17:10 -0600 Subject: [uClinux-dev] General qn on task_struct Message-ID: All : After mount -t nfs 10.0.0.18:/home/jsl/uClinus-dist/user/app /tmp then cd /tmp and issue exec app which is very simple testing code ( hello world !) and I got hello world but quit telnet with message " Connection closed by foreign host ". also it seems that if I issue exec either in /tmp or in root " Connection closed by foreign host " will shows up then quit telnet. If I issue app because app in /bin I got hello world ! without any problem. Does any one met the same problem I had ? Thank you in advance Jason Liu This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From justin.wojdacki at analog.com Tue Mar 26 13:12:39 2002 From: justin.wojdacki at analog.com (Justin Wojdacki) Date: Tue, 26 Mar 2002 10:12:39 -0800 Subject: [uClinux-dev] General qn on task_struct References: <20020326171011.A5750@localhost.localdomain> Message-ID: <3CA0BA17.82DAA03B@analog.com> Chia Shee Liang wrote: > > This is a general newbie question on struct task_struct. I read > somewhere that it is not a good idea to change the members of this struct > as the offsets are often hardcoded in assembly language routines. > This is true. On some platforms, such as MIPS, the offsets are converted into numbers that can be used in assembly via an offset.h file (see arch/mips/tools/). I don't see anything like this for the M68K port, but ARM definitely has it too (arch/arm/tools/), as does PA-RISC. Altering this could cause problems. > > So is it safe to add members towards the end of the struct, thereby > increasing the size, but not touching anything else? > It's safe if nobody has hardcoded sizeof(struct task_struct) anywhere in their code. :) Are you sure you need to modify struct task_struct? I ask only because this could become a maintenance nightmare for you. -- ------------------------------------------------- Justin Wojdacki justin.wojdacki at analog.com (408) 350-5032 Communications Processors Group -- Analog Devices This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From pgccoop at hotmail.com Tue Mar 26 13:58:46 2002 From: pgccoop at hotmail.com (Jonathan Schmidt) Date: Tue, 26 Mar 2002 18:58:46 +0000 Subject: [uClinux-dev] More on "delay loops" and timers! Message-ID: I am still having problems getting the CVS version of the uClinux kernel to run on my board. It still crashes at "Calibrating delay loop...". I have looked everywhere for a reason the timer interrupt wouldn't be happening, and I have checked the .ld file. I have started to suspect the last few commands that my Makefile runs to get the final kernel image. Here are the commands: $(CROSS_COMPILE)objcopy -O binary --remove-section=.romvec \ --remove-section=.text --remove-section=.ramvec \ --remove-section=.bss --remove-section=.eram \ $(ROOTDIR)/$(LINUXDIR)/linux $(IMAGEDIR)/linux.data $(CROSS_COMPILE)objcopy -O binary --remove-section=.ramvec \ --remove-section=.bss --remove-section=.data \ --remove-section=.eram \ --set-section-flags=.romvec=CONTENTS,ALLOC,LOAD,READONLY,CODE \ $(ROOTDIR)/$(LINUXDIR)/linux $(IMAGEDIR)/linux.text cat $(IMAGEDIR)/linux.text $(IMAGEDIR)/linux.data $(ROMFSIMG) > $(IMAGE) Do these seem correct? Perhaps my .ld file is missing some globals? I have attached it as well. I have 4MB of flash and 16MB of SDRAM. Please let me know if you have any ideas! (is there any other files of interest I can attach?) Thanks, Jonathan Schmidt _________________________________________________________________ Send and receive Hotmail on your mobile device: http://mobile.msn.com -------------- next part -------------- A non-text attachment was scrubbed... Name: rom.ld Type: application/octet-stream Size: 1391 bytes Desc: not available URL: From damnyosh at yahoo.com Tue Mar 26 14:12:06 2002 From: damnyosh at yahoo.com (Yosh K) Date: Tue, 26 Mar 2002 11:12:06 -0800 (PST) Subject: [uClinux-dev] tftp problem Message-ID: <20020326191206.17122.qmail@web20607.mail.yahoo.com> Hello, I am trying to upload the image to MCF5272 eval board using tftp. I've tested the tftp server and works fine. Here is how the board is setup base: 16 baud: 19200 server: 192.168.99.13 client: 192.168.99.14 gateway: 0.0.0.0 netmask: 255.255.0.0 filename: image.bin filetype: ELF mac: 00:CF:52:72:C3:01 When I do dn -i image.bin i get the error "unable to locate 192.168.99.13" I know the tftp server is running on 192.168.99.13. So is there something I missed to setup properly. Thanks Yosh __________________________________________________ Do You Yahoo!? Yahoo! Movies - coverage of the 74th Academy Awards? http://movies.yahoo.com/ This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From Subashk at ami.com Tue Mar 26 15:13:49 2002 From: Subashk at ami.com (Subash Kalbarga) Date: Tue, 26 Mar 2002 15:13:49 -0500 Subject: [uClinux-dev] spawning processes without having a file system Message-ID: <8CCBDD5583C50E4196F012E79439B45C3644FE@atl-ms1.megatrends.com> Hi all I am on a M5372 and am running into space (on ROM) problems because I have a lot of programs who all link to the same libraries. I was wondering if there is any easy way to make all the programs into one big image (not unlike the kernel) where all functions are available to everyone else? If I do this the kernel should be able to start processes that are defined by function entry points rather than as files in the file system For eg: I would just have a function alert_process() that is the entry point into the process and maybe a process table where the kernel looks up and starts these. Is this a feasible thought? Any suggestions on how I should approach it? Thanks in advance Subash -------------- next part -------------- An HTML attachment was scrubbed... URL: From Jim.Henderson at endevco.com Tue Mar 26 15:47:17 2002 From: Jim.Henderson at endevco.com (Henderson, Jim) Date: Tue, 26 Mar 2002 12:47:17 -0800 Subject: [uClinux-dev] M5272C3 gdb bdm Message-ID: <412A28C967692A438B59223143A7268FE1B146@ALTAIR.endevco.com> I have an M5272C3 eval board. When I start my GDB, I get this: # gdb-5.0/gdb/gdb GNU gdb 5.0 Copyright 2000 Free Software Foundation, Inc. GDB is free software, covered by the GNU General Public License, and you are welcome to change it and/or distribute copies of it under certain conditions. Type "show copying" to see the conditions. There is absolutely no warranty for GDB. Type "show warranty" for details. This GDB was configured as "--host=i686-pc-linux-gnu --target=m68k-bdm-elf". Coldfire debug module version is 0 (5206(e)) (gdb) Am I running the correct version? Shouldn't the "Coldfire debug module version" be 5272 instead of 5206? ------------------------------------------------------------------------ This email may contain confidential information and/or copyright material. This email is intended for the use of the addressee only. Any unauthorized use may be unlawful. If you receive this email by mistake, please advise the sender immediately by using the reply facility in your email software.You should be aware that the contents of this email may be monitored to ensure compliance with the Meggitt IT User policy. This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From Jim.Henderson at endevco.com Tue Mar 26 16:03:29 2002 From: Jim.Henderson at endevco.com (Henderson, Jim) Date: Tue, 26 Mar 2002 13:03:29 -0800 Subject: [uClinux-dev] Gdb+bdm+5272C3 Message-ID: <412A28C967692A438B59223143A7268FE1B147@ALTAIR.endevco.com> I tried patching gdb-5.1.1 with gdb-bdm-20010901 (from ftp://www.cybertec.com.au/pub/bdm) and it didn't work for me. I had get gdb-5.0 from ftp://sources.redhat.com/pub/gdb/old-releases/ and patch it. Is gdb-bdm-20010901 supposed to patch gdb-5.1.1? -----Original Message----- From: Julien Boibessot [mailto:julienb at actimage.fr] Sent: Monday, March 18, 2002 12:04 AM To: uclinux-dev at uclinux.org Subject: Re: [uClinux-dev] Gdb+bdm+5272C3 Hi Greg ! Well, in fact my gdb init script is the one you posted on this list some times ago.....:-) And it initializes DRAM controller...... Thanks Julien ----- Original Message ----- From: "Greg Ungerer" To: Sent: Monday, March 18, 2002 12:53 AM Subject: Re: [uClinux-dev] Gdb+bdm+5272C3 > > Hi Julien, > > Julien Boibessot wrote: > > > I'm trying to use GDB thrue BDM (P&E Adaptator) with the Motorola 5272C3. > > > > I have compile and insert BDM driver ( gdb-bdm-20010901 from cybertec) in my > > host linux and it works (chk /dev/bdmcf0 is running and all tests are ok). > > I've got GDB 5.1.1 sources, patche it for m68k/coldfire BDM debugging and > > recompile it to get a running "m68k-bdm-elf-gdb". > > > > Then, I launch gdb with: "m68k-bdm-elf-gdb --command=5272.gdb image.elf". > > 5272.gdb is my init script for GDB and image.elf is my linux image. > > > > gdb is running and I can do whatever I want. So I launch the "load" command, > > take a coffee to wait the end of the download, and "set $pc=0x20000". But > > then when I try to execute the program by typing command "c", gdb crashes (I > > get a "segmentation fault"). My board is reseted. > > > > Does anybody knows what is happening ??? > > > Does you gdbinit script initialize RAM/SDRAM? > > If it doesn't then your load cannot possibly work, since > it will be trying to load the binary into RAM. > > Regards > Greg > > > ------------------------------------------------------------------------ > Greg Ungerer -- Chief Software Wizard EMAIL: gerg at snapgear.com > SnapGear PHONE: +61 7 3435 2888 > 825 Stanley St, FAX: +61 7 3891 3630 > Woolloongabba, QLD, 4102, Australia WEB: www.snapgear.com > > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ ------------------------------------------------------------------------ This email may contain confidential information and/or copyright material. This email is intended for the use of the addressee only. Any unauthorized use may be unlawful. If you receive this email by mistake, please advise the sender immediately by using the reply facility in your email software.You should be aware that the contents of this email may be monitored to ensure compliance with the Meggitt IT User policy. This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From hamilton at SEDSystems.ca Tue Mar 26 16:09:40 2002 From: hamilton at SEDSystems.ca (Kendrick Hamilton) Date: Tue, 26 Mar 2002 15:09:40 -0600 (Canada Central Standard Time) Subject: [uClinux-dev] tftp problem In-Reply-To: <20020326191206.17122.qmail@web20607.mail.yahoo.com> Message-ID: What is dn -i ? I have been using tftp in busybox. The command I use to download to the board is tftp get. On Tue, 26 Mar 2002, Yosh K wrote: > Hello, > I am trying to upload the image to MCF5272 eval board > using tftp. I've tested the tftp server and works > fine. Here is how the board is setup > base: 16 > baud: 19200 > server: 192.168.99.13 > client: 192.168.99.14 > gateway: 0.0.0.0 > netmask: 255.255.0.0 > filename: image.bin > filetype: ELF > mac: 00:CF:52:72:C3:01 > > When I do dn -i image.bin i get the error "unable to > locate 192.168.99.13" I know the tftp server is > running on 192.168.99.13. So is there something I > missed to setup properly. > Thanks > Yosh > > __________________________________________________ > Do You Yahoo!? > Yahoo! Movies - coverage of the 74th Academy Awards? > http://movies.yahoo.com/ > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From hamilton at SEDSystems.ca Tue Mar 26 16:14:58 2002 From: hamilton at SEDSystems.ca (Kendrick Hamilton) Date: Tue, 26 Mar 2002 15:14:58 -0600 (Canada Central Standard Time) Subject: [uClinux-dev] spawning processes without having a file system In-Reply-To: <8CCBDD5583C50E4196F012E79439B45C3644FE@atl-ms1.megatrends.com> Message-ID: Have you looked at how busybox works. It is one executable that does multiple different jobs. It looks at the command line parameters to figure out how it should be have. That way, all the different little programs use one copy of the library functions. Another thing you could try is zflat (zipped flat) format. On Tue, 26 Mar 2002, Subash Kalbarga wrote: > Hi all > > I am on a M5372 and am running into space (on ROM) problems because I have a > lot of programs who all link to the same libraries. > > I was wondering if there is any easy way to make all the programs into one > big image (not unlike the kernel) where all functions are available to > everyone else? > > If I do this the kernel should be able to start processes that are defined > by function entry points rather than as files in the file system > > For eg: I would just have a function alert_process() that is the entry point > into the process and maybe a process table where the kernel looks up and > starts these. > > Is this a feasible thought? Any suggestions on how I should approach it? > > Thanks in advance > Subash > > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From justin.wojdacki at analog.com Tue Mar 26 16:20:58 2002 From: justin.wojdacki at analog.com (Justin Wojdacki) Date: Tue, 26 Mar 2002 13:20:58 -0800 Subject: [uClinux-dev] spawning processes without having a file system References: <8CCBDD5583C50E4196F012E79439B45C3644FE@atl-ms1.megatrends.com> Message-ID: <3CA0E63A.3EB8B3C9@analog.com> Take a look at Busybox. This program already does some of what you're talking about. You'll have to modify the loader in the kernel, however, as the kernel expects programs to be in some form of filesystem. Also, some functions use temporary files in /var/tmp, so you'd need to be sure that none of your code does this. Oh yeah, and if the kernel can't find a root filesystem when it boots, it will panic. It needs it for things like /dev. You might be able to get around this by leveraging some of the netboot code, but you'll still have to have some root filesystem somewhere for the kernel. -- ------------------------------------------------- Justin Wojdacki justin.wojdacki at analog.com (408) 350-5032 Communications Processors Group -- Analog Devices This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From philip at engarts.com Tue Mar 26 16:24:27 2002 From: philip at engarts.com (Philip Nye) Date: Tue, 26 Mar 2002 21:24:27 -0000 Subject: [uClinux-dev] tftp problem References: <20020326191206.17122.qmail@web20607.mail.yahoo.com> Message-ID: <045f01c1d50c$9c6b8320$5801a8c0@localdomain> Hi Yosh, I am not sure what your trouble is but there are some differences from my setup. Your netmask should be 255.255.255.0 since 192.168.x.x is actually 256 class C nets rather than 1 class B (I doubt that the eval board checks but it might). Your filetype should be "Image" not "ELF" to match image.bin. I also have a gateway address on the local net (e.g. 192.168.99.1) - I don't know whether this would make a difference. Philip Nye ----- Original Message ----- From: "Yosh K" To: Sent: Tuesday, March 26, 2002 7:12 PM Subject: [uClinux-dev] tftp problem > Hello, > I am trying to upload the image to MCF5272 eval board > using tftp. I've tested the tftp server and works > fine. Here is how the board is setup > base: 16 > baud: 19200 > server: 192.168.99.13 > client: 192.168.99.14 > gateway: 0.0.0.0 > netmask: 255.255.0.0 > filename: image.bin > filetype: ELF > mac: 00:CF:52:72:C3:01 > > When I do dn -i image.bin i get the error "unable to > locate 192.168.99.13" I know the tftp server is > running on 192.168.99.13. So is there something I > missed to setup properly. > Thanks > Yosh > > __________________________________________________ > Do You Yahoo!? > Yahoo! Movies - coverage of the 74th Academy Awards. > http://movies.yahoo.com/ > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From philip at engarts.com Tue Mar 26 16:29:07 2002 From: philip at engarts.com (Philip Nye) Date: Tue, 26 Mar 2002 21:29:07 -0000 Subject: [uClinux-dev] spawning processes without having a file system References: <8CCBDD5583C50E4196F012E79439B45C3644FE@atl-ms1.megatrends.com> Message-ID: <047401c1d50d$430e8240$5801a8c0@localdomain> Hello Subash, This is the way that Busybox works so have a look at that. It builds all the "apps" into a single executable which shares as much code as possible. It determines what to do by what name it is called by - all its "applictions" in /bin are actually just links to the same busybox executable. Philip Nye ----- Original Message ----- From: "Subash Kalbarga" To: Sent: Tuesday, March 26, 2002 8:13 PM Subject: [uClinux-dev] spawning processes without having a file system > Hi all > > I am on a M5372 and am running into space (on ROM) problems because I have a > lot of programs who all link to the same libraries. > > I was wondering if there is any easy way to make all the programs into one > big image (not unlike the kernel) where all functions are available to > everyone else? > > If I do this the kernel should be able to start processes that are defined > by function entry points rather than as files in the file system > > For eg: I would just have a function alert_process() that is the entry point > into the process and maybe a process table where the kernel looks up and > starts these. > > Is this a feasible thought? Any suggestions on how I should approach it? > > Thanks in advance > Subash > > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From bruce at tele-ip.com Tue Mar 26 18:22:15 2002 From: bruce at tele-ip.com (Bruce Paterson) Date: Wed, 27 Mar 2002 10:22:15 +1100 Subject: [uClinux-dev] Gdb+bdm+5272C3 References: <412A28C967692A438B59223143A7268FE1B147@ALTAIR.endevco.com> Message-ID: <3CA102A7.3B46E3F5@tele-ip.com> "Henderson, Jim" wrote: > > I tried patching gdb-5.1.1 with gdb-bdm-20010901 (from > ftp://www.cybertec.com.au/pub/bdm) and it didn't work for me. I had get > gdb-5.0 from ftp://sources.redhat.com/pub/gdb/old-releases/ and patch it. > > Is gdb-bdm-20010901 supposed to patch gdb-5.1.1? I used the 5.0 patches to patch gdb 5.1.0 (part of Insight). A few patches don't apply directly, but with a bit of manual assistance based on the intent of the patch it was possible to get them all through. -- Cheers, Bruce ------------------------------------------------------------------- This email and any files transmitted with it are confidential and intended solely for the use of the individual or entity to whom they are addressed. If you have received this email in error please notify the system manager. /\\\/\\\/\\\ / / Bruce Paterson / \\\ \\\ \\\ / / Senior Design Engineer / /\\\/\\\/\\\/ / 87 Peters Ave, Mulgrave, Vic, 3170 / / \\\ \\\ \\\ / PO Box 4112, Mulgrave, Vic, 3170, Australia / / \\\/\\\ \\\/ Ph: +61 3 8561 4232 Fax: +61 3 9560 9055 Tele-IP Ltd. Email: bruce at tele-ip.com Icq: #32015991 WWW: http://www.tele-ip.com VK3TJN ------------------------------------------------------------------- This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From davidm at snapgear.com Tue Mar 26 19:17:26 2002 From: davidm at snapgear.com (David McCullough) Date: Wed, 27 Mar 2002 10:17:26 +1000 Subject: [uClinux-dev] Picogui - compiled and linked.... In-Reply-To: <3CA0A377.CC036AF5@cern.ch>; from jody.pearson@cern.ch on Tue, Mar 26, 2002 at 05:36:07PM +0100 References: <3CA0A377.CC036AF5@cern.ch> Message-ID: <20020327101726.A3613@beast.internal.moreton.com.au> Jivin Jody Pearson lays it down ... ... > -rwxr--r-- 1 root 5764 299808 Mar 26 17:24 pgserver > # file pgserver* > pgserver: data > pgserver.gdb: ELF 32-bit MSB executable, Motorola 68000, version 1, > statically linked, not stripped > # > > Then i copy that into me romfs, and remake the image. > > SO when I try to run the bin, i get; > > Sash command shell (version 1.1.1) > /> pgserver > MAP_SHARED not supported (cannot write mappings to disk) There is an "mmap" call somewhere in the code that is using mmap in a way that isn't supported under uClinux. You probably need to use an anonymous mapping to get it going. Cheers, Davidm -- David McCullough: Ph: +61 7 3435 2815 http://www.SnapGear.com davidm at snapgear.com Fx: +61 7 3891 3630 825 Stanley St., W'gabba QLD 4102, Oz This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From gerg at snapgear.com Tue Mar 26 19:45:06 2002 From: gerg at snapgear.com (Greg Ungerer) Date: Wed, 27 Mar 2002 10:45:06 +1000 Subject: [uClinux-dev] Re: v850 update patch 20020326 References: <3CA019F0.2050402@snapgear.com> Message-ID: <3CA11612.9000505@snapgear.com> Hi Miles, Done. Regards Greg Miles Bader wrote: > Thanks > > I forgot one file, though; could you apply this patch too? > > Patch: > > Index: include/asm-v850/dma.h > =================================================================== > RCS file: /var/cvs/uClinux-2.4.x/include/asm-v850/dma.h,v > retrieving revision 1.1 > diff -u -r1.1 dma.h > --- include/asm-v850/dma.h 2001/10/22 13:39:02 1.1 > +++ include/asm-v850/dma.h 2002/03/26 08:09:31 > @@ -9,6 +9,10 @@ > /* release it again */ > extern void free_dma (unsigned int dmanr); > > +#ifdef CONFIG_PCI > +extern int isa_dma_bridge_buggy; > +#else > #define isa_dma_bridge_buggy (0) > +#endif > > #endif /* __V850_DMA_H__ */ > > > -Miles > -- ------------------------------------------------------------------------ Greg Ungerer -- Chief Software Wizard EMAIL: gerg at snapgear.com SnapGear PHONE: +61 7 3435 2888 825 Stanley St, FAX: +61 7 3891 3630 Woolloongabba, QLD, 4102, Australia WEB: www.snapgear.com This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From gerg at snapgear.com Tue Mar 26 20:19:02 2002 From: gerg at snapgear.com (Greg Ungerer) Date: Wed, 27 Mar 2002 11:19:02 +1000 Subject: [uClinux-dev] patch to add v850 support to busybox insmod References: Message-ID: <3CA11E06.8040705@snapgear.com> Hi Miles, Rolled into the uClinux-dist code. Thanks Greg Miles Bader wrote: > Hi, > > This patch adds support for the v850e processor to busybox's insmod > program. The patch is against busybox 0.60.0 from uClinux-dist-20020214. > > Could you apply it? > > Thanks, > > -Miles > > Patch: > ------------------------------------------------------------------------ Greg Ungerer -- Chief Software Wizard EMAIL: gerg at snapgear.com SnapGear PHONE: +61 7 3435 2888 825 Stanley St, FAX: +61 7 3891 3630 Woolloongabba, QLD, 4102, Australia WEB: www.snapgear.com This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From jeff at arcturusnetworks.com Tue Mar 26 20:24:56 2002 From: jeff at arcturusnetworks.com (D. Jeff Dionne) Date: Wed, 27 Mar 2002 01:24:56 -0000 Subject: [uClinux-dev] Picogui - compiled and linked.... Message-ID: <200203270124.g2R1OuF30646@arcturusnetworks.com> David McCullough said: > Jivin Jody Pearson lays it down ... .. > > /> pgserver > > MAP_SHARED not supported (cannot write mappings to disk) > > > There is an "mmap" call somewhere in the code that is using mmap in a way > that isn't supported under uClinux. You probably need to use an > anonymous mapping to get it going. It will likely be in the frame buffer driver when it tries to map /dev/fb into memory for direct access. Look there, possibly with an eye to hacking it since under uClinux it's mapped all the time anyway ;^) J > > Cheers, > Davidm > > -- > David McCullough: Ph: +61 7 3435 2815 http://www.SnapGear.com > davidm at snapgear.com Fx: +61 7 3891 3630 825 Stanley St., W'gabba QLD 4102, Oz > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ > -- D. Jeff Dionne Jeff at ArcturusNetworks.com -------------- next part -------------- A non-text attachment was scrubbed... Name: Type: application/octet-stream Size: 2 bytes Desc: not available URL: From Manigandanv at myw.ltindia.com Tue Mar 26 22:02:15 2002 From: Manigandanv at myw.ltindia.com (Manigandan.V) Date: Wed, 27 Mar 2002 08:32:15 +0530 Subject: [uClinux-dev] tftp problem References: <20020326191206.17122.qmail@web20607.mail.yahoo.com> Message-ID: <009301c1d53b$f1a69ee0$4806a8c0@emsysdom> Hello Yosh, Even we had a similar problem .You can do the following to check up 1.Isolate the board from intranet i.e., only host and target with crosswire ethernet cable 2.Download NBTFTP Server for windows and try the same in windows hyperterminal . 3.Change netmask to class c, change filetype to image and give proper gateway PS: Dbug doesn't support ping so dont try ping for netwrok testing Regards Manigandan ----- Original Message ----- From: "Yosh K" To: Sent: Wednesday, March 27, 2002 12:42 AM Subject: [uClinux-dev] tftp problem Hello, I am trying to upload the image to MCF5272 eval board using tftp. I've tested the tftp server and works fine. Here is how the board is setup base: 16 baud: 19200 server: 192.168.99.13 client: 192.168.99.14 gateway: 0.0.0.0 netmask: 255.255.0.0 filename: image.bin filetype: ELF mac: 00:CF:52:72:C3:01 When I do dn -i image.bin i get the error "unable to locate 192.168.99.13" I know the tftp server is running on 192.168.99.13. So is there something I missed to setup properly. Thanks Yosh __________________________________________________ Do You Yahoo!? Yahoo! Movies - coverage of the 74th Academy Awards? http://movies.yahoo.com/ This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From daniel.haensse at fhk.usz.ch Wed Mar 27 02:06:49 2002 From: daniel.haensse at fhk.usz.ch (Daniel Haensse) Date: Wed, 27 Mar 2002 08:06:49 +0100 Subject: [uClinux-dev] More on "delay loops" and timers! In-Reply-To: References: Message-ID: <200203270711.IAA01003@ds-00122.usz.ch> Hi Jonathan, the ld looks good. We had in the past some trouble setting up the development environment right. Maybe it is something like that. Kitkat is going to release a distro cd for the dragonix mainboard this week, that should cleanly install and set all paths on your linux box. Do you have two boards, do they behave the same way? Dani Am Dienstag, 26. M?rz 2002 19.58 schrieben Sie: > I am still having problems getting the CVS version of the uClinux kernel to > run on my board. It still crashes at "Calibrating delay loop...". I have > looked everywhere for a reason the timer interrupt wouldn't be happening, > and I have checked the .ld file. I have started to suspect the last few > commands that my Makefile runs to get the final kernel image. Here are the > commands: > > $(CROSS_COMPILE)objcopy -O binary --remove-section=.romvec \ > --remove-section=.text --remove-section=.ramvec \ > --remove-section=.bss --remove-section=.eram \ > $(ROOTDIR)/$(LINUXDIR)/linux $(IMAGEDIR)/linux.data > $(CROSS_COMPILE)objcopy -O binary --remove-section=.ramvec \ > --remove-section=.bss --remove-section=.data \ > --remove-section=.eram \ > > --set-section-flags=.romvec=CONTENTS,ALLOC,LOAD,READONLY,CODE \ > $(ROOTDIR)/$(LINUXDIR)/linux $(IMAGEDIR)/linux.text > cat $(IMAGEDIR)/linux.text $(IMAGEDIR)/linux.data $(ROMFSIMG) > > $(IMAGE) > > > Do these seem correct? Perhaps my .ld file is missing some globals? I > have attached it as well. I have 4MB of flash and 16MB of SDRAM. Please > let me know if you have any ideas! (is there any other files of interest I > can attach?) > > Thanks, > > Jonathan Schmidt > > _________________________________________________________________ > Send and receive Hotmail on your mobile device: http://mobile.msn.com This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From petero at cvs.com.au Wed Mar 27 02:32:40 2002 From: petero at cvs.com.au (Peter Ogilvy) Date: Wed, 27 Mar 2002 18:32:40 +1100 Subject: [uClinux-dev] Cheap Coldfire 5272 Dev Board Message-ID: <5.1.0.14.1.20020327182620.00a79cf0@localhost> Hi All, We are looking for a cheap 5272 board with 100baseT ethernet, which runs uClinux. It is mainly for us to test the tool chain while we debug our own 5272 hardware design, so its other specs are not the most critical aspect. What is important to us is that it is imediately available and has already been used to run uClinux so we can get it up and running quickly. Any suggestions? Thanks Peter Ogilvy This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From julienb at actimage.fr Wed Mar 27 02:59:55 2002 From: julienb at actimage.fr (Julien Boibessot) Date: Wed, 27 Mar 2002 08:59:55 +0100 Subject: [uClinux-dev] Gdb+bdm+5272C3 References: <412A28C967692A438B59223143A7268FE1B147@ALTAIR.endevco.com> Message-ID: <04ee01c1d565$61ce2e80$8947ec95@bruker.fr> No this patch doesn't work with 5.1.x gdb...... I was faced to this problem too...... 5.0 gdb is the right choice....... Julien ----- Original Message ----- From: "Henderson, Jim" To: Sent: Tuesday, March 26, 2002 10:03 PM Subject: RE: [uClinux-dev] Gdb+bdm+5272C3 > I tried patching gdb-5.1.1 with gdb-bdm-20010901 (from > ftp://www.cybertec.com.au/pub/bdm) and it didn't work for me. I had get > gdb-5.0 from ftp://sources.redhat.com/pub/gdb/old-releases/ and patch it. > > Is gdb-bdm-20010901 supposed to patch gdb-5.1.1? > > -----Original Message----- > From: Julien Boibessot [mailto:julienb at actimage.fr] > Sent: Monday, March 18, 2002 12:04 AM > To: uclinux-dev at uclinux.org > Subject: Re: [uClinux-dev] Gdb+bdm+5272C3 > > > Hi Greg ! > > Well, in fact my gdb init script is the one you posted on this list some > times ago.....:-) > And it initializes DRAM controller...... > > Thanks > Julien > > ----- Original Message ----- > From: "Greg Ungerer" > To: > Sent: Monday, March 18, 2002 12:53 AM > Subject: Re: [uClinux-dev] Gdb+bdm+5272C3 > > > > > > Hi Julien, > > > > Julien Boibessot wrote: > > > > > I'm trying to use GDB thrue BDM (P&E Adaptator) with the Motorola > 5272C3. > > > > > > I have compile and insert BDM driver ( gdb-bdm-20010901 from cybertec) > in my > > > host linux and it works (chk /dev/bdmcf0 is running and all tests are > ok). > > > I've got GDB 5.1.1 sources, patche it for m68k/coldfire BDM debugging > and > > > recompile it to get a running "m68k-bdm-elf-gdb". > > > > > > Then, I launch gdb with: "m68k-bdm-elf-gdb --command=5272.gdb > image.elf". > > > 5272.gdb is my init script for GDB and image.elf is my linux image. > > > > > > gdb is running and I can do whatever I want. So I launch the "load" > command, > > > take a coffee to wait the end of the download, and "set $pc=0x20000". > But > > > then when I try to execute the program by typing command "c", gdb > crashes (I > > > get a "segmentation fault"). My board is reseted. > > > > > > Does anybody knows what is happening ??? > > > > > > Does you gdbinit script initialize RAM/SDRAM? > > > > If it doesn't then your load cannot possibly work, since > > it will be trying to load the binary into RAM. > > > > Regards > > Greg > > > > > > ------------------------------------------------------------------------ > > Greg Ungerer -- Chief Software Wizard EMAIL: gerg at snapgear.com > > SnapGear PHONE: +61 7 3435 2888 > > 825 Stanley St, FAX: +61 7 3891 3630 > > Woolloongabba, QLD, 4102, Australia WEB: www.snapgear.com > > > > This message resent by the uclinux-dev at uclinux.org list server > http://www.uClinux.org/ > > > > This message resent by the uclinux-dev at uclinux.org list server > http://www.uClinux.org/ > > > ------------------------------------------------------------------------ > This email may contain confidential information and/or copyright material. > This email is intended for the use of the addressee only. Any unauthorized > use may be unlawful. If you receive this email by mistake, please advise the > sender immediately by using the reply facility in your email software.You > should be aware that the contents of this email may be monitored to ensure > compliance with the Meggitt IT User policy. > > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From eric.bosch at fourtress.nl Wed Mar 27 03:41:31 2002 From: eric.bosch at fourtress.nl (Eric Bosch (Fourtress)) Date: Wed, 27 Mar 2002 09:41:31 +0100 Subject: [uClinux-dev] Problems when booting from uCLinux Message-ID: If I try the go command in Minicom to execute the OS image in the flash ROM, I don't see a login screen. What's wrong! Thanks, Eric -------------- next part -------------- An HTML attachment was scrubbed... URL: From philippe.ney at smartdata.ch Wed Mar 27 05:27:58 2002 From: philippe.ney at smartdata.ch (Philippe Ney) Date: Wed, 27 Mar 2002 11:27:58 +0100 Subject: [uClinux-dev] Picogui - compiled and linked.... In-Reply-To: <3CA0A377.CC036AF5@cern.ch> References: <3CA0A377.CC036AF5@cern.ch> Message-ID: <20020327112758.4af62733.philippe.ney@smartdata.ch> > Hi All, > > Well I have made serious changes to the normal linking process. > [deleted stuff] > > Sash command shell (version 1.1.1) > /> pgserver > MAP_SHARED not supported (cannot write mappings to disk) Yes, as said David and Jeff, try using another video driver. BTW, we use PicoGUI with uClinux but with the pic compiler and with uClibc. Also, you can subscribe to ... -philippe This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From jody.pearson at cern.ch Wed Mar 27 06:24:22 2002 From: jody.pearson at cern.ch (Jody Pearson) Date: Wed, 27 Mar 2002 12:24:22 +0100 Subject: [uClinux-dev] Picogui - compiled and linked.... References: <3CA0A377.CC036AF5@cern.ch> <20020327112758.4af62733.philippe.ney@smartdata.ch> Message-ID: <3CA1ABE6.A4645A10@cern.ch> Hi all, I appreciate all of the help. I tracked it down to the opening of the framebuffer device, in video/fbdev.c /* Map it */ // FB_MEM = mmap(0,fbdev_mapsize = fixinfo.smem_len, // PROT_READ|PROT_WRITE,MAP_SHARED,fbdev_fd,0); FB_MEM = mmap(0,fbdev_mapsize = fixinfo.smem_len, PROT_READ|PROT_WRITE, 0,fbdev_fd,0); I believe (if i am reading this correctly) than the MAP_SHARED is something to do with the depth ? Anyway - I can get picogui to run as an elf binary, under an elf kernel, and display a black background with a little cursor. I intend to get the input (touchscreen) working, then introduce the changes as patches to picogui. Thanks all again for your help Jody PS - I am also considering a website of screenshots/info/patches and so on, anyone interested in that (ie - is it worth me bothering ?) This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From Manigandanv at myw.ltindia.com Wed Mar 27 07:14:51 2002 From: Manigandanv at myw.ltindia.com (Manigandan.V) Date: Wed, 27 Mar 2002 17:44:51 +0530 Subject: [uClinux-dev] How to convert bin file to elf References: <5.1.0.14.1.20020327182620.00a79cf0@localhost> Message-ID: <003401c1d588$ff3c6060$4806a8c0@emsysdom> Hi, I am trying to debug uclinux kernel using gdb-bdm-elf but for that i want image.elf (linux image) .But after make i could not find this .How to form image.elf from image.bin . I am assuming i should have image.elf alone for kernel debugging . I think we cant do kernel debugging for the kernel compiled for "execute from rom " We have to have 'execute from RAM' am i right ? Regards Mani This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From gerg at snapgear.com Wed Mar 27 08:14:46 2002 From: gerg at snapgear.com (Greg Ungerer) Date: Wed, 27 Mar 2002 23:14:46 +1000 Subject: [uClinux-dev] Re: Cheap Coldfire 5272 Dev Board References: <5.1.0.14.1.20020327182620.00a79cf0@localhost> Message-ID: <3CA1C5C6.1070904@snapgear.com> Hi Peter, Peter Ogilvy wrote: > We are looking for a cheap 5272 board with 100baseT > ethernet, which runs uClinux. > > It is mainly for us to test the tool chain while > we debug our own 5272 hardware design, so its other > specs are not the most critical aspect. > > What is important to us is that it is imediately > available and has already been used to run uClinux > so we can get it up and running quickly. > > Any suggestions? Heres 2: 1. Motorola M5272C3 dev board (not sure how much these cost) 2. SnapGear Lite, http://www.snapgear.com/lite.html (US$249 retail) - ColdFire 5272 @ 66MHz - 4MB SDRAM - 2MB FLASH - 10/100 MB ethernet - 2nd 10MB only ethernet - 1 serial port Both run uClinux very well. The SnapGear/lite ships as standard running uClinux out of FLASH. Regards Greg ------------------------------------------------------------------------ Greg Ungerer -- Chief Software Wizard EMAIL: gerg at snapgear.com SnapGear PHONE: +61 7 3279 1822 825 Stanley St, FAX: +61 7 3279 1820 Woolloogabba, QLD, 4102, Australia WEB: www.snapgear.com This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From philippe.ney at smartdata.ch Wed Mar 27 08:39:11 2002 From: philippe.ney at smartdata.ch (Philippe Ney) Date: Wed, 27 Mar 2002 14:39:11 +0100 Subject: [uClinux-dev] Picogui - compiled and linked.... In-Reply-To: <3CA1ABE6.A4645A10@cern.ch> References: <3CA0A377.CC036AF5@cern.ch> <20020327112758.4af62733.philippe.ney@smartdata.ch> <3CA1ABE6.A4645A10@cern.ch> Message-ID: <20020327143911.05de804b.philippe.ney@smartdata.ch> > Hi all, > > Anyway - I can get picogui to run as an elf binary, under an elf kernel, and display a > black background with a little cursor. cool_ > I intend to get the input (touchscreen) working, then introduce the changes as patches to > picogui. If you use something like the mc68328digi from uClinux, just some little modifications of the chipslicets.c in pgserver/input ought to be sufficient to have it working. -philippe This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From damnyosh at yahoo.com Wed Mar 27 11:01:14 2002 From: damnyosh at yahoo.com (Yosh K) Date: Wed, 27 Mar 2002 08:01:14 -0800 (PST) Subject: [uClinux-dev] Kunh's 5272 site Message-ID: <20020327160114.9539.qmail@web20609.mail.yahoo.com> Hello, Does some one have the readme files that were at this site site? http://www.rcs.ei.tum.de/~kuhn/uclinux/coldfire/tarifa/20011119/README Or where I can find information of putting the uClinux image into flash on MCF5272 eval board. Thanks Yosh __________________________________________________ Do You Yahoo!? Yahoo! Movies - coverage of the 74th Academy Awards? http://movies.yahoo.com/ This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From pgccoop at hotmail.com Wed Mar 27 12:37:37 2002 From: pgccoop at hotmail.com (Jonathan Schmidt) Date: Wed, 27 Mar 2002 17:37:37 +0000 Subject: [uClinux-dev] More on "delay loops" and timers! Message-ID: This is not a commercial product, no. It is going to be used by organizations such as the Pacific Geoscience Center (Canadian Government) in scientific research (seismometers and magnetometers, mainly) as a general purpose data-logger. I have a MAX3100 chip onboard in case there is need for 3 serial ports (as I understand it, there is need for more like 8 serial ports, but 3 will have to do). There is 4MB of flash and 16MB of SDRAM to give the developers plenty of room to play in. The destination of the data being logged is either another computer over the internet (hence the CS8900) or CompactFlash. Do you know of any drivers for CompactFlash (memory mapped mode) in uClinux that I can adapt to my board? I have looked briefly and found nothing obvious. I imagine that I will have a website available with schematics, board layouts, etc... at some point. Perhaps OpenHardware is interested in hosting a page for another project (ie, mine :-)? Perhaps I will ask Tom W. Thanks for all your (and others) help over the last few months -- you have been very resourceful! >From: Daniel Haensse >Reply-To: daniel.haensse at alumni.ethz.ch >To: "Jonathan Schmidt" >Subject: [uClinux-dev] More on "delay loops" and timers! > >Hi Jonathan, > >is this an commercial project you're doing? > >Dani _________________________________________________________________ Chat with friends online, try MSN Messenger: http://messenger.msn.com This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From MGroden at hxi.com Wed Mar 27 13:15:45 2002 From: MGroden at hxi.com (Mike Groden) Date: Wed, 27 Mar 2002 13:15:45 -0500 Subject: [uClinux-dev] Kunh's 5272 site Message-ID: <879EA880A0FED411996B00B0D0B082EC2B4104@hxi_exch01.hxi.com> I happened to make a copy. -----Original Message----- From: Yosh K [mailto:damnyosh at yahoo.com] Sent: Wednesday, March 27, 2002 11:01 AM To: uclinux-dev at uclinux.org Subject: [uClinux-dev] Kunh's 5272 site Hello, Does some one have the readme files that were at this site site? http://www.rcs.ei.tum.de/~kuhn/uclinux/coldfire/tarifa/20011119/README Or where I can find information of putting the uClinux image into flash on MCF5272 eval board. Thanks Yosh __________________________________________________ Do You Yahoo!? Yahoo! Movies - coverage of the 74th Academy Awards? http://movies.yahoo.com/ This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ -------------- next part -------------- An embedded and charset-unspecified text was scrubbed... Name: good boot loader stuff.txt URL: From Jim.Henderson at endevco.com Wed Mar 27 16:06:19 2002 From: Jim.Henderson at endevco.com (Henderson, Jim) Date: Wed, 27 Mar 2002 13:06:19 -0800 Subject: [uClinux-dev] GDM: "You can't do that when your target is `exec'" Message-ID: <412A28C967692A438B59223143A7268FE1B14B@ALTAIR.endevco.com> > What does `exec' mean here? > > # /home2/downloads/gdb-5.0/gdb/gdb /tftpboot/tbc_dbug4.elf > GNU gdb 5.0 > Copyright 2000 Free Software Foundation, Inc. > GDB is free software, covered by the GNU General Public License, and you > are > welcome to change it and/or distribute copies of it under certain > conditions. > Type "show copying" to see the conditions. > There is absolutely no warranty for GDB. Type "show warranty" for > details. > This GDB was configured as "--host=i686-pc-linux-gnu > --target=m68k-bdm-elf"... > Coldfire debug module version is 0 (5206(e)) > A program is being debugged already. Kill it? (y or n) y > (no debugging symbols found)... > (gdb) load > You can't do that when your target is `exec' > > ------------------------------------------------------------------------ This email may contain confidential information and/or copyright material. This email is intended for the use of the addressee only. Any unauthorized use may be unlawful. If you receive this email by mistake, please advise the sender immediately by using the reply facility in your email software.You should be aware that the contents of this email may be monitored to ensure compliance with the Meggitt IT User policy. This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From marky at crestaudio.com Wed Mar 27 16:30:12 2002 From: marky at crestaudio.com (Mark Yeck) Date: Wed, 27 Mar 2002 15:30:12 -0600 Subject: [uClinux-dev] NET+ARM with SDRAM Message-ID: Is anyone using the Net+40 with SDRAM? I dont know much about SDRAM and I'm trying to figure out how to get it to work right with gdb. It appears to be hooked up correctly, but I'm not sure about my .gdbinit file. I'm using the Net+40 with 2 x Micron 48LC4M16A2-75. -mark This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From Fabrice_Gautier at sdesigns.com Wed Mar 27 19:10:22 2002 From: Fabrice_Gautier at sdesigns.com (Fabrice Gautier) Date: Wed, 27 Mar 2002 16:10:22 -0800 Subject: [uClinux-dev] GDM: "You can't do that when your target is `ex ec'" Message-ID: "Your target is exec" means that your trying to debug a native executable file. You need to set your target to remote (or whatever your remote debugging method require) in order to load something on your target. -- Fabrice Gautier, Fabrice_Gautier at sdesigns.com > -----Original Message----- > From: Henderson, Jim [mailto:Jim.Henderson at endevco.com] > Sent: Wednesday, March 27, 2002 1:06 PM > To: uClinux-Dev (E-mail) > Subject: [uClinux-dev] GDM: "You can't do that when your target is > `exec'" > > > > What does `exec' mean here? > > > > # /home2/downloads/gdb-5.0/gdb/gdb /tftpboot/tbc_dbug4.elf > > GNU gdb 5.0 > > Copyright 2000 Free Software Foundation, Inc. > > GDB is free software, covered by the GNU General Public > License, and you > > are > > welcome to change it and/or distribute copies of it under certain > > conditions. > > Type "show copying" to see the conditions. > > There is absolutely no warranty for GDB. Type "show warranty" for > > details. > > This GDB was configured as "--host=i686-pc-linux-gnu > > --target=m68k-bdm-elf"... > > Coldfire debug module version is 0 (5206(e)) > > A program is being debugged already. Kill it? (y or n) y > > (no debugging symbols found)... > > (gdb) load > > You can't do that when your target is `exec' > > > > > -------------------------------------------------------------- > ---------- > This email may contain confidential information and/or > copyright material. > This email is intended for the use of the addressee only. Any > unauthorized > use may be unlawful. If you receive this email by mistake, > please advise the > sender immediately by using the reply facility in your email > software.You > should be aware that the contents of this email may be > monitored to ensure > compliance with the Meggitt IT User policy. > > This message resent by the uclinux-dev at uclinux.org list > server http://www.uClinux.org/ > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From Jim.Henderson at endevco.com Wed Mar 27 18:23:37 2002 From: Jim.Henderson at endevco.com (Henderson, Jim) Date: Wed, 27 Mar 2002 15:23:37 -0800 Subject: [uClinux-dev] ELF with "no debugging symbols found" Message-ID: <412A28C967692A438B59223143A7268FE1B14D@ALTAIR.endevco.com> I've tried the -g and the -ggdb compiler switches but I still get "no debugging symbols found". What am I doing wrong? # /home2/downloads/gdb-5.0/gdb/gdb /tftpboot/tbc_dbug4.elf GNU gdb 5.0 Copyright 2000 Free Software Foundation, Inc. GDB is free software, covered by the GNU General Public License, and you are welcome to change it and/or distribute copies of it under certain conditions. Type "show copying" to see the conditions. There is absolutely no warranty for GDB. Type "show warranty" for details. This GDB was configured as "--host=i686-pc-linux-gnu --target=m68k-bdm-elf"... Coldfire debug module version is 0 (5206(e)) A program is being debugged already. Kill it? (y or n) y (no debugging symbols found)... ------------------------------------------------------------------------ This email may contain confidential information and/or copyright material. This email is intended for the use of the addressee only. Any unauthorized use may be unlawful. If you receive this email by mistake, please advise the sender immediately by using the reply facility in your email software.You should be aware that the contents of this email may be monitored to ensure compliance with the Meggitt IT User policy. This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From Jim.Henderson at endevco.com Wed Mar 27 20:10:04 2002 From: Jim.Henderson at endevco.com (Henderson, Jim) Date: Wed, 27 Mar 2002 17:10:04 -0800 Subject: [uClinux-dev] No debugging information Message-ID: <412A28C967692A438B59223143A7268FE1B14F@ALTAIR.endevco.com> The m68k-elf-gcc switches are: ... /usr/local/bin/m68k-elf-gcc -m5200 -S -I. -I../../src/include -I../../src/dbug/v2 -I. -fno-builtin -ansi -Wall -Dmcf5272_DBUG -ggdb ../../src/dbug/v2/uif/net/tftp.c -o obj/mcf5272/gnugcc2953/Linux/uif/net/tftp.s ... The m68k-elf-as switches are: ... /usr/local/bin/m68k-elf-as -m5206e --register-prefix-optional -ahls=obj/mcf5272/gnugcc2953/Linux/uif/net/tftp.lst -o obj/mcf5272/gnugcc2953/Linux/uif/net/tftp.o obj/mcf5272/gnugcc2953/Linux/uif/net/tftp.s ... The m68k-elf-ar switches are: /usr/local/bin/m68k-elf-ar -rcs obj/mcf5272/gnugcc2953/Linux/libdbug.a obj/mcf5272/gnugcc2953/Linux/cpu/m68k/mcf5200/mcf5200_lo.o ... The m68k-elf-ld switches are: /usr/local/bin/m68k-elf-ld --verbose -E -T src/dbug/gnu/evs4.ld -Map obj/gnu/tbc_dbug4.map -E -L../libdbug/obj/mcf5272/gnugcc2953/Linux -L/usr/local/lib/gcc-lib/m68k-elf/2.95.3/m5200 obj/gnu/cpu/m68k/mcf5272/vectors.o ... obj/gnu/board/build.o -ldbug -lgcc And this generates an ELF file with no degugging information: /usr/local/bin/m68k-elf-objdump -g -S -d a.out > obj/gnu/tbc_dbug4.dis /usr/local/bin/m68k-elf-objdump: a.out: no recognized debugging information What am I doing wrong? ------------------------------------------------------------------------ This email may contain confidential information and/or copyright material. This email is intended for the use of the addressee only. Any unauthorized use may be unlawful. If you receive this email by mistake, please advise the sender immediately by using the reply facility in your email software.You should be aware that the contents of this email may be monitored to ensure compliance with the Meggitt IT User policy. This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From miles at lsi.nec.co.jp Wed Mar 27 22:31:59 2002 From: miles at lsi.nec.co.jp (Miles Bader) Date: 28 Mar 2002 12:31:59 +0900 Subject: [uClinux-dev] Re: v850 update patch 20020326 In-Reply-To: <3CA11612.9000505@snapgear.com> References: <3CA019F0.2050402@snapgear.com> <3CA11612.9000505@snapgear.com> Message-ID: Thank you... ... and ... could you also apply this patch: -------------- next part -------------- A non-text attachment was scrubbed... Name: rte_mb_a_pci-kmods.patch Type: text/x-patch Size: 3076 bytes Desc: Patch to add kernel-module support to arch/v850/kernel/rte_mb_a_pci.c URL: -------------- next part -------------- Thanks, -Miles -- .Numeric stability is probably not all that important when you're guessing. From davidm at snapgear.com Wed Mar 27 23:54:38 2002 From: davidm at snapgear.com (David McCullough) Date: Thu, 28 Mar 2002 14:54:38 +1000 Subject: [uClinux-dev] Re: v850 update patch 20020326 In-Reply-To: ; from miles@lsi.nec.co.jp on Thu, Mar 28, 2002 at 12:31:59PM +0900 References: <3CA019F0.2050402@snapgear.com> <3CA11612.9000505@snapgear.com> Message-ID: <20020328145438.A29463@beast.internal.moreton.com.au> Jivin Miles Bader lays it down ... > Thank you... > > ... and ... could you also apply this patch: Done, Cheers, Davidm -- David McCullough: Ph: +61 7 3435 2815 http://www.SnapGear.com davidm at snapgear.com Fx: +61 7 3891 3630 825 Stanley St., W'gabba QLD 4102, Oz This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From miles at lsi.nec.co.jp Thu Mar 28 00:29:46 2002 From: miles at lsi.nec.co.jp (Miles Bader) Date: 28 Mar 2002 14:29:46 +0900 Subject: [uClinux-dev] Re: v850 update patch 20020326 In-Reply-To: <20020328145438.A29463@beast.internal.moreton.com.au> References: <3CA019F0.2050402@snapgear.com> <3CA11612.9000505@snapgear.com> <20020328145438.A29463@beast.internal.moreton.com.au> Message-ID: Thank you again. BTW, how much effort does it take for you to apply one these patches, usually? I.e., should I feel a little guilty when I keep sending little patches like this, or is merely 3 keystrokes and snap of the fingers to apply them? [I usually prefer to wait a while and bundle things up, but sometimes ...] -Miles -- 97% of everything is grunge This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From damnyosh at yahoo.com Thu Mar 28 01:16:22 2002 From: damnyosh at yahoo.com (Yosh K) Date: Wed, 27 Mar 2002 22:16:22 -0800 (PST) Subject: [uClinux-dev] Kunh's 5272 site In-Reply-To: <879EA880A0FED411996B00B0D0B082EC2B4104@hxi_exch01.hxi.com> Message-ID: <20020328061622.96492.qmail@web20610.mail.yahoo.com> Thank you Mike. Good thinkin :) --- Mike Groden wrote: > I happened to make a copy. > > > -----Original Message----- > From: Yosh K [mailto:damnyosh at yahoo.com] > Sent: Wednesday, March 27, 2002 11:01 AM > To: uclinux-dev at uclinux.org > Subject: [uClinux-dev] Kunh's 5272 site > > > Hello, > Does some one have the readme files that were at > this > site site? > http://www.rcs.ei.tum.de/~kuhn/uclinux/coldfire/tarifa/20011119/README > Or where I can find information of putting the > uClinux image into flash on MCF5272 eval board. > Thanks > Yosh > > > __________________________________________________ > Do You Yahoo!? > Yahoo! Movies - coverage of the 74th Academy Awards? > http://movies.yahoo.com/ > This message resent by the uclinux-dev at uclinux.org > list server > http://www.uClinux.org/ > > > uClinux flash device mini howto > ******************************* > > Author: Bernhard Kuhn, Lineo Inc. > Revision: 0.1 > > Abstract: > This document describes how to set up the Motorola > M5272C3 > evaluation board 'Tarifa' to boot from flash and how > to > boot from a cramfs partition instead of an > kernel-attached > romfs image. This paper presumes that you already > know how > to boot Linux on Tarifa from RAM with an romfs image > as > root filesystem attached to the kernel image. > > Content: > o Choosing a bootloader > o The Flash Device Partition layout > o Installing Colilo > o Attaching the Linux-Image to Colilo, saving flash > memory > o Other Boot strategies > o Generating a cramfs images > o Creating an empty romfs image > o Creating a compressed kernel for flashing > o System recovery using the BDM interface cable > o Unresoled Issues > o List of Resources > o (No) Warranty > o Thanks > > > Choosing a bootloader > ===================== > > The Tarifa board (Motorola M5272C3 evaluation board) > is > shipped with a pre-installed bootloader called > "dBUG". This bootloader allows the user to download > images from a tftp server to memory (RAM) starting > at 0x20000 > (refer to the "dn" command in the dBUG manual) and > to > execute the kernel from there ("go 20000"). > Additionaly, downloaded > images can be moved to flash memory (7 x 256KB > starting at > 0xffe40000, see "upuser" in the Tarifa/dBUG manual). > > Those who don't want to remove the pre-installed > bootloader > can choose to install a bootloader-kernel-rootfs > combination > in the topmost one megabyte (0xfff00000 to > 0xffffffff) of the > flash device and after altering jumper 13 to > position 2-3, the M5272C3 > boots from there (but this option limits the > available amount > of flash memory to only one megabyte: the area from > 0xffe00000 > to 0xffeffffff is just a mirror of the topmost one > megabyte > in this mode). As dBUG (as pre-installed) cannot be > configured > to automatically start the kernel, you will have to > install > another bootloader to have a kernel-kickstarter and > in order to > make fully use of the two megabytes of flash memory > the Tarifa > board provides. The bootloader prefered in this > document > is called "Colilo" and is available under the terms > of the GPL. > Unfortnuatly, Colilo is not able to load images via > the > tftp mechanism, so dBUG is probably the better > choice during development. For the first steps, you > might > want to have dBUG booted by default (which is > located > at the first 256KB of the flash memory from > 0xffe00000 to > 0xffe3ffff) and invoke colilo (that resides at the > first 256KB > of the topmost one megabyte from 0xfff00000 to > 0xfff3ffff) > by typing "go fff00400". So switching jumper 13 > allows > you to either choose booting from dBUG (and > subsequently > starting colilo) or colilo directly. > > > The Flash Device Partition layout > ================================= > > The uClinux-distribution (as is) from > www.uclinux.org makes > use of the block memory device driver > (linux/drivers/block/blkmem.c). > This driver already provides flashing algorithms > that allows > the user to write to a broad range of flash device > (refer to > blkmem.c, and the flashw/netflash utilities in the > uClinux-distribution user tree for more information > on that topic). But the Memory Technology Device > drivers > (linux/drivers/mtd) are offering more options and > probably > will be the better supported drivers for flash > devices > in the future. This paper concentrates on MTD. > In this context, please note that the > journaling flash file system (jffs) apparently does > only work in > combination with MTD drivers, but not with the > blkmem driver. > > The standard uClinux-2.4.x kernel (CVS version) can > be > compiled that way that it recognises the 2 MByte > flash > on the Tarifa board using the "physmap" driver > (start=0xffe0000, size=0x200000, buswidth=2). > But in this case, there is only one big 2MB > partition. > This would make the process of flashing the > bootloader, kernel and rootfs seperatly more > complex. > For this reason, i have introduces a dedicated > mapping > driver for the Tarifa board [TARIFAMAP]. BTW.: > generating > a mapping device driver file is rather easy: just > copy "physmap.c", > add the partition table and replace the > add_mtd_device() > call in init_physmap() by an apropriate > add_mtd_partitions()-call (compare physmap.c and > tarifa.c > for details). Here is the list of the pre-defined > partitions: > > mtd0 / rom0: 0xffe00000-0xffe03fff : "boot (16K)" > mtd2 / rom1: 0xffe80000-0xffefffff : "kernel > (512K)" > mtd4 / rom2: 0xfff00000-0xffffffff : "rootfs > (1024K)" > mtd6 / rom3: 0xffe04000-0xffe05fff : "spare (8K)" > mtd8 / rom4: 0xffe06000-0xffe07fff : "spare (8K)" > mtd10 / rom5: 0xffe40000-0xffe7ffff : "spare (256K)" > mtd12 / rom6: 0xffe00000-0xffffffff : "complete > (2048K)" > mtd14 / rom7: 0xfff00000-0xfff3ffff : "boot J13 > (256K)" > mtd16 / rom8: 0xfff40000-0xfffbffff : "kernel J13 > (512K)" > mtd18 / rom9: 0xfffc0000-0xffffffff : "rootfs J13 > (256K)" > > First of all, you may have recognized that the minor > numbers for the MTD character device files (major > 90) > do not correspond to those of the mtdblock device > files > (major 31, also known as /dev/rom*) for a reason i > don't > know - maybe you can tell me :-) > > Additionaly, the mapping looks a little bit > fragmented. > This is due to the fact that the first 256KB memory > block > on the flash device of the Tarifa board is > fragmented in > in smaller eraseable flash blocks for bootloaders > and > configuration space. BTW.: the 224KByte from > 0xffe08000 > to 0xfffe3ffff is not directly mappable (the MTD > driver > will complain that the mapping is not block > aligned), > but can be accessed via mtd12/rom6, if necessary. > > mtd14, mtd16 and mtd18 are only necessary when > playing > around with jumper 13: if you are just using the > topmost > one megabyte of the flash, then you don't have the > need to delete the pre-installed dBUG bootloader > (if you don't have a BDM interface cable, this might > be > the best option for you, since otherwise you > probably === message truncated === __________________________________________________ Do You Yahoo!? Yahoo! Movies - coverage of the 74th Academy Awards? http://movies.yahoo.com/ This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From wangwb98 at sohu.com Thu Mar 28 01:30:35 2002 From: wangwb98 at sohu.com (wangwb98 at sohu.com) Date: Thu, 28 Mar 2002 14:30:35 +0800 (CST) Subject: [uClinux-dev] problem while compile egcs for arm7tdmi Message-ID: <604300.1017297035963.JavaMail.postfix@mailsrv11.mail.sohu.com> Hi, everyone I'm working on pentium PC building a arm7tdmi cross compile environmen. When I found the arm7tdmi ports of uclinux, I downloaded all the files. First, I compiled binutils using: ./configure --target=arm-uclinux --prefix=/opt/a7tdmi and make; make install, succeeded. Then, I make a symbol link /usr/src/linux/include point to the patched uclinux/include(I only made this patch uclinux-patch-export2.patch). Then, I make egcs using this command: ./configure --target=arm-uclinux --prefix=/opt/a7tdmi --with-headers=/usr/src/linux/include --enable-languages=c --disable-threads when I type make, it failed, the last few lines of error message is: /tmp/cco7Z7Jw.s: Assembler messages: /tmp/cco7Z7Jw.s:1971: Error: Subtraction of two symbols in different sections "strtol" {*UND* section} - ".L44" {*UND* section} at file address 3892. /tmp/cco7Z7Jw.s:1975: Error: Subtraction of two symbols in different sections "strtoul" {*UND* section} - ".L54" {*UND* section} at file address 3908. make[5]: *** [vfscanf.o] Error 1 make[5]: Leaving directory `/root/arm7tdmi/egcs/arm-uclinux/pic/newlib/libc/stdio' make[4]: *** [stdio/lib.a] Error 1 make[4]: Leaving directory `/root/arm7tdmi/egcs/arm-uclinux/pic/newlib/libc' make[3]: *** [libc/libc.a] Error 2 make[3]: Leaving directory `/root/arm7tdmi/egcs/arm-uclinux/pic/newlib' make[2]: *** [multi-do] Error 1 make[2]: Leaving directory `/root/arm7tdmi/egcs/arm-uclinux/newlib' make[1]: *** [all] Error 2 make[1]: Leaving directory `/root/arm7tdmi/egcs/arm-uclinux/newlib' make: *** [all-target-newlib] Error 2 I don't know why this happen, could anybody help me? thx in advance. This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From Joachim.Schweidler at gmx.de Thu Mar 28 01:52:50 2002 From: Joachim.Schweidler at gmx.de (Joachim Schweidler) Date: Thu, 28 Mar 2002 07:52:50 +0100 (MET) Subject: [uClinux-dev] M68360: IRQ 68 from ttyS is not replaceable, LSR security check engaged References: <1481.1017139814@www20.rz2.gmx.net> Message-ID: <19098.1017298370@www51.gmx.net> Hi all, some months ago I got uclinux 2.0.38 up and running on an 68EN360 Board (SBS V360). Now I upgraded to Kernel 2.4.x and the kernel starts and comes till opening an initial console and then I get the error "M68360_request_irq: IRQ 68 from ttyS is not replaceable". Another error that aperars sometimes is "LSR security check engaged". It's a CVS - snapshot from 03/16/2002. Doas anyone have the same problem or knows a solution to this problem. Many thanks in advance. Joe Here is a log I captured over the serial console (BDM Interface): ********************************************************************* Linux version 2.4.17-uc0 (root at for02067) (gcc version 2.95.3 20010315 (release)(Cold2 uClinux/MC68360 Flat model support (C) 1998,1999 Kenneth Albanowski, D. Jeff Dionne QUICC port done by SED Systems , based on 2.0.38 port by Lineo Inc. . On node 0 totalpages: 1023 zone(0): 0 pages. zone(1): 1023 pages. zone(2): 0 pages. Kernel command line: CPM UART driver version 0.03 ttyS00 at irq 0x04 is an SMC ttyS01 at irq 0x1d is an SCC ttyS02 at irq 0x1c is an SCC ttyS03 at irq 0x1b is an SCC Calibrating delay loop... 4.37 BogoMIPS Memory available: 2700k/3332k RAM, 394k/1024k ROM (630k kernel code, 549k data) kmem_create: Forcing size word alignment - vm_area_struct kmem_create: Forcing size word alignment - mm_struct kmem_create: Forcing size word alignment - filp Dentry-cache hash table entries: 512 (order: 0, 4096 bytes) Inode-cache hash table entries: 512 (order: 0, 4096 bytes) kmem_create: Forcing size word alignment - inode_cache Mount-cache hash table entries: 512 (order: 0, 4096 bytes) kmem_create: Forcing size word alignment - bdev_cache kmem_create: Forcing size word alignment - cdev_cache Buffer-cache hash table entries: 1024 (order: 0, 4096 bytes) Page-cache hash table entries: 1024 (order: 0, 4096 bytes) POSIX conformance testing by UNIFIX Linux NET4.0 for Linux 2.4 Based upon Swansea University Computer Society NET3.039 kmem_create: Forcing size word alignment - sock Initializing RT netlink socket Starting kswapd kmem_create: Forcing size word alignment - file lock cache Serial driver version 5.05c (2001-07-08) with no serial options enabled ttyS00 at 0x03f8 (irq = 4) is a 16450 ttyS01 at 0x02f8 (irq = 3) is a 16450 ttyS02 at 0x03e8 (irq = 4) is a 16450 kmem_create: Forcing size word alignment - blkdev_requests block: 64 slots per queue, batch=16 RAMDISK driver initialized: 16 RAM disks of 4096K size 1024 blocksize Blkmem copyright 1998,1999 D. Jeff Dionne Blkmem copyright 1998 Kenneth Albanowski Blkmem 1 disk images: 0: C6A08-124207 [VIRTUAL C6A08-124207] (RO) PPP generic driver version 2.4.1 NET4: Linux TCP/IP 1.0 for NET4.0 IP Protocols: ICMP, UDP, TCP kmem_create: Forcing size word alignment - ip_dst_cache IP: routing cache hash table of 512 buckets, 4Kbytes TCP: Hash tables configured (established 512 bind 512) NET4: Unix domain sockets 1.0/SMP for Linux NET4.0. VFS: Mounted root (romfs filesystem) readonly. Freeing unused kernel memory: 0k freed (0x127000 - 0x126000) M68360_request_irq: IRQ 68 from ttyS is not replaceable M68360_request_irq: IRQ 68 from ttyS is not replaceable M68360_request_irq: IRQ 68 from ttyS is not replaceable M68360_request_irq: IRQ 68 from ttyS is not replaceable -- GMX - Die Kommunikationsplattform im Internet. http://www.gmx.net This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From miles at lsi.nec.co.jp Thu Mar 28 02:57:16 2002 From: miles at lsi.nec.co.jp (Miles Bader) Date: 28 Mar 2002 16:57:16 +0900 Subject: [uClinux-dev] Re: v850 update patch 20020326 In-Reply-To: References: <3CA019F0.2050402@snapgear.com> <3CA11612.9000505@snapgear.com> <20020328145438.A29463@beast.internal.moreton.com.au> Message-ID: Er, speaking of which, I kinda forgot to update the Makefile... If you could: Index: arch/v850/kernel/Makefile =================================================================== RCS file: /var/cvs/uClinux-2.4.x/arch/v850/kernel/Makefile,v retrieving revision 1.3 diff -u -r1.3 Makefile --- arch/v850/kernel/Makefile 2002/03/26 06:51:28 1.3 +++ arch/v850/kernel/Makefile 2002/03/28 07:53:59 @@ -16,7 +16,7 @@ O_TARGET := kernel.o obj-y := intv.o entry.o process.o syscalls.o time.o semaphore.o setup.o \ signal.o irq.o mach.o ptrace.o bug.o -export-objs := v850_ksyms.o +export-objs := v850_ksyms.o rte_mb_a_pci.o obj-$(CONFIG_MODULES) += v850_ksyms.o # chip-specific code Thanks, -Miles -- Suburbia: where they tear out the trees and then name streets after them. This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From davidm at snapgear.com Thu Mar 28 03:13:33 2002 From: davidm at snapgear.com (David McCullough) Date: Thu, 28 Mar 2002 18:13:33 +1000 Subject: [uClinux-dev] Re: v850 update patch 20020326 In-Reply-To: ; from miles@lsi.nec.co.jp on Thu, Mar 28, 2002 at 02:29:46PM +0900 References: <3CA019F0.2050402@snapgear.com> <3CA11612.9000505@snapgear.com> <20020328145438.A29463@beast.internal.moreton.com.au> Message-ID: <20020328181333.A4859@beast.internal.moreton.com.au> Jivin Miles Bader lays it down ... > Thank you again. > > BTW, how much effort does it take for you to apply one these patches, > usually? I.e., should I feel a little guilty when I keep sending > little patches like this, or is merely 3 keystrokes and snap of the > fingers to apply them? When the patches are clean and up to date (like yours :-) it doesn't take long at all. Not much more than: # cd srcdir # patch -p0 < latest-patch-from-miles # cvs commit 'touched files' > [I usually prefer to wait a while and bundle things up, but sometimes ...] What you have been doing so far is fine by me. Bundling does help, but if you think something should go up sooner I'm sure Gerg or I will oblige :-) Cheers, Davidm -- David McCullough: Ph: +61 7 3435 2815 http://www.SnapGear.com davidm at snapgear.com Fx: +61 7 3891 3630 825 Stanley St., W'gabba QLD 4102, Oz This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From stefan.jonsson at faltcom.se Thu Mar 28 03:02:16 2002 From: stefan.jonsson at faltcom.se (Stefan Jonsson) Date: Thu, 28 Mar 2002 09:02:16 +0100 Subject: SV: [uClinux-dev] How to convert bin file to elf In-Reply-To: <003401c1d588$ff3c6060$4806a8c0@emsysdom> Message-ID: Hello, If I'm not wrong the .elf is also made along the 'make' and should exist somewhere if it is not stripped out somewhere in some makefile. Check the Makefiles. But if it is as I think the .elf is probably put somewhere else (it was for me). Browse your directories and u might find it :-) Yes, I think u have to execute from ram (I have to when debuging with lauterbach anyway). Regards, Stefan Jonsson. -----Ursprungligt meddelande----- Fr?n: owner-uclinux-dev at uclinux.org [mailto:owner-uclinux-dev at uclinux.org]F?r Manigandan.V Skickat: den 27 mars 2002 13:15 Till: uclinux-dev at uclinux.org ?mne: [uClinux-dev] How to convert bin file to elf Hi, I am trying to debug uclinux kernel using gdb-bdm-elf but for that i want image.elf (linux image) .But after make i could not find this .How to form image.elf from image.bin . I am assuming i should have image.elf alone for kernel debugging . I think we cant do kernel debugging for the kernel compiled for "execute from rom " We have to have 'execute from RAM' am i right ? Regards Mani This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From miles at lsi.nec.co.jp Thu Mar 28 03:08:15 2002 From: miles at lsi.nec.co.jp (Miles Bader) Date: 28 Mar 2002 17:08:15 +0900 Subject: [uClinux-dev] Re: out-of-line BUG In-Reply-To: References: Message-ID: Miles Bader writes: > Those of you who read the linux-kernel mailing list may have noticed Andrew > Morton's patch to shrink the kernel by changing calls to the `BUG' macro in > inline functions in header files to call a normal function `out_of_line_bug'. Any comment on this? If you don't want to apply the more pervasive patch, could you apply the following, which is the v850-specific part of it (so it's guaranteed not to conflict down the line; it should be compatible with Andrew's patch as it uses a weak symbol define `out_of_line_bug')? [I'll keep Andrew's patch in my personal tree, but I want to avoid differences with the repository as much as I can] -------------- next part -------------- A non-text attachment was scrubbed... Name: oolb-v850.patch Type: text/x-patch Size: 4573 bytes Desc: v850-specific out-of-line-BUG patch URL: -------------- next part -------------- Thanks, -Miles -- I'm beginning to think that life is just one long Yoko Ono album; no rhyme or reason, just a lot of incoherent shrieks and then it's over. --Ian Wolff From davidm at snapgear.com Thu Mar 28 03:23:21 2002 From: davidm at snapgear.com (David McCullough) Date: Thu, 28 Mar 2002 18:23:21 +1000 Subject: [uClinux-dev] problem while compile egcs for arm7tdmi In-Reply-To: <604300.1017297035963.JavaMail.postfix@mailsrv11.mail.sohu.com>; from wangwb98@sohu.com on Thu, Mar 28, 2002 at 02:30:35PM +0800 References: <604300.1017297035963.JavaMail.postfix@mailsrv11.mail.sohu.com> Message-ID: <20020328182321.B4859@beast.internal.moreton.com.au> Jivin wangwb98 at sohu.com lays it down ... > Hi, everyone > I'm working on pentium PC building a arm7tdmi cross compile environmen. When I found the arm7tdmi ports of uclinux, I downloaded all the files. > First, I compiled binutils using: ./configure --target=arm-uclinux --prefix=/opt/a7tdmi and make; make install, succeeded. > Then, I make a symbol link /usr/src/linux/include point to the patched uclinux/include(I only made this patch uclinux-patch-export2.patch). > Then, I make egcs using this command: > ./configure --target=arm-uclinux --prefix=/opt/a7tdmi --with-headers=/usr/src/linux/include --enable-languages=c --disable-threads > > when I type make, it failed, the last few lines of error message is: Have a look at http://www.uclinux.org/pub/uClinux/arm-elf-tools/ There are patches, source and a build script there that should help you makes tools for the arm7tdmi. Checkout build-uclinux-tools.sh for the build sequence. You can set PREFIX in the script to put the tools in another dir though I would let them go in /usr/local myself, Cheers, Davidm -- David McCullough: Ph: +61 7 3435 2815 http://www.SnapGear.com davidm at snapgear.com Fx: +61 7 3891 3630 825 Stanley St., W'gabba QLD 4102, Oz This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From sujz at cad.csie.ncku.edu.tw Thu Mar 28 03:58:33 2002 From: sujz at cad.csie.ncku.edu.tw (SuCC) Date: Thu, 28 Mar 2002 16:58:33 +0800 Subject: [uClinux-dev] about kswapd Message-ID: <002501c1d636$c1a0bf40$2df7748c@SCC> Hi, I have a memory ran off problem!! The memory size will reduced slowly when I run some programs.. So, I wnat to make sure if the kswapd work well and I add some message in kswapd!! After I start the system and run some programs, system memory still reduced and I can't see the messages in kswapd for a long time!! Is this mean that my kswapd doesn'y work or how can I know if the kswapd work well?? Thanks!! -------------- next part -------------- An HTML attachment was scrubbed... URL: From lpapageorgiou at patras.atmel.com Thu Mar 28 04:04:34 2002 From: lpapageorgiou at patras.atmel.com (Lycourgus Papageorgiou) Date: Thu, 28 Mar 2002 11:04:34 +0200 Subject: [uClinux-dev] problem while compile egcs for arm7tdmi In-Reply-To: <604300.1017297035963.JavaMail.postfix@mailsrv11.mail.sohu.com> References: <604300.1017297035963.JavaMail.postfix@mailsrv11.mail.sohu.com> Message-ID: <20020328110434.51a0b4e2.lpapageorgiou@patras.atmel.com> On Thu, 28 Mar 2002 14:30:35 +0800 (CST) wrote: > Hi, everyone > I'm working on pentium PC building a arm7tdmi cross compile > environmen. When I found the arm7tdmi ports of uclinux, I > downloaded all the files. First, I compiled binutils using: > ./configure --target=arm-uclinux --prefix=/opt/a7tdmi and > make; make install, succeeded. Then, I make a symbol link > /usr/src/linux/include point to the patched uclinux/include(I > only made this patch uclinux-patch-export2.patch). Then, I make > egcs using this command: ./configure --target=arm-uclinux > --prefix=/opt/a7tdmi --with-headers=/usr/src/linux/include > --enable-languages=c --disable-threads > > when I type make, it failed, the last few lines of error > message is: > > /tmp/cco7Z7Jw.s: Assembler messages: > /tmp/cco7Z7Jw.s:1971: Error: Subtraction of two symbols in different > sections "strtol" {*UND* section} - ".L44" {*UND* section} at file > address 3892./tmp/cco7Z7Jw.s:1975: Error: Subtraction of two symbols > in different sections "strtoul" {*UND* section} - ".L54" {*UND* > section} at file address 3908. make[5]: *** [vfscanf.o] Error 1 > make[5]: Leaving directory > `/root/arm7tdmi/egcs/arm-uclinux/pic/newlib/libc/stdio' make[4]: *** > [stdio/lib.a] Error 1 make[4]: Leaving directory > `/root/arm7tdmi/egcs/arm-uclinux/pic/newlib/libc' make[3]: *** > [libc/libc.a] Error 2 make[3]: Leaving directory > `/root/arm7tdmi/egcs/arm-uclinux/pic/newlib' make[2]: *** [multi-do] > Error 1 make[2]: Leaving directory > `/root/arm7tdmi/egcs/arm-uclinux/newlib' make[1]: *** [all] Error 2 > make[1]: Leaving directory `/root/arm7tdmi/egcs/arm-uclinux/newlib' > make: *** [all-target-newlib] Error 2 > > I don't know why this happen, could anybody help me? thx in > advance. This message resent by the uclinux-dev at uclinux.org > list server http://www.uClinux.org/ Hi... You probably downloaded the toolchain from http://www.uclinux.org/pub/uClinux/ports/arm7tdmi This toolchain has this problem with vfscanf... There are two solutions.... The first one is to issue the command "make -i" instead of "make" (in order to tell the compiler not to stop at errors...). The problem is that with this option you wouldn't be able to use any function based on vfscanf (ex. fscanf, scanf). The second, is to download another, newer and better toolchain for arm7tdmi ....Take a look at http://www.uclinux.org/pub/uClinux/arm-elf-tools/ and download the newer arm package.. It seems that the directory http://www.uclinux.org/pub/uClinux/ports/arm7tdmi is a bit old... For further information you can checkout the mailing list archives.. It is an issue that confused many of us, in the past, and caused many questions and answers.... Greetings from Patras... Lyc . ______________________________________ Lycourgus Papageorgiou Firmware Engineer Atmel Hellas S.A. Patras Greece Tel: (+3)0944 909452 e-mail: lpapageorgiou at patras.atmel.com ______________________________________ May the OpenSource Force be with you!! ______________________________________ This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From uclinux at schoeldgen.de Thu Mar 28 04:17:19 2002 From: uclinux at schoeldgen.de (Matthias) Date: Thu, 28 Mar 2002 10:17:19 +0100 Subject: [uClinux-dev] Problems when booting from uCLinux References: Message-ID: <3CA2DF9F.921D5298@schoeldgen.de> > "Eric Bosch (Fourtress)" wrote: > > If I try the go command in Minicom to execute the OS image in the > flash ROM, I don't see a login screen. > > What's wrong! > > Thanks, > > Eric > Hi Eric! Do you see any of the boot messages ? If not, maybe there is something wrong with your CONSOLE environment variable. Please be a bit more specific about the board and version of uclinux , if the problem persists. Matthias This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From uclinux at schoeldgen.de Thu Mar 28 04:21:12 2002 From: uclinux at schoeldgen.de (Matthias) Date: Thu, 28 Mar 2002 10:21:12 +0100 Subject: [uClinux-dev] Picogui - compiled and linked.... References: <3CA0A377.CC036AF5@cern.ch> <20020327112758.4af62733.philippe.ney@smartdata.ch> <3CA1ABE6.A4645A10@cern.ch> Message-ID: <3CA2E087.CDF95452@schoeldgen.de> Jody Pearson wrote: > > Hi all, > > I appreciate all of the help. > > I tracked it down to the opening of the framebuffer device, in video/fbdev.c > /* Map it */ > // FB_MEM = mmap(0,fbdev_mapsize = fixinfo.smem_len, > // PROT_READ|PROT_WRITE,MAP_SHARED,fbdev_fd,0); > > FB_MEM = mmap(0,fbdev_mapsize = fixinfo.smem_len, > PROT_READ|PROT_WRITE, 0,fbdev_fd,0); > > I believe (if i am reading this correctly) than the MAP_SHARED is something to do with > the depth ? > > Anyway - I can get picogui to run as an elf binary, under an elf kernel, and display a > black background with a little cursor. > > I intend to get the input (touchscreen) working, then introduce the changes as patches to > picogui. > > Thanks all again for your help > > Jody > > PS - I am also considering a website of screenshots/info/patches and so on, anyone > interested in that (ie - is it worth me bothering ?) > > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ Hi there! I modified fbmem.c for NO_MM and my framebuffer now initializes correctly. Download the modified file from http://www.schoeldgen.de/robot The changes were stolen from the old console code of uClinux 2.0.38 by Kenneth Albanowski and seem to run good. Greetings matthias This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From barros at smar.com.br Thu Mar 28 10:36:44 2002 From: barros at smar.com.br (Marcelo Barros de Almeida) Date: Thu, 28 Mar 2002 07:36:44 -0800 Subject: [uClinux-dev] NET+ARM with SDRAM References: Message-ID: <3CA3388C.B758C262@smar.com.br> Hi Mark, > Is anyone using the Net+40 with SDRAM? I am not using uClinux but NetOS instead. But my OCD file has the following initial registers configuration: run halt load delay slow memwrite long 0xffb00000 0x4004a800 memwrite long 0xffb00008 0x06000000 memwrite long 0xffb00028 0x00000000 memwrite long 0xffb00030 0x00000000 memwrite long 0xffc00000 0x0c8c0000 memwrite long 0xffc00010 0x02000004 memwrite long 0xffc00014 0xffe003B0 memwrite long 0xffc00020 0x0000022c memwrite long 0xffc00024 0xff800070 memwrite long 0xffc00030 0x03000000 memwrite long 0xffc00034 0xff0005d8 memwrite long 0xffc00040 0x02000004 memwrite long 0xffc00044 0xffe003B0 memwrite long 0xffc00040 0x02000005 memwrite long 0xffc00010 0x02000004 memwrite long 0xffc00020 0x0000022d memwrite long 0xffc00030 0x03000000 delay fast reg cpsr 0xd3 I am using a Micron MT48LC2M32B2. My SRAM is the chip select 1 (CS1), 0xFFC00020/24. See hardware manual, page 10-1, to more details. Bye, Marcelo Barros This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From davidm at snapgear.com Thu Mar 28 06:18:05 2002 From: davidm at snapgear.com (David McCullough) Date: Thu, 28 Mar 2002 21:18:05 +1000 Subject: [uClinux-dev] about kswapd In-Reply-To: <002501c1d636$c1a0bf40$2df7748c@SCC>; from sujz@cad.csie.ncku.edu.tw on Thu, Mar 28, 2002 at 04:58:33PM +0800 References: <002501c1d636$c1a0bf40$2df7748c@SCC> Message-ID: <20020328211805.B5170@beast.internal.moreton.com.au> Jivin SuCC lays it down ... > Hi, > I have a memory ran off problem!! > The memory size will reduced slowly when I run some programs.. > > So, I wnat to make sure if the kswapd work well and I add some message in kswapd!! > After I start the system and run some programs, system memory still reduced and I can't see the messages in kswapd for a long time!! > Is this mean that my kswapd doesn'y work or how can I know if the kswapd work well?? How recent is your kernel ? There was a bug in 2.4 for a short while where the cleanup wasn't running, but this was a long time ago :-) If you allocate more then it will run when memory gets low. Maybe wait a little while longer, run a few more programs, and see if the trace comes out :-) You have to be careful with 2.4, it always looks as though there is less memory than you expect. It does free up when it's needed. If you want to check on a program's memory usage, use the size reported by "ps" or look at the individual /proc/XX/* entries, Cheers, Davidm -- David McCullough: Ph: +61 7 3435 2815 http://www.SnapGear.com davidm at snapgear.com Fx: +61 7 3891 3630 825 Stanley St., W'gabba QLD 4102, Oz This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From davidm at snapgear.com Thu Mar 28 06:25:14 2002 From: davidm at snapgear.com (David McCullough) Date: Thu, 28 Mar 2002 21:25:14 +1000 Subject: [uClinux-dev] Re: v850 update patch 20020326 In-Reply-To: ; from miles@lsi.nec.co.jp on Thu, Mar 28, 2002 at 04:57:16PM +0900 References: <3CA019F0.2050402@snapgear.com> <3CA11612.9000505@snapgear.com> <20020328145438.A29463@beast.internal.moreton.com.au> Message-ID: <20020328212514.A5246@beast.internal.moreton.com.au> Jivin Miles Bader lays it down ... > Er, speaking of which, I kinda forgot to update the Makefile... > > If you could: Yep, done -- David McCullough: Ph: +61 7 3435 2815 http://www.SnapGear.com davidm at snapgear.com Fx: +61 7 3891 3630 825 Stanley St., W'gabba QLD 4102, Oz This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From davidm at snapgear.com Thu Mar 28 06:33:07 2002 From: davidm at snapgear.com (David McCullough) Date: Thu, 28 Mar 2002 21:33:07 +1000 Subject: [uClinux-dev] Re: out-of-line BUG In-Reply-To: ; from miles@lsi.nec.co.jp on Thu, Mar 28, 2002 at 05:08:15PM +0900 References: Message-ID: <20020328213307.B5246@beast.internal.moreton.com.au> Jivin Miles Bader lays it down ... > Miles Bader writes: > > Those of you who read the linux-kernel mailing list may have noticed Andrew > > Morton's patch to shrink the kernel by changing calls to the `BUG' macro in > > inline functions in header files to call a normal function `out_of_line_bug'. > > Any comment on this? Sorry, I've been meaning to. We have a kernel config option for m68k that sort of already gives us this by only printing the file/line number for BUG errors. SUPRESS_BUG_MESSAGES or something. This means all the strings are identical and get combined by the compiler (maybe ?). Either way it saves a similar amount of space yet still allows you to find which BUG occured. If this patch goes into the kernel proper we will pick it up when we import it. At that time we would just convert everything else over to use it. I am happy to wait unless other feel we need to preempt this going into the kernel proper ? > If you don't want to apply the more pervasive patch, could you apply the > following, which is the v850-specific part of it (so it's guaranteed not > to conflict down the line; it should be compatible with Andrew's patch > as it uses a weak symbol define `out_of_line_bug')? > > [I'll keep Andrew's patch in my personal tree, but I want to avoid > differences with the repository as much as I can] Applied. Cheers, Davidm -- David McCullough: Ph: +61 7 3435 2815 http://www.SnapGear.com davidm at snapgear.com Fx: +61 7 3891 3630 825 Stanley St., W'gabba QLD 4102, Oz This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From uclinux at schoeldgen.de Thu Mar 28 06:40:02 2002 From: uclinux at schoeldgen.de (Matthias) Date: Thu, 28 Mar 2002 12:40:02 +0100 Subject: [uClinux-dev] M68360: IRQ 68 from ttyS is not replaceable, LSR security check engaged References: <1481.1017139814@www20.rz2.gmx.net> <19098.1017298370@www51.gmx.net> Message-ID: <3CA30111.E4BEF03C@schoeldgen.de> Joachim Schweidler wrote: > > Hi all, > > some months ago I got uclinux 2.0.38 up and running on an 68EN360 Board > (SBS > V360). Now I upgraded to Kernel 2.4.x and the kernel starts and comes till > opening an initial console and then I get the error "M68360_request_irq: > IRQ 68 from ttyS is not replaceable". > > Another error that aperars sometimes is "LSR security check engaged". > > It's a CVS - snapshot from 03/16/2002. > > Doas anyone have the same problem or knows a solution to this problem. > > Many thanks in advance. > Joe > > Here is a log I captured over the serial console (BDM Interface): > > ********************************************************************* > > Linux version 2.4.17-uc0 (root at for02067) (gcc version 2.95.3 20010315 > (release)(Cold2 > > uClinux/MC68360 > Flat model support (C) 1998,1999 Kenneth Albanowski, D. Jeff Dionne > QUICC port done by SED Systems , > based on 2.0.38 port by Lineo Inc. . > On node 0 totalpages: 1023 > zone(0): 0 pages. > zone(1): 1023 pages. > zone(2): 0 pages. > Kernel command line: > CPM UART driver version 0.03 > ttyS00 at irq 0x04 is an SMC > ttyS01 at irq 0x1d is an SCC > ttyS02 at irq 0x1c is an SCC > ttyS03 at irq 0x1b is an SCC > Calibrating delay loop... 4.37 BogoMIPS > Memory available: 2700k/3332k RAM, 394k/1024k ROM (630k kernel code, 549k > data) > kmem_create: Forcing size word alignment - vm_area_struct > kmem_create: Forcing size word alignment - mm_struct > kmem_create: Forcing size word alignment - filp > Dentry-cache hash table entries: 512 (order: 0, 4096 bytes) > Inode-cache hash table entries: 512 (order: 0, 4096 bytes) > kmem_create: Forcing size word alignment - inode_cache > Mount-cache hash table entries: 512 (order: 0, 4096 bytes) > kmem_create: Forcing size word alignment - bdev_cache > kmem_create: Forcing size word alignment - cdev_cache > Buffer-cache hash table entries: 1024 (order: 0, 4096 bytes) > Page-cache hash table entries: 1024 (order: 0, 4096 bytes) > POSIX conformance testing by UNIFIX > Linux NET4.0 for Linux 2.4 > Based upon Swansea University Computer Society NET3.039 > kmem_create: Forcing size word alignment - sock > Initializing RT netlink socket > Starting kswapd > kmem_create: Forcing size word alignment - file lock cache > Serial driver version 5.05c (2001-07-08) with no serial options enabled > ttyS00 at 0x03f8 (irq = 4) is a 16450 > ttyS01 at 0x02f8 (irq = 3) is a 16450 > ttyS02 at 0x03e8 (irq = 4) is a 16450 > kmem_create: Forcing size word alignment - blkdev_requests > block: 64 slots per queue, batch=16 > RAMDISK driver initialized: 16 RAM disks of 4096K size 1024 blocksize > Blkmem copyright 1998,1999 D. Jeff Dionne > Blkmem copyright 1998 Kenneth Albanowski > Blkmem 1 disk images: > 0: C6A08-124207 [VIRTUAL C6A08-124207] (RO) > PPP generic driver version 2.4.1 > NET4: Linux TCP/IP 1.0 for NET4.0 > IP Protocols: ICMP, UDP, TCP > kmem_create: Forcing size word alignment - ip_dst_cache > IP: routing cache hash table of 512 buckets, 4Kbytes > TCP: Hash tables configured (established 512 bind 512) > NET4: Unix domain sockets 1.0/SMP for Linux NET4.0. > VFS: Mounted root (romfs filesystem) readonly. > Freeing unused kernel memory: 0k freed (0x127000 - 0x126000) > M68360_request_irq: IRQ 68 from ttyS is not replaceable > M68360_request_irq: IRQ 68 from ttyS is not replaceable > M68360_request_irq: IRQ 68 from ttyS is not replaceable > M68360_request_irq: IRQ 68 from ttyS is not replaceable > > -- > GMX - Die Kommunikationsplattform im Internet. > http://www.gmx.net > > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ Hello Joachim! It seems, that you included the "Standard serial port support" in the configuration for your kernel. This is only for the 16450/550 found on (IBM-)PC-hardware. Take it out and hopefully your terminal will work. BTW, irq 68 should be the machine dependent bit position of the irq mask bit. I'm not familiar with the 68EN360 , but it seems a bit high , isn't it? Greetings matthias This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From Joachim.Schweidler at gmx.de Thu Mar 28 07:16:34 2002 From: Joachim.Schweidler at gmx.de (Joachim Schweidler) Date: Thu, 28 Mar 2002 13:16:34 +0100 (MET) Subject: [uClinux-dev] M68360: IRQ 68 from ttyS is not replaceable, LSR security check engaged References: <3CA30111.E4BEF03C@schoeldgen.de> Message-ID: <18363.1017317794@www54.gmx.net> Hi Matthias, I thought I need standard serial support for the serial console ... Good tip, thank you very much. But now I got another error: "THRE...rs_interrupt_single(0, 2)...end." Do you know something about this error ? IRQ 68 ist right. This platform hat over 70 IRQs, and this one worked fine under 2.0.38. Joe > Joachim Schweidler wrote: > > > > Hi all, > > > > some months ago I got uclinux 2.0.38 up and running on an 68EN360 Board > > (SBS > > V360). Now I upgraded to Kernel 2.4.x and the kernel starts and comes > till > > opening an initial console and then I get the error "M68360_request_irq: > > IRQ 68 from ttyS is not replaceable". > > > > Another error that aperars sometimes is "LSR security check engaged". > > > > It's a CVS - snapshot from 03/16/2002. > > > > Doas anyone have the same problem or knows a solution to this problem. > > > > Many thanks in advance. > > Joe > > > > Here is a log I captured over the serial console (BDM Interface): > > > > ********************************************************************* > > > > Linux version 2.4.17-uc0 (root at for02067) (gcc version 2.95.3 20010315 > > (release)(Cold2 > > > > uClinux/MC68360 > > Flat model support (C) 1998,1999 Kenneth Albanowski, D. Jeff Dionne > > QUICC port done by SED Systems , > > based on 2.0.38 port by Lineo Inc. . > > On node 0 totalpages: 1023 > > zone(0): 0 pages. > > zone(1): 1023 pages. > > zone(2): 0 pages. > > Kernel command line: > > CPM UART driver version 0.03 > > ttyS00 at irq 0x04 is an SMC > > ttyS01 at irq 0x1d is an SCC > > ttyS02 at irq 0x1c is an SCC > > ttyS03 at irq 0x1b is an SCC > > Calibrating delay loop... 4.37 BogoMIPS > > Memory available: 2700k/3332k RAM, 394k/1024k ROM (630k kernel code, > 549k > > data) > > kmem_create: Forcing size word alignment - vm_area_struct > > kmem_create: Forcing size word alignment - mm_struct > > kmem_create: Forcing size word alignment - filp > > Dentry-cache hash table entries: 512 (order: 0, 4096 bytes) > > Inode-cache hash table entries: 512 (order: 0, 4096 bytes) > > kmem_create: Forcing size word alignment - inode_cache > > Mount-cache hash table entries: 512 (order: 0, 4096 bytes) > > kmem_create: Forcing size word alignment - bdev_cache > > kmem_create: Forcing size word alignment - cdev_cache > > Buffer-cache hash table entries: 1024 (order: 0, 4096 bytes) > > Page-cache hash table entries: 1024 (order: 0, 4096 bytes) > > POSIX conformance testing by UNIFIX > > Linux NET4.0 for Linux 2.4 > > Based upon Swansea University Computer Society NET3.039 > > kmem_create: Forcing size word alignment - sock > > Initializing RT netlink socket > > Starting kswapd > > kmem_create: Forcing size word alignment - file lock cache > > Serial driver version 5.05c (2001-07-08) with no serial options enabled > > ttyS00 at 0x03f8 (irq = 4) is a 16450 > > ttyS01 at 0x02f8 (irq = 3) is a 16450 > > ttyS02 at 0x03e8 (irq = 4) is a 16450 > > kmem_create: Forcing size word alignment - blkdev_requests > > block: 64 slots per queue, batch=16 > > RAMDISK driver initialized: 16 RAM disks of 4096K size 1024 blocksize > > Blkmem copyright 1998,1999 D. Jeff Dionne > > Blkmem copyright 1998 Kenneth Albanowski > > Blkmem 1 disk images: > > 0: C6A08-124207 [VIRTUAL C6A08-124207] (RO) > > PPP generic driver version 2.4.1 > > NET4: Linux TCP/IP 1.0 for NET4.0 > > IP Protocols: ICMP, UDP, TCP > > kmem_create: Forcing size word alignment - ip_dst_cache > > IP: routing cache hash table of 512 buckets, 4Kbytes > > TCP: Hash tables configured (established 512 bind 512) > > NET4: Unix domain sockets 1.0/SMP for Linux NET4.0. > > VFS: Mounted root (romfs filesystem) readonly. > > Freeing unused kernel memory: 0k freed (0x127000 - 0x126000) > > M68360_request_irq: IRQ 68 from ttyS is not replaceable > > M68360_request_irq: IRQ 68 from ttyS is not replaceable > > M68360_request_irq: IRQ 68 from ttyS is not replaceable > > M68360_request_irq: IRQ 68 from ttyS is not replaceable > > > > -- > > GMX - Die Kommunikationsplattform im Internet. > > http://www.gmx.net > > > > This message resent by the uclinux-dev at uclinux.org list server > http://www.uClinux.org/ > > Hello Joachim! > It seems, that you included the "Standard serial port support" in the > configuration for your kernel. This is only for the 16450/550 found on > (IBM-)PC-hardware. Take it out and hopefully your terminal will work. > BTW, irq 68 should be the machine dependent bit position of the irq mask > bit. I'm not familiar with the 68EN360 , but it seems a bit high , isn't > it? > > Greetings > matthias > This message resent by the uclinux-dev at uclinux.org list server > http://www.uClinux.org/ > -- Joachim Schweidler GMX - Die Kommunikationsplattform im Internet. http://www.gmx.net This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From Joachim.Schweidler at gmx.de Thu Mar 28 08:47:07 2002 From: Joachim.Schweidler at gmx.de (Joachim Schweidler) Date: Thu, 28 Mar 2002 14:47:07 +0100 (MET) Subject: [uClinux-dev] M68360: init: fork failed References: <18363.1017317794@www54.gmx.net> Message-ID: <7251.1017323227@www54.gmx.net> Hello, now I got the serial port up and running. Everything seems to be fine but when init starts I get the message: ... Freeing unused kernel memory: 0k freed (0x123000 - 0x122000) init: fork of rc shell failed init: Booting to single user mode. init: fork of single user shell failed init: fork failed ... Any ideas, Joe Log: Linux version 2.4.17-uc0 (root at for02067) (gcc version 2.95.3 20010315 (release)(ColdFire patches - 20010318 from http://fiddes.net/coldfire/)(-msep-data patches)) 2 uClinux/MC68360 Flat model support (C) 1998,1999 Kenneth Albanowski, D. Jeff Dionne QUICC port done by SED Systems , based on 2.0.38 port by Lineo Inc. . On node 0 totalpages: 1023 zone(0): 0 pages. zone(1): 1023 pages. zone(2): 0 pages. Kernel command line: CPM UART driver version 0.03 ttyS00 at irq 0x04 is an SMC .ttyS01 at irq 0x1d is an SCC ttyS02 at irq 0x1c is an SCC ttyS03 at irq 0x1b is an SCC Calibrating delay loop... 4.37 BogoMIPS Memory available: 2716k/3351k RAM, 413k/1024k ROM (611k kernel code, 551k data) kmem_create: Forcing size word alignment - vm_area_struct kmem_create: Forcing size word alignment - mm_struct kmem_create: Forcing size word alignment - filp Dentry-cache hash table entries: 512 (order: 0, 4096 bytes) Inode-cache hash table entries: 512 (order: 0, 4096 bytes) kmem_create: Forcing size word alignment - inode_cache Mount-cache hash table entries: 512 (order: 0, 4096 bytes) kmem_create: Forcing size word alignment - bdev_cache kmem_create: Forcing size word alignment - cdev_cache Buffer-cache hash table entries: 1024 (order: 0, 4096 bytes) Page-cache hash table entries: 1024 (order: 0, 4096 bytes) POSIX conformance testing by UNIFIX Linux NET4.0 for Linux 2.4 Based upon Swansea University Computer Society NET3.039 kmem_create: Forcing size word alignment - sock Initializing RT netlink socket Starting kswapd kmem_create: Forcing size word alignment - file lock cache kmem_create: Forcing size word alignment - blkdev_requests block: 64 slots per queue, batch=16 RAMDISK driver initialized: 16 RAM disks of 4096K size 1024 blocksize Blkmem copyright 1998,1999 D. Jeff Dionne Blkmem copyright 1998 Kenneth Albanowski Blkmem 1 disk images: 0: C2614-11FE13 [VIRTUAL C2614-11FE13] (RO) PPP generic driver version 2.4.1 NET4: Linux TCP/IP 1.0 for NET4.0 IP Protocols: ICMP, UDP, TCP kmem_create: Forcing size word alignment - ip_dst_cache IP: routing cache hash table of 512 buckets, 4Kbytes TCP: Hash tables configured (established 512 bind 512) NET4: Unix domain sockets 1.0/SMP for Linux NET4.0. VFS: Mounted root (romfs filesystem) readonly. Freeing unused kernel memory: 0k freed (0x123000 - 0x122000) init: fork of rc shell failed init: Booting to single user mode. init: fork of single user shell failed init: fork failed init: fork failed init: fork failed init: fork failed -- GMX - Die Kommunikationsplattform im Internet. http://www.gmx.net This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From Joachim.Schweidler at gmx.de Thu Mar 28 08:44:26 2002 From: Joachim.Schweidler at gmx.de (Joachim Schweidler) Date: Thu, 28 Mar 2002 14:44:26 +0100 (MET) Subject: [uClinux-dev] M68360: init: fork failed References: <18363.1017317794@www54.gmx.net> Message-ID: <2394.1017323066@www54.gmx.net> Sorry, I forgot to disable debugging messages. Joe > Hi Matthias, > > I thought I need standard serial support for the serial console ... Good > tip, thank you very much. > > But now I got another error: "THRE...rs_interrupt_single(0, 2)...end." > > Do you know something about this error ? > > IRQ 68 ist right. This platform hat over 70 IRQs, and this one worked fine > under 2.0.38. > > Joe > > > Joachim Schweidler wrote: > > > > > > Hi all, > > > > > > some months ago I got uclinux 2.0.38 up and running on an 68EN360 > Board > > > (SBS > > > V360). Now I upgraded to Kernel 2.4.x and the kernel starts and comes > > till > > > opening an initial console and then I get the error > "M68360_request_irq: > > > IRQ 68 from ttyS is not replaceable". > > > > > > Another error that aperars sometimes is "LSR security check engaged". > > > > > > It's a CVS - snapshot from 03/16/2002. > > > > > > Doas anyone have the same problem or knows a solution to this problem. > > > > > > Many thanks in advance. > > > Joe > > > > > > Here is a log I captured over the serial console (BDM Interface): > > > > > > ********************************************************************* > > > > > > Linux version 2.4.17-uc0 (root at for02067) (gcc version 2.95.3 20010315 > > > (release)(Cold2 > > > > > > uClinux/MC68360 > > > Flat model support (C) 1998,1999 Kenneth Albanowski, D. Jeff Dionne > > > QUICC port done by SED Systems , > > > based on 2.0.38 port by Lineo Inc. . > > > On node 0 totalpages: 1023 > > > zone(0): 0 pages. > > > zone(1): 1023 pages. > > > zone(2): 0 pages. > > > Kernel command line: > > > CPM UART driver version 0.03 > > > ttyS00 at irq 0x04 is an SMC > > > ttyS01 at irq 0x1d is an SCC > > > ttyS02 at irq 0x1c is an SCC > > > ttyS03 at irq 0x1b is an SCC > > > Calibrating delay loop... 4.37 BogoMIPS > > > Memory available: 2700k/3332k RAM, 394k/1024k ROM (630k kernel code, > > 549k > > > data) > > > kmem_create: Forcing size word alignment - vm_area_struct > > > kmem_create: Forcing size word alignment - mm_struct > > > kmem_create: Forcing size word alignment - filp > > > Dentry-cache hash table entries: 512 (order: 0, 4096 bytes) > > > Inode-cache hash table entries: 512 (order: 0, 4096 bytes) > > > kmem_create: Forcing size word alignment - inode_cache > > > Mount-cache hash table entries: 512 (order: 0, 4096 bytes) > > > kmem_create: Forcing size word alignment - bdev_cache > > > kmem_create: Forcing size word alignment - cdev_cache > > > Buffer-cache hash table entries: 1024 (order: 0, 4096 bytes) > > > Page-cache hash table entries: 1024 (order: 0, 4096 bytes) > > > POSIX conformance testing by UNIFIX > > > Linux NET4.0 for Linux 2.4 > > > Based upon Swansea University Computer Society NET3.039 > > > kmem_create: Forcing size word alignment - sock > > > Initializing RT netlink socket > > > Starting kswapd > > > kmem_create: Forcing size word alignment - file lock cache > > > Serial driver version 5.05c (2001-07-08) with no serial options > enabled > > > ttyS00 at 0x03f8 (irq = 4) is a 16450 > > > ttyS01 at 0x02f8 (irq = 3) is a 16450 > > > ttyS02 at 0x03e8 (irq = 4) is a 16450 > > > kmem_create: Forcing size word alignment - blkdev_requests > > > block: 64 slots per queue, batch=16 > > > RAMDISK driver initialized: 16 RAM disks of 4096K size 1024 blocksize > > > Blkmem copyright 1998,1999 D. Jeff Dionne > > > Blkmem copyright 1998 Kenneth Albanowski > > > Blkmem 1 disk images: > > > 0: C6A08-124207 [VIRTUAL C6A08-124207] (RO) > > > PPP generic driver version 2.4.1 > > > NET4: Linux TCP/IP 1.0 for NET4.0 > > > IP Protocols: ICMP, UDP, TCP > > > kmem_create: Forcing size word alignment - ip_dst_cache > > > IP: routing cache hash table of 512 buckets, 4Kbytes > > > TCP: Hash tables configured (established 512 bind 512) > > > NET4: Unix domain sockets 1.0/SMP for Linux NET4.0. > > > VFS: Mounted root (romfs filesystem) readonly. > > > Freeing unused kernel memory: 0k freed (0x127000 - 0x126000) > > > M68360_request_irq: IRQ 68 from ttyS is not replaceable > > > M68360_request_irq: IRQ 68 from ttyS is not replaceable > > > M68360_request_irq: IRQ 68 from ttyS is not replaceable > > > M68360_request_irq: IRQ 68 from ttyS is not replaceable > > > > > > -- > > > GMX - Die Kommunikationsplattform im Internet. > > > http://www.gmx.net > > > > > > This message resent by the uclinux-dev at uclinux.org list server > > http://www.uClinux.org/ > > > > Hello Joachim! > > It seems, that you included the "Standard serial port support" in the > > configuration for your kernel. This is only for the 16450/550 found on > > (IBM-)PC-hardware. Take it out and hopefully your terminal will work. > > BTW, irq 68 should be the machine dependent bit position of the irq mask > > bit. I'm not familiar with the 68EN360 , but it seems a bit high , isn't > > it? > > > > Greetings > > matthias > > This message resent by the uclinux-dev at uclinux.org list server > > http://www.uClinux.org/ > > > > -- > Joachim Schweidler > > GMX - Die Kommunikationsplattform im Internet. > http://www.gmx.net > > This message resent by the uclinux-dev at uclinux.org list server > http://www.uClinux.org/ > -- GMX - Die Kommunikationsplattform im Internet. http://www.gmx.net This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From hamilton at SEDSystems.ca Thu Mar 28 09:48:23 2002 From: hamilton at SEDSystems.ca (Kendrick Hamilton) Date: Thu, 28 Mar 2002 08:48:23 -0600 (CST) Subject: [uClinux-dev] M68360: IRQ 68 from ttyS is not replaceable, LSR security check engaged In-Reply-To: <18363.1017317794@www54.gmx.net> Message-ID: There are 256 interrupts on the 68360 and Irq 68 is correct for SMC2. Send me your configuration file. Kendrick Hamilton E.I.T. SED Systems, a division of Calian Ltd. 18 Innovation Blvd. PO Box 1464 Saskatoon, Saskatchewan Canada S7N 3R1 Hamilton at sedsystems.ca Tel: (306) 933-1453 Fax: (306) 933-1486 On Thu, 28 Mar 2002, Joachim Schweidler wrote: > Hi Matthias, > > I thought I need standard serial support for the serial console ... Good > tip, thank you very much. > > But now I got another error: "THRE...rs_interrupt_single(0, 2)...end." > > Do you know something about this error ? > > IRQ 68 ist right. This platform hat over 70 IRQs, and this one worked fine > under 2.0.38. > > Joe > > > Joachim Schweidler wrote: > > > > > > Hi all, > > > > > > some months ago I got uclinux 2.0.38 up and running on an 68EN360 Board > > > (SBS > > > V360). Now I upgraded to Kernel 2.4.x and the kernel starts and comes > > till > > > opening an initial console and then I get the error "M68360_request_irq: > > > IRQ 68 from ttyS is not replaceable". > > > > > > Another error that aperars sometimes is "LSR security check engaged". > > > > > > It's a CVS - snapshot from 03/16/2002. > > > > > > Doas anyone have the same problem or knows a solution to this problem. > > > > > > Many thanks in advance. > > > Joe > > > > > > Here is a log I captured over the serial console (BDM Interface): > > > > > > ********************************************************************* > > > > > > Linux version 2.4.17-uc0 (root at for02067) (gcc version 2.95.3 20010315 > > > (release)(Cold2 > > > > > > uClinux/MC68360 > > > Flat model support (C) 1998,1999 Kenneth Albanowski, D. Jeff Dionne > > > QUICC port done by SED Systems , > > > based on 2.0.38 port by Lineo Inc. . > > > On node 0 totalpages: 1023 > > > zone(0): 0 pages. > > > zone(1): 1023 pages. > > > zone(2): 0 pages. > > > Kernel command line: > > > CPM UART driver version 0.03 > > > ttyS00 at irq 0x04 is an SMC > > > ttyS01 at irq 0x1d is an SCC > > > ttyS02 at irq 0x1c is an SCC > > > ttyS03 at irq 0x1b is an SCC > > > Calibrating delay loop... 4.37 BogoMIPS > > > Memory available: 2700k/3332k RAM, 394k/1024k ROM (630k kernel code, > > 549k > > > data) > > > kmem_create: Forcing size word alignment - vm_area_struct > > > kmem_create: Forcing size word alignment - mm_struct > > > kmem_create: Forcing size word alignment - filp > > > Dentry-cache hash table entries: 512 (order: 0, 4096 bytes) > > > Inode-cache hash table entries: 512 (order: 0, 4096 bytes) > > > kmem_create: Forcing size word alignment - inode_cache > > > Mount-cache hash table entries: 512 (order: 0, 4096 bytes) > > > kmem_create: Forcing size word alignment - bdev_cache > > > kmem_create: Forcing size word alignment - cdev_cache > > > Buffer-cache hash table entries: 1024 (order: 0, 4096 bytes) > > > Page-cache hash table entries: 1024 (order: 0, 4096 bytes) > > > POSIX conformance testing by UNIFIX > > > Linux NET4.0 for Linux 2.4 > > > Based upon Swansea University Computer Society NET3.039 > > > kmem_create: Forcing size word alignment - sock > > > Initializing RT netlink socket > > > Starting kswapd > > > kmem_create: Forcing size word alignment - file lock cache > > > Serial driver version 5.05c (2001-07-08) with no serial options enabled > > > ttyS00 at 0x03f8 (irq = 4) is a 16450 > > > ttyS01 at 0x02f8 (irq = 3) is a 16450 > > > ttyS02 at 0x03e8 (irq = 4) is a 16450 > > > kmem_create: Forcing size word alignment - blkdev_requests > > > block: 64 slots per queue, batch=16 > > > RAMDISK driver initialized: 16 RAM disks of 4096K size 1024 blocksize > > > Blkmem copyright 1998,1999 D. Jeff Dionne > > > Blkmem copyright 1998 Kenneth Albanowski > > > Blkmem 1 disk images: > > > 0: C6A08-124207 [VIRTUAL C6A08-124207] (RO) > > > PPP generic driver version 2.4.1 > > > NET4: Linux TCP/IP 1.0 for NET4.0 > > > IP Protocols: ICMP, UDP, TCP > > > kmem_create: Forcing size word alignment - ip_dst_cache > > > IP: routing cache hash table of 512 buckets, 4Kbytes > > > TCP: Hash tables configured (established 512 bind 512) > > > NET4: Unix domain sockets 1.0/SMP for Linux NET4.0. > > > VFS: Mounted root (romfs filesystem) readonly. > > > Freeing unused kernel memory: 0k freed (0x127000 - 0x126000) > > > M68360_request_irq: IRQ 68 from ttyS is not replaceable > > > M68360_request_irq: IRQ 68 from ttyS is not replaceable > > > M68360_request_irq: IRQ 68 from ttyS is not replaceable > > > M68360_request_irq: IRQ 68 from ttyS is not replaceable > > > > > > -- > > > GMX - Die Kommunikationsplattform im Internet. > > > http://www.gmx.net > > > > > > This message resent by the uclinux-dev at uclinux.org list server > > http://www.uClinux.org/ > > > > Hello Joachim! > > It seems, that you included the "Standard serial port support" in the > > configuration for your kernel. This is only for the 16450/550 found on > > (IBM-)PC-hardware. Take it out and hopefully your terminal will work. > > BTW, irq 68 should be the machine dependent bit position of the irq mask > > bit. I'm not familiar with the 68EN360 , but it seems a bit high , isn't > > it? > > > > Greetings > > matthias > > This message resent by the uclinux-dev at uclinux.org list server > > http://www.uClinux.org/ > > > > -- > Joachim Schweidler > > GMX - Die Kommunikationsplattform im Internet. > http://www.gmx.net > > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From Joachim.Schweidler at gmx.de Thu Mar 28 09:57:23 2002 From: Joachim.Schweidler at gmx.de (Joachim Schweidler) Date: Thu, 28 Mar 2002 15:57:23 +0100 (MET) Subject: [uClinux-dev] M68360: IRQ 68 from ttyS is not replaceable, LSR security check engaged References: Message-ID: <14503.1017327443@www54.gmx.net> which configuration file do you mean ? > There are 256 interrupts on the 68360 and Irq 68 is correct for SMC2. Send > me your configuration file. > > Kendrick Hamilton E.I.T. > SED Systems, a division of Calian Ltd. > 18 Innovation Blvd. > PO Box 1464 > Saskatoon, Saskatchewan > Canada > S7N 3R1 > > Hamilton at sedsystems.ca > Tel: (306) 933-1453 > Fax: (306) 933-1486 > > On Thu, 28 Mar 2002, Joachim Schweidler wrote: > > > Hi Matthias, > > > > I thought I need standard serial support for the serial console ... Good > > tip, thank you very much. > > > > But now I got another error: "THRE...rs_interrupt_single(0, 2)...end." > > > > Do you know something about this error ? > > > > IRQ 68 ist right. This platform hat over 70 IRQs, and this one worked > fine > > under 2.0.38. > > > > Joe > > > > > Joachim Schweidler wrote: > > > > > > > > Hi all, > > > > > > > > some months ago I got uclinux 2.0.38 up and running on an 68EN360 > Board > > > > (SBS > > > > V360). Now I upgraded to Kernel 2.4.x and the kernel starts and > comes > > > till > > > > opening an initial console and then I get the error > "M68360_request_irq: > > > > IRQ 68 from ttyS is not replaceable". > > > > > > > > Another error that aperars sometimes is "LSR security check > engaged". > > > > > > > > It's a CVS - snapshot from 03/16/2002. > > > > > > > > Doas anyone have the same problem or knows a solution to this > problem. > > > > > > > > Many thanks in advance. > > > > Joe > > > > > > > > Here is a log I captured over the serial console (BDM Interface): > > > > > > > > > ********************************************************************* > > > > > > > > Linux version 2.4.17-uc0 (root at for02067) (gcc version 2.95.3 > 20010315 > > > > (release)(Cold2 > > > > > > > > uClinux/MC68360 > > > > Flat model support (C) 1998,1999 Kenneth Albanowski, D. Jeff Dionne > > > > QUICC port done by SED Systems , > > > > based on 2.0.38 port by Lineo Inc. . > > > > On node 0 totalpages: 1023 > > > > zone(0): 0 pages. > > > > zone(1): 1023 pages. > > > > zone(2): 0 pages. > > > > Kernel command line: > > > > CPM UART driver version 0.03 > > > > ttyS00 at irq 0x04 is an SMC > > > > ttyS01 at irq 0x1d is an SCC > > > > ttyS02 at irq 0x1c is an SCC > > > > ttyS03 at irq 0x1b is an SCC > > > > Calibrating delay loop... 4.37 BogoMIPS > > > > Memory available: 2700k/3332k RAM, 394k/1024k ROM (630k kernel code, > > > 549k > > > > data) > > > > kmem_create: Forcing size word alignment - vm_area_struct > > > > kmem_create: Forcing size word alignment - mm_struct > > > > kmem_create: Forcing size word alignment - filp > > > > Dentry-cache hash table entries: 512 (order: 0, 4096 bytes) > > > > Inode-cache hash table entries: 512 (order: 0, 4096 bytes) > > > > kmem_create: Forcing size word alignment - inode_cache > > > > Mount-cache hash table entries: 512 (order: 0, 4096 bytes) > > > > kmem_create: Forcing size word alignment - bdev_cache > > > > kmem_create: Forcing size word alignment - cdev_cache > > > > Buffer-cache hash table entries: 1024 (order: 0, 4096 bytes) > > > > Page-cache hash table entries: 1024 (order: 0, 4096 bytes) > > > > POSIX conformance testing by UNIFIX > > > > Linux NET4.0 for Linux 2.4 > > > > Based upon Swansea University Computer Society NET3.039 > > > > kmem_create: Forcing size word alignment - sock > > > > Initializing RT netlink socket > > > > Starting kswapd > > > > kmem_create: Forcing size word alignment - file lock cache > > > > Serial driver version 5.05c (2001-07-08) with no serial options > enabled > > > > ttyS00 at 0x03f8 (irq = 4) is a 16450 > > > > ttyS01 at 0x02f8 (irq = 3) is a 16450 > > > > ttyS02 at 0x03e8 (irq = 4) is a 16450 > > > > kmem_create: Forcing size word alignment - blkdev_requests > > > > block: 64 slots per queue, batch=16 > > > > RAMDISK driver initialized: 16 RAM disks of 4096K size 1024 > blocksize > > > > Blkmem copyright 1998,1999 D. Jeff Dionne > > > > Blkmem copyright 1998 Kenneth Albanowski > > > > Blkmem 1 disk images: > > > > 0: C6A08-124207 [VIRTUAL C6A08-124207] (RO) > > > > PPP generic driver version 2.4.1 > > > > NET4: Linux TCP/IP 1.0 for NET4.0 > > > > IP Protocols: ICMP, UDP, TCP > > > > kmem_create: Forcing size word alignment - ip_dst_cache > > > > IP: routing cache hash table of 512 buckets, 4Kbytes > > > > TCP: Hash tables configured (established 512 bind 512) > > > > NET4: Unix domain sockets 1.0/SMP for Linux NET4.0. > > > > VFS: Mounted root (romfs filesystem) readonly. > > > > Freeing unused kernel memory: 0k freed (0x127000 - 0x126000) > > > > M68360_request_irq: IRQ 68 from ttyS is not replaceable > > > > M68360_request_irq: IRQ 68 from ttyS is not replaceable > > > > M68360_request_irq: IRQ 68 from ttyS is not replaceable > > > > M68360_request_irq: IRQ 68 from ttyS is not replaceable > > > > > > > > -- > > > > GMX - Die Kommunikationsplattform im Internet. > > > > http://www.gmx.net > > > > > > > > This message resent by the uclinux-dev at uclinux.org list server > > > http://www.uClinux.org/ > > > > > > Hello Joachim! > > > It seems, that you included the "Standard serial port support" in the > > > configuration for your kernel. This is only for the 16450/550 found on > > > (IBM-)PC-hardware. Take it out and hopefully your terminal will work. > > > BTW, irq 68 should be the machine dependent bit position of the irq > mask > > > bit. I'm not familiar with the 68EN360 , but it seems a bit high , > isn't > > > it? > > > > > > Greetings > > > matthias > > > This message resent by the uclinux-dev at uclinux.org list server > > > http://www.uClinux.org/ > > > > > > > -- > > Joachim Schweidler > > > > GMX - Die Kommunikationsplattform im Internet. > > http://www.gmx.net > > > > This message resent by the uclinux-dev at uclinux.org list server > http://www.uClinux.org/ > > > > This message resent by the uclinux-dev at uclinux.org list server > http://www.uClinux.org/ > -- Joachim Schweidler GMX - Die Kommunikationsplattform im Internet. http://www.gmx.net This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From tahir at snom.de Thu Mar 28 10:49:08 2002 From: tahir at snom.de (Muhammad Usman Tahir) Date: Thu, 28 Mar 2002 16:49:08 +0100 Subject: [uClinux-dev] User Application: Permission denied. Message-ID: <010801c1d670$19399590$2200a8c0@flamenco> Hi, I've an MCF 5206 system with 8MB RAM and having 2.0.38 uClinux kernel running fine with basic apps like shell, dhcpcd etc. I've added one of my applications in '/user/my_app' and it has been successfully compiled and linked to generate a binary. I then modified 'rc' script to run this app by do_rc from init. On execution it says : Command: my_app my_app: Permission denied. and then continues with the next item in 'rc'. I have checked all the permissions for the binary of my_app in the 'romfs/bin' and have even forced it to be executable but the error message remains the same. I tried to copy the file attribs from the running dhcpcd app but still ended up with the same result. I should add here that I bypassed the generation of elf file while compilation and linked to get a flat binary directly with -elf2flt option in m68k-elf-gcc. (I had lots of problems with elf2flt and searching through the archives I found this work around to what seems to be a common problem). I think this should not have caused any problem as long as I don't need an elf file for my_app, which I don't. Any idea how I can get my_app really working without this message? Shall I try some other method to start my_app? Thanks in advance for the all the help. Regards, Usman. This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From julienb at actimage.fr Thu Mar 28 11:04:43 2002 From: julienb at actimage.fr (Julien Boibessot) Date: Thu, 28 Mar 2002 17:04:43 +0100 Subject: [uClinux-dev] 2nd serial port on M5272C3 devt board References: <002501c1d636$c1a0bf40$2df7748c@SCC> Message-ID: <022001c1d672$46377560$8947ec95@bruker.fr> Hi all !!! does anyone can confirm that uClinux is handling the 2nd serial port on M5272C3 devt board ? if yes how should I configure the kernel to get access to this port with "/dev/ttyS1" (for example) ?? thanks for any answer Julien Boibessot -------------- next part -------------- An HTML attachment was scrubbed... URL: From fskivee at skynet.be Thu Mar 28 11:40:51 2002 From: fskivee at skynet.be (=?iso-8859-1?Q?Fabian_Skiv=E9e?=) Date: Thu, 28 Mar 2002 17:40:51 +0100 Subject: [uClinux-dev] Init chip registers : Palm m500 specific Message-ID: <001f01c1d677$53187a90$418688d9@galia> Hello, I made a Real-time OS and it works on xcopilot. I find the initialisation code for the Palm Vx. Now, I try to port it to my palm m500. I'm looking for the exact intialisation values of the chip registers for this device. I take a look in the source from the alliance web site but this code is not present. Examples of registers : SCR, PLLCR, PLLFSR, PCTLR, PBSEL, PESEL, DRAMMC, .... Where can I find this values ? Anyone can help me ? Thanks, Fabian Skiv?e --------------------------------------------------------------------------------- Fabian Skiv?e [a] Rue Petite Spinette 8, 4630 Soumagne, Belgium [t] + 32 (0) 4 377 23 76 [g] + 32 (0) 498 62 47 03 [e] fskivee at skynet.be -------------- next part -------------- An HTML attachment was scrubbed... URL: From hamilton at SEDSystems.ca Thu Mar 28 12:02:20 2002 From: hamilton at SEDSystems.ca (Kendrick Hamilton) Date: Thu, 28 Mar 2002 11:02:20 -0600 (CST) Subject: [uClinux-dev] M68360: IRQ 68 from ttyS is not replaceable, LSR security check engaged In-Reply-To: <14503.1017327443@www54.gmx.net> Message-ID: The config file generated by make config/menuconfig/xconfig Kendrick Hamilton E.I.T. SED Systems, a division of Calian Ltd. 18 Innovation Blvd. PO Box 1464 Saskatoon, Saskatchewan Canada S7N 3R1 Hamilton at sedsystems.ca Tel: (306) 933-1453 Fax: (306) 933-1486 On Thu, 28 Mar 2002, Joachim Schweidler wrote: > > which configuration file do you mean ? > > > There are 256 interrupts on the 68360 and Irq 68 is correct for SMC2. Send > > me your configuration file. > > > > Kendrick Hamilton E.I.T. > > SED Systems, a division of Calian Ltd. > > 18 Innovation Blvd. > > PO Box 1464 > > Saskatoon, Saskatchewan > > Canada > > S7N 3R1 > > > > Hamilton at sedsystems.ca > > Tel: (306) 933-1453 > > Fax: (306) 933-1486 > > > > On Thu, 28 Mar 2002, Joachim Schweidler wrote: > > > > > Hi Matthias, > > > > > > I thought I need standard serial support for the serial console ... Good > > > tip, thank you very much. > > > > > > But now I got another error: "THRE...rs_interrupt_single(0, 2)...end." > > > > > > Do you know something about this error ? > > > > > > IRQ 68 ist right. This platform hat over 70 IRQs, and this one worked > > fine > > > under 2.0.38. > > > > > > Joe > > > > > > > Joachim Schweidler wrote: > > > > > > > > > > Hi all, > > > > > > > > > > some months ago I got uclinux 2.0.38 up and running on an 68EN360 > > Board > > > > > (SBS > > > > > V360). Now I upgraded to Kernel 2.4.x and the kernel starts and > > comes > > > > till > > > > > opening an initial console and then I get the error > > "M68360_request_irq: > > > > > IRQ 68 from ttyS is not replaceable". > > > > > > > > > > Another error that aperars sometimes is "LSR security check > > engaged". > > > > > > > > > > It's a CVS - snapshot from 03/16/2002. > > > > > > > > > > Doas anyone have the same problem or knows a solution to this > > problem. > > > > > > > > > > Many thanks in advance. > > > > > Joe > > > > > > > > > > Here is a log I captured over the serial console (BDM Interface): > > > > > > > > > > > > ********************************************************************* > > > > > > > > > > Linux version 2.4.17-uc0 (root at for02067) (gcc version 2.95.3 > > 20010315 > > > > > (release)(Cold2 > > > > > > > > > > uClinux/MC68360 > > > > > Flat model support (C) 1998,1999 Kenneth Albanowski, D. Jeff Dionne > > > > > QUICC port done by SED Systems , > > > > > based on 2.0.38 port by Lineo Inc. . > > > > > On node 0 totalpages: 1023 > > > > > zone(0): 0 pages. > > > > > zone(1): 1023 pages. > > > > > zone(2): 0 pages. > > > > > Kernel command line: > > > > > CPM UART driver version 0.03 > > > > > ttyS00 at irq 0x04 is an SMC > > > > > ttyS01 at irq 0x1d is an SCC > > > > > ttyS02 at irq 0x1c is an SCC > > > > > ttyS03 at irq 0x1b is an SCC > > > > > Calibrating delay loop... 4.37 BogoMIPS > > > > > Memory available: 2700k/3332k RAM, 394k/1024k ROM (630k kernel code, > > > > 549k > > > > > data) > > > > > kmem_create: Forcing size word alignment - vm_area_struct > > > > > kmem_create: Forcing size word alignment - mm_struct > > > > > kmem_create: Forcing size word alignment - filp > > > > > Dentry-cache hash table entries: 512 (order: 0, 4096 bytes) > > > > > Inode-cache hash table entries: 512 (order: 0, 4096 bytes) > > > > > kmem_create: Forcing size word alignment - inode_cache > > > > > Mount-cache hash table entries: 512 (order: 0, 4096 bytes) > > > > > kmem_create: Forcing size word alignment - bdev_cache > > > > > kmem_create: Forcing size word alignment - cdev_cache > > > > > Buffer-cache hash table entries: 1024 (order: 0, 4096 bytes) > > > > > Page-cache hash table entries: 1024 (order: 0, 4096 bytes) > > > > > POSIX conformance testing by UNIFIX > > > > > Linux NET4.0 for Linux 2.4 > > > > > Based upon Swansea University Computer Society NET3.039 > > > > > kmem_create: Forcing size word alignment - sock > > > > > Initializing RT netlink socket > > > > > Starting kswapd > > > > > kmem_create: Forcing size word alignment - file lock cache > > > > > Serial driver version 5.05c (2001-07-08) with no serial options > > enabled > > > > > ttyS00 at 0x03f8 (irq = 4) is a 16450 > > > > > ttyS01 at 0x02f8 (irq = 3) is a 16450 > > > > > ttyS02 at 0x03e8 (irq = 4) is a 16450 > > > > > kmem_create: Forcing size word alignment - blkdev_requests > > > > > block: 64 slots per queue, batch=16 > > > > > RAMDISK driver initialized: 16 RAM disks of 4096K size 1024 > > blocksize > > > > > Blkmem copyright 1998,1999 D. Jeff Dionne > > > > > Blkmem copyright 1998 Kenneth Albanowski > > > > > Blkmem 1 disk images: > > > > > 0: C6A08-124207 [VIRTUAL C6A08-124207] (RO) > > > > > PPP generic driver version 2.4.1 > > > > > NET4: Linux TCP/IP 1.0 for NET4.0 > > > > > IP Protocols: ICMP, UDP, TCP > > > > > kmem_create: Forcing size word alignment - ip_dst_cache > > > > > IP: routing cache hash table of 512 buckets, 4Kbytes > > > > > TCP: Hash tables configured (established 512 bind 512) > > > > > NET4: Unix domain sockets 1.0/SMP for Linux NET4.0. > > > > > VFS: Mounted root (romfs filesystem) readonly. > > > > > Freeing unused kernel memory: 0k freed (0x127000 - 0x126000) > > > > > M68360_request_irq: IRQ 68 from ttyS is not replaceable > > > > > M68360_request_irq: IRQ 68 from ttyS is not replaceable > > > > > M68360_request_irq: IRQ 68 from ttyS is not replaceable > > > > > M68360_request_irq: IRQ 68 from ttyS is not replaceable > > > > > > > > > > -- > > > > > GMX - Die Kommunikationsplattform im Internet. > > > > > http://www.gmx.net > > > > > > > > > > This message resent by the uclinux-dev at uclinux.org list server > > > > http://www.uClinux.org/ > > > > > > > > Hello Joachim! > > > > It seems, that you included the "Standard serial port support" in the > > > > configuration for your kernel. This is only for the 16450/550 found on > > > > (IBM-)PC-hardware. Take it out and hopefully your terminal will work. > > > > BTW, irq 68 should be the machine dependent bit position of the irq > > mask > > > > bit. I'm not familiar with the 68EN360 , but it seems a bit high , > > isn't > > > > it? > > > > > > > > Greetings > > > > matthias > > > > This message resent by the uclinux-dev at uclinux.org list server > > > > http://www.uClinux.org/ > > > > > > > > > > -- > > > Joachim Schweidler > > > > > > GMX - Die Kommunikationsplattform im Internet. > > > http://www.gmx.net > > > > > > This message resent by the uclinux-dev at uclinux.org list server > > http://www.uClinux.org/ > > > > > > > This message resent by the uclinux-dev at uclinux.org list server > > http://www.uClinux.org/ > > > > -- > Joachim Schweidler > > GMX - Die Kommunikationsplattform im Internet. > http://www.gmx.net > > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From Joachim.Schweidler at gmx.de Thu Mar 28 13:20:43 2002 From: Joachim.Schweidler at gmx.de (Joachim Schweidler) Date: Thu, 28 Mar 2002 19:20:43 +0100 (MET) Subject: [uClinux-dev] M68360: IRQ 68 from ttyS is not replaceable, LSR security check engaged References: Message-ID: <13034.1017339643@www29.gmx.net> .config: # # Automatically generated make config: don't edit # CONFIG_UCLINUX=y CONFIG_UID16=y CONFIG_RWSEM_GENERIC_SPINLOCK=y # CONFIG_RWSEM_XCHGADD_ALGORITHM is not set # # Code maturity level options # CONFIG_EXPERIMENTAL=y # # Loadable module support # # CONFIG_MODULES is not set # # Processor type and features # # CONFIG_M68000 is not set # CONFIG_M68EN302 is not set # CONFIG_M68328 is not set # CONFIG_M68EZ328 is not set # CONFIG_M68VZ328 is not set # CONFIG_M68332 is not set CONFIG_M68360=y # CONFIG_M5204 is not set # CONFIG_M5206 is not set # CONFIG_M5206e is not set # CONFIG_M5272 is not set # CONFIG_M5307 is not set # CONFIG_M5407 is not set CONFIG_360QUICC=y # CONFIG_CLOCK_AUTO is not set # CONFIG_CLOCK_16MHz is not set # CONFIG_CLOCK_20MHz is not set CONFIG_CLOCK_25MHz=y # CONFIG_CLOCK_33MHz is not set # CONFIG_CLOCK_40MHz is not set # CONFIG_CLOCK_45MHz is not set # CONFIG_CLOCK_48MHz is not set # CONFIG_CLOCK_50MHz is not set # CONFIG_CLOCK_54MHz is not set # CONFIG_CLOCK_60MHz is not set # CONFIG_CLOCK_66MHz is not set # # Platform # # CONFIG_UCQUICC is not set CONFIG_SBSV360=y # CONFIG_RAMAUTO is not set CONFIG_RAM4MB=y # CONFIG_RAM8MB is not set # CONFIG_RAM16MB is not set # CONFIG_RAM32MB is not set # CONFIG_AUTOBIT is not set # CONFIG_RAM8BIT is not set # CONFIG_RAM16BIT is not set CONFIG_RAM32bit=y CONFIG_RAMKERNEL=y # CONFIG_ROMKERNEL is not set # CONFIG_HIMEMKERNEL is not set # # General setup # CONFIG_NET=y # CONFIG_VISWS is not set # CONFIG_PCI is not set # CONFIG_MCA is not set # CONFIG_HOTPLUG is not set # CONFIG_PCMCIA is not set # CONFIG_SYSVIPC is not set # CONFIG_BSD_PROCESS_ACCT is not set # CONFIG_SYSCTL is not set # CONFIG_KCORE_ELF is not set CONFIG_KCORE_AOUT=y CONFIG_BINFMT_FLAT=y # CONFIG_BINFMT_ZFLAT is not set # CONFIG_BINFMT_AOUT is not set # CONFIG_BINFMT_ELF is not set # CONFIG_BINFMT_MISC is not set # CONFIG_PM is not set # # Memory Technology Devices (MTD) # # CONFIG_MTD is not set # # Parallel port support # # CONFIG_PARPORT is not set # # Plug and Play configuration # # CONFIG_PNP is not set # # Block devices # # CONFIG_BLK_DEV_FD is not set # CONFIG_BLK_DEV_XD is not set # CONFIG_BLK_DEV_LOOP is not set # CONFIG_BLK_DEV_NBD is not set CONFIG_BLK_DEV_RAM=y CONFIG_BLK_DEV_RAM_SIZE=4096 # CONFIG_BLK_DEV_INITRD is not set CONFIG_BLK_DEV_BLKMEM=y CONFIG_NOFLASH=y # CONFIG_AMDFLASH is not set # CONFIG_INTELFLASH is not set # # Networking options # CONFIG_PACKET=y # CONFIG_PACKET_MMAP is not set # CONFIG_NETLINK_DEV is not set # CONFIG_NETFILTER is not set # CONFIG_FILTER is not set CONFIG_UNIX=y CONFIG_INET=y # CONFIG_IP_MULTICAST is not set # CONFIG_IP_ADVANCED_ROUTER is not set # CONFIG_IP_PNP is not set # CONFIG_NET_IPIP is not set # CONFIG_NET_IPGRE is not set # CONFIG_ARPD is not set # CONFIG_INET_ECN is not set # CONFIG_SYN_COOKIES is not set # CONFIG_IPV6 is not set # CONFIG_KHTTPD is not set # CONFIG_ATM is not set # CONFIG_VLAN_8021Q is not set # # # # CONFIG_IPX is not set # CONFIG_ATALK is not set # CONFIG_DECNET is not set # CONFIG_BRIDGE is not set # CONFIG_X25 is not set # CONFIG_LAPB is not set # CONFIG_LLC is not set # CONFIG_NET_DIVERT is not set # CONFIG_ECONET is not set # CONFIG_WAN_ROUTER is not set # CONFIG_NET_FASTROUTE is not set # CONFIG_NET_HW_FLOWCONTROL is not set # # QoS and/or fair queueing # # CONFIG_NET_SCHED is not set # # Telephony Support # # CONFIG_PHONE is not set # # ATA/IDE/MFM/RLL support # # CONFIG_IDE is not set # CONFIG_BLK_DEV_IDE_MODES is not set # CONFIG_BLK_DEV_HD is not set # # SCSI support # # CONFIG_SCSI is not set # # I2O device support # # CONFIG_I2O is not set # # Network device support # CONFIG_NETDEVICES=y # # ARCnet devices # # CONFIG_ARCNET is not set # CONFIG_DUMMY is not set # CONFIG_BONDING is not set # CONFIG_EQUALIZER is not set # CONFIG_TUN is not set # CONFIG_ETHERTAP is not set # # Ethernet (10 or 100Mbit) # # CONFIG_NET_ETHERNET is not set # # Ethernet (1000 Mbit) # # CONFIG_MYRI_SBUS is not set # CONFIG_FDDI is not set # CONFIG_HIPPI is not set CONFIG_PPP=y # CONFIG_PPP_MULTILINK is not set # CONFIG_PPP_ASYNC is not set # CONFIG_PPP_SYNC_TTY is not set # CONFIG_PPP_DEFLATE is not set # CONFIG_PPP_BSDCOMP is not set # CONFIG_PPPOE is not set # CONFIG_SLIP is not set # # Wireless LAN (non-hamradio) # # CONFIG_NET_RADIO is not set # # Token Ring devices # # CONFIG_TR is not set # CONFIG_NET_FC is not set # CONFIG_SHAPER is not set # # Wan interfaces # # CONFIG_WAN is not set # # Amateur Radio support # # CONFIG_HAMRADIO is not set # # IrDA (infrared) support # # CONFIG_IRDA is not set # # ISDN subsystem # # CONFIG_ISDN is not set # # Old CD-ROM drivers (not SCSI, not IDE) # # CONFIG_CD_NO_IDESCSI is not set # # Character devices # CONFIG_M68360_SMC_UART=y CONFIG_M68360_SCC_UART=y CONFIG_M68360_UART=y CONFIG_SERIAL_CONSOLE=y CONFIG_CONSOLE_9600=y # CONFIG_CONSOLE_19200 is not set # CONFIG_CONSOLE_115200 is not set # CONFIG_LEDMAN is not set # CONFIG_VT is not set # CONFIG_SERIAL is not set # CONFIG_SERIAL_NONSTANDARD is not set # CONFIG_UNIX98_PTYS is not set # # I2C support # # CONFIG_I2C is not set # # Mice # # CONFIG_BUSMOUSE is not set # CONFIG_MOUSE is not set # # Joysticks # # CONFIG_INPUT_GAMEPORT is not set # CONFIG_INPUT_SERIO is not set # # Joysticks # # CONFIG_QIC02_TAPE is not set # # Watchdog Cards # # CONFIG_WATCHDOG is not set # CONFIG_NVRAM is not set # CONFIG_RTC is not set # CONFIG_DTLK is not set # CONFIG_R3964 is not set # CONFIG_APPLICOM is not set # # Ftape, the floppy tape device driver # # CONFIG_FTAPE is not set # CONFIG_AGP is not set # CONFIG_DRM is not set # # File systems # # CONFIG_QUOTA is not set # CONFIG_AUTOFS_FS is not set # CONFIG_AUTOFS4_FS is not set # CONFIG_REISERFS_FS is not set # CONFIG_ADFS_FS is not set # CONFIG_AFFS_FS is not set # CONFIG_HFS_FS is not set # CONFIG_BFS_FS is not set # CONFIG_EXT3_FS is not set # CONFIG_JBD is not set # CONFIG_FAT_FS is not set # CONFIG_EFS_FS is not set # CONFIG_CRAMFS is not set # CONFIG_TMPFS is not set # CONFIG_RAMFS is not set # CONFIG_ISO9660_FS is not set # CONFIG_MINIX_FS is not set # CONFIG_VXFS_FS is not set # CONFIG_NTFS_FS is not set # CONFIG_HPFS_FS is not set CONFIG_PROC_FS=y # CONFIG_DEVFS_FS is not set # CONFIG_QNX4FS_FS is not set CONFIG_ROMFS_FS=y CONFIG_EXT2_FS=y # CONFIG_SYSV_FS is not set # CONFIG_UDF_FS is not set # CONFIG_UFS_FS is not set # # Network File Systems # # CONFIG_CODA_FS is not set # CONFIG_INTERMEZZO_FS is not set CONFIG_NFS_FS=y # CONFIG_NFS_V3 is not set # CONFIG_NFSD is not set CONFIG_SUNRPC=y CONFIG_LOCKD=y # CONFIG_SMB_FS is not set # CONFIG_NCP_FS is not set # CONFIG_ZISOFS_FS is not set # CONFIG_ZLIB_FS_INFLATE is not set # # Partition Types # # CONFIG_PARTITION_ADVANCED is not set CONFIG_MSDOS_PARTITION=y # CONFIG_SMB_NLS is not set # CONFIG_NLS is not set # # Sound # # CONFIG_SOUND is not set # # USB support # # # USB Controllers # # # USB Device Class drivers # # # USB Human Interface Devices (HID) # # # USB Imaging devices # # # USB Multimedia devices # # # USB Network adaptors # # # USB port drivers # # # USB Serial Converter support # # # USB Miscellaneous drivers # # # Kernel hacking # CONFIG_FULLDEBUG=y # CONFIG_MAGIC_SYSRQ is not set # CONFIG_CONTIGUOUS_PAGE_ALLOC is not set # CONFIG_PROFILE is not set CONFIG_NO_KERNEL_MSG=y # CONFIG_SMALL_TASKS is not set > The config file generated by make config/menuconfig/xconfig > > Kendrick Hamilton E.I.T. > SED Systems, a division of Calian Ltd. > 18 Innovation Blvd. > PO Box 1464 > Saskatoon, Saskatchewan > Canada > S7N 3R1 > > Hamilton at sedsystems.ca > Tel: (306) 933-1453 > Fax: (306) 933-1486 > > On Thu, 28 Mar 2002, Joachim Schweidler wrote: > > > > > which configuration file do you mean ? > > > > > There are 256 interrupts on the 68360 and Irq 68 is correct for SMC2. > Send > > > me your configuration file. > > > > > > Kendrick Hamilton E.I.T. > > > SED Systems, a division of Calian Ltd. > > > 18 Innovation Blvd. > > > PO Box 1464 > > > Saskatoon, Saskatchewan > > > Canada > > > S7N 3R1 > > > > > > Hamilton at sedsystems.ca > > > Tel: (306) 933-1453 > > > Fax: (306) 933-1486 > > > > > > On Thu, 28 Mar 2002, Joachim Schweidler wrote: > > > > > > > Hi Matthias, > > > > > > > > I thought I need standard serial support for the serial console ... > Good > > > > tip, thank you very much. > > > > > > > > But now I got another error: "THRE...rs_interrupt_single(0, > 2)...end." > > > > > > > > Do you know something about this error ? > > > > > > > > IRQ 68 ist right. This platform hat over 70 IRQs, and this one > worked > > > fine > > > > under 2.0.38. > > > > > > > > Joe > > > > > > > > > Joachim Schweidler wrote: > > > > > > > > > > > > Hi all, > > > > > > > > > > > > some months ago I got uclinux 2.0.38 up and running on an > 68EN360 > > > Board > > > > > > (SBS > > > > > > V360). Now I upgraded to Kernel 2.4.x and the kernel starts and > > > comes > > > > > till > > > > > > opening an initial console and then I get the error > > > "M68360_request_irq: > > > > > > IRQ 68 from ttyS is not replaceable". > > > > > > > > > > > > Another error that aperars sometimes is "LSR security check > > > engaged". > > > > > > > > > > > > It's a CVS - snapshot from 03/16/2002. > > > > > > > > > > > > Doas anyone have the same problem or knows a solution to this > > > problem. > > > > > > > > > > > > Many thanks in advance. > > > > > > Joe > > > > > > > > > > > > Here is a log I captured over the serial console (BDM > Interface): > > > > > > > > > > > > > > > ********************************************************************* > > > > > > > > > > > > Linux version 2.4.17-uc0 (root at for02067) (gcc version 2.95.3 > > > 20010315 > > > > > > (release)(Cold2 > > > > > > > > > > > > uClinux/MC68360 > > > > > > Flat model support (C) 1998,1999 Kenneth Albanowski, D. Jeff > Dionne > > > > > > QUICC port done by SED Systems , > > > > > > based on 2.0.38 port by Lineo Inc. . > > > > > > On node 0 totalpages: 1023 > > > > > > zone(0): 0 pages. > > > > > > zone(1): 1023 pages. > > > > > > zone(2): 0 pages. > > > > > > Kernel command line: > > > > > > CPM UART driver version 0.03 > > > > > > ttyS00 at irq 0x04 is an SMC > > > > > > ttyS01 at irq 0x1d is an SCC > > > > > > ttyS02 at irq 0x1c is an SCC > > > > > > ttyS03 at irq 0x1b is an SCC > > > > > > Calibrating delay loop... 4.37 BogoMIPS > > > > > > Memory available: 2700k/3332k RAM, 394k/1024k ROM (630k kernel > code, > > > > > 549k > > > > > > data) > > > > > > kmem_create: Forcing size word alignment - vm_area_struct > > > > > > kmem_create: Forcing size word alignment - mm_struct > > > > > > kmem_create: Forcing size word alignment - filp > > > > > > Dentry-cache hash table entries: 512 (order: 0, 4096 bytes) > > > > > > Inode-cache hash table entries: 512 (order: 0, 4096 bytes) > > > > > > kmem_create: Forcing size word alignment - inode_cache > > > > > > Mount-cache hash table entries: 512 (order: 0, 4096 bytes) > > > > > > kmem_create: Forcing size word alignment - bdev_cache > > > > > > kmem_create: Forcing size word alignment - cdev_cache > > > > > > Buffer-cache hash table entries: 1024 (order: 0, 4096 bytes) > > > > > > Page-cache hash table entries: 1024 (order: 0, 4096 bytes) > > > > > > POSIX conformance testing by UNIFIX > > > > > > Linux NET4.0 for Linux 2.4 > > > > > > Based upon Swansea University Computer Society NET3.039 > > > > > > kmem_create: Forcing size word alignment - sock > > > > > > Initializing RT netlink socket > > > > > > Starting kswapd > > > > > > kmem_create: Forcing size word alignment - file lock cache > > > > > > Serial driver version 5.05c (2001-07-08) with no serial options > > > enabled > > > > > > ttyS00 at 0x03f8 (irq = 4) is a 16450 > > > > > > ttyS01 at 0x02f8 (irq = 3) is a 16450 > > > > > > ttyS02 at 0x03e8 (irq = 4) is a 16450 > > > > > > kmem_create: Forcing size word alignment - blkdev_requests > > > > > > block: 64 slots per queue, batch=16 > > > > > > RAMDISK driver initialized: 16 RAM disks of 4096K size 1024 > > > blocksize > > > > > > Blkmem copyright 1998,1999 D. Jeff Dionne > > > > > > Blkmem copyright 1998 Kenneth Albanowski > > > > > > Blkmem 1 disk images: > > > > > > 0: C6A08-124207 [VIRTUAL C6A08-124207] (RO) > > > > > > PPP generic driver version 2.4.1 > > > > > > NET4: Linux TCP/IP 1.0 for NET4.0 > > > > > > IP Protocols: ICMP, UDP, TCP > > > > > > kmem_create: Forcing size word alignment - ip_dst_cache > > > > > > IP: routing cache hash table of 512 buckets, 4Kbytes > > > > > > TCP: Hash tables configured (established 512 bind 512) > > > > > > NET4: Unix domain sockets 1.0/SMP for Linux NET4.0. > > > > > > VFS: Mounted root (romfs filesystem) readonly. > > > > > > Freeing unused kernel memory: 0k freed (0x127000 - 0x126000) > > > > > > M68360_request_irq: IRQ 68 from ttyS is not replaceable > > > > > > M68360_request_irq: IRQ 68 from ttyS is not replaceable > > > > > > M68360_request_irq: IRQ 68 from ttyS is not replaceable > > > > > > M68360_request_irq: IRQ 68 from ttyS is not replaceable > > > > > > > > > > > > -- > > > > > > GMX - Die Kommunikationsplattform im Internet. > > > > > > http://www.gmx.net > > > > > > > > > > > > This message resent by the uclinux-dev at uclinux.org list server > > > > > http://www.uClinux.org/ > > > > > > > > > > Hello Joachim! > > > > > It seems, that you included the "Standard serial port support" in > the > > > > > configuration for your kernel. This is only for the 16450/550 > found on > > > > > (IBM-)PC-hardware. Take it out and hopefully your terminal will > work. > > > > > BTW, irq 68 should be the machine dependent bit position of the > irq > > > mask > > > > > bit. I'm not familiar with the 68EN360 , but it seems a bit high , > > > isn't > > > > > it? > > > > > > > > > > Greetings > > > > > matthias > > > > > This message resent by the uclinux-dev at uclinux.org list server > > > > > http://www.uClinux.org/ > > > > > > > > > > > > > -- > > > > Joachim Schweidler > > > > > > > > GMX - Die Kommunikationsplattform im Internet. > > > > http://www.gmx.net > > > > > > > > This message resent by the uclinux-dev at uclinux.org list server > > > http://www.uClinux.org/ > > > > > > > > > > This message resent by the uclinux-dev at uclinux.org list server > > > http://www.uClinux.org/ > > > > > > > -- > > Joachim Schweidler > > > > GMX - Die Kommunikationsplattform im Internet. > > http://www.gmx.net > > > > This message resent by the uclinux-dev at uclinux.org list server > http://www.uClinux.org/ > > > > This message resent by the uclinux-dev at uclinux.org list server > http://www.uClinux.org/ > -- GMX - Die Kommunikationsplattform im Internet. http://www.gmx.net This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From damnyosh at yahoo.com Thu Mar 28 13:29:06 2002 From: damnyosh at yahoo.com (Yosh K) Date: Thu, 28 Mar 2002 10:29:06 -0800 (PST) Subject: [uClinux-dev] MCF5272 Message-ID: <20020328182906.95357.qmail@web20603.mail.yahoo.com> Hello, I'm working on my senior project and need some guidance. I am working with MCF5272 eval board. I need to put the kernel in flash and execute in RAM. I've read Bernhard Kuhn's README on how to do this. Since his site is no longer there, the patches he talks about are no longer there either. I guess what I'm trying learn is that what do I need to do in order to get the MTD partitions. I got colilo uploaded into the topmost first meg of flash, but I lost about what changes I need to make to the kernel in order for it to stay in flash and decompress to 0x20000. Any pointer is much appreciated. Yosh __________________________________________________ Do You Yahoo!? Yahoo! Movies - coverage of the 74th Academy Awards? http://movies.yahoo.com/ This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From ktreis at keyww.com Thu Mar 28 13:52:45 2002 From: ktreis at keyww.com (Ken Treis) Date: Thu, 28 Mar 2002 10:52:45 -0800 Subject: [uClinux-dev] [uClibc] fprintf and non-blocking IO Message-ID: <3CA3667D.4020800@keyww.com> We're using fdopen() on a non-blocking socket FD, and then we're using fprintf() on that. Some questions: - is this sane? - Does uClibc properly deal with underlying file descriptors which are non-blocking? In particular, I was looking at this code in _uClibc_fwrite: while (bytes) { if ((rv = write(fp->fd, p, bytes)) < 0) { rv = 0; if (errno != EINTR) { break; } } p += rv; bytes -= rv; } Seems at first glance that we shouldn't break on EAGAIN. Thoughts? -- Ken Treis / Software Engineer Key Technology, Inc. ktreis at keyww.com ---- This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From andersen at codepoet.org Thu Mar 28 14:24:49 2002 From: andersen at codepoet.org (Erik Andersen) Date: Thu, 28 Mar 2002 12:24:49 -0700 Subject: [uClinux-dev] [uClibc] fprintf and non-blocking IO In-Reply-To: <3CA3667D.4020800@keyww.com> References: <3CA3667D.4020800@keyww.com> Message-ID: <20020328192449.GA27938@codepoet.org> On Thu Mar 28, 2002 at 10:52:45AM -0800, Ken Treis wrote: > We're using fdopen() on a non-blocking socket FD, and then we're using > fprintf() on that. Some questions: > > - is this sane? > - Does uClibc properly deal with underlying file descriptors which are > non-blocking? > > In particular, I was looking at this code in _uClibc_fwrite: > > while (bytes) { > if ((rv = write(fp->fd, p, bytes)) < 0) { > rv = 0; > if (errno != EINTR) { > break; That code is wrong. I fixed that in uClibc's CVS tree on Feb 23: http://uclibc.org/cgi-bin/cvsweb/uClibc/libc/stdio/stdio.c.diff?r1=1.36&r2=1.37 -Erik -- Erik B. Andersen http://codepoet-consulting.com/ --This message was written using 73% post-consumer electrons-- This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From ktreis at keyww.com Thu Mar 28 14:47:54 2002 From: ktreis at keyww.com (Ken Treis) Date: Thu, 28 Mar 2002 11:47:54 -0800 Subject: [uClinux-dev] [uClibc] fprintf and non-blocking IO References: <3CA3667D.4020800@keyww.com> <20020328192449.GA27938@codepoet.org> Message-ID: <3CA3736A.9070604@keyww.com> Erik Andersen wrote: > On Thu Mar 28, 2002 at 10:52:45AM -0800, Ken Treis wrote: > > We're using fdopen() on a non-blocking socket FD, and then we're using > > fprintf() on that. Some questions: > > > > - is this sane? > > - Does uClibc properly deal with underlying file descriptors which are > > non-blocking? > > > > In particular, I was looking at this code in _uClibc_fwrite: > > > > while (bytes) { > > if ((rv = write(fp->fd, p, bytes)) < 0) { > > rv = 0; > > if (errno != EINTR) { > > break; > > That code is wrong. I fixed that in uClibc's CVS tree on Feb 23: > > http://uclibc.org/cgi-bin/cvsweb/uClibc/libc/stdio/stdio.c.diff?r1=1.36&r2=1.37 > > > So, given that, does it make any sense to try to use fprintf() with a FILE* that's been fdopen()ed from a non-blocking descriptor? It seems that at any point, our _uClibc_fwrite could fail (i.e., it could be half-way through the format of a float), and that's going to be a nasty situation to try to recover from. -- Ken Treis / Software Engineer Key Technology, Inc. ktreis at keyww.com ---- -------------- next part -------------- An HTML attachment was scrubbed... URL: From alexx at ueidaq.com Thu Mar 28 15:34:18 2002 From: alexx at ueidaq.com (Alex Ivchenko) Date: Thu, 28 Mar 2002 15:34:18 -0500 Subject: [uClinux-dev] MCF5272 References: <20020328182906.95357.qmail@web20603.mail.yahoo.com> Message-ID: <3CA37E4A.45188020@ueidaq.com> Yosh, do you by any chance have a copy of Bernhard Kuhn's RTAI patch for MCF5272? BTW, I had the same problem with loading Linux from the flash before. I patched dBUG to copy data from the flash into RAM and start from 0x20000. -- Regards, Alex -- Alex Ivchenko, Ph.D. United Electronic Industries, Inc. "The High-Performance Alternative (tm)" -- 611 Neponset Street Canton, MA 02021 Tel:(781)821-2890 x222 Fax:(781)821-2891 http://www.ueidaq.com This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From Subashk at ami.com Thu Mar 28 15:55:22 2002 From: Subashk at ami.com (Subash Kalbarga) Date: Thu, 28 Mar 2002 15:55:22 -0500 Subject: [uClinux-dev] problems building uclibc Message-ID: <8CCBDD5583C50E4196F012E79439B45C364562@atl-ms1.megatrends.com> Hi, I am using the uclinux-dist 20020306 and compiling for M5272 I had trouble building uClibc with io.h and I followed the path suggested by an earlier mail in this forum But now it fails while copying crti.o to a non existent directory. Can anyone suggest a solution? Thanks in advance Subash -------------- next part -------------- An HTML attachment was scrubbed... URL: From damnyosh at yahoo.com Thu Mar 28 16:00:47 2002 From: damnyosh at yahoo.com (Yosh K) Date: Thu, 28 Mar 2002 13:00:47 -0800 (PST) Subject: [uClinux-dev] MCF5272 In-Reply-To: <3CA37E4A.45188020@ueidaq.com> Message-ID: <20020328210047.17707.qmail@web20601.mail.yahoo.com> Alex, Unfortunately I don't have any patches that Bernhard refers to. Right now I am using the CF Flasher for Windows to just directly upload images to flash. That little utility seems to work good. Yosh --- Alex Ivchenko wrote: > Yosh, > > do you by any chance have a copy of Bernhard Kuhn's > RTAI patch for MCF5272? > > BTW, I had the same problem with loading Linux from > the flash before. > I patched dBUG to copy data from the flash into RAM > and start from 0x20000. > > -- > Regards, > Alex > > -- > Alex Ivchenko, Ph.D. > United Electronic Industries, Inc. > "The High-Performance Alternative (tm)" > -- > 611 Neponset Street > Canton, MA 02021 > Tel:(781)821-2890 x222 Fax:(781)821-2891 > http://www.ueidaq.com > This message resent by the uclinux-dev at uclinux.org > list server http://www.uClinux.org/ __________________________________________________ Do You Yahoo!? Yahoo! Movies - coverage of the 74th Academy Awards? http://movies.yahoo.com/ This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From andersen at codepoet.org Thu Mar 28 16:20:24 2002 From: andersen at codepoet.org (Erik Andersen) Date: Thu, 28 Mar 2002 14:20:24 -0700 Subject: [uClinux-dev] [uClibc] fprintf and non-blocking IO In-Reply-To: <3CA3736A.9070604@keyww.com> References: <3CA3667D.4020800@keyww.com> <20020328192449.GA27938@codepoet.org> <3CA3736A.9070604@keyww.com> Message-ID: <20020328212024.GA29433@codepoet.org> On Thu Mar 28, 2002 at 11:47:54AM -0800, Ken Treis wrote: > > > So, given that, does it make any sense to try to use fprintf() with a > FILE* that's been fdopen()ed from a non-blocking descriptor? It seems > that at any point, our _uClibc_fwrite could fail (i.e., it could be > half-way through the format of a float), and that's going to be a nasty > situation to try to recover from. If you set the FILE to non-blocking, stdio doesn't know about that, and so yes, you need to expect that any operation using that FILE can fail. This is not a uClibc or uClinux specific problem though. Basically, using stdio on a non-blocking descriptor is going to troublesome, and will need lots of extra error checking and application support, regardless of OS/libc... -Erik -- Erik B. Andersen http://codepoet-consulting.com/ --This message was written using 73% post-consumer electrons-- This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From JLiu at PacesetterElectronics.com Thu Mar 28 17:01:29 2002 From: JLiu at PacesetterElectronics.com (JINGSONG LIU) Date: Thu, 28 Mar 2002 16:01:29 -0600 Subject: [uClinux-dev] BDM driver error ! Message-ID: All: I'm new to Linux and uclinux and I met problem when I want to install BDM for debugging. The problem is: I try to install a BDM driver in my redhat 7.2 for MCF5272C3 but I got many when issue insmod bdm.o from reahat 7.2. Is something wrong when building the driver or something else ? Thank you in advance! Jason Liu -------------- next part -------------- An HTML attachment was scrubbed... URL: From chico.santanna at uol.com.br Thu Mar 28 14:33:37 2002 From: chico.santanna at uol.com.br (Francisco Sant'Anna) Date: Thu, 28 Mar 2002 19:33:37 +0000 Subject: [uClinux-dev] Run in EB-40A Message-ID: <20020328193337.4254e320.chico.santanna@uol.com.br> Hello, I have an EB-40A modified by Atmel with 1MB-RAM / 4MB-Flash. Will it run ucLinux ? The board distributed with uclinux (EB-40LS) has 2MB-RAM, is it nedded or 1MB-RAM is enough ? If needed, is it possible to run from Flash ? thanks. -- -chico This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From alexx at ueidaq.com Thu Mar 28 17:46:33 2002 From: alexx at ueidaq.com (Alex Ivchenko) Date: Thu, 28 Mar 2002 17:46:33 -0500 Subject: [uClinux-dev] RTAI and uCLinux for ColdFire References: <20020328210047.17707.qmail@web20601.mail.yahoo.com> Message-ID: <3CA39D49.A39DFEB2@ueidaq.com> Guys, does anybody have by any chance a copy of Bernhard Kuhn's RTAI patch for MCF5272? His site seems to be down :-( -- Regards, Alex -- Alex Ivchenko, Ph.D. United Electronic Industries, Inc. "The High-Performance Alternative (tm)" -- 611 Neponset Street Canton, MA 02021 Tel:(781)821-2890 x222 Fax:(781)821-2891 http://www.ueidaq.com This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From rwehrli at azpower.com Thu Mar 28 18:42:28 2002 From: rwehrli at azpower.com (Rob Wehrli) Date: Thu, 28 Mar 2002 16:42:28 -0700 Subject: [uClinux-dev] New H8S uClinux dev platform... Message-ID: <3CA3AA64.AF952458@azpower.com> I'm trying to learn how many people would be interested in the following uClinux development platform...to get an idea of how many to build! 33MHz Hitachi H8S 10Mbit/s Ethernet 8MB SDRAM 4MB Flash 512KB Boot ROM RTC RS-232 Serial Port User LED Coin Cell Back-up Battery Pinned headers bring out CPU pins for extending HW $199 end-user pricing quantity 1 (due out May 2002) Take Care. Rob! This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From smackeron at caltech.edu Thu Mar 28 19:11:49 2002 From: smackeron at caltech.edu (Matt Bergeron) Date: Thu, 28 Mar 2002 16:11:49 -0800 Subject: [uClinux-dev] Very large images In-Reply-To: Message-ID: <90A73BE1-42A9-11D6-8272-000502777EB1@caltech.edu> Hi, Was this large images issue on the uCsimm ever resolved? I'm using distribution 20020220 and trying to build a 2.0.39 kernel with the ELF toolchain, and getting an image over 200 MB in size. I'm getting the exact same warning in the linker stage that Steven was. Is there a problem with the distro? Thanks, Matt On Sunday, October 21, 2001, at 05:10 PM, Steven Merrifield wrote: > Hi all, > > I am still having problems getting the elf tools to create an image > that is > less > than 200M. I am getting two warnings at the linker stage, shown below. > Is > the > last one due to a missing flag somewhere, and is it the cause of my > problem? > > Using distribution 20010622, kernel 2.0.38, EZ328, elf, patched > Rules.make > to > disable the coff tools. > > Thanks, > steve > > m68k-elf-ld -g -T arch/m68knommu/platform/68EZ328/ucsimm/rom.ld > arch/m68knommu/p > latform/68EZ328/ucsimm/crt0_rom.o init/main.o init/version.o \ > arch/m68knommu/kernel/kernel.o arch/m68knommu/mm/mm.o > arch/m68knommu/platform/68 > EZ328/platform.o kernel/kernel.o fs/fs.o ipc/ipc.o net/network.a > mmnommu/mm.o \ > fs/filesystems.a \ > drivers/block/block.a drivers/char/char.a drivers/net/net.a \ > /win/uClinux-distribution/linux/lib/lib.a arch/m68knommu/lib/lib.a > `m68k-elf-gcc > -g -D__KERNEL__ -I/win/uClinux-distribution/linux/include -v 2>&1 | > grep > specs > | sed -e "s/Reading specs from //" | sed -e > s/specs/m68000\\\/libgcc.a/` -o > linu > x > /usr/local/bin/m68k-elf-ld.real: Warning: size of symbol `irq2dev_map' > changed f > rom 4 to 128 in auto_irq.o > /usr/local/bin/m68k-elf-ld.real: warning: no memory region specified for > section > `.rodata' > m68k-elf-nm linux | grep -v '\(compiled\)\|\(\.o$\)\|\( a \)' | sort > > System.ma > p > > This message resent by the uclinux-dev at uclinux.org list server > http://www.uClinux.org/ This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From miles at lsi.nec.co.jp Thu Mar 28 20:05:47 2002 From: miles at lsi.nec.co.jp (Miles Bader) Date: 29 Mar 2002 10:05:47 +0900 Subject: [uClinux-dev] Re: out-of-line BUG In-Reply-To: <20020328213307.B5246@beast.internal.moreton.com.au> References: <20020328213307.B5246@beast.internal.moreton.com.au> Message-ID: David McCullough writes: > This means all the strings are identical and get combined by the > compiler (maybe ?). No, gcc before 3.x doesn't do this (that's the reason for Andrew's patch). > If this patch goes into the kernel proper we will pick it up when we import > it. At that time we would just convert everything else over to use it. > > I am happy to wait unless other feel we need to preempt this going into the > kernel proper ? Naw, it's just a nice space savings. Thanks, -Miles -- `The suburb is an obsolete and contradictory form of human settlement' This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From jeff at arcturusnetworks.com Thu Mar 28 20:10:11 2002 From: jeff at arcturusnetworks.com (D. Jeff Dionne) Date: Fri, 29 Mar 2002 01:10:11 -0000 Subject: [uClinux-dev] Very large images Message-ID: <200203290110.g2T1ABF15153@arcturusnetworks.com> Matt Bergeron said: > Hi, > > Was this large images issue on the uCsimm ever resolved? I'm using > distribution 20020220 and trying to build a 2.0.39 kernel with the ELF > toolchain, and getting an image over 200 MB in size. Something is wrong with your .ld file or your object copy from ELF to binary. > I'm getting the > exact same warning in the linker stage that Steven was. Is there a > problem with the distro? Distro? Not sure what that is (WRT uClinux) but you are having a problem with the kernel (build system or .ld files) interacting with the toolchain. On the linux ELF file, run m68k-elf-nm or m68k-elf-objdump --headers and have a look at what sections are where (and how big). Then look at what the .ld file is trying to do and what the Makefiles (or Rules.make) are trying to strip out to make the binary image. Likely just something 'extra' that the ELF toolchain is putting into the file that COFF didn't... being aware that objcopy has to put space in the binary file for any holes in the memory map that remain (say, between .text+.rodata and .data+.bss) J > > Thanks, > Matt > > > On Sunday, October 21, 2001, at 05:10 PM, Steven Merrifield wrote: > > > Hi all, > > > > I am still having problems getting the elf tools to create an image > > that is > > less > > than 200M. I am getting two warnings at the linker stage, shown below. > > Is > > the > > last one due to a missing flag somewhere, and is it the cause of my > > problem? > > > > Using distribution 20010622, kernel 2.0.38, EZ328, elf, patched > > Rules.make > > to > > disable the coff tools. > > > > Thanks, > > steve > > > > m68k-elf-ld -g -T arch/m68knommu/platform/68EZ328/ucsimm/rom.ld > > arch/m68knommu/p > > latform/68EZ328/ucsimm/crt0_rom.o init/main.o init/version.o \ > > arch/m68knommu/kernel/kernel.o arch/m68knommu/mm/mm.o > > arch/m68knommu/platform/68 > > EZ328/platform.o kernel/kernel.o fs/fs.o ipc/ipc.o net/network.a > > mmnommu/mm.o \ > > fs/filesystems.a \ > > drivers/block/block.a drivers/char/char.a drivers/net/net.a \ > > /win/uClinux-distribution/linux/lib/lib.a arch/m68knommu/lib/lib.a > > `m68k-elf-gcc > > -g -D__KERNEL__ -I/win/uClinux-distribution/linux/include -v 2>&1 | > > grep > > specs > > | sed -e "s/Reading specs from //" | sed -e > > s/specs/m68000\\\/libgcc.a/` -o > > linu > > x > > /usr/local/bin/m68k-elf-ld.real: Warning: size of symbol `irq2dev_map' > > changed f > > rom 4 to 128 in auto_irq.o > > /usr/local/bin/m68k-elf-ld.real: warning: no memory region specified for > > section > > `.rodata' > > m68k-elf-nm linux | grep -v '\(compiled\)\|\(\.o$\)\|\( a \)' | sort > > > System.ma > > p > > > > This message resent by the uclinux-dev at uclinux.org list server > > http://www.uClinux.org/ > > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ > -- D. Jeff Dionne Jeff at ArcturusNetworks.com -------------- next part -------------- A non-text attachment was scrubbed... Name: Type: application/octet-stream Size: 2 bytes Desc: not available URL: From sujz at cad.csie.ncku.edu.tw Fri Mar 29 04:48:39 2002 From: sujz at cad.csie.ncku.edu.tw (SuCC) Date: Fri, 29 Mar 2002 17:48:39 +0800 Subject: [uClinux-dev] pppd supported Message-ID: <000a01c1d706$ef268820$2df7748c@SCC> Hi, I have pppd supported by uClinux-dist Is this pppd support MS-CHAP "server"?? or how can I do so that pppd can support MS-CHAP "server"!! thanks!! -------------- next part -------------- An HTML attachment was scrubbed... URL: From robert at schwebel.de Fri Mar 29 05:25:37 2002 From: robert at schwebel.de (Robert Schwebel) Date: Fri, 29 Mar 2002 11:25:37 +0100 (CET) Subject: [uClinux-dev] MCF5272 In-Reply-To: <3CA37E4A.45188020@ueidaq.com> Message-ID: On Thu, 28 Mar 2002, Alex Ivchenko wrote: > do you by any chance have a copy of Bernhard Kuhn's RTAI patch for > MCF5272? I've incorporated Bernhard's m68knommu stuff into the RTAI CVS. Please be sure to checkout the m68knommu branch. 24.1.9 will probably have the patch included. Robert -- +--------------------------------------------------------+ | Dipl.-Ing. Robert Schwebel | http://www.pengutronix.de | | Pengutronix - Linux Solutions for Science and Industry | | Braunschweiger Str. 79, 31134 Hildesheim, Germany | | Phone: +49-5121-28619-0 | Fax: +49-5121-28619-4 | +--------------------------------------------------------+ This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From shankarr at connexus.com.sg Fri Mar 29 08:32:55 2002 From: shankarr at connexus.com.sg (Shankar) Date: Fri, 29 Mar 2002 21:32:55 +0800 Subject: [uClinux-dev] Signal 11 problem.... In-Reply-To: <000b01c1d225$900d53e0$1a000a0a@cxs8> Message-ID: <000001c1d726$3ccecf00$1a000a0a@cxs8> Hi all, I have download and and compiled the jpeg compression using arm-uclinux-gcc compiler. When I tried to execute the cjpeg exec.. I am getting killed(Signal 11) error and the exec gets terminated.. But the same code is working if compiled it for x86.. The following is the makefile I have used...Please help me to solve this.. Thank you.. with regards, woodmissile. srcdir = . # The name of your C compiler: CC= arm-uclinux-gcc # You may need to adjust these cc options: INC = -I- -I. -I../../home/rama/lh79531r04Feb2002/usr.20010706/include -I../../h ome/rama/lh79531r04Feb2002/usr.20010706/local/arm-uclinux/include DEFS = -DL_openraw -DL_setifaddr -DL_setifflags -DL_addroute -DL_closeraw CFLAGS = -fpic -mdisable-got -Wa,-k -mcpu=arm7 $(DEFS) -O1 -I$(srcdir) -ggdb3 $(INC) # Link-time cc options: LDFLAGS= -v # To link any special libraries, add the necessary -l commands here. LDLIBS= ../../home/rama/lh79531r04Feb2002/usr.20010706/local/arm-uclinux/lib/libc.a \ ../../home/rama/lh79531r04Feb2002/usr.20010706/local/lib/gcc-lib/arm-uclinux /2.96/libgcc.a CRT0 = -T ../../home/rama/lh79531r04Feb2002/usr.20010706/local/arm-uclinux/lib/ldscrip ts/ucarm-app.x ../../home/rama/lh79531r04Feb2002/usr.20010706/local/arm-uclinux/lib/crt0.o # If using GNU libtool, LIBTOOL references it; if not, LIBTOOL is empty. LIBTOOL = # $(O) expands to "lo" if using libtool, plain "o" if not. # Similarly, $(A) expands to "la" or "a". O = o A = a # Library version ID; libtool uses this for the shared library version number. # Note: we suggest this match the macro of the same name in jpeglib.h. JPEG_LIB_VERSION = 62 # Put here the object file name for the correct system-dependent memory # manager file. For Unix this is usually jmemnobs.o, but you may want # to use jmemansi.o or jmemname.o if you have limited swap space. SYSDEPMEM= jmemname.$(O) # miscellaneous OS-dependent stuff SHELL= /bin/sh # linker # LN= $(CC) LN = arm-uclinux-ld # file deletion command RM= rm -f # directory creation command MKDIR= mkdir # library (.a) file creation command AR= arm-uclinux-ar rc # second step in .a creation (use "touch" if not needed) AR2= arm-uclinux-ranlib # installation program INSTALL= /usr/bin/install -c INSTALL_PROGRAM= ${INSTALL} INSTALL_LIB= ${INSTALL} -m 644 INSTALL_DATA= ${INSTALL} -m 644 # End of configurable options. # source files: JPEG library proper LIBSOURCES= jcapimin.c jcapistd.c jccoefct.c jccolor.c jcdctmgr.c jchuff.c \ jcinit.c jcmainct.c jcmarker.c jcmaster.c jcomapi.c jcparam.c \ jcphuff.c jcprepct.c jcsample.c jctrans.c jdapimin.c jdapistd.c \ jdatadst.c jdatasrc.c jdcoefct.c jdcolor.c jddctmgr.c jdhuff.c \ jdinput.c jdmainct.c jdmarker.c jdmaster.c jdmerge.c jdphuff.c \ jdpostct.c jdsample.c jdtrans.c jerror.c jfdctflt.c jfdctfst.c \ jfdctint.c jidctflt.c jidctfst.c jidctint.c jidctred.c jquant1.c \ jquant2.c jutils.c jmemmgr.c # memmgr back ends: compile only one of these into a working library SYSDEPSOURCES= jmemansi.c jmemname.c jmemnobs.c jmemdos.c jmemmac.c # source files: cjpeg/djpeg/jpegtran applications, also rdjpgcom/wrjpgcom APPSOURCES= cjpeg.c djpeg.c jpegtran.c rdjpgcom.c wrjpgcom.c cdjpeg.c \ rdcolmap.c rdswitch.c transupp.c rdppm.c wrppm.c rdgif.c wrgif.c \ rdtarga.c wrtarga.c rdbmp.c wrbmp.c rdrle.c wrrle.c SOURCES= $(LIBSOURCES) $(SYSDEPSOURCES) $(APPSOURCES) # files included by source files INCLUDES= jchuff.h jdhuff.h jdct.h jerror.h jinclude.h jmemsys.h jmorecfg.h \ jpegint.h jpeglib.h jversion.h cdjpeg.h cderror.h transupp.h # documentation, test, and support files DOCS= README install.doc usage.doc cjpeg.1 djpeg.1 jpegtran.1 rdjpgcom.1 \ wrjpgcom.1 wizard.doc example.c libjpeg.doc structure.doc \ coderules.doc filelist.doc change.log MKFILES= configure makefile.cfg makefile.ansi makefile.unix makefile.bcc \ makefile.mc6 makefile.dj makefile.wat makefile.vc makelib.ds \ makeapps.ds makeproj.mac makcjpeg.st makdjpeg.st makljpeg.st \ maktjpeg.st makefile.manx makefile.sas makefile.mms makefile.vms \ makvms.opt CONFIGFILES= jconfig.cfg jconfig.bcc jconfig.mc6 jconfig.dj jconfig.wat \ jconfig.vc jconfig.mac jconfig.st jconfig.manx jconfig.sas \ jconfig.vms CONFIGUREFILES= config.guess config.sub install-sh ltconfig ltmain.sh OTHERFILES= jconfig.doc ckconfig.c ansi2knr.c ansi2knr.1 jmemdosa.asm TESTFILES= testorig.jpg testimg.ppm testimg.bmp testimg.jpg testprog.jpg \ testimgp.jpg DISTFILES= $(DOCS) $(MKFILES) $(CONFIGFILES) $(SOURCES) $(INCLUDES) \ $(CONFIGUREFILES) $(OTHERFILES) $(TESTFILES) # library object files common to compression and decompression COMOBJECTS= jcomapi.$(O) jutils.$(O) jerror.$(O) jmemmgr.$(O) $(SYSDEPMEM) # compression library object files CLIBOBJECTS= jcapimin.$(O) jcapistd.$(O) jctrans.$(O) jcparam.$(O) \ jdatadst.$(O) jcinit.$(O) jcmaster.$(O) jcmarker.$(O) jcmainct.$(O) \ jcprepct.$(O) jccoefct.$(O) jccolor.$(O) jcsample.$(O) jchuff.$(O) \ jcphuff.$(O) jcdctmgr.$(O) jfdctfst.$(O) jfdctflt.$(O) \ jfdctint.$(O) # decompression library object files DLIBOBJECTS= jdapimin.$(O) jdapistd.$(O) jdtrans.$(O) jdatasrc.$(O) \ jdmaster.$(O) jdinput.$(O) jdmarker.$(O) jdhuff.$(O) jdphuff.$(O) \ jdmainct.$(O) jdcoefct.$(O) jdpostct.$(O) jddctmgr.$(O) \ jidctfst.$(O) jidctflt.$(O) jidctint.$(O) jidctred.$(O) \ jdsample.$(O) jdcolor.$(O) jquant1.$(O) jquant2.$(O) jdmerge.$(O) # These objectfiles are included in libjpeg.a LIBOBJECTS= $(CLIBOBJECTS) $(DLIBOBJECTS) $(COMOBJECTS) # object files for sample applications (excluding library files) COBJECTS= cjpeg.$(O) rdppm.$(O) rdgif.$(O) rdtarga.$(O) rdrle.$(O) \ rdbmp.$(O) rdswitch.$(O) cdjpeg.$(O) DOBJECTS= djpeg.$(O) wrppm.$(O) wrgif.$(O) wrtarga.$(O) wrrle.$(O) \ wrbmp.$(O) rdcolmap.$(O) cdjpeg.$(O) TROBJECTS= jpegtran.$(O) rdswitch.$(O) cdjpeg.$(O) transupp.$(O) all: libjpeg.$(A) cjpeg djpeg jpegtran rdjpgcom wrjpgcom # Special compilation rules to support ansi2knr and libtool. .SUFFIXES: .lo .la # How to compile with libtool. # .c.lo: # $(LIBTOOL) --mode=compile $(CC) $(CFLAGS) -c $(srcdir)/$*.c # How to use ansi2knr, when not using libtool. # .c.o: # ./ansi2knr $(srcdir)/$*.c knr/$*.c # $(CC) $(CFLAGS) -c knr/$*.c # $(RM) knr/$*.c # How to use ansi2knr AND libtool. # .c.lo: # ./ansi2knr $(srcdir)/$*.c knr/$*.c # $(LIBTOOL) --mode=compile $(CC) $(CFLAGS) -c knr/$*.c # $(RM) knr/$*.c ansi2knr: ansi2knr.c $(CC) $(CFLAGS) $(ANSI2KNRFLAGS) -o ansi2knr $(srcdir)/ansi2knr.c $(MKDIR) knr # the library: # without libtool: libjpeg.a: $(LIBOBJECTS) $(RM) libjpeg.a $(AR) libjpeg.a $(LIBOBJECTS) $(AR2) libjpeg.a # with libtool: libjpeg.la: $(LIBOBJECTS) $(LIBTOOL) --mode=link $(CC) -o libjpeg.la $(LIBOBJECTS) \ -rpath $(libdir) -version-info $(JPEG_LIB_VERSION) # sample programs: cjpeg: $(COBJECTS) libjpeg.$(A) $(LN) $(LDFLAGS) -o cjpeg.elf $(CRT0) $(COBJECTS) libjpeg.$(A) $(LDLIBS) elf2flt cjpeg.elf -o cjpeg djpeg: $(DOBJECTS) libjpeg.$(A) $(LN) $(LDFLAGS) -o djpeg.elf $(CRT0) $(DOBJECTS) libjpeg.$(A) $(LDLIBS) elf2flt djpeg.elf -o djpeg jpegtran: $(TROBJECTS) libjpeg.$(A) $(LN) $(LDFLAGS) -o jpegtran.elf $(CRT0) $(TROBJECTS) libjpeg.$(A) $(LDLIBS) elf2flt jpegtran.elf -o jpegtran rdjpgcom: rdjpgcom.$(O) $(LN) $(LDFLAGS) -o rdjpgcom.elf $(CRT0) rdjpgcom.$(O) $(LDLIBS) elf2flt rdjpgcom.elf -o rdjpgcom wrjpgcom: wrjpgcom.$(O) $(LN) $(LDFLAGS) -o wrjpgcom.elf $(CRT0) wrjpgcom.$(O) $(LDLIBS) elf2flt wrjpgcom.elf -o wrjpgcom # Installation rules: install: cjpeg djpeg jpegtran rdjpgcom wrjpgcom $(INSTALL_PROGRAM) cjpeg $(bindir)/$(binprefix)cjpeg $(INSTALL_PROGRAM) djpeg $(bindir)/$(binprefix)djpeg $(INSTALL_PROGRAM) jpegtran $(bindir)/$(binprefix)jpegtran $(INSTALL_PROGRAM) rdjpgcom $(bindir)/$(binprefix)rdjpgcom $(INSTALL_PROGRAM) wrjpgcom $(bindir)/$(binprefix)wrjpgcom $(INSTALL_DATA) $(srcdir)/cjpeg.1 $(mandir)/$(manprefix)cjpeg.$(manext) $(INSTALL_DATA) $(srcdir)/djpeg.1 $(mandir)/$(manprefix)djpeg.$(manext) $(INSTALL_DATA) $(srcdir)/jpegtran.1 $(mandir)/$(manprefix)jpegtran.$(manext) $(INSTALL_DATA) $(srcdir)/rdjpgcom.1 $(mandir)/$(manprefix)rdjpgcom.$(manext) $(INSTALL_DATA) $(srcdir)/wrjpgcom.1 $(mandir)/$(manprefix)wrjpgcom.$(manext) install-lib: libjpeg.$(A) install-headers $(INSTALL_LIB) libjpeg.$(A) $(libdir)/$(binprefix)libjpeg.$(A) install-headers: jconfig.h $(INSTALL_DATA) jconfig.h $(includedir)/jconfig.h $(INSTALL_DATA) $(srcdir)/jpeglib.h $(includedir)/jpeglib.h $(INSTALL_DATA) $(srcdir)/jmorecfg.h $(includedir)/jmorecfg.h $(INSTALL_DATA) $(srcdir)/jerror.h $(includedir)/jerror.h clean: $(RM) *.o *.lo libjpeg.a libjpeg.la $(RM) cjpeg djpeg jpegtran rdjpgcom wrjpgcom $(RM) ansi2knr core testout* config.log config.status $(RM) -r knr .libs _libs distclean: clean $(RM) Makefile jconfig.h libtool config.cache test: cjpeg djpeg jpegtran $(RM) testout* ./djpeg -dct int -ppm -outfile testout.ppm $(srcdir)/testorig.jpg ./djpeg -dct int -bmp -colors 256 -outfile testout.bmp $(srcdir)/testorig.jpg ./cjpeg -dct int -outfile testout.jpg $(srcdir)/testimg.ppm ./djpeg -dct int -ppm -outfile testoutp.ppm $(srcdir)/testprog.jpg ./cjpeg -dct int -progressive -opt -outfile testoutp.jpg $(srcdir)/testimg.ppm ./jpegtran -outfile testoutt.jpg $(srcdir)/testprog.jpg cmp $(srcdir)/testimg.ppm testout.ppm cmp $(srcdir)/testimg.bmp testout.bmp cmp $(srcdir)/testimg.jpg testout.jpg cmp $(srcdir)/testimg.ppm testoutp.ppm cmp $(srcdir)/testimgp.jpg testoutp.jpg cmp $(srcdir)/testorig.jpg testoutt.jpg check: test # Mistake catcher: jconfig.h: jconfig.doc echo You must prepare a system-dependent jconfig.h file. echo Please read the installation directions in install.doc. exit 1 # GNU Make likes to know which target names are not really files to be made: .PHONY: all install install-lib install-headers clean distclean test check jcapimin.$(O): jcapimin.c jinclude.h jconfig.h jpeglib.h jmorecfg.h jpegint.h jerror.h jcapistd.$(O): jcapistd.c jinclude.h jconfig.h jpeglib.h jmorecfg.h jpegint.h jerror.h jccoefct.$(O): jccoefct.c jinclude.h jconfig.h jpeglib.h jmorecfg.h jpegint.h jerror.h jccolor.$(O): jccolor.c jinclude.h jconfig.h jpeglib.h jmorecfg.h jpegint.h jerror.h jcdctmgr.$(O): jcdctmgr.c jinclude.h jconfig.h jpeglib.h jmorecfg.h jpegint.h jerror.h jdct.h jchuff.$(O): jchuff.c jinclude.h jconfig.h jpeglib.h jmorecfg.h jpegint.h jerror.h jchuff.h jcinit.$(O): jcinit.c jinclude.h jconfig.h jpeglib.h jmorecfg.h jpegint.h jerror.h jcmainct.$(O): jcmainct.c jinclude.h jconfig.h jpeglib.h jmorecfg.h jpegint.h jerror.h jcmarker.$(O): jcmarker.c jinclude.h jconfig.h jpeglib.h jmorecfg.h jpegint.h jerror.h jcmaster.$(O): jcmaster.c jinclude.h jconfig.h jpeglib.h jmorecfg.h jpegint.h jerror.h jcomapi.$(O): jcomapi.c jinclude.h jconfig.h jpeglib.h jmorecfg.h jpegint.h jerror.h jcparam.$(O): jcparam.c jinclude.h jconfig.h jpeglib.h jmorecfg.h jpegint.h jerror.h jcphuff.$(O): jcphuff.c jinclude.h jconfig.h jpeglib.h jmorecfg.h jpegint.h jerror.h jchuff.h jcprepct.$(O): jcprepct.c jinclude.h jconfig.h jpeglib.h jmorecfg.h jpegint.h jerror.h jcsample.$(O): jcsample.c jinclude.h jconfig.h jpeglib.h jmorecfg.h jpegint.h jerror.h jctrans.$(O): jctrans.c jinclude.h jconfig.h jpeglib.h jmorecfg.h jpegint.h jerror.h jdapimin.$(O): jdapimin.c jinclude.h jconfig.h jpeglib.h jmorecfg.h jpegint.h jerror.h jdapistd.$(O): jdapistd.c jinclude.h jconfig.h jpeglib.h jmorecfg.h jpegint.h jerror.h jdatadst.$(O): jdatadst.c jinclude.h jconfig.h jpeglib.h jmorecfg.h jerror.h jdatasrc.$(O): jdatasrc.c jinclude.h jconfig.h jpeglib.h jmorecfg.h jerror.h jdcoefct.$(O): jdcoefct.c jinclude.h jconfig.h jpeglib.h jmorecfg.h jpegint.h jerror.h jdcolor.$(O): jdcolor.c jinclude.h jconfig.h jpeglib.h jmorecfg.h jpegint.h jerror.h jddctmgr.$(O): jddctmgr.c jinclude.h jconfig.h jpeglib.h jmorecfg.h jpegint.h jerror.h jdct.h jdhuff.$(O): jdhuff.c jinclude.h jconfig.h jpeglib.h jmorecfg.h jpegint.h jerror.h jdhuff.h jdinput.$(O): jdinput.c jinclude.h jconfig.h jpeglib.h jmorecfg.h jpegint.h jerror.h jdmainct.$(O): jdmainct.c jinclude.h jconfig.h jpeglib.h jmorecfg.h jpegint.h jerror.h jdmarker.$(O): jdmarker.c jinclude.h jconfig.h jpeglib.h jmorecfg.h jpegint.h jerror.h jdmaster.$(O): jdmaster.c jinclude.h jconfig.h jpeglib.h jmorecfg.h jpegint.h jerror.h jdmerge.$(O): jdmerge.c jinclude.h jconfig.h jpeglib.h jmorecfg.h jpegint.h jerror.h jdphuff.$(O): jdphuff.c jinclude.h jconfig.h jpeglib.h jmorecfg.h jpegint.h jerror.h jdhuff.h jdpostct.$(O): jdpostct.c jinclude.h jconfig.h jpeglib.h jmorecfg.h jpegint.h jerror.h jdsample.$(O): jdsample.c jinclude.h jconfig.h jpeglib.h jmorecfg.h jpegint.h jerror.h jdtrans.$(O): jdtrans.c jinclude.h jconfig.h jpeglib.h jmorecfg.h jpegint.h jerror.h jerror.$(O): jerror.c jinclude.h jconfig.h jpeglib.h jmorecfg.h jversion.h jerror.h jfdctflt.$(O): jfdctflt.c jinclude.h jconfig.h jpeglib.h jmorecfg.h jpegint.h jerror.h jdct.h jfdctfst.$(O): jfdctfst.c jinclude.h jconfig.h jpeglib.h jmorecfg.h jpegint.h jerror.h jdct.h jfdctint.$(O): jfdctint.c jinclude.h jconfig.h jpeglib.h jmorecfg.h jpegint.h jerror.h jdct.h jidctflt.$(O): jidctflt.c jinclude.h jconfig.h jpeglib.h jmorecfg.h jpegint.h jerror.h jdct.h jidctfst.$(O): jidctfst.c jinclude.h jconfig.h jpeglib.h jmorecfg.h jpegint.h jerror.h jdct.h jidctint.$(O): jidctint.c jinclude.h jconfig.h jpeglib.h jmorecfg.h jpegint.h jerror.h jdct.h jidctred.$(O): jidctred.c jinclude.h jconfig.h jpeglib.h jmorecfg.h jpegint.h jerror.h jdct.h jquant1.$(O): jquant1.c jinclude.h jconfig.h jpeglib.h jmorecfg.h jpegint.h jerror.h jquant2.$(O): jquant2.c jinclude.h jconfig.h jpeglib.h jmorecfg.h jpegint.h jerror.h jutils.$(O): jutils.c jinclude.h jconfig.h jpeglib.h jmorecfg.h jpegint.h jerror.h jmemmgr.$(O): jmemmgr.c jinclude.h jconfig.h jpeglib.h jmorecfg.h jpegint.h jerror.h jmemsys.h jmemansi.$(O): jmemansi.c jinclude.h jconfig.h jpeglib.h jmorecfg.h jpegint.h jerror.h jmemsys.h jmemname.$(O): jmemname.c jinclude.h jconfig.h jpeglib.h jmorecfg.h jpegint.h jerror.h jmemsys.h jmemnobs.$(O): jmemnobs.c jinclude.h jconfig.h jpeglib.h jmorecfg.h jpegint.h jerror.h jmemsys.h jmemdos.$(O): jmemdos.c jinclude.h jconfig.h jpeglib.h jmorecfg.h jpegint.h jerror.h jmemsys.h jmemmac.$(O): jmemmac.c jinclude.h jconfig.h jpeglib.h jmorecfg.h jpegint.h jerror.h jmemsys.h cjpeg.$(O): cjpeg.c cdjpeg.h jinclude.h jconfig.h jpeglib.h jmorecfg.h jerror.h cderror.h jversion.h djpeg.$(O): djpeg.c cdjpeg.h jinclude.h jconfig.h jpeglib.h jmorecfg.h jerror.h cderror.h jversion.h jpegtran.$(O): jpegtran.c cdjpeg.h jinclude.h jconfig.h jpeglib.h jmorecfg.h jerror.h cderror.h transupp.h jversion.h rdjpgcom.$(O): rdjpgcom.c jinclude.h jconfig.h wrjpgcom.$(O): wrjpgcom.c jinclude.h jconfig.h cdjpeg.$(O): cdjpeg.c cdjpeg.h jinclude.h jconfig.h jpeglib.h jmorecfg.h jerror.h cderror.h rdcolmap.$(O): rdcolmap.c cdjpeg.h jinclude.h jconfig.h jpeglib.h jmorecfg.h jerror.h cderror.h rdswitch.$(O): rdswitch.c cdjpeg.h jinclude.h jconfig.h jpeglib.h jmorecfg.h jerror.h cderror.h transupp.$(O): transupp.c jinclude.h jconfig.h jpeglib.h jmorecfg.h jpegint.h jerror.h transupp.h rdppm.$(O): rdppm.c cdjpeg.h jinclude.h jconfig.h jpeglib.h jmorecfg.h jerror.h cderror.h wrppm.$(O): wrppm.c cdjpeg.h jinclude.h jconfig.h jpeglib.h jmorecfg.h jerror.h cderror.h rdgif.$(O): rdgif.c cdjpeg.h jinclude.h jconfig.h jpeglib.h jmorecfg.h jerror.h cderror.h wrgif.$(O): wrgif.c cdjpeg.h jinclude.h jconfig.h jpeglib.h jmorecfg.h jerror.h cderror.h rdtarga.$(O): rdtarga.c cdjpeg.h jinclude.h jconfig.h jpeglib.h jmorecfg.h jerror.h cderror.h wrtarga.$(O): wrtarga.c cdjpeg.h jinclude.h jconfig.h jpeglib.h jmorecfg.h jerror.h cderror.h rdbmp.$(O): rdbmp.c cdjpeg.h jinclude.h jconfig.h jpeglib.h jmorecfg.h jerror.h cderror.h wrbmp.$(O): wrbmp.c cdjpeg.h jinclude.h jconfig.h jpeglib.h jmorecfg.h jerror.h cderror.h rdrle.$(O): rdrle.c cdjpeg.h jinclude.h jconfig.h jpeglib.h jmorecfg.h jerror.h cderror.h wrrle.$(O): wrrle.c cdjpeg.h jinclude.h jconfig.h jpeglib.h jmorecfg.h jerror.h cderror.h This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From gerg at snapgear.com Fri Mar 29 08:54:09 2002 From: gerg at snapgear.com (Greg Ungerer) Date: Fri, 29 Mar 2002 23:54:09 +1000 Subject: [uClinux-dev] 2nd serial port on M5272C3 devt board References: <002501c1d636$c1a0bf40$2df7748c@SCC> <022001c1d672$46377560$8947ec95@bruker.fr> Message-ID: <3CA47201.8030202@snapgear.com> Hi Julien, Julien Boibessot wrote: > does anyone can confirm that uClinux is handling the 2nd serial port on > M5272C3 devt board ? Yes, it can handle the 2nd serial port. > if yes how should I configure the kernel to get access to this port with > "/dev/ttyS1" (for example) ?? Yes, device is /dev/ttyS1. Make sure you have the latest code version. There was a problem in older ColdFire serial driver that meant the 2nd serial port was not enabled at the chip pins. Regards Greg ------------------------------------------------------------------------ Greg Ungerer -- Chief Software Wizard EMAIL: gerg at snapgear.com SnapGear PHONE: +61 7 3279 1822 825 Stanley St, FAX: +61 7 3279 1820 Woolloogabba, QLD, 4102, Australia WEB: www.snapgear.com This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From gerg at snapgear.com Fri Mar 29 09:07:05 2002 From: gerg at snapgear.com (Greg Ungerer) Date: Sat, 30 Mar 2002 00:07:05 +1000 Subject: [uClinux-dev] Re: User Application: Permission denied. References: <010801c1d670$19399590$2200a8c0@flamenco> Message-ID: <3CA47509.3000209@snapgear.com> Hi Usman, Muhammad Usman Tahir wrote: > I've an MCF 5206 system with 8MB RAM and having 2.0.38 uClinux kernel > running fine with basic apps like shell, dhcpcd etc. I've added one of my > applications in '/user/my_app' and it has been successfully compiled and > linked to generate a binary. I then modified 'rc' script to run this app by > do_rc from init. On execution it says : > > Command: my_app > my_app: Permission denied. > > and then continues with the next item in 'rc'. > > I have checked all the permissions for the binary of my_app in the > 'romfs/bin' and have even forced it to be executable but the error message > remains the same. I tried to copy the file attribs from the running dhcpcd > app but still ended up with the same result. > > I should add here that I bypassed the generation of elf file while > compilation and linked to get a flat binary directly with -elf2flt option in > m68k-elf-gcc. (I had lots of problems with elf2flt and searching through the > archives I found this work around to what seems to be a common problem). I > think this should not have caused any problem as long as I don't need an elf > file for my_app, which I don't. Can you reply with what the exact command line was? Have you check that you really do have a FLAT format file? Hex dump the file and check that it has a valid looking header. You could use the "flthdr" program too to check if it is a valid FLAT file. > Any idea how I can get my_app really working without this message? Shall I > try some other method to start my_app? Thanks in advance for the all the > help. Try running it from the command line. Regards Greg ------------------------------------------------------------------------ Greg Ungerer -- Chief Software Wizard EMAIL: gerg at snapgear.com SnapGear PHONE: +61 7 3279 1822 825 Stanley St, FAX: +61 7 3279 1820 Woolloogabba, QLD, 4102, Australia WEB: www.snapgear.com This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From siddons at bnl.gov Fri Mar 29 09:47:12 2002 From: siddons at bnl.gov (D. Peter Siddons) Date: Fri, 29 Mar 2002 09:47:12 -0500 Subject: [uClinux-dev] RTAI and uCLinux for ColdFire References: <20020328210047.17707.qmail@web20601.mail.yahoo.com> <3CA39D49.A39DFEB2@ueidaq.com> Message-ID: <3CA47E70.21094EDC@bnl.gov> Hi Alex, Here they are. Pete. Alex Ivchenko wrote: > > Guys, > > does anybody have by any chance a copy of Bernhard Kuhn's > RTAI patch for MCF5272? > > His site seems to be down :-( > > -- > Regards, > Alex > > -- > Alex Ivchenko, Ph.D. > United Electronic Industries, Inc. > "The High-Performance Alternative (tm)" > -- > 611 Neponset Street > Canton, MA 02021 > Tel:(781)821-2890 x222 Fax:(781)821-2891 > http://www.ueidaq.com > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ -- ----------------- D. Peter Siddons National Synchrotron Light Source, Bldg. 725D Brookhaven National Laboratory Upton New York 11973 Email: siddons at bnl.gov -------------- next part -------------- A non-text attachment was scrubbed... Name: rtai_patches.tgz Type: application/x-gtar Size: 31349 bytes Desc: not available URL: From uclinux at schoeldgen.de Fri Mar 29 16:00:13 2002 From: uclinux at schoeldgen.de (Matthias) Date: Fri, 29 Mar 2002 22:00:13 +0100 Subject: [uClinux-dev] M68360: init: fork failed References: <18363.1017317794@www54.gmx.net> <7251.1017323227@www54.gmx.net> Message-ID: <3CA4D5DD.EBB30746@schoeldgen.de> Joachim Schweidler wrote: > > Hello, > > now I got the serial port up and running. Everything seems to be fine but > when init starts I get the message: > > ... > init: fork of rc shell failed > init: Booting to single user mode. > init: fork of single user shell failed > init: fork failed > init: fork failed > init: fork failed > init: fork failed > > -- > GMX - Die Kommunikationsplattform im Internet. > http://www.gmx.net > > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ Hello Joachim! Try the following: enable DEBUGGING in simpleinit.c , but remove the exit(0) statement after the output of "toks", "tty" and the like. For some unknown reason this worked for me. Make sure you have a valid 'rc' file in your /etc directory, and you should be up and running. I also had problems with the first /dev/console entry configured inside simpleinit.c, so i removed this too and relied only on my inittab. There is another problem: the inittab was incorrect for my platform (ucsimm). Please compare this: inet:unknown:/bin/inetd ttyS0:vt100:/bin/agetty -l /bin/login ttyS0 9600 vt100 with yours. Leave out the inetd thing, if you don't need it. good luck matthias This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From gmenie at akamai.com Tue Mar 26 03:40:12 2002 From: gmenie at akamai.com (Menie, Georges) Date: Tue, 26 Mar 2002 03:40:12 -0500 Subject: [uClinux-dev] 68*328 timer was: "Calibrating delay loop..." c rashes Message-ID: Attached is the patch for the ucdimm; could you try it, then if everything is ok, I will generate a patch for the whole 68*328 arch, merging the RTC patch of Bernard Kuhn so everyone will have the choice to use timer 1 or the rtc to generate ticks. After applying this patch, you will have to edit your linux-2.4.x/include/m68knommu/param.h to change the HZ constant to 128 to get a better accuracy. Cheers, Georges > -----Original Message----- > From: Daniel Haensse [mailto:daniel.haensse at fhk.usz.ch] > Sent: lundi 25 mars 2002 14:28 > To: uclinux-dev at uclinux.org > Subject: Re: [uClinux-dev] 68*328 timer was: "Calibrating > delay loop..." > c rashes > > > Am Montag, 25. M?rz 2002 14.04 schrieben Sie: > > thanks for the try, it seems that 128Hz give enough accuracy, > > I did try it also, and I had to add a mach_gettimeoffset() > > function in order to have an acceptable accuracy. > > > > I'm going to send a new patch for the DragenEngine board, > > which includes the timer changes and other de2 specific changes. > > If there is some interest, I can generate a timer patch > > for the 68328 arch. > Yes, please, we're interested. > > > > Cheers, > > Georges -------------- next part -------------- An embedded and charset-unspecified text was scrubbed... Name: patch-timer.txt URL: From davidm at snapgear.com Tue Mar 26 19:17:26 2002 From: davidm at snapgear.com (David McCullough) Date: Wed, 27 Mar 2002 10:17:26 +1000 Subject: [uClinux-dev] Picogui - compiled and linked.... In-Reply-To: <3CA0A377.CC036AF5@cern.ch>; from jody.pearson@cern.ch on Tue, Mar 26, 2002 at 05:36:07PM +0100 References: <3CA0A377.CC036AF5@cern.ch> Message-ID: <20020327101726.A3613@beast.internal.moreton.com.au> Jivin Jody Pearson lays it down ... ... > -rwxr--r-- 1 root 5764 299808 Mar 26 17:24 pgserver > # file pgserver* > pgserver: data > pgserver.gdb: ELF 32-bit MSB executable, Motorola 68000, version 1, > statically linked, not stripped > # > > Then i copy that into me romfs, and remake the image. > > SO when I try to run the bin, i get; > > Sash command shell (version 1.1.1) > /> pgserver > MAP_SHARED not supported (cannot write mappings to disk) There is an "mmap" call somewhere in the code that is using mmap in a way that isn't supported under uClinux. You probably need to use an anonymous mapping to get it going. Cheers, Davidm -- David McCullough: Ph: +61 7 3435 2815 http://www.SnapGear.com davidm at snapgear.com Fx: +61 7 3891 3630 825 Stanley St., W'gabba QLD 4102, Oz This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From julienb at actimage.fr Wed Mar 27 02:59:55 2002 From: julienb at actimage.fr (Julien Boibessot) Date: Wed, 27 Mar 2002 08:59:55 +0100 Subject: [uClinux-dev] Gdb+bdm+5272C3 References: <412A28C967692A438B59223143A7268FE1B147@ALTAIR.endevco.com> Message-ID: <04ee01c1d565$61ce2e80$8947ec95@bruker.fr> No this patch doesn't work with 5.1.x gdb...... I was faced to this problem too...... 5.0 gdb is the right choice....... Julien ----- Original Message ----- From: "Henderson, Jim" To: Sent: Tuesday, March 26, 2002 10:03 PM Subject: RE: [uClinux-dev] Gdb+bdm+5272C3 > I tried patching gdb-5.1.1 with gdb-bdm-20010901 (from > ftp://www.cybertec.com.au/pub/bdm) and it didn't work for me. I had get > gdb-5.0 from ftp://sources.redhat.com/pub/gdb/old-releases/ and patch it. > > Is gdb-bdm-20010901 supposed to patch gdb-5.1.1? > > -----Original Message----- > From: Julien Boibessot [mailto:julienb at actimage.fr] > Sent: Monday, March 18, 2002 12:04 AM > To: uclinux-dev at uclinux.org > Subject: Re: [uClinux-dev] Gdb+bdm+5272C3 > > > Hi Greg ! > > Well, in fact my gdb init script is the one you posted on this list some > times ago.....:-) > And it initializes DRAM controller...... > > Thanks > Julien > > ----- Original Message ----- > From: "Greg Ungerer" > To: > Sent: Monday, March 18, 2002 12:53 AM > Subject: Re: [uClinux-dev] Gdb+bdm+5272C3 > > > > > > Hi Julien, > > > > Julien Boibessot wrote: > > > > > I'm trying to use GDB thrue BDM (P&E Adaptator) with the Motorola > 5272C3. > > > > > > I have compile and insert BDM driver ( gdb-bdm-20010901 from cybertec) > in my > > > host linux and it works (chk /dev/bdmcf0 is running and all tests are > ok). > > > I've got GDB 5.1.1 sources, patche it for m68k/coldfire BDM debugging > and > > > recompile it to get a running "m68k-bdm-elf-gdb". > > > > > > Then, I launch gdb with: "m68k-bdm-elf-gdb --command=5272.gdb > image.elf". > > > 5272.gdb is my init script for GDB and image.elf is my linux image. > > > > > > gdb is running and I can do whatever I want. So I launch the "load" > command, > > > take a coffee to wait the end of the download, and "set $pc=0x20000". > But > > > then when I try to execute the program by typing command "c", gdb > crashes (I > > > get a "segmentation fault"). My board is reseted. > > > > > > Does anybody knows what is happening ??? > > > > > > Does you gdbinit script initialize RAM/SDRAM? > > > > If it doesn't then your load cannot possibly work, since > > it will be trying to load the binary into RAM. > > > > Regards > > Greg > > > > > > ------------------------------------------------------------------------ > > Greg Ungerer -- Chief Software Wizard EMAIL: gerg at snapgear.com > > SnapGear PHONE: +61 7 3435 2888 > > 825 Stanley St, FAX: +61 7 3891 3630 > > Woolloongabba, QLD, 4102, Australia WEB: www.snapgear.com > > > > This message resent by the uclinux-dev at uclinux.org list server > http://www.uClinux.org/ > > > > This message resent by the uclinux-dev at uclinux.org list server > http://www.uClinux.org/ > > > ------------------------------------------------------------------------ > This email may contain confidential information and/or copyright material. > This email is intended for the use of the addressee only. Any unauthorized > use may be unlawful. If you receive this email by mistake, please advise the > sender immediately by using the reply facility in your email software.You > should be aware that the contents of this email may be monitored to ensure > compliance with the Meggitt IT User policy. > > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ > This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/ From Subashk at ami.com Thu Mar 28 15:55:22 2002 From: Subashk at ami.com (Subash Kalbarga) Date: Thu, 28 Mar 2002 15:55:22 -0500 Subject: [uClinux-dev] problems building uclibc Message-ID: <8CCBDD5583C50E4196F012E79439B45C364562@atl-ms1.megatrends.com> Hi, I am using the uclinux-dist 20020306 and compiling for M5272 I had trouble building uClibc with io.h and I followed the path suggested by an earlier mail in this forum But now it fails while copying crti.o to a non existent directory. Can anyone suggest a solution? Thanks in advance Subash -------------- next part -------------- An HTML attachment was scrubbed... URL: From shiwucai at 163.com Sat Mar 23 03:18:53 2002 From: shiwucai at 163.com (cai) Date: Sat, 23 Mar 2002 16:18:53 +0800 Subject: [uClinux-dev] EB55 Board Message-ID: <000d01c1d2f0$95f3a040$f60aa8c0@wizvisioko3yuo> hi, I want to port uclinux to ATMEL's EB55 board , you see , I am a beginner to embed system , so anyboy can tell me how I should do step by step? -------------- next part -------------- An HTML attachment was scrubbed... URL: