Re: [PATCH 1/3] gpio: dln2: fix issue when an IRQ is unmasked then enabled

From: Linus Walleij
Date: Mon Jan 05 2015 - 03:55:32 EST


On Wed, Dec 31, 2014 at 9:55 PM, Linus Walleij <linus.walleij@xxxxxxxxxx> wrote:
> On Thu, Dec 11, 2014 at 2:02 PM, Octavian Purdila
> <octavian.purdila@xxxxxxxxx> wrote:
>
>> As noticed during suspend/resume operations, the IRQ can be unmasked
>> then disabled in suspend and eventually enabled in resume, but without
>> being unmasked.
>>
>> The current implementation does not take into account interactions
>> between mask/unmask and enable/disable interrupts, and thus in the
>> above scenarios the IRQs remain unactive.
>>
>> To fix this we removed the enable/disable operations as they fallback
>> to mask/unmask anyway.
>>
>> We also remove the pending bitmaks as it is already done in irq_data
>> (i.e. IRQS_PENDING).
>>
>> Signed-off-by: Octavian Purdila <octavian.purdila@xxxxxxxxx>
>
> Patch applied for fixes.

Bah now that I see there are several versions of the patch set
floating around and also MFD patches I don't quite understand
how acute this is or how it's to be applied.

- Are these regression fixes or nice to have for next kernel
release?

- Are the GPIO patches independent of the MFD patch?

Yours,
Linus Walleij
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/