[v3,3/8] VFIO: platform: forwarded state tested when selecting IRQ handler

Message ID 1416767760-14487-4-git-send-email-eric.auger@linaro.org
State New
Headers show

Commit Message

Auger Eric Nov. 23, 2014, 6:35 p.m.
In case the IRQ is forwarded, the VFIO platform IRQ handler does not
need to disable the IRQ anymore.

When setting the IRQ handler we now also test the forwarded state. In
case the IRQ is forwarded we select the edge handler (no automaske).

Signed-off-by: Eric Auger <eric.auger@linaro.org>

---

v2 -> v3:
- forwarded state was tested in the handler. Now the forwarded state
  is tested before setting the handler. This definitively limits
  the dynamics of forwarded state changes but I don't think there is
  a use case where we need to be able to change the state at any time.
---
 drivers/vfio/platform/vfio_platform_irq.c | 7 ++++++-
 1 file changed, 6 insertions(+), 1 deletion(-)

Comments

Christoffer Dall Nov. 30, 2014, 12:47 p.m. | #1
The subject reads strangely, perhaps just:

VFIO: platform: test forward state when selecting IRQ handler

On Sun, Nov 23, 2014 at 07:35:55PM +0100, Eric Auger wrote:
> In case the IRQ is forwarded, the VFIO platform IRQ handler does not
> need to disable the IRQ anymore.
> 
> When setting the IRQ handler we now also test the forwarded state. In
> case the IRQ is forwarded we select the edge handler (no automaske).
> 
> Signed-off-by: Eric Auger <eric.auger@linaro.org>
> 
> ---
> 
> v2 -> v3:
> - forwarded state was tested in the handler. Now the forwarded state
>   is tested before setting the handler. This definitively limits
>   the dynamics of forwarded state changes but I don't think there is
>   a use case where we need to be able to change the state at any time.

user space can change this by calling the VFIO_IRQ_SET_ACTION_TRIGGER
whenever it wants, right?

> ---
>  drivers/vfio/platform/vfio_platform_irq.c | 7 ++++++-
>  1 file changed, 6 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/vfio/platform/vfio_platform_irq.c b/drivers/vfio/platform/vfio_platform_irq.c
> index 08d400e..61a2920 100644
> --- a/drivers/vfio/platform/vfio_platform_irq.c
> +++ b/drivers/vfio/platform/vfio_platform_irq.c
> @@ -230,8 +230,13 @@ static int vfio_platform_set_irq_trigger(struct vfio_platform_device *vdev,
>  {
>  	struct vfio_platform_irq *irq = &vdev->irqs[index];
>  	irq_handler_t handler;
> +	struct irq_data *d;
> +	bool is_forwarded;
>  
> -	if (vdev->irqs[index].flags & VFIO_IRQ_INFO_MASKABLE)
> +	d = irq_get_irq_data(irq->hwirq);
> +	is_forwarded = irqd_irq_forwarded(d);
> +
> +	if ((vdev->irqs[index].flags & VFIO_IRQ_INFO_MASKABLE) && !is_forwarded)
>  		handler = vfio_maskable_irq_handler;
>  	else
>  		handler = vfio_irq_handler;
> -- 
> 1.9.1
>
Auger Eric Dec. 1, 2014, 2:39 p.m. | #2
Hi Christoffer,
On 11/30/2014 01:47 PM, Christoffer Dall wrote:
> The subject reads strangely, perhaps just:
> 
> VFIO: platform: test forward state when selecting IRQ handler
OK
> 
> On Sun, Nov 23, 2014 at 07:35:55PM +0100, Eric Auger wrote:
>> In case the IRQ is forwarded, the VFIO platform IRQ handler does not
>> need to disable the IRQ anymore.
>>
>> When setting the IRQ handler we now also test the forwarded state. In
>> case the IRQ is forwarded we select the edge handler (no automaske).
>>
>> Signed-off-by: Eric Auger <eric.auger@linaro.org>
>>
>> ---
>>
>> v2 -> v3:
>> - forwarded state was tested in the handler. Now the forwarded state
>>   is tested before setting the handler. This definitively limits
>>   the dynamics of forwarded state changes but I don't think there is
>>   a use case where we need to be able to change the state at any time.
> 
> user space can change this by calling the VFIO_IRQ_SET_ACTION_TRIGGER
> whenever it wants, right?
yes the user can set/unset the VFIO signaling (and request_irq/free_irq)
using VFIO_IRQ_SET_ACTION_TRIGGER. In this new version I do not allow
changing the forwarded state when the handler is attached (request_irq).

Does it answer your interrogation?

Best Regards

Eric

> 
>> ---
>>  drivers/vfio/platform/vfio_platform_irq.c | 7 ++++++-
>>  1 file changed, 6 insertions(+), 1 deletion(-)
>>
>> diff --git a/drivers/vfio/platform/vfio_platform_irq.c b/drivers/vfio/platform/vfio_platform_irq.c
>> index 08d400e..61a2920 100644
>> --- a/drivers/vfio/platform/vfio_platform_irq.c
>> +++ b/drivers/vfio/platform/vfio_platform_irq.c
>> @@ -230,8 +230,13 @@ static int vfio_platform_set_irq_trigger(struct vfio_platform_device *vdev,
>>  {
>>  	struct vfio_platform_irq *irq = &vdev->irqs[index];
>>  	irq_handler_t handler;
>> +	struct irq_data *d;
>> +	bool is_forwarded;
>>  
>> -	if (vdev->irqs[index].flags & VFIO_IRQ_INFO_MASKABLE)
>> +	d = irq_get_irq_data(irq->hwirq);
>> +	is_forwarded = irqd_irq_forwarded(d);
>> +
>> +	if ((vdev->irqs[index].flags & VFIO_IRQ_INFO_MASKABLE) && !is_forwarded)
>>  		handler = vfio_maskable_irq_handler;
>>  	else
>>  		handler = vfio_irq_handler;
>> -- 
>> 1.9.1
>>
Christoffer Dall Dec. 1, 2014, 8:10 p.m. | #3
On Mon, Dec 01, 2014 at 03:39:24PM +0100, Eric Auger wrote:
> Hi Christoffer,
> On 11/30/2014 01:47 PM, Christoffer Dall wrote:
> > The subject reads strangely, perhaps just:
> > 
> > VFIO: platform: test forward state when selecting IRQ handler
> OK
> > 
> > On Sun, Nov 23, 2014 at 07:35:55PM +0100, Eric Auger wrote:
> >> In case the IRQ is forwarded, the VFIO platform IRQ handler does not
> >> need to disable the IRQ anymore.
> >>
> >> When setting the IRQ handler we now also test the forwarded state. In
> >> case the IRQ is forwarded we select the edge handler (no automaske).
> >>
> >> Signed-off-by: Eric Auger <eric.auger@linaro.org>
> >>
> >> ---
> >>
> >> v2 -> v3:
> >> - forwarded state was tested in the handler. Now the forwarded state
> >>   is tested before setting the handler. This definitively limits
> >>   the dynamics of forwarded state changes but I don't think there is
> >>   a use case where we need to be able to change the state at any time.
> > 
> > user space can change this by calling the VFIO_IRQ_SET_ACTION_TRIGGER
> > whenever it wants, right?
> yes the user can set/unset the VFIO signaling (and request_irq/free_irq)
> using VFIO_IRQ_SET_ACTION_TRIGGER. In this new version I do not allow
> changing the forwarded state when the handler is attached (request_irq).
> 
> Does it answer your interrogation?
> 
interrogation?  hopefully it wasn't that bad ;)

Yes, that answers my question.

Thanks,
-Christoffer
Auger Eric Dec. 1, 2014, 9:15 p.m. | #4
On 12/01/2014 09:10 PM, Christoffer Dall wrote:
> On Mon, Dec 01, 2014 at 03:39:24PM +0100, Eric Auger wrote:
>> Hi Christoffer,
>> On 11/30/2014 01:47 PM, Christoffer Dall wrote:
>>> The subject reads strangely, perhaps just:
>>>
>>> VFIO: platform: test forward state when selecting IRQ handler
>> OK
>>>
>>> On Sun, Nov 23, 2014 at 07:35:55PM +0100, Eric Auger wrote:
>>>> In case the IRQ is forwarded, the VFIO platform IRQ handler does not
>>>> need to disable the IRQ anymore.
>>>>
>>>> When setting the IRQ handler we now also test the forwarded state. In
>>>> case the IRQ is forwarded we select the edge handler (no automaske).
>>>>
>>>> Signed-off-by: Eric Auger <eric.auger@linaro.org>
>>>>
>>>> ---
>>>>
>>>> v2 -> v3:
>>>> - forwarded state was tested in the handler. Now the forwarded state
>>>>   is tested before setting the handler. This definitively limits
>>>>   the dynamics of forwarded state changes but I don't think there is
>>>>   a use case where we need to be able to change the state at any time.
>>>
>>> user space can change this by calling the VFIO_IRQ_SET_ACTION_TRIGGER
>>> whenever it wants, right?
>> yes the user can set/unset the VFIO signaling (and request_irq/free_irq)
>> using VFIO_IRQ_SET_ACTION_TRIGGER. In this new version I do not allow
>> changing the forwarded state when the handler is attached (request_irq).
>>
>> Does it answer your interrogation?
>>
> interrogation?  hopefully it wasn't that bad ;)
Oups sorry "faux ami" in french. I wanted to say question but wanted to
say it smarter. another time ;-)
> 
> Yes, that answers my question.
> 
> Thanks,
> -Christoffer
>

Patch

diff --git a/drivers/vfio/platform/vfio_platform_irq.c b/drivers/vfio/platform/vfio_platform_irq.c
index 08d400e..61a2920 100644
--- a/drivers/vfio/platform/vfio_platform_irq.c
+++ b/drivers/vfio/platform/vfio_platform_irq.c
@@ -230,8 +230,13 @@  static int vfio_platform_set_irq_trigger(struct vfio_platform_device *vdev,
 {
 	struct vfio_platform_irq *irq = &vdev->irqs[index];
 	irq_handler_t handler;
+	struct irq_data *d;
+	bool is_forwarded;
 
-	if (vdev->irqs[index].flags & VFIO_IRQ_INFO_MASKABLE)
+	d = irq_get_irq_data(irq->hwirq);
+	is_forwarded = irqd_irq_forwarded(d);
+
+	if ((vdev->irqs[index].flags & VFIO_IRQ_INFO_MASKABLE) && !is_forwarded)
 		handler = vfio_maskable_irq_handler;
 	else
 		handler = vfio_irq_handler;