[all-commits] [llvm/llvm-project] 28f12a: [RISCV] Teach ComputeNumSignBitsForTargetNode abou...
Alex Bradbury via All-commits
all-commits at lists.llvm.org
Wed Aug 3 05:46:07 PDT 2022
Author: Alex Bradbury <asb at igalia.com>
Date: 2022-08-03 (Wed, 03 Aug 2022)
[RISCV] Teach ComputeNumSignBitsForTargetNode about masked atomic intrinsics
An unnecessary sext.w is generated when masking the result of the
riscv_masked_cmpxchg_i64 intrinsic. Implementing handling of the
intrinsic in ComputeNumSignBitsForTargetNode allows it to be removed.
Although this isn't a particularly important optimisation, removing the
sext.w simplifies implementation of an additional cmpxchg-related
optimisation in D130192.
Although I can't produce a test with different codegen for the other
atomics intrinsics, these are added as well for completeness.
Differential Revision: https://reviews.llvm.org/D130191
More information about the All-commits