On Sun, 30 Nov 2003, matthias geissert wrote: > > Looks like what you forwarded was a log from 2.4.something. Can you > Yes, that's right. > > forward the corresponding output from 2.6.0-test11 after enabling both > > USB mass storage verbose debug (to show the higher level protocol), and > > USB debug (to show if anything's going wrong at the lower levels)?
> Ok, here's dmesg of kernel 2.6.0test11. I firstly loaded the ohci-driver > and copy some data. There wasen't any error during copy. Afterwards used > the ehci driver. Mounting seems to work fine. If I copy some data the > DVD device starts reading for a few seconds, then stops for a while and > starts reading again. This time I checked the copied files. It seems > that even the ehci driver copies the data but it takes much longer than > the ohci driver (due to the stops of the drive and reseting the bus). > > If you need more information I would be delighted to help you. > > Thanks > matthias > > p.s. I won't read my emails till next Friday. >From the log it's easy to see why the transfer took so long. There were 6 occurrences of this sort of thing: > Nov 30 13:29:10 nb09 kernel: usb-storage: queuecommand called > Nov 30 13:29:10 nb09 kernel: usb-storage: *** thread awakened. > Nov 30 13:29:11 nb09 kernel: usb-storage: Command READ_10 (10 bytes) > Nov 30 13:29:11 nb09 kernel: usb-storage: 28 00 00 00 4f d1 00 00 34 00 > Nov 30 13:29:11 nb09 kernel: usb-storage: Bulk Command S 0x43425355 T 0xda L 106496 > F 128 Trg 0 LUN 0 CL 12 > Nov 30 13:29:11 nb09 kernel: usb-storage: usb_stor_bulk_transfer_buf: xfer 31 bytes > Nov 30 13:29:11 nb09 kernel: usb-storage: Status code 0; transferred 31/31 > Nov 30 13:29:11 nb09 kernel: usb-storage: -- transfer complete > Nov 30 13:29:11 nb09 kernel: usb-storage: Bulk command transfer result=0 > Nov 30 13:29:11 nb09 kernel: usb-storage: usb_stor_bulk_transfer_sglist: xfer 106496 > bytes, 26 entries > Nov 30 13:29:11 nb09 kernel: usb-storage: Status code 0; transferred 106496/106496 > Nov 30 13:29:11 nb09 kernel: usb-storage: -- transfer complete > Nov 30 13:29:11 nb09 kernel: usb-storage: Bulk data transfer result 0x0 > Nov 30 13:29:11 nb09 kernel: usb-storage: Attempting to get CSW... > Nov 30 13:29:11 nb09 kernel: usb-storage: usb_stor_bulk_transfer_buf: xfer 13 bytes > Nov 30 13:29:11 nb09 kernel: usb-storage: Status code 0; transferred 0/13 > Nov 30 13:29:11 nb09 kernel: usb-storage: -- short transfer > Nov 30 13:29:11 nb09 kernel: usb-storage: Bulk status result = 1 The CSW transfer returned 0 bytes, not 13. Another Genesyslogic bug? > Nov 30 13:29:11 nb09 kernel: usb-storage: -- transport indicates error, resetting > Nov 30 13:29:11 nb09 kernel: usb-storage: usb_stor_Bulk_reset called > Nov 30 13:29:11 nb09 kernel: usb-storage: usb_stor_control_msg: rq=ff rqtype=21 > value=0000 index=00 len=0 > Nov 30 13:29:16 nb09 kernel: usb-storage: Soft reset: clearing bulk-in endpoint halt Although this looks normal, other instances of Bulk resets in the log do not show this 6-second delay. I don't understand that; the delay is built into the usb-storage driver. > Nov 30 13:29:16 nb09 kernel: usb-storage: usb_stor_control_msg: rq=01 rqtype=02 > value=0000 index=81 len=0 > Nov 30 13:29:16 nb09 kernel: usb-storage: usb_stor_clear_halt: result = 0 > Nov 30 13:29:16 nb09 kernel: usb-storage: Soft reset: clearing bulk-out endpoint halt > Nov 30 13:29:16 nb09 kernel: usb-storage: usb_stor_control_msg: rq=01 rqtype=02 > value=0000 index=02 len=0 > Nov 30 13:29:16 nb09 kernel: usb-storage: usb_stor_clear_halt: result = 0 > Nov 30 13:29:16 nb09 kernel: usb-storage: Soft reset done > Nov 30 13:29:16 nb09 kernel: usb-storage: scsi cmd done, result=0x70000 > Nov 30 13:29:16 nb09 kernel: usb-storage: *** thread sleeping. > Nov 30 13:29:16 nb09 kernel: usb-storage: queuecommand called > Nov 30 13:29:16 nb09 kernel: usb-storage: *** thread awakened. > Nov 30 13:29:16 nb09 kernel: usb-storage: Command READ_10 (10 bytes) > Nov 30 13:29:16 nb09 kernel: usb-storage: 28 00 00 00 4f d1 00 00 34 00 The SCSI layer retries the command. > Nov 30 13:29:16 nb09 kernel: usb-storage: Bulk Command S 0x43425355 T 0xdb L 106496 > F 128 Trg 0 LUN 0 CL 12 > Nov 30 13:29:16 nb09 kernel: usb-storage: usb_stor_bulk_transfer_buf: xfer 31 bytes > Nov 30 13:29:16 nb09 kernel: usb-storage: Status code 0; transferred 31/31 > Nov 30 13:29:16 nb09 kernel: usb-storage: -- transfer complete > Nov 30 13:29:16 nb09 kernel: usb-storage: Bulk command transfer result=0 > Nov 30 13:29:16 nb09 kernel: usb-storage: usb_stor_bulk_transfer_sglist: xfer 106496 > bytes, 26 entries > Nov 30 13:29:16 nb09 kernel: usb-storage: Status code 0; transferred 106496/106496 > Nov 30 13:29:16 nb09 kernel: usb-storage: -- transfer complete > Nov 30 13:29:16 nb09 kernel: usb-storage: Bulk data transfer result 0x0 > Nov 30 13:29:16 nb09 kernel: usb-storage: Attempting to get CSW... > Nov 30 13:29:16 nb09 kernel: usb-storage: usb_stor_bulk_transfer_buf: xfer 13 bytes > Nov 30 13:29:16 nb09 kernel: usb-storage: Status code 0; transferred 13/13 > Nov 30 13:29:16 nb09 kernel: usb-storage: -- transfer complete > Nov 30 13:29:16 nb09 kernel: usb-storage: Bulk status result = 0 > Nov 30 13:29:16 nb09 kernel: usb-storage: Bulk Status S 0x53425355 T 0xdb R 0 Stat > 0x0 > Nov 30 13:29:16 nb09 kernel: usb-storage: scsi cmd done, result=0x0 > Nov 30 13:29:16 nb09 kernel: usb-storage: *** thread sleeping. And the second time it worked, fortunately. The same thing happened each time. Those long reset delays will destroy your throughput, but at least the data is getting transferred without error. Alan Stern ------------------------------------------------------- This SF.net email is sponsored by: SF.net Giveback Program. Does SourceForge.net help you be more productive? Does it help you create better code? SHARE THE LOVE, and help us help YOU! Click Here: http://sourceforge.net/donate/ _______________________________________________ [EMAIL PROTECTED] To unsubscribe, use the last form field at: https://lists.sourceforge.net/lists/listinfo/linux-usb-devel
