search for: kick_cpu

...possibilities here, considering that in undercommit cases we should not exceed HEAD_SPIN_THRESHOLD, 1. the looping vcpu in pv_head_spin_check() should do halt() considering that we have done enough spinning (more than typical lock-hold time), and hence we are in potential overcommit. 2. multiplex kick_cpu to do directed yield in qspinlock case. But this may result in some ping ponging?

...possibilities here, considering that in undercommit cases we should not exceed HEAD_SPIN_THRESHOLD, 1. the looping vcpu in pv_head_spin_check() should do halt() considering that we have done enough spinning (more than typical lock-hold time), and hence we are in potential overcommit. 2. multiplex kick_cpu to do directed yield in qspinlock case. But this may result in some ping ponging?

This series forms the kvm host part of paravirtual spinlock based against kvm tree. Please refer to https://lkml.org/lkml/2013/8/9/265 for kvm guest and Xen, x86 part merged to -tip spinlocks. Please note that: kvm uapi: Add KICK_CPU and PV_UNHALT definition to uapi is a common patch for both guest and host. Changes since V12: fold the patch 3 into patch 2 for bisection. (Eric Northup) Raghavendra K T (3): kvm uapi: Add KICK_CPU and PV_UNHALT definition to uapi kvm hypervisor: Simplify kvm_for_each_vcpu with kvm_i...

This series forms the kvm host part of paravirtual spinlock based against kvm tree. Please refer to https://lkml.org/lkml/2013/8/9/265 for kvm guest and Xen, x86 part merged to -tip spinlocks. Please note that: kvm uapi: Add KICK_CPU and PV_UNHALT definition to uapi is a common patch for both guest and host. Changes since V12: fold the patch 3 into patch 2 for bisection. (Eric Northup) Raghavendra K T (3): kvm uapi: Add KICK_CPU and PV_UNHALT definition to uapi kvm hypervisor: Simplify kvm_for_each_vcpu with kvm_i...

...ghavendra K T wrote: > > This series forms the kvm host part of paravirtual spinlock > based against kvm tree. > > Please refer to https://lkml.org/lkml/2013/8/9/265 for > kvm guest and Xen, x86 part merged to -tip spinlocks. > > Please note that: > kvm uapi: Add KICK_CPU and PV_UNHALT definition to uapi is a common patch > for both guest and host. > Thanks, applied. The patchset is not against kvm.git queue though, so I had to fix one minor conflict manually. > Changes since V12: > fold the patch 3 into patch 2 for bisection. (Eric Northup) >...

Il 12/03/2014 19:54, Waiman Long ha scritto: > @@ -807,8 +889,13 @@ void __init kvm_spinlock_init(void) > if (!kvm_para_has_feature(KVM_FEATURE_PV_UNHALT)) > return; > > +#ifdef CONFIG_QUEUE_SPINLOCK > + pv_lock_ops.kick_cpu = kvm_kick_cpu_type; > + pv_lock_ops.hibernate = kvm_hibernate; > +#else > pv_lock_ops.lock_spinning = PV_CALLEE_SAVE(kvm_lock_spinning); > pv_lock_ops.unlock_kick = kvm_unlock_kick; > +#endif This should also disable the unfair path. Paolo

Il 12/03/2014 19:54, Waiman Long ha scritto: > @@ -807,8 +889,13 @@ void __init kvm_spinlock_init(void) > if (!kvm_para_has_feature(KVM_FEATURE_PV_UNHALT)) > return; > > +#ifdef CONFIG_QUEUE_SPINLOCK > + pv_lock_ops.kick_cpu = kvm_kick_cpu_type; > + pv_lock_ops.hibernate = kvm_hibernate; > +#else > pv_lock_ops.lock_spinning = PV_CALLEE_SAVE(kvm_lock_spinning); > pv_lock_ops.unlock_kick = kvm_unlock_kick; > +#endif This should also disable the unfair path. Paolo

...atic_key_slow_inc(&paravirt_unfairlocks_enabled); > + printk(KERN_INFO "KVM setup unfair spinlock\n"); > + > + return 0; > +} > +early_initcall(kvm_unfair_locks_init_jump); > +#endif > I think this should apply to all paravirt implementations, unless pv_lock_ops.kick_cpu != NULL. Paolo

...atic_key_slow_inc(&paravirt_unfairlocks_enabled); > + printk(KERN_INFO "KVM setup unfair spinlock\n"); > + > + return 0; > +} > +early_initcall(kvm_unfair_locks_init_jump); > +#endif > I think this should apply to all paravirt implementations, unless pv_lock_ops.kick_cpu != NULL. Paolo

...include/asm/paravirt.h +++ b/arch/x86/include/asm/paravirt.h @@ -711,7 +711,12 @@ static inline void __set_fixmap(unsigned /* enum fixed_addresses */ idx, } #if defined(CONFIG_SMP) && defined(CONFIG_PARAVIRT_SPINLOCKS) - +#ifdef CONFIG_QUEUE_SPINLOCK +static __always_inline void __queue_kick_cpu(int cpu, enum pv_kick_type type) +{ + PVOP_VCALL2(pv_lock_ops.kick_cpu, cpu, type); +} +#else static __always_inline void __ticket_lock_spinning(struct arch_spinlock *lock, __ticket_t ticket) { @@ -723,7 +728,7 @@ static __always_inline void __ticket_unlock_kick(struct arch_spinlock *lock...

This series forms the kvm host part of paravirtual spinlock based against kvm tree. Please refer https://lkml.org/lkml/2013/8/6/178 for kvm guest part of the series. Please note that: kvm uapi: Add KICK_CPU and PV_UNHALT definition to uapi is a common patch for both guest and host. Srivatsa Vaddagiri (1): kvm hypervisor : Add a hypercall to KVM hypervisor to support pv-ticketlocks Raghavendra K T (4): kvm uapi: Add KICK_CPU and PV_UNHALT definition to uapi kvm : Fold pv_unhalt flag into GET...

This series forms the kvm host part of paravirtual spinlock based against kvm tree. Please refer https://lkml.org/lkml/2013/8/6/178 for kvm guest part of the series. Please note that: kvm uapi: Add KICK_CPU and PV_UNHALT definition to uapi is a common patch for both guest and host. Srivatsa Vaddagiri (1): kvm hypervisor : Add a hypercall to KVM hypervisor to support pv-ticketlocks Raghavendra K T (4): kvm uapi: Add KICK_CPU and PV_UNHALT definition to uapi kvm : Fold pv_unhalt flag into GET...

...6/include/asm/paravirt.h +++ b/arch/x86/include/asm/paravirt.h @@ -711,7 +711,17 @@ static inline void __set_fixmap(unsigned /* enum fixed_addresses */ idx, } #if defined(CONFIG_SMP) && defined(CONFIG_PARAVIRT_SPINLOCKS) +#ifdef CONFIG_QUEUE_SPINLOCK +static __always_inline void __queue_kick_cpu(int cpu, enum pv_kick_type type) +{ + PVOP_VCALL2(pv_lock_ops.kick_cpu, cpu, type); +} +static __always_inline void __queue_hibernate(void) +{ + PVOP_VCALL0(pv_lock_ops.hibernate); +} +#else static __always_inline void __ticket_lock_spinning(struct arch_spinlock *lock, __ticket_t ticket)...

...h/x86/include/asm/paravirt.h +++ b/arch/x86/include/asm/paravirt.h @@ -712,6 +712,24 @@ static inline void __set_fixmap(unsigned /* enum fixed_addresses */ idx, #if defined(CONFIG_SMP) && defined(CONFIG_PARAVIRT_SPINLOCKS) +#ifdef CONFIG_QUEUE_SPINLOCK + +static __always_inline void pv_kick_cpu(int cpu) +{ + PVOP_VCALLEE1(pv_lock_ops.kick_cpu, cpu); +} + +static __always_inline void pv_lockwait(u8 *lockbyte) +{ + PVOP_VCALLEE1(pv_lock_ops.lockwait, lockbyte); +} + +static __always_inline void pv_lockstat(enum pv_lock_stats type) +{ + PVOP_VCALLEE1(pv_lock_ops.lockstat, type); +} + +#else...

...h/x86/include/asm/paravirt.h +++ b/arch/x86/include/asm/paravirt.h @@ -712,6 +712,24 @@ static inline void __set_fixmap(unsigned /* enum fixed_addresses */ idx, #if defined(CONFIG_SMP) && defined(CONFIG_PARAVIRT_SPINLOCKS) +#ifdef CONFIG_QUEUE_SPINLOCK + +static __always_inline void pv_kick_cpu(int cpu) +{ + PVOP_VCALLEE1(pv_lock_ops.kick_cpu, cpu); +} + +static __always_inline void pv_lockwait(u8 *lockbyte) +{ + PVOP_VCALLEE1(pv_lock_ops.lockwait, lockbyte); +} + +static __always_inline void pv_lockstat(enum pv_lock_stats type) +{ + PVOP_VCALLEE1(pv_lock_ops.lockstat, type); +} + +#else...

...OCKED_OFFSET) > + > +struct pv_node { > + struct mcs_spinlock mcs; > + struct mcs_spinlock __offset[3]; > + int cpu, head; > +}; I am wondering why you need the separate cpu and head variables. I thought one will be enough here. The wait code put the cpu number in head, the the kick_cpu code kick the one in cpu which is just the cpu # of the tail. > + > +#define INVALID_HEAD -1 > +#define NO_HEAD nr_cpu_ids > + I think it is better to use a constant like -2 for NO_HEAD instead of an external variable. > +void __pv_init_node(struct mcs_spinlock *node) > +{ &...

...OCKED_OFFSET) > + > +struct pv_node { > + struct mcs_spinlock mcs; > + struct mcs_spinlock __offset[3]; > + int cpu, head; > +}; I am wondering why you need the separate cpu and head variables. I thought one will be enough here. The wait code put the cpu number in head, the the kick_cpu code kick the one in cpu which is just the cpu # of the tail. > + > +#define INVALID_HEAD -1 > +#define NO_HEAD nr_cpu_ids > + I think it is better to use a constant like -2 for NO_HEAD instead of an external variable. > +void __pv_init_node(struct mcs_spinlock *node) > +{ &...

...hat in undercommit cases we should not exceed > HEAD_SPIN_THRESHOLD, > > 1. the looping vcpu in pv_head_spin_check() should do halt() > considering that we have done enough spinning (more than typical > lock-hold time), and hence we are in potential overcommit. > > 2. multiplex kick_cpu to do directed yield in qspinlock case. > But this may result in some ping ponging? Actually, I think the qspinlock can work roughly the same as the pvticketlock, using the same lock_spinning and unlock_lock hooks. The x86-specific codepath can use bit 1 in the ->wait byte as "I have...

...aolo Bonzini wrote: > Il 12/03/2014 19:54, Waiman Long ha scritto: >> @@ -807,8 +889,13 @@ void __init kvm_spinlock_init(void) >> if (!kvm_para_has_feature(KVM_FEATURE_PV_UNHALT)) >> return; >> >> +#ifdef CONFIG_QUEUE_SPINLOCK >> + pv_lock_ops.kick_cpu = kvm_kick_cpu_type; >> + pv_lock_ops.hibernate = kvm_hibernate; >> +#else >> pv_lock_ops.lock_spinning = PV_CALLEE_SAVE(kvm_lock_spinning); >> pv_lock_ops.unlock_kick = kvm_unlock_kick; >> +#endif > > This should also disable the unfair path. > &...

...+++++++++++++++++++++++++++++++++ kernel/Kconfig.locks | 2 +- 2 files changed, 55 insertions(+), 1 deletions(-) diff --git a/arch/x86/kernel/kvm.c b/arch/x86/kernel/kvm.c index f318e78..3ddc436 100644 --- a/arch/x86/kernel/kvm.c +++ b/arch/x86/kernel/kvm.c @@ -568,6 +568,7 @@ static void kvm_kick_cpu(int cpu) kvm_hypercall2(KVM_HC_KICK_CPU, flags, apicid); } +#ifndef CONFIG_QUEUE_SPINLOCK enum kvm_contention_stat { TAKEN_SLOW, TAKEN_SLOW_PICKUP, @@ -795,6 +796,55 @@ static void kvm_unlock_kick(struct arch_spinlock *lock, __ticket_t ticket) } } } +#else /* !CONFIG_QUEUE_SPINLOCK...