Re: linux-next: suspend tree build warnings

[Andy Walls]

On Wed, 2009-08-19 at 16:36 -0700, Greg KH wrote:
> On Wed, Aug 19, 2009 at 11:38:03PM +0200, Rafael J. Wysocki wrote:
> > On Wednesday 19 August 2009, Stephen Rothwell wrote:
> > > Hi Rafael,
> > 
> > Hi,
> > 
> > > Today's linux-next build (x86_64 allmodconfig) produced these warnings:
> > > 
> > > drivers/media/dvb/frontends/dib7000p.c: In function âdib7000p_i2c_enumerationâ:
> > > drivers/media/dvb/frontends/dib7000p.c:1315: warning: the frame size of 2256 bytes is larger than 2048 bytes
> > > drivers/media/dvb/frontends/dib3000mc.c: In function âdib3000mc_i2c_enumerationâ:
> > > drivers/media/dvb/frontends/dib3000mc.c:853: warning: the frame size of 2160 bytes is larger than 2048 bytes
> > > 
> > > Introduced by commit 99307958cc9c1b0b2e0dad4bbefdafaf9ac5a681 ("PM:
> > > Introduce core framework for run-time PM of I/O devices (rev. 17)").
> > 
> > Well.
> > 
> > This commit increases the size of struct device quite a bit and both of the
> > drivers above create a "state" object on the stack that contains struct device
> > among other things.
> 
> Ick.  struct device should _never_ be on the stack, why would this code
> want to do such a thing?

It appears that the state object is a dummy being used to detect and
twiddle some identical chips on the i2c bus.  The functions called only
use the "i2c_adapter" and "cfg" member of the dummy state object, but
those functions want that state object as an input argument.

<obvious>
The simplest fix is dynamic allocation of the dummy state object with
kmalloc() and then to free it before exiting the function.
</obvious>

Regards,
Andy


> thanks,
> 
> greg k-h


--
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/