Actually, all I was referring to was the trace command: $ trace <pid> Or you can do it from Activity Monitor.app On Dec 8, 2008, at 03:28, Gene Selkov wrote:
On Mon, Dec 8, 2008 at 12:19 AM, Jeremy Huddleston <jeremyhu@apple.com> wrote:
There's nothing suspicious in your system log. A trace of the X11 process could help out here.
You mean something like strace, or should I build it with symbols and run it in gdb?
I caught it in this condition once again this morning, after the system woke up and swapped in the existing X11 process. I just shut the lid yesterday with X11 running (it was running fine then); today I opened the lid and saw this. Restarting X11 didn't help; it started up with the same problem. The condition really has something to do with it being idle. I watched: if it is displaced by other processes while idle, the CPU usage goes down from 100% to anywhere between 30 and 60%; better yet, any activity in any X11 window brings it down to 2-3%.
The effect disappears after a system restart.
I just tried to see if I could cause it to misbehave by putting my mac to sleep, but a few short sleep-wake cycles didn't do it. Maybe it needs a more substantial clock mismatch to trigger this, if the clock has anything to do with it. I will keep watching it during my normal work cycle.
The question is: what's the right way to trace it when I see it again. Obviously, I can't run the whole session in gdb. But when I see it spin, can I connect gdb to it and find out at least where it spins (if not what causes it). Alternatively, is there a way to make it dump a more verbose log?
Thanks,
--Gene _______________________________________________ Xquartz-dev mailing list Xquartz-dev@lists.macosforge.org http://lists.macosforge.org/mailman/listinfo.cgi/xquartz-dev