On Thu, Jul 09, 2009 at 03:33:41PM -0700, Steve Calfee wrote: > I cannot actually give a vid/pid of a device that hangs. I may be > wrong. I have merely observed the usb bus with a catc under windows 98 > and xp, and they don't keep asking for strings once fetch for string 0 > fails. (After all, a gadget without strings are supposed to have > string indexes of 0.) Gadget bugs always lurk. You're right. If the LANGID transaction stalls, there shouldn't be any string get requests to that device. > However, if Linux has been working asking for strings without a langid > (therefore that don't exist) and failing with stalls, that probably is > good enough proof that I am wrong and the "old" Linux way was correct. > I don't like the idea of adding flags to solve "potential" problems. See my last patch. That should both fix the current regression and the unnecessary additional LANGID get requests which occured before my patch. Daniel -- 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
