diff --git a/compiler/src/iree/compiler/Codegen/Common/EmulateNarrowType.cpp b/compiler/src/iree/compiler/Codegen/Common/EmulateNarrowType.cpp
index 1824fb08bb17..e2c8805d270c 100644
--- a/compiler/src/iree/compiler/Codegen/Common/EmulateNarrowType.cpp
+++ b/compiler/src/iree/compiler/Codegen/Common/EmulateNarrowType.cpp
@@ -146,7 +146,7 @@ struct EmulateNarrowTypePass final
     }
 
     RewritePatternSet sinkBroadcast(ctx);
-    vector::populateSinkVectorBroadcastPatterns(sinkBroadcast);
+    vector::populateSinkVectorOpsPatterns(sinkBroadcast);
     if (failed(applyPatternsAndFoldGreedily(getOperation(),
                                             std::move(sinkBroadcast)))) {
       getOperation()->emitOpError("failed in sinking of broadcasts");
diff --git a/compiler/src/iree/compiler/Codegen/Common/GenericVectorization.cpp b/compiler/src/iree/compiler/Codegen/Common/GenericVectorization.cpp
index 98814d134268..8aee5ba2c0e4 100644
--- a/compiler/src/iree/compiler/Codegen/Common/GenericVectorization.cpp
+++ b/compiler/src/iree/compiler/Codegen/Common/GenericVectorization.cpp
@@ -409,6 +409,7 @@ void GenericVectorizationPass::runOnOperation() {
     vector::populateVectorTransferPermutationMapLoweringPatterns(
         vectorizationPatterns);
     vector::populateVectorReductionToContractPatterns(vectorizationPatterns);
+    vector::populateSinkVectorOpsPatterns(vectorizationPatterns);
   }
   if (foldCastIntoContract) {
     vector::populateFoldArithExtensionPatterns(vectorizationPatterns);
diff --git a/compiler/src/iree/compiler/Codegen/LLVMGPU/ConvertToNVVM.cpp b/compiler/src/iree/compiler/Codegen/LLVMGPU/ConvertToNVVM.cpp
index b1ac58b1b62f..16cc62f9a734 100644
--- a/compiler/src/iree/compiler/Codegen/LLVMGPU/ConvertToNVVM.cpp
+++ b/compiler/src/iree/compiler/Codegen/LLVMGPU/ConvertToNVVM.cpp
@@ -167,7 +167,7 @@ struct ConvertToNVVMPass final
           LLVM::RoundEvenOp, LLVM::RoundOp, LLVM::SinOp, LLVM::SqrtOp>();
 
       // TODO: Remove once we support replacing non-root ops.
-      target.addLegalOp<gpu::YieldOp, gpu::GPUModuleOp, gpu::ModuleEndOp>();
+      target.addLegalOp<gpu::YieldOp, gpu::GPUModuleOp>();
 
       if (failed(applyPartialConversion(m, target, std::move(llvmPatterns)))) {
         signalPassFailure();
diff --git a/compiler/src/iree/compiler/Codegen/LLVMGPU/LLVMGPUTensorCoreVectorization.cpp b/compiler/src/iree/compiler/Codegen/LLVMGPU/LLVMGPUTensorCoreVectorization.cpp
index aa72280e5fcc..f00af4284543 100644
--- a/compiler/src/iree/compiler/Codegen/LLVMGPU/LLVMGPUTensorCoreVectorization.cpp
+++ b/compiler/src/iree/compiler/Codegen/LLVMGPU/LLVMGPUTensorCoreVectorization.cpp
@@ -96,6 +96,7 @@ class LLVMGPUTensorCoreVectorizationPass final
       vector::populateVectorTransferPermutationMapLoweringPatterns(
           contractionPatterns);
       vector::populateVectorReductionToContractPatterns(contractionPatterns);
+      vector::populateSinkVectorOpsPatterns(contractionPatterns);
       if (failed(applyPatternsAndFoldGreedily(
               funcOp, std::move(contractionPatterns)))) {
         return signalPassFailure();
diff --git a/compiler/src/iree/compiler/Codegen/SPIRV/SPIRVInitialVectorLowering.cpp b/compiler/src/iree/compiler/Codegen/SPIRV/SPIRVInitialVectorLowering.cpp
index 493894675cd3..c24da496891d 100644
--- a/compiler/src/iree/compiler/Codegen/SPIRV/SPIRVInitialVectorLowering.cpp
+++ b/compiler/src/iree/compiler/Codegen/SPIRV/SPIRVInitialVectorLowering.cpp
@@ -314,6 +314,7 @@ class SPIRVInitialLoweringPass final
       // cancel them or embed into contract ops. Embedding in the flexible
       // contract ops will help to sustain the structure through various
       // transformations.
+      vector::populateSinkVectorOpsPatterns(patterns);
       vector::populateVectorReductionToContractPatterns(patterns);
       // Pull in patterns to canonicalize transfer ops.
       vector::populateVectorTransferPermutationMapLoweringPatterns(patterns);
diff --git a/compiler/src/iree/compiler/Tools/BUILD.bazel b/compiler/src/iree/compiler/Tools/BUILD.bazel
index ee5d60dccfff..94d655486068 100644
--- a/compiler/src/iree/compiler/Tools/BUILD.bazel
+++ b/compiler/src/iree/compiler/Tools/BUILD.bazel
@@ -105,6 +105,7 @@ iree_compiler_cc_library(
         "@llvm-project//mlir:GPUTransforms",
         "@llvm-project//mlir:IR",
         "@llvm-project//mlir:LLVMDialect",
+        "@llvm-project//mlir:LLVMIRTransforms",
         "@llvm-project//mlir:LinalgDialect",
         "@llvm-project//mlir:LinalgPassIncGen",
         "@llvm-project//mlir:LinalgTransforms",
diff --git a/compiler/src/iree/compiler/Tools/CMakeLists.txt b/compiler/src/iree/compiler/Tools/CMakeLists.txt
index c8531f6ca1f5..d9033ba99cee 100644
--- a/compiler/src/iree/compiler/Tools/CMakeLists.txt
+++ b/compiler/src/iree/compiler/Tools/CMakeLists.txt
@@ -90,6 +90,7 @@ iree_cc_library(
     MLIRGPUToSPIRV
     MLIRIR
     MLIRLLVMDialect
+    MLIRLLVMIRTransforms
     MLIRLinalgDialect
     MLIRLinalgTransforms
     MLIRMLProgramDialect
diff --git a/compiler/src/iree/compiler/Tools/init_mlir_dialects.h b/compiler/src/iree/compiler/Tools/init_mlir_dialects.h
index 2148ff2e150a..6dde522a94ee 100644
--- a/compiler/src/iree/compiler/Tools/init_mlir_dialects.h
+++ b/compiler/src/iree/compiler/Tools/init_mlir_dialects.h
@@ -29,6 +29,7 @@
 #include "mlir/Dialect/GPU/IR/GPUDialect.h"
 #include "mlir/Dialect/GPU/TransformOps/GPUTransformOps.h"
 #include "mlir/Dialect/LLVMIR/LLVMDialect.h"
+#include "mlir/Dialect/LLVMIR/Transforms/InlinerInterfaceImpl.h"
 #include "mlir/Dialect/Linalg/IR/Linalg.h"
 #include "mlir/Dialect/Linalg/TransformOps/DialectExtension.h"
 #include "mlir/Dialect/MLProgram/IR/MLProgram.h"
@@ -94,6 +95,7 @@ inline void registerMlirDialects(DialectRegistry &registry) {
   // clang-format on
   cf::registerBufferizableOpInterfaceExternalModels(registry);
   func::registerInlinerExtension(registry);
+  LLVM::registerInlinerInterface(registry);
   tensor::registerInferTypeOpInterfaceExternalModels(registry);
   tensor::registerTilingInterfaceExternalModels(registry);
 
diff --git a/llvm-external-projects/iree-dialects/BUILD.bazel b/llvm-external-projects/iree-dialects/BUILD.bazel
index 680c4b038013..f1c279450fff 100644
--- a/llvm-external-projects/iree-dialects/BUILD.bazel
+++ b/llvm-external-projects/iree-dialects/BUILD.bazel
@@ -400,6 +400,8 @@ cc_binary(
         "@llvm-project//mlir:FuncExtensions",
         "@llvm-project//mlir:FunctionInterfaces",
         "@llvm-project//mlir:IR",
+        "@llvm-project//mlir:LLVMDialect",
+        "@llvm-project//mlir:LLVMIRTransforms",
         "@llvm-project//mlir:LinalgDialect",
         "@llvm-project//mlir:LinalgTransformOps",
         "@llvm-project//mlir:LinalgTransforms",
diff --git a/llvm-external-projects/iree-dialects/tools/iree-dialects-opt/CMakeLists.txt b/llvm-external-projects/iree-dialects/tools/iree-dialects-opt/CMakeLists.txt
index 6515519f38c4..c687de7c0f0f 100644
--- a/llvm-external-projects/iree-dialects/tools/iree-dialects-opt/CMakeLists.txt
+++ b/llvm-external-projects/iree-dialects/tools/iree-dialects-opt/CMakeLists.txt
@@ -13,6 +13,8 @@ set(LIBS
   MLIRFuncDialect
   MLIRFuncInlinerExtension
   MLIRIndexToLLVM
+  MLIRLLVMDialect
+  MLIRLLVMIRTransforms
   MLIRLinalgDialect
   MLIRLinalgTransformOps
   MLIRMemRefDialect
diff --git a/llvm-external-projects/iree-dialects/tools/iree-dialects-opt/iree-dialects-opt.cpp b/llvm-external-projects/iree-dialects/tools/iree-dialects-opt/iree-dialects-opt.cpp
index 885d5839cb48..d42bfb624d93 100644
--- a/llvm-external-projects/iree-dialects/tools/iree-dialects-opt/iree-dialects-opt.cpp
+++ b/llvm-external-projects/iree-dialects/tools/iree-dialects-opt/iree-dialects-opt.cpp
@@ -13,6 +13,7 @@
 #include "mlir/Dialect/ControlFlow/IR/ControlFlow.h"
 #include "mlir/Dialect/Func/Extensions/InlinerExtension.h"
 #include "mlir/Dialect/Func/IR/FuncOps.h"
+#include "mlir/Dialect/LLVMIR/Transforms/InlinerInterfaceImpl.h"
 #include "mlir/Dialect/Linalg/IR/Linalg.h"
 #include "mlir/Dialect/Linalg/TransformOps/DialectExtension.h"
 #include "mlir/Dialect/Linalg/TransformOps/LinalgTransformOps.h"
@@ -74,6 +75,7 @@ int main(int argc, char **argv) {
 
   // External models.
   mlir::func::registerInlinerExtension(registry);
+  mlir::LLVM::registerInlinerInterface(registry);
   mlir::linalg::registerTilingInterfaceExternalModels(registry);
 
   registry.addExtensions<transform_ext::StructuredTransformOpsExtension>();
diff --git a/third_party/llvm-project b/third_party/llvm-project
index 6b7afaa9db8f..8835909fd7a9 160000
--- a/third_party/llvm-project
+++ b/third_party/llvm-project
@@ -1 +1 @@
-Subproject commit 6b7afaa9db8f904ebf0262774e38e54b36598782
+Subproject commit 8835909fd7a905712143e3f833b1fb57e5c6d2a7