Add AffineScope trait to scf.index_switch

Adding the `AffineScope` trait to `scf.index_switch` solves #64287. This defines a new top level scope for symbols, which proves highly practical as it enables a broader range of things to be represented as sequences of affine loop nests.
I suggest that we add this trait to other SCF operations like `scf.for`, `scf.while` as well, as that will be very useful, and should not have any major issues that we cannot solve.

Author: gptsarthak

mlir/include/mlir/Dialect/SCF/IR/SCFOps.td

@@ -1101,7 +1101,7 @@ def WhileOp : SCF_Op<"while",
 // IndexSwitchOp
 //===----------------------------------------------------------------------===//
 
-def IndexSwitchOp : SCF_Op<"index_switch", [RecursiveMemoryEffects,
+def IndexSwitchOp : SCF_Op<"index_switch", [RecursiveMemoryEffects, AffineScope,
     SingleBlockImplicitTerminator<"scf::YieldOp">,
     DeclareOpInterfaceMethods<RegionBranchOpInterface,
                               ["getRegionInvocationBounds",

mlir/test/Dialect/Affine/parallelize.mlir

@@ -323,3 +323,22 @@ func.func @iter_arg_memrefs(%in: memref<10xf32>) {
   }
   return
 }
+
+// CHECK-LABEL: @scf_affine_scope
+func.func @scf_affine_scope() {
+  %c0 = arith.constant 0 : index
+  %0 = tensor.empty(%c0) : tensor<?xi1>
+  %1 = bufferization.to_memref %0 : memref<?xi1>
+  %alloc = memref.alloc(%c0) : memref<?xi1>
+  %2 = scf.index_switch %c0 -> tensor<?x31x6xf16>
+  default {
+    %dim = memref.dim %1, %c0 : memref<?xi1>
+    affine.for %arg0 = 0 to %dim {
+      %3 = affine.load %1[%arg0] : memref<?xi1>
+      affine.store %3, %alloc[%arg0] : memref<?xi1>
+    }
+    %3 = tensor.empty(%c0) : tensor<?x31x6xf16>
+    scf.yield %3 : tensor<?x31x6xf16>
+  } 
+  return
+}