[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] [Xen-changelog] [xen stable-4.5] xen: Have schedulers revise initial placement
commit c421378a8d14c811e5467d535bc71adc0328a316 Author: George Dunlap <george.dunlap@xxxxxxxxxx> AuthorDate: Fri Aug 5 14:07:27 2016 +0200 Commit: Jan Beulich <jbeulich@xxxxxxxx> CommitDate: Fri Aug 5 14:07:27 2016 +0200 xen: Have schedulers revise initial placement The generic domain creation logic in xen/common/domctl.c:default_vcpu0_location() attempts to try to do initial placement load-balancing by placing vcpu 0 on the least-busy non-primary hyperthread available. Unfortunately, the logic can end up picking a pcpu that's not in the online mask. When this is passed to a scheduler such which assumes that the initial assignment is valid, it causes a null pointer dereference looking up the runqueue. Furthermore, this initial placement doesn't take into account hard or soft affinity, or any scheduler-specific knowledge (such as historic runqueue load, as in credit2). To solve this, when inserting a vcpu, always call the per-scheduler "pick" function to revise the initial placement. This will automatically take all knowledge the scheduler has into account. csched2_cpu_pick ASSERTs that the vcpu's pcpu scheduler lock has been taken. Grab and release the lock to minimize time spend with irqs disabled. Signed-off-by: George Dunlap <george.dunlap@xxxxxxxxxx> Reviewed-by: Meng Xu <mengxu@xxxxxxxxxxxxx> Reviwed-by: Dario Faggioli <dario.faggioli@xxxxxxxxxx> master commit: 9f358ddd69463fa8fb65cf67beb5f6f0d3350e32 master date: 2016-07-26 10:42:49 +0100 --- xen/common/sched_credit.c | 3 +++ xen/common/sched_credit2.c | 10 +++++++++- xen/common/sched_rt.c | 5 +++++ 3 files changed, 17 insertions(+), 1 deletion(-) diff --git a/xen/common/sched_credit.c b/xen/common/sched_credit.c index 54fd830..2e11aa2 100644 --- a/xen/common/sched_credit.c +++ b/xen/common/sched_credit.c @@ -892,6 +892,9 @@ csched_vcpu_insert(const struct scheduler *ops, struct vcpu *vc) spinlock_t *lock; unsigned long flags; + /* This is safe because vc isn't yet being scheduled */ + vc->processor = csched_cpu_pick(ops, vc); + lock = vcpu_schedule_lock_irqsave(vc, &flags); if ( !__vcpu_on_runq(svc) && vcpu_runnable(vc) && !vc->is_running ) diff --git a/xen/common/sched_credit2.c b/xen/common/sched_credit2.c index db9e1c4..7a81db5 100644 --- a/xen/common/sched_credit2.c +++ b/xen/common/sched_credit2.c @@ -269,6 +269,7 @@ struct csched2_dom { uint16_t nr_vcpus; }; +static int csched2_cpu_pick(const struct scheduler *ops, struct vcpu *vc); /* * Time-to-credit, credit-to-time. @@ -874,9 +875,16 @@ csched2_vcpu_insert(const struct scheduler *ops, struct vcpu *vc) /* FIXME: Do we need the private lock here? */ list_add_tail(&svc->sdom_elem, &svc->sdom->vcpu); - /* Add vcpu to runqueue of initial processor */ + /* csched2_cpu_pick() expects the pcpu lock to be held */ + lock = vcpu_schedule_lock_irq(vc); + + vc->processor = csched2_cpu_pick(ops, vc); + + spin_unlock_irq(lock); + lock = vcpu_schedule_lock_irq(vc); + /* Add vcpu to runqueue of initial processor */ runq_assign(ops, vc); vcpu_schedule_unlock_irq(lock, vc); diff --git a/xen/common/sched_rt.c b/xen/common/sched_rt.c index a77e808..bb972fa 100644 --- a/xen/common/sched_rt.c +++ b/xen/common/sched_rt.c @@ -169,6 +169,8 @@ struct rt_dom { struct domain *dom; /* pointer to upper domain */ }; +static int rt_cpu_pick(const struct scheduler *ops, struct vcpu *vc); + /* * Useful inline functions */ @@ -554,6 +556,9 @@ rt_vcpu_insert(const struct scheduler *ops, struct vcpu *vc) if ( is_idle_vcpu(vc) ) return; + /* This is safe because vc isn't yet being scheduled */ + vc->processor = rt_cpu_pick(ops, vc); + lock = vcpu_schedule_lock_irq(vc); now = NOW(); -- generated by git-patchbot for /home/xen/git/xen.git#stable-4.5 _______________________________________________ Xen-changelog mailing list Xen-changelog@xxxxxxxxxxxxx https://lists.xenproject.org/xen-changelog
|
Lists.xenproject.org is hosted with RackSpace, monitoring our |