[llvm] 4c2ad82 - [AMDGPU] Do not preserve UniformityInfo (#76696)

[via llvm-commits]

Author: Sameer Sahasrabuddhe
Date: 2024-01-02T13:42:56+05:30
New Revision: 4c2ad82b9dff537f7c8f0a2abb0d5dc7e6435741

URL: https://github.com/llvm/llvm-project/commit/4c2ad82b9dff537f7c8f0a2abb0d5dc7e6435741
DIFF: https://github.com/llvm/llvm-project/commit/4c2ad82b9dff537f7c8f0a2abb0d5dc7e6435741.diff

LOG: [AMDGPU] Do not preserve UniformityInfo (#76696)

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.

Added: 
    

Modified: 
    llvm/include/llvm/ADT/GenericUniformityImpl.h
    llvm/lib/Target/AMDGPU/AMDGPURewriteUndefForPHI.cpp
    llvm/lib/Target/AMDGPU/AMDGPUUnifyDivergentExitNodes.cpp

Removed: 
    


################################################################################
diff  --git a/llvm/include/llvm/ADT/GenericUniformityImpl.h b/llvm/include/llvm/ADT/GenericUniformityImpl.h
index b7d0a1228ebfc0..d397b937d78ccc 100644
--- a/llvm/include/llvm/ADT/GenericUniformityImpl.h
+++ b/llvm/include/llvm/ADT/GenericUniformityImpl.h
@@ -33,6 +33,12 @@
 /// the propagation of the impact of divergent control flow on the divergence of
 /// values (sync dependencies).
 ///
+/// NOTE: In general, no interface exists for a transform to update
+/// (Machine)UniformityInfo. Additionally, (Machine)CycleAnalysis is a
+/// transitive dependence, but it also does not provide an interface for
+/// updating itself. Given that, transforms should not preserve uniformity in
+/// their getAnalysisUsage() callback.
+///
 //===----------------------------------------------------------------------===//
 
 #ifndef LLVM_ADT_GENERICUNIFORMITYIMPL_H

diff  --git a/llvm/lib/Target/AMDGPU/AMDGPURewriteUndefForPHI.cpp b/llvm/lib/Target/AMDGPU/AMDGPURewriteUndefForPHI.cpp
index 459400e3359ca1..79e9312034da45 100644
--- a/llvm/lib/Target/AMDGPU/AMDGPURewriteUndefForPHI.cpp
+++ b/llvm/lib/Target/AMDGPU/AMDGPURewriteUndefForPHI.cpp
@@ -85,7 +85,6 @@ class AMDGPURewriteUndefForPHILegacy : public FunctionPass {
     AU.addRequired<DominatorTreeWrapperPass>();
 
     AU.addPreserved<DominatorTreeWrapperPass>();
-    AU.addPreserved<UniformityInfoWrapperPass>();
     AU.setPreservesCFG();
   }
 };

diff  --git a/llvm/lib/Target/AMDGPU/AMDGPUUnifyDivergentExitNodes.cpp b/llvm/lib/Target/AMDGPU/AMDGPUUnifyDivergentExitNodes.cpp
index 9bc3ba161c9ebe..1bfb7c0edd80a9 100644
--- a/llvm/lib/Target/AMDGPU/AMDGPUUnifyDivergentExitNodes.cpp
+++ b/llvm/lib/Target/AMDGPU/AMDGPUUnifyDivergentExitNodes.cpp
@@ -109,9 +109,6 @@ void AMDGPUUnifyDivergentExitNodes::getAnalysisUsage(AnalysisUsage &AU) const {
     // FIXME: preserve PostDominatorTreeWrapperPass
   }
 
-  // No divergent values are changed, only blocks and branch edges.
-  AU.addPreserved<UniformityInfoWrapperPass>();
-
   // We preserve the non-critical-edgeness property
   AU.addPreservedID(BreakCriticalEdgesID);