Re: Re: Still struggling with HVM: tx timeouts on emulated nics

[Stefan Bader <stefan.bader@canonical.com>]

On 30.09.2011 16:09, Stefano Stabellini wrote:
> @@ -270,6 +271,18 @@ ret_t do_physdev_op(int cmd, XEN_GUEST_H
>          if ( !is_hvm_domain(v->domain) ||
>               domain_pirq_to_irq(v->domain, eoi.irq) > 0 )
>              pirq_guest_eoi(pirq);
> +        if ( is_hvm_domain(v->domain) &&
> +                domain_pirq_to_emuirq(v->domain, eoi.irq) > 0 )
> +        {
> +            struct hvm_irq *hvm_irq = &v->domain->arch.hvm_domain.irq;
> +            int gsi = domain_pirq_to_emuirq(v->domain, eoi.irq);
> +
> +            /* if this is a level irq and count > 0, send another
> +             * notification */ 
> +            if ( gsi >= NR_ISAIRQS /* ISA irqs are edge triggered */
> +                    && hvm_irq->gsi_assert_count[gsi] )
> +                send_guest_pirq(v->domain, pirq);
> +        }
>          spin_unlock(&v->domain->event_lock);
>          ret = 0;
>          break;

This hunk looks substantially different from my 4.1.1 based code. There is no
spin_lock acquired. Not sure that could be a reason for the different behaviour,
too. I'll add that spinlock too.

    case PHYSDEVOP_eoi: {
        struct physdev_eoi eoi;
        ret = -EFAULT;
        if ( copy_from_guest(&eoi, arg, 1) != 0 )
            break;
        ret = -EINVAL;
        if ( eoi.irq >= v->domain->nr_pirqs )
            break;
        if ( v->domain->arch.pirq_eoi_map )
            evtchn_unmask(v->domain->pirq_to_evtchn[eoi.irq]);
        if ( !is_hvm_domain(v->domain) ||
             domain_pirq_to_irq(v->domain, eoi.irq) > 0 )
            ret = pirq_guest_eoi(v->domain, eoi.irq);
        else
            ret = 0;
        break;
    }