-
Notifications
You must be signed in to change notification settings - Fork 769
[SYCL] Switch to preview implementation of sycl::vec
#14317
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.
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
Conversation
This comment was marked as outdated.
This comment was marked as outdated.
e5ccd59
to
325d9c2
Compare
325d9c2
to
519e114
Compare
519e114
to
24e2b8a
Compare
@@ -1429,8 +1429,7 @@ class SwizzleOp { | |||
DataT getValue(EnableIfMultipleIndexes<IdxNum, size_t> Index) const { | |||
if (std::is_same_v<OperationCurrentT<DataT>, GetOp<DataT>>) { | |||
std::array<int, getNumElements()> Idxs{Indexes...}; | |||
// Cast required for int8_t -> std::byte | |||
return static_cast<DataT>(m_Vector->getValue(Idxs[Index])); | |||
return (*m_Vector)[Idxs[Index]]; |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
This is the bugfix mentioned in the description.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
LGTM. Thanks :)
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
LGTM!
CUDA testing succeeded in https://github.com/intel/llvm/actions/runs/9963598138, re-run was triggered by re-run of the AMDGPU task that was necessary due to the infrastructure issues on the runner (GPU not visible). |
I'm keeping the
vector_preview.hpp
file name to avoid misleading diff in this PR.Also, a minor fix in swizzles implementation is included in this change fixing a bug that would have affected
SubGroup/load_store.cpp
E2E test otherwise.