Here is the top of a gprof analysis. I'm wondering if the Xkb references indicate that it is the problem. I can hypothesise that some changes in the general Xserver code touched upon Xkb, expecting the DDX to initialise it appropriately. Since Xprint does not use Xkb, it doesn't handle it in any explicit way. Could Xkb be spinning, waiting for the Xprint DDX to tell it to go away and stop being annoying? This Xkb hypothesis is corroborated by https://bugs.freedesktop.org/show_bug.cgi?id=13063
Function/Method Count Total (s) % Self (s) Total ms/call Self ms/call Dispatch 1 0.150 68.200 0.030 150.000 30.000 WaitForSomething 1444561 0.060 27.300 0.060 0.000 0.000 __i686.get_pc_thunk.bx 0 0.060 27.300 0.060 0.000 0.000 FlushIfCriticalOutputPending 1444561 0.060 27.300 0.060 0.000 0.000 XkbRF_LoadRules 1 0.010 4.500 0.000 10.000 0.000 XkbInitKeyboardDeviceStruct 1 0.010 4.500 0.000 10.000 0.000 XkbDDXNamesFromRules 1 0.010 4.500 0.000 10.000 0.000 GetInputLine 833 0.010 4.500 0.010 0.010 0.010 CoreKeyboardProc 3 0.010 4.500 0.000 3.330 0.000 InitAndStartDevices 1 0.010 2.300 0.000 5.000 0.000 EnableDevice 4 0.000 1.500 0.000 0.830 0.000 CloseDownDevices 1 0.000 1.500 0.000 3.330 0.000 CloseDevice 4 0.000 1.500 0.000 0.830 0.000 ActivateDevice 4 0.000 1.500 0.000 0.830 0.000 InitCoreDevices 1 0.000 0.800 0.000 1.670 0.000 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
