[Bug target/91489] misplaced stack pointer when __ms_hook_prologue__ attribute is used
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=91489 Jeffrey A. Law changed: What|Removed |Added Resolution|--- |FIXED Status|WAITING |RESOLVED CC||law at redhat dot com --- Comment #6 from Jeffrey A. Law --- Should be fixed on trunk now
[Bug target/91489] misplaced stack pointer when __ms_hook_prologue__ attribute is used
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=91489 --- Comment #5 from CVS Commits --- The master branch has been updated by Jeff Law : https://gcc.gnu.org/g:659ba632e440280e8d61e1ae870e917765db5ae5 commit r11-4806-g659ba632e440280e8d61e1ae870e917765db5ae5 Author: Jeff Law Date: Fri Nov 6 17:38:00 2020 -0500 Fix stack pointer handling in ms_hook_prologue functions for i386 target. gcc/ PR target/91489 * config/i386/i386.md (simple_return): Also check for ms_hook_prologue function attribute. * config/i386/i386.c (ix86_can_use_return_insn_p): Also check for ms_hook_prologue function attribute. * config/i386/i386-protos.h (ix86_function_ms_hook_prologue): Declare. gcc/testsuite PR target/91489 * gcc.target/i386/ms_hook_prologue.c: Expand testcase to reproduce PR target/91489 issue.
[Bug target/91489] misplaced stack pointer when __ms_hook_prologue__ attribute is used
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=91489 --- Comment #4 from Paul Gofman --- I suppose I figured a better way to fix this and sent the patch to the mailing list: https://gcc.gnu.org/ml/gcc-patches/2020-02/msg00554.html
[Bug target/91489] misplaced stack pointer when __ms_hook_prologue__ attribute is used
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=91489
--- Comment #3 from Paul Gofman ---
(In reply to Martin Liška from comment #2)
> @Paul: Can you please send the patch to GCC patches mailing list?
I think I found later that unfortunately this patch is not quite correct: the
stack was still wrong under certain conditions, IIRC at least (but probably not
only) if __attribute__((optimize("omit-frame-pointer"))) is additionally used
on function. I thought that such sort of fix looks more like a workaround
anyway. Probably a better fix would be to either track stack pointer properly
upon generating ms hook prologue or maybe enforce some function attributes if
ms_hook_prologue is used, but that time I did not dig gcc code deep enough to
find how to do that right.
I am currently away for about 2 weeks, after that I could get back to it and at
least update the test to indicate how it is still wrong with such a patch.
[Bug target/91489] misplaced stack pointer when __ms_hook_prologue__ attribute is used
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=91489 Martin Liška changed: What|Removed |Added Status|UNCONFIRMED |WAITING Last reconfirmed||2020-01-29 CC||marxin at gcc dot gnu.org Ever confirmed|0 |1 --- Comment #2 from Martin Liška --- @Paul: Can you please send the patch to GCC patches mailing list?
[Bug target/91489] misplaced stack pointer when __ms_hook_prologue__ attribute is used
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=91489 --- Comment #1 from Paul Gofman --- Created attachment 46739 --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=46739&action=edit Pop frame pointer in function label if it is not needed instead of prologue for "ms_hook_prologue" functions on i386 The attached patch fixes the issue for me. I am not sure if that is the best way to fix it though.
