Re: [Bug: 21.5-b34] Hang with gnuserv on Mavericks
Ar an chéad lá de mí na Nollaig, scríobh Stephen J. Turnbull:
Aidan Kehoe writes:
> Attaching with lldb I see that XEmacs is waiting in
> XtAppProcessEvent, as it should;
Are you sure it is "as it should"? In particular, the bug I reported
(described in http://tracker.xemacs.org/issue565,
http://trac.macports.org/ticket/18491, and
https://bugs.freedesktop.org/show_bug.cgi?format=multiple&id=20048)
was never fixed properly, although it seems they did apply my patch
"Patch to prevent reentering...". The symptom I had was an infloop in
_XtWaitForSomething, which is fixed by that patch. But maybe the real
bug also affects something else in XtAppPending.
It *may* be inflooping, now you mention it. We were seeing XEmacs taking 94%
of the CPU at random, but more often without -vanilla. At that point lldb
wasn’t working for me because I wasn’t in the admin group, so I couldn’t
work out exactly where it was.
Jeremy Huddleston, in that bug, says he backed out your patch in March 2012.
It might well be that this is the first Apple X11 release I’ve seen since
that was done.
> when I call p Feval (Fcar (Fread_from_string (build_ascstring
> ("(gnuserv-start)"), Qnil, Qnil))) within lldb, gnuserv restarts and the
> screen TTY becomes reactive again.
Is the original gnuserv process dead at that point, or does restarting
gnuserv kill it? Is screen itself dead, or just the XEmacs screen?
Restarting gnuserv kills it. Screen is not dead, rather the XEmacs process
running within it is not responsive, neither to TTY input via screen, nor to
GNUserv input, until I execute the above code within the debugger. The
XEmacs within the TTY wakes up immediately when i do that.
I’ll rebuild without X11 support (my big project at the moment is editing an
OCRed scan of a Persian-English dictionary, so the TTY is where I’m spending
most of my time) and see if it goes away, which is what your bug suggests it
should do.
--
‘Liston operated so fast that he once accidentally amputated an assistant’s
fingers along with a patient’s leg, […] The patient and the assistant both
died of sepsis, and a spectator reportedly died of shock, resulting in the
only known procedure with a 300% mortality.’ (Atul Gawande, NEJM, 2012)
_______________________________________________
XEmacs-Beta mailing list
XEmacs-Beta(a)xemacs.org
http://lists.xemacs.org/mailman/listinfo/xemacs-beta