llvm-project.git/llvm/lib/Target/AMDGPU/AMDGPUUnifyDivergentExitNodes.cpp, branch users/nico/python-2 Unnamed repository; edit this file 'description' to name the repository. [NFC][LLVM][AMDGPU] Cleanup pass initialization for AMDGPU (#134410) 2025-04-08T00:27:50+00:00 Rahul Joshi rjoshi@nvidia.com 2025-04-08T00:27:50+00:00 a3754ade637f71d36495eb95ea2d8cab9885c9d9 - Remove calls to pass initialization from pass constructors. - https://github.com/llvm/llvm-project/issues/111767 
- Remove calls to pass initialization from pass constructors.
- https://github.com/llvm/llvm-project/issues/111767
 [AMDGPU] Don't unify divergent exit nodes with `musttail` calls (#126395) 2025-02-10T02:48:24+00:00 Shilei Tian i@tianshilei.me 2025-02-10T02:48:24+00:00 967973512b9eba99dd8b04db42dbafcc50d94728 Fixes SWDEV-512254. 
Fixes SWDEV-512254.
 [AMDGPU] Remove unused includes (NFC) (#116154) 2024-11-14T05:10:03+00:00 Kazu Hirata kazu@google.com 2024-11-14T05:10:03+00:00 be187369a03bf2df8bdbc76ecd381377b3bb6074 Identified with misc-include-cleaner. 
Identified with misc-include-cleaner.
 [NFC] Rename `Intrinsic::getDeclaration` to `getOrInsertDeclaration` (#111752) 2024-10-11T12:26:03+00:00 Rahul Joshi rjoshi@nvidia.com 2024-10-11T12:26:03+00:00 fa789dffb1e12c2aece0187aeacc48dfb1768340 Rename the function to reflect its correct behavior and to be consistent with `Module::getOrInsertFunction`. This is also in preparation of adding a new `Intrinsic::getDeclaration` that will have behavior similar to `Module::getFunction` (i.e, just lookup, no creation). 
Rename the function to reflect its correct behavior and to be consistent
with `Module::getOrInsertFunction`. This is also in preparation of
adding a new `Intrinsic::getDeclaration` that will have behavior similar
to `Module::getFunction` (i.e, just lookup, no creation).
 [AMDGPU] clang-tidy: use emplace_back instead of push_back. NFC. 2024-07-17T07:27:35+00:00 Jay Foad jay.foad@amd.com 2024-07-17T07:25:15+00:00 5e338f1f4ae28b9dd7d722a77ab204e358006a86 






[AMDGPU] Do not preserve UniformityInfo (#76696)
2024-01-02T08:12:56+00:00

Sameer Sahasrabuddhe
sameer.sahasrabuddhe@amd.com

2024-01-02T08:12:56+00:00

4c2ad82b9dff537f7c8f0a2abb0d5dc7e6435741

UniformityInfo has a transitive dependence on CycleInfo. A transform may
change the CFG in trivial ways that do not affect uniformity, but that
can leave cycles in a slightly inconsistent state. In the absence of
updates to CycleInfo, it's cleaner to just invalidate both analyses.




UniformityInfo has a transitive dependence on CycleInfo. A transform may
change the CFG in trivial ways that do not affect uniformity, but that
can leave cycles in a slightly inconsistent state. In the absence of
updates to CycleInfo, it's cleaner to just invalidate both analyses.






[LowerSwitch] Don't let pass manager handle the dependency (#68662)
2023-10-25T01:24:36+00:00

Ruiling, Song
ruiling.song@amd.com

2023-10-25T01:24:36+00:00

ac24238002076e36b2d33d18d1bf47a9de59fab4

Some passes has limitation that only support simple terminators:
branch/unreachable/return. Right now, they ask the pass manager to add
LowerSwitch pass to eliminate `switch`. Let's manage such kind of pass
dependency by ourselves. Also add the assertion in the related passes.




Some passes has limitation that only support simple terminators:
branch/unreachable/return. Right now, they ask the pass manager to add
LowerSwitch pass to eliminate `switch`. Let's manage such kind of pass
dependency by ourselves. Also add the assertion in the related passes.






Revert "[AMDGPU][NFC] Preserve PDTWrapperPass in UnifyDivergentExitNodes"
2023-05-03T15:34:55+00:00

Anshil Gandhi
gandhi21299@gmail.com

2023-05-03T15:34:11+00:00

f4f6eed62749d7cf3ecffa524c6c4c3eeff6f613

This reverts commit 40ed87a0515dd72047ec8e78dcf0d8084d0891dd.





This reverts commit 40ed87a0515dd72047ec8e78dcf0d8084d0891dd.







[AMDGPU][NFC] Preserve PDTWrapperPass in UnifyDivergentExitNodes
2023-05-03T02:27:44+00:00

Anshil Gandhi
gandhi21299@gmail.com

2023-05-03T02:26:23+00:00

40ed87a0515dd72047ec8e78dcf0d8084d0891dd

Reviewed By: arsenm

Differential Revision: https://reviews.llvm.org/D149568





Reviewed By: arsenm

Differential Revision: https://reviews.llvm.org/D149568







[AMDGPUUnifyDivergentExitNodes] Add NewPM support
2023-03-25T20:04:36+00:00

Anshil Gandhi
Anshil.Gandhi@amd.com

2023-03-25T19:42:31+00:00

b48e7c2d01a329977da32ef7d5f0feba95d201f7

Meanwhile, use UniformityAnalysis instead of LegacyDivergenceAnalysis to collect divergence info.

Reviewed By: arsenm, sameerds

Differential Revision: https://reviews.llvm.org/D141355





Meanwhile, use UniformityAnalysis instead of LegacyDivergenceAnalysis to collect divergence info.

Reviewed By: arsenm, sameerds

Differential Revision: https://reviews.llvm.org/D141355