[OE-core] [PATCH] insane: detect and warn about relocations in .text

Mark Hatle mark.hatle at windriver.com
Wed Oct 3 15:56:36 UTC 2012


As an FYI -- if you have text relocations, then the prelinker will skip that 
binary, resulting in lower startup performance.  So as a general QA check, it's 
well worth having.

--Mark

On 10/3/12 10:42 AM, Phil Blundell wrote:
> On Wed, 2012-10-03 at 11:24 +0100, Phil Blundell wrote:
>> Signed-off-by: Phil Blundell <pb at pbcl.net>
>> ---
>> This requires qa.elf.run_objdump() so needs to be applied after the
>> patch which adds that function.
>
> By the way, the background motivation for this was that the current
> oe-core version of gcc seems to be somewhat broken in this respect on
> MIPS; it seems to have reverted to generating an .eh_frame with a
> relocation in it which results in DT_TEXTREL on the binary.
>
> With -fasynchronous-unwind-tables it seems to get even worse.  I don't
> quite recall exactly what the failure was there but it was bad enough to
> make it unusable.
>
> I think both of these problems are regressions from the gcc-4.6 that we
> were using previously.  I'm not quite sure at what point it went wrong
> though.
>
> p.
>
>
>
> _______________________________________________
> Openembedded-core mailing list
> Openembedded-core at lists.openembedded.org
> http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-core
>





More information about the Openembedded-core mailing list