On Wednesday 23 March 2005 22:57, Johannes Stezenbach wrote: > Tony Glader wrote: > > Is there some good reason why there is mdelay(10) in ves1820_writereg(): > > That is a good question. > > > static int ves1820_writereg(struct ves1820_state *state, u8 reg, u8 data) <nip> > > msleep(10); <snip> > > Doesn't it slow down lot of i2c communication? > > Yes, but maybe that's the intention (bad hardware). > Does it work for you without the sleep? It feels somewhat large tho, don't it Johannes? If some hardware would require a delay there it would be because of a reset or some such (with special time constraints) was not handeled very well. A verification of the driver from it's datasheet is required! Kenneth
