Re: [Xen-devel] [PATCH v2] xen/spinlock: Don't use pvqspinlock if only 1 vCPU

2018-07-31 Thread Boris Ostrovsky
On 07/19/2018 05:39 PM, Waiman Long wrote: > 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. > > The xen_pvspin veriable is

Re: [Xen-devel] [PATCH v2] xen/spinlock: Don't use pvqspinlock if only 1 vCPU

2018-07-23 Thread Waiman Long
On 07/23/2018 12:42 AM, Davidlohr Bueso wrote: > On Mon, 23 Jul 2018, Wanpeng Li wrote: > >> On Fri, 20 Jul 2018 at 06:03, Waiman Long wrote: >>> >>> On 07/19/2018 05:54 PM, Davidlohr Bueso wrote: >>> > On Thu, 19 Jul 2018, Waiman Long wrote: >>> > >>> >> On a VM with only 1 vCPU, the locking

Re: [Xen-devel] [PATCH v2] xen/spinlock: Don't use pvqspinlock if only 1 vCPU

2018-07-23 Thread Waiman Long
On 07/22/2018 11:31 PM, Wanpeng Li wrote: > On Fri, 20 Jul 2018 at 06:03, Waiman Long wrote: >> On 07/19/2018 05:54 PM, Davidlohr Bueso wrote: >>> On Thu, 19 Jul 2018, Waiman Long wrote: >>> On a VM with only 1 vCPU, the locking fast paths will always be successful. In this case, there

Re: [Xen-devel] [PATCH v2] xen/spinlock: Don't use pvqspinlock if only 1 vCPU

2018-07-22 Thread Davidlohr Bueso
On Sun, 22 Jul 2018, Davidlohr Bueso wrote: On Mon, 23 Jul 2018, Wanpeng Li wrote: On Fri, 20 Jul 2018 at 06:03, Waiman Long wrote: On 07/19/2018 05:54 PM, Davidlohr Bueso wrote: On Thu, 19 Jul 2018, Waiman Long wrote: On a VM with only 1 vCPU, the locking fast paths will always be

Re: [Xen-devel] [PATCH v2] xen/spinlock: Don't use pvqspinlock if only 1 vCPU

2018-07-22 Thread Davidlohr Bueso
On Mon, 23 Jul 2018, Wanpeng Li wrote: On Fri, 20 Jul 2018 at 06:03, Waiman Long wrote: On 07/19/2018 05:54 PM, Davidlohr Bueso wrote: > On Thu, 19 Jul 2018, Waiman Long wrote: > >> On a VM with only 1 vCPU, the locking fast paths will always be >> successful. In this case, there is no need

Re: [Xen-devel] [PATCH v2] xen/spinlock: Don't use pvqspinlock if only 1 vCPU

2018-07-22 Thread Wanpeng Li
On Fri, 20 Jul 2018 at 06:03, Waiman Long wrote: > > On 07/19/2018 05:54 PM, Davidlohr Bueso wrote: > > On Thu, 19 Jul 2018, Waiman Long wrote: > > > >> 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 > >>

Re: [Xen-devel] [PATCH v2] xen/spinlock: Don't use pvqspinlock if only 1 vCPU

2018-07-19 Thread Boris Ostrovsky
On 07/19/2018 05:39 PM, Waiman Long wrote: > 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. > > The xen_pvspin veriable is

Re: [Xen-devel] [PATCH v2] xen/spinlock: Don't use pvqspinlock if only 1 vCPU

2018-07-19 Thread Waiman Long
On 07/19/2018 05:54 PM, Davidlohr Bueso wrote: > On Thu, 19 Jul 2018, Waiman Long wrote: > >> 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

Re: [Xen-devel] [PATCH v2] xen/spinlock: Don't use pvqspinlock if only 1 vCPU

2018-07-19 Thread Davidlohr Bueso
On Thu, 19 Jul 2018, Waiman Long wrote: 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. The xen_pvspin veriable is also

[Xen-devel] [PATCH v2] xen/spinlock: Don't use pvqspinlock if only 1 vCPU

2018-07-19 Thread 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. The xen_pvspin veriable is also turned off in this 1 vCPU case to eliminate