diff options
| author | Dan Williams <dan.j.williams@intel.com> | 2019-03-11 12:37:55 -0700 |
|---|---|---|
| committer | Dan Williams <dan.j.williams@intel.com> | 2019-03-11 12:37:55 -0700 |
| commit | 4083014e32699af04a8e6eaa4855b08dba36a47a (patch) | |
| tree | fa37f9f9691fe64ca8a3c0cdc0315dc12462e6e4 /kernel/sched/psi.c | |
| parent | 6fd96ff557963de8e62842a0dc360a6e3610d2bb (diff) | |
| parent | 78153dd45e7e0596ba32b15d02bda08e1513111e (diff) | |
| download | linux-4083014e32699af04a8e6eaa4855b08dba36a47a.tar.bz2 | |
Merge branch 'for-5.1/nfit/ars' into libnvdimm-for-next
Merge several updates to the ARS implementation. Highlights include:
* Support retrieval of short-ARS results if the ARS state is "requires
continuation", and even if the "no_init_ars" module parameter is
specified.
* Allow busy-polling of the kernel ARS state by allowing root to reset
the exponential back-off timer.
* Filter potentially stale ARS results by tracking query-ARS relative to
the previous start-ARS.
Diffstat (limited to 'kernel/sched/psi.c')
| -rw-r--r-- | kernel/sched/psi.c | 21 |
1 files changed, 17 insertions, 4 deletions
diff --git a/kernel/sched/psi.c b/kernel/sched/psi.c index fe24de3fbc93..c3484785b179 100644 --- a/kernel/sched/psi.c +++ b/kernel/sched/psi.c @@ -124,6 +124,7 @@ * sampling of the aggregate task states would be. */ +#include "../workqueue_internal.h" #include <linux/sched/loadavg.h> #include <linux/seq_file.h> #include <linux/proc_fs.h> @@ -480,9 +481,6 @@ static void psi_group_change(struct psi_group *group, int cpu, groupc->tasks[t]++; write_seqcount_end(&groupc->seq); - - if (!delayed_work_pending(&group->clock_work)) - schedule_delayed_work(&group->clock_work, PSI_FREQ); } static struct psi_group *iterate_groups(struct task_struct *task, void **iter) @@ -513,6 +511,7 @@ void psi_task_change(struct task_struct *task, int clear, int set) { int cpu = task_cpu(task); struct psi_group *group; + bool wake_clock = true; void *iter = NULL; if (!task->pid) @@ -530,8 +529,22 @@ void psi_task_change(struct task_struct *task, int clear, int set) task->psi_flags &= ~clear; task->psi_flags |= set; - while ((group = iterate_groups(task, &iter))) + /* + * Periodic aggregation shuts off if there is a period of no + * task changes, so we wake it back up if necessary. However, + * don't do this if the task change is the aggregation worker + * itself going to sleep, or we'll ping-pong forever. + */ + if (unlikely((clear & TSK_RUNNING) && + (task->flags & PF_WQ_WORKER) && + wq_worker_last_func(task) == psi_update_work)) + wake_clock = false; + + while ((group = iterate_groups(task, &iter))) { psi_group_change(group, cpu, clear, set); + if (wake_clock && !delayed_work_pending(&group->clock_work)) + schedule_delayed_work(&group->clock_work, PSI_FREQ); + } } void psi_memstall_tick(struct task_struct *task, int cpu) |