Skip to content
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

Jump to bottom

TOSA: add folder for tosa.sub #72

Merged
merged 4 commits into from
Aug 16, 2023
Merged

TOSA: add folder for tosa.sub #72

Show file tree
Hide file tree
Changes from 3 commits
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
aeff0ff
TOSA: add folder for tosa.sub
mgehre-amd Aug 11, 2023
62d497e
Merge commit '226f162d17f3' into matthias.tosa_sub_folder
mgehre-amd Aug 14, 2023
34c90dd
mlir/lib/Dialect/Tosa/Transforms/TosaFolders.cpp: review comments
mgehre-amd Aug 16, 2023
473b2b8
Update mlir/lib/Dialect/Tosa/Transforms/TosaFolders.cpp
mgehre-amd Aug 16, 2023
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
34 changes: 33 additions & 1 deletion mlir/lib/Dialect/Tosa/Transforms/TosaFolders.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1005,6 +1005,37 @@ struct TosaFoldConstantAdd : public TosaFoldConstantBinary<TosaFoldConstantAdd,
}
};

struct TosaFoldConstantSub : public TosaFoldConstantBinary<TosaFoldConstantSub, SubOp> {
using TosaFoldConstantBinary<TosaFoldConstantSub, SubOp>::TosaFoldConstantBinary;

DenseElementsAttr computeInteger(DenseElementsAttr lhsValues,
DenseElementsAttr rhsValues,
PatternRewriter &rewriter, SubOp op) const {
bool overflowed = false;
auto newTensor = applyElementWise<APInt, APInt>(lhsValues, rhsValues,
op.getType(), [&overflowed](const APInt &first, const APInt &second) {
bool didOverflow;
auto res = first.ssub_ov(second, didOverflow);
overflowed |= didOverflow;
return res;
});

if (overflowed) {
op->emitWarning("Subtraction did overflow. The results are unspecified.");
}
return newTensor;
}

DenseElementsAttr computeFloat(DenseElementsAttr lhsValues,
DenseElementsAttr rhsValues,
PatternRewriter &rewriter, SubOp op) const {
return applyElementWise<APFloat, APFloat>(lhsValues, rhsValues,
op.getType(), [](const APFloat &first, const APFloat &second) {
return first - second;
});
}
};

struct TosaFoldConstantGreater : public TosaFoldConstantBinary<TosaFoldConstantGreater, GreaterOp> {
using TosaFoldConstantBinary<TosaFoldConstantGreater, GreaterOp>::TosaFoldConstantBinary;

Expand Down Expand Up @@ -1382,6 +1413,7 @@ void mlir::tosa::populateTosaFoldConstantPatterns(
patterns.add<TosaFoldConstantFloatCasts>(ctx, foldSplatOrSingleUseOnly);
}
patterns.add<TosaFoldConstantAdd>(ctx, foldSplatOrSingleUseOnly);
patterns.add<TosaFoldConstantSub>(ctx, foldSplatOrSingleUseOnly);
patterns.add<TosaFoldConstantGreater>(ctx, foldSplatOrSingleUseOnly);
patterns.add<TosaFoldConstantBitwiseNot>(ctx, foldSplatOrSingleUseOnly);
patterns.add<TosaFoldConstantCeil>(ctx, foldSplatOrSingleUseOnly);
Expand All @@ -1394,4 +1426,4 @@ void mlir::tosa::populateTosaFoldConstantPatterns(
patterns.add<TosaFoldConstantMinimum>(ctx, foldSplatOrSingleUseOnly);
patterns.add<TosaFoldConstantMaximum>(ctx, foldSplatOrSingleUseOnly);
patterns.add<TosaFoldConstantPad>(ctx, foldSplatOrSingleUseOnly);
}
}
mgehre-amd marked this conversation as resolved.
Show resolved Hide resolved
150 changes: 150 additions & 0 deletions mlir/test/Dialect/Tosa/constant-sub.mlir
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,150 @@
// RUN: mlir-opt --split-input-file -verify-diagnostics --tosa-layerwise-constant-fold %s | FileCheck %s

// Float subtractions

// CHECK-LABEL: @sub_fold_float
func.func @sub_fold_float() -> tensor<4xf16> {
// CHECK: [[RES:]] ={{.*}}tosa.const{{.*}}1.152500e+02, 7.988280e+00, 0.000000e+00, -5.000000e+00
// CHECK-NOT: tosa.sub
// CHECK: return [[RES]]
%0 = "tosa.const"() {value =
dense<[-17.4978, 4.9882, 0.0, -0.0]> :
tensor<4xf16>
} : () -> tensor<4xf16>
%1 = "tosa.const"() {value =
dense<[-132.7, -3.0, -0.0, 5.0]> :
tensor<4xf16>
} : () -> tensor<4xf16>
%2 = "tosa.sub"(%0, %1) : (tensor<4xf16>, tensor<4xf16>) -> tensor<4xf16>
return %2 : tensor<4xf16>
}

// CHECK-LABEL: @sub_fold_float_infinity_nan
func.func @sub_fold_float_infinity_nan() -> tensor<6xf32> {
// CHECK: [[RES:]] ={{.*}}tosa.const{{.*}}0x7F800000, 0xFF800000, 0x7F800000, 0xFF800000, 0x7FC00000, 0x7F800000
// CHECK-NOT: tosa.sub
// CHECK: return [[RES]]
%0 = "tosa.const"() {value =
dense<[0x7F800000, 0xFF800000, 0x7F800000, 0xFF800000, 0x7FC00000, 0x7F800000]> :
tensor<6xf32>
} : () -> tensor<6xf32>
%1 = "tosa.const"() {value =
dense<[3.0, -3.0, -3.0, 3.0, 1.0, 0xFF800000]> :
tensor<6xf32>
} : () -> tensor<6xf32>
%2 = "tosa.sub"(%0, %1) : (tensor<6xf32>, tensor<6xf32>) -> tensor<6xf32>
return %2 : tensor<6xf32>
}

// CHECK-LABEL: @sub_fold_float_overflow
func.func @sub_fold_float_overflow() -> tensor<2xf32> {
// CHECK: [[RES:]] ={{.*}}tosa.const{{.*}}0x7F800000, 0xFF800000
// CHECK-NOT: tosa.sub
// CHECK: return [[RES]]
%0 = "tosa.const"() {value =
dense<[3.1e+38, -3.1e+38]> :
tensor<2xf32>
} : () -> tensor<2xf32>
%1 = "tosa.const"() {value =
dense<[-2.1e+38, 1.1e+38]> :
tensor<2xf32>
} : () -> tensor<2xf32>
%2 = "tosa.sub"(%0, %1) : (tensor<2xf32>, tensor<2xf32>) -> tensor<2xf32>
return %2 : tensor<2xf32>
}

// -----
// Int subtraction

// CHECK-LABEL: @sub_fold_int
func.func @sub_fold_int() -> tensor<4xi32> {
// CHECK: [[RES:]] ={{.*}}tosa.const{{.*}}-149, 1, 0, 5
// CHECK-NOT: tosa.sub
// CHECK: return [[RES]]
%0 = "tosa.const"() {value =
dense<[-17, 4, 0, 0]> :
tensor<4xi32>
} : () -> tensor<4xi32>
%1 = "tosa.const"() {value =
dense<[132, 3, 0, -5]> :
tensor<4xi32>
} : () -> tensor<4xi32>
%2 = "tosa.sub"(%0, %1) : (tensor<4xi32>, tensor<4xi32>) -> tensor<4xi32>
return %2 : tensor<4xi32>
}

// CHECK-LABEL: @sub_fold_int_overflow
func.func @sub_fold_int_overflow() -> tensor<4xi32> {
// Don't expect any specific results for the overflowing subtraction, just
// expect that it is folded.
// CHECK: [[RES:]] ={{.*}}tosa.const
// CHECK-NOT: tosa.sub
// CHECK: return [[RES]]
%0 = "tosa.const"() {value =
dense<[2147483647, 2147483640, -2147483648, -2147483640]> :
tensor<4xi32>
} : () -> tensor<4xi32>
%1 = "tosa.const"() {value =
dense<[-1, -10, 1, 30]> :
tensor<4xi32>
} : () -> tensor<4xi32>
// expected-warning@below {{Subtraction did overflow. The results are unspecified.}}
%2 = "tosa.sub"(%0, %1) : (tensor<4xi32>, tensor<4xi32>) -> tensor<4xi32>
return %2 : tensor<4xi32>
}

// -----
// self-subtraction

// CHECK-LABEL: @sub_fold_equal_args
func.func @sub_fold_equal_args() -> tensor<3xi32> {
// CHECK: [[RES:]] ={{.*}}tosa.const{{.*}}<0>
// CHECK-NOT: tosa.sub
// CHECK: return [[RES]]
%0 = "tosa.const"() {value =
dense<[-17, 4, 0]> :
tensor<3xi32>
} : () -> tensor<3xi32>
%2 = "tosa.sub"(%0, %0) : (tensor<3xi32>, tensor<3xi32>) -> tensor<3xi32>
return %2 : tensor<3xi32>
}

// -----
// Broadcasted subtractions

// CHECK-LABEL: @sub_fold_int_broadcast_simple
func.func @sub_fold_int_broadcast_simple() -> tensor<3xi32> {
// CHECK: [[RES:]] ={{.*}}tosa.const{{.*}}-29, -8, -12
// CHECK-NOT: tosa.sub
// CHECK: return [[RES]]
%0 = "tosa.const"() {value =
dense<[-17, 4, 0]> :
tensor<3xi32>
} : () -> tensor<3xi32>
%1 = "tosa.const"() {value =
dense<12> :
tensor<1xi32>
} : () -> tensor<1xi32>
%2 = "tosa.sub"(%0, %1) : (tensor<3xi32>, tensor<1xi32>) -> tensor<3xi32>
return %2 : tensor<3xi32>
}

// CHECK-LABEL: @sub_fold_int_broadcast_complex
func.func @sub_fold_int_broadcast_complex() -> tensor<3x3xi32> {
// CHECK: [[RES:]] ={{.*}}tosa.const
// CHECK-SAME{LITERAL}: [[-29, -10, -13],
// CHECK-SAME{LITERAL}: [-11, 8, 5],
// CHECK-SAME{LITERAL}: [7, 26, 23]]
// CHECK-NOT: tosa.sub
// CHECK: return [[RES]]
%0 = "tosa.const"() {value =
dense<[[-17], [1], [19]]> :
tensor<3x1xi32>
} : () -> tensor<3x1xi32>
%1 = "tosa.const"() {value =
dense<[[12, -7, -4]]> :
tensor<1x3xi32>
} : () -> tensor<1x3xi32>
%2 = "tosa.sub"(%0, %1) : (tensor<3x1xi32>, tensor<1x3xi32>) -> tensor<3x3xi32>
return %2 : tensor<3x3xi32>
}