rcu-tasks: Shorten per-grace-period sleep for RCU Tasks Trace

The various RCU tasks flavors currently wait 100 milliseconds between each
grace period in order to prevent CPU-bound loops and to favor efficiency
over latency.  However, RCU Tasks Trace needs to have a grace-period
latency of roughly 25 milliseconds, which is completely infeasible given
the 100-millisecond per-grace-period sleep.  This commit therefore reduces
this sleep duration to 5 milliseconds (or one jiffy, whichever is longer)
in kernels built with CONFIG_TASKS_TRACE_RCU_READ_MB=y.

Link: https://lore.kernel.org/bpf/CAADnVQK_AiX+S_L_A4CQWT11XyveppBbQSQgH_qWGyzu_E8Yeg@mail.gmail.com/
Reported-by: Alexei Starovoitov <alexei.starovoitov@gmail.com>
Cc: Daniel Borkmann <daniel@iogearbox.net>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: <bpf@vger.kernel.org>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
This commit is contained in:
Paul E. McKenney 2020-09-09 22:05:41 -07:00
parent 574de8766f
commit 4fe192dfbe

View File

@ -28,6 +28,7 @@ typedef void (*postgp_func_t)(struct rcu_tasks *rtp);
* @kthread_ptr: This flavor's grace-period/callback-invocation kthread. * @kthread_ptr: This flavor's grace-period/callback-invocation kthread.
* @gp_func: This flavor's grace-period-wait function. * @gp_func: This flavor's grace-period-wait function.
* @gp_state: Grace period's most recent state transition (debugging). * @gp_state: Grace period's most recent state transition (debugging).
* @gp_sleep: Per-grace-period sleep to prevent CPU-bound looping.
* @init_fract: Initial backoff sleep interval. * @init_fract: Initial backoff sleep interval.
* @gp_jiffies: Time of last @gp_state transition. * @gp_jiffies: Time of last @gp_state transition.
* @gp_start: Most recent grace-period start in jiffies. * @gp_start: Most recent grace-period start in jiffies.
@ -49,6 +50,7 @@ struct rcu_tasks {
struct wait_queue_head cbs_wq; struct wait_queue_head cbs_wq;
raw_spinlock_t cbs_lock; raw_spinlock_t cbs_lock;
int gp_state; int gp_state;
int gp_sleep;
int init_fract; int init_fract;
unsigned long gp_jiffies; unsigned long gp_jiffies;
unsigned long gp_start; unsigned long gp_start;
@ -233,7 +235,7 @@ static int __noreturn rcu_tasks_kthread(void *arg)
cond_resched(); cond_resched();
} }
/* Paranoid sleep to keep this from entering a tight loop */ /* Paranoid sleep to keep this from entering a tight loop */
schedule_timeout_idle(HZ/10); schedule_timeout_idle(rtp->gp_sleep);
set_tasks_gp_state(rtp, RTGS_WAIT_CBS); set_tasks_gp_state(rtp, RTGS_WAIT_CBS);
} }
@ -557,6 +559,7 @@ EXPORT_SYMBOL_GPL(rcu_barrier_tasks);
static int __init rcu_spawn_tasks_kthread(void) static int __init rcu_spawn_tasks_kthread(void)
{ {
rcu_tasks.gp_sleep = HZ / 10;
rcu_tasks.init_fract = 10; rcu_tasks.init_fract = 10;
rcu_tasks.pregp_func = rcu_tasks_pregp_step; rcu_tasks.pregp_func = rcu_tasks_pregp_step;
rcu_tasks.pertask_func = rcu_tasks_pertask; rcu_tasks.pertask_func = rcu_tasks_pertask;
@ -690,6 +693,7 @@ EXPORT_SYMBOL_GPL(rcu_barrier_tasks_rude);
static int __init rcu_spawn_tasks_rude_kthread(void) static int __init rcu_spawn_tasks_rude_kthread(void)
{ {
rcu_tasks_rude.gp_sleep = HZ / 10;
rcu_spawn_tasks_kthread_generic(&rcu_tasks_rude); rcu_spawn_tasks_kthread_generic(&rcu_tasks_rude);
return 0; return 0;
} }
@ -1170,8 +1174,12 @@ EXPORT_SYMBOL_GPL(rcu_barrier_tasks_trace);
static int __init rcu_spawn_tasks_trace_kthread(void) static int __init rcu_spawn_tasks_trace_kthread(void)
{ {
if (IS_ENABLED(CONFIG_TASKS_TRACE_RCU_READ_MB)) { if (IS_ENABLED(CONFIG_TASKS_TRACE_RCU_READ_MB)) {
rcu_tasks_trace.gp_sleep = HZ / 10;
rcu_tasks_trace.init_fract = 10; rcu_tasks_trace.init_fract = 10;
} else { } else {
rcu_tasks_trace.gp_sleep = HZ / 200;
if (rcu_tasks_trace.gp_sleep <= 0)
rcu_tasks_trace.gp_sleep = 1;
rcu_tasks_trace.init_fract = HZ / 5; rcu_tasks_trace.init_fract = HZ / 5;
if (rcu_tasks_trace.init_fract <= 0) if (rcu_tasks_trace.init_fract <= 0)
rcu_tasks_trace.init_fract = 1; rcu_tasks_trace.init_fract = 1;