On Mon, Mar 11, 2002 at 12:47:58PM -0500, Pete Zaitcev wrote: > Guys, > > The patch came about as offshot of my anti-usbkbd crusade. > (Or shall I say Vojtech's anti-usbkbd policies? :)
:))) > This patch is truly a kludge, but it seems to work. > Some hardware by a certain company called "ATEN" crashes > if GET_REPORT is sent. I tried to modify our GET_REPORT, but > it seems that nothing helps. So, I simply switched GET_REPORT off. Well, they have this "patented usb-sniffing technology", perhaps they really don't handle the GET_REPORT request at all ... > It works for the piece of junk that Red Hat bought for my testing, > and I got several "thank you" notes since we shipped the patch. > The usbkbd works because it never sends GET_REPORT. > > So, what do you think? I'm for including it. I'd prefer having one blacklist with the quirks (QUIRK_NOGET, QUIRK_IGNORE, ... etc), and I'll generate a patch for that. > @@ -1025,6 +1031,7 @@ > if (hid->outhead == hid->outtail) > hid->outtail = (hid->outtail + 1) & (HID_CONTROL_FIFO_SIZE - 1); > > + /* P3 XXX buggy - use separate flag. */ > if (hid->urbout.status != -EINPROGRESS) > hid_submit_out(hid); Btw, this comment is about using the -EINPROGRESS being racy, right? This is fixed in a newer version that has yet to make it into 2.4. -- Vojtech Pavlik SuSE Labs _______________________________________________ [EMAIL PROTECTED] To unsubscribe, use the last form field at: https://lists.sourceforge.net/lists/listinfo/linux-usb-devel
