On Sat, 30 Mar 2013, Andy Lutomirski wrote: > On Sat, Mar 30, 2013 at 8:21 PM, Andy Lutomirski <luto@xxxxxxxxxxxxxx> wrote: > > I have a Corsair Flash Voyager GT and a Lenovo x220 running Fedora's > > 3.8.2 kernel. It doesn't work very well, and the problems vary > > depending on whether I'm using usb2 or usb3. > > > > - On usb3, the device node can't be opened with O_DIRECT (using, for > > example, dd oflag=direct). It returns -EINVAL. (On usb2, O_DIRECT > > works.) > > - Also on usb3, if I dd if=/dev/zero of=/dev/sdb bs=1M, dd finishes > > in a second or two and no I/O occurs. This system only has 2GB of RAM > > -- caching shouldn't be that extreme. > > *facepalm*. The issue is that one of the I/O errors caused the sdb > block device to get sufficiently wedged that it still exists in /sys, > but udev removed sdb (and I replaced it with a file). If I manually > recreate the sdb device node, ejecting it gets an IO error. (It's not > plugged in any more.) rmmoding and modprobing usb_storage got rid of > the bogus sdb device. Does this mean that none of the problems mentioned in the first email require any attention? What about the "sufficiently wedged" issue? 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
