intel/llvm
1
0
Fork 0
mirror of https://github.com/intel/llvm.git synced 2026-02-05 13:21:04 +08:00
Code Issues Packages Projects Releases Wiki Activity

[InstCombine] cttz(sext(x)) -> cttz(zext(x))

Browse Source 
```

----------------------------------------
define i32 @src(i16 %x, i1 %b) {
%0:
  %z = sext i16 %x to i32
  %p = cttz i32 %z, %b
  ret i32 %p
}
=>
define i32 @tgt(i16 %x, i1 %b) {
%0:
  %z = zext i16 %x to i32
  %p = cttz i32 %z, %b
  ret i32 %p
}
Transformation seems to be correct!
```

https://alive2.llvm.org/ce/z/evomeg

Reviewed By: spatel

Differential Revision: https://reviews.llvm.org/D101764
This commit is contained in:
Dávid Bolvanský
2021-05-03 23:59:19 +02:00
parent 1c1406f24d
commit 08c08577f9
2 changed files with 21 additions and 13 deletions
Download Patch File Download Diff File Expand all files Collapse all files

8
llvm/lib/Transforms/InstCombine/InstCombineCalls.cpp
View File

@@ -439,6 +439,14 @@ static Instruction *foldCttzCtlz(IntrinsicInst &II, InstCombinerImpl &IC) {
if (match(Op0, m_Neg(m_Value(X))))
return IC.replaceOperand(II, 0, X);
// cttz(sext(x)) -> cttz(zext(x))
if (match(Op0, m_OneUse(m_SExt(m_Value(X))))) {
auto *Zext = IC.Builder.CreateZExt(X, II.getType());
auto *CttzZext =
IC.Builder.CreateBinaryIntrinsic(Intrinsic::cttz, Zext, Op1);
return IC.replaceInstUsesWith(II, CttzZext);
}
// Zext doesn't change the number of trailing zeros, so narrow:
// cttz(zext(x)) -> zext(cttz(x)) if the 'ZeroIsUndef' parameter is 'true'.
if (match(Op0, m_OneUse(m_ZExt(m_Value(X)))) && match(Op1, m_One())) {

26
llvm/test/Transforms/InstCombine/cttz.ll
View File

@@ -64,9 +64,9 @@ define <2 x i64> @cttz_zext_zero_def_vec(<2 x i32> %x) {
define i32 @cttz_sext_zero_undef(i16 %x) {
; CHECK-LABEL: @cttz_sext_zero_undef(
; CHECK-NEXT: [[S:%.*]] = sext i16 [[X:%.*]] to i32
; CHECK-NEXT: [[TZ:%.*]] = call i32 @llvm.cttz.i32(i32 [[S]], i1 true), !range [[RNG1]]
; CHECK-NEXT: ret i32 [[TZ]]
; CHECK-NEXT: [[TMP1:%.*]] = call i16 @llvm.cttz.i16(i16 [[X:%.*]], i1 true), !range [[RNG0]]
; CHECK-NEXT: [[TMP2:%.*]] = zext i16 [[TMP1]] to i32
; CHECK-NEXT: ret i32 [[TMP2]]
;
%s = sext i16 %x to i32
%tz = call i32 @llvm.cttz.i32(i32 %s, i1 true)
@@ -75,11 +75,11 @@ define i32 @cttz_sext_zero_undef(i16 %x) {
define i32 @cttz_sext_zero_def(i16 %x) {
; CHECK-LABEL: @cttz_sext_zero_def(
; CHECK-NEXT: [[S:%.*]] = zext i16 [[X:%.*]] to i32
; CHECK-NEXT: [[TZ:%.*]] = call i32 @llvm.cttz.i32(i32 [[S]], i1 false), !range [[RNG1]]
; CHECK-NEXT: ret i32 [[TZ]]
; CHECK-NEXT: [[TMP1:%.*]] = zext i16 [[X:%.*]] to i32
; CHECK-NEXT: [[TMP2:%.*]] = call i32 @llvm.cttz.i32(i32 [[TMP1]], i1 false), !range [[RNG1]]
; CHECK-NEXT: ret i32 [[TMP2]]
;
%s = zext i16 %x to i32
%s = sext i16 %x to i32
%tz = call i32 @llvm.cttz.i32(i32 %s, i1 false)
ret i32 %tz
}
@@ -99,9 +99,9 @@ define i32 @cttz_sext_zero_undef_extra_use(i16 %x) {
define <2 x i64> @cttz_sext_zero_undef_vec(<2 x i32> %x) {
; CHECK-LABEL: @cttz_sext_zero_undef_vec(
; CHECK-NEXT: [[S:%.*]] = sext <2 x i32> [[X:%.*]] to <2 x i64>
; CHECK-NEXT: [[TZ:%.*]] = tail call <2 x i64> @llvm.cttz.v2i64(<2 x i64> [[S]], i1 true)
; CHECK-NEXT: ret <2 x i64> [[TZ]]
; CHECK-NEXT: [[TMP1:%.*]] = call <2 x i32> @llvm.cttz.v2i32(<2 x i32> [[X:%.*]], i1 true)
; CHECK-NEXT: [[TMP2:%.*]] = zext <2 x i32> [[TMP1]] to <2 x i64>
; CHECK-NEXT: ret <2 x i64> [[TMP2]]
;
%s = sext <2 x i32> %x to <2 x i64>
%tz = tail call <2 x i64> @llvm.cttz.v2i64(<2 x i64> %s, i1 true)
@@ -110,9 +110,9 @@ define <2 x i64> @cttz_sext_zero_undef_vec(<2 x i32> %x) {
define <2 x i64> @cttz_sext_zero_def_vec(<2 x i32> %x) {
; CHECK-LABEL: @cttz_sext_zero_def_vec(
; CHECK-NEXT: [[S:%.*]] = sext <2 x i32> [[X:%.*]] to <2 x i64>
; CHECK-NEXT: [[TZ:%.*]] = tail call <2 x i64> @llvm.cttz.v2i64(<2 x i64> [[S]], i1 false)
; CHECK-NEXT: ret <2 x i64> [[TZ]]
; CHECK-NEXT: [[TMP1:%.*]] = zext <2 x i32> [[X:%.*]] to <2 x i64>
; CHECK-NEXT: [[TMP2:%.*]] = call <2 x i64> @llvm.cttz.v2i64(<2 x i64> [[TMP1]], i1 false)
; CHECK-NEXT: ret <2 x i64> [[TMP2]]
;
%s = sext <2 x i32> %x to <2 x i64>
%tz = tail call <2 x i64> @llvm.cttz.v2i64(<2 x i64> %s, i1 false)