m_forcenick.c causes core on 7.0.3

[Rachel Llorenna]

For your clarification, he sent me a full backtrace privately, which I
responded to with a fix. He then re-sent it to the list again. I
understand however that there may have been some latency-related
issues here, though.

I also understand that not everyone knows how to write modules, so I
pointed out where the error was, albeit a bit unclearly. Yes, I agree,
my explanation of what to fix was terrible. I did however follow that
up with a better explanation of how to fix it, which I am confident
would have fixed the issue.

To anyone with the authority to do so that is still reading this
thread: add this issue to the FAQ somewhere?


On Tue, 22 Feb 2005 19:45:37 -0600, Caleb <caleb at x-ess.org> wrote:
> Please keep in mind that not everybody who use modules knows how to write
> them.
> And he resent it to the list with a full backtrace as YOU had asked him to.
> 
> 
> ----- Original Message -----
> From: "Rachel Llorenna" <rachies at gmail.com>
> To: "Andrey Sheykhot" <sheyh at xtra.co.nz>; "General IRCD-Hybrid Discussion"
> <hybrid at lists.ircd-hybrid.org>
> Sent: Tuesday, February 22, 2005 7:23 PM
> Subject: Re: Re[2]: m_forcenick.c causes core on 7.0.3
> 
> > I'm not quite sure why you re-sent this to the list, but I already
> > answered this question in my reply to you. In case you've forgotten:
> >
> > I've already told you to take a look at the message tabs, since there
> > are now 6 handler fields.  But, since you don't seem able to do that,
> > either through incompetence or ignorance, I've determined that that is
> > indeed the case.
> > The module specifies:
> > {m_ignore, m_not_oper, mo_forcenick, mo_forcenick}
> > Which is only 4 fields of the required 6.
> >
> > Problem solved. Maybe we should put that in an FAQ somewhere too, it
> > seems to come up often with non-stock modules.
> >
> > If you're still confused about it, try changing this line:
> > {m_ignore, m_not_oper, mo_forcenick, mo_forcenick}
> > to:
> > {m_ignore, m_not_oper, mo_forcenick, m_ignore, mo_forcenick,m_ignore}
> >
> >
> > On Wed, 23 Feb 2005 14:19:19 +1300, Andrey Sheykhot <sheyh at xtra.co.nz>
> > wrote:
> >>
> >> yes I understand that this module is not related to hybrid itself,
> >> however people in this list seem to be interested in solving ircd
> >> related issues like this.
> >>
> >> dgb full output:
> >>
> >> (gdb) bt full
> >> #0  0x00000000 in ?? ()
> >> No symbol table info available.
> >> #1  0x1c01939b in handle_command (mptr=0x2375a120, client_p=0x7d4eee98,
> >> from=0x7d4eee98, i=3,
> >>     hpara=0x3c1d7ec0) at parse.c:394
> >>         mptr = (struct Message *) 0x2375a120
> >>         client_p = (struct Client *) 0x7d4eee98
> >>         hpara = (char **) 0x3c1d7ec0
> >>         handler = 0
> >> #2  0x1c0191da in parse (client_p=0x7d4eee98, pbuffer=0x3c1d3ea0
> >> "FORCENICK", bufend=0x3c1d3eb4 "")
> >>     at parse.c:300
> >>         from = (struct Client *) 0x7d4eee98
> >>         ch = 0x3 ""
> >>         s = 0x3c1d3eaa "test"
> >>         end = 0x3c1d3eb3 "2"
> >>         i = 3
> >>         numeric = 0x0
> >>         mptr = (struct Message *) 0x2375a120
> >> #3  0x1c018ec2 in client_dopacket (client_p=0x7d4eee98, buffer=0x3c1d3ea0
> >> "FORCENICK", length=20)
> >>     at packet.c:528
> >>         client_p = (struct Client *) 0x3c1d7ec0
> >>         buffer = 0x3c1d3ea0 "FORCENICK"
> >>         length = 20
> >> #4  0x1c018980 in parse_client_queued (client_p=0x7d4eee98) at
> >> packet.c:174
> >>         client_p = (struct Client *) 0x7d4eee98
> >>         dolen = 3
> >> ---Type <return> to continue, or q <return> to quit---
> >>         checkflood = 0
> >>         lclient_p = (struct LocalUser *) 0x818a063c
> >> #5  0x1c018d94 in read_packet (fd=6, data=0x7d4eee98) at packet.c:442
> >>         client_p = (struct Client *) 0x7d4eee98
> >>         lclient_p = (struct LocalUser *) 0x818a063c
> >>         length = 22
> >>         lbuf_len = 3
> >>         fd_r = 6
> >>         binary = 0
> >> #6  0x1c01c18b in comm_select (delay=0) at s_bsd_poll.c:247
> >>         F = (fde_t *) 0x3c4e7570
> >>         revents = 64
> >>         delay = 469863596
> >>         num = 6
> >>         ci = 1
> >>         hdl = (PF *) 0x1c018cac <read_packet>
> >> #7  0x1c014ce6 in io_loop () at ircd.c:350
> >>         st = 3
> >>         delay = 3
> >> #8  0x1c015618 in main (argc=0, argv=0xcfbfb6ec) at ircd.c:780
> >> No locals.
> >> (gdb)
> >>
> >> Regards,
> >> Andrey
> >>
> >> Wednesday, February 23, 2005, 1:59:34 PM, you wrote:
> >>
> >> RL> Firstly, if you want help, you'll need to provide a full backtrace
> >> (bt full)
> >> RL> Also, it's noted that contrib modules are unsupported by the official
> >> RL> hybrid team, so I'm not sure who you'd talk to about getting that
> >> RL> fixed. FORCENICK is also a HybServ-provided module, so I recommend
> >> you
> >> RL> speak to the HybServ development team about that.
> >>
> >> RL> It probably has something to do with the new message tabs, with
> >> RL> additional handler fields (there are now a total of 6)
> >>
> >> RL> On Wed, 23 Feb 2005 13:47:19 +1300, Andrey Sheykhot
> >> <sheyh at xtra.co.nz> wrote:
> >> >> Hi all
> >> >>
> >> >> I'm having problem with forcenick module for hybrid ircd. It worked
> >> >> fine on 7.0.1 but it causes ircd to crash on 7.0.3.
> >> >>
> >> >> Module is located:
> >> >> http://www.wohmart.com/ircd/pub/hybrid/3-Feature/hybserv/m_forcenick.c
> >> >>
> >> >> Whenever oper issues FORCENICK command ircd core the same moment.
> >> >>
> >> >> If anyone could help me fix this module or make some workaround
> >> >> that would be highly appreciated.
> >> >>
> >> >> Best regards,
> >> >>  Andrey
> >> >>
> >> >>
> >>
> >>
> >
> >
> > --
> > Regards,
> >
> > Rachel Llorenna (frequency)
> 
> 


-- 
Regards,

Rachel Llorenna (frequency)