benchgc: support transpose op

scripts/correctness.sh

@@ -23,6 +23,7 @@ python3 -m benchgc --verbose 0 --driver linalg --case reduce.l2_square --md 0:12
 python3 -m benchgc --verbose 0 --driver linalg --case fill --md 0:f32 --md 1:32x4096xf32 --cmp 1:P:0:0 || FAIL=1
 python3 -m benchgc --verbose 0 --driver linalg --case copy --md 0:1024x1024xf32 --md 1:1024x1024xbf16 || FAIL=1
 python3 -m benchgc --verbose 0 --driver linalg --case broadcast --md 0:1024xf32 --md 1:2x32x1024xf32 --dimensions=0 --dimensions=1 || FAIL=1
+python3 -m benchgc --verbose 0 --driver linalg --case transpose --md 0:32x64x128xf32 --md 1:64x128x32xf32 --permutation=1 --permutation=2 --permutation=0 || FAIL=1
 
 # matmul
 python3 -m benchgc --verbose 0 --driver linalg --case batch_matmul --md 0:16x512x64xf32 --md 1:16x64x32xf32 --md 2:16x512x32xf32 || FAIL=1

test/benchgc/src/benchgc/__main__.py

@@ -187,6 +187,14 @@ def add_common_options(parser: argparse.ArgumentParser):
             type=int,
         )
 
+        parser.add_argument(
+            "--permutation",
+            required=False,
+            default=None,
+            action="append",
+            help="define the permutation attribute in linalg op",
+            type=int,
+        )
 
 def add_bench_options(parser: argparse.ArgumentParser):
     """add options for bench mode"""

test/benchgc/src/benchgc/linalg/misc.py

@@ -97,3 +97,23 @@ def mlir_copy(flags: argparse.Namespace, args: List[Arg]) -> ir.Module:
             )
         ],
     )
+
+
+def ref_transpose(
+    cache: MLIRCache, op: ir.OpView, var: Dict[str, torch.Tensor]
+) -> Tuple[torch.Tensor, ...]:
+    permutation: List[int] = [int(d) for d in op.attributes["permutation"]]
+    return (var[cache.opr[0]].permute(permutation).contiguous(),)
+
+
+def mlir_transpose(flags: argparse.Namespace, args: List[Arg]) -> ir.Module:
+    return init_module(
+        flags.entry,
+        (args[0],),
+        (args[1],),
+        lambda ctx, arg0: [
+            linalg.transpose(
+                arg0, outs=[args[1].get_zero_op(ctx)], permutation=flags.permutation
+            )
+        ],
+    )