On Dec 29, 2007, at 3:38 AM, Hans Petter Selasky wrote:
On Saturday 29 December 2007, Ben Kelly wrote:
It didn't look like there had been any USB changes in there
recently.  Has the USB development moved completely to p4?

Just for the time being. I will update my SVN repo soon so that you can get
the source code and compile for yourself.

Thanks!

I would like to try the new stack, but I am guessing I will need to
patch it for usb/95173 in order to use my umass device.

Usually it is best if patches go into FreeBSD current first, because else I will have to keep track of the patches I take in. Regarding your patch, it looks OK, except it would be nice if the patch could be modified so that it automatically detects these quirks. Because adding quirks is a bad solution
in the long run.

I agree that quirks are not the best solution. Unfortunately, though, I don't think it will be possible to detect the problem and dynamically disable sync'ing. At least this particular USB-to-disk controller seems to lockup when the sync command is issued. It requires a power cycle to recover the device.

It did occur to me, however, that perhaps it would make sense to disable sync'ing for all umass devices by default and then add a quirk to enable the feature on devices where its known to be beneficial. This is based on the (perhaps dubious) theory that most umass devices probably do not do a lot of caching because of the risk of being disconnected at any time. I didn't pursue this option, though, since I thought it would be more controversial and harder to get applied to the tree.

Maybe you have another USB mass storage device you can test ?

I think I can probably find one. Eventually I would like to test my hard drive enclosure, though, since I use it the most often currently. I probably won't be able to run the tests until after the new year.

Thanks again.

- Ben
_______________________________________________
freebsd-usb@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-usb
To unsubscribe, send any mail to "[EMAIL PROTECTED]"

Reply via email to