Merge OpenAI Triton commit 9182231 (#4601)

This PR change the Triton base from
e21efcb to
9182231 (Jun 25).
Pass rate: 97.12%

Author: whitneywhtsang

cmake/llvm-hash.txt

@@ -1 +1 @@
-8957e64a20fc7f4277565c6cfe3e555c119783ce
+570885128351868c1308bb22e8ca351d318bc4a1

lib/Conversion/TritonGPUToLLVM/FuncOpToLLVM.cpp

@@ -1,15 +1,9 @@
+#include "mlir/Conversion/FuncToLLVM/ConvertFuncToLLVM.h"
 #include "mlir/Dialect/LLVMIR/NVVMDialect.h"
 #include "mlir/IR/BuiltinAttributes.h"
 #include "triton/Conversion/TritonGPUToLLVM/PatternTritonGPUOpToLLVM.h"
 #include "triton/Conversion/TritonGPUToLLVM/Utility.h"
 
-namespace mlir {
-FailureOr<LLVM::LLVMFuncOp>
-convertFuncOpToLLVMFuncOp(FunctionOpInterface funcOp,
-                          ConversionPatternRewriter &rewriter,
-                          const LLVMTypeConverter &converter);
-}
-
 namespace {
 
 using namespace mlir;

third_party/amd/lib/TritonAMDGPUToLLVM/ScalarizePackedFOps.cpp

@@ -1,4 +1,5 @@
 #include "TritonAMDGPUToLLVM/Passes.h"
+#include "llvm/IR/IRBuilder.h"
 #include "llvm/IR/Instructions.h"
 #include "llvm/IR/PatternMatch.h"
 #include "llvm/IR/Verifier.h"
@@ -79,7 +80,7 @@ struct ScalarizePackedFOps : FunctionPass {
   ScalarizePackedFOps() : FunctionPass(ID) {}
 
   bool runOnFunction(Function &F) override {
-    IRBuilder builder(F.getContext());
+    IRBuilder<> builder(F.getContext());
     bool changed = false;
     SmallVector<Instruction *> instsToErase;
     for (BasicBlock &BB : F) {

third_party/intel/lib/TritonIntelGPUToLLVM/PipelineManager.h

@@ -38,7 +38,8 @@ namespace mlir {
 FailureOr<LLVM::LLVMFuncOp>
 convertFuncOpToLLVMFuncOp(FunctionOpInterface funcOp,
                           ConversionPatternRewriter &rewriter,
-                          const LLVMTypeConverter &converter);
+                          const LLVMTypeConverter &converter,
+                          SymbolTableCollection *symbolTables = nullptr);
 }
 
 namespace mlir::triton::intel {

third_party/intel/lib/TritonIntelGPUToLLVM/TritonGPUToLLVM.cpp

@@ -25,13 +25,6 @@ namespace mlir::triton::gpu::intel {
 #include "intel/include/TritonIntelGPUToLLVM/Passes.h.inc"
 } // namespace mlir::triton::gpu::intel
 
-namespace mlir {
-FailureOr<LLVM::LLVMFuncOp>
-convertFuncOpToLLVMFuncOp(FunctionOpInterface funcOp,
-                          ConversionPatternRewriter &rewriter,
-                          const LLVMTypeConverter &converter);
-}
-
 using namespace mlir;
 
 namespace {

third_party/intel/triton_xpu.cc

@@ -3,6 +3,7 @@
 #include "passes.h"
 
 #include "llvm/IR/InstIterator.h"
+#include "llvm/IR/Operator.h"
 #include "llvm/IRReader/IRReader.h"
 #include "llvm/Passes/PassBuilder.h"
 #include "llvm/Passes/PassPlugin.h"