-
Notifications
You must be signed in to change notification settings - Fork 787
[SYCL][UR] Fix device partition queries (2) #10592
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
Fix CSLICE partitioning for PVC. Also, fix level_zero_ext_intel_cslice to correctly test with and without immediate command list. Signed-off-by: Jaime Arteaga <[email protected]> Signed-off-by: Piotr Balcer <[email protected]> Signed-off-by: Fábio Mestre <[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.
Test change LGTM.
@@ -469,6 +469,10 @@ inline pi_result ur2piDeviceInfoValue(ur_device_info_t ParamName, | |||
*pValuePI = pValueUR->value.affinity_domain; | |||
break; | |||
} | |||
case UR_DEVICE_PARTITION_BY_CSLICE: { | |||
*pValuePI = 0; |
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.
what does it mean to return "0" value? why is this a literal "0" and not some enum value?
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.
UR_DEVICE_PARTITION_BY_CSLICE
does not have a value in the corresponding UR union type (or, maybe more accurately, its value is empty). Because the PI's data is as an array of key-value pairs, there's no way to represent an absence of value for a key. That's why the 0 assignment is here, which matches the old behavior.
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 restores the old behavior, but we might want to later revisit the level-zero adapter implementation so that it doesn't hardcode the partition type in the get info method.
@intel/llvm-gatekeepers please merge when possible. |
@jandres742 , post-commit seems to be failing after this: https://github.com/intel/llvm/actions/runs/5731890807/job/15533756331 |
thanks @aelovikov-intel . checking. |
@aelovikov-intel : posted fix on #10649 |
@aelovikov-intel seems like post build is still failing, but logs show something unrelated to these changes https://github.com/intel/llvm/actions/runs/5736337996/job/15545731350
what do you advise? |
We expect #10652 to fix it. |
Fix CSLICE partitioning for PVC. Also, fix level_zero_ext_intel_cslice to correctly test with and without immediate command list. Signed-off-by: Jaime Arteaga <[email protected]> Signed-off-by: Piotr Balcer <[email protected]> Signed-off-by: Fábio Mestre <[email protected]> Signed-off-by: Jaime Arteaga <[email protected]> Signed-off-by: Piotr Balcer <[email protected]> Signed-off-by: Fábio Mestre <[email protected]>
Fix CSLICE partitioning for PVC. Also, fix level_zero_ext_intel_cslice to correctly test with and without immediate command list. Signed-off-by: Jaime Arteaga <[email protected]> Signed-off-by: Piotr Balcer <[email protected]> Signed-off-by: Fábio Mestre <[email protected]> Signed-off-by: Jaime Arteaga <[email protected]> Signed-off-by: Piotr Balcer <[email protected]> Signed-off-by: Fábio Mestre <[email protected]>
Fix CSLICE partitioning for PVC.
Also, fix level_zero_ext_intel_cslice to correctly test with
and without immediate command list.
Signed-off-by: Jaime Arteaga [email protected]
Signed-off-by: Piotr Balcer [email protected]
Signed-off-by: Fábio Mestre [email protected]