Skip to content

[SYCL] Update the kernel function pointer on function copy/move #20389

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 4 commits into
base: sycl
Choose a base branch
from
Open

[SYCL] Update the kernel function pointer on function copy/move #20389

Changes from 1 commit
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
2708774
[SYCL] Update the kernel function pointer on function copy/move
slawekptak Oct 17, 2025
aa5910c
Add a test for capturing kernel arguments, if the kernel lambda is de…
slawekptak Oct 20, 2025
6613b2f
Change test to use malloc_device
slawekptak Oct 20, 2025
0980dbc
Leave the int kernel arg type
slawekptak Oct 20, 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
9 changes: 8 additions & 1 deletion sycl/source/detail/queue_impl.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -572,12 +572,15 @@ EventImplPtr queue_impl::submit_kernel_direct_impl(
KernelData KData;

KData.setDeviceKernelInfoPtr(DeviceKernelInfo);
KData.setKernelFunc(HostKernel.getPtr());
KData.setNDRDesc(NDRDesc);

auto SubmitKernelFunc = [&](detail::CG::StorageInitHelper &CGData,
bool SchedulerBypass) -> EventImplPtr {
if (SchedulerBypass) {
// No need to copy/move the kernel function, so we set
// the function pointer to the original function
KData.setKernelFunc(HostKernel.getPtr());

return submit_kernel_scheduler_bypass(KData, CGData.MEvents,
CallerNeedsEvent, nullptr, nullptr,
CodeLoc, IsTopCodeLoc);
Expand All @@ -589,6 +592,10 @@ EventImplPtr queue_impl::submit_kernel_direct_impl(
std::shared_ptr<detail::HostKernelBase> HostKernelPtr =
HostKernel.takeOrCopyOwnership();

// When the kernel function is stored for future use,
// set the function pointer to the stored function
KData.setKernelFunc(HostKernelPtr->getPtr());

KData.extractArgsAndReqsFromLambda();

CommandGroup.reset(new detail::CGExecKernel(
Expand Down
Loading