[Klone-users] Re: Segmentation fault in kloned

[Steven Van Ingelgem]

Just do an empty post, I think that is what triggers the crash.
I think even a simple "POST" with no parameters to the "index.kl1" script
will crash the server, though I didn't tested it.

2008/9/29 Stefano Barbato <barbato at koanlogic.com>

> Hi Steven,
>
> could you provide a sample webapp/ dir with html files that fire-up the
> error? can you?
>
> thanks,
> s
>
>
>
>
>
>
>
>
>
> On 29/set/08, at 16:57, Steven Van Ingelgem wrote:
>
>  Ok, I have been able to reproduce the error.
>>
>> Change in request.c "REQUEST_DEFAULT_POST_TIMEOUT" into for example 5, and
>> if you sent it requests, the kloned will crash after 5 seconds. But seems
>> only if you send it "posts". Not when sending gets.
>>
>> Greetings,
>> Steven
>>
>> 2008/9/29 Steven Van Ingelgem <steven at vaningelgem.be>
>>
>> =====================================================================
>> Program received signal SIGSEGV, Segmentation fault.
>> 0x0810f7c7 in ?? ()
>> (gdb) bt
>> #0  0x0810f7c7 in ?? ()
>> #1  0x0807c15e in io_close (io=0x8115600) at io.c:447
>> #2  0x0806ef94 in request_cb_close_socket (al=0x810fc50, arg=0x8115600) at
>> request.c:1545
>> #3  0x0805d1d8 in timerm_sigalrm (sigalrm=14) at timer.c:100
>> #4  <signal handler called>
>> #5  0xb7efd410 in ?? ()
>> #6  0xbfcb5138 in ?? ()
>> #7  0xbfcb5028 in ?? ()
>> #8  0xbfcb50a8 in ?? ()
>> #9  0xb7bd40fd in select () from /lib/tls/i686/cmov/libc.so.6
>> #10 0x0804fb0c in server_loop (s=0x810fa18) at server.c:953
>> #11 0x08052084 in app_run () at main.c:157
>> #12 0x08052558 in main (argc=3, argv=0xbfcb5244) at entry.c:408
>> =====================================================================
>>
>> This occurs when I'm running an AJAX-called javascript function with a
>> timeout of 1500ms.
>> It's not always at the same call this crashes... But generally it's like
>> after half a minute or so.
>>
>> =====================================================================
>> (gdb) up
>> #1  0x0807c15e in io_close (io=0x8115600) at io.c:447
>> 447             dbg_err_if(io->close(io));
>> (gdb) up
>> #2  0x0806ef94 in request_cb_close_socket (al=0x810fc50, arg=0x8115600) at
>> request.c:1545
>> 1545        io_close(io);
>> (gdb) up
>> #3  0x0805d1d8 in timerm_sigalrm (sigalrm=14) at timer.c:100
>> 100             al->cb(al, al->arg);
>> (gdb) down
>> #2  0x0806ef94 in request_cb_close_socket (al=0x810fc50, arg=0x8115600) at
>> request.c:1545
>> 1545        io_close(io);
>> (gdb) display io
>> 1: io = (io_t *) 0x8115600
>> (gdb) display io->name
>> 2: io->name = 0x8113fc0 "\230?\021\b"
>> (gdb) display io->eof
>> 3: io->eof = 1222693008
>> (gdb) display io->type
>> 4: io->type = 134672157
>> =====================================================================
>>
>> I think the there is somewhere not done a correct cleanup of the structure
>> which the " timerm_sigalrm " is trying to destroy...
>>
>> Could you check this out?
>>
>>
>> Thanks!
>>
>> _______________________________________________
>> Klone-users mailing list
>> Klone-users at koanlogic.com
>> http://koanlogic.com/cgi-bin/mailman/listinfo/klone-users
>>
>
>
> _______________________________________________
> Klone-users mailing list
> Klone-users at koanlogic.com
> http://koanlogic.com/cgi-bin/mailman/listinfo/klone-users
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://koanlogic.com/pipermail/klone-users/attachments/20080929/d070465f/attachment.html