From 8bdf7a252b6398dbd4beaa7a2ba0904cea0abd04 Mon Sep 17 00:00:00 2001 From: "Paul E. McKenney" Date: Tue, 18 Jun 2013 11:15:21 -0700 Subject: nohz_full: Add testing information to documentation This commit adds information about testing nohz_full, and also emphasizes the fact that you need a multi-CPU system to get any benefit from nohz_full. Signed-off-by: Paul E. McKenney Acked-by: Frederic Weisbecker Reviewed-by: Josh Triplett --- Documentation/timers/NO_HZ.txt | 44 ++++++++++++++++++++++++++++++++---------- 1 file changed, 34 insertions(+), 10 deletions(-) (limited to 'Documentation') diff --git a/Documentation/timers/NO_HZ.txt b/Documentation/timers/NO_HZ.txt index 88697584242..cca122f2512 100644 --- a/Documentation/timers/NO_HZ.txt +++ b/Documentation/timers/NO_HZ.txt @@ -24,8 +24,8 @@ There are three main ways of managing scheduling-clock interrupts workloads, you will normally -not- want this option. These three cases are described in the following three sections, followed -by a third section on RCU-specific considerations and a fourth and final -section listing known issues. +by a third section on RCU-specific considerations, a fourth section +discussing testing, and a fifth and final section listing known issues. NEVER OMIT SCHEDULING-CLOCK TICKS @@ -121,14 +121,15 @@ boot parameter specifies the adaptive-ticks CPUs. For example, "nohz_full=1,6-8" says that CPUs 1, 6, 7, and 8 are to be adaptive-ticks CPUs. Note that you are prohibited from marking all of the CPUs as adaptive-tick CPUs: At least one non-adaptive-tick CPU must remain -online to handle timekeeping tasks in order to ensure that system calls -like gettimeofday() returns accurate values on adaptive-tick CPUs. -(This is not an issue for CONFIG_NO_HZ_IDLE=y because there are no -running user processes to observe slight drifts in clock rate.) -Therefore, the boot CPU is prohibited from entering adaptive-ticks -mode. Specifying a "nohz_full=" mask that includes the boot CPU will -result in a boot-time error message, and the boot CPU will be removed -from the mask. +online to handle timekeeping tasks in order to ensure that system +calls like gettimeofday() returns accurate values on adaptive-tick CPUs. +(This is not an issue for CONFIG_NO_HZ_IDLE=y because there are no running +user processes to observe slight drifts in clock rate.) Therefore, the +boot CPU is prohibited from entering adaptive-ticks mode. Specifying a +"nohz_full=" mask that includes the boot CPU will result in a boot-time +error message, and the boot CPU will be removed from the mask. Note that +this means that your system must have at least two CPUs in order for +CONFIG_NO_HZ_FULL=y to do anything for you. Alternatively, the CONFIG_NO_HZ_FULL_ALL=y Kconfig parameter specifies that all CPUs other than the boot CPU are adaptive-ticks CPUs. This @@ -232,6 +233,29 @@ scheduler will decide where to run them, which might or might not be where you want them to run. +TESTING + +So you enable all the OS-jitter features described in this document, +but do not see any change in your workload's behavior. Is this because +your workload isn't affected that much by OS jitter, or is it because +something else is in the way? This section helps answer this question +by providing a simple OS-jitter test suite, which is available on branch +master of the following git archive: + +git://git.kernel.org/pub/scm/linux/kernel/git/frederic/dynticks-testing.git + +Clone this archive and follow the instructions in the README file. +This test procedure will produce a trace that will allow you to evaluate +whether or not you have succeeded in removing OS jitter from your system. +If this trace shows that you have removed OS jitter as much as is +possible, then you can conclude that your workload is not all that +sensitive to OS jitter. + +Note: this test requires that your system have at least two CPUs. +We do not currently have a good way to remove OS jitter from single-CPU +systems. + + KNOWN ISSUES o Dyntick-idle slows transitions to and from idle slightly. -- cgit v1.2.3-70-g09d2 From 2ec1f2d98752293f4831ce7d7bdbc3fc36bdd114 Mon Sep 17 00:00:00 2001 From: "Paul E. McKenney" Date: Wed, 12 Jun 2013 15:12:21 -0700 Subject: rcu: Increase rcutorture test coverage Currently, rcutorture has separate torture_types to test synchronous, asynchronous, and expedited grace-period primitives. This has two disadvantages: (1) Three times the number of runs to cover the combinations and (2) Little testing of concurrent combinations of the three options. This commit therefore adds a pair of module parameters that control normal and expedited state, with the default being both types, randomly selected, by the fakewriter processes, thus reducing source-code size and increasing test coverage. In addtion, the writer task switches between asynchronous-normal and expedited grace-period primitives driven by the same pair of module parameters. Signed-off-by: Paul E. McKenney Reviewed-by: Josh Triplett --- Documentation/RCU/torture.txt | 10 ++ kernel/rcutorture.c | 226 ++++++++++++------------------------------ 2 files changed, 73 insertions(+), 163 deletions(-) (limited to 'Documentation') diff --git a/Documentation/RCU/torture.txt b/Documentation/RCU/torture.txt index d8a50238739..dac02a6219b 100644 --- a/Documentation/RCU/torture.txt +++ b/Documentation/RCU/torture.txt @@ -42,6 +42,16 @@ fqs_holdoff Holdoff time (in microseconds) between consecutive calls fqs_stutter Wait time (in seconds) between consecutive bursts of calls to force_quiescent_state(). +gp_normal Make the fake writers use normal synchronous grace-period + primitives. + +gp_exp Make the fake writers use expedited synchronous grace-period + primitives. If both gp_normal and gp_exp are set, or + if neither gp_normal nor gp_exp are set, then randomly + choose the primitive so that about 50% are normal and + 50% expedited. By default, neither are set, which + gives best overall test coverage. + irqreader Says to invoke RCU readers from irq level. This is currently done via timers. Defaults to "1" for variants of RCU that permit this. (Or, more accurately, variants of RCU that do diff --git a/kernel/rcutorture.c b/kernel/rcutorture.c index c898f14a5b7..ddef6187187 100644 --- a/kernel/rcutorture.c +++ b/kernel/rcutorture.c @@ -65,6 +65,8 @@ static int irqreader = 1; /* RCU readers from irq (timers). */ static int fqs_duration; /* Duration of bursts (us), 0 to disable. */ static int fqs_holdoff; /* Hold time within burst (us). */ static int fqs_stutter = 3; /* Wait time between bursts (s). */ +static bool gp_exp; /* Use expedited GP wait primitives. */ +static bool gp_normal; /* Use normal GP wait primitives. */ static int n_barrier_cbs; /* Number of callbacks to test RCU barriers. */ static int object_debug; /* Test object-debug double call_rcu()?. */ static int onoff_interval; /* Wait time between CPU hotplugs, 0=disable. */ @@ -99,6 +101,10 @@ module_param(fqs_holdoff, int, 0444); MODULE_PARM_DESC(fqs_holdoff, "Holdoff time within fqs bursts (us)"); module_param(fqs_stutter, int, 0444); MODULE_PARM_DESC(fqs_stutter, "Wait time between fqs bursts (s)"); +module_param(gp_normal, bool, 0444); +MODULE_PARM_DESC(gp_normal, "Use normal (non-expedited) GP wait primitives"); +module_param(gp_exp, bool, 0444); +MODULE_PARM_DESC(gp_exp, "Use expedited GP wait primitives"); module_param(n_barrier_cbs, int, 0444); MODULE_PARM_DESC(n_barrier_cbs, "# of callbacks/kthreads for barrier testing"); module_param(object_debug, int, 0444); @@ -363,6 +369,7 @@ struct rcu_torture_ops { int (*completed)(void); void (*deferred_free)(struct rcu_torture *p); void (*sync)(void); + void (*exp_sync)(void); void (*call)(struct rcu_head *head, void (*func)(struct rcu_head *rcu)); void (*cb_barrier)(void); void (*fqs)(void); @@ -446,81 +453,27 @@ static void rcu_torture_deferred_free(struct rcu_torture *p) call_rcu(&p->rtort_rcu, rcu_torture_cb); } -static struct rcu_torture_ops rcu_ops = { - .init = NULL, - .readlock = rcu_torture_read_lock, - .read_delay = rcu_read_delay, - .readunlock = rcu_torture_read_unlock, - .completed = rcu_torture_completed, - .deferred_free = rcu_torture_deferred_free, - .sync = synchronize_rcu, - .call = call_rcu, - .cb_barrier = rcu_barrier, - .fqs = rcu_force_quiescent_state, - .stats = NULL, - .irq_capable = 1, - .can_boost = rcu_can_boost(), - .name = "rcu" -}; - -static void rcu_sync_torture_deferred_free(struct rcu_torture *p) -{ - int i; - struct rcu_torture *rp; - struct rcu_torture *rp1; - - cur_ops->sync(); - list_add(&p->rtort_free, &rcu_torture_removed); - list_for_each_entry_safe(rp, rp1, &rcu_torture_removed, rtort_free) { - i = rp->rtort_pipe_count; - if (i > RCU_TORTURE_PIPE_LEN) - i = RCU_TORTURE_PIPE_LEN; - atomic_inc(&rcu_torture_wcount[i]); - if (++rp->rtort_pipe_count >= RCU_TORTURE_PIPE_LEN) { - rp->rtort_mbtest = 0; - list_del(&rp->rtort_free); - rcu_torture_free(rp); - } - } -} - static void rcu_sync_torture_init(void) { INIT_LIST_HEAD(&rcu_torture_removed); } -static struct rcu_torture_ops rcu_sync_ops = { +static struct rcu_torture_ops rcu_ops = { .init = rcu_sync_torture_init, .readlock = rcu_torture_read_lock, .read_delay = rcu_read_delay, .readunlock = rcu_torture_read_unlock, .completed = rcu_torture_completed, - .deferred_free = rcu_sync_torture_deferred_free, + .deferred_free = rcu_torture_deferred_free, .sync = synchronize_rcu, - .call = NULL, - .cb_barrier = NULL, - .fqs = rcu_force_quiescent_state, - .stats = NULL, - .irq_capable = 1, - .can_boost = rcu_can_boost(), - .name = "rcu_sync" -}; - -static struct rcu_torture_ops rcu_expedited_ops = { - .init = rcu_sync_torture_init, - .readlock = rcu_torture_read_lock, - .read_delay = rcu_read_delay, /* just reuse rcu's version. */ - .readunlock = rcu_torture_read_unlock, - .completed = rcu_no_completed, - .deferred_free = rcu_sync_torture_deferred_free, - .sync = synchronize_rcu_expedited, - .call = NULL, - .cb_barrier = NULL, + .exp_sync = synchronize_rcu_expedited, + .call = call_rcu, + .cb_barrier = rcu_barrier, .fqs = rcu_force_quiescent_state, .stats = NULL, .irq_capable = 1, .can_boost = rcu_can_boost(), - .name = "rcu_expedited" + .name = "rcu" }; /* @@ -549,13 +502,14 @@ static void rcu_bh_torture_deferred_free(struct rcu_torture *p) } static struct rcu_torture_ops rcu_bh_ops = { - .init = NULL, + .init = rcu_sync_torture_init, .readlock = rcu_bh_torture_read_lock, .read_delay = rcu_read_delay, /* just reuse rcu's version. */ .readunlock = rcu_bh_torture_read_unlock, .completed = rcu_bh_torture_completed, .deferred_free = rcu_bh_torture_deferred_free, .sync = synchronize_rcu_bh, + .exp_sync = synchronize_rcu_bh_expedited, .call = call_rcu_bh, .cb_barrier = rcu_barrier_bh, .fqs = rcu_bh_force_quiescent_state, @@ -564,38 +518,6 @@ static struct rcu_torture_ops rcu_bh_ops = { .name = "rcu_bh" }; -static struct rcu_torture_ops rcu_bh_sync_ops = { - .init = rcu_sync_torture_init, - .readlock = rcu_bh_torture_read_lock, - .read_delay = rcu_read_delay, /* just reuse rcu's version. */ - .readunlock = rcu_bh_torture_read_unlock, - .completed = rcu_bh_torture_completed, - .deferred_free = rcu_sync_torture_deferred_free, - .sync = synchronize_rcu_bh, - .call = NULL, - .cb_barrier = NULL, - .fqs = rcu_bh_force_quiescent_state, - .stats = NULL, - .irq_capable = 1, - .name = "rcu_bh_sync" -}; - -static struct rcu_torture_ops rcu_bh_expedited_ops = { - .init = rcu_sync_torture_init, - .readlock = rcu_bh_torture_read_lock, - .read_delay = rcu_read_delay, /* just reuse rcu's version. */ - .readunlock = rcu_bh_torture_read_unlock, - .completed = rcu_bh_torture_completed, - .deferred_free = rcu_sync_torture_deferred_free, - .sync = synchronize_rcu_bh_expedited, - .call = NULL, - .cb_barrier = NULL, - .fqs = rcu_bh_force_quiescent_state, - .stats = NULL, - .irq_capable = 1, - .name = "rcu_bh_expedited" -}; - /* * Definitions for srcu torture testing. */ @@ -670,6 +592,11 @@ static int srcu_torture_stats(char *page) return cnt; } +static void srcu_torture_synchronize_expedited(void) +{ + synchronize_srcu_expedited(&srcu_ctl); +} + static struct rcu_torture_ops srcu_ops = { .init = rcu_sync_torture_init, .readlock = srcu_torture_read_lock, @@ -678,45 +605,13 @@ static struct rcu_torture_ops srcu_ops = { .completed = srcu_torture_completed, .deferred_free = srcu_torture_deferred_free, .sync = srcu_torture_synchronize, + .exp_sync = srcu_torture_synchronize_expedited, .call = srcu_torture_call, .cb_barrier = srcu_torture_barrier, .stats = srcu_torture_stats, .name = "srcu" }; -static struct rcu_torture_ops srcu_sync_ops = { - .init = rcu_sync_torture_init, - .readlock = srcu_torture_read_lock, - .read_delay = srcu_read_delay, - .readunlock = srcu_torture_read_unlock, - .completed = srcu_torture_completed, - .deferred_free = rcu_sync_torture_deferred_free, - .sync = srcu_torture_synchronize, - .call = NULL, - .cb_barrier = NULL, - .stats = srcu_torture_stats, - .name = "srcu_sync" -}; - -static void srcu_torture_synchronize_expedited(void) -{ - synchronize_srcu_expedited(&srcu_ctl); -} - -static struct rcu_torture_ops srcu_expedited_ops = { - .init = rcu_sync_torture_init, - .readlock = srcu_torture_read_lock, - .read_delay = srcu_read_delay, - .readunlock = srcu_torture_read_unlock, - .completed = srcu_torture_completed, - .deferred_free = rcu_sync_torture_deferred_free, - .sync = srcu_torture_synchronize_expedited, - .call = NULL, - .cb_barrier = NULL, - .stats = srcu_torture_stats, - .name = "srcu_expedited" -}; - /* * Definitions for sched torture testing. */ @@ -745,6 +640,8 @@ static struct rcu_torture_ops sched_ops = { .completed = rcu_no_completed, .deferred_free = rcu_sched_torture_deferred_free, .sync = synchronize_sched, + .exp_sync = synchronize_sched_expedited, + .call = call_rcu_sched, .cb_barrier = rcu_barrier_sched, .fqs = rcu_sched_force_quiescent_state, .stats = NULL, @@ -752,35 +649,6 @@ static struct rcu_torture_ops sched_ops = { .name = "sched" }; -static struct rcu_torture_ops sched_sync_ops = { - .init = rcu_sync_torture_init, - .readlock = sched_torture_read_lock, - .read_delay = rcu_read_delay, /* just reuse rcu's version. */ - .readunlock = sched_torture_read_unlock, - .completed = rcu_no_completed, - .deferred_free = rcu_sync_torture_deferred_free, - .sync = synchronize_sched, - .cb_barrier = NULL, - .fqs = rcu_sched_force_quiescent_state, - .stats = NULL, - .name = "sched_sync" -}; - -static struct rcu_torture_ops sched_expedited_ops = { - .init = rcu_sync_torture_init, - .readlock = sched_torture_read_lock, - .read_delay = rcu_read_delay, /* just reuse rcu's version. */ - .readunlock = sched_torture_read_unlock, - .completed = rcu_no_completed, - .deferred_free = rcu_sync_torture_deferred_free, - .sync = synchronize_sched_expedited, - .cb_barrier = NULL, - .fqs = rcu_sched_force_quiescent_state, - .stats = NULL, - .irq_capable = 1, - .name = "sched_expedited" -}; - /* * RCU torture priority-boost testing. Runs one real-time thread per * CPU for moderate bursts, repeatedly registering RCU callbacks and @@ -930,9 +798,11 @@ rcu_torture_fqs(void *arg) static int rcu_torture_writer(void *arg) { + bool exp; int i; long oldbatch = rcu_batches_completed(); struct rcu_torture *rp; + struct rcu_torture *rp1; struct rcu_torture *old_rp; static DEFINE_RCU_RANDOM(rand); @@ -957,7 +827,31 @@ rcu_torture_writer(void *arg) i = RCU_TORTURE_PIPE_LEN; atomic_inc(&rcu_torture_wcount[i]); old_rp->rtort_pipe_count++; - cur_ops->deferred_free(old_rp); + if (gp_normal == gp_exp) + exp = !!(rcu_random(&rand) & 0x80); + else + exp = gp_exp; + if (!exp) { + cur_ops->deferred_free(old_rp); + } else { + cur_ops->exp_sync(); + list_add(&old_rp->rtort_free, + &rcu_torture_removed); + list_for_each_entry_safe(rp, rp1, + &rcu_torture_removed, + rtort_free) { + i = rp->rtort_pipe_count; + if (i > RCU_TORTURE_PIPE_LEN) + i = RCU_TORTURE_PIPE_LEN; + atomic_inc(&rcu_torture_wcount[i]); + if (++rp->rtort_pipe_count >= + RCU_TORTURE_PIPE_LEN) { + rp->rtort_mbtest = 0; + list_del(&rp->rtort_free); + rcu_torture_free(rp); + } + } + } } rcutorture_record_progress(++rcu_torture_current_version); oldbatch = cur_ops->completed(); @@ -986,10 +880,18 @@ rcu_torture_fakewriter(void *arg) schedule_timeout_uninterruptible(1 + rcu_random(&rand)%10); udelay(rcu_random(&rand) & 0x3ff); if (cur_ops->cb_barrier != NULL && - rcu_random(&rand) % (nfakewriters * 8) == 0) + rcu_random(&rand) % (nfakewriters * 8) == 0) { cur_ops->cb_barrier(); - else + } else if (gp_normal == gp_exp) { + if (rcu_random(&rand) & 0x80) + cur_ops->sync(); + else + cur_ops->exp_sync(); + } else if (gp_normal) { cur_ops->sync(); + } else { + cur_ops->exp_sync(); + } rcu_stutter_wait("rcu_torture_fakewriter"); } while (!kthread_should_stop() && fullstop == FULLSTOP_DONTSTOP); @@ -2000,11 +1902,9 @@ rcu_torture_init(void) int cpu; int firsterr = 0; int retval; - static struct rcu_torture_ops *torture_ops[] = - { &rcu_ops, &rcu_sync_ops, &rcu_expedited_ops, - &rcu_bh_ops, &rcu_bh_sync_ops, &rcu_bh_expedited_ops, - &srcu_ops, &srcu_sync_ops, &srcu_expedited_ops, - &sched_ops, &sched_sync_ops, &sched_expedited_ops, }; + static struct rcu_torture_ops *torture_ops[] = { + &rcu_ops, &rcu_bh_ops, &srcu_ops, &sched_ops, + }; mutex_lock(&fullstop_mutex); -- cgit v1.2.3-70-g09d2