llvm-project/llvm/test/Analysis/DivergenceAnalysis/AMDGPU
Jay Foad c92e51d84b [SDA] Don't stop divergence propagation at the IPD.
Summary:
This fixes B42473 and B42706.

This patch makes the SDA propagate branch divergence until the end of the RPO traversal. Before, the SyncDependenceAnalysis propagated divergence only until the IPD in rpo order. RPO is incompatible with post dominance in the presence of loops. This made the SDA crash because blocks were missed in the propagation.

Reviewers: foad, nhaehnle

Reviewed By: foad

Subscribers: jvesely, llvm-commits

Tags: #llvm

Differential Revision: https://reviews.llvm.org/D65274

llvm-svn: 372223
2019-09-18 13:40:22 +00:00
..
always_uniform.ll
atomics.ll
b42473-r1-crash.ll [SDA] Don't stop divergence propagation at the IPD. 2019-09-18 13:40:22 +00:00
hidden_diverge.ll [SDA] Bug fix: Use IPD outside the loop as divergence bound 2019-04-18 16:17:35 +00:00
hidden_loopdiverge.ll
interp_f16.ll [AMDGPU] Add intrinsics for 16 bit interpolation 2019-01-28 13:48:59 +00:00
intrinsics.ll [AMDGPU] ImmArg and SourceOfDivergence for permlane/dpp 2019-06-13 16:31:51 +00:00
irreducible.ll
kernel-args.ll
lit.local.cfg
llvm.amdgcn.buffer.atomic.ll [AMDGPU] Flag new raw/struct atomic ops as source of divergence 2019-04-17 14:04:31 +00:00
llvm.amdgcn.image.atomic.ll
no-return-blocks.ll
phi-undef.ll
temporal_diverge.ll
workitem-intrinsics.ll