Re: winXP "Standard PC" HAL and qemu-kvm >= 0.15

[Michael Tokarev <mjt@tls.msk.ru>]

On 06.12.2011 20:57, Michael Tokarev wrote:
> On 06.12.2011 20:38, Jan Kiszka wrote:
>> On 2011-12-06 17:29, Michael Tokarev wrote:
> []
>>> It appears there are two issues here, one is fixed by
>>> 09de0f469c3c2a277c7874f6c60992c8b94719a9 and is 32bit-only, and
>>> another bisect leads to this commit:
> 
> Or 3... :)
> 
>>> commit 59539c913383fdd3350681301b44f02fa7ee2757
>>> Author: Jan Kiszka <jan.kiszka@siemens.com>
>>> Date:   Mon Jun 27 12:22:28 2011 +0200
>>>
>>>     qemu-kvm: Fix in-kernel PIC reset
> 
> []
>>> Anything wrong with this patch?
>>
>> I tend to say "no". It may just reveals some issue elsewhere.
>>
>> To cross-check: Does this series [1] expose the same issue with vanilla
>> QEMU when enabling that in-kernel irqchip version?
> 
> I cross-checked it the other way, which really should have been done
> at the very beginning instead of wasting so much time of so many
> people at once.  I just tried qemu-kvm-1.0, which I wasn't due to
> an unrelated issue (1.0 does not work for me in my funky 32/64bit
> environment).  Also, since the bug has been reported especially
> against 0.15 version (triggering by upgrading from 0.14 to 0.15),
> I didn't insist on trying 1.0, which was my biggest mistake.
> 
> And in qemu-kvm-1.0, the problem guest Just Works.  So i t must be
> something else which were fixed between 0.15 and 1.0.

For the 0.15 .. 1.0 change, the first commit which restores the (broken
in 0.15) functionality is this one:

commit 86fbf97ceb4a9c46a609dd4ae053ba4262b68fe8
Author: Jan Kiszka <jan.kiszka@siemens.com>
Date:   Fri Oct 7 09:19:45 2011 +0200

    i8259: Clear ELCR on reset

    The ELCR is actually part of the chipset but we model it here for
    simplicity reasons. The PIIX3 clears the ELCR on reset, which was once
    broken by 4dbe19e181. Fix this by splitting up pic_init_reset from
    pic_reset and clearing the register in the latter.

    Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
    Signed-off-by: Blue Swirl <blauwirbel@gmail.com>

Which is quite expected having in mind the commit which "broke"
it for 0.15.

> Upstream qemu 1.0 does not have this issue too.
> 
>> [1] http://thread.gmane.org/gmane.comp.emulators.kvm.devel/82871
> 
> I'll try to verify this series in a moment, but hopefully it will work
> fine too :)

And that turned out to be not so easy for me.  Which is this
series against?  Is there may be a git tree with this series
applied already?

Thanks!

/mjt