On 28/07/14 16:22, Alexander Graf wrote: > > On 28.07.2014, at 16:16, David Hildenbrand <dahi@xxxxxxxxxxxxxxxxxx> wrote: > >>> >>> On 10.07.14 15:10, Christian Borntraeger wrote: >>>> From: David Hildenbrand <dahi@xxxxxxxxxxxxxxxxxx> >>>> >>>> If a cpu is stopped, it must never be allowed to run and no interrupt may wake it >>>> up. A cpu also has to be unhalted if it is halted and has work to do - this >>>> scenario wasn't hit in kvm case yet, as only "disabled wait" is processed within >>>> QEMU. >>>> >>>> Signed-off-by: David Hildenbrand <dahi@xxxxxxxxxxxxxxxxxx> >>>> Reviewed-by: Cornelia Huck <cornelia.huck@xxxxxxxxxx> >>>> Reviewed-by: Christian Borntraeger <borntraeger@xxxxxxxxxx> >>>> Signed-off-by: Christian Borntraeger <borntraeger@xxxxxxxxxx> >>> >>> This looks like it's something that generic infrastructure should take >>> care of, no? How does this work for the other archs? They always get an >>> interrupt on the transition between !has_work -> has_work. Why don't we >>> get one for s390x? >>> >>> >>> Alex >>> >>> >> >> Well, we have the special case on s390 as a CPU that is in the STOPPED or the >> CHECK STOP state may never run - even if there is an interrupt. It's >> basically like this CPU has been switched off. >> >> Imagine that it is tried to inject an interrupt into a stopped vcpu. It >> will kick the stopped vcpu and thus lead to a call to >> "kvm_arch_process_async_events()". We have to deny that this vcpu will ever >> run as long as it is stopped. It's like a way to "suppress" the >> interrupt for such a transition you mentioned. > > An interrupt kick usually just means we go back into the main loop. From there we check the interrupt bitmap which interrupt to handle. Check out the handling code here: > > http://git.qemu.org/?p=qemu.git;a=blob;f=cpu-exec.c;h=38e5f02a307523d99134f4e2e6c51683bb10b45b;hb=HEAD#l580 > > If you just check for the stopped state in here, do_interrupt() will never get called and thus the CPU shouldn't ever get executed. Unless I'm heavily mistaken :). > >> >> Later, another vcpu might decide to turn that vcpu back on (by e.g. sending a >> SIGP START to that vcpu). > > Yes, in that case that other CPU generates a signal (a different bit in interrupt_request) and the first CPU would see that it has to wake up and wake up. > >> I am not sure if such a mechanism/scenario is applicable to any other arch. They >> all seem to reset the cs->halted flag if they know they are able to run (e.g. >> due to an interrupt) - they have no such thing as "stopped cpus", only >> "halted/waiting cpus". > > There's not really much difference between the two. The only difference from a software point of view is that a "stopped" CPU has its external interrupt bits masked off, no? We have - wait (wait bit in PSW) - disabled wait (wait bit and interrupt fencing in PSW) - STOPPED (not related to PSW, state change usually handled via service processor or hypervisor) I think we have to differentiate between KVM/TCG. On KVM we always do in kernel halt and qemu sees a halted only for STOPPED or disabled wait. TCG has to take care of the normal wait as well. >From a first glimpse, a disabled wait and STOPPED look similar, but there are (important) differences, e.g. other CPUs get a different a different result from a SIGP SENSE. This makes a big difference, e.g. for Linux guests, that send a SIGP STOP, followed by a SIGP SENSE loop until the CPU is down on hotplug (and shutdown, kexec..) So I think we agree, that handling the cpu states natively makes sense. The question is now only how to model it correctly without breaking TCG/KVM and reuse as much common code as possible. Correct? Do I understand you correctly, that your collapsing of stopped and halted is only in the qemu coding sense, IOW maybe we could just modify kvm_arch_process_async_events to consider the STOPPED state, as TCGs sigp implementation does not support SMP anyway? David would that work? Christian -- To unsubscribe from this list: send the line "unsubscribe kvm" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html