This repository has been archived by the owner on Jun 18, 2024. It is now read-only.
Allow clean exits from the scheduler with scx_bpf_exit() #166
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
|
Spoke offline with Tejun. Going to change a few things around and send a v2 |
Byte-Lab
force-pushed
the
clean_exit_bpf
branch
from
b80ef80
to
b300b36
Compare
|
Oops, missed fixing up scx_ops_exit_kind(). Will resend another patch shortly |
There may be cases where a scheduler wants to cause itself to exit under normal circumstances. For example, if a scheduler does not support hotplug, it may want to exit to user space if a CPU is hotplugged or unplugged to let it know that it should restart itself. To enable this, we implement a new kfunc called scx_bpf_exit_bstr(). This kfunc shares a good amount of logic with scx_bpf_error_bstr(), as both of them can plumb an exit message from BPF to user space. In addition, we also augment struct scx_exit_info to include an exit code field. Signed-off-by: David Vernet <[email protected]>
Byte-Lab
force-pushed
the
clean_exit_bpf
branch
from
b300b36
to
ae74764
Compare
As with scx_bpf_error(), we want schedulers to have an ergonomic API for exiting execution. Let's add an scx_bpf_exit() macro wrapper that does nearly the same thing as scx_bpf_error(), but invokes scx_bpf_exit_bstr() instead of scx_bpf_error_bstr(). Signed-off-by: David Vernet <[email protected]>
We no longer have scx_bpf_switch_all(). Let's update the test to use __COMPAT_SCX_OPS_SWITCH_PARTIAL. Along the way, make it less flaky. Signed-off-by: David Vernet <[email protected]>
Let's add a selftest that verifies us using scx_bpf_exit() to exit from various callbacks in the program. Signed-off-by: David Vernet <[email protected]>
Byte-Lab
force-pushed
the
clean_exit_bpf
branch
from
ae74764
to
14162f7
Compare
|
@htejun, this should be ready for another look |
htejun
approved these changes
Mar 28, 2024
Sign up for free
to subscribe to this conversation on GitHub.
Already have an account?
Sign in.
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
There may be cases where a scheduler wants to cause itself to exit under
normal circumstances. For example, if a scheduler does not support
hotplug, it may want to exit to user space if a CPU is hotplugged or
unplugged to let it know that it should restart itself.
To enable this, we implement a new kfunc called scx_bpf_exit_bstr().
This kfunc shares a good amount of logic with scx_bpf_error_bstr(), as
both of them can plumb an exit message from BPF to user space. In
addition, we also augment struct scx_exit_info to include an exit code
field.