Re: [Qemu-devel] [PATCH] kvm: First step to push iothread lock out of inner run loop

From: Avi Kivity <avi@redhat.com>
To: Jan Kiszka <jan.kiszka@web.de>
Cc: Marcelo Tosatti <mtosatti@redhat.com>,
	Liu Ping Fan <pingfank@linux.vnet.ibm.com>,
	kvm <kvm@vger.kernel.org>, qemu-devel <qemu-devel@nongnu.org>,
	Alexander Graf <agraf@suse.de>,
	Anthony Liguori <anthony@codemonkey.ws>,
	Gleb Natapov <gleb@redhat.com>
Subject: Re: [Qemu-devel] [PATCH] kvm: First step to push iothread lock out of inner run loop
Date: Sun, 24 Jun 2012 17:31:18 +0300	[thread overview]
Message-ID: <4FE724B6.8010607@redhat.com> (raw)
In-Reply-To: <4FE71F44.9020800@web.de>

On 06/24/2012 05:08 PM, Jan Kiszka wrote:
> On 2012-06-24 10:49, Avi Kivity wrote:
>> On 06/23/2012 02:45 PM, Jan Kiszka wrote:
>>>
>>> Hmm, we may need the iothread lock around cpu_set_apic_tpr for
>>> !kvm_irqchip_in_kernel(). And as we are at it, apic_base manipulation
>>> can be but there as well.
>>>
>>> With in-kernel irqchip, there is no such need. Also, no one accesses
>>> eflags outside of the vcpu thread, independent of the irqchip mode.
>> 
>> In fact !kvm_irqchip_in_kernel() is broken wrt the tpr.  Interrupt
>> injection needs to be done atomically, but currently we check the tpr
>> from the injecting thread, which means the cpu thread can race with it.
>>  We need to move the check to the vcpu thread so that the guest vcpu is
>> halted.
> 
> So apic_set_irq basically needs to be deferred to vcpu context, right?
> Will have a look.

Correct. IIRC, the kernel's 0a5fff192388d2 made the problem much worse,
but did not create it.  It was either Vista or XP-64 which triggered the
problem reliably.  Copying Gleb in case he remembers more.

-- 
error compiling committee.c: too many arguments to function