Re: [Xen-devel] [PATCH] credit2: Add more debugging

[George Dunlap <george.dunlap@xxxxxxxxxx>]

The purpose of this was to catch more information next time you see this
bug in the automated testing; do you run it with debug=y?

If so, I'll change it to #ifndef NDEBUG ... #else BUG_ON() ... #endif

 -George

On Thu, 2011-03-03 at 16:38 +0000, Ian Jackson wrote:
> George Dunlap writes ("[Xen-devel] [PATCH] credit2: Add more debugging"):
> > Signed-off-by: George Dunlap <george.dunlap@xxxxxxxxxxxxx>
> ...
> > +    /* DEBUG */
> 
> Shouldn't this be in an #ifdef or something rather than just marked
> with a comment ?
> 
> > +    if ( !is_idle_vcpu(scurr->vcpu) && scurr->rqd != rqd)
> > +    {
> > +        int other_rqi=-1, this_rqi;
> > +
> > +        this_rqi=c2r(ops, cpu);
> > +
> > +        if ( scurr->rqd )
> > +        {
> > +            int rq;
> > +            other_rqi=-2;
> > +            for_each_cpu_mask(rq, CSCHED_PRIV(ops)->active_queues)
> > +            {
> > +                if ( scurr->rqd == &CSCHED_PRIV(ops)->rqd[rq] )
> > +                {
> > +                    other_rqi = rq;
> > +                    break;
> > +                }
> > +            }
> > +        }
> > +        printk("%s: pcpu %d rq %d, but scurr d%dv%d assigned to pcpu %d rq 
> > %d!\n",
> > +               __func__,
> > +               cpu, this_rqi,
> > +               scurr->vcpu->domain->domain_id, scurr->vcpu->vcpu_id,
> > +               scurr->vcpu->processor, other_rqi);
> > +    }
> >      BUG_ON(!is_idle_vcpu(scurr->vcpu) && scurr->rqd != rqd);
> 
> And in general this seems like a somewhat odd construct.
> 
> Ian.



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