Displaying 3 results from an estimated 3 matches for "gpudivergenceanalysis".
2019 Jul 22
3
Fwd: bugpoint can't automatically select a safe interpreter!
...pagateBranchDivergence(llvm::Instruction
const&) /home/jayfoad2/git/llvm-project/llvm/lib/Analysis/DivergenceAnalysis.cpp:311:36
#12 0x0000000004c1b106 llvm::DivergenceAnalysis::compute()
/home/jayfoad2/git/llvm-project/llvm/lib/Analysis/DivergenceAnalysis.cpp:386:9
#13 0x0000000004c1b54b
llvm::GPUDivergenceAnalysis::GPUDivergenceAnalysis(llvm::Function&,
llvm::DominatorTree const&, llvm::PostDominatorTree const&,
llvm::LoopInfo const&, llvm::TargetTransformInfo const&)
/home/jayfoad2/git/llvm-project/llvm/lib/Analysis/DivergenceAnalysis.cpp:446:1
#14 0x0000000004c162c5
std::enable_if<!(...
2018 May 28
0
[RFC] A New Divergence Analysis for LLVM
...8: Invariant }
-- The new DivergenceAnalysis --
The proposed DivergenceAnalysis detects divergence in unstructured,
reducible control flow and nested divergent loops. The implementation
provides light-weight frontends for use in VPlan
(LoopDivergenceAnalysis) and the analysis of GPU kernels
(GPUDivergenceAnalysis). Both frontends build on a single, shared
implementation of the actual divergence analysis algorithm.
Apart from solving current issues with divergence detection in LLVM, the
new analysis will enable the adoption of advanced vectorization
techniques in VPlan that are already implemented in RV,...
2019 Jan 31
3
[RFC] Adding thread group semantics to LangRef (motivated by GPUs)
Strong agree with Mehdi, I am also not really sure what is the proposal at
this point so it's hard to comment further.
> There are a number of questions that I have. Do we need better machine
descriptions so that various resources can be considered? Do we need the
capability to reason about the machine state for the cross-lane operations
to enable more optimizations? Are intrinsics the