Re: [PATCH 6/6 V2] Add age out of guest paused flag

[Eric B Munson]

On Tue, 01 Nov 2011, Anthony Liguori wrote:

> On 11/01/2011 02:51 PM, Eric B Munson wrote:
> >On Tue, 01 Nov 2011, Anthony Liguori wrote:
> >
> >>On 10/31/2011 03:07 PM, Eric B Munson wrote:
> >>>The KVM_GUEST_PAUSED flag will prevent a guest from compaining about a soft
> >>>lockup but it can mask real soft lockups if the flag isn't cleared when it is
> >>>no longer relevant.  This patch adds a kvm ioctl that the hypervisor will use
> >>>when it resumes a guest to start a timer for aging out the flag.  The time out
> >>>will be specified by the hypervisor in the ioctl call.
> >>>
> >>>Signed-off-by: Eric B Munson<emunson@xxxxxxxxx>
> >>
> >>Why not have the guest clear the flag when it acknowledges it?
> >>
> >>The hypervisor would unconditionally set the bit, and the guest
> >>would do a testandclear to check if the bit is set.  I think that
> >>avoids the whole aging business.
> >>
> >>Regards,
> >>
> >>Anthony Liguori
> >
> >If you have a look at patch 5 of this series, the flag is cleared when the
> >guest checks the validity of a soft lockup.  However, the aging is meant to
> >cover the case where the guest never sees a soft lockup.  We don't want this
> >flag to be stored for ever and end up delaying real soft lockup messages.  With
> >that case in mind, I thought this was a good/simple compramise.
> 
> If the guest clears the flag, then I don't think you have to worry about this.
> 
> Regards,
> 
> Anthony Liguori

Except that the soft lockup doesn't trigger on every qemu "stop", only some of
them.  The flag isn't cleared until the detector is triggered.

Eric

Attachment: signature.asc
Description: Digital signature