Re: [Xen-devel] [PATCH, v2] x86/HVM: assorted RTC emulation adjustments

[Stefano Stabellini <stefano.stabellini@xxxxxxxxxxxxx>]

On Fri, 24 Aug 2012, Jan Beulich wrote:
> >>> On 23.08.12 at 15:44, Stefano Stabellini 
> >>> <stefano.stabellini@xxxxxxxxxxxxx>
> wrote:
> > On Wed, 22 Aug 2012, Jan Beulich wrote:
> >> - don't call rtc_timer_update() on REG_A writes when the value didn't
> >>   change (doing the call always was reported to cause wall clock time
> >>   lagging with the JVM running on Windows)
> >> - don't call rtc_timer_update() on REG_B writes at all
> >> - only call alarm_timer_update() on REG_B writes when relevant bits
> >>   change
> >> - only call check_update_timer() on REG_B writes when SET changes
> >> - instead properly handle AF and PF when the guest is not also setting
> >>   AIE/PIE respectively (for UF this was already the case, only a
> >>   comment was slightly inaccurate)
> >> - raise the RTC IRQ not only when UIE gets set while UF was already
> >>   set, but generalize this to cover AIE and PIE as well
> >> - properly mask off bit 7 when retrieving the hour values in
> >>   alarm_timer_update(), and properly use RTC_HOURS_ALARM's bit 7 when
> >>   converting from 12- to 24-hour value
> >> - also handle the two other possible clock bases
> >> - use RTC_* names in a couple of places where literal numbers were used
> >>   so far
> >> 
> >> Note that this only improves the situation described in the thread at
> >> http://lists.xen.org/archives/html/xen-devel/2012-08/msg00664.html,
> >> there are still problems with the emulation when invoked at a high rate
> >> as described there.
> >> 
> >> Signed-off-by: Jan Beulich <jbeulich@xxxxxxxx>
> > 
> > Although this patch solves a real problem and should probably go in at
> > some point, I am a bit worried about drifting too much from the original
> > RTC emulator (that was taken from QEMU),
> 
> Then does that emulator have similar problems?

I am not sure, it probably does.
I am afraid the code already drifted too much to make comparisons.


> > because it would be nice to be able to backport features like this one:
> > 
> > http://marc.info/?l=qemu-devel&m=134392375010304 
> 
> I agree this would be nice to have (albeit I'm not sure how much
> the original problem is actually present in the Xen code, particularly
> with the patch here applied, as I think it may implicitly clean up some
> of the unneccesary active timers).

Maybe, but with your patch applied, are there going to be any timers
running if the guest is not making use of the RTC?

_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxx
http://lists.xen.org/xen-devel