[llvm] 1640582 - [InstCombine] replace undef elements in vector constant when doing icmp folds (PR44383)
Sanjay Patel via llvm-commits
llvm-commits at lists.llvm.org
Fri Jan 3 06:17:04 PST 2020
Author: Sanjay Patel
Date: 2020-01-03T09:16:57-05:00
New Revision: 164058274364670c8b59055b4689ae7e1720268e
URL: https://github.com/llvm/llvm-project/commit/164058274364670c8b59055b4689ae7e1720268e
DIFF: https://github.com/llvm/llvm-project/commit/164058274364670c8b59055b4689ae7e1720268e.diff
LOG: [InstCombine] replace undef elements in vector constant when doing icmp folds (PR44383)
As shown in P44383:
https://bugs.llvm.org/show_bug.cgi?id=44383
...we can't safely propagate a vector constant through this icmp fold
if that vector constant contains undefined elements.
We know that each defined element of the constant is safe though, so
find the first of those and replicate it into the formerly undef lanes.
Differential Revision: https://reviews.llvm.org/D72101
Added:
Modified:
llvm/lib/Transforms/InstCombine/InstCombineCompares.cpp
llvm/test/Transforms/InstCombine/canonicalize-constant-low-bit-mask-and-icmp-ne-to-icmp-ugt.ll
llvm/test/Transforms/InstCombine/canonicalize-constant-low-bit-mask-and-icmp-sgt-to-icmp-sgt.ll
llvm/test/Transforms/InstCombine/canonicalize-constant-low-bit-mask-and-icmp-slt-to-icmp-sgt.ll
llvm/test/Transforms/InstCombine/canonicalize-constant-low-bit-mask-and-icmp-ugt-to-icmp-ugt.ll
llvm/test/Transforms/InstCombine/canonicalize-constant-low-bit-mask-and-icmp-ult-to-icmp-ugt.ll
Removed:
################################################################################
diff --git a/llvm/lib/Transforms/InstCombine/InstCombineCompares.cpp b/llvm/lib/Transforms/InstCombine/InstCombineCompares.cpp
index 4abac988c2e4..f38dc436722d 100644
--- a/llvm/lib/Transforms/InstCombine/InstCombineCompares.cpp
+++ b/llvm/lib/Transforms/InstCombine/InstCombineCompares.cpp
@@ -3364,6 +3364,23 @@ static Value *foldICmpWithLowBitMaskedVal(ICmpInst &I,
llvm_unreachable("All possible folds are handled.");
}
+ // The mask value may be a vector constant that has undefined elements. But it
+ // may not be safe to propagate those undefs into the new compare, so replace
+ // those elements by copying an existing, defined, and safe scalar constant.
+ Type *OpTy = M->getType();
+ auto *VecC = dyn_cast<Constant>(M);
+ if (OpTy->isVectorTy() && VecC && VecC->containsUndefElement()) {
+ Constant *SafeReplacementConstant = nullptr;
+ for (unsigned i = 0, e = OpTy->getVectorNumElements(); i != e; ++i) {
+ if (!isa<UndefValue>(VecC->getAggregateElement(i))) {
+ SafeReplacementConstant = VecC->getAggregateElement(i);
+ break;
+ }
+ }
+ assert(SafeReplacementConstant && "Failed to find undef replacement");
+ M = Constant::replaceUndefsWith(VecC, SafeReplacementConstant);
+ }
+
return Builder.CreateICmp(DstPred, X, M);
}
diff --git a/llvm/test/Transforms/InstCombine/canonicalize-constant-low-bit-mask-and-icmp-ne-to-icmp-ugt.ll b/llvm/test/Transforms/InstCombine/canonicalize-constant-low-bit-mask-and-icmp-ne-to-icmp-ugt.ll
index f1f2528fe3d2..49f76fd9f407 100644
--- a/llvm/test/Transforms/InstCombine/canonicalize-constant-low-bit-mask-and-icmp-ne-to-icmp-ugt.ll
+++ b/llvm/test/Transforms/InstCombine/canonicalize-constant-low-bit-mask-and-icmp-ne-to-icmp-ugt.ll
@@ -82,7 +82,7 @@ define <2 x i1> @p2_vec_nonsplat_edgecase1(<2 x i8> %x) {
define <3 x i1> @p3_vec_splat_undef(<3 x i8> %x) {
; CHECK-LABEL: @p3_vec_splat_undef(
-; CHECK-NEXT: [[TMP1:%.*]] = icmp ugt <3 x i8> [[X:%.*]], <i8 3, i8 undef, i8 3>
+; CHECK-NEXT: [[TMP1:%.*]] = icmp ugt <3 x i8> [[X:%.*]], <i8 3, i8 3, i8 3>
; CHECK-NEXT: ret <3 x i1> [[TMP1]]
;
%tmp0 = and <3 x i8> %x, <i8 3, i8 undef, i8 3>
@@ -92,7 +92,7 @@ define <3 x i1> @p3_vec_splat_undef(<3 x i8> %x) {
define <3 x i1> @p3_vec_nonsplat_undef(<3 x i8> %x) {
; CHECK-LABEL: @p3_vec_nonsplat_undef(
-; CHECK-NEXT: [[TMP1:%.*]] = icmp ugt <3 x i8> [[X:%.*]], <i8 -1, i8 undef, i8 3>
+; CHECK-NEXT: [[TMP1:%.*]] = icmp ugt <3 x i8> [[X:%.*]], <i8 -1, i8 -1, i8 3>
; CHECK-NEXT: ret <3 x i1> [[TMP1]]
;
%tmp0 = and <3 x i8> %x, <i8 -1, i8 undef, i8 3>
diff --git a/llvm/test/Transforms/InstCombine/canonicalize-constant-low-bit-mask-and-icmp-sgt-to-icmp-sgt.ll b/llvm/test/Transforms/InstCombine/canonicalize-constant-low-bit-mask-and-icmp-sgt-to-icmp-sgt.ll
index d3ef5e54f010..16d063ee0552 100644
--- a/llvm/test/Transforms/InstCombine/canonicalize-constant-low-bit-mask-and-icmp-sgt-to-icmp-sgt.ll
+++ b/llvm/test/Transforms/InstCombine/canonicalize-constant-low-bit-mask-and-icmp-sgt-to-icmp-sgt.ll
@@ -76,7 +76,7 @@ define <2 x i1> @p2_vec_nonsplat_edgecase() {
define <3 x i1> @p3_vec_splat_undef() {
; CHECK-LABEL: @p3_vec_splat_undef(
; CHECK-NEXT: [[X:%.*]] = call <3 x i8> @gen3x8()
-; CHECK-NEXT: [[TMP1:%.*]] = icmp sgt <3 x i8> [[X]], <i8 3, i8 undef, i8 3>
+; CHECK-NEXT: [[TMP1:%.*]] = icmp sgt <3 x i8> [[X]], <i8 3, i8 3, i8 3>
; CHECK-NEXT: ret <3 x i1> [[TMP1]]
;
%x = call <3 x i8> @gen3x8()
@@ -88,7 +88,7 @@ define <3 x i1> @p3_vec_splat_undef() {
define <3 x i1> @p3_vec_nonsplat_undef() {
; CHECK-LABEL: @p3_vec_nonsplat_undef(
; CHECK-NEXT: [[X:%.*]] = call <3 x i8> @gen3x8()
-; CHECK-NEXT: [[TMP1:%.*]] = icmp sgt <3 x i8> [[X]], <i8 15, i8 3, i8 undef>
+; CHECK-NEXT: [[TMP1:%.*]] = icmp sgt <3 x i8> [[X]], <i8 15, i8 3, i8 15>
; CHECK-NEXT: ret <3 x i1> [[TMP1]]
;
%x = call <3 x i8> @gen3x8()
diff --git a/llvm/test/Transforms/InstCombine/canonicalize-constant-low-bit-mask-and-icmp-slt-to-icmp-sgt.ll b/llvm/test/Transforms/InstCombine/canonicalize-constant-low-bit-mask-and-icmp-slt-to-icmp-sgt.ll
index 6366fed23c3f..b71172b42634 100644
--- a/llvm/test/Transforms/InstCombine/canonicalize-constant-low-bit-mask-and-icmp-slt-to-icmp-sgt.ll
+++ b/llvm/test/Transforms/InstCombine/canonicalize-constant-low-bit-mask-and-icmp-slt-to-icmp-sgt.ll
@@ -61,7 +61,7 @@ define <2 x i1> @p2_vec_nonsplat_edgecase(<2 x i8> %x) {
define <3 x i1> @p3_vec_splat_undef(<3 x i8> %x) {
; CHECK-LABEL: @p3_vec_splat_undef(
-; CHECK-NEXT: [[TMP1:%.*]] = icmp sgt <3 x i8> [[X:%.*]], <i8 3, i8 undef, i8 3>
+; CHECK-NEXT: [[TMP1:%.*]] = icmp sgt <3 x i8> [[X:%.*]], <i8 3, i8 3, i8 3>
; CHECK-NEXT: ret <3 x i1> [[TMP1]]
;
%tmp0 = and <3 x i8> %x, <i8 3, i8 undef, i8 3>
@@ -71,7 +71,7 @@ define <3 x i1> @p3_vec_splat_undef(<3 x i8> %x) {
define <3 x i1> @p3_vec_nonsplat_undef(<3 x i8> %x) {
; CHECK-LABEL: @p3_vec_nonsplat_undef(
-; CHECK-NEXT: [[TMP1:%.*]] = icmp sgt <3 x i8> [[X:%.*]], <i8 undef, i8 15, i8 3>
+; CHECK-NEXT: [[TMP1:%.*]] = icmp sgt <3 x i8> [[X:%.*]], <i8 15, i8 15, i8 3>
; CHECK-NEXT: ret <3 x i1> [[TMP1]]
;
%tmp0 = and <3 x i8> %x, <i8 undef, i8 15, i8 3>
diff --git a/llvm/test/Transforms/InstCombine/canonicalize-constant-low-bit-mask-and-icmp-ugt-to-icmp-ugt.ll b/llvm/test/Transforms/InstCombine/canonicalize-constant-low-bit-mask-and-icmp-ugt-to-icmp-ugt.ll
index 9e0008611b58..9b1d405935d0 100644
--- a/llvm/test/Transforms/InstCombine/canonicalize-constant-low-bit-mask-and-icmp-ugt-to-icmp-ugt.ll
+++ b/llvm/test/Transforms/InstCombine/canonicalize-constant-low-bit-mask-and-icmp-ugt-to-icmp-ugt.ll
@@ -99,7 +99,7 @@ define <2 x i1> @p2_vec_nonsplat_edgecase1() {
define <3 x i1> @p3_vec_splat_undef() {
; CHECK-LABEL: @p3_vec_splat_undef(
; CHECK-NEXT: [[X:%.*]] = call <3 x i8> @gen3x8()
-; CHECK-NEXT: [[TMP1:%.*]] = icmp ugt <3 x i8> [[X]], <i8 3, i8 undef, i8 3>
+; CHECK-NEXT: [[TMP1:%.*]] = icmp ugt <3 x i8> [[X]], <i8 3, i8 3, i8 3>
; CHECK-NEXT: ret <3 x i1> [[TMP1]]
;
%x = call <3 x i8> @gen3x8()
@@ -111,7 +111,7 @@ define <3 x i1> @p3_vec_splat_undef() {
define <3 x i1> @p3_vec_nonsplat_undef() {
; CHECK-LABEL: @p3_vec_nonsplat_undef(
; CHECK-NEXT: [[X:%.*]] = call <3 x i8> @gen3x8()
-; CHECK-NEXT: [[TMP1:%.*]] = icmp ugt <3 x i8> [[X]], <i8 3, i8 undef, i8 15>
+; CHECK-NEXT: [[TMP1:%.*]] = icmp ugt <3 x i8> [[X]], <i8 3, i8 3, i8 15>
; CHECK-NEXT: ret <3 x i1> [[TMP1]]
;
%x = call <3 x i8> @gen3x8()
diff --git a/llvm/test/Transforms/InstCombine/canonicalize-constant-low-bit-mask-and-icmp-ult-to-icmp-ugt.ll b/llvm/test/Transforms/InstCombine/canonicalize-constant-low-bit-mask-and-icmp-ult-to-icmp-ugt.ll
index f6a85d4860ee..728f9e0ddf8a 100644
--- a/llvm/test/Transforms/InstCombine/canonicalize-constant-low-bit-mask-and-icmp-ult-to-icmp-ugt.ll
+++ b/llvm/test/Transforms/InstCombine/canonicalize-constant-low-bit-mask-and-icmp-ult-to-icmp-ugt.ll
@@ -83,7 +83,7 @@ define <2 x i1> @p2_vec_nonsplat_edgecase1(<2 x i8> %x) {
define <3 x i1> @p3_vec_splat_undef(<3 x i8> %x) {
; CHECK-LABEL: @p3_vec_splat_undef(
-; CHECK-NEXT: [[TMP1:%.*]] = icmp ugt <3 x i8> [[X:%.*]], <i8 3, i8 undef, i8 3>
+; CHECK-NEXT: [[TMP1:%.*]] = icmp ugt <3 x i8> [[X:%.*]], <i8 3, i8 3, i8 3>
; CHECK-NEXT: ret <3 x i1> [[TMP1]]
;
%tmp0 = and <3 x i8> %x, <i8 3, i8 undef, i8 3>
@@ -93,7 +93,7 @@ define <3 x i1> @p3_vec_splat_undef(<3 x i8> %x) {
define <3 x i1> @p3_vec_nonsplat_undef(<3 x i8> %x) {
; CHECK-LABEL: @p3_vec_nonsplat_undef(
-; CHECK-NEXT: [[TMP1:%.*]] = icmp ugt <3 x i8> [[X:%.*]], <i8 7, i8 31, i8 undef>
+; CHECK-NEXT: [[TMP1:%.*]] = icmp ugt <3 x i8> [[X:%.*]], <i8 7, i8 31, i8 7>
; CHECK-NEXT: ret <3 x i1> [[TMP1]]
;
%tmp0 = and <3 x i8> %x, <i8 7, i8 31, i8 undef>
More information about the llvm-commits
mailing list