kern/115614: Recent changes in atapi-cam?

[Thomas Quinot]

* Patrick Hajek, 2007-09-11 :

> Commenting out atapicam_load="YES" in loader.conf and not precompiled it in
> the kernel, the errors disappear.
> 
> dmesg shows:
> 
> d0: 95205MB <TOSHIBA MK1032GAX AB211A> at ata0-master UDMA100
> acd0: DVDR <MATSHITADVD-RAM UJ-840S/1.52> at ata1-master UDMA33
> 
> No sense errors. So, it appears to point to the atapi-cam module.

Don't jump to conclusions hastily. Just because an error shows up only
when ATAPI/CAM is loaded doesn't mean it's necessarily a bug there.

> Suggestions?

It looks to me like this is similar to the issue described in
kern/115614. In general, it's a good habit to first check for an already
open PR when you encounter a problem, and to open one if there isn't an
existing one. This ensure that the issue and its resolution are properly
tracked and do not fall into the cracks. It is all to easy to miss or
lose track of an untracked discussion on -current.

To help investigate this issue, it would be useful to identify as
precisely as possible at what point of 6.2-STABLE it appeared.

Also, boot -v output and CAM debugging traces (see "man 4 cam") will be
useful. Please preserve subject line and include
bug-followup at freebsd.org in all correspondence to ensure a proper audit
trail. Thanks!

Thomas.