On 3 April 2018 at 20:12, Wolfram Sang <wsa@xxxxxxxxxxxxx> wrote: > >> Yeah, good point. So when exactly should this be called? Only on error >> conditions that could be caused by a misbehaving slave? > > Chapter 3.1.16 of the I2C specification, only when a slave keeps SDA > low. Usually you detect that before you start transmitting. > > Really, please remove it for now and add it later, properly tested! > OK, I will remove it for now. > BTW what kind of tests did you do with this driver so far? > I have used it with a ATSHA204A crypto chip, using a userland driver and i2c-dev, and with a NXP PCF8563 RTC using the kernel driver.
