This is the mail archive of the
binutils@sourceware.org
mailing list for the binutils project.
Re: [ARM] fix cortex-a8-fix bug
- From: Nitin Garg <nitingarg98 at gmail dot com>
- To: Nathan Sidwell <nathan at codesourcery dot com>
- Cc: binutils <binutils at sources dot redhat dot com>
- Date: Thu, 6 Aug 2009 16:50:24 -0500
- Subject: Re: [ARM] fix cortex-a8-fix bug
- References: <4A7AD4B1.80906@codesourcery.com>
Is this relevant to Cortex-A9 as well?
Regards,
Nitin
On Thu, Aug 6, 2009 at 8:03 AM, Nathan Sidwell<nathan@codesourcery.com> wrote:
> I've applied this obvious patch to fix a cortex-a8-fix bug. ?The nature of
> the erratum means that section alignments must be preserved modulo the page
> size, and rather than force the section alignment to be the page size (and
> introduce padding), we must ensure they don't move after we've applied the
> fix.
>
> Unfortunately, we only call layout_sections_again if we find at least one
> stub to insert, so when there are none to insert (on the initial pass), we
> leave it to a later linker stage to do the layout. ?That subsequent layout
> can move sections around, if ELF headers are to be placed in the first
> segment.
>
> Fixed thusly, tested on arm-eabi and arm-linux-gnueabi.
>
> nathan
> --
> Nathan Sidwell ? ?:: ? http://www.codesourcery.com ? :: ? ? ? ? CodeSourcery
>
>