Date: Sat, 5 Jun 2010 18:18:16 +0100
From: Marc Zyngier <maz@misterjones.org>
To: Esben Haabendal <esbenhaabendal@gmail.com>
Cc: Thomas Gleixner <tglx@linutronix.de>,
       Esben Haabendal <eha@doredevelopment.dk>, linux-kernel@vger.kernel.org,
       mingo@elte.hu, joachim.eastwood@jotron.com
Subject: Re: [RFC][PATCH] irq: support IRQ_NESTED_THREAD with non-threaded 
 interrupt handlers
Message-ID: <20100605181816.7aae71c7@hina.wild-wind.fr.eu.org>
In-Reply-To: <AANLkTilUGOVPBtdl2KB2nrbra8uBO8q4AX0v57hCtr_9@mail.gmail.com>
References: <1275686352.2970.2.camel@eha.doredevelopment.dk>
	<alpine.LFD.2.00.1006042343150.2933@localhost.localdomain>
	<AANLkTilb7---Qg1oLwzanKP1RGe04fkAXDtZzz0w9MGz@mail.gmail.com>
	<20100605151031.2d562268@hina.wild-wind.fr.eu.org>
	<AANLkTilUGOVPBtdl2KB2nrbra8uBO8q4AX0v57hCtr_9@mail.gmail.com>
Organization: Metropolis -- Nowhere
Mime-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 1896
Lines: 42

On Sat, 5 Jun 2010 18:53:53 +0200
Esben Haabendal <esbenhaabendal@gmail.com> wrote:

> On Sat, Jun 5, 2010 at 4:10 PM, Marc Zyngier <maz@misterjones.org> wrote:
> > On Sat, 5 Jun 2010 15:56:01 +0200
> > Esben Haabendal <esbenhaabendal@gmail.com> wrote:
> >
> >> I have a board with an I2C PCA9535 chip with two PHY interrupt lines
> >> hooked up to. The pca953x driver calls set_irq_nested_thread on all
> >> irq's on initialization. The PHY driver then calls request_irq, and has
> >> no idea that it should actually be using a threaded handler.
> >>
> >> With this patch, the PHY driver is able to work in this scenario
> >> without changes (and so should any other driver using request_irq).
> >
> > You may want to give request_any_context_irq() a try (available since
> > the latest merge window). It still requires your driver to be changed,
> > but it should then work in both threaded and non-threaded cases.
> 
> The problem is not in "my" driver, but in the phy driver framework in this
> particular case.

To me, any driver/subsystem/whatever qualifies as "yours" if you're taking
care of it ;-).

> What is the plan here, should all drivers change from
> request_any_context_irq() at some point in time?

All? Probably not. Only the few where the need arises, and after carefully
checking that you're not introducing a horrible race condition somewhere
(threaded handlers are running with interrupts enabled...).

	M.
-- 
Enforcement officers may use a hand-held detection device to measure both
the direction and the strength of a TV signal. This makes it easy for us
to locate TV receiving equipment in even the hardest-to-reach places.
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/