[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Xen-devel] [PATCH v7 06/11] x86, paravirt: Add interface to support kvm/xen vcpu preempted check
- To: Pan Xinhui <xinhui@xxxxxxxxxxxxxxxxxx>
- From: Peter Zijlstra <peterz@xxxxxxxxxxxxx>
- Date: Wed, 16 Nov 2016 11:23:55 +0100
- Cc: kvm@xxxxxxxxxxxxxxx, rkrcmar@xxxxxxxxxx, benh@xxxxxxxxxxxxxxxxxxx, bsingharora@xxxxxxxxx, will.deacon@xxxxxxx, virtualization@xxxxxxxxxxxxxxxxxxxxxxxxxx, paulus@xxxxxxxxx, kernellwp@xxxxxxxxx, linux-s390@xxxxxxxxxxxxxxx, dave@xxxxxxxxxxxx, mpe@xxxxxxxxxxxxxx, x86@xxxxxxxxxx, borntraeger@xxxxxxxxxx, mingo@xxxxxxxxxx, Pan Xinhui <xinhui.pan@xxxxxxxxxxxxxxxxxx>, xen-devel@xxxxxxxxxxxxxxxxxxxx, paulmck@xxxxxxxxxxxxxxxxxx, boqun.feng@xxxxxxxxx, jgross@xxxxxxxx, linux-kernel@xxxxxxxxxxxxxxx, David.Laight@xxxxxxxxxx, xen-devel-request@xxxxxxxxxxxxxxxxxxxx, pbonzini@xxxxxxxxxx, linuxppc-dev@xxxxxxxxxxxxxxxx
- Delivery-date: Wed, 16 Nov 2016 10:24:39 +0000
- List-id: Xen developer discussion <xen-devel.lists.xen.org>
On Wed, Nov 16, 2016 at 12:19:09PM +0800, Pan Xinhui wrote:
> Hi, Peter.
> I think we can avoid a function call in a simpler way. How about below
>
> static inline bool vcpu_is_preempted(int cpu)
> {
> /* only set in pv case*/
> if (pv_lock_ops.vcpu_is_preempted)
> return pv_lock_ops.vcpu_is_preempted(cpu);
> return false;
> }
That is still more expensive. It needs to do an actual load and makes it
hard to predict the branch, you'd have to actually wait for the load to
complete etc.
Also, it generates more code.
Paravirt muck should strive to be as cheap as possible when ran on
native hardware.
_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxx
https://lists.xen.org/xen-devel
|
