[llvm] goldsteinn/is truly negatable (PR #72767)
via llvm-commits
llvm-commits at lists.llvm.org
Sat Nov 18 15:29:42 PST 2023
llvmbot wrote:
<!--LLVM PR SUMMARY COMMENT-->
@llvm/pr-subscribers-llvm-transforms
@llvm/pr-subscribers-llvm-analysis
Author: None (goldsteinn)
<details>
<summary>Changes</summary>
- [InstCombine] Add tests for improving `sub X, ~Y` -> `add X, -Y`; NFC
- [InstCombine] Don't transform `sub X, ~Y` -> `add X, -Y` unless `Y` is actually negatable
---
Full diff: https://github.com/llvm/llvm-project/pull/72767.diff
3 Files Affected:
- (modified) llvm/lib/Transforms/InstCombine/InstCombineNegator.cpp (+5-3)
- (modified) llvm/test/Analysis/ValueTracking/knownbits-and-or-xor-lowbit.ll (+2-3)
- (modified) llvm/test/Transforms/InstCombine/fold-sub-of-not-to-inc-of-add.ll (+20-7)
``````````diff
diff --git a/llvm/lib/Transforms/InstCombine/InstCombineNegator.cpp b/llvm/lib/Transforms/InstCombine/InstCombineNegator.cpp
index 53b34b03ca78070..6e0b44c11f22a45 100644
--- a/llvm/lib/Transforms/InstCombine/InstCombineNegator.cpp
+++ b/llvm/lib/Transforms/InstCombine/InstCombineNegator.cpp
@@ -447,9 +447,11 @@ std::array<Value *, 2> Negator::getSortedOperandsOfBinOp(Instruction *I) {
// `xor` is negatible if one of its operands is invertible.
// FIXME: InstCombineInverter? But how to connect Inverter and Negator?
if (auto *C = dyn_cast<Constant>(Ops[1])) {
- Value *Xor = Builder.CreateXor(Ops[0], ConstantExpr::getNot(C));
- return Builder.CreateAdd(Xor, ConstantInt::get(Xor->getType(), 1),
- I->getName() + ".neg");
+ if (IsTrulyNegation) {
+ Value *Xor = Builder.CreateXor(Ops[0], ConstantExpr::getNot(C));
+ return Builder.CreateAdd(Xor, ConstantInt::get(Xor->getType(), 1),
+ I->getName() + ".neg");
+ }
}
return nullptr;
}
diff --git a/llvm/test/Analysis/ValueTracking/knownbits-and-or-xor-lowbit.ll b/llvm/test/Analysis/ValueTracking/knownbits-and-or-xor-lowbit.ll
index 7fbc78079c24029..f44948e6a089f88 100644
--- a/llvm/test/Analysis/ValueTracking/knownbits-and-or-xor-lowbit.ll
+++ b/llvm/test/Analysis/ValueTracking/knownbits-and-or-xor-lowbit.ll
@@ -106,9 +106,8 @@ define <2 x i1> @sub_XY_and_bit0_is_zero_fail(<2 x i8> %x, <2 x i8> %C) nounwind
define i1 @sub_XY_xor_bit0_is_one_fail(i8 %x, i8 %C) nounwind {
; CHECK-LABEL: @sub_XY_xor_bit0_is_one_fail(
-; CHECK-NEXT: [[TMP1:%.*]] = xor i8 [[C:%.*]], -2
-; CHECK-NEXT: [[C1_NEG:%.*]] = add i8 [[TMP1]], 1
-; CHECK-NEXT: [[Y:%.*]] = add i8 [[C1_NEG]], [[X:%.*]]
+; CHECK-NEXT: [[C1:%.*]] = xor i8 [[C:%.*]], 1
+; CHECK-NEXT: [[Y:%.*]] = sub i8 [[X:%.*]], [[C1]]
; CHECK-NEXT: [[W:%.*]] = xor i8 [[Y]], [[X]]
; CHECK-NEXT: [[R:%.*]] = icmp eq i8 [[W]], 10
; CHECK-NEXT: ret i1 [[R]]
diff --git a/llvm/test/Transforms/InstCombine/fold-sub-of-not-to-inc-of-add.ll b/llvm/test/Transforms/InstCombine/fold-sub-of-not-to-inc-of-add.ll
index 1e315696c3110c3..6f311f05fb0176e 100644
--- a/llvm/test/Transforms/InstCombine/fold-sub-of-not-to-inc-of-add.ll
+++ b/llvm/test/Transforms/InstCombine/fold-sub-of-not-to-inc-of-add.ll
@@ -22,6 +22,19 @@ define i32 @p0_scalar(i32 %x, i32 %y) {
ret i32 %t1
}
+define i8 @p0_scalar_not_truly_negatable(i8 %x, i8 %y) {
+; CHECK-LABEL: @p0_scalar_not_truly_negatable(
+; CHECK-NEXT: [[XX:%.*]] = xor i8 [[X:%.*]], 123
+; CHECK-NEXT: [[YY:%.*]] = xor i8 [[Y:%.*]], 45
+; CHECK-NEXT: [[R:%.*]] = sub i8 [[XX]], [[YY]]
+; CHECK-NEXT: ret i8 [[R]]
+;
+ %xx = xor i8 %x, 123
+ %yy = xor i8 %y, 45
+ %r = sub i8 %xx, %yy
+ ret i8 %r
+}
+
;------------------------------------------------------------------------------;
; Vector tests
;------------------------------------------------------------------------------;
@@ -79,28 +92,28 @@ define i32 @n4(i32 %x, i32 %y) {
; CHECK-NEXT: ret i32 [[T1]]
;
%t0 = xor i32 %x, -1
- %t1 = sub i32 %t0, %y ; swapped
+ %t1 = sub i32 %t0, %y ; swapped
ret i32 %t1
}
define i32 @n5_is_not_not(i32 %x, i32 %y) {
; CHECK-LABEL: @n5_is_not_not(
-; CHECK-NEXT: [[T0_NEG:%.*]] = add i32 [[X:%.*]], -2147483647
-; CHECK-NEXT: [[T1:%.*]] = add i32 [[T0_NEG]], [[Y:%.*]]
+; CHECK-NEXT: [[T0:%.*]] = xor i32 [[X:%.*]], 2147483647
+; CHECK-NEXT: [[T1:%.*]] = sub i32 [[Y:%.*]], [[T0]]
; CHECK-NEXT: ret i32 [[T1]]
;
- %t0 = xor i32 %x, 2147483647 ; not -1
+ %t0 = xor i32 %x, 2147483647 ; not -1
%t1 = sub i32 %y, %t0
ret i32 %t1
}
define <2 x i32> @n5_is_not_not_vec_splat(<2 x i32> %x, <2 x i32> %y) {
; CHECK-LABEL: @n5_is_not_not_vec_splat(
-; CHECK-NEXT: [[T0_NEG:%.*]] = add <2 x i32> [[X:%.*]], <i32 -2147483647, i32 -2147483647>
-; CHECK-NEXT: [[T1:%.*]] = add <2 x i32> [[T0_NEG]], [[Y:%.*]]
+; CHECK-NEXT: [[T0:%.*]] = xor <2 x i32> [[X:%.*]], <i32 2147483647, i32 2147483647>
+; CHECK-NEXT: [[T1:%.*]] = sub <2 x i32> [[Y:%.*]], [[T0]]
; CHECK-NEXT: ret <2 x i32> [[T1]]
;
- %t0 = xor <2 x i32> %x, <i32 2147483647, i32 2147483647> ; signmask, but not -1
+ %t0 = xor <2 x i32> %x, <i32 2147483647, i32 2147483647> ; signmask, but not -1
%t1 = sub <2 x i32> %y, %t0
ret <2 x i32> %t1
}
``````````
</details>
https://github.com/llvm/llvm-project/pull/72767
More information about the llvm-commits
mailing list