[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[Xen-devel] Re: [PATCH 08/13] xen/pvticketlock: disable interrupts while blocking
- To: Avi Kivity <avi@xxxxxxxxxx>
- From: Don Zickus <dzickus@xxxxxxxxxx>
- Date: Wed, 7 Sep 2011 11:56:57 -0400
- Cc: Jeremy Fitzhardinge <jeremy@xxxxxxxx>, Marcelo Tosatti <mtosatti@xxxxxxxxxx>, Nick Piggin <npiggin@xxxxxxxxx>, KVM <kvm@xxxxxxxxxxxxxxx>, Stefano Stabellini <stefano.stabellini@xxxxxxxxxxxxx>, Peter Zijlstra <peterz@xxxxxxxxxxxxx>, the arch/x86 maintainers <x86@xxxxxxxxxx>, Linux Kernel Mailing List <linux-kernel@xxxxxxxxxxxxxxx>, Andi Kleen <andi@xxxxxxxxxxxxxx>, Jeremy Fitzhardinge <jeremy.fitzhardinge@xxxxxxxxxx>, "H. Peter Anvin" <hpa@xxxxxxxxx>, Ingo Molnar <mingo@xxxxxxx>, Linus Torvalds <torvalds@xxxxxxxxxxxxxxxxxxxx>, Xen Devel <xen-devel@xxxxxxxxxxxxxxxxxxx>
- Delivery-date: Wed, 07 Sep 2011 09:04:12 -0700
- List-id: Xen developer discussion <xen-devel.lists.xensource.com>
On Wed, Sep 07, 2011 at 06:11:14PM +0300, Avi Kivity wrote:
> On 09/07/2011 04:44 PM, Don Zickus wrote:
> >> Is there a way to tell whether an NMI was internally or externally
> >> generated?
> >> I don't think so, especially as two or more NMIs can be coalesced.
> >> So any NMI received on this first cpu has to check the NMI reason
> >> port?
> >Well we cheat and execute all the nmi handlers first. If they come back
> >as handled, we skip the check for the external NMI.
> And hope that no other NMI was generated while we're handling this
> one. It's a little... fragile?
No. If another NMI is generated while we are processing the current one
it should get latched. Upon completion of the current one, the cpu should
jump right back into the nmi exception routine again. The only downside
is when multiple NMIs come in during the processing of the current one.
Only one can be latched, so the others get dropped. But we are addressing
> >But you are right, other than checking the reason port, there isn't a way
> >to determine if an NMI is internally or externally generated.
> >> >>
> >> >> But on the other hand, I don't really care if you can say that this
> >> path
> >> >> will never be called in a virtual machine.
> >> >
> >> >Does virtual machines support hot remove of cpus? Probably not
> >> >considering bare-metal barely supports it.
> >> >
> >> They do.
> >But vcpus probably don't have the notion of a bsp cpu, so perhaps virtual
> >machines can get away with it easier? (I don't know enough about the hot
> >cpu remove code to really explain it, just enough to know it can cause
> >problems and people are trying to address it).
> The concept of a bsp exists in exactly the same way as on real hardware.
> error compiling committee.c: too many arguments to function
Xen-devel mailing list