Re: [XenPPC] [pushed][ppc] cleanup deliver_ee()

[Hollis Blanchard <hollisb@xxxxxxxxxx>]

On Mar 29, 2006, at 5:58 AM, jimix@xxxxxxxxxxxxxx wrote:

> changeset:   9661:f1726cfcf72f
> tag:         tip
> user:        jimix@xxxxxxxxxxxxxxxxxxxxx
> date:        Wed Mar 29 06:57:49 2006 -0500
> summary:     [ppc] cleanup deliver_ee()
>
> diff -r b7c9d7c30cf4 -r f1726cfcf72f xen/arch/ppc/external.c
> --- a/xen/arch/ppc/external.c   Tue Mar 28 10:16:38 2006 -0500
> +++ b/xen/arch/ppc/external.c   Wed Mar 29 06:57:49 2006 -0500
> @@ -53,18 +53,9 @@ void deliver_ee(struct cpu_user_regs *re
>
>      /* trigger exception only if we have a pending irq, we're 
> resuming a guest
>       * (not the hypervisor), the guest is dom0, and it has MSR:EE 
> set. */
> -    if (!event_pending(current)) {
> -        if (current->vcpu_info->evtchn_upcall_pending) {
> -            printk("%s:0x%lx,0x%x\n", __func__,
> -                   current->vcpu_info->evtchn_upcall_pending,
> -                   current->vcpu_info->evtchn_upcall_mask);
> -            for (;;);
> -        }
> -        return;
> -    }
> -    if (regs->msr & MSR_HV) return;
> -    if (current->domain->domain_id != 0) return;
> +    if (!event_pending(current)) return;
>      if (!(regs->msr & MSR_EE)) return;
> +    BUG_ON(regs->msr & MSR_HV);
How is this a bug? deliver_ee() is called every time we return from an 
exception, including when resuming into the hypervisor from a nested 
interrupt.
--
Hollis Blanchard
IBM Linux Technology Center


_______________________________________________
Xen-ppc-devel mailing list
Xen-ppc-devel@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-ppc-devel