Hi Dave,
I had a chance to take look at your patch(es). They didn't apply cleanly on my tree, and I noticed some issues ...
What kind of issues? You mean the lack of locking in devio.c and friends? That's mostly taken care of in my usbfs patches (which need to go along with this patch, to avoid instant deadlocks in devio.c when disconnect is called).
Well, the driver model locking issues in particular. Also the issue I mentioned with the set_configuration diagnostics being issued at the wrong time. Plus there was one deadlock I saw with "usbtest".
I like the idea of having the driver model issues resolved before the ones relating to dev->serialize or ps->devsem, strictly as a layering issue -- it should help dis-entangle the problems better.
Given that usbfs has so many locking problems, there needs to be some strategy for phasing in the fixes. Does that one make for problems I didn't anticipate?
so here's an updated version, without the driver-specific patches.
Great. I will try to get everything together and send it this weekend.
I merged your earlier usbnet patches into my tree, FWIW. It was good to finally see /sys listing all the interfaces claimed by that driver, instead of only half of them!
- Dave
All the best,
Duncan.
------------------------------------------------------- The SF.Net email is sponsored by EclipseCon 2004 Premiere Conference on Open Tools Development and Integration See the breadth of Eclipse activity. February 3-5 in Anaheim, CA. http://www.eclipsecon.org/osdn _______________________________________________ [EMAIL PROTECTED] To unsubscribe, use the last form field at: https://lists.sourceforge.net/lists/listinfo/linux-usb-devel
