timers: Exclude isolated cpus from timer migration
Subsystems (7)
v13 cover timers: Exclude isolated cpus from timer migration
The timer migration mechanism allows active CPUs to pull timers from idle ones to improve the overall idle time. This is however undesired when CPU intensive workloads run on isolated cores, as the algorithm would move the timers from housekeeping to isolated cores, negatively affecting the isolation.
Exclude isolated cores from the timer migration algorithm, extend the concept of unavailable cores, currently used for offline ones, to isolated ones:
* A core is unavailable if isolated or offline;
* A core is available if non isolated and online;
A core is considered unavailable as isolated if it belongs to:
* the isolcpus (domain) list
* an isolated cpuset
Except if it is:
* in the nohz_full list (already idle for the hierarchy)
* the nohz timekeeper core (must be available to handle global timers)
CPUs are added to the hierarchy during late boot, excluding isolated ones, the hierarchy is also adapted when the cpuset isolation changes.
Due to how the timer migration algorithm works, any CPU part of the hierarchy can have their global timers pulled by remote CPUs and have to pull remote timers, only skipping pulling remote timers would break the logic.
For this reason, prevent isolated CPUs from pulling remote global timers, but also the other way around: any global timer started on an isolated CPU will run there. This does not break the concept of isolation (global timers don't come from outside the CPU) and, if considered inappropriate, can usually be mitigated with other isolation techniques (e.g. IRQ pinning).
This effect was noticed on a 128 cores machine running oslat on the isolated cores (1-31,33-63,65-95,97-127). The tool monopolises CPUs, and the CPU with lowest count in a timer migration hierarchy (here 1 and 65) appears as always active and continuously pulls global timers, from the housekeeping CPUs. This ends up moving driver work (e.g. delayed work) to isolated CPUs and causes latency spikes:
before the change:
# oslat -c 1-31,33-63,65-95,97-127 -D 62s ... Maximum: 1203 10 3 4 ... 5 (us)
after the change:
# oslat -c 1-31,33-63,65-95,97-127 -D 62s ... Maximum: 10 4 3 4 3 ... 5 (us)
The same behaviour was observed on a machine with as few as 20 cores / 40 threads with isocpus set to: 1-9,11-39 with rtla-osnoise-top.
The first 5 patches are preparatory work to change the concept of online/offline to available/unavailable, keep track of those in a separate cpumask cleanup the setting/clearing functions and change a function name in cpuset code.
Patch 6 and 7 adapt isolation and cpuset to prevent domain isolated and nohz_full from covering all CPUs not leaving any housekeeping one. This can lead to problems with the changes introduced in this series because no CPU would remain to handle global timers.
Patch 9 extends the unavailable status to domain isolated CPUs, which is the main contribution of the series.
This series is equivalent to v13 but rebased on v6.18-rc2.
Changes since v12:
* Pick and adapt patch by Yury Norov to initialise cpumasks
* Reorganise accesses to tmigr_available_cpumask to avoid races
Changes since v11:
* Rename isolcpus_nohz_conflict() to isolated_cpus_can_update()
* Move tick_nohz_cpu_hotpluggable() check to tmigr_is_isolated()
* Use workqueues in tmigr_isolated_exclude_cpumask() to avoid sleeping while atomic
* Add cpumask initialiser to safely use cpumask cleanup helpers
Changes since v10:
* Simplify housekeeping conflict condition
* Reword commit (Frederic Weisbecker)
Changes since v9:
* Fix total housekeeping enforcement to focus only on nohz and domain
* Avoid out of bound access in the housekeeping array if no flag is set
* Consider isolated_cpus while checking for nohz conflicts in cpuset
* Improve comment about why nohz CPUs are not excluded by tmigr
Changes since v8 [1]:
* Postpone hotplug registration to late initcall (Frederic Weisbecker)
* Move main activation logic in _tmigr_set_cpu_available() and call it after checking for isolation on hotplug and cpusets changes
* Call _tmigr_set_cpu_available directly to force enable tick CPU if required (this saves checking for that on every hotplug change).
Changes since v7:
* Move tmigr_available_cpumask out of tmc lock and specify conditions.
* Initialise tmigr isolation despite the state of isolcpus.
* Move tick CPU check to condition to run SMP call.
* Fix descriptions.
Changes since v6 [2]:
* Prevent isolation checks from running during early boot
* Prevent double (de)activation while setting cpus (un)available
* Use synchronous smp calls from the isolation path
* General cleanup
Changes since v5:
* Remove fallback if no housekeeping is left by isolcpus and nohz_full
* Adjust condition not to activate CPUs in the migration hierarchy
* Always force the nohz tick CPU active in the hierarchy
Changes since v4 [3]:
* use on_each_cpu_mask() with changes on isolated CPUs to avoid races
* keep nohz_full CPUs included in the timer migration hierarchy
* prevent domain isolated and nohz_full to cover all CPUs
Changes since v3:
* add parameter to function documentation
* split into multiple straightforward patches
Changes since v2:
* improve comments about handling CPUs isolated at boot
* minor cleanup
Changes since v1 [4]:
* split into smaller patches
* use available mask instead of unavailable
* simplification and cleanup
[1] - https://lore.kernel.org/lkml/[email protected]
[2] - https://lore.kernel.org/lkml/[email protected]
[3] - https://lore.kernel.org/lkml/[email protected]
[4] - https://lore.kernel.org/lkml/[email protected]
Frederic Weisbecker (1):
timers/migration: Postpone online/offline callbacks registration to
late initcall
Gabriele Monaco (7):
timers: Rename tmigr 'online' bit to 'available'
timers: Add the available mask in timer migration
timers: Use scoped_guard when setting/clearing the tmigr available
flag
cgroup/cpuset: Rename update_unbound_workqueue_cpumask() to
update_exclusion_cpumasks()
sched/isolation: Force housekeeping if isolcpus and nohz_full don't
leave any
cgroup/cpuset: Fail if isolated and nohz_full don't leave any
housekeeping
timers: Exclude isolated cpus from timer migration
Yury Norov (1):
cpumask: Add initialiser to use cleanup helpers
include/linux/cpumask.h | 2 + include/linux/timer.h | 9 ++ include/trace/events/timer_migration.h | 4 +- kernel/cgroup/cpuset.c | 78 +++++++++- kernel/sched/isolation.c | 23 +++ kernel/time/timer_migration.c | 205 ++++++++++++++++++++----- kernel/time/timer_migration.h | 2 +- 7 files changed, 278 insertions(+), 45 deletions(-)
base-commit: 211ddde0823f1442e4ad052a2f30f050145ccada
Signature · 1 line
2.51.0
PATCH RESEND v13 1/9 timers/migration: Postpone online/offline callbacks registration to late initcall
From: Frederic Weisbecker <[email protected]>
During the early boot process, the default clocksource used for timekeeping is the jiffies. Better clocksources can only be selected once clocksource_done_booting() is called as an fs initcall.
NOHZ can only be enabled after that stage, making global timer migration irrelevant up to that point.
The tree remains inactive before NOHZ is enabled anyway. Therefore it makes sense to enable each CPUs to the tree only once that is setup.
Make the CPUs available to the tree on late initcall, after the right clocksource had a chance to be selected. This will also simplify the handling of domain isolated CPUs on further patches.
--- kernel/time/timer_migration.c | 24 +++++++++++++----------- 1 file changed, 13 insertions(+), 11 deletions(-)
kernel/time/timer_migration.c +13 -11
@@ -1481,6 +1481,16 @@ static int tmigr_cpu_online(unsigned int cpu)return 0 ; } +/* +* NOHZ can only be enabled after clocksource_done_booting(). Don't +* bother trashing the cache in the tree before. +*/ +static int __init tmigr_late_init (void ) +{ +return cpuhp_setup_state (CPUHP_AP_TMIGR_ONLINE ,"tmigr:online" , + tmigr_cpu_online, tmigr_cpu_offline); +} +static void tmigr_init_group (struct tmigr_group *group,unsigned int lvl,int node) {
@@ -1843,18 +1853,10 @@ static int __init tmigr_init(void) ret =cpuhp_setup_state (CPUHP_TMIGR_PREPARE ,"tmigr:prepare" , tmigr_cpu_prepare,NULL ); -if (ret) - goto err; - - ret =cpuhp_setup_state (CPUHP_AP_TMIGR_ONLINE ,"tmigr:online" , - tmigr_cpu_online, tmigr_cpu_offline); -if (ret) - goto err; - -return 0 ; - err: -pr_err ("Timer migration setup failed\n" ); +if (ret) +pr_err ("Timer migration setup failed\n" );return ret; }early_initcall (tmigr_init); +late_initcall (tmigr_late_init);
Signature · 1 line
2.51.0
Frederic Weisbecker
maintainerRe: [RESEND PATCH v13 1/9] timers/migration: Postpone online/offline callbacks registration to late initcall
Quoted 29 lines · Gabriele Monaco
From: Frederic Weisbecker <[email protected]>
During the early boot process, the default clocksource used for timekeeping is the jiffies. Better clocksources can only be selected once
clocksource_done_booting()is called as an fs initcall.NOHZ can only be enabled after that stage, making global timer migration irrelevant up to that point.
The tree remains inactive before NOHZ is enabled anyway. Therefore it makes sense to enable each CPUs to the tree only once that is setup.
Make the CPUs available to the tree on late initcall, after the right clocksource had a chance to be selected. This will also simplify the handling of domain isolated CPUs on further patches.
Signed-off-by Frederic Weisbecker <[email protected]>Signed-off-by Gabriele Monaco <[email protected]>--- kernel/time/timer_migration.c | 24 +++++++++++++----------- 1 file changed, 13 insertions(+), 11 deletions(-)kernel/time/timer_migration.c +0
-0@@ -1481,6 +1481,16 @@ static int tmigr_cpu_online(unsigned int cpu)return 0 ; }
kernel/time/timer_migration.c static int tmigr_cpu_online(unsigned int cpu) · line 1492
Quoted 10 lines
kernel/time/timer_migration.c +9
-0+/* +* NOHZ can only be enabled after clocksource_done_booting(). Don't +* bother trashing the cache in the tree before. +*/ +static int __init tmigr_late_init (void ) +{ +return cpuhp_setup_state (CPUHP_AP_TMIGR_ONLINE ,"tmigr:online" , + tmigr_cpu_online, tmigr_cpu_offline); +}
I just worked on a fix in the timer migration code and this made me realize my suggestion was plain wrong.
The CPU doing the prepare work for the target must be online because we must guarantee that the old top level is active since we unconditionally propagate its active state to the new top.
And if we do that late call to tmigr_cpu_online (after most CPUs have booted) then we break that guarantee.
So I fear we can't do that and we must go back to your previous idea which consisted in sending IPIs to apply isolation on late stage.
Sorry about the late brain :-s
Signature · 2 lines
Frederic Weisbecker SUSE Labs
PATCH RESEND v13 2/9 timers: Rename tmigr 'online' bit to 'available'
In reply to Gabriele MonacoThe timer migration hierarchy excludes offline CPUs via the tmigr_is_not_available function, which is essentially checking the online bit for the CPU.
Rename the online bit to available and all references in function names and tracepoint to generalise the concept of available CPUs.
--- include/trace/events/timer_migration.h | 4 ++-- kernel/time/timer_migration.c | 22 +++++++++++----------- kernel/time/timer_migration.h | 2 +- 3 files changed, 14 insertions(+), 14 deletions(-)
include/trace/events/timer_migration.h +2 -2
@@ -173,14 +173,14 @@ DEFINE_EVENT(tmigr_cpugroup, tmigr_cpu_active,TP_ARGS (tmc) ); -DEFINE_EVENT (tmigr_cpugroup, tmigr_cpu_online, +DEFINE_EVENT (tmigr_cpugroup, tmigr_cpu_available, TP_PROTO(struct tmigr_cpu *tmc),TP_ARGS (tmc) ); -DEFINE_EVENT(tmigr_cpugroup, tmigr_cpu_offline, +DEFINE_EVENT(tmigr_cpugroup, tmigr_cpu_unavailable, TP_PROTO(struct tmigr_cpu *tmc),
kernel/time/timer_migration.c +11 -11
@@ -427,7 +427,7 @@ static DEFINE_PER_CPU(struct tmigr_cpu, tmigr_cpu);static inline bool tmigr_is_not_available (struct tmigr_cpu *tmc) { -return !(tmc->tmgroup && tmc->online); +return !(tmc->tmgroup && tmc->available); }/*
@@ -926,7 +926,7 @@ static void tmigr_handle_remote_cpu(unsigned int cpu, u64 now,* updated the event takes care when hierarchy is completely * idle. Otherwise the migrator does it as the event is enqueued. */ -if (!tmc->online || tmc->remote || tmc->cpuevt.ignore || +if (!tmc->available || tmc->remote || tmc->cpuevt.ignore || now < tmc->cpuevt.nextevt.expires) {raw_spin_unlock_irq (&tmc->lock);return ;
@@ -973,7 +973,7 @@ static void tmigr_handle_remote_cpu(unsigned int cpu, u64 now,* (See also section "Required event and timerqueue update after a * remote expiry" in the documentation at the top) */ -if (!tmc->online || !tmc->idle) { +if (!tmc->available || !tmc->idle) {timer_unlock_remote_bases (cpu); goto unlock; }
@@ -1432,19 +1432,19 @@ static long tmigr_trigger_active(void *unused) {struct tmigr_cpu *tmc =this_cpu_ptr (&tmigr_cpu); -WARN_ON_ONCE (!tmc->online || tmc->idle); +WARN_ON_ONCE (!tmc->available || tmc->idle);return 0 ; } -static int tmigr_cpu_offline (unsigned int cpu) +static int tmigr_clear_cpu_available (unsigned int cpu) {struct tmigr_cpu *tmc =this_cpu_ptr (&tmigr_cpu);int migrator;u64 firstexp;raw_spin_lock_irq (&tmc->lock); - tmc->online = false; + tmc->available = false;WRITE_ONCE (tmc->wakeup,KTIME_MAX );/*
@@ -1452,7 +1452,7 @@ static int tmigr_cpu_offline(unsigned int cpu)* offline; Therefore nextevt value is set to KTIME_MAX */ firstexp =__tmigr_cpu_deactivate (tmc,KTIME_MAX ); -trace_tmigr_cpu_offline (tmc); +trace_tmigr_cpu_unavailable (tmc);raw_spin_unlock_irq (&tmc->lock);if (firstexp !=KTIME_MAX ) {
@@ -1463,7 +1463,7 @@ static int tmigr_cpu_offline(unsigned int cpu)return 0 ; } -static int tmigr_cpu_online (unsigned int cpu) +static int tmigr_set_cpu_available (unsigned int cpu) {struct tmigr_cpu *tmc =this_cpu_ptr (&tmigr_cpu);
@@ -1472,11 +1472,11 @@ static int tmigr_cpu_online(unsigned int cpu)return -EINVAL ;raw_spin_lock_irq (&tmc->lock); -trace_tmigr_cpu_online (tmc); +trace_tmigr_cpu_available (tmc); tmc->idle =timer_base_is_idle ();if (!tmc->idle)__tmigr_cpu_activate (tmc); - tmc->online = true; + tmc->available = true;raw_spin_unlock_irq (&tmc->lock);return 0 ; }
@@ -1488,7 +1488,7 @@ static int tmigr_cpu_online(unsigned int cpu)static int __init tmigr_late_init (void ) {return cpuhp_setup_state (CPUHP_AP_TMIGR_ONLINE ,"tmigr:online" , - tmigr_cpu_online, tmigr_cpu_offline); + tmigr_set_cpu_available, tmigr_clear_cpu_available); } static void tmigr_init_group(struct tmigr_group *group, unsigned int lvl,
kernel/time/timer_migration.h +1 -1
@@ -97,7 +97,7 @@ struct tmigr_group {*/ struct tmigr_cpu {raw_spinlock_t lock; -bool online; +bool available;bool idle;bool remote;struct tmigr_group *tmgroup;
Signature · 1 line
2.51.0
PATCH RESEND v13 3/9 timers: Add the available mask in timer migration
In reply to Gabriele MonacoKeep track of the CPUs available for timer migration in a cpumask. This prepares the ground to generalise the concept of unavailable CPUs.
--- kernel/time/timer_migration.c | 15 ++++++++++++++- 1 file changed, 14 insertions(+), 1 deletion(-)
kernel/time/timer_migration.c +14 -1
@@ -422,6 +422,12 @@ static unsigned int tmigr_crossnode_level __read_mostly; static DEFINE_PER_CPU(struct tmigr_cpu, tmigr_cpu); +/* +* CPUs available for timer migration. +* Protected by cpuset_mutex (with cpus_read_lock held) or cpus_write_lock. +*/ +static cpumask_var_t tmigr_available_cpumask; +#define TMIGR_NONE 0xFF#define BIT_CNT 8
@@ -1443,6 +1449,7 @@ static int tmigr_clear_cpu_available(unsigned int cpu)int migrator;u64 firstexp; +cpumask_clear_cpu (cpu, tmigr_available_cpumask);raw_spin_lock_irq (&tmc->lock); tmc->available = false;WRITE_ONCE (tmc->wakeup,KTIME_MAX );
@@ -1456,7 +1463,7 @@ static int tmigr_clear_cpu_available(unsigned int cpu)raw_spin_unlock_irq (&tmc->lock);if (firstexp !=KTIME_MAX ) { - migrator =cpumask_any_but (cpu_online_mask, cpu); + migrator =cpumask_any (tmigr_available_cpumask);work_on_cpu (migrator, tmigr_trigger_active,NULL ); }
@@ -1471,6 +1478,7 @@ static int tmigr_set_cpu_available(unsigned int cpu)if (WARN_ON_ONCE (!tmc->tmgroup))return -EINVAL ; +cpumask_set_cpu (cpu, tmigr_available_cpumask);raw_spin_lock_irq (&tmc->lock);trace_tmigr_cpu_available (tmc); tmc->idle =timer_base_is_idle ();
@@ -1808,6 +1816,11 @@ static int __init tmigr_init(void)if (ncpus ==1 )return 0 ; +if (!zalloc_cpumask_var (&tmigr_available_cpumask,GFP_KERNEL )) { + ret = -ENOMEM ; + goto err; + } +/* * Calculate the required hierarchy levels. Unfortunately there is no * reliable information available, unless all possible CPUs have been
Signature · 1 line
2.51.0
PATCH RESEND v13 4/9 timers: Use scoped_guard when setting/clearing the tmigr available flag
In reply to Gabriele MonacoCleanup tmigr_clear_cpu_available() and tmigr_set_cpu_available() to prepare for easier checks on the available flag.
--- kernel/time/timer_migration.c | 34 +++++++++++++++++----------------- 1 file changed, 17 insertions(+), 17 deletions(-)
kernel/time/timer_migration.c +17 -17
@@ -1450,17 +1450,17 @@ static int tmigr_clear_cpu_available(unsigned int cpu)u64 firstexp;cpumask_clear_cpu (cpu, tmigr_available_cpumask); -raw_spin_lock_irq (&tmc->lock); - tmc->available = false; -WRITE_ONCE (tmc->wakeup,KTIME_MAX ); + scoped_guard(raw_spinlock_irq, &tmc->lock) { + tmc->available = false; +WRITE_ONCE (tmc->wakeup,KTIME_MAX ); -/* -* CPU has to handle the local events on his own, when on the way to -* offline; Therefore nextevt value is set to KTIME_MAX -*/ - firstexp =__tmigr_cpu_deactivate (tmc,KTIME_MAX ); -trace_tmigr_cpu_unavailable (tmc); -raw_spin_unlock_irq (&tmc->lock); +/* +* CPU has to handle the local events on his own, when on the way to +* offline; Therefore nextevt value is set to KTIME_MAX +*/ + firstexp =__tmigr_cpu_deactivate (tmc,KTIME_MAX ); +trace_tmigr_cpu_unavailable (tmc); + }if (firstexp !=KTIME_MAX ) { migrator =cpumask_any (tmigr_available_cpumask);
@@ -1479,13 +1479,13 @@ static int tmigr_set_cpu_available(unsigned int cpu)return -EINVAL ;cpumask_set_cpu (cpu, tmigr_available_cpumask); -raw_spin_lock_irq (&tmc->lock); -trace_tmigr_cpu_available (tmc); - tmc->idle =timer_base_is_idle (); -if (!tmc->idle) -__tmigr_cpu_activate (tmc); - tmc->available = true; -raw_spin_unlock_irq (&tmc->lock); + scoped_guard(raw_spinlock_irq, &tmc->lock) { +trace_tmigr_cpu_available (tmc); + tmc->idle =timer_base_is_idle (); +if (!tmc->idle) +__tmigr_cpu_activate (tmc); + tmc->available = true; + }return 0 ; }
Signature · 1 line
2.51.0
PATCH RESEND v13 5/9 cgroup/cpuset: Rename update_unbound_workqueue_cpumask() to update_exclusion_cpumasks()
In reply to Gabriele Monacoupdate_unbound_workqueue_cpumask() updates unbound workqueues settings when there's a change in isolated CPUs, but it can be used for other subsystems requiring updated when isolated CPUs change.
Generalise the name to update_exclusion_cpumasks() to prepare for other functions unrelated to workqueues to be called in that spot.
--- kernel/cgroup/cpuset.c | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-)
kernel/cgroup/cpuset.c +6 -6
@@ -1391,7 +1391,7 @@ static bool partition_xcpus_del(int old_prs, struct cpuset *parent,return isolcpus_updated; } -static void update_unbound_workqueue_cpumask (bool isolcpus_updated) +static void update_exclusion_cpumasks (bool isolcpus_updated) {int ret;
@@ -1555,7 +1555,7 @@ static int remote_partition_enable(struct cpuset *cs, int new_prs,list_add (&cs->remote_sibling, &remote_children);cpumask_copy (cs->effective_xcpus, tmp->new_cpus);spin_unlock_irq (&callback_lock); -update_unbound_workqueue_cpumask (isolcpus_updated); +update_exclusion_cpumasks (isolcpus_updated);cpuset_force_rebuild (); cs->prs_err =0 ;
@@ -1596,7 +1596,7 @@ static void remote_partition_disable(struct cpuset *cs, struct tmpmasks *tmp)compute_excpus (cs, cs->effective_xcpus);reset_partition_data (cs);spin_unlock_irq (&callback_lock); -update_unbound_workqueue_cpumask (isolcpus_updated); +update_exclusion_cpumasks (isolcpus_updated);cpuset_force_rebuild ();/*
@@ -1665,7 +1665,7 @@ static void remote_cpus_update(struct cpuset *cs, struct cpumask *xcpus,if (xcpus)cpumask_copy (cs->exclusive_cpus, xcpus);spin_unlock_irq (&callback_lock); -update_unbound_workqueue_cpumask (isolcpus_updated); +update_exclusion_cpumasks (isolcpus_updated);if (adding || deleting)cpuset_force_rebuild ();
@@ -2023,7 +2023,7 @@ static int update_parent_effective_cpumask(struct cpuset *cs, int cmd,WARN_ON_ONCE (parent->nr_subparts <0 ); }spin_unlock_irq (&callback_lock); -update_unbound_workqueue_cpumask (isolcpus_updated); +update_exclusion_cpumasks (isolcpus_updated);if ((old_prs != new_prs) && (cmd == partcmd_update))update_partition_exclusive_flag (cs, new_prs);
@@ -3043,7 +3043,7 @@ static int update_prstate(struct cpuset *cs, int new_prs)else if (isolcpus_updated)isolated_cpus_update (old_prs, new_prs, cs->effective_xcpus);spin_unlock_irq (&callback_lock); -update_unbound_workqueue_cpumask (isolcpus_updated); +update_exclusion_cpumasks (isolcpus_updated);/* Force update if switching back to member & update effective_xcpus */ update_cpumasks_hier (cs, &tmpmask, !new_prs);
Signature · 1 line
2.51.0
PATCH RESEND v13 6/9 sched/isolation: Force housekeeping if isolcpus and nohz_full don't leave any
In reply to Gabriele MonacoCurrently the user can set up isolcpus and nohz_full in such a way that leaves no housekeeping CPU (i.e. no CPU that is neither domain isolated nor nohz full). This can be a problem for other subsystems (e.g. the timer wheel imgration).
Prevent this configuration by invalidating the last setting in case the union of isolcpus (domain) and nohz_full covers all CPUs.
--- kernel/sched/isolation.c | 23 +++++++++++++++++++++++ 1 file changed, 23 insertions(+)
kernel/sched/isolation.c +23 -0
@@ -167,6 +167,29 @@ static int __init housekeeping_setup(char *str, unsigned long flags) } } +/* +* Check the combination of nohz_full and isolcpus=domain, +* necessary to avoid problems with the timer migration +* hierarchy. managed_irq is ignored by this check since it +* isn't considered in the timer migration logic. +*/ + iter_flags = housekeeping.flags & (HK_FLAG_KERNEL_NOISE |HK_FLAG_DOMAIN ); + type =find_first_bit (&iter_flags,HK_TYPE_MAX ); +/* +* Pass the check if none of these flags were previously set or +* are not in the current selection. +*/ + iter_flags = flags & (HK_FLAG_KERNEL_NOISE |HK_FLAG_DOMAIN ); + first_cpu = (type ==HK_TYPE_MAX || !iter_flags) ?0 : +cpumask_first_and_and (cpu_present_mask, + housekeeping_staging, housekeeping.cpumasks[type]); +if (first_cpu >=min (nr_cpu_ids, setup_max_cpus)) { +pr_warn ("Housekeeping: must include one present CPU " +"neither in nohz_full= nor in isolcpus=domain, " +"ignoring setting %s\n" , str); + goto free_housekeeping_staging; + } + iter_flags = flags & ~housekeeping.flags;for_each_set_bit (type, &iter_flags,HK_TYPE_MAX )
Signature · 1 line
2.51.0
PATCH RESEND v13 7/9 cgroup/cpuset: Fail if isolated and nohz_full don't leave any housekeeping
In reply to Gabriele MonacoCurrently the user can set up isolated cpus via cpuset and nohz_full in such a way that leaves no housekeeping CPU (i.e. no CPU that is neither domain isolated nor nohz full). This can be a problem for other subsystems (e.g. the timer wheel imgration).
Prevent this configuration by blocking any assignation that would cause the union of domain isolated cpus and nohz_full to covers all CPUs.
--- kernel/cgroup/cpuset.c | 63 ++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 63 insertions(+)
kernel/cgroup/cpuset.c +63 -0
@@ -1327,6 +1327,19 @@ static void isolated_cpus_update(int old_prs, int new_prs, struct cpumask *xcpuscpumask_andnot (isolated_cpus, isolated_cpus, xcpus); } +/* +* isolated_cpus_should_update - Returns if the isolated_cpus mask needs update +* @prs: new or old partition_root_state +* @parent: parent cpuset +* Return: true if isolated_cpus needs modification, false otherwise +*/ +static bool isolated_cpus_should_update (int prs,struct cpuset *parent) +{ +if (!parent) + parent = &top_cpuset; +return prs != parent->partition_root_state; +} +/* * partition_xcpus_add - Add new exclusive CPUs to partition * @new_prs: new partition_root_state
@@ -1391,6 +1404,42 @@ static bool partition_xcpus_del(int old_prs, struct cpuset *parent,return isolcpus_updated; } +/* +* isolated_cpus_can_update - check for isolated & nohz_full conflicts +* @add_cpus: cpu mask for cpus that are going to be isolated +* @del_cpus: cpu mask for cpus that are no longer isolated, can be NULL +* Return: false if there is conflict, true otherwise +* +* If nohz_full is enabled and we have isolated CPUs, their combination must +* still leave housekeeping CPUs. +*/ +static bool isolated_cpus_can_update (struct cpumask *add_cpus, +struct cpumask *del_cpus) +{ +cpumask_var_t full_hk_cpus; +int res = true; + +if (!housekeeping_enabled (HK_TYPE_KERNEL_NOISE )) +return true; + +if (del_cpus &&cpumask_weight_and (del_cpus, +housekeeping_cpumask (HK_TYPE_KERNEL_NOISE ))) +return true; + +if (!alloc_cpumask_var (&full_hk_cpus,GFP_KERNEL )) +return false; + +cpumask_and (full_hk_cpus,housekeeping_cpumask (HK_TYPE_KERNEL_NOISE ), +housekeeping_cpumask (HK_TYPE_DOMAIN )); +cpumask_andnot (full_hk_cpus, full_hk_cpus, isolated_cpus); +cpumask_and (full_hk_cpus, full_hk_cpus, cpu_active_mask); +if (!cpumask_weight_andnot (full_hk_cpus, add_cpus)) + res = false; + +free_cpumask_var (full_hk_cpus); +return res; +} +static void update_exclusion_cpumasks (bool isolcpus_updated) {int ret;
@@ -1549,6 +1598,9 @@ static int remote_partition_enable(struct cpuset *cs, int new_prs,if (!cpumask_intersects (tmp->new_cpus, cpu_active_mask) ||cpumask_subset (top_cpuset.effective_cpus, tmp->new_cpus))return PERR_INVCPUS ; +if (isolated_cpus_should_update (new_prs,NULL ) && + !isolated_cpus_can_update (tmp->new_cpus,NULL )) +return PERR_HKEEPING ;spin_lock_irq (&callback_lock); isolcpus_updated =partition_xcpus_add (new_prs,NULL , tmp->new_cpus);
@@ -1648,6 +1700,9 @@ static void remote_cpus_update(struct cpuset *cs, struct cpumask *xcpus,else if (cpumask_intersects (tmp->addmask, subpartitions_cpus) ||cpumask_subset (top_cpuset.effective_cpus, tmp->addmask)) cs->prs_err =PERR_NOCPUS ; +else if (isolated_cpus_should_update (prs,NULL ) && + !isolated_cpus_can_update (tmp->addmask, tmp->delmask)) + cs->prs_err =PERR_HKEEPING ;if (cs->prs_err) goto invalidate; }
@@ -1994,6 +2049,12 @@ static int update_parent_effective_cpumask(struct cpuset *cs, int cmd,return err; } +if (deleting &&isolated_cpus_should_update (new_prs, parent) && + !isolated_cpus_can_update (tmp->delmask, tmp->addmask)) { + cs->prs_err =PERR_HKEEPING ; +return PERR_HKEEPING ; + } +/* * Change the parent's effective_cpus & effective_xcpus (top cpuset * only).
@@ -3009,6 +3070,8 @@ static int update_prstate(struct cpuset *cs, int new_prs)* Need to update isolated_cpus. */ isolcpus_updated = true; +if (!isolated_cpus_can_update (cs->effective_xcpus,NULL )) + err =PERR_HKEEPING ; }else {/* * Switching back to member is always allowed even if it
Signature · 1 line
2.51.0
PATCH RESEND v13 8/9 cpumask: Add initialiser to use cleanup helpers
In reply to Gabriele MonacoFrom: Yury Norov <[email protected]>
Now we can simplify a code that allocates cpumasks for local needs.
Automatic variables have to be initialized at declaration, or at least before any possibility for the logic to return, so that compiler wouldn't try to call an associate destructor function on a random stack number.
Because cpumask_var_t, depending on the CPUMASK_OFFSTACK config, is either a pointer or an array, we have to have a macro for initialization.
So define a CPUMASK_VAR_NULL macro, which allows to init struct cpumask pointer with NULL when CPUMASK_OFFSTACK is enabled, and effectively a no-op when CPUMASK_OFFSTACK is disabled (initialisation optimised out with -O2).
--- include/linux/cpumask.h | 2 ++ 1 file changed, 2 insertions(+)
include/linux/cpumask.h +2 -0
@@ -1005,6 +1005,7 @@ static __always_inline unsigned int cpumask_size(void)#define this_cpu_cpumask_var_ptr (x) this_cpu_read(x)#define __cpumask_var_read_mostly __read_mostly +#define CPUMASK_VAR_NULL NULLbool alloc_cpumask_var_node (cpumask_var_t *mask,gfp_t flags,int node);
@@ -1051,6 +1052,7 @@ static __always_inline bool cpumask_available(cpumask_var_t mask)#define this_cpu_cpumask_var_ptr (x) this_cpu_ptr(x)#define __cpumask_var_read_mostly +#define CPUMASK_VAR_NULL {} static __always_inline bool alloc_cpumask_var(cpumask_var_t *mask, gfp_t flags) {
Signature · 1 line
2.51.0
PATCH RESEND v13 9/9 timers: Exclude isolated cpus from timer migration
In reply to Gabriele MonacoThe timer migration mechanism allows active CPUs to pull timers from idle ones to improve the overall idle time. This is however undesired when CPU intensive workloads run on isolated cores, as the algorithm would move the timers from housekeeping to isolated cores, negatively affecting the isolation.
Exclude isolated cores from the timer migration algorithm, extend the concept of unavailable cores, currently used for offline ones, to isolated ones:
* A core is unavailable if isolated or offline;
* A core is available if non isolated and online;
A core is considered unavailable as isolated if it belongs to:
* the isolcpus (domain) list
* an isolated cpuset
Except if it is:
* in the nohz_full list (already idle for the hierarchy)
* the nohz timekeeper core (must be available to handle global timers)
CPUs are added to the hierarchy during late boot, excluding isolated ones, the hierarchy is also adapted when the cpuset isolation changes.
Due to how the timer migration algorithm works, any CPU part of the hierarchy can have their global timers pulled by remote CPUs and have to pull remote timers, only skipping pulling remote timers would break the logic.
For this reason, prevent isolated CPUs from pulling remote global timers, but also the other way around: any global timer started on an isolated CPU will run there. This does not break the concept of isolation (global timers don't come from outside the CPU) and, if considered inappropriate, can usually be mitigated with other isolation techniques (e.g. IRQ pinning).
This effect was noticed on a 128 cores machine running oslat on the isolated cores (1-31,33-63,65-95,97-127). The tool monopolises CPUs, and the CPU with lowest count in a timer migration hierarchy (here 1 and 65) appears as always active and continuously pulls global timers, from the housekeeping CPUs. This ends up moving driver work (e.g. delayed work) to isolated CPUs and causes latency spikes:
before the change:
# oslat -c 1-31,33-63,65-95,97-127 -D 62s ... Maximum: 1203 10 3 4 ... 5 (us)
after the change:
# oslat -c 1-31,33-63,65-95,97-127 -D 62s ... Maximum: 10 4 3 4 3 ... 5 (us)
The same behaviour was observed on a machine with as few as 20 cores / 40 threads with isocpus set to: 1-9,11-39 with rtla-osnoise-top.
--- include/linux/timer.h | 9 +++ kernel/cgroup/cpuset.c | 3 + kernel/time/timer_migration.c | 128 ++++++++++++++++++++++++++++++++-- 3 files changed, 135 insertions(+), 5 deletions(-)
include/linux/timer.h +9 -0
@@ -188,4 +188,13 @@ int timers_dead_cpu(unsigned int cpu);#define timers_dead_cpu NULL#endif +#if defined(CONFIG_SMP ) && defined(CONFIG_NO_HZ_COMMON ) +extern int tmigr_isolated_exclude_cpumask (struct cpumask *exclude_cpumask); +#else +static inline int tmigr_isolated_exclude_cpumask (struct cpumask *exclude_cpumask) +{ +return 0 ; +} +#endif +#endif
kernel/cgroup/cpuset.c +3 -0
@@ -1451,6 +1451,9 @@ static void update_exclusion_cpumasks(bool isolcpus_updated) ret =workqueue_unbound_exclude_cpumask (isolated_cpus);WARN_ON_ONCE (ret <0 ); + + ret =tmigr_isolated_exclude_cpumask (isolated_cpus); +WARN_ON_ONCE (ret <0 ); }/**
kernel/time/timer_migration.c +123 -5
175 lines not on this page · raw message
Signature · 1 line
2.51.0
Quoted 65 lines · Gabriele Monaco
The timer migration mechanism allows active CPUs to pull timers from idle ones to improve the overall idle time. This is however undesired when CPU intensive workloads run on isolated cores, as the algorithm would move the timers from housekeeping to isolated cores, negatively affecting the isolation.
Exclude isolated cores from the timer migration algorithm, extend the concept of unavailable cores, currently used for offline ones, to isolated ones:
* A core is unavailable if isolated or offline;
* A core is available if non isolated and online;A core is considered unavailable as isolated if it belongs to:
* the isolcpus (domain) list
* an isolated cpuset
Except if it is:
* in thenohz_fulllist (already idle for the hierarchy)
* the nohz timekeeper core (must be available to handle global timers)CPUs are added to the hierarchy during late boot, excluding isolated ones, the hierarchy is also adapted when the cpuset isolation changes.
Due to how the timer migration algorithm works, any CPU part of the hierarchy can have their global timers pulled by remote CPUs and have to pull remote timers, only skipping pulling remote timers would break the logic.
For this reason, prevent isolated CPUs from pulling remote global timers, but also the other way around: any global timer started on an isolated CPU will run there. This does not break the concept of isolation (global timers don't come from outside the CPU) and, if considered inappropriate, can usually be mitigated with other isolation techniques (e.g. IRQ pinning).This effect was noticed on a 128 cores machine running oslat on the isolated cores (1-31,33-63,65-95,97-127). The tool monopolises CPUs, and the CPU with lowest count in a timer migration hierarchy (here 1 and 65) appears as always active and continuously pulls global timers, from the housekeeping CPUs. This ends up moving driver work (e.g. delayed work) to isolated CPUs and causes latency spikes:
before the change:
# oslat -c 1-31,33-63,65-95,97-127 -D 62s ... Maximum: 1203 10 3 4 ... 5 (us)after the change:
# oslat -c 1-31,33-63,65-95,97-127 -D 62s ... Maximum: 10 4 3 4 3 ... 5 (us)The same behaviour was observed on a machine with as few as 20 cores / 40 threads with isocpus set to: 1-9,11-39 with rtla-osnoise-top.
The first 5 patches are preparatory work to change the concept of online/offline to available/unavailable, keep track of those in a separate cpumask cleanup the setting/clearing functions and change a function name in cpuset code.
Patch 6 and 7 adapt isolation and cpuset to prevent domain isolated and
nohz_fullfrom covering all CPUs not leaving any housekeeping one. This can lead to problems with the changes introduced in this series because no CPU would remain to handle global timers.
Quoted 4 lines
Patch 9 extends the unavailable status to domain isolated CPUs, which is the main contribution of the series.
This series is equivalent to v13 but rebased on v6.18-rc2.
Thomas,
This patch series have undergone multiple round of reviews. Do you think it is good enough to be merged into tip?
It does contain some cpuset code, but most of the changes are in the timer code. So I think it is better to go through the tip tree. It does have some minor conflicts with the current for-6.19 branch of the cgroup tree, but it can be easily resolved during merge.
What do you think?
Thanks,
Longman
Frederic Weisbecker
maintainerHi Waiman,
Quoted 76 lines · Waiman Long
Quoted 65 lines · Gabriele Monaco
The timer migration mechanism allows active CPUs to pull timers from idle ones to improve the overall idle time. This is however undesired when CPU intensive workloads run on isolated cores, as the algorithm would move the timers from housekeeping to isolated cores, negatively affecting the isolation.
Exclude isolated cores from the timer migration algorithm, extend the concept of unavailable cores, currently used for offline ones, to isolated ones:
* A core is unavailable if isolated or offline;
* A core is available if non isolated and online;A core is considered unavailable as isolated if it belongs to:
* the isolcpus (domain) list
* an isolated cpuset
Except if it is:
* in thenohz_fulllist (already idle for the hierarchy)
* the nohz timekeeper core (must be available to handle global timers)CPUs are added to the hierarchy during late boot, excluding isolated ones, the hierarchy is also adapted when the cpuset isolation changes.
Due to how the timer migration algorithm works, any CPU part of the hierarchy can have their global timers pulled by remote CPUs and have to pull remote timers, only skipping pulling remote timers would break the logic.
For this reason, prevent isolated CPUs from pulling remote global timers, but also the other way around: any global timer started on an isolated CPU will run there. This does not break the concept of isolation (global timers don't come from outside the CPU) and, if considered inappropriate, can usually be mitigated with other isolation techniques (e.g. IRQ pinning).This effect was noticed on a 128 cores machine running oslat on the isolated cores (1-31,33-63,65-95,97-127). The tool monopolises CPUs, and the CPU with lowest count in a timer migration hierarchy (here 1 and 65) appears as always active and continuously pulls global timers, from the housekeeping CPUs. This ends up moving driver work (e.g. delayed work) to isolated CPUs and causes latency spikes:
before the change:
# oslat -c 1-31,33-63,65-95,97-127 -D 62s ... Maximum: 1203 10 3 4 ... 5 (us)after the change:
# oslat -c 1-31,33-63,65-95,97-127 -D 62s ... Maximum: 10 4 3 4 3 ... 5 (us)The same behaviour was observed on a machine with as few as 20 cores / 40 threads with isocpus set to: 1-9,11-39 with rtla-osnoise-top.
The first 5 patches are preparatory work to change the concept of online/offline to available/unavailable, keep track of those in a separate cpumask cleanup the setting/clearing functions and change a function name in cpuset code.
Patch 6 and 7 adapt isolation and cpuset to prevent domain isolated and
nohz_fullfrom covering all CPUs not leaving any housekeeping one. This can lead to problems with the changes introduced in this series because no CPU would remain to handle global timers.Quoted 4 lines
Patch 9 extends the unavailable status to domain isolated CPUs, which is the main contribution of the series.
This series is equivalent to v13 but rebased on v6.18-rc2.
Thomas,
This patch series have undergone multiple round of reviews. Do you think it is good enough to be merged into tip?
Quoted 6 lines
It does contain some cpuset code, but most of the changes are in the timer code. So I think it is better to go through the tip tree. It does have some minor conflicts with the current for-6.19 branch of the cgroup tree, but it can be easily resolved during merge.
What do you think?
Just wait a little, I realize I made a buggy suggestion to Gabriele and a detail needs to be fixed.
My bad...
Signature · 2 lines
Frederic Weisbecker SUSE Labs
A message that is not in the archive (sent off-list or not yet received): [email protected]
Quoted 103 lines · Waiman Long
On 10/30/25 10:12 AM, Frederic Weisbecker wrote:Quoted 99 lines
Hi Waiman,Le Wed, Oct 29, 2025 at 10:56:06PM -0400, Waiman Long a écrit :
Quoted 88 lines
On 10/20/25 7:27 AM, Gabriele Monaco wrote:Quoted 66 lines
The timer migration mechanism allows active CPUs to pull timers from idle ones to improve the overall idle time. This is however undesired when CPU intensive workloads run on isolated cores, as the algorithm would move the timers from housekeeping to isolated cores, negatively affecting the isolation.Exclude isolated cores from the timer migration algorithm, extend the concept of unavailable cores, currently used for offline ones, to isolated ones:
* A core is unavailable if isolated or offline;
* A core is available if non isolated and online;A core is considered unavailable as isolated if it belongs to:
* the isolcpus (domain) list
* an isolated cpuset
Except if it is:
* in thenohz_fulllist (already idle for the hierarchy)
* the nohz timekeeper core (must be available to handle global timers)CPUs are added to the hierarchy during late boot, excluding isolated ones, the hierarchy is also adapted when the cpuset isolation changes.
Due to how the timer migration algorithm works, any CPU part of the hierarchy can have their global timers pulled by remote CPUs and have to pull remote timers, only skipping pulling remote timers would break the logic.
For this reason, prevent isolated CPUs from pulling remote global timers, but also the other way around: any global timer started on an isolated CPU will run there. This does not break the concept of isolation (global timers don't come from outside the CPU) and, if considered inappropriate, can usually be mitigated with other isolation techniques (e.g. IRQ pinning).This effect was noticed on a 128 cores machine running oslat on the isolated cores (1-31,33-63,65-95,97-127). The tool monopolises CPUs, and the CPU with lowest count in a timer migration hierarchy (here 1 and 65) appears as always active and continuously pulls global timers, from the housekeeping CPUs. This ends up moving driver work (e.g. delayed work) to isolated CPUs and causes latency spikes:
before the change:
# oslat -c 1-31,33-63,65-95,97-127 -D 62s ... Maximum: 1203 10 3 4 ... 5 (us)after the change:
# oslat -c 1-31,33-63,65-95,97-127 -D 62s ... Maximum: 10 4 3 4 3 ... 5 (us)The same behaviour was observed on a machine with as few as 20 cores / 40 threads with isocpus set to: 1-9,11-39 with rtla-osnoise-top.
The first 5 patches are preparatory work to change the concept of online/offline to available/unavailable, keep track of those in a separate cpumask cleanup the setting/clearing functions and change a function name in cpuset code.
Patch 6 and 7 adapt isolation and cpuset to prevent domain isolated and
nohz_fullfrom covering all CPUs not leaving any housekeeping one. This can lead to problems with the changes introduced in this series because no CPU would remain to handle global timers.Quoted 5 lines
Patch 9 extends the unavailable status to domain isolated CPUs, which is the main contribution of the series.
This series is equivalent to v13 but rebased on v6.18-rc2.
Thomas,This patch series have undergone multiple round of reviews. Do you think it
is good enough to be merged into tip?It does contain some cpuset code, but most of the changes are in the timer code. So I think it is better to go through the tip tree. It does have some
minor conflicts with the current for-6.19 branch of the cgroup tree, but it
can be easily resolved during merge.Quoted 2 lines
What do you think?
Just wait a little, I realize I made a buggy suggestion to Gabriele and a detail needs to be fixed.
Quoted 5 lines
Quoted 2 lines
My bad...
OK, I thought you were OK with the timer changes. I guess Gabriele will have to send out a new version to address your finding.
Sure, I'm going to have a look at this next week and send a V14.
Thanks,
Gabriele
Quoted 105 lines · Gabriele Monaco
Quoted 103 lines · Waiman Long
Quoted 99 lines · Frederic Weisbecker
Hi Waiman,
Quoted 81 lines · Waiman Long
Quoted 66 lines · Gabriele Monaco
The timer migration mechanism allows active CPUs to pull timers from idle ones to improve the overall idle time. This is however undesired when CPU intensive workloads run on isolated cores, as the algorithm would move the timers from housekeeping to isolated cores, negatively affecting the isolation.
Exclude isolated cores from the timer migration algorithm, extend the concept of unavailable cores, currently used for offline ones, to isolated ones:
* A core is unavailable if isolated or offline;
* A core is available if non isolated and online;A core is considered unavailable as isolated if it belongs to:
* the isolcpus (domain) list
* an isolated cpuset
Except if it is:
* in thenohz_fulllist (already idle for the hierarchy)
* the nohz timekeeper core (must be available to handle global timers)CPUs are added to the hierarchy during late boot, excluding isolated ones, the hierarchy is also adapted when the cpuset isolation changes.
Due to how the timer migration algorithm works, any CPU part of the hierarchy can have their global timers pulled by remote CPUs and have to pull remote timers, only skipping pulling remote timers would break the logic.
For this reason, prevent isolated CPUs from pulling remote global timers, but also the other way around: any global timer started on an isolated CPU will run there. This does not break the concept of isolation (global timers don't come from outside the CPU) and, if considered inappropriate, can usually be mitigated with other isolation techniques (e.g. IRQ pinning).This effect was noticed on a 128 cores machine running oslat on the isolated cores (1-31,33-63,65-95,97-127). The tool monopolises CPUs, and the CPU with lowest count in a timer migration hierarchy (here 1 and 65) appears as always active and continuously pulls global timers, from the housekeeping CPUs. This ends up moving driver work (e.g. delayed work) to isolated CPUs and causes latency spikes:
before the change:
# oslat -c 1-31,33-63,65-95,97-127 -D 62s ... Maximum: 1203 10 3 4 ... 5 (us)after the change:
# oslat -c 1-31,33-63,65-95,97-127 -D 62s ... Maximum: 10 4 3 4 3 ... 5 (us)The same behaviour was observed on a machine with as few as 20 cores / 40 threads with isocpus set to: 1-9,11-39 with rtla-osnoise-top.
The first 5 patches are preparatory work to change the concept of online/offline to available/unavailable, keep track of those in a separate cpumask cleanup the setting/clearing functions and change a function name in cpuset code.
Patch 6 and 7 adapt isolation and cpuset to prevent domain isolated and
nohz_fullfrom covering all CPUs not leaving any housekeeping one. This can lead to problems with the changes introduced in this series because no CPU would remain to handle global timers.Quoted 5 lines
Patch 9 extends the unavailable status to domain isolated CPUs, which is the main contribution of the series.
This series is equivalent to v13 but rebased on v6.18-rc2.
Thomas,
This patch series have undergone multiple round of reviews. Do you think it
is good enough to be merged into tip?Quoted 9 lines
It does contain some cpuset code, but most of the changes are in the timer code. So I think it is better to go through the tip tree. It does have some
minor conflicts with the current for-6.19 branch of the cgroup tree, but it
can be easily resolved during merge.What do you think?
Just wait a little, I realize I made a buggy suggestion to Gabriele and a detail needs to be fixed.
Quoted 6 lines
Quoted 5 lines
Quoted 2 lines
My bad...
OK, I thought you were OK with the timer changes. I guess Gabriele will have to send out a new version to address your finding.
Sure, I'm going to have a look at this next week and send a V14.
I am going to extract out your 2 cpuset patches and send them to the cgroup mailing list separately. So you don't need to include them in your next version.
Cheers,
Longman
Frederic Weisbecker
maintainerQuoted 94 lines · Waiman Long
Quoted 93 lines · Gabriele Monaco
Quoted 91 lines · Waiman Long
Quoted 90 lines · Frederic Weisbecker
Hi Waiman,
Quoted 76 lines · Waiman Long
Quoted 65 lines · Gabriele Monaco
The timer migration mechanism allows active CPUs to pull timers from idle ones to improve the overall idle time. This is however undesired when CPU intensive workloads run on isolated cores, as the algorithm would move the timers from housekeeping to isolated cores, negatively affecting the isolation.
Exclude isolated cores from the timer migration algorithm, extend the concept of unavailable cores, currently used for offline ones, to isolated ones:
* A core is unavailable if isolated or offline;
* A core is available if non isolated and online;A core is considered unavailable as isolated if it belongs to:
* the isolcpus (domain) list
* an isolated cpuset
Except if it is:
* in thenohz_fulllist (already idle for the hierarchy)
* the nohz timekeeper core (must be available to handle global timers)CPUs are added to the hierarchy during late boot, excluding isolated ones, the hierarchy is also adapted when the cpuset isolation changes.
Due to how the timer migration algorithm works, any CPU part of the hierarchy can have their global timers pulled by remote CPUs and have to pull remote timers, only skipping pulling remote timers would break the logic.
For this reason, prevent isolated CPUs from pulling remote global timers, but also the other way around: any global timer started on an isolated CPU will run there. This does not break the concept of isolation (global timers don't come from outside the CPU) and, if considered inappropriate, can usually be mitigated with other isolation techniques (e.g. IRQ pinning).This effect was noticed on a 128 cores machine running oslat on the isolated cores (1-31,33-63,65-95,97-127). The tool monopolises CPUs, and the CPU with lowest count in a timer migration hierarchy (here 1 and 65) appears as always active and continuously pulls global timers, from the housekeeping CPUs. This ends up moving driver work (e.g. delayed work) to isolated CPUs and causes latency spikes:
before the change:
# oslat -c 1-31,33-63,65-95,97-127 -D 62s ... Maximum: 1203 10 3 4 ... 5 (us)after the change:
# oslat -c 1-31,33-63,65-95,97-127 -D 62s ... Maximum: 10 4 3 4 3 ... 5 (us)The same behaviour was observed on a machine with as few as 20 cores / 40 threads with isocpus set to: 1-9,11-39 with rtla-osnoise-top.
The first 5 patches are preparatory work to change the concept of online/offline to available/unavailable, keep track of those in a separate cpumask cleanup the setting/clearing functions and change a function name in cpuset code.
Patch 6 and 7 adapt isolation and cpuset to prevent domain isolated and
nohz_fullfrom covering all CPUs not leaving any housekeeping one. This can lead to problems with the changes introduced in this series because no CPU would remain to handle global timers.Quoted 4 lines
Patch 9 extends the unavailable status to domain isolated CPUs, which is the main contribution of the series.
This series is equivalent to v13 but rebased on v6.18-rc2.
Thomas,
This patch series have undergone multiple round of reviews. Do you think it
is good enough to be merged into tip?Quoted 8 lines
It does contain some cpuset code, but most of the changes are in the timer code. So I think it is better to go through the tip tree. It does have some
minor conflicts with the current for-6.19 branch of the cgroup tree, but it
can be easily resolved during merge.What do you think?
Just wait a little, I realize I made a buggy suggestion to Gabriele and a detail needs to be fixed.
Quoted 8 lines
Quoted 4 lines
Quoted 3 lines
Quoted 1 line
My bad...
OK, I thought you were OK with the timer changes. I guess Gabriele will have to send out a new version to address your finding.
Sure, I'm going to have a look at this next week and send a V14.
I am going to extract out your 2 cpuset patches and send them to the cgroup mailing list separately. So you don't need to include them in your next version.
I'm not sure this will help if you apply those to an external tree if the plan is to apply the whole to the timer tree. Or we'll create a dependency issue...
Quoted 3 lines
Cheers,
Longman
Signature · 2 lines
Frederic Weisbecker SUSE Labs
Quoted 97 lines · Frederic Weisbecker
Quoted 96 lines · Waiman Long
Quoted 95 lines · Gabriele Monaco
Quoted 94 lines · Waiman Long
Quoted 86 lines · Frederic Weisbecker
Hi Waiman,
Quoted 83 lines · Waiman Long
Quoted 65 lines · Gabriele Monaco
The timer migration mechanism allows active CPUs to pull timers from idle ones to improve the overall idle time. This is however undesired when CPU intensive workloads run on isolated cores, as the algorithm would move the timers from housekeeping to isolated cores, negatively affecting the isolation.
Exclude isolated cores from the timer migration algorithm, extend the concept of unavailable cores, currently used for offline ones, to isolated ones:
* A core is unavailable if isolated or offline;
* A core is available if non isolated and online;A core is considered unavailable as isolated if it belongs to:
* the isolcpus (domain) list
* an isolated cpuset
Except if it is:
* in thenohz_fulllist (already idle for the hierarchy)
* the nohz timekeeper core (must be available to handle global timers)CPUs are added to the hierarchy during late boot, excluding isolated ones, the hierarchy is also adapted when the cpuset isolation changes.
Due to how the timer migration algorithm works, any CPU part of the hierarchy can have their global timers pulled by remote CPUs and have to pull remote timers, only skipping pulling remote timers would break the logic.
For this reason, prevent isolated CPUs from pulling remote global timers, but also the other way around: any global timer started on an isolated CPU will run there. This does not break the concept of isolation (global timers don't come from outside the CPU) and, if considered inappropriate, can usually be mitigated with other isolation techniques (e.g. IRQ pinning).This effect was noticed on a 128 cores machine running oslat on the isolated cores (1-31,33-63,65-95,97-127). The tool monopolises CPUs, and the CPU with lowest count in a timer migration hierarchy (here 1 and 65) appears as always active and continuously pulls global timers, from the housekeeping CPUs. This ends up moving driver work (e.g. delayed work) to isolated CPUs and causes latency spikes:
before the change:
# oslat -c 1-31,33-63,65-95,97-127 -D 62s ... Maximum: 1203 10 3 4 ... 5 (us)after the change:
# oslat -c 1-31,33-63,65-95,97-127 -D 62s ... Maximum: 10 4 3 4 3 ... 5 (us)The same behaviour was observed on a machine with as few as 20 cores / 40 threads with isocpus set to: 1-9,11-39 with rtla-osnoise-top.
The first 5 patches are preparatory work to change the concept of online/offline to available/unavailable, keep track of those in a separate cpumask cleanup the setting/clearing functions and change a function name in cpuset code.
Patch 6 and 7 adapt isolation and cpuset to prevent domain isolated and
nohz_fullfrom covering all CPUs not leaving any housekeeping one. This can lead to problems with the changes introduced in this series because no CPU would remain to handle global timers.Quoted 4 lines
Patch 9 extends the unavailable status to domain isolated CPUs, which is the main contribution of the series.
This series is equivalent to v13 but rebased on v6.18-rc2.
Thomas,
This patch series have undergone multiple round of reviews. Do you think it
is good enough to be merged into tip?It does contain some cpuset code, but most of the changes are in the timer code. So I think it is better to go through the tip tree. It does have some
minor conflicts with the current for-6.19 branch of the cgroup tree, but it
can be easily resolved during merge.Quoted 5 lines
Quoted 1 line
What do you think?
Just wait a little, I realize I made a buggy suggestion to Gabriele and a detail needs to be fixed.
My bad...
OK, I thought you were OK with the timer changes. I guess Gabriele will have to send out a new version to address your finding.
Quoted 7 lines
Quoted 4 lines
Quoted 1 line
Sure, I'm going to have a look at this next week and send a V14.
I am going to extract out your 2 cpuset patches and send them to the cgroup mailing list separately. So you don't need to include them in your next version.
I'm not sure this will help if you apply those to an external tree if the plan is to apply the whole to the timer tree. Or we'll create a dependency issue...
These 2 cpuset patches are actually independent of the timer related changes. The purpose of these two patches are to prevent the cpuset code from adding isolated CPUs in such a way that all the nohz_full HK CPUs become domain-isolated. This is a corner case that normal users won't try to do. The patches are just an insurance policy to ensure that users can't do that. This is complementary to the sched/isolation patch that limits what CPUs can be put to the isolcpus and nohz_full boot parameters. All these patches are independent of the timer related changes, though you can say that the solution will only be complete if all the pieces are in place.
There are another set of pending cpuset patches from Chen Ridong that does some restructuring of the cpuset code that will likely have some conflicts with these 2 patches. So I would like to settle the cpuset changes to avoid future conflicts.
Cheers,
Longman
Frederic Weisbecker
maintainerQuoted 106 lines · Waiman Long
Quoted 97 lines · Frederic Weisbecker
Quoted 96 lines · Waiman Long
Quoted 95 lines · Gabriele Monaco
Quoted 94 lines · Waiman Long
Quoted 86 lines · Frederic Weisbecker
Hi Waiman,
Quoted 83 lines · Waiman Long
Quoted 65 lines · Gabriele Monaco
The timer migration mechanism allows active CPUs to pull timers from idle ones to improve the overall idle time. This is however undesired when CPU intensive workloads run on isolated cores, as the algorithm would move the timers from housekeeping to isolated cores, negatively affecting the isolation.
Exclude isolated cores from the timer migration algorithm, extend the concept of unavailable cores, currently used for offline ones, to isolated ones:
* A core is unavailable if isolated or offline;
* A core is available if non isolated and online;A core is considered unavailable as isolated if it belongs to:
* the isolcpus (domain) list
* an isolated cpuset
Except if it is:
* in thenohz_fulllist (already idle for the hierarchy)
* the nohz timekeeper core (must be available to handle global timers)CPUs are added to the hierarchy during late boot, excluding isolated ones, the hierarchy is also adapted when the cpuset isolation changes.
Due to how the timer migration algorithm works, any CPU part of the hierarchy can have their global timers pulled by remote CPUs and have to pull remote timers, only skipping pulling remote timers would break the logic.
For this reason, prevent isolated CPUs from pulling remote global timers, but also the other way around: any global timer started on an isolated CPU will run there. This does not break the concept of isolation (global timers don't come from outside the CPU) and, if considered inappropriate, can usually be mitigated with other isolation techniques (e.g. IRQ pinning).This effect was noticed on a 128 cores machine running oslat on the isolated cores (1-31,33-63,65-95,97-127). The tool monopolises CPUs, and the CPU with lowest count in a timer migration hierarchy (here 1 and 65) appears as always active and continuously pulls global timers, from the housekeeping CPUs. This ends up moving driver work (e.g. delayed work) to isolated CPUs and causes latency spikes:
before the change:
# oslat -c 1-31,33-63,65-95,97-127 -D 62s ... Maximum: 1203 10 3 4 ... 5 (us)after the change:
# oslat -c 1-31,33-63,65-95,97-127 -D 62s ... Maximum: 10 4 3 4 3 ... 5 (us)The same behaviour was observed on a machine with as few as 20 cores / 40 threads with isocpus set to: 1-9,11-39 with rtla-osnoise-top.
The first 5 patches are preparatory work to change the concept of online/offline to available/unavailable, keep track of those in a separate cpumask cleanup the setting/clearing functions and change a function name in cpuset code.
Patch 6 and 7 adapt isolation and cpuset to prevent domain isolated and
nohz_fullfrom covering all CPUs not leaving any housekeeping one. This can lead to problems with the changes introduced in this series because no CPU would remain to handle global timers.Quoted 4 lines
Patch 9 extends the unavailable status to domain isolated CPUs, which is the main contribution of the series.
This series is equivalent to v13 but rebased on v6.18-rc2.
Thomas,
This patch series have undergone multiple round of reviews. Do you think it
is good enough to be merged into tip?It does contain some cpuset code, but most of the changes are in the timer code. So I think it is better to go through the tip tree. It does have some
minor conflicts with the current for-6.19 branch of the cgroup tree, but it
can be easily resolved during merge.Quoted 5 lines
Quoted 1 line
What do you think?
Just wait a little, I realize I made a buggy suggestion to Gabriele and a detail needs to be fixed.
My bad...
OK, I thought you were OK with the timer changes. I guess Gabriele will have to send out a new version to address your finding.
Quoted 7 lines
Quoted 4 lines
Quoted 1 line
Sure, I'm going to have a look at this next week and send a V14.
I am going to extract out your 2 cpuset patches and send them to the cgroup mailing list separately. So you don't need to include them in your next version.
I'm not sure this will help if you apply those to an external tree if the plan is to apply the whole to the timer tree. Or we'll create a dependency issue...
Quoted 9 lines
These 2 cpuset patches are actually independent of the timer related changes. The purpose of these two patches are to prevent the cpuset code from adding isolated CPUs in such a way that all the
nohz_fullHK CPUs become domain-isolated. This is a corner case that normal users won't try to do. The patches are just an insurance policy to ensure that users can't do that. This is complementary to the sched/isolation patch that limits what CPUs can be put to the isolcpus andnohz_fullboot parameters. All these patches are independent of the timer related changes, though you can say that the solution will only be complete if all the pieces are in place.
Right but there will be a conflict if the timer patches don't have the rename of update_unbound_workqueue_cpumask().
Quoted 4 lines
There are another set of pending cpuset patches from Chen Ridong that does some restructuring of the cpuset code that will likely have some conflicts with these 2 patches. So I would like to settle the cpuset changes to avoid future conflicts.
Ok so it looks like there will be conflicts eventually during the merge window. In that case it makes sense to take Gabriel cpuset patches but he'll need to rebase the rest on top of the timer tree.
Thanks.
Quoted 4 lines
Cheers,
Longman
Signature · 2 lines
Frederic Weisbecker SUSE Labs
Quoted 11 lines · Frederic Weisbecker
Quoted 9 lines · Waiman Long
These 2 cpuset patches are actually independent of the timer related changes. The purpose of these two patches are to prevent the cpuset code from adding isolated CPUs in such a way that all the
nohz_fullHK CPUs become domain-isolated. This is a corner case that normal users won't try to do. The patches are just an insurance policy to ensure that users can't do that. This is complementary to the sched/isolation patch that limits what CPUs can be put to the isolcpus andnohz_fullboot parameters. All these patches are independent of the timer related changes, though you can say that the solution will only be complete if all the pieces are in place.
Quoted 3 lines
Right but there will be a conflict if the timer patches don't have the rename of
update_unbound_workqueue_cpumask().
Waiman, are you referring to [1]?
Since that is an RFC, couldn't you just take in those patches before merging [1] and adapt just that one directly in the cpuset tree? I guess git should figure it out if we keep my cpuset patches in both trees (or at least 5/9), as long as the conflicts come in later commits.
Different story is if you already took some conflicting patches in, then I can look into rebasing as Frederic suggests.
Thanks,
Gabriele
[1] -
https://lore.kernel.org/lkml/[email protected]
Quoted 15 lines
Quoted 4 lines
There are another set of pending cpuset patches from Chen Ridong that does some restructuring of the cpuset code that will likely have some conflicts with these 2 patches. So I would like to settle the cpuset changes to avoid future conflicts.
Ok so it looks like there will be conflicts eventually during the merge window. In that case it makes sense to take Gabriel cpuset patches but he'll need to rebase the rest on top of the timer tree.
Thanks.
Quoted 4 lines
Cheers,
Longman
Quoted 115 lines · Frederic Weisbecker
Quoted 105 lines · Waiman Long
Quoted 97 lines · Frederic Weisbecker
Quoted 96 lines · Waiman Long
Quoted 95 lines · Gabriele Monaco
Quoted 94 lines · Waiman Long
Quoted 86 lines · Frederic Weisbecker
Hi Waiman,
Quoted 83 lines · Waiman Long
Quoted 65 lines · Gabriele Monaco
The timer migration mechanism allows active CPUs to pull timers from idle ones to improve the overall idle time. This is however undesired when CPU intensive workloads run on isolated cores, as the algorithm would move the timers from housekeeping to isolated cores, negatively affecting the isolation.
Exclude isolated cores from the timer migration algorithm, extend the concept of unavailable cores, currently used for offline ones, to isolated ones:
* A core is unavailable if isolated or offline;
* A core is available if non isolated and online;A core is considered unavailable as isolated if it belongs to:
* the isolcpus (domain) list
* an isolated cpuset
Except if it is:
* in thenohz_fulllist (already idle for the hierarchy)
* the nohz timekeeper core (must be available to handle global timers)CPUs are added to the hierarchy during late boot, excluding isolated ones, the hierarchy is also adapted when the cpuset isolation changes.
Due to how the timer migration algorithm works, any CPU part of the hierarchy can have their global timers pulled by remote CPUs and have to pull remote timers, only skipping pulling remote timers would break the logic.
For this reason, prevent isolated CPUs from pulling remote global timers, but also the other way around: any global timer started on an isolated CPU will run there. This does not break the concept of isolation (global timers don't come from outside the CPU) and, if considered inappropriate, can usually be mitigated with other isolation techniques (e.g. IRQ pinning).This effect was noticed on a 128 cores machine running oslat on the isolated cores (1-31,33-63,65-95,97-127). The tool monopolises CPUs, and the CPU with lowest count in a timer migration hierarchy (here 1 and 65) appears as always active and continuously pulls global timers, from the housekeeping CPUs. This ends up moving driver work (e.g. delayed work) to isolated CPUs and causes latency spikes:
before the change:
# oslat -c 1-31,33-63,65-95,97-127 -D 62s ... Maximum: 1203 10 3 4 ... 5 (us)after the change:
# oslat -c 1-31,33-63,65-95,97-127 -D 62s ... Maximum: 10 4 3 4 3 ... 5 (us)The same behaviour was observed on a machine with as few as 20 cores / 40 threads with isocpus set to: 1-9,11-39 with rtla-osnoise-top.
The first 5 patches are preparatory work to change the concept of online/offline to available/unavailable, keep track of those in a separate cpumask cleanup the setting/clearing functions and change a function name in cpuset code.
Patch 6 and 7 adapt isolation and cpuset to prevent domain isolated and
nohz_fullfrom covering all CPUs not leaving any housekeeping one. This can lead to problems with the changes introduced in this series because no CPU would remain to handle global timers.Quoted 4 lines
Patch 9 extends the unavailable status to domain isolated CPUs, which is the main contribution of the series.
This series is equivalent to v13 but rebased on v6.18-rc2.
Thomas,
This patch series have undergone multiple round of reviews. Do you think it
is good enough to be merged into tip?It does contain some cpuset code, but most of the changes are in the timer code. So I think it is better to go through the tip tree. It does have some
minor conflicts with the current for-6.19 branch of the cgroup tree, but it
can be easily resolved during merge.Quoted 5 lines
Quoted 1 line
What do you think?
Just wait a little, I realize I made a buggy suggestion to Gabriele and a detail needs to be fixed.
My bad...
OK, I thought you were OK with the timer changes. I guess Gabriele will have to send out a new version to address your finding.
Quoted 7 lines
Quoted 4 lines
Quoted 1 line
Sure, I'm going to have a look at this next week and send a V14.
I am going to extract out your 2 cpuset patches and send them to the cgroup mailing list separately. So you don't need to include them in your next version.
I'm not sure this will help if you apply those to an external tree if the plan is to apply the whole to the timer tree. Or we'll create a dependency issue...
Quoted 9 lines
These 2 cpuset patches are actually independent of the timer related changes. The purpose of these two patches are to prevent the cpuset code from adding isolated CPUs in such a way that all the
nohz_fullHK CPUs become domain-isolated. This is a corner case that normal users won't try to do. The patches are just an insurance policy to ensure that users can't do that. This is complementary to the sched/isolation patch that limits what CPUs can be put to the isolcpus andnohz_fullboot parameters. All these patches are independent of the timer related changes, though you can say that the solution will only be complete if all the pieces are in place.
Quoted 2 lines
Right but there will be a conflict if the timer patches don't have the rename of
update_unbound_workqueue_cpumask().
Yes, I missed the fact that patch 9 does have a cpuset.c hunk. Yes, it will have a dependency on the prior cpuset patches.
Quoted 7 lines
Quoted 4 lines
There are another set of pending cpuset patches from Chen Ridong that does some restructuring of the cpuset code that will likely have some conflicts with these 2 patches. So I would like to settle the cpuset changes to avoid future conflicts.
Ok so it looks like there will be conflicts eventually during the merge window. In that case it makes sense to take Gabriel cpuset patches but he'll need to rebase the rest on top of the timer tree.
It depends on whether the patch set can be merged in time into tip for the next v6.19 merge windows.
Cheers,
Longman
Quoted 6 lines
Thanks.
Quoted 3 lines
Cheers,
Longman
Frederic Weisbecker
maintainerQuoted 84 lines · Waiman Long
Quoted 83 lines · Frederic Weisbecker
Hi Waiman,
Quoted 80 lines · Waiman Long
Quoted 65 lines · Gabriele Monaco
The timer migration mechanism allows active CPUs to pull timers from idle ones to improve the overall idle time. This is however undesired when CPU intensive workloads run on isolated cores, as the algorithm would move the timers from housekeeping to isolated cores, negatively affecting the isolation.
Exclude isolated cores from the timer migration algorithm, extend the concept of unavailable cores, currently used for offline ones, to isolated ones:
* A core is unavailable if isolated or offline;
* A core is available if non isolated and online;A core is considered unavailable as isolated if it belongs to:
* the isolcpus (domain) list
* an isolated cpuset
Except if it is:
* in thenohz_fulllist (already idle for the hierarchy)
* the nohz timekeeper core (must be available to handle global timers)CPUs are added to the hierarchy during late boot, excluding isolated ones, the hierarchy is also adapted when the cpuset isolation changes.
Due to how the timer migration algorithm works, any CPU part of the hierarchy can have their global timers pulled by remote CPUs and have to pull remote timers, only skipping pulling remote timers would break the logic.
For this reason, prevent isolated CPUs from pulling remote global timers, but also the other way around: any global timer started on an isolated CPU will run there. This does not break the concept of isolation (global timers don't come from outside the CPU) and, if considered inappropriate, can usually be mitigated with other isolation techniques (e.g. IRQ pinning).This effect was noticed on a 128 cores machine running oslat on the isolated cores (1-31,33-63,65-95,97-127). The tool monopolises CPUs, and the CPU with lowest count in a timer migration hierarchy (here 1 and 65) appears as always active and continuously pulls global timers, from the housekeeping CPUs. This ends up moving driver work (e.g. delayed work) to isolated CPUs and causes latency spikes:
before the change:
# oslat -c 1-31,33-63,65-95,97-127 -D 62s ... Maximum: 1203 10 3 4 ... 5 (us)after the change:
# oslat -c 1-31,33-63,65-95,97-127 -D 62s ... Maximum: 10 4 3 4 3 ... 5 (us)The same behaviour was observed on a machine with as few as 20 cores / 40 threads with isocpus set to: 1-9,11-39 with rtla-osnoise-top.
The first 5 patches are preparatory work to change the concept of online/offline to available/unavailable, keep track of those in a separate cpumask cleanup the setting/clearing functions and change a function name in cpuset code.
Patch 6 and 7 adapt isolation and cpuset to prevent domain isolated and
nohz_fullfrom covering all CPUs not leaving any housekeeping one. This can lead to problems with the changes introduced in this series because no CPU would remain to handle global timers.Quoted 4 lines
Patch 9 extends the unavailable status to domain isolated CPUs, which is the main contribution of the series.
This series is equivalent to v13 but rebased on v6.18-rc2.
Thomas,
This patch series have undergone multiple round of reviews. Do you think it is good enough to be merged into tip?
It does contain some cpuset code, but most of the changes are in the timer code. So I think it is better to go through the tip tree. It does have some minor conflicts with the current for-6.19 branch of the cgroup tree, but it can be easily resolved during merge.
Quoted 7 lines
Quoted 5 lines
Quoted 1 line
What do you think?
Just wait a little, I realize I made a buggy suggestion to Gabriele and a detail needs to be fixed.
My bad...
OK, I thought you were OK with the timer changes.
I was ok until...just a few days ago, and I should have written about it right away but you know, being wrong is a process that takes time :o)
Quoted 1 line
I guess Gabriele will have to send out a new version to address your finding.
Right.
Thanks!
Signature · 2 lines
Frederic Weisbecker SUSE Labs