Hi Dragan,
Sure, but I think that having such an additional debug facility
can only help and save the people from adding temporary printk()s
while debugging.
Mileages, I guess. I like temporary printouts for temporary debug
sessions. This way the upstream source code stays slim. In my experience
with I2C over the years, debugging happens with developers anyhow.
Logfiles from users which help developers create patches are very rare.
And those users are usually capable enough to add debug patches.
Given these experiences (which may be different in other subsystems), I
don't see why we should carry the bloat.
That said, I'll leave the final decision to the driver maintainers.