On Wed, 2 Nov 2011, Anton Alekseev wrote: > > Still no better. > > > > All right, two more things to try. Set the delays back to 9 * 125 and > > replace the previous patch with the updated version below. If it still > > doesn't work, try an extreme test: Set both delays to a much larger > > value such as 20000. > > Both solutions didn't help. Sending you dmesg for second solution > (long delays). Actually it did help. The messages in the log are now different: [ 1131.655074] ehci_hcd 0000:00:12.2: start disable periodic [ 1131.659056] ehci_hcd 0000:00:12.2: disable periodic 0 [ 1131.663035] ehci_hcd 0000:00:12.2: start enable periodic [ 1131.666990] ehci_hcd 0000:00:12.2: enable periodic 0 [ 1131.694712] cannot submit datapipe for urb 2, error -27: internal error You didn't get the "force halt; handshake..." message. It looks like the periodic schedule can take as long as 8 ms to stop running! Error -27 is -EFBIG, which means the driver tried to submit too much data for ehci-hcd to handle. > Even when I initialize my USB soundcard with parameters: 48000 Hz and > buffer_size 1024 samples I get XRUNs. The overruns are probably caused by the 8 ms latency, although the driver should be able to cope with that. > Am I first ATI EHCI owner with such problems? You're the first I have heard of. Ming Lei seems to have found another. Alan Stern -- 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
