Re: [Qemu-devel] [PATCH v4 5/5] i8259: fix dynamically masking slave IRQs with IMR register

[Avi Kivity <avi@redhat.com>]

On 09/03/2012 06:54 PM, Jan Kiszka wrote:
> On 2012-09-03 17:52, Avi Kivity wrote:
>> On 09/03/2012 06:42 PM, Juan Quintela wrote:
>>> Avi Kivity <avi@redhat.com> wrote:
>>>> On 09/03/2012 11:40 AM, Andreas Färber wrote:
>>>>> Am 03.09.2012 04:56, schrieb Matthew Ogilvie:
>>>>>> diff --git a/hw/i8259_common.c b/hw/i8259_common.c
>>>>>> index ab3d98b..dcde5f2 100644
>>>>>> --- a/hw/i8259_common.c
>>>>>> +++ b/hw/i8259_common.c
>>>>> [...]
>>>>>> @@ -111,6 +112,7 @@ static const VMStateDescription vmstate_pic_common = {
>>>>>>          VMSTATE_UINT8(isr, PICCommonState),
>>>>>>          VMSTATE_UINT8(priority_add, PICCommonState),
>>>>>>          VMSTATE_UINT8(irq_base, PICCommonState),
>>>>>> +        VMSTATE_UINT8(icw3, PICCommonState),
>>>>>>          VMSTATE_UINT8(read_reg_select, PICCommonState),
>>>>>>          VMSTATE_UINT8(poll, PICCommonState),
>>>>>>          VMSTATE_UINT8(special_mask, PICCommonState),
>>>>>
>>>>> Additional VMState needs to be versioned by incrementing .version_id and
>>>>> by specifying the new version number here. Otherwise it breaks migration.
>>>
>>> For the subsection, only sending this when icw3 != 0 is enough?  I am
>>> searching for a test about when we need to send it.
>> 
>> Looks like the optimal condition is ((s->icw3 & ~s->eclr) != 0) (i.e.
>> bit set in icw3 but clear in eclr).
> 
> The standard PC values are optimal: 4 for master, 2 for slave.

Can you explain why?  I saw that icw3 is always ORed with eclr, so my
condition will catch exactly those cases where a change in behaviour
occurs, and no more.

-- 
error compiling committee.c: too many arguments to function