[AMDGPU][NewPM] Port SIPostRABundler to NPM #123717
Merged
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Member
|
@llvm/pr-subscribers-backend-amdgpu Author: Akshat Oke (optimisan) ChangesFull diff: https://github.com/llvm/llvm-project/pull/123717.diff 6 Files Affected:
diff --git a/llvm/lib/Target/AMDGPU/AMDGPU.h b/llvm/lib/Target/AMDGPU/AMDGPU.h
index 5d9a830f041a74..846b7c98399a4a 100644
--- a/llvm/lib/Target/AMDGPU/AMDGPU.h
+++ b/llvm/lib/Target/AMDGPU/AMDGPU.h
@@ -428,8 +428,8 @@ extern char &SIInsertWaitcntsID;
void initializeSIFormMemoryClausesPass(PassRegistry&);
extern char &SIFormMemoryClausesID;
-void initializeSIPostRABundlerPass(PassRegistry&);
-extern char &SIPostRABundlerID;
+void initializeSIPostRABundlerLegacyPass(PassRegistry&);
+extern char &SIPostRABundlerLegacyID;
void initializeGCNCreateVOPDPass(PassRegistry &);
extern char &GCNCreateVOPDID;
diff --git a/llvm/lib/Target/AMDGPU/AMDGPUPassRegistry.def b/llvm/lib/Target/AMDGPU/AMDGPUPassRegistry.def
index 09a39d23d801b9..e0ff0dfc97fa9e 100644
--- a/llvm/lib/Target/AMDGPU/AMDGPUPassRegistry.def
+++ b/llvm/lib/Target/AMDGPU/AMDGPUPassRegistry.def
@@ -108,5 +108,6 @@ MACHINE_FUNCTION_PASS("si-opt-vgpr-liverange", SIOptimizeVGPRLiveRangePass())
MACHINE_FUNCTION_PASS("si-optimize-exec-masking", SIOptimizeExecMaskingPass())
MACHINE_FUNCTION_PASS("si-peephole-sdwa", SIPeepholeSDWAPass())
MACHINE_FUNCTION_PASS("si-pre-allocate-wwm-regs", SIPreAllocateWWMRegsPass())
+MACHINE_FUNCTION_PASS("si-post-ra-bundler", SIPostRABundlerPass())
MACHINE_FUNCTION_PASS("si-shrink-instructions", SIShrinkInstructionsPass())
#undef MACHINE_FUNCTION_PASS
diff --git a/llvm/lib/Target/AMDGPU/AMDGPUTargetMachine.cpp b/llvm/lib/Target/AMDGPU/AMDGPUTargetMachine.cpp
index 53ec80b8f72049..f3b8f5926a3aa4 100644
--- a/llvm/lib/Target/AMDGPU/AMDGPUTargetMachine.cpp
+++ b/llvm/lib/Target/AMDGPU/AMDGPUTargetMachine.cpp
@@ -45,6 +45,7 @@
#include "SIOptimizeExecMasking.h"
#include "SIOptimizeVGPRLiveRange.h"
#include "SIPeepholeSDWA.h"
+#include "SIPostRABundler.h"
#include "SIPreAllocateWWMRegs.h"
#include "SIShrinkInstructions.h"
#include "TargetInfo/AMDGPUTargetInfo.h"
@@ -532,7 +533,7 @@ extern "C" LLVM_EXTERNAL_VISIBILITY void LLVMInitializeAMDGPUTarget() {
initializeSIOptimizeExecMaskingLegacyPass(*PR);
initializeSIPreAllocateWWMRegsLegacyPass(*PR);
initializeSIFormMemoryClausesPass(*PR);
- initializeSIPostRABundlerPass(*PR);
+ initializeSIPostRABundlerLegacyPass(*PR);
initializeGCNCreateVOPDPass(*PR);
initializeAMDGPUUnifyDivergentExitNodesPass(*PR);
initializeAMDGPUAAWrapperPassPass(*PR);
@@ -1642,7 +1643,7 @@ void GCNPassConfig::addPostRegAlloc() {
void GCNPassConfig::addPreSched2() {
if (TM->getOptLevel() > CodeGenOptLevel::None)
addPass(createSIShrinkInstructionsLegacyPass());
- addPass(&SIPostRABundlerID);
+ addPass(&SIPostRABundlerLegacyID);
}
void GCNPassConfig::addPreEmitPass() {
diff --git a/llvm/lib/Target/AMDGPU/SIPostRABundler.cpp b/llvm/lib/Target/AMDGPU/SIPostRABundler.cpp
index 8464cb3d6fc43d..cbd8f3ad5c5e61 100644
--- a/llvm/lib/Target/AMDGPU/SIPostRABundler.cpp
+++ b/llvm/lib/Target/AMDGPU/SIPostRABundler.cpp
@@ -12,6 +12,7 @@
///
//===----------------------------------------------------------------------===//
+#include "SIPostRABundler.h"
#include "AMDGPU.h"
#include "GCNSubtarget.h"
#include "llvm/ADT/SmallSet.h"
@@ -23,13 +24,13 @@ using namespace llvm;
namespace {
-class SIPostRABundler : public MachineFunctionPass {
+class SIPostRABundlerLegacy : public MachineFunctionPass {
public:
static char ID;
public:
- SIPostRABundler() : MachineFunctionPass(ID) {
- initializeSIPostRABundlerPass(*PassRegistry::getPassRegistry());
+ SIPostRABundlerLegacy() : MachineFunctionPass(ID) {
+ initializeSIPostRABundlerLegacyPass(*PassRegistry::getPassRegistry());
}
bool runOnMachineFunction(MachineFunction &MF) override;
@@ -42,7 +43,11 @@ class SIPostRABundler : public MachineFunctionPass {
AU.setPreservesAll();
MachineFunctionPass::getAnalysisUsage(AU);
}
+};
+class SIPostRABundler {
+public:
+ bool run(MachineFunction &MF);
private:
const SIRegisterInfo *TRI;
@@ -62,14 +67,14 @@ constexpr uint64_t MemFlags = SIInstrFlags::MTBUF | SIInstrFlags::MUBUF |
} // End anonymous namespace.
-INITIALIZE_PASS(SIPostRABundler, DEBUG_TYPE, "SI post-RA bundler", false, false)
+INITIALIZE_PASS(SIPostRABundlerLegacy, DEBUG_TYPE, "SI post-RA bundler", false, false)
-char SIPostRABundler::ID = 0;
+char SIPostRABundlerLegacy::ID = 0;
-char &llvm::SIPostRABundlerID = SIPostRABundler::ID;
+char &llvm::SIPostRABundlerLegacyID = SIPostRABundlerLegacy::ID;
FunctionPass *llvm::createSIPostRABundlerPass() {
- return new SIPostRABundler();
+ return new SIPostRABundlerLegacy();
}
bool SIPostRABundler::isDependentLoad(const MachineInstr &MI) const {
@@ -121,9 +126,18 @@ bool SIPostRABundler::canBundle(const MachineInstr &MI,
!isDependentLoad(NextMI));
}
-bool SIPostRABundler::runOnMachineFunction(MachineFunction &MF) {
+bool SIPostRABundlerLegacy::runOnMachineFunction(MachineFunction &MF) {
if (skipFunction(MF.getFunction()))
return false;
+ return SIPostRABundler().run(MF);
+}
+
+PreservedAnalyses SIPostRABundlerPass::run(MachineFunction &MF, MachineFunctionAnalysisManager &) {
+ SIPostRABundler().run(MF);
+ return PreservedAnalyses::all();
+}
+
+bool SIPostRABundler::run(MachineFunction &MF) {
TRI = MF.getSubtarget<GCNSubtarget>().getRegisterInfo();
BitVector BundleUsedRegUnits(TRI->getNumRegUnits());
diff --git a/llvm/lib/Target/AMDGPU/SIPostRABundler.h b/llvm/lib/Target/AMDGPU/SIPostRABundler.h
new file mode 100644
index 00000000000000..2038bbe1617f4f
--- /dev/null
+++ b/llvm/lib/Target/AMDGPU/SIPostRABundler.h
@@ -0,0 +1,22 @@
+//===- SIPostRABundler.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_SIPOSTRABUNDLER_H
+#define LLVM_LIB_TARGET_AMDGPU_SIPOSTRABUNDLER_H
+
+#include "llvm/CodeGen/MachinePassManager.h"
+
+namespace llvm {
+class SIPostRABundlerPass : public PassInfoMixin<SIPostRABundlerPass> {
+public:
+ PreservedAnalyses run(MachineFunction &MF,
+ MachineFunctionAnalysisManager &MFAM);
+};
+} // namespace llvm
+
+#endif // LLVM_LIB_TARGET_AMDGPU_SIPOSTRABUNDLER_H
diff --git a/llvm/test/CodeGen/AMDGPU/postra-bundle-memops.mir b/llvm/test/CodeGen/AMDGPU/postra-bundle-memops.mir
index b878ee51cdb433..458afca384911e 100644
--- a/llvm/test/CodeGen/AMDGPU/postra-bundle-memops.mir
+++ b/llvm/test/CodeGen/AMDGPU/postra-bundle-memops.mir
@@ -1,5 +1,6 @@
# NOTE: Assertions have been autogenerated by utils/update_mir_test_checks.py
# RUN: llc -mtriple=amdgcn -mcpu=gfx900 -verify-machineinstrs -run-pass=si-post-ra-bundler %s -o - | FileCheck -check-prefix=GCN %s
+# RUN: llc -mtriple=amdgcn -mcpu=gfx900 -passes=si-post-ra-bundler %s -o - | FileCheck -check-prefix=GCN %s
---
name: bundle_memops
|
|
✅ With the latest revision this PR passed the C/C++ code formatter. |
arsenm
reviewed
Jan 21, 2025
optimisan
force-pushed
the
port-si-postra-bundler
branch
from
5a45870 to
1d828c0
Compare
arsenm
approved these changes
Feb 21, 2025
Collaborator
|
LLVM Buildbot has detected a new failure on builder Full details are available at: https://lab.llvm.org/buildbot/#/builders/73/builds/13398 Here is the relevant piece of the build log for the reference |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
4 participants
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
No description provided.