projects / platform / upstream / llvm.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: 1f67dc8) | patch
[AMDGPU] Remove the assertion for MUBUF instruction with voffset
authorChristudasan Devadasan <Christudasan.Devadasan@amd.com>
Sat, 12 Nov 2022 13:29:48 +0000 (18:59 +0530)
committerChristudasan Devadasan <Christudasan.Devadasan@amd.com>
Wed, 16 Nov 2022 06:36:37 +0000 (12:06 +0530)
commit349b6afef47ba34ee3328f7ec6becf1d8c1be2ea
tree5afca91c55b4b548b69ab5f29e13ff79a6631f06tree | snapshot
parent1f67dc8b7c225290d1b3eb93d90e2c9861aeefc0commit | diff
[AMDGPU] Remove the assertion for MUBUF instruction with voffset

Currently, there is an assertion that limits the MUBUF instruction
with voffset used for a VGPR spill inside kernel functions when
the frame pointer is really used. It seemed more like an unwanted
limitation and hence removing the assertion.

Reviewed By: arsenm

Differential Revision: https://reviews.llvm.org/D137892
llvm/lib/Target/AMDGPU/SIRegisterInfo.cpp diff | blob | history
llvm/test/CodeGen/AMDGPU/kernel-mubuf-with-voffset.mir [new file with mode: 0644] blob
llvm/test/CodeGen/AMDGPU/kernel-vgpr-spill-mubuf-with-voffset.ll [new file with mode: 0644] blob
Domain: System / Toolchain;