On Mon, 8 Jun 2009, Stefan Sassenberg wrote: > >> The device is not physically a card reader with a removable medium, it's > >> an 8GB stick. What's wrong with it? > > > > In short, what's wrong is the device is buggy. However you probably > > already know that... > > > > There's no way to tell exactly what's wrong from the information you > > provided. We need to see more. Can you provide a usbmon trace > > starting from the time you plug the device in and ending with the > > failed mount? > > > > And can you provide a similar trace showing it working with 2.6.24? > > I put the bad trace to <http://coolmule.de/bad.tar.gz>. When creating > the good trace on the 2.6.24 vm the fdisk also failed. I repeated the > fdisk command and noticed, that it sometimes works, sometimes not. I > made a 'watch "fdisk -l /dev/sdc"' for 40 seconds on the 2.6.29 machine > and had a success rate of 7 out of 20 with no special pattern. The same > watch on the 2.6.24 (sda) had 3 failures and then 17 good runs in a row. > If you'd like to see the trace of the 2.6.24 connect, > <http://coolmule.de/good.tar.gz>, but it's a lot shorter. Unfortunately the logs don't show the cause of the problem. There is a different pattern of sector accesses in the two systems, probably because they have different versions of udev and hal. That might explain why you see different error rates. Still, the errors do appear to come more or less at random. (For instance, even your "good" log contains a bunch of "Medium Not Present" errors near the start.) I can't tell why. It just looks like the stick is unreliable. Have you tried using this stick with other operating systems? 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
