[Prev][Next][Index][Thread]
9500 crashes (still going)
I'm running cl-http on a 9500 with all the latest patches and fixes,
etc. I haven't tried the experimental mactcp interface, other than
that I believe I've tried everything.
Each fix whittles away a little bit at my problems. I think it's
been narrowed down to one or two things. Still, my machine rarely
stays up for more than a day, and never more than two days. The load
is still pretty light.
By installing and playing around with the preallocation "patch" I
eliminated the most frequent crash type as reported by macsbug.
The crashes I get now are characterized by a bad heap, in response to
the ht and hc commands in macsbug I get something like:
>The Rclassic-httpdS heap at 020FCB6C is bad
> Block length is bad
> Block header
> 021246C4 0A3C 4C49 3E41 5254 4943 4C45 533A 204E %<LI>ARTICLES: N
>
> Totaling the Rclassic-httpdS heap at 020FCB6C
> The Rclassic-httpdS heap at 020FCB6C is bad
> Block length is bad
> Block header
> 021246C4 0A3C 4C49 3E41 5254 4943 4C45 533A 204E %<LI>ARTICLES: N
The crashes (8 recorded so far in macsbug logs) all occur within calls
to the cl-http function enumerate-normal-item (that is to say, it's
the topmost common function in the active process stack for each
crash). I don't know how relevant that is since I do a lot of
enumerating in my application. If it's a random thing chances are
pretty good it would happen during an enumeration. Still, it's
something in common.
Is there any data I can collect that would be more meaningful? Any
suggestions on how to deal with this? This is a cry for help.
=====
Christopher Welty Vassar College Computer Science Dept.
weltyc@cs.vassar.edu Poughkeepsie, NY 12601
http://www.cs.vassar.edu/faculty/welty (914) 437-5992
Follow-Ups: