[SYCL][E2E] Update some incompatible requirements #18007
Conversation
There test are neved launched due to the REQUIRES set. Update these sets and remove accelerator-related parts if needed, otherwise some test may fail due to lack of device.
KornevNikita
temporarily deployed
to
WindowsCILock
GitHub Actions
Inactive
KornevNikita
temporarily deployed
to
WindowsCILock
GitHub Actions
Inactive
KornevNikita
temporarily deployed
to
WindowsCILock
GitHub Actions
Inactive
KornevNikita
temporarily deployed
to
WindowsCILock
GitHub Actions
Inactive
KornevNikita
temporarily deployed
to
WindowsCILock
GitHub Actions
Inactive
KornevNikita
had a problem deploying
to
WindowsCILock
GitHub Actions
Error
KornevNikita
had a problem deploying
to
WindowsCILock
GitHub Actions
Error
KornevNikita
force-pushed
the
upd-some-e2e
branch
from
daaf695 to
e856d63
Compare
KornevNikita
temporarily deployed
to
WindowsCILock
GitHub Actions
Inactive
KornevNikita
temporarily deployed
to
WindowsCILock
GitHub Actions
Inactive
KornevNikita
temporarily deployed
to
WindowsCILock
GitHub Actions
Inactive
KornevNikita
temporarily deployed
to
WindowsCILock
GitHub Actions
Inactive
| @@ -1,4 +1,4 @@ | |||
| // REQUIRES: linux, cpu && (gpu && level_zero) | |||
| // REQUIRES: linux, cpu || (gpu && level_zero) | |||
There was a problem hiding this comment.
According to other tests in this dir - I believe it's just a typo.
There was a problem hiding this comment.
No, this is not a typo, but maybe I used wrong pattern.
I want this test can be tested on machines which have both CPU and GPU, as you can see
sycl::queue gpu_queue(sycl::gpu_selector_v);
sycl::queue cpu_queue(sycl::cpu_selector_v);
There was a problem hiding this comment.
Okay, I got it, I'll update the REQUIRES string
There was a problem hiding this comment.
I've updated this string - 7163eb7
Anyways, looks like this test wasn't launched in pre-commit on gen12 - it's in the list of unsupported tests.
@intel/dpcpp-devops-reviewers could you help me to figure out if it's true?
There was a problem hiding this comment.
yep, definitely looks like it didn't run. i dont think cpu and gpu will ever both be true, i think we need any-device-is-cpu && any-device-is-gpu
There was a problem hiding this comment.
@sarnex take a look at this test - https://github.com/intel/llvm/blob/sycl/sycl/test-e2e/AddressSanitizer/aot/cpu.cpp
the REQUIRES there is even simpler and it also uses run-aux, but it's also in the list of unsupported tests
There was a problem hiding this comment.
ah ok yeah good find, let me actually debug it, thanks
There was a problem hiding this comment.
There was a problem hiding this comment.
lol okay, then I think we can leave the test as is. it should work when the issue with gen12 is fixed.
@intel/dpcpp-sanitizers-review take a look please.
There was a problem hiding this comment.
Thank you for correcting the require pattern.
Yep, we haven't fixed the GEN12 issue yet, because it's low priority.
We will take a look at GEN12 in the future.
|
I've deleted selector-tests-related changes, as these tests require refactoring / converting to unit tests. @maarquitos14 could you please take one more look just in case. |
|
Unrelated failure |
These test are never launched due to the REQUIRES set.