On 06/27/12 17:45, Boris Samorodov wrote:
27.06.2012 15:51, Alexander Motin пишет:

The only change after that I see potentially related is r237478. It adds
more checks when fetching SCSI sense data, that for some reason are not
working in your case. I still can not completely understand why there
was no any READ CAPACITY errors reported before, but may be I am missing
something. You can try to revert that revision for check.

Confirm. Reverting this commit alone helps here. Both my system with
patched kernel uses /dev/da0 and patched kernel works if the system
is booted from the stick.

OK. But I am not sure what to do about it. I don't see problem in my code. I believe it is either hardware or umass problem, or both.

Though it is a little bit noisy. ;-) Since now I know that it shouldn't
here is a question: should I file a PR on this noisiness (i.e. error
reporting, etc.)?

These are real errors for CAM. There would be no noise if device correctly reported SCSI senses, as drivers are already instructed to not wine about unsupported command codes. But in this case CAM doesn't know what are the errors and prefers to report them.

Alexander Motin
freebsd-usb@freebsd.org mailing list
To unsubscribe, send any mail to "freebsd-usb-unsubscr...@freebsd.org"

Reply via email to