This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: [RS6000] PR 45807 r2 eh_frame info error
- From: David Edelsohn <dje dot gcc at gmail dot com>
- To: GCC Patches <gcc-patches at gcc dot gnu dot org>, Alan Modra <amodra at gmail dot com>
- Date: Mon, 27 Sep 2010 18:56:21 -0400
- Subject: Re: [RS6000] PR 45807 r2 eh_frame info error
- References: <20100927142714.GS16376@bubble.grove.modra.org>
On Mon, Sep 27, 2010 at 10:27 AM, Alan Modra <amodra@gmail.com> wrote:
> This patch makes sure that we emit eh_frame info for r2 at the place
> we actually save it. ?Previously, __builtin_unwind_init saved r2,
> while the prologue emitted the eh_frame info for r2. ?This opened a
> window for error, and Murphy did the rest. ?In unwind-dw2.c
> _Unwind_Resume_or_Rethrow we manage to insert a call before r2 is
> saved. ?See the PR for the ugly details.
>
> Powerpc64-linux bootstrap and regression test in progress. ?OK to
> apply assuming no regressions? ?I'd like to backport this to active
> branches too.
>
> ? ? ? ?PR target/45807
> ? ? ? ?* config/rs6000/aix.h (SETUP_FRAME_ADDRESSES): Delete.
> ? ? ? ?* config/rs6000/linux64.h (SETUP_FRAME_ADDRESSES): Delete.
> ? ? ? ?* config/rs6000/rs6000-protos.h (rs6000_aix_emit_builtin_unwind_init):
> ? ? ? ?Delete.
> ? ? ? ?* config/rs6000/rs6000.c (rs6000_aix_emit_builtin_unwind_init): Delete.
> ? ? ? ?(rs6000_emit_prologue): Don't just create frame save info for r2,
> ? ? ? ?actually save r2.
Okay.
Thanks, David