Profiler bugs in 3-HEAD (help?)

From: Amos Jeffries <squid3@dont-contact.us>
Date: Mon, 04 Feb 2008 23:07:52 +1300

Anyone have any ideas?

As far as I can tell tonight, the call sequence in both
src/comm_epoll.cc (comm_check_incoming) and lib/util.c (xfree) appear to
be correct :-(

Heres the segfault trace:

Program received signal SIGABRT, Aborted.
[Switching to Thread 0xb7c4b940 (LWP 24800)]
0xb7c76f06 in raise () from /lib/libc.so.6
#0 0xb7c76f06 in raise () from /lib/libc.so.6
#1 0xb7c78811 in abort () from /lib/libc.so.6
#2 0x08091490 in xassert (msg=0x818d744 "cstack[cstack_head].timer ==
type",
     file=0x818d6f1 "Profiler.c", line=213) at debug.cc:575
#3 0x08160f35 in xprof_stop (type=XPROF_comm_check_incoming,
     timer=0x816c610 "comm_check_incoming") at Profiler.c:213
#4 0x0808fa97 in comm_select (msec=10) at comm_epoll.cc:277
#5 0x08130e4e in CommSelectEngine::checkEvents (this=0xbff5c39c,
timeout=10)
     at comm.cc:2686
#6 0x0809ad81 in EventLoop::checkEngine (this=0xbff5c35c,
engine=0xbff5c39c,
     primary=true) at EventLoop.cc:50
#7 0x0809aed8 in EventLoop::runOnce (this=0xbff5c35c) at EventLoop.cc:124
#8 0x0809b04a in EventLoop::run (this=0xbff5c35c) at EventLoop.cc:100
#9 0x080e13dd in main (argc=2, argv=0xbff5c454) at main.cc:1321
(gdb) f 3
#3 0x08160f35 in xprof_stop (type=XPROF_comm_check_incoming,
     timer=0x816c610 "comm_check_incoming") at Profiler.c:213
(gdb) p timer
$1 = 0x816c610 "comm_check_incoming"
(gdb) p type
$3 = XPROF_comm_check_incoming
(gdb) p (xprof_type) cstack[cstack_head].timer
$4 = XPROF_xfree
(gdb) f 4
#4 0x0808fa97 in comm_select (msec=10) at comm_epoll.cc:277
277 PROF_stop(comm_check_incoming);

Amos

-- 
Please use Squid 2.6STABLE17+ or 3.0STABLE1+
There are serious security advisories out on all earlier releases.
Received on Mon Feb 04 2008 - 03:07:50 MST

This archive was generated by hypermail pre-2.1.9 : Sat Mar 01 2008 - 12:00:09 MST