From mboxrd@z Thu Jan  1 00:00:00 1970
From: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Date: Sat, 10 Oct 2009 23:14:17 +0000
Subject: Re: [PATCH 16/27] Add desktop PowerPC specific emulation
Message-Id: <1255216457.2192.16.camel@pasglop>
List-Id: <kvm-ppc.vger.kernel.org>
References: <1254212303-8737-17-git-send-email-agraf@suse.de>
In-Reply-To: <1254212303-8737-17-git-send-email-agraf@suse.de>
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit
To: kvm-ppc@vger.kernel.org

On Fri, 2009-10-09 at 13:57 -0700, Hollis Blanchard wrote:

> > +#ifdef CONFIG_PPC64
> > +#ifdef DEBUG_EMUL
> > +	printk(KERN_INFO "mtDEC: %x\n", vcpu->arch.dec);
> > +#endif
> > +	/* POWER4+ triggers a dec interrupt if the value is < 0 */
> > +	if (vcpu->arch.dec & 0x80000000) {
> > +		del_timer(&vcpu->arch.dec_timer);
> > +		kvmppc_core_queue_dec(vcpu);
> > +	}
> > +	else if (true) {
> > +#else
> >  	if (vcpu->arch.tcr & TCR_DIE) {
> > +#endif
> >  		/* The decrementer ticks at the same rate as the timebase, so
> >  		 * that's how we convert the guest DEC value to the number of
> >  		 * host ticks. */
> 
> Ifdefs through the middle of control syntax makes my head hurt. :)

Note that your original BookE DEC emulation looks fishy :-)

I may have missed something in your code... but I don't think it
emulates the expected HW behaviour:

Basically, when the BookE DEC hits the 1 -> 0 transition it latches an
event in TSE:DIS always, whether TCR:DIE is set or not. If DIE is not
set, the interrupt is sent. It will then stop counting if auto-reload
isn't enabled.

That means that if TSR:DIS is set from a previous event while TCR:DIE is
clear, then setting TCR:DIE (with MSR:EE enabled of course) will trigger
a DEC interrupt on BookE.

The BookE DEC interrupt is basically a level sensitive thing sourced
from to (TSR:DIS && TCR:DIE), and TSR:DIS remains set until explicitely
cleared.

Cheers,
Ben.