sched/pelt: Use rq_clock_task() for hw_pressure
PATCH v5 1/1 sched/pelt: Use rq_clock_task() for hw_pressure
commit 97450eb90965 ("sched/pelt: Remove shift of thermal clock") removed the decay_shift for hw_pressure. This commit uses the sched_clock_task() in sched_tick() while it replaces the sched_clock_task() with rq_clock_pelt() in __update_blocked_others(). This could bring inconsistence. One possible scenario I can think of is in ___update_load_sum():
u64 delta = now - sa->last_update_time
'now' could be calculated by rq_clock_pelt() from __update_blocked_others(), and last_update_time was calculated by rq_clock_task() previously from sched_tick(). Usually the former chases after the latter, it cause a very large 'delta' and brings unexpected behavior.
---v4->v5:
Added Vincent's Reviewed-by tag.v3->v4:
Revert to v2 to keep the clock outside of pelt functions. (Vincent Guittot)
Added comments for hw_pressure within __update_blocked_others(). (Qais Yousef)
v2>v3:
call rq_clock_task() inside update_hw_load_avg() to avoid any inconsistence in the future. (Qais Yousef)
Added comments around update_hw_load_avg(). (Qais Yousef)v1->v2:
Added Hongyan's Reviewed-by tag.
Removed the Reported-by/Closes tags because they are not related to this fix.(Hongyan Xia)
--- kernel/sched/fair.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-)
kernel/sched/fair.c +2 -1
@@ -9360,9 +9360,10 @@ static bool __update_blocked_others(struct rq *rq, bool *done) hw_pressure =arch_scale_hw_pressure (cpu_of (rq)); +/* hw_pressure doesn't care about invariance */ decayed =update_rt_rq_load_avg (now, rq, curr_class == &rt_sched_class) |update_dl_rq_load_avg (now, rq, curr_class == &dl_sched_class) | -update_hw_load_avg (now, rq, hw_pressure) | +update_hw_load_avg (rq_clock_task (rq), rq, hw_pressure) |update_irq_load_avg (rq,0 );if (others_have_blocked (rq))
Signature · 1 line
2.25.1
The following commit has been merged into the sched/core branch of tip:
The following commit has been merged into the sched/core branch of tip:
Commit-ID: 84d265281d6cea65353fc24146280e0d86ac50cb Gitweb: https://git.kernel.org/tip/84d265281d6cea65353fc24146280e0d86ac50cb Author: Chen Yu <[email protected]> AuthorDate: Tue, 27 Aug 2024 19:26:07 +08:00 Committer: Peter Zijlstra <[email protected]> CommitterDate: Tue, 10 Sep 2024 09:51:14 +02:00
sched/pelt: Use rq_clock_task() for hw_pressure
commit 97450eb90965 ("sched/pelt: Remove shift of thermal clock") removed the decay_shift for hw_pressure. This commit uses the sched_clock_task() in sched_tick() while it replaces the sched_clock_task() with rq_clock_pelt() in __update_blocked_others(). This could bring inconsistence. One possible scenario I can think of is in ___update_load_sum():
u64 delta = now - sa->last_update_time
'now' could be calculated by rq_clock_pelt() from __update_blocked_others(), and last_update_time was calculated by rq_clock_task() previously from sched_tick(). Usually the former chases after the latter, it cause a very large 'delta' and brings unexpected behavior.
--- kernel/sched/fair.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-)
kernel/sched/fair.c +2 -1
@@ -9719,9 +9719,10 @@ static bool __update_blocked_others(struct rq *rq, bool *done) hw_pressure =arch_scale_hw_pressure (cpu_of (rq)); +/* hw_pressure doesn't care about invariance */ decayed =update_rt_rq_load_avg (now, rq, curr_class == &rt_sched_class) |update_dl_rq_load_avg (now, rq, curr_class == &dl_sched_class) | -update_hw_load_avg (now, rq, hw_pressure) | +update_hw_load_avg (rq_clock_task (rq), rq, hw_pressure) |update_irq_load_avg (rq,0 );if (others_have_blocked (rq))