Ethereal-dev: [Ethereal-dev] ethereal-0.10.4: crash on HP-UX 11.11 - workaround: reduce -O lev

Note: This archive is from the project's previous web site, ethereal.com. This list is no longer active.

From: "Falk Hayn" <Falk.Hayn@xxxxxx>
Date: Fri, 28 May 2004 15:53:26 +0200 (MEST)
Dear Ladies and Sirs, Dear Guy Harris,Dear Olivier Biot,
Dear core-developpers,

searching the error message pattern fv->ftype->set_value
in the dev mailing list archive leads to at least 9 hits.
( 1st occurance of problem in Dec. 2003 )
-> Problem is outstanding for at least 6 month.

******
Recommended workarounds: reduce optimization level
******

I reduced the optimization level from -O2 to -O, this modified 
ethereal build completed and the resulting binaries do not crash.
( I was lazy: vi configure, search O2, replace O)


All dev mailing list threads I read, seems to have the same pattern:
- does not depend on GTK version, I�m still using GTL 1.2.10
- 64 bit enabled operating system (64 bit linux, 64 bit HP-UX 11.11)
- gcc 3.3.x    (I do have gcc 3.3.3)


important question: Who of the core developpers has a 64 bit machine (SPARC,
ITANIUM, AMD64, with gcc 3.3.x  *and* does not (!) see the problem ?


Pls. let me know what You think.


With regards
F.

--- Weitergeleitete Nachricht / Forwarded Message ---
Date: Tue, 18 May 2004 16:33:42 +0200 (MEST)
From: "Falk Hayn" <Falk.Hayn@xxxxxx>
To: ethereal-dev@xxxxxxxxxxxx
Subject: ethereal-0.10.4: crash on HP-UX 11.11 - with backtrace in debugger
session

Dear Ladies and Sirs,

pls. find a backtrace from active debugger session:

Who has an idea what the root cause of the 1st "dangerous looking" message
in the backtrace may be ?
(Well, runnig a pgm in debugger may create bogus msgs, but this looks really
true... ;-)  :

#7  0x7a557a28 in proto_tree_set_ether (fi=0x6, value=0x46b <Error reading
address 0x46b: Bad address>) at proto.c:1442

Who can give advice�s to drill down the problem further ?


Full backtrace under running debugger.
======================================

# gdb gethereal
.... omitted ...
(gdb) show env DISPLAY
DISPLAY = mydisp:0.0

(gdb) run
Starting program: /usr/local/bin/ethereal 
Detaching after fork from process 3704

** ERROR **: file ftypes.c: line 431 (fvalue_set): assertion failed:
(fv->ftype->set_value)
aborting...

Program received signal SIGABRT, Aborted.
0x7b00fad0 in kill () from /usr/lib/libc.2

(gdb) backtrace                                                             
                                                                      
#0  0x7b00fad0 in kill () from /usr/lib/libc.2
#1  0x7afaa55c in raise () from /usr/lib/libc.2
#2  0x7afea9a8 in abort_C () from /usr/lib/libc.2
#3  0x7afeaa04 in abort () from /usr/lib/libc.2
#4  0x7ad43850 in g_logv (log_domain=0x0, log_level=G_LOG_LEVEL_ERROR,
format=0x0, args1=0x6) at gmessages.c:391
#5  0x7ad438ac in g_log (log_domain=0x6 <Error reading address 0x6: Bad
address>, log_level=1131, format=0x432cc040 <Error reading address
0x432cc040: Bad address>) at gmessages.c:408
#6  0x7a9c8740 in fvalue_set (fv=0x79c6a0fc, value=0x4002061e,
already_copied=0) at ftypes.c:432
#7  0x7a557a28 in proto_tree_set_ether (fi=0x6, value=0x46b <Error reading
address 0x46b: Bad address>) at proto.c:1442
#8  0x7a557898 in proto_tree_add_ether (tree=0x401d1330, hfindex=3366,
tvb=0x401cfde8, start=6, length=6, value=0x4002061e "") at proto.c:1400
#9  0x7a557958 in proto_tree_add_ether_hidden (tree=0x6, hfindex=1131,
tvb=0x432cc040, start=134217759, length=6, value=0x4002061e "") at
proto.c:1411
#10 0x7a6a9118 in dissect_eth_common (tvb=0x401cfde8, pinfo=0x40169d98,
tree=0x401d1408, fcs_len=-1) at ../packet-eth.c:266
#11 0x7a6a9438 in dissect_eth_maybefcs () at ../packet-eth.c:358
#12 0x7a551da8 in call_dissector_work (handle=0x4006b9c0, tvb=0x401cfde8,
pinfo=0x40169d98, tree=0x401d1408) at packet.c:513
#13 0x7a552434 in dissector_try_port (sub_dissectors=0x6, port=1, tvb=0x0,
pinfo=0x40169d98, tree=0x401d1408) at packet.c:776
#14 0x7a6d0b8c in dissect_frame (tvb=0x401cfde8, pinfo=0x40169d98,
tree=0x401d1408) at ../packet-frame.c:184
#15 0x7a6d0558 in dissect_frame () at ../packet-frame.c:74
#16 0x7a6d0558 in dissect_frame () at ../packet-frame.c:74
Previous frame identical to this frame (corrupt stack?)
(gdb) 




--- Weitergeleitete Nachricht / Forwarded Message ---
Date: Tue, 18 May 2004 15:56:24 +0200 (MEST)
From: "Falk Hayn" <Falk.Hayn@xxxxxx>
To: ethereal-dev@xxxxxxxxxxxx
Subject: ethereal-0.10.4: crash on HP-UX 11.11

Platform: HP-UX 11.11
=====================
perhaps platform specific issue, HP-UX 11.11


Symptom: ethereal-0.10.4 crash :
=================================
** ERROR **: file ftypes.c: line 431 (fvalue_set): i
assertion failed: (fv->ftype->set_value)
aborting...
ethereal: 19117 Abort(coredump)



remark symptom is ethereal 0.10.xx only
=======================================
These crashes only occur with ethereal 0.10.xx train.
1st reported with ethereal 0.10.3

Ethereal 0.9.16 works stable in the same enviroment, 
same machine, same libs, same gcc version



gdb backtrace session:
======================

gdb ethereal core
GNU gdb 6.1
...
warning: Couldn't read general-purpose registers from `.reg' section in core
file.


warning: The shared libraries were not privately mapped; setting a
breakpoint in a shared library will not work until you rerun the program.

Debugging dynamic executables loaded via the hpux8 dld.sl is not supported.


warning: Couldn't read general-purpose registers from `.reg' section in core
file.

#0  0x00000000 in ?? ()
(gdb) backtrace
#0  0x00000000 in ?? ()
(gdb) 

libraries versions
==================
all libs are used with ethereal 0.9.16 and ethereal 0.10.3 without any
changes
 GTK 1.2.10
 TCPDUMP 0.8.3
 ADNS 1.1
 UCD SNMP 4.2.6
 PCRE 4.5


gcc version
===========
gcc -v
Reading specs from /usr/local/lib/gcc-lib/hppa2.0w-hp-hpux11.11/3.3.3/specs
Configured with: ../gcc-3.3.3/configure --with-gnu-as
--enable-languages=c,c++,objc
Thread model: single
gcc version 3.3.3

 
machine:
========
J2240, 2 Processors, 1 Gig RAM, 1 Gig Swap., HP-UX 11.11


Can pls. someone give me advices, how to proceed ?

Do not hesitate to contact me via falk.hayn@xxxxxx if You have
further questions.

With kindly regards


-- 
NEU : GMX Internet.FreeDSL
Ab sofort DSL-Tarif ohne Grundgeb�hr: http://www.gmx.net/dsl



-- 
NEU : GMX Internet.FreeDSL
Ab sofort DSL-Tarif ohne Grundgeb�hr: http://www.gmx.net/dsl



-- 
+++ Jetzt WLAN-Router f�r alle DSL-Einsteiger und Wechsler +++
GMX DSL-Powertarife zudem 3 Monate gratis* http://www.gmx.net/dsl