ping: [patch] Fix zero registers core files w/gcc-4.7

H.J. Lu hjl.tools@gmail.com
Wed Jan 4 22:52:00 GMT 2012


On Wed, Jan 4, 2012 at 5:46 AM, Jan Kratochvil
<jan.kratochvil@redhat.com> wrote:
> ping
>
> It is a dependency for GDB 7.4 release:
>        http://sourceware.org/gdb/wiki/GDB_7.4_Release
>
> Probably obvious but still...
>
>
> On Fri, 23 Dec 2011 22:15:32 +0100, Jan Kratochvil wrote:
> Hello H.J.,
>
> there is a regression by:
>        commit 2c9c556617a7de8657c25b512d272c26b070ae22
>        Author: H.J. Lu <hjl.tools@gmail.com>
>        Date:   Thu Jun 16 22:08:10 2011 +0000
>            Suport x32 gcore.
>
> when built with -O2 by
>        gcc (GCC) 4.7.0 20111223 (experimental)
>        (not by 4.6.x)
>
> as the code violates ISO C99 6.2.4 item 5 by using local variable outside of
> its block, GCC optimizes out the second memcpy, keeping there only that
> memset.
>
> ./gdb -nx ./gdb -ex start -ex 'gcore 1' -ex 'set confirm no' -ex q;gdb -nx ./gdb ./1 -ex q
> will print:
> Core was generated by `.../gdb/gdb'.
> #0  0x0000000000000000 in ?? ()
>
> No regressions on x86_64-fedora16-linux-gnu.
>
> Probably obvious, OK to check it in?
>
>
> Thanks,
> Jan
>
>
> bfd/
> 2011-12-23  Jan Kratochvil  <jan.kratochvil@redhat.com>
>
>        Fix zero registers core files when built by gcc-4.7.
>        * elf64-x86-64.c (elf_x86_64_write_core_note): Remove variables p and
>        size.  Call elfcore_write_note for the local variables.  Remove the
>        final elfcore_write_note call.  Add NOTREACHED comments.
>
>

OK.  Sorry for the delay.

-- 
H.J.



More information about the Binutils mailing list