Re: randomly dying xterms
Re: randomly dying xterms
- Subject: Re: randomly dying xterms
- From: George Peter Staplin <email@hidden>
- Date: Fri, 24 Oct 2008 18:41:14 -0600
Quoted Ronald Cohen <email@hidden>:
With the latest X11 (2.3.1) and Mac OS X 10.5.5, I seem to find xterms
randomly dying. Not frequently, but seemingly about once a day.
Whatever process was running in that xterm (which can be almost
anything) is killed at the same time. It is not confined to xterms,
but seemingly any xwindow; I've also had the experience with xemacs.
X11 itself usually stays alive (though I've had less frequent crashes
of X11 as well).
I'm not sure if I'm seeing something useful here, but I do notice, at
about the right time in the console log, the following:
10/24/08 2:26:01 PM [0x0-0xad0ad].org.x.X11[4134] Xlib: sequence lost
(0x12020 > 0x389a) in reply type 0x20!
10/24/08 2:26:01 PM [0x0-0xad0ad].org.x.X11[4134] Xlib: sequence lost
(0x1206f > 0x389a) in reply type 0x20!
10/24/08 2:26:01 PM [0x0-0xad0ad].org.x.X11[4134] Xlib: sequence lost
(0x16568 > 0x389a) in reply type 0x20!
10/24/08 2:34:05 PM [0x0-0xad0ad].org.x.X11[4134] Xlib: sequence lost
(0x12020 > 0x389a) in reply type 0x20!
10/24/08 2:34:05 PM [0x0-0xad0ad].org.x.X11[4134] Xlib: sequence lost
(0x16570 > 0x389a) in reply type 0xa!
10/24/08 2:34:05 PM [0x0-0xad0ad].org.x.X11[4134] Xlib: sequence lost
(0x12074 > 0x389a) in reply type 0x20!
10/24/08 2:34:15 PM [0x0-0xad0ad].org.x.X11[4134] Xlib: sequence lost
(0x10000 > 0x389a) in reply type 0x20!
I think these may just be the result of my attempt to make this window
active by selecting the window from the "window" menu in X11. I
don't see anything else terribly relevant. Nothing for example in a
crash log.
Anyone else having such experiences, or ideas?
I've seen the same "sequence lost" error messages and experienced the
same problems before. I saw them a lot more often when I was working
on xpbproxy with 2.3.1. I suspect a race in XQuartz, otherwise I
don't see how the sequence is getting lost. In any case, I've been
unable to reproduce it precisely. I'm glad you were able to (as you
mentioned in another message).
I think one way of tracking it down might be to insert (temporarily)
an abort() so that when a sequence is lost (after after printing the
message), we get a backtrace that might be usable -- hopefully with
other thread state too.
Does it occur with the latest 2.3.2 beta 1?
George
--
http://people.freedesktop.org/~gstaplin/
_______________________________________________
Do not post admin requests to the list. They will be ignored.
X11-users mailing list (email@hidden)
This email sent to email@hidden