Ack: Re: [Oneiric][SRU][PATCH 0/1] Bug #974664 - Issue with ite_cir drive

[Luis Henriques]

On Thu, Apr 19, 2012 at 11:32:30AM -0300, Herton R. Krzesinski wrote:
> On Thu, Apr 19, 2012 at 03:22:23PM +0100, Luis Henriques wrote:
> > BugLink: http://bugs.launchpad.net/bugs/974664
> > 
> > == Oneiric SRU Justification ==
> > 
> > A bug in the upstream ite_cir driver has been triggered recently on Oneiric
> > kernels, causing some systems (in particular Dell XPS 1340) to crash during
> > boot.
> > 
> > The problem is related to an early registration of an ISR during the device
> > probe process.  This caused IRQs to be served before the driver initialisation
> > had been completed.
> > 
> > == Fix ==
> > 
> > Moving the invocation of request_irq() later in the probe solved the problem.
> 
> Ack, since the move after itdev->params.init_hardware should ensure that
> no IRQs come in (init_hardware callback looking at the code should
> ensure hardware is initialized make sure interrupts are disabled, but
> before it seems an undefined state). I believe the key point here is
> moving after init_hardware is safe, since it initializes the hardware
> with interrupts disabled.
> 
> Is this already upstream? The fix should be sent, with Cc: stable.

Ugh!  I've sent it to lkml already but realise now I forgot the Cc:
stable.

Cheers,
--
Luis