Re: [PATCH] thermal: cpu_cooling: fix ptr_ret.cocci warnings

[Javi Merino]

On Thu, Mar 05, 2015 at 04:46:13AM +0000, Eduardo Valentin wrote:
> On Thu, Mar 05, 2015 at 04:34:02AM +0800, kbuild test robot wrote:
> > drivers/thermal/cpu_cooling.c:463:18-24: WARNING: PTR_ERR_OR_ZERO can be used
> > 
> > 
> >  Use PTR_ERR_OR_ZERO rather than if(IS_ERR(...)) + PTR_ERR
> > 
> > Generated by: scripts/coccinelle/api/ptr_ret.cocci
> > 
> > CC: Javi Merino <javi.merino@xxxxxxx>
> > Signed-off-by: Fengguang Wu <fengguang.wu@xxxxxxxxx>
> > ---
> > 
> >  cpu_cooling.c |    2 +-
> >  1 file changed, 1 insertion(+), 1 deletion(-)
> > 
> > --- a/drivers/thermal/cpu_cooling.c
> > +++ b/drivers/thermal/cpu_cooling.c
> > @@ -460,7 +460,7 @@ static int get_static_power(struct cpufr
> >  	if (voltage == 0) {
> >  		dev_warn_ratelimited(cpufreq_device->cpu_dev,
> >  				     "Failed to get voltage for frequency %lu: %ld\n",
> > -				     freq_hz, IS_ERR(opp) ? PTR_ERR(opp) : 0);
> > +				     freq_hz, PTR_ERR_OR_ZERO(opp));
> 
> This patch causes this compilation warning.
> 
> In file included from include/linux/device.h:27:0,
>                  from include/linux/thermal.h:30,
>                  from drivers/thermal/cpu_cooling.c:26:
> drivers/thermal/cpu_cooling.c: In function âget_static_powerâ:
> include/linux/ratelimit.h:31:9: warning: format â%ldâ expects argument of type âlong intâ, but argument 4 has type âintâ [-Wformat=]
>   struct ratelimit_state name =     \
>          ^
> include/linux/device.h:1162:9: note: in expansion of macro âDEFINE_RATELIMIT_STATEâ
>   static DEFINE_RATELIMIT_STATE(_rs,    \
>          ^
> include/linux/device.h:1178:2: note: in expansion of macro âdev_level_ratelimitedâ
>   dev_level_ratelimited(dev_warn, dev, fmt, ##__VA_ARGS__)
>   ^
> drivers/thermal/cpu_cooling.c:463:3: note: in expansion of macro âdev_warn_ratelimitedâ
>    dev_warn_ratelimited(cpufreq_device->cpu_dev,
>    ^

PTR_ERR_OR_ZERO() has the wrong signature in my opinion.  PTR_ERR()
returns a long but PTR_ERR_OR_ZERO() returns an int.  I think
PTR_ERR_OR_ZERO() should return a long so that it can be a truly
replacement for "IS_ERR(x) ? PTR_ERR(x) : 0".  It's probably not
worth the churn though.

Find the correct fix below.  Do you want me to send a proper patch?

---8<---
--- a/drivers/thermal/cpu_cooling.c
+++ b/drivers/thermal/cpu_cooling.c
@@ -461,8 +461,8 @@ static int get_static_power(struct cpufreq_cooling_device *cpufreq_device,
 
 	if (voltage == 0) {
 		dev_warn_ratelimited(cpufreq_device->cpu_dev,
-				     "Failed to get voltage for frequency %lu: %ld\n",
-				     freq_hz, IS_ERR(opp) ? PTR_ERR(opp) : 0);
+				     "Failed to get voltage for frequency %lu: %d\n",
+				     freq_hz, PTR_ERR_OR_ZERO(opp));
 		return -EINVAL;
 	}
 

--
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/