Aw: gpg related crash/hang on 2.0.11

[albrecht dress arcor de]

Hi Nuno,

I appended your broken message to my mailbox, but it's signature is checked *without* any hang using balsa 2.0.11. This works both with automagic and manual key retreival.

I can confirm that the linking order of your binary is correct, but it's definitely a hang in gpgme (I think the SIGPIPE comes if esd is not running - shouldn't be critical?).

All I can imagine is that you have an old or somehow broken gpg version (which in turn is gpgme's backend). Here I use gpg 1.2.1, and, again, this works fine with this mail.

Hth,

Albrecht.

----- Original Nachricht ----
Von:     Nuno Monteiro <nuno@itsari.org>
An:      balsa-list@gnome.org
Datum:   12.05.03 23:42
Betreff: gpg related crash/hang on 2.0.11

> Hi all,
> 
> 
> I'm experiencing a weird hang/crash with Balsa 2.0.11 when opening a 
> _specific_ gpg signed email. I tried to reproduce it with every other 
> signed email I had around, but it seems it'll only choke on this one. 
> Whenever I double click on that particular email from the list, Balsa 
> hangs and stays like that forever (well, until Gnome asks me if I want 
> to kill the window, at least). The system is a fairly stock Mdk 9.1 
> installed a week ago, gpgme is version 0.3.15, and works otherwise 
> great. I got this backtrace from gdb:
> 
> (gdb) r
> Starting program: /home/nuno/src/balsa/src/balsa [New Thread 16384 (LWP 
> 20906)]
> ** Message: init gpgme version 0.3.15
> Opening 1 mailboxes on startup.
> [New Thread 32769 (LWP 20909)]
> [New Thread 16386 (LWP 20910)]
> opening Inbox..
> [New Thread 32771 (LWP 20911)]
> [New Thread 49156 (LWP 20912)]
> 
> Program received signal SIGPIPE, Broken pipe.
> [Switching to Thread 16384 (LWP 20906)]
> 0x4027474b in write () from /lib/i686/libpthread.so.0
> (gdb) bt
> #0  0x4027474b in write () from /lib/i686/libpthread.so.0
> #1  0x40e0c498 in __JCR_LIST__ () from /usr/lib/libgpgme.so.6
> #2  0x40e07ccf in _gpgme_io_write (fd=30, buffer=0xbffff1e0, count=511)
>      at posix-io.c:75
> #3  0x40dfcda1 in write_cb_data (dh=0x86e9b88, fd=30) at data.c:969
> #4  0x40dfceb8 in _gpgme_data_outbound_handler (opaque=0x86e9b88, fd=30)
>      at data.c:1004
> #5  0x40dfd5dc in do_select (fdt=0x86e9418) at wait.c:168
> #6  0x40dfd919 in _gpgme_wait_on_condition (ctx=0x86e9390, cond=0x0)
>      at wait.c:281
> #7  0x40dfd8ae in _gpgme_wait_one (ctx=0x86e9390) at wait.c:269
> #8  0x40dff2dc in gpgme_op_verify (ctx=0x86e9390, sig=0x86e6c30,     
> text=0x86e9b88, r_stat=0x86e9eb0) at verify.c:488
> #9  0x080c9103 in libbalsa_body_check_signature (body=0x86e9390)
>      at rfc3156.c:355
> #10 0x08074dac in balsa_message_scan_signatures (body=0x86ebb38,     
> message=0x8625be8) at balsa-message.c:532
> #11 0x08074ebf in balsa_message_set (bm=0x86ee4e8, message=0x8625be8)
>      at balsa-message.c:659
> #12 0x0808d991 in message_window_idle_handler (mw=0x86ea168)
>      at message-window.c:228
> 
> 
> libpthread is linked before libgpgme as it is supposed to be:
> 
> nuno@hobbes:~/src/balsa/src$ ldd balsa|grep -n '\(pthread\|gpgme\)'
> 14:     libpthread.so.0 => /lib/i686/libpthread.so.0 (0x40269000)
> 43:     libgpgme.so.6 => /usr/lib/libgpgme.so.6 (0x40df6000)
> 
> 
> I'm also attaching a gzipped version of the offending email, in case it 
> is needed to debug the problem further. Let me know if you need 
> anything else.
> 
> 
> Cheers,
> 
> 
> 			Nuno
>