[PATCH] D115536: [RISCV] Don't use VLMAX for start value splat in reduction lowering.

[Craig Topper via Phabricator via llvm-commits]

craig.topper created this revision.
craig.topper added reviewers: frasercrmck, rogfer01, HsiangKai, arcbbb, khchen.
Herald added subscribers: VincentWu, luke957, achieveartificialintelligence, StephenFan, vkmr, evandro, luismarques, apazos, sameer.abuasal, s.egerton, Jim, benna, psnobl, jocewei, PkmX, the_o, brucehoult, MartinMosbeck, edward-jones, zzheng, jrtc27, kito-cheng, niosHD, sabuasal, simoncook, johnrusso, rbar, asb, hiraditya.
craig.topper requested review of this revision.
Herald added a subscriber: MaskRay.
Herald added a project: LLVM.

The reduction instructions only reads the first element. The
execution time for a splat may take longer with a larger VL.
We should use the smallest VL we can.


Repository:
  rG LLVM Github Monorepo

https://reviews.llvm.org/D115536

Files:
  llvm/lib/Target/RISCV/RISCVISelLowering.cpp
  llvm/test/CodeGen/RISCV/rvv/fixed-vectors-reduction-fp-vp.ll
  llvm/test/CodeGen/RISCV/rvv/fixed-vectors-reduction-fp.ll
  llvm/test/CodeGen/RISCV/rvv/fixed-vectors-reduction-int-vp.ll
  llvm/test/CodeGen/RISCV/rvv/fixed-vectors-reduction-int.ll
  llvm/test/CodeGen/RISCV/rvv/vreductions-fp-sdnode.ll
  llvm/test/CodeGen/RISCV/rvv/vreductions-fp-vp.ll
  llvm/test/CodeGen/RISCV/rvv/vreductions-int-rv32.ll
  llvm/test/CodeGen/RISCV/rvv/vreductions-int-rv64.ll
  llvm/test/CodeGen/RISCV/rvv/vreductions-int-vp.ll

-------------- next part --------------
A non-text attachment was scrubbed...
Name: D115536.393521.patch
Type: text/x-patch
Size: 350800 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20211210/61b5a155/attachment-0001.bin>