Skip to content

[TOSA] Lower boolean aten.bitwise_not to logical_not #4364

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
wants to merge 1 commit into
base: main
Choose a base branch
from
Open

[TOSA] Lower boolean aten.bitwise_not to logical_not #4364

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
11 changes: 11 additions & 0 deletions lib/Conversion/TorchToTosa/TorchToTosa.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -28,6 +28,7 @@
#include <numeric>
#include <optional>
#include <random>
#include <type_traits>

#include "mlir/Dialect/Tosa/Utils/QuantUtils.h"

Expand Down Expand Up @@ -91,6 +92,16 @@ class ConvertAtenUnaryOp : public OpConversionPattern<AtenOpT> {

self = tosa::tosaCastTensorToType(rewriter, self, outType).value();

if constexpr (std::is_same_v<AtenOpT, AtenBitwiseNotOp>) {
if (auto intTy = dyn_cast<IntegerType>(outType.getElementType())) {
if (intTy.getWidth() == 1) {
rewriter.replaceOpWithNewOp<tosa::LogicalNotOp>(op, outType, self);
return success();
}
}
// otherwise fall through to standard emission
}

rewriter.replaceOpWithNewOp<TosaOpT>(op, outType, self);

return success();
Expand Down
23 changes: 23 additions & 0 deletions projects/pt1/python/torch_mlir_e2e_test/test_suite/elementwise.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -5175,6 +5175,29 @@ def ElementwiseBitwiseNotInt32Module_basic(module, tu: TestUtils):
# ==============================================================================


class ElementwiseBitwiseNotBoolModule(torch.nn.Module):
def __init__(self):
super().__init__()

@export
@annotate_args(
[
None,
([-1, -1], torch.bool, True),
]
)
def forward(self, x):
return torch.bitwise_not(x)


@register_test_case(module_factory=lambda: ElementwiseBitwiseNotBoolModule())
def ElementwiseBitwiseNotBoolModule_basic(module, tu: TestUtils):
module.forward(tu.randint(3, 4, low=0, high=2).to(torch.bool))


# ==============================================================================


class ElementwiseSubTensorInt8Module(torch.nn.Module):
def __init__(self):
super().__init__()
Expand Down
14 changes: 14 additions & 0 deletions test/Conversion/TorchToTosa/basic.mlir
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -135,6 +135,20 @@ func.func @torch.aten.bitwise_not$basic(%arg0: !torch.vtensor<[?,?],f32>) -> !to

// -----

// CHECK-LABEL: func.func @torch.aten.bitwise_not$bool(
// CHECK-SAME: %[[ARG:.*]]: !torch.vtensor<[2,3],i1>) -> !torch.vtensor<[2,3],i1> {
// CHECK: %[[ARG_BUILTIN:.*]] = torch_c.to_builtin_tensor %[[ARG]] : !torch.vtensor<[2,3],i1> -> tensor<2x3xi1>
// CHECK: %[[RESULT_BUILTIN:.*]] = tosa.logical_not %[[ARG_BUILTIN]] : (tensor<2x3xi1>) -> tensor<2x3xi1>
// CHECK: %[[RESULT:.*]] = torch_c.from_builtin_tensor %[[RESULT_BUILTIN]] : tensor<2x3xi1> -> !torch.vtensor<[2,3],i1>
// CHECK: return %[[RESULT]] : !torch.vtensor<[2,3],i1>
// CHECK: }
func.func @torch.aten.bitwise_not$bool(%arg0: !torch.vtensor<[2,3],i1>) -> !torch.vtensor<[2,3],i1> {
%0 = torch.aten.bitwise_not %arg0 : !torch.vtensor<[2,3],i1> -> !torch.vtensor<[2,3],i1>
return %0 : !torch.vtensor<[2,3],i1>
}

// -----

// CHECK-LABEL: func.func @torch.aten.ceil$basic(
// CHECK-SAME: %[[VAL_0:.*]]: !torch.vtensor<[?,?],f32>) -> !torch.vtensor<[?,?],f32> {
// CHECK: %[[VAL_1:.*]] = torch_c.to_builtin_tensor %[[VAL_0]] : !torch.vtensor<[?,?],f32> -> tensor<?x?xf32>
Expand Down
Loading