Hi Andrew, Andrew Stubbs wrote:
OK for mainline ?This looks fine to me. I know there will be things that need fixing for both experimental architectures.
Indeed. I tried to be a bit more verbose also to avoid too high expectations by occasional gcc-patches@ readers.
P.S. Apologies, but I think my commits today conflict a little; you should be able to drop the hunks that patch deleted code.
I did so - but I then realized that I should have also added gfx1100 to the new chunk.
Committed as r14-7006-g97a52f69d209f6 (see attachment) - as follow up to the original r14-7005-g52a2c659ae6c21
Tobias
commit 97a52f69d209f69e755ffad6897c7176da9ac686 Author: Tobias Burnus <tob...@codesourcery.com> Date: Mon Jan 8 15:18:10 2024 +0100 amdgcn: Add gfx1100 to new XNACK defaults in mkoffload Commit r14-6997-g78dff4c25c1b95 added an arch-dependent SET_XNACK_OFF vs. SET_XNACK_ANY check; that was added between writing and committing the add-gfx1100 commit r14-7005-g52a2c659ae6c21 - and I missed to add it there. gcc/ChangeLog: * config/gcn/mkoffload.cc (main): Handle gfx1100 when setting the default XNACK. --- gcc/config/gcn/mkoffload.cc | 1 + 1 file changed, 1 insertion(+) diff --git a/gcc/config/gcn/mkoffload.cc b/gcc/config/gcn/mkoffload.cc index 2cd201d56ca..d4cd509089e 100644 --- a/gcc/config/gcn/mkoffload.cc +++ b/gcc/config/gcn/mkoffload.cc @@ -1018,6 +1018,7 @@ main (int argc, char **argv) case EF_AMDGPU_MACH_AMDGCN_GFX906: case EF_AMDGPU_MACH_AMDGCN_GFX908: case EF_AMDGPU_MACH_AMDGCN_GFX1030: + case EF_AMDGPU_MACH_AMDGCN_GFX1100: SET_XNACK_OFF (elf_flags); break; case EF_AMDGPU_MACH_AMDGCN_GFX90a: