lkml.org 
[lkml]   [2012]   [Feb]   [2]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectRe: [PATCH v2] add LED driver for PCA9663 I2C chip

> Andrew Morton <akpm@linux-foundation.org> wrote:
> > > +exit:
> > > + while (i--) {
> > > + led_classdev_unregister(&pca9633[i].led_cdev);
> > > + cancel_work_sync(&pca9633[i].work);
> > > + }
> > This (untested!) loop has an off-by-one error.

> Well, it's partly wrong. It's wrong for the cancel_work_sync() but not
> for the led_classdev_unregister(). The cancel_work_sync() can just be
> removed: we haven't scheduled any works yet.

there are 4 leds with get led_classdev_register()ed

wouldn't it be possible that someone is starting to use led1 while led3 has
not been registered yet? if registration for led3 fails, the exit code
(above) is called and there might be work scheduled

so (hypothetically):
register led1 -> ok
register led2 -> ok
use led1 (schedule work)
register led3 -> fail -> exit code

the code has been copied from other led drivers
I think there is no harm to keep cancel_work_sync()

p.

--

Peter Meerwald
+43-664-2444418 (mobile)

\
 
 \ /
  Last update: 2012-02-02 11:45    [from the cache]
©2003-2011 Jasper Spaans