[mlir] Remove -finalizing-bufferize from tests

This pass is not needed when the IR was bufferized with `-one-shot-bufferize`.

mlir/test/Integration/Dialect/Linalg/CPU/test-collapse-tensor.mlir

@@ -1,6 +1,6 @@
 // RUN: mlir-opt %s \
 // RUN: -one-shot-bufferize="bufferize-function-boundaries" \
-// RUN: -finalizing-bufferize -buffer-deallocation-pipeline -convert-bufferization-to-memref \
+// RUN: -buffer-deallocation-pipeline -convert-bufferization-to-memref \
 // RUN: -convert-scf-to-cf -expand-strided-metadata -lower-affine -convert-cf-to-llvm -convert-arith-to-llvm \
 // RUN: -finalize-memref-to-llvm -convert-func-to-llvm -reconcile-unrealized-casts | \
 // RUN: mlir-cpu-runner -e main -entry-point-result=void \

mlir/test/Integration/Dialect/Linalg/CPU/test-expand-tensor.mlir

@@ -1,6 +1,6 @@
 // RUN: mlir-opt %s \
 // RUN: -one-shot-bufferize="bufferize-function-boundaries" \
-// RUN: -finalizing-bufferize -buffer-deallocation-pipeline -convert-bufferization-to-memref \
+// RUN: -buffer-deallocation-pipeline -convert-bufferization-to-memref \
 // RUN: -convert-scf-to-cf -expand-strided-metadata -lower-affine -convert-cf-to-llvm -convert-arith-to-llvm \
 // RUN: -finalize-memref-to-llvm -convert-func-to-llvm -reconcile-unrealized-casts | \
 // RUN: mlir-cpu-runner -e main -entry-point-result=void \

mlir/test/Integration/Dialect/Linalg/CPU/test-padtensor.mlir

@@ -1,6 +1,6 @@
 // RUN: mlir-opt %s -test-linalg-transform-patterns=test-linalg-to-vector-patterns \
 // RUN: -one-shot-bufferize="bufferize-function-boundaries" \
-// RUN: -finalizing-bufferize -buffer-deallocation-pipeline -convert-bufferization-to-memref \
+// RUN: -buffer-deallocation-pipeline -convert-bufferization-to-memref \
 // RUN: -convert-linalg-to-loops -convert-scf-to-cf -expand-strided-metadata \
 // RUN: -lower-affine -convert-arith-to-llvm -finalize-memref-to-llvm -convert-func-to-llvm -reconcile-unrealized-casts | \
 // RUN: mlir-cpu-runner -e main -entry-point-result=void \

mlir/test/Integration/Dialect/Linalg/CPU/test-subtensor-insert-multiple-uses.mlir

@@ -1,6 +1,6 @@
 // RUN: mlir-opt %s \
 // RUN: -one-shot-bufferize="bufferize-function-boundaries" \
-// RUN: -finalizing-bufferize -buffer-deallocation-pipeline -convert-bufferization-to-memref \
+// RUN: -buffer-deallocation-pipeline -convert-bufferization-to-memref \
 // RUN: -convert-linalg-to-loops -convert-scf-to-cf -expand-strided-metadata  \
 // RUN: -lower-affine -convert-arith-to-llvm --finalize-memref-to-llvm \
 // RUN: -convert-func-to-llvm -reconcile-unrealized-casts | \

mlir/test/Integration/Dialect/Linalg/CPU/test-subtensor-insert.mlir

@@ -1,6 +1,6 @@
 // RUN: mlir-opt %s \
 // RUN: -one-shot-bufferize="bufferize-function-boundaries" \
-// RUN: -finalizing-bufferize -buffer-deallocation-pipeline -convert-bufferization-to-memref \
+// RUN: -buffer-deallocation-pipeline -convert-bufferization-to-memref \
 // RUN: -convert-linalg-to-loops -convert-scf-to-cf -expand-strided-metadata \
 // RUN: -lower-affine -convert-arith-to-llvm --finalize-memref-to-llvm \
 // RUN: -convert-func-to-llvm -reconcile-unrealized-casts | \

mlir/test/Integration/Dialect/Linalg/CPU/test-tensor-e2e.mlir

@@ -1,6 +1,6 @@
 // RUN: mlir-opt %s \
 // RUN: -one-shot-bufferize="bufferize-function-boundaries" \
-// RUN: -finalizing-bufferize -buffer-deallocation-pipeline -convert-bufferization-to-memref -convert-linalg-to-loops \
+// RUN: -buffer-deallocation-pipeline -convert-bufferization-to-memref -convert-linalg-to-loops \
 // RUN: -convert-arith-to-llvm -convert-scf-to-cf -convert-cf-to-llvm --finalize-memref-to-llvm -convert-func-to-llvm -reconcile-unrealized-casts | \
 // RUN: mlir-cpu-runner -e main -entry-point-result=void \
 // RUN:   -shared-libs=%mlir_runner_utils \

mlir/test/Integration/Dialect/Linalg/CPU/test-tensor-matmul.mlir

@@ -1,15 +1,15 @@
 // UNSUPPORTED: asan
 // RUN: mlir-opt %s -test-transform-dialect-erase-schedule \
 // RUN: -one-shot-bufferize="bufferize-function-boundaries" \
-// RUN: -finalizing-bufferize -buffer-deallocation-pipeline -convert-bufferization-to-memref -convert-linalg-to-loops -convert-scf-to-cf \
+// RUN: -buffer-deallocation-pipeline -convert-bufferization-to-memref -convert-linalg-to-loops -convert-scf-to-cf \
 // RUN: -expand-strided-metadata -lower-affine -convert-arith-to-llvm -convert-scf-to-cf --finalize-memref-to-llvm -convert-func-to-llvm -reconcile-unrealized-casts | \
 // RUN: mlir-cpu-runner -e main -entry-point-result=void \
 // RUN:   -shared-libs=%mlir_c_runner_utils,%mlir_runner_utils \
 // RUN: | FileCheck %s
 
 // RUN: mlir-opt %s -transform-interpreter -test-transform-dialect-erase-schedule \
 // RUN: -one-shot-bufferize="bufferize-function-boundaries" \
-// RUN: -finalizing-bufferize -convert-linalg-to-loops -convert-scf-to-cf -convert-scf-to-cf \
+// RUN: -convert-linalg-to-loops -convert-scf-to-cf -convert-scf-to-cf \
 // RUN:  -expand-strided-metadata -lower-affine -convert-arith-to-llvm -convert-scf-to-cf --finalize-memref-to-llvm -convert-func-to-llvm -reconcile-unrealized-casts | \
 // RUN: mlir-cpu-runner -e main -entry-point-result=void \
 // RUN:   -shared-libs=%mlir_c_runner_utils,%mlir_runner_utils \

mlir/test/Integration/GPU/CUDA/all-reduce-and.mlir

@@ -15,6 +15,9 @@
 // RUN:   --entry-point-result=void \
 // RUN: | FileCheck %s
 
+  memref.global "private" constant @internal_global : memref<3x7xi32,4> = dense<467>
+
+
 func.func @main() {
   %data = memref.alloc() : memref<2x6xi32>
   %sum = memref.alloc() : memref<2xi32>
@@ -61,9 +64,12 @@ func.func @main() {
   // AND
   gpu.launch blocks(%bx, %by, %bz) in (%grid_x = %c2, %grid_y = %c1, %grid_z = %c1)
              threads(%tx, %ty, %tz) in (%block_x = %c6, %block_y = %c1, %block_z = %c1) {
-    %val = memref.load %data[%bx, %tx] : memref<2x6xi32>
-    %reduced = gpu.all_reduce and %val uniform {} : (i32) -> (i32)
-    memref.store %reduced, %sum[%bx] : memref<2xi32>
+    //%val = memref.load %data[%bx, %tx] : memref<2x6xi32>
+    //%reduced = gpu.all_reduce and %val uniform {} : (i32) -> (i32)
+    %0 = memref.get_global @internal_global : memref<3x7xi32,4>
+    //%c0 = arith.constant 0 : index
+    %v = memref.load %0[%c0, %c0] : memref<3x7xi32,4>
+    memref.store %v, %sum[%bx] : memref<2xi32>
     gpu.terminator
   }