07.09.2020 11:13, Andy Shevchenko пишет: > On Sun, Sep 6, 2020 at 9:51 PM Dmitry Osipenko <digetx@xxxxxxxxx> wrote: >> >> Use a single reset_control_reset() instead of assert/deasset couple in >> order to make code cleaner a tad. Note that the reset_control_reset() >> uses 1 microsecond delay instead of 2 that was used previously, but this >> shouldn't matter. In addition don't ignore potential error of the reset >> control by emitting a noisy warning if it fails, which shouldn't ever >> happen in practice. > > Still it's not clear if you check the datasheet or not. Some > elaboration would be good to have. I'll update the commit message with more details. Thanks! > ... > >> + WARN_ON_ONCE(err); > > Why screaming here? Wouldn't be dev_warn() enough? The error condition is an indicator of a severe problem because the reset shouldn't ever fail in practice, hence screaming is a preferred behavour. I'll a add a comment to the code, telling this.
