[PATCH] D124657: [AMDGPU][ISEL] Directly custom lower INSERT_SUBVECTOR instead of via INSERT_VECTOR_ELT

[Mahesha S via Phabricator via llvm-commits]

hsmhsm created this revision.
hsmhsm added reviewers: arsenm, foad, rampitec.
Herald added subscribers: kerbowa, hiraditya, t-tye, tpr, dstuttard, yaxunl, nhaehnle, jvesely, kzhuravl.
Herald added a project: All.
hsmhsm requested review of this revision.
Herald added subscribers: llvm-commits, wdng.
Herald added a project: LLVM.

When the target vector size is <= 64, we can directly custom lower INSERT_SUBVECTOR
instead taking it through INSERT_VECTOR_ELT.

Custom lowering of INSERT_SUBVECTOR for the cases where target vector size > 64 is not
really happening at the moment. We need to handle it in a generic way for all possible
allowed vec sizes. This patch is a kind of pre-checkin patch to handle it.


Repository:
  rG LLVM Github Monorepo

https://reviews.llvm.org/D124657

Files:
  llvm/lib/Target/AMDGPU/SIISelLowering.cpp
  llvm/test/CodeGen/AMDGPU/vector_shuffle.packed.ll

-------------- next part --------------
A non-text attachment was scrubbed...
Name: D124657.425974.patch
Type: text/x-patch
Size: 6988 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20220429/c3ebc9e5/attachment.bin>