xen/spinlock: Don't use pvqspinlock if only 1 vCPU
Subsystems (2)
PATCH 1/1 xen/spinlock: Don't use pvqspinlock if only 1 vCPU
On a VM with only 1 vCPU, the locking fast paths will always be successful. In this case, there is no need to use the the PV qspinlock code which has higher overhead on the unlock side than the native qspinlock code.
--- arch/x86/xen/spinlock.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-)
arch/x86/xen/spinlock.c +2 -1
@@ -130,7 +130,8 @@ void xen_uninit_lock_cpu(int cpu)void __init xen_init_spinlocks (void ) { -if (!xen_pvspin) { +/* Don't need to use pvqspinlock code if there is only 1 vCPU. */ +if (!xen_pvspin ||num_possible_cpus () ==1 ) {printk (KERN_DEBUG "xen: PV spinlocks disabled\n" );return ; }
Signature · 1 line
1.8.3.1
Boris Ostrovsky
reviewerQuoted 15 lines · Waiman Long
On a VM with only 1 vCPU, the locking fast paths will always be successful. In this case, there is no need to use the the PV qspinlock code which has higher overhead on the unlock side than the native qspinlock code.
Signed-off-by Waiman Long <[email protected]>--- arch/x86/xen/spinlock.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-)arch/x86/xen/spinlock.c +0
-0
arch/x86/xen/spinlock.c void xen_uninit_lock_cpu(int cpu) · line 137
Quoted 9 lines
arch/x86/xen/spinlock.c +2
-1@@ -130,7 +130,8 @@ void xen_uninit_lock_cpu(int cpu)void __init xen_init_spinlocks (void ) { -if (!xen_pvspin) { +/* Don't need to use pvqspinlock code if there is only 1 vCPU. */ +if (!xen_pvspin ||num_possible_cpus () ==1 ) {printk (KERN_DEBUG "xen: PV spinlocks disabled\n" );return ; }
I think we need to set xen_pvspin to false for such configurations. Notice that xen_init_lock_cpu() will try to perform some additional pvspinlock initializations.
-boris
_______________________________________________
Xen-devel mailing list
[email protected]
https://lists.xenproject.org/mailman/listinfo/xen-devel
Boris Ostrovsky
reviewerQuoted 15 lines · Waiman Long
On a VM with only 1 vCPU, the locking fast paths will always be successful. In this case, there is no need to use the the PV qspinlock code which has higher overhead on the unlock side than the native qspinlock code.
Signed-off-by Waiman Long <[email protected]>--- arch/x86/xen/spinlock.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-)arch/x86/xen/spinlock.c +0
-0
arch/x86/xen/spinlock.c void xen_uninit_lock_cpu(int cpu) · line 137
Quoted 9 lines
arch/x86/xen/spinlock.c +2
-1@@ -130,7 +130,8 @@ void xen_uninit_lock_cpu(int cpu)void __init xen_init_spinlocks (void ) { -if (!xen_pvspin) { +/* Don't need to use pvqspinlock code if there is only 1 vCPU. */ +if (!xen_pvspin ||num_possible_cpus () ==1 ) {printk (KERN_DEBUG "xen: PV spinlocks disabled\n" );return ; }
I think we need to set xen_pvspin to false for such configurations. Notice that xen_init_lock_cpu() will try to perform some additional pvspinlock initializations.
-boris
Quoted 31 lines · Boris Ostrovsky
Quoted 15 lines · Waiman Long
On a VM with only 1 vCPU, the locking fast paths will always be successful. In this case, there is no need to use the the PV qspinlock code which has higher overhead on the unlock side than the native qspinlock code.
Signed-off-by Waiman Long <[email protected]>--- arch/x86/xen/spinlock.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-)arch/x86/xen/spinlock.c +0
-0arch/x86/xen/spinlock.c void xen_uninit_lock_cpu(int cpu) · line 137
Quoted 9 lines
arch/x86/xen/spinlock.c +2
-1@@ -130,7 +130,8 @@ void xen_uninit_lock_cpu(int cpu)void __init xen_init_spinlocks (void ) { -if (!xen_pvspin) { +/* Don't need to use pvqspinlock code if there is only 1 vCPU. */ +if (!xen_pvspin ||num_possible_cpus () ==1 ) {printk (KERN_DEBUG "xen: PV spinlocks disabled\n" );return ; }I think we need to set
xen_pvspinto false for such configurations. Notice thatxen_init_lock_cpu()will try to perform some additional pvspinlock initializations.
Quoted 1 line
-boris
The other pvqspinlock initialization has no runtime impact other than allocating a bit of extra memory. Anyway, I will revise the patch to disable xen_pvspin under such condition.
-Longman
_______________________________________________
Xen-devel mailing list
[email protected]
https://lists.xenproject.org/mailman/listinfo/xen-devel
Quoted 31 lines · Boris Ostrovsky
Quoted 15 lines · Waiman Long
On a VM with only 1 vCPU, the locking fast paths will always be successful. In this case, there is no need to use the the PV qspinlock code which has higher overhead on the unlock side than the native qspinlock code.
Signed-off-by Waiman Long <[email protected]>--- arch/x86/xen/spinlock.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-)arch/x86/xen/spinlock.c +0
-0arch/x86/xen/spinlock.c void xen_uninit_lock_cpu(int cpu) · line 137
Quoted 9 lines
arch/x86/xen/spinlock.c +2
-1@@ -130,7 +130,8 @@ void xen_uninit_lock_cpu(int cpu)void __init xen_init_spinlocks (void ) { -if (!xen_pvspin) { +/* Don't need to use pvqspinlock code if there is only 1 vCPU. */ +if (!xen_pvspin ||num_possible_cpus () ==1 ) {printk (KERN_DEBUG "xen: PV spinlocks disabled\n" );return ; }I think we need to set
xen_pvspinto false for such configurations. Notice thatxen_init_lock_cpu()will try to perform some additional pvspinlock initializations.
Quoted 1 line
-boris
The other pvqspinlock initialization has no runtime impact other than allocating a bit of extra memory. Anyway, I will revise the patch to disable xen_pvspin under such condition.
-Longman