cvs commit: src/sys/cam/scsi scsi_da.c src/sys/dev/usb umass.c usbdevs

Florent Thoumie flz at xbsd.org
Mon Jan 30 17:51:36 PST 2006


On Tuesday 31 January 2006 02:00, Nate Lawson wrote:
> Florent Thoumie wrote:
> > On Tuesday 31 January 2006 01:24, Nate Lawson wrote:
> >>I'm not concerned about the revision.  I'm concerned about the vendor
> >>(Generic*) and device name (STORAGE DEVICE*).  Why are the *'s needed?
> >
> > Seemed common practice reading the other entries.
>
> No, that's definitely not it.  In fact, the most recent entries should
> be audited to see if they really need the *'s.  This wildcard might
> overly match the wrong devices.
>
> >>(Again, a PR would help track this kind of conversation as shown in
> >>previous PRs about quirks.  Submitters often match way too much.)
> >>
> >>>Do you want me to create a PR just for tracking purposes?
> >>>[1] http://docs.freebsd.org/cgi/mid.cgi?20060116193024.GA95183
> >>
> >>That would be nice, especially since some of the requested info is
> >>missing (dmesg, usbdevs -v).  However, if you cited a email in the
> >>commit msg (maybe SMTP Message-ID) such that we could find it in the
> >>future, that would probably be enough.  I'm not trying to create a
> >>bureaucracy, just make sure we don't lose information like we used to on
> >>why a quirk was added in the first place.
> >
> > I only mentioned the freebsd-usb mailing list. I'll contact Anders to get
> > additional details and I (or he) will fill a PR so that we can add it to
> > the comment.
>
> Thanks.
>
> > It seems a lot of devices are concerned by the sync cache problem, would
> > it be harmful to just remove this part of the code or could there be a
> > way to detect if the device supports it or not?
>
> Well, it's important to run SYNC_CACHE in shutdown or possibly when
> unmounting a filesystem.  Otherwise, data could be lost on boot.
> However, I support adding a USB-specific mechanism that says SYNC_CACHE
> should only be run on shutdown or device_eject, that way devices that
> hang after this command is run would still work at runtime.  And SCSI
> devices that support multiple calls to SYNC_CACHE (i.e. most non-USB
> devs) would still work too.
>
> However, the first step is to investigate what windows and Linux do.

Linux only sends the SYNCHRONIZE_CACHE command if the WCE (Write Cache Enable) 
bit of the disk is set. I can't seem to find something equivalent to this in 
our CAM framework. I have no particular SCSI knowledge but I guess I can have 
a look at this tomorrow.

I'm forwarding this to freebsd-scsi@ (keep me CC'ed, as I'm not subscribed to 
this list, yet).

-- 
Florent Thoumie
flz at FreeBSD.org
FreeBSD Committer
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 187 bytes
Desc: not available
Url : http://lists.freebsd.org/pipermail/cvs-src/attachments/20060131/80db4971/attachment.bin


More information about the cvs-src mailing list