Re: + spi-imx-use-platform_driver_probe-as-probe-lives-in-inittext.patch added to -mm tree

[Grant Likely <grant.likely-s3s/WqlpOiPyB63q8FvJNQ@public.gmane.org>]

2009/12/9 Grant Likely <grant.likely-s3s/WqlpOiPyB63q8FvJNQ@public.gmane.org>:
> 2009/12/9 Uwe Kleine-König <u.kleine-koenig-bIcnvbaLZ9MEGnE8C9+IrQ@public.gmane.org>:
>> On Tue, Dec 08, 2009 at 05:34:21PM -0700, Grant Likely wrote:
>>> Hi Uwe.
>>>
>>> On Tue, Dec 8, 2009 at 4:30 PM,  <akpm-de/tnXTf+JLsfHDXvbKv3WD2FQJk+8+b@public.gmane.org> wrote:
>>> > Subject: spi-imx: use platform_driver_probe as probe lives in .init.text
>>> > From: Uwe Kleine-König <u.kleine-koenig-bIcnvbaLZ9MEGnE8C9+IrQ@public.gmane.org>
>>> >
>>> > Using platform_driver_register with a probe function defined using __init
>>> > is wrong.
>>> >
>>> > This fixes an oops after:
>>> >
>>> >        cd /sys/bus/platform/drivers/spi_imx
>>> >        echo -n spi_imx.0 > unbind
>>> >        echo -n spi_imx.0 > bind
>>> >
>>> > Signed-off-by: Uwe Kleine-König <u.kleine-koenig-bIcnvbaLZ9MEGnE8C9+IrQ@public.gmane.org>
>>> > Cc: David Brownell <dbrownell-Rn4VEauK+AKRv+LV9MX5uipxlwaOVQ5f@public.gmane.org>
>>> > Cc: Daniel Mack <daniel-rDUAYElUppE@public.gmane.org>
>>> > Cc: Sascha Hauer <s.hauer-bIcnvbaLZ9MEGnE8C9+IrQ@public.gmane.org>
>>> > Cc: Grant Likely <grant.likely-s3s/WqlpOiPyB63q8FvJNQ@public.gmane.org>
>>> > Signed-off-by: Andrew Morton <akpm-de/tnXTf+JLsfHDXvbKv3WD2FQJk+8+b@public.gmane.org>
>>> > ---
>>> >
>>> >  drivers/spi/spi_imx.c |    3 +--
>>> >  1 file changed, 1 insertion(+), 2 deletions(-)
>>> >
>>> > diff -puN drivers/spi/spi_imx.c~spi-imx-use-platform_driver_probe-as-probe-lives-in-inittext drivers/spi/spi_imx.c
>>> > --- a/drivers/spi/spi_imx.c~spi-imx-use-platform_driver_probe-as-probe-lives-in-inittext
>>> > +++ a/drivers/spi/spi_imx.c
>>> > @@ -666,13 +666,12 @@ static struct platform_driver spi_imx_dr
>>> >                   .name = DRIVER_NAME,
>>> >                   .owner = THIS_MODULE,
>>> >                   },
>>> > -       .probe = spi_imx_probe,
>>> >        .remove = __exit_p(spi_imx_remove),
>>>
>>> This looks wrong.  Why can't spi_imx_probe and spi_imx_remove be
>>> changed to be __devinit and __devexit?
>> I don't care much, but I thought this to be in accordance to
>>
>>        d1e44d9 (SPI driver runtime footprint shrinkage)
>>
>> (http://git.kernel.org/linus/d1e44d9).  I have a whole bunch of such
>> patches and already sent it three times (I think).  There is no
>> accordance in the community that one of the two possibilities is better
>> than the other.
>
> hmmm.  I don't like the pattern, but I'm not going to reject it out of
> hand.  Let me think about it for a bit.

But regardless, I'll make a decision before the end of the merge
window because I know it fixes an oops.

g.

-- 
Grant Likely, B.Sc., P.Eng.
Secret Lab Technologies Ltd.

------------------------------------------------------------------------------
Return on Information:
Google Enterprise Search pays you back
Get the facts.
http://p.sf.net/sfu/google-dev2dev