[MLIR] Move warp_execute_on_lane_0 from vector to gpu (#116994)

Please see the related RFC here:
https://discourse.llvm.org/t/rfc-move-execute-on-lane-0-from-vector-to-gpu-dialect/82989.

This patch does exactly one thing - moves the op to gpu.

1 files changed, 6 insertions, 5 deletions

diff --git a/mlir/test/lib/Dialect/Vector/TestVectorTransforms.cpp b/mlir/test/lib/Dialect/Vector/TestVectorTransforms.cpp
index 72aaa7dc4f89..9d8969edfd90 100644
--- a/mlir/test/lib/Dialect/Vector/TestVectorTransforms.cpp
+++ b/mlir/test/lib/Dialect/Vector/TestVectorTransforms.cpp
@@ -519,7 +519,7 @@ struct TestVectorScanLowering
 /// Allocate shared memory for a single warp to test lowering of
 /// WarpExecuteOnLane0Op.
 static Value allocateGlobalSharedMemory(Location loc, OpBuilder &builder,
-                                        WarpExecuteOnLane0Op warpOp,
+                                        gpu::WarpExecuteOnLane0Op warpOp,
                                         Type type) {
   static constexpr int64_t kSharedMemorySpace = 3;
   // Compute type of shared memory buffer.
@@ -583,8 +583,9 @@ struct TestVectorDistribution
   MLIR_DEFINE_EXPLICIT_INTERNAL_INLINE_TYPE_ID(TestVectorDistribution)
 
   void getDependentDialects(DialectRegistry &registry) const override {
-    registry.insert<scf::SCFDialect, memref::MemRefDialect, gpu::GPUDialect,
-                    affine::AffineDialect>();
+    registry
+        .insert<vector::VectorDialect, scf::SCFDialect, memref::MemRefDialect,
+                gpu::GPUDialect, affine::AffineDialect>();
   }
 
   StringRef getArgument() const final { return "test-vector-warp-distribute"; }
@@ -622,7 +623,7 @@ struct TestVectorDistribution
     RewritePatternSet patterns(&getContext());
 
     getOperation().walk([&](Operation *op) {
-      if (auto warpOp = dyn_cast<WarpExecuteOnLane0Op>(op)) {
+      if (auto warpOp = dyn_cast<gpu::WarpExecuteOnLane0Op>(op)) {
         if (hoistUniform) {
           moveScalarUniformCode(warpOp);
         }
@@ -677,7 +678,7 @@ struct TestVectorDistribution
     WarpExecuteOnLane0LoweringOptions options;
     options.warpAllocationFn = allocateGlobalSharedMemory;
     options.warpSyncronizationFn = [](Location loc, OpBuilder &builder,
-                                      WarpExecuteOnLane0Op warpOp) {
+                                      gpu::WarpExecuteOnLane0Op warpOp) {
       builder.create<gpu::BarrierOp>(loc);
     };
     // Test on one pattern in isolation.