This is the mail archive of the
mailing list for the GCC project.
Re: [RFA patch i386]: Prepare x64 prologue using positive offsets for frame-pointer
On Mon, Jul 5, 2010 at 2:17 AM, Kai Tietz <firstname.lastname@example.org> wrote:
> 2010/7/4 Kai Tietz <email@example.com>:
>> this patch prepares for x64 SEH, that frame-pointer points at end of
>> stack-prologue. This is required for x64, because the SEH unwind
>> information can't express negative offsets for stack.
>> 2010-07-04 ?Kai Tietz
>> ? ? ? ?* config/i386/i386.c (ix86_compute_frame_layout): Let frame's
>> ? ? ? ?hard_frame_pointer_offset point to end of prologue.
>> ? ? ? ?(ix86_expand_prologue): Handle x64 frame-pointer prologue.
>> ? ? ? ?(ix86_expand_epilogue): Handle x64 frame-pointer epilogue.
>> Tested for i686-pc-cygwin, x86_64-pc-mingw32. Ok for apply?
> This update patch avoids, that for profiling-built the x64
> frame-layout gets used. It would lead to issues, caused by profiling
> code, which assumes old stack-frame layout.
> | ?(\_/) This is Bunny. Copy and paste
> | (='.'=) Bunny into your signature to help
> | (")_(") him gain world domination