-
Notifications
You must be signed in to change notification settings - Fork 769
Task sequence revert #14359
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
Task sequence revert #14359
Conversation
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.
Would be good to understand motivation for the revert - did we decide to drop the feature, or did it cause some regressions which are hard to debug?
cb2de31
to
515cd40
Compare
I agree with Alexey, please include in the description the reasoning for the revert: i.e.: "This change was checked in prematurely as the backend support isn't in yet. When the backend support is in we will re-submit this change." |
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.
commented
8f742bc
to
17b5648
Compare
@AlexeySachkov @GarveyJoe I added to the description the reason behind the revert. |
This reverts commit 92f5b98.
…add new fpga_cluster kernel property (intel#12453)" This reverts commit 7b9001e.
f1b3ba4
to
956da95
Compare
// __SYCL_ASPECT(ext_intel_fpga_task_sequence__, 62) | ||
#define __SYCL_ANY_DEVICE_HAS_ext_intel_fpga_task_sequence__ 0 | ||
#endif | ||
|
||
#ifndef __SYCL_ANY_DEVICE_HAS_ext_oneapi_limited_graph__ | ||
// __SYCL_ASPECT(ext_oneapi_limited_graph, 63) |
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 and the following defines should all have been updated, similar to ALL_DEVICE_HAS
.
The ID specified for ext_oneapi_limited_graph
now does not match the ID defined in aspects.def
anymore. The same applies for all following defines.
@aejjehint Can you please fix this in a follow-up PR?
This PR reverts the aspect numbering that was modified in PR #14359 to avoid breaking existing code.
This change re-apply the task sequence changes from PR #12453 that were reverted in #14359 now that the spec has been approved and the back-end implementation is ready. The spec is also moved from proposed to experimental. --------- Co-authored-by: bowenxue-intel <[email protected]> Co-authored-by: Steffen Larsen <[email protected]>
This reverts PR #12453 and #13080
The original PR introducing task sequence was checked in prematurely as the backend support isn't in yet. When the backend support is in we will re-submit this change.