[PATCH] D92056: [LoopVec] Support global addresses as argument to uniform mem ops

Philip Reames via Phabricator via llvm-commits llvm-commits at lists.llvm.org
Tue Nov 24 12:34:13 PST 2020


reames created this revision.
reames added reviewers: fhahn, anna.
Herald added subscribers: dantrushin, bollu, hiraditya, mcrosier.
Herald added a project: LLVM.
reames requested review of this revision.

The initial step of the uniform-after-vectorization (lane-0 demanded only) analysis was very awkwardly written.  It would revisit use list of each pointer operand of a widened load/store.  As a result, it was in the worst case O(N^2) where N was the number of instructions in a loop, and had restricted operand Value types to reduce the size of use lists.

This patch replaces the original algorithm with one which is at most O(2N) in the number of instructions in the loop.  (The key observation is that each use of a potentially interesting pointer is visited at most twice, once on first scan, once in the use list of *it's* operand.  Only instructions within the loop have their uses scanned.)


Repository:
  rG LLVM Github Monorepo

https://reviews.llvm.org/D92056

Files:
  llvm/lib/Transforms/Vectorize/LoopVectorize.cpp
  llvm/test/Transforms/LoopVectorize/X86/cost-model-assert.ll
  llvm/test/Transforms/LoopVectorize/X86/uniform_mem_op.ll
  llvm/test/Transforms/LoopVectorize/pr44488-predication.ll

-------------- next part --------------
A non-text attachment was scrubbed...
Name: D92056.307435.patch
Type: text/x-patch
Size: 24831 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20201124/338a4d48/attachment.bin>


More information about the llvm-commits mailing list