Skip to content

Remove unused tsan ignores #360

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
wants to merge 3 commits into
base: rocm-jaxlib-v0.6.0
Choose a base branch
from
+6 −8
Open

Remove unused tsan ignores #360

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
7438305
[ROCm] Disable Cudnn fusions
draganmladjenovic Sep 17, 2025
ca0f4c6
Remove unused ignore
alekstheod Sep 18, 2025
0082e18
Remove more ignores
alekstheod Sep 18, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
4 changes: 0 additions & 4 deletions build_tools/rocm/tsan_ignore_list.txt
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -15,17 +15,13 @@ race:xla::gpu::AllocateDestinationBuffer
race:xla::LocalDeviceState::ThenRelease

# To be fixed
race:xla::LiteralBase::Piece::DeallocateBuffers
race:xla::PjRtStreamExecutorLoadedExecutable::ExecuteHelper
race:xla::PjRtStreamExecutorClient::BufferFromHostBufferInternal
race:xla::HloRunnerPjRt::TransferLiteralsFromDevice
race:xla::MutableLiteralBase::~MutableLiteralBase
race:xla::MutableLiteralBase::PopulateR1<int>
race:xla::xla::gpu::GpuCompiler::CompileSingleModule
race:xla::LiteralBase::Piece::Storage::Storage
race:xla::LocalClient::TransferFromOutfeedLocal
race:llvm::cl::opt_storage<bool, false, false>::setValue<int>
race:xla::gpu::(anonymous namespace)::RecoverExp2Pattern::initStaticsIfNeeded<std::tuple<mlir::Float32Type, mlir::FloatTF32Type, mlir::BFloat16Type>>
race:lld::lldMain
race:llvm::*
race:xla::gpu::GpuExecutable::ExecuteAsyncOnStream
4 changes: 2 additions & 2 deletions xla/service/gpu/amdgpu_compiler.cc
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -122,8 +122,8 @@ absl::Status AMDGPUCompiler::OptimizeHloConvolutionCanonicalization(
stream_executor::RocmSolverContext::Create);
pipeline.AddPass<ConvRewriter>(gpu_version);
pipeline.AddPass<ConvPaddingLegalization>();
auto rcc = std::get<se::RocmComputeCapability>(gpu_version);
pipeline.AddPass<CudnnFusedConvRewriter>(rcc, dnn_version, toolkit_version);
// auto rcc = std::get<se::RocmComputeCapability>(gpu_version);
// pipeline.AddPass<CudnnFusedConvRewriter>(rcc, dnn_version, toolkit_version);
Copy link

Copilot AI Sep 18, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Commented-out code should be removed rather than left as comments. If this code needs to be preserved for future reference, consider documenting the reason in a TODO comment or tracking it in an issue.

Suggested change
// pipeline.AddPass<CudnnFusedConvRewriter>(rcc, dnn_version, toolkit_version);

Copilot uses AI. Check for mistakes.

// The conv padding/vectorization passes which we need to get rid of. They
// also leave behind unnecessary tuple/get-tuple-element pairs that
Expand Down
6 changes: 4 additions & 2 deletions xla/service/gpu/transforms/BUILD
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1009,6 +1009,9 @@ cc_library(
xla_test(
name = "cudnn_fused_conv_rewriter_test",
srcs = ["cudnn_fused_conv_rewriter_test.cc"],
tags = [
"cuda-only",
],
backend_tags = {
"gpu_a100": [
"noasan",
Expand All @@ -1017,8 +1020,7 @@ xla_test(
},
backends = [
"gpu_a100",
"gpu_amd_any",
] + if_oss(["gpu_any"]),
],
shard_count = 10,
deps = [
":conv_rewriter",
Expand Down
Loading