10.03.2020 14:37, Wolfram Sang пишет: > On Mon, Mar 02, 2020 at 08:35:12PM +0300, Dmitry Osipenko wrote: >> The I2C timeout error message doesn't tell us what exactly failed and some >> I2C client drivers do not clarify the error either. Adding WARN_ON_ONCE() >> results in a stacktrace being dumped into KMSG, which is very useful for >> debugging purposes. > > This is good for debugging, in deed, yet not good in the generic case. > Timeouts are not an exception on the I2C bus (think of an EEPROM which > is busy during an erase cycle), so it shouldn't be printed at all. > > This prinout should rather be dropped or at least be dev_dbg. Oh, well. I'll keep this debugging applied locally then, it's quite unfortunate when something fails silently :)
