Skip to content
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

Jump to bottom

[SYCL][Fusion][NFC] Remove name and ticket ID from TODOs #12620

Merged
merged 1 commit into from
Feb 6, 2024
Merged

[SYCL][Fusion][NFC] Remove name and ticket ID from TODOs #12620

Changes from all commits
Commits
Show all changes
1 commit
Select commit
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
16 changes: 7 additions & 9 deletions sycl/source/detail/jit_compiler.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -667,8 +667,8 @@ jit_compiler::fuseKernels(QueueImplPtr Queue,
unsigned KernelIndex = 0;
ParamList FusedParams;
PromotionMap PromotedAccs;
// TODO(Lukas, ONNX-399): Collect information about streams and auxiliary
// resources (which contain reductions) and figure out how to fuse them.
// TODO: Collect information about streams and auxiliary resources (which
// contain reductions) and figure out how to fuse them.
for (auto &RawCmd : InputKernels) {
auto *KernelCmd = static_cast<ExecCGCommand *>(RawCmd);
auto &CG = KernelCmd->getCG();
Expand Down Expand Up @@ -760,8 +760,7 @@ jit_compiler::fuseKernels(QueueImplPtr Queue,
}
}

// TODO(Lukas, ONNX-399): Check for the correct kernel bundle state of the
// device image?
// TODO: Check for the correct kernel bundle state of the device image?
auto &RawDeviceImage = DeviceImage->getRawData();
auto DeviceImageSize = static_cast<size_t>(RawDeviceImage.BinaryEnd -
RawDeviceImage.BinaryStart);
Expand Down Expand Up @@ -803,8 +802,7 @@ jit_compiler::fuseKernels(QueueImplPtr Queue,
// Not all overloads of parallel_for_work_group only specify the number of
// work-groups, so the above mechanism might not detect all hierarchical
// parallelism.
// TODO(Lukas, CRD-6): Find a more reliable way to detect hierarchical
// parallelism.
// TODO: Find a more reliable way to detect hierarchical parallelism.
}

// We need to copy the storages here. The input CGs might be eliminated
Expand All @@ -815,9 +813,9 @@ jit_compiler::fuseKernels(QueueImplPtr Queue,
KernelCG->getArgsStorage().end());
AccStorage.insert(AccStorage.end(), KernelCG->getAccStorage().begin(),
KernelCG->getAccStorage().end());
// TODO(Lukas, ONNX-399): Does the MSharedPtrStorage contain any
// information about actual shared pointers beside the kernel bundle and
// handler impl? If yes, we might need to copy it here.
// TODO: Does the MSharedPtrStorage contain any information about actual
// shared pointers beside the kernel bundle and handler impl? If yes, we
// might need to copy it here.
Requirements.insert(Requirements.end(), KernelCG->getRequirements().begin(),
KernelCG->getRequirements().end());
Events.insert(Events.end(), KernelCG->getEvents().begin(),
Expand Down
Loading