Skip to contentSkip to conversation

pinctrl: sunxi: Deal with configless pins

bug report Dan Carpenter 2 participants 3 messages Nov 16, 2016 – Nov 17, 2016 GPIO

lore
Share & tools
lore
#0 lore raw

Hello Maxime Ripard,

The patch e11dee2e98f8: "pinctrl: sunxi: Deal with configless pins" from Oct 20, 2016, leads to the following static checker warning:

drivers/pinctrl/sunxi/pinctrl-sunxi.c:397 sunxi_pctrl_dt_node_to_map()
warn: variable dereferenced before check 'map' (see line 323)

drivers/pinctrl/sunxi/pinctrl-sunxi.c
394 *num_maps = i;
395
396 /*

   397           * We know have the number of maps we need, we can resize our
   398           * map array
   399           */

400 *map = krealloc(*map, i * sizeof(struct pinctrl_map), GFP_KERNEL);
401 if (!map)

This should obviously be "if (*map) " here.

   402                  return -ENOMEM;
   403  
   404          return 0;
   405  
   406  err_free_map:
   407          kfree(map);

I would send a patch, but I wasn't sure about this. Should this be kfree(*map);? I don't have my cross function DB built for ARM so I don't know how this is called... Maybe set *map to NULL as well before returning?

408 return ret;
409 }

regards,
dan carpenter

#1 lore raw

Hi Dan,

Quoted 18 lines · Dan Carpenter

Hello Maxime Ripard,

The patch e11dee2e98f8: "pinctrl: sunxi: Deal with configless pins" from Oct 20, 2016, leads to the following static checker warning:

drivers/pinctrl/sunxi/pinctrl-sunxi.c:397 sunxi_pctrl_dt_node_to_map()
warn: variable dereferenced before check 'map' (see line 323)

drivers/pinctrl/sunxi/pinctrl-sunxi.c

   394          *num_maps = i;
   395  
   396          /*
   397           * We know have the number of maps we need, we can resize our
   398           * map array
   399           */
   400          *map = krealloc(*map, i * sizeof(struct pinctrl_map), GFP_KERNEL);
   401          if (!map)
Quoted 1 line

This should obviously be "if (*map) " here.

Erm, yes, totally.

Quoted 10 lines
   402                  return -ENOMEM;
   403  
   404          return 0;
   405  
   406  err_free_map:
   407          kfree(map);

I would send a patch, but I wasn't sure about this. Should this be kfree(*map);?

Yes, that one too.

Quoted 3 lines

I don't have my cross function DB built for ARM so I
don't know how this is called... Maybe set *map to NULL as well before returning?

And that would be even better yes :)

Will you send a patch?

Thanks,
Maxime

Signature · 3 lines
Maxime Ripard, Free Electrons
Embedded Linux and Kernel engineering
http://free-electrons.com
#2 lore raw
Quoted 1 line · Maxime Ripard

Will you send a patch?

Sure. I'll send it tomorrow. Thanks.

regards,
dan carpenter