Hi!

 I seem to have found a quite reproducible segfault (some other users
confirmed it to me on different architectures like amd64 and i386) so it
doesn't seem to be only powerpc:

 w3m https://db.debian.org/ (notice https, not http)
 o (for getting to the options)
 /ssl (searching for ssl)
 j (for going down a line)

 This is the recipe that made it possible for others to reproduce it.
Please notice that it doesn't seem to happen when one uses http://
instead of https:// and searching in the page itself doesn't trigger it
neither.

 Here some gdb backtrace:

#v+
Forward: ssl
Program received signal SIGPIPE, Broken pipe.
[Switching to Thread 0x48022ea0 (LWP 30959)]
0x0fb280a4 in write () from /lib/libc.so.6
(gdb) bt
#0  0x0fb280a4 in write () from /lib/libc.so.6
#1  0x0facb434 in _IO_file_write () from /lib/libc.so.6
#2  0x0facaf80 in ?? () from /lib/libc.so.6
#3  0x0facb3b4 in _IO_do_write () from /lib/libc.so.6
#4  0x0facbc78 in _IO_file_sync () from /lib/libc.so.6
#5  0x0fabbf9c in fflush () from /lib/libc.so.6
#6  0x1003730c in ?? ()
#7  0x10037964 in ?? ()
#8  0x1000c3e0 in ?? ()
#9  0x1000c788 in ?? ()
#10 0x1000493c in ?? ()
#11 0x100140d8 in ?? ()
#12 0x0fa6f720 in ?? () from /lib/libc.so.6
#13 0x0fa6f8e0 in __libc_start_main () from /lib/libc.so.6
#14 0x00000000 in ?? ()
(gdb)
#v-

 Find also attached an strace -rf from one session. I waited a while
between the relevant keypresses (the enter after /ssl and before the j
and the segfault) so you can more easier find the place with the timing
informations in the strace output.

 If I can help in any way, pretty please let me know. I'm not sure if
this is related to #372171 somehow, though moving ~/.w3m away doesn't
gain me anything. #305879 sounds pretty much like this so it doesn't
seem to be fixed with 0.5.1-5.

 So long,
Rhonda

* Gerfried Fuchs <rhonda@debian.at> [2008-04-09 18:16:48 CEST]:

 It's even worse, it seems to be somehow related to https pages ...

w3m https://nm.debian.org/nmlist.php
/blub
j
<segfault>

 I haven't heard back from anyone, and the problem still holds. :(

 So long,
Rhonda

retitle 475199 w3m segfaults after search in https mode
thanks

* Gerfried Fuchs <rhonda@deb.at> [2008-07-28 13:17:46 CEST]:

 Erm, still someone around? This makes it absolutely annoying and quite
unusable for https pages (like nagios setups) and searching in there.

 Pretty please, any response would be appreciated, I am here and willing
to provide more informations if only one tells me what's further needed
...

 So long,
Rhonda

Hmm, on my system (sid i386, updated yesterday), the above problem
doesn't occur, even though I remove ~/.w3m.  The search feature in
https page with w3m 0.5.2-2+b1 works for me.

Thanks,
--
Tatsuya Kinoshita

* Tatsuya Kinoshita <tats@debian.org> [2008-11-10 13:10:27 CET]:

 Another way that worked relyable for me:

w3m 'https://db.debian.org/search.cgi?country=de&dosearch=Search...'
/joey
(twice return)
<segfault>

 Again, others were able to reproduce it. I've w3m-img running, but I
purged that package, moved my .w3m directory aside to start off with a
clean environment, it didn't change much ...  :(

 Yes, 0.5.2-2+b1 also here, current up2date lenny.
Rhonda

Hmm, it also works on my system, with or without w3m-img, with or
without libc6-i686.  I have never seen a segfault with w3m 0.5.2.

Hi,

I did what was described in the bug reports first mail:

zsh: segmentation fault  w3m -no-mouse https://db.debian.org/
2:carsten@taz:~> dpkg -l w3m
Desired=Unknown/Install/Remove/Purge/Hold
| Status=Not/Inst/Cfg-files/Unpacked/Failed-cfg/Half-inst/trig-aWait/Trig-pend
|/ Err?=(none)/Hold/Reinst-required/X=both-problems (Status,Err: uppercase=bad)
||/ Name           Version        Description
+++-==============-==============-============================================
ii  w3m            0.5.2-2+b1     WWW browsable pager with excellent tables/fr
3:carsten@taz:~> dpkg -l w3m-img
Desired=Unknown/Install/Remove/Purge/Hold
| Status=Not/Inst/Cfg-files/Unpacked/Failed-cfg/Half-inst/trig-aWait/Trig-pend
|/ Err?=(none)/Hold/Reinst-required/X=both-problems (Status,Err: uppercase=bad)
||/ Name           Version        Description
+++-==============-==============-============================================
pn  w3m-img        <none>         (no description available)
4:carsten@taz:~> dpkg -l w3m-ssl
Desired=Unknown/Install/Remove/Purge/Hold
| Status=Not/Inst/Cfg-files/Unpacked/Failed-cfg/Half-inst/trig-aWait/Trig-pend
|/ Err?=(none)/Hold/Reinst-required/X=both-problems (Status,Err: uppercase=bad)
||/ Name           Version        Description
+++-==============-==============-============================================
un  w3m-ssl        <none>         (no description available)
5:carsten@taz:~> uname -a
Linux taz 2.6.18-6-686 #1 SMP Mon Oct 13 16:13:09 UTC 2008 i686 GNU/Linux
6:carsten@taz:~> cat .w3m/keymap
keymap q EXIT
keymap j UP
keymap k DOWN
keymap h LEFT
keymap l RIGHT


Regards,
Carsten

Some German speaking people are able to reproduce this but Japanese
speaking people not? If yes then this could be helpful:

7:carsten@taz:~> locale
LANG=de_DE.UTF-8
LC_CTYPE="de_DE.UTF-8"
LC_NUMERIC=C
LC_TIME=C
LC_COLLATE=C
LC_MONETARY="de_DE.UTF-8"
LC_MESSAGES=C
LC_PAPER="de_DE.UTF-8"
LC_NAME="de_DE.UTF-8"
LC_ADDRESS="de_DE.UTF-8"
LC_TELEPHONE="de_DE.UTF-8"
LC_MEASUREMENT="de_DE.UTF-8"
LC_IDENTIFICATION="de_DE.UTF-8"
LC_ALL=

	Hi!

* Tatsuya Kinoshita <tats@debian.org> [2008-11-12 13:57:49 CET]:
to receive a segfault from a w3m session within gdb and am able to offer
you a backtrace:

Forward: security
Program received signal SIGPIPE, Broken pipe.
[Switching to Thread 0x48022f10 (LWP 8330)]
0x0fb2f524 in __write_nocancel () from /lib/libc.so.6
(gdb) bt
#0  0x0fb2f524 in __write_nocancel () from /lib/libc.so.6
#1  0x0fad03a4 in _IO_new_file_write (f=0x10172a10, data=0x48250000, n=9) at fileops.c:1268
#2  0x0facfef0 in new_do_write (fp=0x10172a10, data=0x48250000 "security\n", to_do=9) at fileops.c:522
#3  0x0fad0324 in _IO_new_do_write (fp=<value optimized out>, data=<value optimized out>, to_do=9) at fileops.c:494
#4  0x0fad0be8 in _IO_new_file_sync (fp=<value optimized out>) at fileops.c:897
#5  0x0fac0d9c in _IO_fflush (fp=0x10172a10) at iofflush.c:43
#6  0x10038cb8 in ?? ()
#7  0x10039474 in ?? ()
#8  0x1000c374 in ?? ()
#9  0x1000c7a8 in ?? ()
#10 0x100141a8 in ?? ()
#11 0x0fa73ae4 in generic_start_main (main=0x10013390 <_init+60548>, argc=2, ubp_av=0xbf8f1454, auxvec=0xbf8f1518, init=0x10172a10, fini=<value optimized out>, rtld_fini=<value optimized out>, stack_end=<value optimized out>) at ../csu/libc-start.c:222
#12 0x0fa73ca0 in __libc_start_main (argc=<value optimized out>, ubp_av=<value optimized out>, ubp_ev=<value optimized out>, auxvec=<value optimized out>, rtld_fini=<value optimized out>, stinfo=<value optimized out>, stack_on_entry=<value optimized out>)
    at ../sysdeps/unix/sysv/linux/powerpc/libc-start.c:92
#13 0x00000000 in ?? ()
(gdb)

 I hope this helps you a bit. If you have any further hint pretty please
let me know.

 So long!
Rhonda

This problem causes pain for me too.  strace-ing w3m showed that it
wants to run some program called "migemo", and I don't have that.  So
i looked it up in the options (search for "migemo") and disabled it.
Guess what: the segfault disappeared.

hth,

	Hi!

* Akos PASZTORY <akos.pasztory@gmail.com> [2009-07-06 12:22:00 CEST]:

 Same for me it seems! Great great great!

 Tatsuya, I take it that you have migemo installed on your system and
that this might be the reason that you were unable to reproduce it?

 Would it be albe to save-guard the call to migemo so that it doesn't
segfault when it's not installed - or disable the option by default? I
rather expect fewer people to have that actually installed and in need
of it than people annoyed by the segfault and not having it installed.

 Thanks, and if you have a fix for it, can you try to get it into lenny
too? That would be really great.

 Thanks, Akos, for finding out about the workaround!
Rhonda

tags 475199 - moreinfo unreproducible
forcemerge 372171 475199
thanks

forcemerge 372171 475199 532921
retitle 372171 w3m: searching without migemo causes segfault
thanks