[OE-core] binutils: regression on kernel builds

[Khem Raj]

yes gold is showing some regressions see another one here
http://sourceware.org/bugzilla/show_bug.cgi?id=14592
cant promise but will take a look. meanwhile if you can triage
it would be nice.


On Mon, Dec 10, 2012 at 3:48 AM, Enrico Scholz <
enrico.scholz at sigma-chemnitz.de> wrote:

> Hi,
>
> with recent binutils, kernel build fails with
>
> | arm-linux-gnueabi-ld: error: arch/arm/boot/compressed/piggy.gzip.o:
> unknown CPU architecture
>
> I am able to reproduce it with
>
> $ > /tmp/x.s
> $ arm-linux-gnueabi-as  -march=armv7-a -march=all -o x0.o /tmp/x.s
> $ arm-linux-gnueabi-as  -march=armv7-a  -o x1.o /tmp/x.s
> $ arm-linux-gnueabi-ld.gold x0.o x1.o
> arm-linux-gnueabi-ld: error: x1.o: unknown CPU architecture
> $ arm-linux-gnueabi-ld.bfd x0.o x1.o
> arm-linux-gnueabi-ld.bfd: warning: cannot find entry symbol _start;
> defaulting to 0000000000008034
>
>
> (kernel sets '-Wa,-march=all' when building piggy.gzip.o)
>
>
> This works fine with binutils 2.22 and when using the BFD linker.
>
>
>
> Enrico
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openembedded.org/pipermail/openembedded-core/attachments/20121210/b452b0f8/attachment-0002.html>