Hi all, Provided that no objections have appeared for last few days, I'm applying the patch to the for-next branch of linux-leds.git. Thanks, Jacek Anaszewski On 08/17/2017 10:56 PM, Jacek Anaszewski wrote: > In case platform data provided empty LED name string the resulting > LED class device name would be crippled. Use corresponding LED chip > bit in place of "function" segment of LED class device name then to > make the LEDs at least distinguishable. > > Signed-off-by: Jacek Anaszewski <jacek.anaszewski@xxxxxxxxx> > Reported-by: Colin King <colin.king@xxxxxxxxxxxxx> > Reported-by: Dan Carpenter <dan.carpenter@xxxxxxxxxx> > Suggested-by: Nate Case <ncase@xxxxxxxxxxx> > --- > drivers/leds/leds-pca955x.c | 12 ++++++++---- > 1 file changed, 8 insertions(+), 4 deletions(-) > > diff --git a/drivers/leds/leds-pca955x.c b/drivers/leds/leds-pca955x.c > index f062d1e..09303fd 100644 > --- a/drivers/leds/leds-pca955x.c > +++ b/drivers/leds/leds-pca955x.c > @@ -473,10 +473,14 @@ static int pca955x_probe(struct i2c_client *client, > * Platform data can specify LED names and > * default triggers > */ > - if (pdata->leds[i].name) > - snprintf(pca955x_led->name, > - sizeof(pca955x_led->name), "pca955x:%s", > - pdata->leds[i].name); > + if (pdata->leds[i].name[0] == '\0') > + snprintf(pdata->leds[i].name, > + sizeof(pdata->leds[i].name), "%d", i); > + > + snprintf(pca955x_led->name, > + sizeof(pca955x_led->name), "pca955x:%s", > + pdata->leds[i].name); > + > if (pdata->leds[i].default_trigger) > pca955x_led->led_cdev.default_trigger = > pdata->leds[i].default_trigger; >
