[X86][AVX] lowerShuffleAsLanePermuteAndShuffle - consistently normalize multi-input...

We only use lowerShuffleAsLanePermuteAndShuffle for unary shuffles at the moment, but we should consistently handle lane index calculations for multiple inputs in both the AVX1 and AVX2 paths.

Minor (almost NFC) tidyup as I'm hoping to use lowerShuffleAsLanePermuteAndShuffle for binary shuffles soon.

diff --git a/llvm/lib/Target/X86/X86ISelLowering.cpp b/llvm/lib/Target/X86/X86ISelLowering.cpp
index 93ef3a0..7efb305 100644 (file)
--- a/llvm/lib/Target/X86/X86ISelLowering.cpp
+++ b/llvm/lib/Target/X86/X86ISelLowering.cpp
@@ -14971,7 +14971,7 @@ static SDValue lowerShuffleAsLanePermuteAndShuffle(
   if (!Subtarget.hasAVX2()) {
     bool LaneCrossing[2] = {false, false};
     for (int i = 0; i < Size; ++i)
-      if (Mask[i] >= 0 && (Mask[i] % Size) / LaneSize != i / LaneSize)
+      if (Mask[i] >= 0 && ((Mask[i] % Size) / LaneSize) != (i / LaneSize))
         LaneCrossing[(Mask[i] % Size) / LaneSize] = true;
     if (!LaneCrossing[0] || !LaneCrossing[1])
       return splitAndLowerShuffle(DL, VT, V1, V2, Mask, DAG);
@@ -14979,7 +14979,7 @@ static SDValue lowerShuffleAsLanePermuteAndShuffle(
     bool LaneUsed[2] = {false, false};
     for (int i = 0; i < Size; ++i)
       if (Mask[i] >= 0)
-        LaneUsed[(Mask[i] / LaneSize)] = true;
+        LaneUsed[(Mask[i] % Size) / LaneSize] = true;
     if (!LaneUsed[0] || !LaneUsed[1])
       return splitAndLowerShuffle(DL, VT, V1, V2, Mask, DAG);
   }