-
Notifications
You must be signed in to change notification settings - Fork 787
[NFC][SYCL] Remove deprecated spelling of IntelReqdSubGroupSize attribute #2399
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
Commit b2da2c8 added support for new spelling [[intel::reqd_sub_group_size()]] and deprecated previous spelling [[cl::intel_reqd_sub_group_size()]] of the IntelReqdSubGroupSize attribute. This patch removes that deprecated spelling and updates several tests of IntelReqdSubGroupSize attribute. Signed-off-by: Soumi Manna <[email protected]>
Signed-off-by: Soumi Manna <[email protected]>
Signed-off-by: Soumi Manna <[email protected]>
Signed-off-by: Soumi Manna <[email protected]>
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.
sycl/
contents LGTM
cl::sycl::range<1>{this->getOutputBufferSize()}, [= | ||
](cl::sycl::id<1> wiID) [[intel::reqd_sub_group_size(16)]] { |
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.
clang-format does bad job on this constructs. I'd prefer it to be like this:
cl::sycl::range<1>{this->getOutputBufferSize()}, [= | |
](cl::sycl::id<1> wiID) [[intel::reqd_sub_group_size(16)]] { | |
// clang-format off | |
cl::sycl::range<1>{this->getOutputBufferSize()}, | |
[=](cl::sycl::id<1> wiID) [[intel::reqd_sub_group_size(16)]] { | |
// clang-format on |
If you agree, please apply to all such cases.
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.
Yes, I agree. I will update all cases.
asm("barrier"); | ||
#endif | ||
}); | ||
}); |
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.
These changes seem to be unrelated.
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.
Yes, this is unrelated. Clang-format complained about this. I will change this to the original.
[[intel::reqd_sub_group_size( | ||
16)]] { |
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.
[[intel::reqd_sub_group_size( | |
16)]] { | |
// clang-format off | |
[[intel::reqd_sub_group_size(16)]] { | |
// clang-format on |
Same as above.
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.
I will update this too.
"rw"(&b[i] + 16), "rw"(&a[i]), "rw"(&a[i] + 16), | ||
"rw"(&c[i]), "rw"(&c[i] + 16)); |
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.
These changes are also irrelevant.
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.
Yes, this is unrelated. Clang-format complained about this. I will change this to the original.
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.
FE changes look fine
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.
FE changes LGTM
@smanna12, could you address the comments from @alexbatashev, please? |
Signed-off-by: Soumi Manna <[email protected]>
d4eba22
Signed-off-by: Soumi Manna <[email protected]>
Signed-off-by: Soumi Manna <[email protected]>
Signed-off-by: Soumi Manna <[email protected]>
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
Signed-off-by: Soumi Manna <[email protected]>
Signed-off-by: Soumi Manna <[email protected]>
Signed-off-by: Soumi Manna <[email protected]>
Signed-off-by: Soumi Manna <[email protected]>
Signed-off-by: Soumi Manna <[email protected]>
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.
I have updated tests. @alexbatashev, could you please review them?
// clang-format off | ||
a[idx[0]]++; | ||
// clang-format on | ||
#endif | ||
}); |
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.
Clang-format makes unrelated change here. I have turned the format off.
// clang-format off | ||
"rw"(&b[i] + 16), "rw"(&a[i]), "rw"(&a[i] + 16), "rw"(&c[i]), | ||
"rw"(&c[i] + 16)); | ||
#else |
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.
Clang-format makes unrelated change here. I have turned the format off.
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.
FE changes LGTM
Thanks everyone for the review. |
Disable CUDA sycl e2e tests
Commit b2da2c8 added support for new spelling [[intel::reqd_sub_group_size()]]
and deprecated previous spelling [[cl::intel_reqd_sub_group_size()]] of the
IntelReqdSubGroupSize attribute to match with spec.
This patch removes that deprecated spelling and updates several tests
of IntelReqdSubGroupSize attribute with new spelling.
Signed-off-by: Soumi Manna [email protected]