[PATCH] D83153: [DAGCombiner] Prevent regression in isMulAddWithConstProfitable

[Sanjay Patel via Phabricator via llvm-commits]

spatel added inline comments.


================
Comment at: llvm/lib/CodeGen/SelectionDAG/DAGCombiner.cpp:15750-15753
+        int64_t C1 = C1Node->getSExtValue();
+        int64_t C2 = C2Node->getSExtValue();
+        // Do sign extension for c1*c2 according to c2's type.
+        int ShiftBits = 64 - ConstNode.getScalarValueSizeInBits();
----------------
How do we know these constants are not wider than 64-bits?


================
Comment at: llvm/lib/CodeGen/SelectionDAG/DAGCombiner.cpp:15755-15759
+        // This transform will introduce regression, if c1 is legal add
+        // immediate while c1*c2 isn't.
+        const TargetLowering &TLI = DAG.getTargetLoweringInfo();
+        if (TLI.isLegalAddImmediate(C1) && !TLI.isLegalAddImmediate(C1C2))
+          return false;
----------------
lebedev.ri wrote:
> You also need a second piece if the puzzle, because we would have
> already performed this transform before ever getting here:
> https://godbolt.org/z/TE_bzk
> 
> You need to add an inverse transform into DAGCombiner.cpp.
Right - I asked if the motivation for this was GEP math:
http://lists.llvm.org/pipermail/llvm-dev/2020-July/142979.html
...but I did not see an answer.

If yes, then we should have GEP tests. If not, then this patch isn't enough.


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D83153/new/

https://reviews.llvm.org/D83153