xiaomi-sm8450/android_kernel_xiaomi_sm8450
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

ANDROID: cpuhp: pause: reduce running priority pause/resume

Browse Source 
In order to get a responsive pause operation, the resume/pause_cpus
are usually called within a RT task prio.

However only the lazy-resume and lazy-pause operations need to
happen quickly, the rest can happen at leisure. Running that
portion at high priority keeps the cpu away from important tasks.

Reduce the priority right after the lazy portion, and restore
it before returning, if it was at RT prio.

Bug: 203115740
Fixes: 683010f555 ("ANDROID: cpu/hotplug: add pause/resume_cpus interface")
Change-Id: I1f3394eb9b5fa1876330fef6e25a203da0fde670
Signed-off-by: Stephen Dickey <dickey@codeaurora.org>
This commit is contained in:
Stephen Dickey
2021-10-14 14:32:39 -07:00
parent 2b8b5e76c7
commit db3e01ecf0
1 changed files with 33 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

33
kernel/cpu.c
View File

@@ -1136,11 +1136,37 @@ void __wait_drain_rq(struct cpumask *cpus)
sched_cpu_drain_rq_wait(cpu);
}
/* if rt task, set to cfs and return previous prio */
static int pause_reduce_prio(void)
{
int prev_prio = -1;
if (current->prio < MAX_RT_PRIO) {
struct sched_param param = { .sched_priority = 0 };
prev_prio = current->prio;
sched_setscheduler_nocheck(current, SCHED_NORMAL, &param);
}
return prev_prio;
}
/* if previous prio was set, restore */
static void pause_restore_prio(int prev_prio)
{
if (prev_prio >= 0 && prev_prio < MAX_RT_PRIO) {
struct sched_param param = { .sched_priority = MAX_RT_PRIO-1-prev_prio };
sched_setscheduler_nocheck(current, SCHED_FIFO, &param);
}
}
int pause_cpus(struct cpumask *cpus)
{
int err = 0;
int cpu;
u64 start_time = 0;
int prev_prio;
start_time = sched_clock();
@@ -1195,6 +1221,8 @@ int pause_cpus(struct cpumask *cpus)
goto err_cpu_maps_update;
}
prev_prio = pause_reduce_prio();
/*
* Slow path deactivation:
*
@@ -1238,6 +1266,7 @@ int pause_cpus(struct cpumask *cpus)
err_cpus_write_unlock:
cpus_write_unlock();
pause_restore_prio(prev_prio);
err_cpu_maps_update:
cpu_maps_update_done();
@@ -1252,6 +1281,7 @@ int resume_cpus(struct cpumask *cpus)
unsigned int cpu;
int err = 0;
u64 start_time = 0;
int prev_prio;
start_time = sched_clock();
@@ -1282,6 +1312,8 @@ int resume_cpus(struct cpumask *cpus)
if (err)
goto err_cpu_maps_update;
prev_prio = pause_reduce_prio();
/* Lazy Resume. Build domains immediately instead of scheduling
* a workqueue. This is so that the cpu can pull load when
* sent a load balancing kick.
@@ -1309,6 +1341,7 @@ int resume_cpus(struct cpumask *cpus)
err_cpus_write_unlock:
cpus_write_unlock();
pause_restore_prio(prev_prio);
err_cpu_maps_update:
cpu_maps_update_done();