diff options
| author | Stanislav Mekhanoshin <Stanislav.Mekhanoshin@amd.com> | 2021-07-08 12:23:52 -0700 |
|---|---|---|
| committer | Stanislav Mekhanoshin <Stanislav.Mekhanoshin@amd.com> | 2021-07-21 12:46:10 -0700 |
| commit | a397c1c82f1c49106d1459c8f755e4a52743b882 (patch) | |
| tree | 4f69879621382cc779e5de3efcd60c88ef0f58e7 /llvm/lib/Support/OptimizedStructLayout.cpp | |
| parent | a467c085707c7b0b3110c21fa35ee8ddde73426e (diff) | |
| download | llvm-a397c1c82f1c49106d1459c8f755e4a52743b882.zip llvm-a397c1c82f1c49106d1459c8f755e4a52743b882.tar.gz llvm-a397c1c82f1c49106d1459c8f755e4a52743b882.tar.bz2 | |
[AMDGPU] Tune perfhint analysis to account access width
A function with less memory instructions but wider access
is the same as a function with more but narrower accesses
in terms of memory boundness. In fact the pass would give
different answers before and after vectorization without
this change.
Differential Revision: https://reviews.llvm.org/D105651
Diffstat (limited to 'llvm/lib/Support/OptimizedStructLayout.cpp')
0 files changed, 0 insertions, 0 deletions