Re: Spaces issues
Re: Spaces issues
- Subject: Re: Spaces issues
- From: Jeff Burley <email@hidden>
- Date: Mon, 28 Apr 2008 14:52:37 -0600
The issue is unresolved in the 1.4.0-apple4. The issue happens when
focus is on the xterm, then a ctrl <-- to change space, click on a
non X11 window, then ctrl --> and click the xterm to regain focus.
Once clicked, keyboard and mouse input do nothing to the xterms and
they're locked up. The process never hangs.
1.4.0-apple4 output:
Sampling process 8523 for 3 seconds with 1 millisecond of run time
between samples
Sampling completed, processing symbols...
Analysis of sampling X (pid 8523) every 1 millisecond
Call graph:
2316 Thread_2503
2316 start
2316 main
2316 X11ControllerMain
2316 X11ApplicationMain
2316 -[NSApplication run]
2312 -[NSApplication
nextEventMatchingMask:untilDate:inMode:dequeue:]
2312 _DPSNextEvent
2305 BlockUntilNextEventMatchingListInMode
2305 ReceiveNextEventCommon
2305 RunCurrentEventLoopInMode
2305 CFRunLoopRunInMode
2305 CFRunLoopRunSpecific
2272 mach_msg
2272 mach_msg_trap
2272 mach_msg_trap
29 __CFMachPortPerform
29
PullEventsFromWindowServerOnConnection(unsigned int, unsigned char)
28 CGEventCreateNextEvent
28 CGSGetNextEventRecordInternal
28 snarfEvents
27 _CGSGetPortStreamInline
27 mach_msg
27 mach_msg_trap
27 mach_msg_trap
1
CGSDispatchDatagramsFromStream
1 CGSDecodeEventRecord
1
CGEventCreateFromDataAndSource
1 free
1 munmap$UNIX2003
1 munmap$UNIX2003
1
ConvertPlatformEventRecordAndPostWithOptions(__CGEvent*,
_CGSEventRecord const*, short, unsigned char, unsigned char)
1 CreateEventWithCGEvent
1 CreateMouseEvent
1 SetEventParameter
1
RetainEventParamData(EventParam*, unsigned char)
1
TypeManager::Retain(unsigned long, unsigned long, unsigned long, void
const*)
1
TypeManager::FindCallbacks(unsigned long, unsigned char,
HITypeCompressedCallBacks*)
1 GetCollectionItem
1 __memcpy
1 __memcpy
2 0xffffffff
2 _sigtramp
2 _sigtramp
2 __CFRunLoopDoObservers
2 _handleWindowsNeedUpdateNote
1 -[NSApplication updateWindows]
1 -[NSNotificationCenter
postNotificationName:object:]
1 -[NSNotificationCenter
postNotificationName:object:userInfo:]
1 +[NSConcreteNotification
newTempNotificationWithName:object:userInfo:]
1 objc_assign_ivar
1 objc_assign_ivar
1 objc_msgSend
1 objc_msgSend
7 SendEventToEventTarget
7
SendEventToEventTargetInternal(OpaqueEventRef*, OpaqueEventTargetRef*,
HandlerCallRec*)
7 DispatchEventToHandlers(EventTargetRec*,
OpaqueEventRef*, HandlerCallRec*)
7
ToolboxEventDispatcherHandler(OpaqueEventHandlerCallRef*,
OpaqueEventRef*, void*)
6 FindWindowAndPartFromMouseEvent
6 FindWindow
6 FindWindowOfClass
6 FindWindowCommon(Point,
OpaqueWindowPtr**, short*)
6 FindPlatformWindow
6 CGSFindWindowAndOwner
6 _CGSFindWindow
6 mach_msg
6 mach_msg_trap
6 mach_msg_trap
1 GetApplicationEventTarget
1 HIObjectGetEventTarget
1 HIObjectGetEventTarget
2 -[X11Application sendEvent:]
2 send_nsevent
2 DarwinSendPointerEvents
2 write$UNIX2003
2 write$UNIX2003
2 NSPopAutoreleasePool
2 -[NSEvent dealloc]
2 ReleaseEvent
2 ReleaseEventParam(EventParam*)
2 ReleaseEventParam(EventParam*)
2316 Thread_2603
2316 thread_start
2316 _pthread_start
2316 CAPThread::Entry(CAPThread*)
2316 HALRunLoop::OwnThread(void*)
2316 CFRunLoopRunInMode
2316 CFRunLoopRunSpecific
2316 mach_msg
2316 mach_msg_trap
2316 mach_msg_trap
2316 Thread_2703
2316 thread_start
2316 _pthread_start
2316 server_thread
2316 dix_main
2316 Dispatch
2316 select$DARWIN_EXTSN
2316 select$DARWIN_EXTSN
2316 Thread_2803
2316 thread_start
2316 _pthread_start
2316 _xp_async_thread
2316 _xp_async_dequeue
2316 pthread_cond_wait$UNIX2003
2316 __semwait_signal
2316 __semwait_signal
Total number in stack (recursive counted multiple, when >=5):
Sort by top of stack, same collapsed (when >= 5):
mach_msg_trap 4621
__semwait_signal 2316
select$DARWIN_EXTSN 2316
Sample analysis of process 8523 written to file /dev/stdout
On Apr 28, 2008, at 1:25 PM, Jeremy Huddleston wrote:
Also, would you mind trying out 1.4.0-apple4 ... I cleaned up some
threading issues which might have "cured" the problem...
http://xquartz.macosforge.org/downloads/Xquartz-1.4.0-apple4.bz2
On Apr 28, 2008, at 12:18, Jeremy Huddleston wrote:
I'm wondering if you're triggering the lockup I so desperately want
to fix for 2.2.1 (http://trac.macosforge.org/projects/xquartz/ticket/41
)... unfortunately, I can't seem to trigger it. Could you please
sample the process and send me the log.
To sample the process, open /Applications/Utilities/Activity
Monitor.app, double click the X11 process (it should be red and say
"not responding"). Then click on the 'sample' button.
Please send me that log.
Thanks,
Jeremy
On Apr 28, 2008, at 11:33, Jeff Burley wrote:
I'm a fairly simple X11 user and don't understand where my
problems reside. I'm your typical unix admin using xterms to ssh
to multiple hosts. I predominantly use ssh -x or y <host> and
rarely revert any display back to my machine.
I'm running 2.2.1-rc1....
I launch xterms using OnMyCommand. Usually like so:
#!/bin/bash
/usr/X11/bin/xterm -T mhdcdog01 -n mhdcdog01 -bg DarkOliveGreen -
fg khaki -ls -e ssh -Y jburley@host
It will launch and sometimes run for days. I can have multiple
windows open. I lock the windows to opening on one space only.
Finally it will stop responding to keyboard input after changing
focus. It doesn't crash, but it seems I've lost connection to the
X server. However, I get nothing in the logs when this happens...
until I quit X11:
4/28/08 12:24:44 PM org.x.startx[8110] font_cache: Scanning user
font directories to generate X11 font caches
4/28/08 12:24:44 PM org.x.privileged_startx[8127] mktemp: mkdtemp
failed on /tmp/.ICE-unix: File exists
4/28/08 12:24:44 PM org.x.privileged_startx[8127] mktemp: mkdtemp
failed on /tmp/.X11-unix: File exists
4/28/08 12:24:44 PM org.x.privileged_startx[8127] mktemp: mkdtemp
failed on /tmp/.font-unix: File exists
4/28/08 12:24:44 PM org.x.privileged_startx[8127] font_cache:
Scanning user font directories to generate X11 font caches
4/28/08 12:24:44 PM org.x.startx[8110] font_cache: Updating FC cache
4/28/08 12:24:44 PM org.x.startx[8110] xinit: Detected Xquartz
startup, setting file=/usr/X11/bin/X, argv[0]=/Applications/
Utilities/X11.app/Contents/MacOS/X11
4/28/08 12:24:44 PM org.x.privileged_startx[8127] font_cache:
Updating FC cache
4/28/08 12:24:45 PM org.x.startx[8110] Xquartz starting:
4/28/08 12:24:45 PM org.x.startx[8110] X.org Release 7.2
4/28/08 12:24:45 PM org.x.startx[8110] X.Org X Server 1.3.0-apple18
4/28/08 12:24:45 PM org.x.startx[8110] Build Date: 20080422
4/28/08 12:24:45 PM org.x.startx[8110] Xquartz: run by launchd for
fd 0
4/28/08 12:24:47 PM org.x.privileged_startx[8127] font_cache: Done
4/28/08 12:24:47 PM org.x.startx[8110] font_cache: Done
Once I quit X11 (due to the unresponsiveness) I get:
4/28/08 12:28:41 PM [0x0-0xc00c].com.apple.finder[116] xterm:
fatal IO error 32 (Broken pipe) or KillClient on X server "/tmp/
launch-JGPWWs/:0"
4/28/08 12:28:41 PM [0x0-0xc00c].com.apple.finder[116] xterm:
fatal IO error 32 (Broken pipe) or KillClient on X server "/tmp/
launch-JGPWWs/:0"
4/28/08 12:28:41 PM [0x0-0xc00c].com.apple.finder[116] xterm:
fatal IO error 32 (Broken pipe) or KillClient on X server "/tmp/
launch-JGPWWs/:0"
4/28/08 12:28:41 PM org.x.startx[8110] xinit: connection to X
server lost.
4/28/08 12:28:41 PM org.x.startx[8110] waiting for X server to
shut down
Is there a way I can place the X server into a verbose debugging
mode to figure out what is going on?
I'm running dual displays at 1600x1200 if that is any consolation.
J
_______________________________________________
Do not post admin requests to the list. They will be ignored.
X11-users mailing list (email@hidden)
This email sent to email@hidden
_______________________________________________
Do not post admin requests to the list. They will be ignored.
X11-users mailing list (email@hidden)
This email sent to email@hidden
_______________________________________________
Do not post admin requests to the list. They will be ignored.
X11-users mailing list (email@hidden)
This email sent to email@hidden