Prev: [PATCH 1/1] perf: add support for arch-dependent symbolic event names to "perf stat"
Next: [RELEASE] Userspace RCU 0.4.2
From: Stefano Stabellini on 4 Mar 2010 14:20
On Thu, 4 Mar 2010, Jeremy Fitzhardinge wrote:
> I think you need to expand this comment to explain the consequences of
> this change on SMP bringup and timers. I still don't see why there are
> (or should be) any differences from PV cpu bringup, aside from the tiny
> details of syscall setup (and MTRR I guess, but that's all a noop
> anyway, right?).
> Also, I think you can probably do the timer changes in a separate patch,
> as even with evtchns, an emulated timer device interrupt will still be
> delivered properly via VIRQ_EMUL_PIN_*.
One of the problems is that some VCPU initialization hypercalls at the
moment are PV specific and don't work with HVM guests, in particular I
find that VCPUOP_initialise doesn't work correctly for me.
But the timers shouldn't be an issue because you should be able to
receive the VIRQ_TIMER or the emulated timer interrupt anyway.
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo(a)vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/