kernel: periodically preempt tasks when platform max concurrency is exceeded #12133
+61
−1
Conversation
This file contains hidden or 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
![copybara-service[bot]](https://avatars.githubusercontent.com/in/44061?s=60&v=4){kind=small}
copybara-service
bot
force-pushed
the
test/cl805619701
branch
4 times, most recently
from
1ca3ad8 to
e612003
Compare
…xceeded Add a new method `ConcurrencyCount()` to the `platform.Platform` interface. This method returns the maximum number of tasks that can run in parallel on the platform. For most platforms, this is effectively infinite (math.MaxInt). For the KVM platform, this is the number of available vCPUs. Introduce a new method `Preempt()` to the `platform.Context` interface. The CPU clock ticker now checks if the number of running tasks is greater than Platform.ConcurrencyCount(). If it is, it calls `t.p.Preempt()` on each running task to trigger a preemption. This mechanism ensures that on platforms with a fixed number of execution units, like KVM with a limited number of vCPUs, tasks are periodically preempted to allow other runnable tasks to be scheduled. This prevents any single task from monopolizing a vCPU and improves overall fairness and responsiveness. The `Preempt()` implementation in most platforms is a no-op. In the KVM platform, `Preempt()` is implemented by leveraging the existing interrupt mechanism, effectively signaling the vCPU to exit. PiperOrigin-RevId: 805926296
copybara-service
bot
force-pushed
the
test/cl805619701
branch
from
e612003 to
57697a8
Compare
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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Outdated suggestions cannot be applied.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.
kernel: periodically preempt tasks when platform max concurrency is exceeded
Add a new method
ConcurrencyCount()to theplatform.Platforminterface. Thismethod returns the maximum number of tasks that can run in parallel on the
platform. For most platforms, this is effectively infinite (math.MaxInt). For
the KVM platform, this is the number of available vCPUs.
Introduce a new method
Preempt()to theplatform.Contextinterface. The CPUclock ticker now checks if the number of running tasks is greater than
Platform.ConcurrencyCount(). If it is, it calls
t.p.Preempt()on each runningtask to trigger a preemption.
This mechanism ensures that on platforms with a fixed number of execution units,
like KVM with a limited number of vCPUs, tasks are periodically preempted to
allow other runnable tasks to be scheduled. This prevents any single task from
monopolizing a vCPU and improves overall fairness and responsiveness.
The
Preempt()implementation in most platforms is a no-op. In the KVMplatform,
Preempt()is implemented by leveraging the existing interruptmechanism, effectively signaling the vCPU to exit.