[all-commits] [llvm/llvm-project] fb0190: [AMDGPU] CallInst -> CallBase for divergence analysis
Robert Imschweiler via All-commits
all-commits at lists.llvm.org
Sat Oct 3 09:47:09 PDT 2026
Branch: refs/heads/users/ro-i/callbase-divergence
Home: https://github.com/llvm/llvm-project
Commit: fb0190dc187d65100f9585e70faf85e7c6351a84
https://github.com/llvm/llvm-project/commit/fb0190dc187d65100f9585e70faf85e7c6351a84
Author: Robert Imschweiler <robert.imschweiler at amd.com>
Date: 2026-10-03 (Sat, 03 Oct 2026)
Changed paths:
M llvm/lib/Target/AMDGPU/AMDGPUTargetTransformInfo.cpp
M llvm/lib/Target/AMDGPU/AMDGPUTargetTransformInfo.h
M llvm/test/CodeGen/AMDGPU/infinite-loop.ll
M llvm/test/CodeGen/AMDGPU/si-unify-exit-multiple-unreachables.ll
Log Message:
-----------
[AMDGPU] CallInst -> CallBase for divergence analysis
This handles non-call instructions like callbr.
The test changes in infinite-loop.ll, where there are now two return
blocks (the original one + the dummy) instead of the unified return
block is a symptom of the callbr in those functions being not marked as
divergent anymore.
This patch was hand-manufactured using Neovim.
To unsubscribe from these emails, change your notification settings at https://github.com/llvm/llvm-project/settings/notifications
More information about the All-commits
mailing list