On Wed, Nov 30, 2016 at 5:49 PM, Linus Walleij <linus.wall...@linaro.org> wrote: > On Sun, Nov 27, 2016 at 11:51 AM, Jonathan Cameron <ji...@kernel.org> wrote: >> On 26/11/16 03:47, Aniroop Mathur wrote: > > [bmp280.c] > >>> /* Wait to make sure we started up properly */ >>> - mdelay(data->start_up_time); >>> + usleep_range(data->start_up_time, data->start_up_time + 100); >> >> As this in probe I doubt we really care. Could just set it longer to shut >> up the warnings. >> Still would like some input from say Linus on this... > > Hm, I don't think it's a big issue... this works too it just looks overworked. > > On the runtime_resume() path we use msleep() instead which I guess is why > it is not changed in this patch, but they have the same purpose. >
I did change msleep to usleep_range in runtime_resume() in bmp280.c as you know resume time is critical indeed. > Yours, > Linus Walleij
