Al Viro wrote:
On Wed, Feb 22, 2006 at 08:08:14AM +0100, Stefan Richter wrote:
...these 6 lines here? We would miss what can be extracted from these
buggy devices but (a) sd_read_cache_type::bad_sense's defaults work IMO
well enough for the so far reported devices
Yeah, right. The same bridge loses its cache contents on reboot.
I.e. missed cache type => dirty fs on every reboot, with actual
corruption in case if there was any recent activity. So... No, thanks.
It might make sense to split that in two patches, but if you really
think that defaults work... I've seen several enclosures where they
don't (different types, at that).
Well, right.
BTW, I never noticed such kind of breakage myself yet. I almost always
break connection not by reboot but by hot-unplug while sbp2 is still
logged in. This would corrupt data too --- *if* these devices really had
a write cache and did not flush it themthelves at some point. Maybe
self-powered devices actually do this at each FireWire bus reset. But
bus powered devices cannot.
--
Stefan Richter
-=====-=-==- --=- =-==-
http://arcgraph.de/sr/
-
: send the line "unsubscribe linux-scsi" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
