Re: [Xen-devel] [PATCH RFC 1/3] x86/vpt: execute callbacks for masked interrupts

[Roger Pau Monné <roger.pau@xxxxxxxxxx>]

On Mon, Feb 26, 2018 at 06:04:51AM -0700, Jan Beulich wrote:
> >>> On 26.02.18 at 13:48, <roger.pau@xxxxxxxxxx> wrote:
> > On Mon, Feb 26, 2018 at 12:35:54PM +0000, Wei Liu wrote:
> >> On Fri, Feb 23, 2018 at 01:27:41PM +0000, Roger Pau Monne wrote:
> >> >  int pt_update_irq(struct vcpu *v)
> >> >  {
> >> >      struct list_head *head = &v->arch.hvm_vcpu.tm_list;
> >> > +    LIST_HEAD(purged);
> >> 
> >> to_purge?
> > 
> > My point is that they have already been purged from the pt->list, but
> > I really don't have a preference.
> > 
> >> >      struct periodic_time *pt, *temp, *earliest_pt;
> >> >      uint64_t max_lag;
> >> >      int irq, is_lapic, pt_vector;
> >> > @@ -267,7 +289,10 @@ int pt_update_irq(struct vcpu *v)
> >> >              {
> >> >                  /* suspend timer emulation */
> >> >                  list_del(&pt->list);
> >> > -                pt->on_list = 0;
> >> > +                if ( pt->cb )
> >> > +                    list_add(&pt->list, &purged);
> >> > +                else
> >> > +                    pt->on_list = 0;
> >> >              }
> >> >              else
> >> >              {
> >> > @@ -283,6 +308,7 @@ int pt_update_irq(struct vcpu *v)
> >> >      if ( earliest_pt == NULL )
> >> >      {
> >> >          spin_unlock(&v->arch.hvm_vcpu.tm_lock);
> >> > +        execute_callbacks(v, &purged);
> >> 
> >> It would be better to check if the list is not empty before calling the
> >> function to avoid the extra lock / unlock.
> > 
> > The lock is also protecting the 'purged' list, so I think that for
> > consistency the lock needs to be held before accessing it.
> 
> But that's a local list, isn't it? No-one else can access it.

destroy_periodic_time can still remove items from this list, if a
timer that's on the 'purged' list is destroyed between added to the
list and executing the callback.

Roger.

_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxxxxxxxxx
https://lists.xenproject.org/mailman/listinfo/xen-devel