Awesome analyzing I have to say! Thank you very much Sergey you really did it! Increasing the timeout did solve the problem. With the higher timeout the device responded after 7 seconds :) Under Windows I didn't recognise the longer time to detection, but I think that 5 sec timeout is VERY short. Without watching exactly this detail I would not be able to notice the difference between 5 or 7 seconds. After all I needed to crawl below my desk to plug it into the windows machine, causing an additional user delay to reach the monitor again ;)) Best, Daniel On Tue, Feb 3, 2009 at 21:34, Sergey Vlasov <vsu@xxxxxxxxxxx> wrote: > On Mon, Feb 02, 2009 at 09:04:06PM +0100, Daniel Eckl wrote: >> Dear list, dear Linux USB Developers, >> >> on the first sight my question seems to be the same like thousands out >> there which normally are "damaged device" or "broken cabeling", but I >> promise my case is somewhat special. >> >> I have a USB Flash drive (Corsair Voyager 8GB) which long time worked >> fine on 3 different PC running Ubuntu (first 8.04, now 8.10). >> >> Today I plugged the device into my laptop, filled it with data and >> later on gave the device to a colleague who plugged it into his Ubuntu >> 8.10. The device wasn't recognised (error messages and debug stuff >> follows below). >> >> Since then it doesn't work anymore on my laptop where I filled it this >> morning, nor on any other Linux machine I tried it. >> >> One would say, okay, the stick is now dead, got too much voltage or >> such, simple as that. But interestingly on Windows it still works >> fine. 2 of the 3 PCs I have in reach are dual boot and as soon as I >> boot into windows, the stick works like a charm! > > But is it detected with a noticeable delay after insertion? > >> So somehow the one Ubuntu system changed something on the stick which >> broke functionality with linux only. > [...] >> Feb 2 19:48:07 azure kernel: [ 435.092045] usb 7-1: new high speed >> USB device using ehci_hcd and address 2 >> Feb 2 19:48:07 azure kernel: [ 435.259426] usb 7-1: configuration #1 >> chosen from 1 choice >> Feb 2 19:48:08 azure kernel: [ 435.386281] usbcore: registered new >> interface driver libusual >> Feb 2 19:48:08 azure kernel: [ 435.417478] Initializing USB Mass >> Storage driver... >> Feb 2 19:48:08 azure kernel: [ 435.418846] scsi5 : SCSI emulation >> for USB Mass Storage devices >> Feb 2 19:48:08 azure kernel: [ 435.419824] usbcore: registered new >> interface driver usb-storage >> Feb 2 19:48:08 azure kernel: [ 435.420662] USB Mass Storage support >> registered. >> Feb 2 19:48:18 azure kernel: [ 446.032074] usb 7-1: reset high speed >> USB device using ehci_hcd and address 2 >> Feb 2 19:48:49 azure kernel: [ 476.580159] usb 7-1: reset high speed >> USB device using ehci_hcd and address 2 >> Feb 2 19:49:19 azure kernel: [ 507.124586] usb 7-1: reset high speed >> USB device using ehci_hcd and address 2 >> Feb 2 19:49:30 azure kernel: [ 517.644073] usb 7-1: reset high speed >> USB device using ehci_hcd and address 2 >> Feb 2 19:49:40 azure kernel: [ 528.052795] usb 7-1: USB disconnect, address 2 >> Feb 2 19:49:40 azure kernel: [ 528.056732] scsi 5:0:0:0: Device >> offlined - not ready after error recovery > > Your logs and usbmon output suggest that the device is too slow to > respond to the INQUIRY command. The default timeout for this command > in Linux is 5 seconds, but it can be increased: > > echo 20 > /sys/module/scsi_mod/parameters/inq_timeout > > The new value takes effect on subsequently detected devices - no need > to reboot or reload modules. > > The A-Data PD16 flash drive also has a similar problem - initially it > works even with the default timeout, but after some use it becomes > slower to initialize (last time I checked, even 10 second timeout was > not enough, with 20 it did not fail yet). > > -----BEGIN PGP SIGNATURE----- > Version: GnuPG v1.4.7 (GNU/Linux) > > iD8DBQFJiKpqW82GfkQfsqIRAgFtAJ93AIZsY4wI97f55opDubqhi/rH3QCdHk2U > 8dTAKxAZ0A7YMx67sWNkpEo= > =v46H > -----END PGP SIGNATURE----- > > -- 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
