[AMDGPU][NewPM] Port AMDGPUOpenCLEnqueuedBlockLowering to NPM #122434

optimisan · 2025-01-10T09:38:31Z

No description provided.

llvmbot · 2025-01-10T09:39:02Z

@llvm/pr-subscribers-backend-amdgpu

Author: Akshat Oke (optimisan)

Changes

Since this is an IR pass, it could be tested as a standalone pass with opt - but it cannot be since the generic PassBuilder does not know about it in opt.

I tested if it is working in the new pipeline with llc by passing -stop-after=amdgpu-lower-enqueued-block.

Full diff: https://github.com/llvm/llvm-project/pull/122434.diff

5 Files Affected:

(modified) llvm/lib/Target/AMDGPU/AMDGPU.h (+3-3)
(modified) llvm/lib/Target/AMDGPU/AMDGPUOpenCLEnqueuedBlockLowering.cpp (+28-9)
(added) llvm/lib/Target/AMDGPU/AMDGPUOpenCLEnqueuedBlockLowering.h (+23)
(modified) llvm/lib/Target/AMDGPU/AMDGPUPassRegistry.def (+1)
(modified) llvm/lib/Target/AMDGPU/AMDGPUTargetMachine.cpp (+4-3)

diff --git a/llvm/lib/Target/AMDGPU/AMDGPU.h b/llvm/lib/Target/AMDGPU/AMDGPU.h
index b9769a1baf4d17..f7a7e48295ae5b 100644
--- a/llvm/lib/Target/AMDGPU/AMDGPU.h
+++ b/llvm/lib/Target/AMDGPU/AMDGPU.h
@@ -440,9 +440,9 @@ void initializeAMDGPUExternalAAWrapperPass(PassRegistry&);
 
 void initializeAMDGPUArgumentUsageInfoPass(PassRegistry &);
 
-ModulePass *createAMDGPUOpenCLEnqueuedBlockLoweringPass();
-void initializeAMDGPUOpenCLEnqueuedBlockLoweringPass(PassRegistry &);
-extern char &AMDGPUOpenCLEnqueuedBlockLoweringID;
+ModulePass *createAMDGPUOpenCLEnqueuedBlockLoweringLegacyPass();
+void initializeAMDGPUOpenCLEnqueuedBlockLoweringLegacyPass(PassRegistry &);
+extern char &AMDGPUOpenCLEnqueuedBlockLoweringLegacyID;
 
 void initializeGCNNSAReassignPass(PassRegistry &);
 extern char &GCNNSAReassignID;
diff --git a/llvm/lib/Target/AMDGPU/AMDGPUOpenCLEnqueuedBlockLowering.cpp b/llvm/lib/Target/AMDGPU/AMDGPUOpenCLEnqueuedBlockLowering.cpp
index 4f5ca08b46c139..fbd15ad176e3bd 100644
--- a/llvm/lib/Target/AMDGPU/AMDGPUOpenCLEnqueuedBlockLowering.cpp
+++ b/llvm/lib/Target/AMDGPU/AMDGPUOpenCLEnqueuedBlockLowering.cpp
@@ -31,6 +31,7 @@
 //
 //===----------------------------------------------------------------------===//
 
+#include "AMDGPUOpenCLEnqueuedBlockLowering.h"
 #include "AMDGPU.h"
 #include "llvm/ADT/DenseSet.h"
 #include "llvm/ADT/SmallString.h"
@@ -48,11 +49,16 @@ using namespace llvm;
 namespace {
 
 /// Lower enqueued blocks.
-class AMDGPUOpenCLEnqueuedBlockLowering : public ModulePass {
+class AMDGPUOpenCLEnqueuedBlockLowering {
+public:
+  bool run(Module &M);
+};
+
+class AMDGPUOpenCLEnqueuedBlockLoweringLegacy : public ModulePass {
 public:
   static char ID;
 
-  explicit AMDGPUOpenCLEnqueuedBlockLowering() : ModulePass(ID) {}
+  explicit AMDGPUOpenCLEnqueuedBlockLoweringLegacy() : ModulePass(ID) {}
 
 private:
   bool runOnModule(Module &M) override;
@@ -60,19 +66,32 @@ class AMDGPUOpenCLEnqueuedBlockLowering : public ModulePass {
 
 } // end anonymous namespace
 
-char AMDGPUOpenCLEnqueuedBlockLowering::ID = 0;
+char AMDGPUOpenCLEnqueuedBlockLoweringLegacy::ID = 0;
 
-char &llvm::AMDGPUOpenCLEnqueuedBlockLoweringID =
-    AMDGPUOpenCLEnqueuedBlockLowering::ID;
+char &llvm::AMDGPUOpenCLEnqueuedBlockLoweringLegacyID =
+    AMDGPUOpenCLEnqueuedBlockLoweringLegacy::ID;
 
-INITIALIZE_PASS(AMDGPUOpenCLEnqueuedBlockLowering, DEBUG_TYPE,
+INITIALIZE_PASS(AMDGPUOpenCLEnqueuedBlockLoweringLegacy, DEBUG_TYPE,
                 "Lower OpenCL enqueued blocks", false, false)
 
-ModulePass* llvm::createAMDGPUOpenCLEnqueuedBlockLoweringPass() {
-  return new AMDGPUOpenCLEnqueuedBlockLowering();
+ModulePass *llvm::createAMDGPUOpenCLEnqueuedBlockLoweringLegacyPass() {
+  return new AMDGPUOpenCLEnqueuedBlockLoweringLegacy();
+}
+
+bool AMDGPUOpenCLEnqueuedBlockLoweringLegacy::runOnModule(Module &M) {
+  AMDGPUOpenCLEnqueuedBlockLowering Impl;
+  return Impl.run(M);
+}
+
+PreservedAnalyses
+AMDGPUOpenCLEnqueuedBlockLoweringPass::run(Module &M, ModuleAnalysisManager &) {
+  AMDGPUOpenCLEnqueuedBlockLowering Impl;
+  if (Impl.run(M))
+    return PreservedAnalyses::none();
+  return PreservedAnalyses::all();
 }
 
-bool AMDGPUOpenCLEnqueuedBlockLowering::runOnModule(Module &M) {
+bool AMDGPUOpenCLEnqueuedBlockLowering::run(Module &M) {
   DenseSet<Function *> Callers;
   auto &C = M.getContext();
   bool Changed = false;
diff --git a/llvm/lib/Target/AMDGPU/AMDGPUOpenCLEnqueuedBlockLowering.h b/llvm/lib/Target/AMDGPU/AMDGPUOpenCLEnqueuedBlockLowering.h
new file mode 100644
index 00000000000000..16ed7c18d8523e
--- /dev/null
+++ b/llvm/lib/Target/AMDGPU/AMDGPUOpenCLEnqueuedBlockLowering.h
@@ -0,0 +1,23 @@
+//===- AMDGPUOpenCLEnqueuedBlockLowering.h -----------------------*- C++-*-===//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+//
+//===----------------------------------------------------------------------===//
+
+#ifndef LLVM_LIB_TARGET_AMDGPU_OPENCLENQUEUEDBLOCKLOWERING_H
+#define LLVM_LIB_TARGET_AMDGPU_OPENCLENQUEUEDBLOCKLOWERING_H
+
+#include "llvm/IR/PassManager.h"
+
+namespace llvm {
+class AMDGPUOpenCLEnqueuedBlockLoweringPass
+    : public PassInfoMixin<AMDGPUOpenCLEnqueuedBlockLoweringPass> {
+public:
+  AMDGPUOpenCLEnqueuedBlockLoweringPass() = default;
+  PreservedAnalyses run(Module &M, ModuleAnalysisManager &MAM);
+};
+} // namespace llvm
+
+#endif // LLVM_LIB_TARGET_AMDGPU_OPENCLENQUEUEDBLOCKLOWERING_H
diff --git a/llvm/lib/Target/AMDGPU/AMDGPUPassRegistry.def b/llvm/lib/Target/AMDGPU/AMDGPUPassRegistry.def
index 182e825a59a41b..1e04f30fa3e7dc 100644
--- a/llvm/lib/Target/AMDGPU/AMDGPUPassRegistry.def
+++ b/llvm/lib/Target/AMDGPU/AMDGPUPassRegistry.def
@@ -22,6 +22,7 @@ MODULE_PASS("amdgpu-lower-buffer-fat-pointers",
 MODULE_PASS("amdgpu-lower-ctor-dtor", AMDGPUCtorDtorLoweringPass())
 MODULE_PASS("amdgpu-sw-lower-lds", AMDGPUSwLowerLDSPass(*this))
 MODULE_PASS("amdgpu-lower-module-lds", AMDGPULowerModuleLDSPass(*this))
+MODULE_PASS("amdgpu-lower-enqueued-block", AMDGPUOpenCLEnqueuedBlockLoweringPass())
 MODULE_PASS("amdgpu-perf-hint",
             AMDGPUPerfHintAnalysisPass(
               *static_cast<const GCNTargetMachine *>(this)))
diff --git a/llvm/lib/Target/AMDGPU/AMDGPUTargetMachine.cpp b/llvm/lib/Target/AMDGPU/AMDGPUTargetMachine.cpp
index 7256eec89008a5..2cba3860cfce09 100644
--- a/llvm/lib/Target/AMDGPU/AMDGPUTargetMachine.cpp
+++ b/llvm/lib/Target/AMDGPU/AMDGPUTargetMachine.cpp
@@ -22,6 +22,7 @@
 #include "AMDGPUIGroupLP.h"
 #include "AMDGPUISelDAGToDAG.h"
 #include "AMDGPUMacroFusion.h"
+#include "AMDGPUOpenCLEnqueuedBlockLowering.h"
 #include "AMDGPUPerfHintAnalysis.h"
 #include "AMDGPUSplitModule.h"
 #include "AMDGPUTargetObjectFile.h"
@@ -499,7 +500,7 @@ extern "C" LLVM_EXTERNAL_VISIBILITY void LLVMInitializeAMDGPUTarget() {
   initializeAMDGPULowerKernelArgumentsPass(*PR);
   initializeAMDGPUPromoteKernelArgumentsPass(*PR);
   initializeAMDGPULowerKernelAttributesPass(*PR);
-  initializeAMDGPUOpenCLEnqueuedBlockLoweringPass(*PR);
+  initializeAMDGPUOpenCLEnqueuedBlockLoweringLegacyPass(*PR);
   initializeAMDGPUPostLegalizerCombinerPass(*PR);
   initializeAMDGPUPreLegalizerCombinerPass(*PR);
   initializeAMDGPURegBankCombinerPass(*PR);
@@ -1172,7 +1173,7 @@ void AMDGPUPassConfig::addIRPasses() {
     addPass(createR600OpenCLImageTypeLoweringPass());
 
   // Replace OpenCL enqueued block function pointers with global variables.
-  addPass(createAMDGPUOpenCLEnqueuedBlockLoweringPass());
+  addPass(createAMDGPUOpenCLEnqueuedBlockLoweringLegacyPass());
 
   // Lower LDS accesses to global memory pass if address sanitizer is enabled.
   if (EnableSwLowerLDS)
@@ -1939,7 +1940,7 @@ void AMDGPUCodeGenPassBuilder::addIRPasses(AddIRPass &addPass) const {
   addPass(AMDGPUAlwaysInlinePass());
   addPass(AlwaysInlinerPass());
 
-  // TODO: Missing OpenCLEnqueuedBlockLowering
+  addPass(AMDGPUOpenCLEnqueuedBlockLoweringPass());
 
   // Runs before PromoteAlloca so the latter can account for function uses
   if (EnableLowerModuleLDS)

arsenm · 2025-01-10T12:56:13Z

Since this is an IR pass, it could be tested as a standalone pass with opt - but it cannot be since the generic PassBuilder does not know about it in opt.

It can be tested, we already test many backend IR passes this way

optimisan · 2025-01-10T13:16:15Z

Oh I see, I did not mention the target triple. Updating the test.

arsenm · 2025-01-13T04:25:42Z

llvm/lib/Target/AMDGPU/AMDGPUPassRegistry.def

            AMDGPULowerBufferFatPointersPass(*this))
 MODULE_PASS("amdgpu-lower-ctor-dtor", AMDGPUCtorDtorLoweringPass())
 MODULE_PASS("amdgpu-sw-lower-lds", AMDGPUSwLowerLDSPass(*this))
+MODULE_PASS("amdgpu-lower-enqueued-block", AMDGPUOpenCLEnqueuedBlockLoweringPass())


amdgpu-sw-lower-lds isn't in alphabetical order but that's already wrong

cdevadas · 2025-01-13T04:13:59Z

llvm/lib/Target/AMDGPU/AMDGPUPassRegistry.def

 MODULE_PASS("amdgpu-lower-buffer-fat-pointers",
            AMDGPULowerBufferFatPointersPass(*this))
 MODULE_PASS("amdgpu-lower-ctor-dtor", AMDGPUCtorDtorLoweringPass())
 MODULE_PASS("amdgpu-sw-lower-lds", AMDGPUSwLowerLDSPass(*this))


Should move this down in a separate PR.

llvm-ci · 2025-01-13T13:44:33Z

LLVM Buildbot has detected a new failure on builder llvm-clang-x86_64-expensive-checks-debian running on gribozavr4 while building llvm at step 6 "test-build-unified-tree-check-all".

Full details are available at: https://lab.llvm.org/buildbot/#/builders/16/builds/11876

Here is the relevant piece of the build log for the reference

Step 6 (test-build-unified-tree-check-all) failure: test (failure)
******************** TEST 'LLVM :: tools/llvm-gsymutil/ARM_AArch64/macho-merged-funcs-dwarf.yaml' FAILED ********************
Exit Code: 1

Command Output (stdout):
--
Input file: /b/1/llvm-clang-x86_64-expensive-checks-debian/build/test/tools/llvm-gsymutil/ARM_AArch64/Output/macho-merged-funcs-dwarf.yaml.tmp.dSYM
Output file (aarch64): /b/1/llvm-clang-x86_64-expensive-checks-debian/build/test/tools/llvm-gsymutil/ARM_AArch64/Output/macho-merged-funcs-dwarf.yaml.tmp.default.gSYM
Loaded 3 functions from DWARF.
Loaded 3 functions from symbol table.
warning: same address range contains different debug info. Removing:
[0x0000000000000248 - 0x0000000000000270): Name=0x00000047
addr=0x0000000000000248, file=  3, line=  5
addr=0x0000000000000254, file=  3, line=  7
addr=0x0000000000000258, file=  3, line=  9
addr=0x000000000000025c, file=  3, line=  8
addr=0x0000000000000260, file=  3, line= 11
addr=0x0000000000000264, file=  3, line= 10
addr=0x0000000000000268, file=  3, line=  6


In favor of this one:
[0x0000000000000248 - 0x0000000000000270): Name=0x00000030
addr=0x0000000000000248, file=  2, line=  5
addr=0x0000000000000254, file=  2, line=  7
addr=0x0000000000000258, file=  2, line=  9
addr=0x000000000000025c, file=  2, line=  8
addr=0x0000000000000260, file=  2, line= 11
addr=0x0000000000000264, file=  2, line= 10
addr=0x0000000000000268, file=  2, line=  6


warning: same address range contains different debug info. Removing:
[0x0000000000000248 - 0x0000000000000270): Name=0x00000030
addr=0x0000000000000248, file=  2, line=  5
addr=0x0000000000000254, file=  2, line=  7
addr=0x0000000000000258, file=  2, line=  9
addr=0x000000000000025c, file=  2, line=  8
addr=0x0000000000000260, file=  2, line= 11
addr=0x0000000000000264, file=  2, line= 10
addr=0x0000000000000268, file=  2, line=  6


In favor of this one:
[0x0000000000000248 - 0x0000000000000270): Name=0x00000001
addr=0x0000000000000248, file=  1, line=  5
addr=0x0000000000000254, file=  1, line=  7
addr=0x0000000000000258, file=  1, line=  9
addr=0x000000000000025c, file=  1, line=  8
addr=0x0000000000000260, file=  1, line= 11
addr=0x0000000000000264, file=  1, line= 10
...

…22434)

optimisan added 3 commits January 10, 2025 09:31

[AMDGPU][NewPM] Port AMDGPUOpenCLEnqueuedBlockLowering to NPM

fb045fe

remove TODO

3d7490d

80 column width fence

0c71adf

llvmbot added the backend:AMDGPU label Jan 10, 2025

optimisan requested review from arsenm and cdevadas January 10, 2025 09:46

alphabetize

993196c

add test for -passes

e134f9d

arsenm approved these changes Jan 13, 2025

View reviewed changes

cdevadas approved these changes Jan 13, 2025

View reviewed changes

optimisan merged commit 73b0e8a into llvm:main Jan 13, 2025
8 checks passed

kazutakahirata pushed a commit to kazutakahirata/llvm-project that referenced this pull request Jan 13, 2025

[AMDGPU][NewPM] Port AMDGPUOpenCLEnqueuedBlockLowering to NPM (llvm#1…

7fcfbd8

…22434)

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

[AMDGPU][NewPM] Port AMDGPUOpenCLEnqueuedBlockLowering to NPM #122434

[AMDGPU][NewPM] Port AMDGPUOpenCLEnqueuedBlockLowering to NPM #122434

Uh oh!

optimisan commented Jan 10, 2025 •

edited

Loading

Uh oh!

llvmbot commented Jan 10, 2025

Uh oh!

arsenm commented Jan 10, 2025

Uh oh!

optimisan commented Jan 10, 2025

Uh oh!

arsenm Jan 13, 2025

Uh oh!

cdevadas Jan 13, 2025

Uh oh!

Uh oh!

llvm-ci commented Jan 13, 2025

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

5 participants

[AMDGPU][NewPM] Port AMDGPUOpenCLEnqueuedBlockLowering to NPM #122434

[AMDGPU][NewPM] Port AMDGPUOpenCLEnqueuedBlockLowering to NPM #122434

Uh oh!

Conversation

optimisan commented Jan 10, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

llvmbot commented Jan 10, 2025

Uh oh!

arsenm commented Jan 10, 2025

Uh oh!

optimisan commented Jan 10, 2025

Uh oh!

arsenm Jan 13, 2025

Choose a reason for hiding this comment

Uh oh!

cdevadas Jan 13, 2025

Choose a reason for hiding this comment

Uh oh!

Uh oh!

llvm-ci commented Jan 13, 2025

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

5 participants

optimisan commented Jan 10, 2025 •

edited

Loading