On Fri, Mar 20, 2009 at 10:34:55AM +0100, Daniel Mack wrote: > On Thu, Mar 19, 2009 at 08:32:21PM -0700, Greg KH wrote: > > On Tue, Mar 17, 2009 at 03:52:17PM +0100, Daniel Mack wrote: > > > When an USB hardware does not provide a valid LANGID, fall back to value > > > zero which is still a reasonable default for most devices. > > > > But that's a broken device, right? What happens without this patch? > > It is broken, yes. Without the patch, both the vendor name and the > product name are swallowed away in kernel space and lsusb. So it doesn't show up at all anymore? Do you have a specific example of this? > > And has this device passed the usb.org test suite? > > Well, the thing is - this bug was always hidden from us because > > a) Windows doesn't care > b) Mac OS X doesn't care > c) Linux used to not care, too. Don't know when that was changed, but it > must have been in the last months. So it's kind of a regression, even > tough the current behaviour is just currect in terms of the USB spec. Can you do a little bit of bisection of different kernel versions to figure out where this changed? > Anyway, with this patch applied, things work fine and I see no reason > why complaining with a warning shouldn't be enough. Failing so hard is > annoying. True. If you let me know when this broke, feel free to resend the patch with that information and I'll be glad to apply it. I need to know when it broke so that we can backport it to older -stable kernel versions if needed as well to make them work again. thanks, greg k-h -- 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
