[PATCH] D70379: [AMDGPU] Reserving VGPR for future SGPR Spill
Austin Kerbow via Phabricator via llvm-commits
llvm-commits at lists.llvm.org
Thu May 7 11:55:26 PDT 2020
kerbowa added inline comments.
================
Comment at: llvm/lib/Target/AMDGPU/SIMachineFunctionInfo.cpp:315
- if (VGPRIndex == 0) {
+ if (VGPRIndex == 0 && !FuncInfo->VGPRReservedForSGPRSpill) {
LaneVGPR = TRI->findUnusedRegister(MRI, &AMDGPU::VGPR_32RegClass, MF);
----------------
saiislam wrote:
> cdevadas wrote:
> > Doesn't it limit the total allowable SGPR spills to 64?
> > What happens if more than 64 callee-saved SGPRs used in the functions?
> Yes, this patch is limited to 64 SGPR spills only.
> Does it make sense to handle that case in a separate patch?
It seems like there is a fallback to spilling to VMEM if there are no free lanes. Doesn't this also limit the number of SGPR to VGPR spills to 32 on Navi?
Repository:
rG LLVM Github Monorepo
CHANGES SINCE LAST ACTION
https://reviews.llvm.org/D70379/new/
https://reviews.llvm.org/D70379
More information about the llvm-commits
mailing list