This prototype extends struct rseq with a new sched_state field, which
contains a "on-cpu" flag kept up-to-date by the scheduler.
It is meant to be used by userspace adaptative mutexes to decide between
busy-wait and futex wait system call (releasing the CPU) behaviors based
on the current state of the mutex owner.
The goal is to improve use-cases where the duration of the critical
sections for a given lock follows a multi-modal distribution, preventing
statistical guesses from doing a good job at choosing between busy-wait
and futex wait behavior.
This is in response to the LWN coverage of 2023 Open Source Summit North
America (https://lwn.net/Articles/931789/) unscheduled slot "Adaptive
spinning in user space" presented by André Almeida.
Feedback is welcome!
Mathieu
Mathieu Desnoyers (4):
rseq: Add sched_state field to struct rseq
selftests/rseq: Add sched_state rseq field and getter
selftests/rseq: Implement sched state test program
selftests/rseq: Implement rseq_mutex test program
include/linux/sched.h | 12 ++
include/uapi/linux/rseq.h | 17 +++
kernel/rseq.c | 14 ++
tools/testing/selftests/rseq/.gitignore | 2 +
tools/testing/selftests/rseq/Makefile | 3 +-
tools/testing/selftests/rseq/rseq-abi.h | 17 +++
tools/testing/selftests/rseq/rseq.h | 5 +
tools/testing/selftests/rseq/rseq_mutex.c | 120 ++++++++++++++++++
.../testing/selftests/rseq/sched_state_test.c | 71 +++++++++++
9 files changed, 260 insertions(+), 1 deletion(-)
create mode 100644 tools/testing/selftests/rseq/rseq_mutex.c
create mode 100644 tools/testing/selftests/rseq/sched_state_test.c
--
2.25.1
+Cc Waiman.
On Wed, 17 May 2023, Mathieu Desnoyers wrote:
>This prototype extends struct rseq with a new sched_state field, which
>contains a "on-cpu" flag kept up-to-date by the scheduler.
>
>It is meant to be used by userspace adaptative mutexes to decide between
>busy-wait and futex wait system call (releasing the CPU) behaviors based
>on the current state of the mutex owner.
>
>The goal is to improve use-cases where the duration of the critical
>sections for a given lock follows a multi-modal distribution, preventing
>statistical guesses from doing a good job at choosing between busy-wait
>and futex wait behavior.
>
>This is in response to the LWN coverage of 2023 Open Source Summit North
>America (https://lwn.net/Articles/931789/) unscheduled slot "Adaptive
>spinning in user space" presented by Andr? Almeida.
>
>Feedback is welcome!
>
>Mathieu
>
>Mathieu Desnoyers (4):
> rseq: Add sched_state field to struct rseq
> selftests/rseq: Add sched_state rseq field and getter
> selftests/rseq: Implement sched state test program
> selftests/rseq: Implement rseq_mutex test program
>
> include/linux/sched.h | 12 ++
> include/uapi/linux/rseq.h | 17 +++
> kernel/rseq.c | 14 ++
> tools/testing/selftests/rseq/.gitignore | 2 +
> tools/testing/selftests/rseq/Makefile | 3 +-
> tools/testing/selftests/rseq/rseq-abi.h | 17 +++
> tools/testing/selftests/rseq/rseq.h | 5 +
> tools/testing/selftests/rseq/rseq_mutex.c | 120 ++++++++++++++++++
> .../testing/selftests/rseq/sched_state_test.c | 71 +++++++++++
> 9 files changed, 260 insertions(+), 1 deletion(-)
> create mode 100644 tools/testing/selftests/rseq/rseq_mutex.c
> create mode 100644 tools/testing/selftests/rseq/sched_state_test.c
>
>--
>2.25.1
>
On Wed, 17 May 2023 11:26:50 -0400
Mathieu Desnoyers <[email protected]> wrote:
> This prototype extends struct rseq with a new sched_state field, which
> contains a "on-cpu" flag kept up-to-date by the scheduler.
>
> It is meant to be used by userspace adaptative mutexes to decide between
> busy-wait and futex wait system call (releasing the CPU) behaviors based
> on the current state of the mutex owner.
Woot!
I'm so glad I talked Jon into hanging around for André's spurious talk.
He's the one that brought up rseq (as he mentioned in his article), and I
guess you read that :-)
Unfortunately, I'm in the process of a lot of trips and
meetings/presentations over the next few weeks, and I will not be able to
look into this. But I hope that André could test it out.
Cheers!
-- Steve