On Monday 22 December 2014 16:55:12 you wrote: > On Sun, 21 Dec 2014 robtongue@xxxxxxxxx wrote: > > On Friday 19 December 2014 17:22:42 you wrote: > > > > You'll have to collect a bunch of additional information. I'd like to > > > > see the entry in /sys/kernel/debug/usb/devices that describes the > > > > sound > > > > > > card, and the contents of the files under > > > > /sys/kernel/debug/usb/ehci/0000:00:1a.0/. > > > > > > /sys/kernel/debug/usb/devices: > > > - http://paste.ubuntu.com/9573759/ > > > > > > /sys/kernel/debug/usb/ehci/0000:00:1a.0/* > > > > > > - http://paste.ubuntu.com/9573789/ > > > > > > > I'd also like to see the dmesg log for when you try the test program, > > > > > > http://paste.ubuntu.com/9573956/ > > > > > > > and a usbmon trace as well. This will require CONFIG_DEBUG_FS, > > > > > > http://paste.ubuntu.com/9573950/ ran a few different things. > > > > > > > Can you do all this under a 3.18 or 3.17 kernel? > > > > > > did it with 3.18.1-gentoo > > > > > > Please excuse the ubuntu pastebins as it is my favorite one. ;) > > > > > > Thanks! > > > > > > -phunyguy > > > > Apparently I am having email client issues as well.... It put Alan Stern's > > name in the FROM box. Not even sure how that is possible. Sorry about > > that. > > The numbers in the bandwidth and periodic files make no sense. I'm > going to have to check on this in detail after the holidays. > > In the meantime, please try the patch below to see if it makes any > difference. Let's see what shows up in the bandwidth file and in the > dmesg log. And this time, include the log starting from when the > device is first plugged in, not from when you start running the test. > > By the way, does your .config enable CONFIG_USB_EHCI_TT_NEWSCHED? If > it doesn't, it should. > > Alan Stern > > > > Index: usb-3.18/drivers/usb/host/ehci-sched.c > =================================================================== > --- usb-3.18.orig/drivers/usb/host/ehci-sched.c > +++ usb-3.18/drivers/usb/host/ehci-sched.c > @@ -892,7 +892,9 @@ static int qh_schedule(struct ehci_hcd * > > for (i = qh->ps.bw_period; i > 0; --i) { > frame = ++ehci->random_frame & (qh->ps.bw_period - 1); > - for (uframe = 0; uframe < 8; uframe++) { > + /* Prefer later microframes over earlier ones */ > + for (uframe = 8; uframe > 0;) { > + --uframe; > status = check_intr_schedule(ehci, > frame, uframe, qh, &c_mask, tt); > if (status == 0) Ok I tried that kernel option first, and that actually worked without issue. I then tried your patch without that kernel option set, and it also worked. Finally I tried that kernel option =y with your patch, and it worked as well. So with that said, is there any need to provide further logs as without the patch seems to work? If so I will provide. As always, thank you for your time. -phunyguy -- 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
