[llvm] a1e0aa7 - [X86] combineMulToPMADDWD - remove useless TODO

[Simon Pilgrim via llvm-commits]

Author: Simon Pilgrim
Date: 2021-11-10T16:56:44Z
New Revision: a1e0aa75ca019e8c4c506640d712b9f0546547b1

URL: https://github.com/llvm/llvm-project/commit/a1e0aa75ca019e8c4c506640d712b9f0546547b1
DIFF: https://github.com/llvm/llvm-project/commit/a1e0aa75ca019e8c4c506640d712b9f0546547b1.diff

LOG: [X86] combineMulToPMADDWD - remove useless TODO

We should always be able to use PMULUDQ/PMULDQ in PMADDWD patterns with greater than 32-bit extended integer sources

Added: 
    

Modified: 
    llvm/lib/Target/X86/X86ISelLowering.cpp

Removed: 
    


################################################################################
diff  --git a/llvm/lib/Target/X86/X86ISelLowering.cpp b/llvm/lib/Target/X86/X86ISelLowering.cpp
index f58455e4cd79..b07a1e8ba21d 100644
--- a/llvm/lib/Target/X86/X86ISelLowering.cpp
+++ b/llvm/lib/Target/X86/X86ISelLowering.cpp
@@ -44441,7 +44441,6 @@ static SDValue combineMulToPMADDWD(SDNode *N, SelectionDAG &DAG,
   EVT VT = N->getValueType(0);
 
   // Only support vXi32 vectors.
-  // TODO: Can we support > 32-bit elements?
   if (!VT.isVector() || VT.getVectorElementType() != MVT::i32)
     return SDValue();