[SYCL] Re-enable several joint matrix E2E tests #15473
Conversation
lbushi25
temporarily deployed
to
WindowsCILock
GitHub Actions
Inactive
lbushi25
temporarily deployed
to
WindowsCILock
GitHub Actions
Inactive
lbushi25
had a problem deploying
to
WindowsCILock
GitHub Actions
Error
lbushi25
had a problem deploying
to
WindowsCILock
GitHub Actions
Error
lbushi25
had a problem deploying
to
WindowsCILock
GitHub Actions
Error
lbushi25
had a problem deploying
to
WindowsCILock
GitHub Actions
Error
lbushi25
had a problem deploying
to
WindowsCILock
GitHub Actions
Error
lbushi25
had a problem deploying
to
WindowsCILock
GitHub Actions
Error
lbushi25
temporarily deployed
to
WindowsCILock
GitHub Actions
Inactive
lbushi25
changed the title
lbushi25
changed the title
lbushi25
temporarily deployed
to
WindowsCILock
GitHub Actions
Inactive
| // RUN: %{run} %t.out | ||
|
|
||
| // Currently row major B fails when annotated_ptr is used | ||
| // XFAIL: gpu |
There was a problem hiding this comment.
Does this test pass on PVC? I did not see any unexpected pass on PVC for this test...
There was a problem hiding this comment.
I forgot to test this on PVC. I just tested it now and yes you are right, it fails.
I'm adding an unsupported directive for pvc on these tests.
There was a problem hiding this comment.
The test is supported on PVC but it is XFAIL.
So you should keep both:
// UNSUPPORTED: gpu-intel-dg2
and
// XFAIL: gpu
(same goes to the other tests)
| // This tests support of col major layout for matrix B which does transpose and | ||
| // then VNNI transform. This is currently only available on AMX | ||
|
|
||
| // XFAIL: gpu |
There was a problem hiding this comment.
Does this test pass on PVC? I did not see any unexpected pass on PVC for this test...
| // RUN: %{build} -o %t.out | ||
| // RUN: %{run} %t.out | ||
|
|
||
| // XFAIL:* |
There was a problem hiding this comment.
Does this test pass on PVC? I did not see any unexpected pass on PVC for this test...
| // This tests support of col major layout for matrix B which does transpose and | ||
| // then VNNI transform. This is currently only available on AMX | ||
|
|
||
| // XFAIL: gpu |
There was a problem hiding this comment.
Does this test pass on PVC? I did not see any unexpected pass on PVC for this test...
| // This tests support of col major layout for matrix B which does transpose and | ||
| // then VNNI transform. This is currently only available on AMX | ||
|
|
||
| // XFAIL: gpu |
There was a problem hiding this comment.
Does this test pass on PVC? I did not see any unexpected pass on PVC for this test...
| // SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception | ||
| // | ||
| //===----------------------------------------------------------------------===// | ||
| // VNNI transform and sub-group size 32 are not supported yet on DG2 by IGC |
There was a problem hiding this comment.
sub-group size 32 part is not relevant for this test. To avoid confusion, I suggest mention only VNNI transform is not supported.
Edit: also, after looking at implementation I see that some VNNI transform is implemented on DG2 in IGC, so in general this statement is not correct. It needs to be more specific. Same is true for other messages below, which mention VNNI transform is not supported.
| // RUN: %{build} -o %t.out | ||
| // RUN: %{run} %t.out | ||
|
|
||
| // XFAIL:* |
There was a problem hiding this comment.
Does this test pass on PVC? I did not see any unexpected pass on PVC for this test...
| // SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception | ||
| // | ||
| //===----------------------------------------------------------------------===// | ||
| // VNNI transform and sub-group size 32 are not supported yet on DG2 by IGC |
There was a problem hiding this comment.
sub-group size 32 part is not relevant for this test. VNNI transform part of the statement needs to be more specific.
| // This tests support of col major layout for matrix B which does transpose and | ||
| // then VNNI transform. This is currently only available on AMX | ||
|
|
||
| // XFAIL: gpu |
There was a problem hiding this comment.
Does this test pass on PVC? I did not see any unexpected pass on PVC for this test...
YuriPlyakhin
left a comment
YuriPlyakhin
left a comment
There was a problem hiding this comment.
I think this change should not be committed. Please, check my comments. Thanks.
| // SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception | ||
| // | ||
| //===----------------------------------------------------------------------===// | ||
| // VNNI transform and sub-group size 32 are not supported yet on DG2 by IGC |
There was a problem hiding this comment.
sub-group size 32 part is not relevant for this test. VNNI transform part needs to be more specific
lbushi25
temporarily deployed
to
WindowsCILock
GitHub Actions
Inactive
lbushi25
temporarily deployed
to
WindowsCILock
GitHub Actions
Inactive
Ok, closing this PR. Instead, I'm creating an issue to track the fail on PVC and later add a reference to that issue as a comment in the test. |
3 participants
Removed the XFAIL directive from several joint matrix tests as they are now passing on all devices except for DG2 GPU because apparently it does not yet support sub-group size of 32 for joint matrix and it also does not support VNNI transforms. Hence, these tests are marked as UNSUPPORTED on DG2 and on all other devices they are now expected to pass.