diff options
| author | Abinav Puthan Purayil <abinavpp@gmail.com> | 2022-07-13 12:10:02 +0530 |
|---|---|---|
| committer | Abinav Puthan Purayil <abinavpp@gmail.com> | 2022-07-19 15:16:28 +0530 |
| commit | 9fa425c1ab2f763da14953f22f005dd1ca581c1c (patch) | |
| tree | 26f1b60d6a225fec2f0a060715141552f7c4ae03 /llvm/docs/CommandGuide | |
| parent | 4baf8f092b47f4f31bda96a7acb7169d389c96fd (diff) | |
| download | llvm-9fa425c1ab2f763da14953f22f005dd1ca581c1c.zip llvm-9fa425c1ab2f763da14953f22f005dd1ca581c1c.tar.gz llvm-9fa425c1ab2f763da14953f22f005dd1ca581c1c.tar.bz2 | |
[AMDGPU] Set amdgpu-memory-bound if a basic block has dense global memory access
AMDGPUPerfHintAnalysis doesn't set the memory bound attribute if
FuncInfo::InstCost outweighs MemInstCost even if we have a basic block
with relatively high global memory access. GCNSchedStrategy could revert
optimal scheduling in favour of occupancy which seems to degrade
performance for some kernels. This change introduces the
HasDenseGlobalMemAcc metric in the heuristic that makes the analysis
more conservative in these cases.
This fixes SWDEV-334259/SWDEV-343932
Differential Revision: https://reviews.llvm.org/D129759
Diffstat (limited to 'llvm/docs/CommandGuide')
0 files changed, 0 insertions, 0 deletions