Yes. Linux cannot (yet) coredump a threaded app.
This is one of the main reasons why I have the DebugCache script to
automate stacktraces.
/Henrik
Duane Wessels wrote:
>
> When I run Squid with Async I/O on Linux (Red Hat 2.2.5-15), it won't
> dump core upon SEGV. Does it happen to you too?
#!/bin/sh
# Modified RunCache script to get automated stacktraces on failures
#
prefix=/usr/local/squid
exec_prefix=${prefix}
logdir=${prefix}/logs
PATH=${exec_prefix}/bin:/bin:/usr/bin
export PATH
conf=""
if test "$1" ; then
conf="-f $1"
shift
fi
cat <<EOF >${prefix}/etc/stacktrace.gdb
handle all pass nostop noprint
handle SIGSEGV stop print
handle SIGBUS stop print
handle SIGABRT stop print
break fatal
run -NDYd3 $conf
bt
c
bt
c
bt
c
bt
c
bt
c
bt
quit
EOF
failcount=0
while : ; do
echo "Running: squid -NDYd3 $conf >> $logdir/squid.out 2>&1"
echo "Startup: `date`" >> $logdir/squid.out
start=`date '+%d%H%M%S'`
gdb --batch -x ${prefix}/etc/stacktrace.gdb ${prefix}/bin/squid >> $logdir/squid.out 2>&1
stop=`date '+%d%H%M%S'`
t=`expr $stop - $start`
if test 0 -le $t -a $t -lt 5 ; then
failcount=`expr $failcount + 1`
else
failcount=0
fi
if test $failcount -gt 5 ; then
echo "DebugCache: EXITING DUE TO REPEATED, FREQUENT FAILURES" >&2
exit 1
fi
sleep 10
done
Received on Wed Oct 13 1999 - 14:52:57 MDT
This archive was generated by hypermail pre-2.1.9 : Tue Dec 09 2003 - 16:12:18 MST