Re: [PATCH 7/7 v3] KVM: PPC: BOOKE: Emulate debug registers and exception

[Scott Wood <scottwood@freescale.com>]

On Tue, 2014-08-12 at 02:36 -0500, Bhushan Bharat-R65777 wrote:
> 
> > -----Original Message-----
> > From: Wood Scott-B07421
> > Sent: Tuesday, August 12, 2014 5:30 AM
> > To: Bhushan Bharat-R65777
> > Cc: agraf@suse.de; kvm-ppc@vger.kernel.org; kvm@vger.kernel.org; Yoder Stuart-
> > B08248
> > Subject: Re: [PATCH 7/7 v3] KVM: PPC: BOOKE: Emulate debug registers and
> > exception
> > 
> > On Wed, 2014-08-06 at 12:08 +0530, Bharat Bhushan wrote:
> > > @@ -1249,6 +1284,7 @@ int kvmppc_subarch_vcpu_init(struct kvm_vcpu *vcpu)
> > >  	setup_timer(&vcpu->arch.wdt_timer, kvmppc_watchdog_func,
> > >  		    (unsigned long)vcpu);
> > >
> > > +	kvmppc_clear_dbsr();
> > >  	return 0;
> > 
> > This could use a comment for why we're doing this.  Also, I'm a bit uneasy about
> > clearing the whole DBSR here, where we haven't yet switched the debug registers
> > to guest context.
> 
> I think we wanted MRR to not cause debug event to guest, So should we only clear MRR ?
> 
> > It shouldn't actually matter except for deferred debug
> > exceptions which are not actually useful (in fact e6500 removed support for
> > them),
> 
> Exactly, that's why I was clearing complete DBSR. Probably we can have a comment
> " Do not let previously set debug events visible to guest. As deferred debug events
>   are not supported, so it is ok to clear complete DBSR.
> " 

This would be affecting host debugging of the host, not guest debugging
of the guest.  Still I don't think it's a huge deal, but clearing only
MRR would be cleaner.

-Scott