[llvm] 34f0a8a - [SLP]Fix comparison in bitwidth check.

[Alexey Bataev via llvm-commits]

Author: Alexey Bataev
Date: 2024-03-21T04:24:34-07:00
New Revision: 34f0a8aaba11bf703ddd2de92eee8ecbb77be5c8

URL: https://github.com/llvm/llvm-project/commit/34f0a8aaba11bf703ddd2de92eee8ecbb77be5c8
DIFF: https://github.com/llvm/llvm-project/commit/34f0a8aaba11bf703ddd2de92eee8ecbb77be5c8.diff

LOG: [SLP]Fix comparison in bitwidth check.

Projected bitwidth should be less than the original, not greater.

Added: 
    llvm/test/Transforms/SLPVectorizer/orig-btiwidth-les-projected.ll

Modified: 
    llvm/lib/Transforms/Vectorize/SLPVectorizer.cpp

Removed: 
    


################################################################################
diff  --git a/llvm/lib/Transforms/Vectorize/SLPVectorizer.cpp b/llvm/lib/Transforms/Vectorize/SLPVectorizer.cpp
index a52064e5417b27..4853c2006fea06 100644
--- a/llvm/lib/Transforms/Vectorize/SLPVectorizer.cpp
+++ b/llvm/lib/Transforms/Vectorize/SLPVectorizer.cpp
@@ -13995,7 +13995,7 @@ bool BoUpSLP::collectValuesToDemote(
     if (MultiNodeScalars.contains(V))
       return false;
     uint32_t OrigBitWidth = DL->getTypeSizeInBits(V->getType());
-    if (OrigBitWidth < BitWidth) {
+    if (OrigBitWidth > BitWidth) {
       APInt Mask = APInt::getBitsSetFrom(OrigBitWidth, BitWidth);
       if (MaskedValueIsZero(V, Mask, SimplifyQuery(*DL)))
         return true;

diff  --git a/llvm/test/Transforms/SLPVectorizer/orig-btiwidth-les-projected.ll b/llvm/test/Transforms/SLPVectorizer/orig-btiwidth-les-projected.ll
new file mode 100644
index 00000000000000..531e964053482e
--- /dev/null
+++ b/llvm/test/Transforms/SLPVectorizer/orig-btiwidth-les-projected.ll
@@ -0,0 +1,22 @@
+; NOTE: Assertions have been autogenerated by utils/update_test_checks.py UTC_ARGS: --version 4
+; RUN: opt -S --passes=slp-vectorizer < %s | FileCheck %s
+
+define i32 @test(i4 %0) {
+; CHECK-LABEL: define i32 @test(
+; CHECK-SAME: i4 [[TMP0:%.*]]) {
+; CHECK-NEXT:  entry:
+; CHECK-NEXT:    [[TMP1:%.*]] = trunc i8 0 to i4
+; CHECK-NEXT:    [[TMP2:%.*]] = trunc i8 0 to i4
+; CHECK-NEXT:    [[ADD_R:%.*]] = or i4 [[TMP1]], [[TMP0]]
+; CHECK-NEXT:    [[ADD_R14:%.*]] = or i4 0, [[TMP2]]
+; CHECK-NEXT:    [[CMP_NOT:%.*]] = icmp eq i4 [[ADD_R]], [[ADD_R14]]
+; CHECK-NEXT:    ret i32 0
+;
+entry:
+  %1 = trunc i8 0 to i4
+  %2 = trunc i8 0 to i4
+  %add.r = or i4 %1, %0
+  %add.r14 = or i4 0, %2
+  %cmp.not = icmp eq i4 %add.r, %add.r14
+  ret i32 0
+}