This series aims to get rid of some places where locks of two RQs are held
at the same time.
Patch [1/5] is a preparation/cleanup. It replaces old (.on_rq == 1) with new
(.on_rq == ONRQ_QUEUED) everywhere. No functional changes.
Patch [2/5] is main in the series. It introduces new state: ONRQ_MIGRATING
and teaches scheduler to understand it (we need a little changes predominantly
in try_to_wake_up()). This will be used in the following way:
(we are changing task's rq)
raw_spin_lock(&src_rq->lock);
dequeue_task(src_rq, p, 0);
p->on_rq = ONRQ_MIGRATING;
set_task_cpu(p, dst_cpu);
raw_spin_unlock(&src_rq->lock);
raw_spin_lock(&dst_rq->lock);
p->on_rq = ONRQ_QUEUED;
enqueue_task(dst_rq, p, 0);
raw_spin_unlock(&dst_rq->lock);
Patches [3-5/5] remove double locks and use new ONRQ_MIGRATING state.
They allow unlocked using of 3-4 function, which looks safe for me.
The series doesn't add any overhead, and it shouldn't worsen performance,
I think. It improves granularity, and it's possible to imagine situations,
which will be happier without double rq lock.
I tested the reliability for 2-3 week (in sum, with breaks) on my work laptop,
and bugs did not happen. Looks like, it's ready for people's review.
Commentaries are welcome. Thanks!
---
Kirill Tkhai (5):
sched: Wrapper for checking task_struct's .on_rq
sched: Teach scheduler to understand ONRQ_MIGRATING state
sched: Remove double_rq_lock() from __migrate_task()
sched/fair: Remove double_lock_balance() from active_load_balance_cpu_stop()
sched/fair: Remove double_lock_balance() from load_balance()
kernel/sched/core.c | 109 +++++++++++++++++++-------------
kernel/sched/deadline.c | 14 ++--
kernel/sched/fair.c | 156 +++++++++++++++++++++++++++++-----------------
kernel/sched/rt.c | 16 ++---
kernel/sched/sched.h | 8 ++
kernel/sched/stop_task.c | 2 -
6 files changed, 187 insertions(+), 118 deletions(-)
--
Signed-off-by: Kirill Tkhai <[email protected]>