[uClinux-dev] relocs problems with arm

Tom Walsh tom at cyberiansoftware.com
Thu Nov 29 19:22:45 EST 2001


Fabrice Gautier wrote:
> 
> > -----Original Message-----
> > From: Greg Ungerer [mailto:gerg at snapgear.com]
> > Subject: Re: [uClinux-dev] relocs problems with arm
> >
> > Hi Fabrice,
> >
> > Here is the one I have been using recently:
> >
> > http://www.uclinux.org/pub/uClinux/utilities/arm-elf-elf2flt-2
> > 0011114.tar.gz
> >
> > It certainly generates relocation entries.
> 
> It seems better indeed (flat_dump no report relocs) but it's not yet good.
> 
> I think i have an endianness problem.
> 
> Once loaded the relocation entries contain 0xa6530040 while the offset of
> the data is 0x46a4,
> and the base offset is 0x020d0000.
> 
> If i do 0x20d0000 + 0xa4460000 (0x46a4 swapped) + 0x40 (size of the header)
> i get 0xa6530040
> so there must something bad in the endiannes.
> 
> My host is a pc and the offset seems to be coded like on a pc (big-endian, i
> guess we call that).  Wheras my target may be little endian.

The X86 PC is little-endian, "little end first" meaning that the low
order part of the word/long is encountered first when going up in
addresses.

Your cross compiler and bintools should be the only determining factor
in how the endian-ess of the code image is formed.


TomW


-- 
Tom Walsh - WN3L - Embedded Systems Consultant
http://openhardware.net, http://cyberiansoftware.com
"Windows? No thanks, I have work to do..."
----------------------------------------------------
This message resent by the uclinux-dev at uclinux.org list server http://www.uClinux.org/



More information about the uClinux-dev mailing list