[PATCH] D104156: [DAGCombine][X86][ARM] EXTRACT_SUBVECTOR(VECTOR_SHUFFLE(?,?,Mask)) -> VECTOR_SHUFFLE(EXTRACT_SUBVECTOR(?, ?), EXTRACT_SUBVECTOR(?, ?), Mask')
Simon Pilgrim via Phabricator via llvm-commits
llvm-commits at lists.llvm.org
Wed Nov 3 05:26:13 PDT 2021
- Previous message: [PATCH] D104156: [DAGCombine][X86][ARM] EXTRACT_SUBVECTOR(VECTOR_SHUFFLE(?,?,Mask)) -> VECTOR_SHUFFLE(EXTRACT_SUBVECTOR(?, ?), EXTRACT_SUBVECTOR(?, ?), Mask')
- Next message: [PATCH] D104156: [DAGCombine][X86][ARM] EXTRACT_SUBVECTOR(VECTOR_SHUFFLE(?,?,Mask)) -> VECTOR_SHUFFLE(EXTRACT_SUBVECTOR(?, ?), EXTRACT_SUBVECTOR(?, ?), Mask')
- Messages sorted by:
[ date ]
[ thread ]
[ subject ]
[ author ]
RKSimon added inline comments.
================
Comment at: llvm/lib/CodeGen/SelectionDAG/DAGCombiner.cpp:20684
+ !TLI.isOperationLegalOrCustom(ISD::VECTOR_SHUFFLE, NarrowVT))
+ return SDValue();
+
----------------
lebedev.ri wrote:
> RKSimon wrote:
> > Shouldn't this be done earlier as soon as we know NarrowVT to early-out ?
> Well, yes and no. As it can be seen in the comment for the previous code block,
> we then would loose `return DAG.getUNDEF(NarrowVT);` constant-fold case.
> Should we not have it?
I suppose it depends if these UNDEF paths are actually active?
Repository:
rG LLVM Github Monorepo
CHANGES SINCE LAST ACTION
https://reviews.llvm.org/D104156/new/
https://reviews.llvm.org/D104156
- Previous message: [PATCH] D104156: [DAGCombine][X86][ARM] EXTRACT_SUBVECTOR(VECTOR_SHUFFLE(?,?,Mask)) -> VECTOR_SHUFFLE(EXTRACT_SUBVECTOR(?, ?), EXTRACT_SUBVECTOR(?, ?), Mask')
- Next message: [PATCH] D104156: [DAGCombine][X86][ARM] EXTRACT_SUBVECTOR(VECTOR_SHUFFLE(?,?,Mask)) -> VECTOR_SHUFFLE(EXTRACT_SUBVECTOR(?, ?), EXTRACT_SUBVECTOR(?, ?), Mask')
- Messages sorted by:
[ date ]
[ thread ]
[ subject ]
[ author ]
More information about the llvm-commits
mailing list