From mboxrd@z Thu Jan  1 00:00:00 1970
From: Jan Kiszka <jan.kiszka@siemens.com>
Subject: Re: [patch -rt 2/2] KVM: lapic: mark LAPIC timer handler as irqsafe
Date: Tue, 25 Nov 2014 20:01:28 +0100
Message-ID: <5474D208.4080009@siemens.com>
References: <20141125172108.992070524@redhat.com> <20141125172329.580701786@redhat.com> <5474BE7C.5040909@redhat.com>
Mime-Version: 1.0
Content-Type: text/plain; charset=windows-1252
Content-Transfer-Encoding: 7bit
Cc: kvm@vger.kernel.org
To: Paolo Bonzini <pbonzini@redhat.com>, linux-kernel@vger.kernel.org
Return-path: <linux-kernel-owner@vger.kernel.org>
In-Reply-To: <5474BE7C.5040909@redhat.com>
Sender: linux-kernel-owner@vger.kernel.org
List-Id: kvm.vger.kernel.org

On 2014-11-25 18:38, Paolo Bonzini wrote:
> 
> 
> On 25/11/2014 18:21, Marcelo Tosatti wrote:
>> +
>> +	if (r == HRTIMER_RESTART) {
>> +		do {
>> +			ret = hrtimer_start_expires(data, HRTIMER_MODE_ABS);
>> +			if (ret == -ETIME)
>> +				hrtimer_add_expires_ns(&ktimer->timer,
>> +							ktimer->period);
> 
> Is it possible to just compute the time where the next interrupt
> happens?  I suspect the printk and WARN_ON below can be easily triggered
> by a guest.

We have a lower bound for the period that a guest can program. Unless
that value is set too low, this should practically not happen if we
avoid disturbances while handling the event and reprogramming the next
one (irqs off?).

Jan

-- 
Siemens AG, Corporate Technology, CT RTC ITP SES-DE
Corporate Competence Center Embedded Linux