Skip to content

release/20.x: [GlobalISel] Do not run verifier after ResetMachineFunctionPass (#124799) #124917

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 1 commit into from
Feb 1, 2025

Conversation

llvmbot
Copy link
Member

@llvmbot llvmbot commented Jan 29, 2025

Backport 66e0498

Requested by: @davemgreen

@llvmbot llvmbot added this to the LLVM 20.X Release milestone Jan 29, 2025
@llvmbot
Copy link
Member Author

llvmbot commented Jan 29, 2025

@arsenm What do you think about merging this PR to the release branch?

@llvmbot
Copy link
Member Author

llvmbot commented Jan 29, 2025

@llvm/pr-subscribers-backend-amdgpu

Author: None (llvmbot)

Changes

Backport 66e0498

Requested by: @davemgreen


Full diff: https://github.com/llvm/llvm-project/pull/124917.diff

2 Files Affected:

  • (modified) llvm/lib/CodeGen/TargetPassConfig.cpp (+4-2)
  • (modified) llvm/test/CodeGen/AMDGPU/llvm.amdgcn.mfma.gfx950.ll (+1-1)
diff --git a/llvm/lib/CodeGen/TargetPassConfig.cpp b/llvm/lib/CodeGen/TargetPassConfig.cpp
index 847a1aef39c5655..cf407b5073035e7 100644
--- a/llvm/lib/CodeGen/TargetPassConfig.cpp
+++ b/llvm/lib/CodeGen/TargetPassConfig.cpp
@@ -1039,11 +1039,13 @@ bool TargetPassConfig::addCoreISelPasses() {
 
     if (addGlobalInstructionSelect())
       return true;
+  }
 
-    // Pass to reset the MachineFunction if the ISel failed.
+  // Pass to reset the MachineFunction if the ISel failed. Outside of the above
+  // if so that the verifier is not added to it.
+  if (Selector == SelectorType::GlobalISel)
     addPass(createResetMachineFunctionPass(
         reportDiagnosticWhenGlobalISelFallback(), isGlobalISelAbortEnabled()));
-  }
 
   // Run the SDAG InstSelector, providing a fallback path when we do not want to
   // abort on not-yet-supported input.
diff --git a/llvm/test/CodeGen/AMDGPU/llvm.amdgcn.mfma.gfx950.ll b/llvm/test/CodeGen/AMDGPU/llvm.amdgcn.mfma.gfx950.ll
index d9ee276c3f076e6..44cb4e803ffad66 100644
--- a/llvm/test/CodeGen/AMDGPU/llvm.amdgcn.mfma.gfx950.ll
+++ b/llvm/test/CodeGen/AMDGPU/llvm.amdgcn.mfma.gfx950.ll
@@ -1,6 +1,6 @@
 ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py UTC_ARGS: --version 4
 ; RUN: llc -mtriple=amdgcn -mcpu=gfx950 -global-isel=0 < %s | FileCheck -enable-var-scope --check-prefixes=GCN,SDAG %s
-; RUN: llc -mtriple=amdgcn -mcpu=gfx950 -global-isel=1 -global-isel-abort=2 < %s | FileCheck -enable-var-scope --check-prefixes=GCN,GISEL %s
+; RUN: llc -mtriple=amdgcn -mcpu=gfx950 -global-isel=1 -global-isel-abort=2 -verify-machineinstrs < %s | FileCheck -enable-var-scope --check-prefixes=GCN,GISEL %s
 
 declare <4 x float> @llvm.amdgcn.mfma.f32.16x16x32.f16(<8 x half>, <8 x half>, <4 x float>, i32 immarg, i32 immarg, i32 immarg)
 declare <16 x float> @llvm.amdgcn.mfma.f32.32x32x16.f16(<8 x half>, <8 x half>, <16 x float>, i32 immarg, i32 immarg, i32 immarg)

…#124799)

After we fall back from GlobalISel to SDAG, the verifier gets called,
which calls getReservedRegs which uses SIMachineFunctionInfo::usesAGPRs
which caches the result of UsesAGPRs. Because we have just fallen-back
the function is empty and it incorrectly gets cached to false. This
patch makes sure we don't try to run the verifier whilst the function is
empty.

(cherry picked from commit 66e0498)
@tstellar tstellar merged commit b23297a into llvm:release/20.x Feb 1, 2025
11 of 13 checks passed
Copy link

github-actions bot commented Feb 1, 2025

@davemgreen (or anyone else). If you would like to add a note about this fix in the release notes (completely optional). Please reply to this comment with a one or two sentence description of the fix. When you are done, please add the release:note label to this PR.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
Development

Successfully merging this pull request may close these issues.

4 participants