> From: Alan Stern [mailto:stern@xxxxxxxxxxxxxxxxxxx] > Sent: Thursday, July 26, 2012 12:32 PM > > On Thu, 26 Jul 2012, Paul Zimmerman wrote: > > > > That's very clear. So Rajaram, when you redo the patch you can simply > > > leave out the checks for SuperSpeed. Set or clear the short_not_ok > > > flag depending only on the size of the bulk transfer, regardless of the > > > speed. > > > > I have a bit of a philosophical problem with that. At SuperSpeed, a short > > packet has to be OK, because of the way the mass-storage protocol works. > > So it seems wrong to set short_not_ok for SuperSpeed when we know we > > will be receiving short packets. > > > > But I guess practicality trumps philosophy :) > > Hmmm. I suppose an impartial approach would be to set short_not_ok > only when the blocksize >= the maxpacket size. That should always work > as a general rule, regardless of speed. Yes. And that might save us from having to rehash all this when HyperSpeed USB (or whatever is next) comes along in a few years. Don't know how difficult that would be to implement, though. Would it just be a matter of comparing fsg_lun.blksize to fsg_common. bulk_out_maxpacket? -- Paul -- To unsubscribe from this list: send the line "unsubscribe linux-usb" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
