Re: panic: tcp_log_addrs: string too long

Norikatsu Shigemura wrote:
> Hi andre!
> 
> 	I have a tcp_log_addrs panic issue like following (with my patch).
> 	I didn't know why 'string too long'.  So I added following patch
> 	to report this issue.  Do you have any idea?

I just woke up from last nights extended stay at the hacking lounge
at BSDCan.  My brain isn't yet working and I have to head out to catch
up with the remaining crew to do some sightseeing here in Ottawa.
Will look at it as soon as we get back later today.

-- 
Andre

> - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
> panic: tcp_log_addrs: string too long:size=178, s=                                                                                                               TCP: [192.168.36.6]:54312 to [219.127.74.122]:22 tcpflags 0x10<ACK>
> cpuid = 0
> KDB: stack backtrace:
> db_trace_self_wrapper(c06b36fd,f88bf8e8,c04a9b31,c06c9a9a,0,...) at db_trace_self_wrapper+0x26
> kdb_backtrace(c06c9a9a,0,c06bbedb,f88bf8f4,0,...) at kdb_backtrace+0x29
> panic(c06bbedb,c06a16de,b2,b2,c6f4c300,...) at panic+0x111
> tcp_log_addrs(f88bfaac,c6bb3828,0,0,2,...) at tcp_log_addrs+0x304
> syncache_expand(f88bfaac,f88bfad4,c6bb3828,f88bfaec,c6ba8800,...) at syncache_expand+0x3dc
> tcp_input(c6ba8800,14,c4e4fc00,1,0,...) at tcp_input+0x7a5
> ip_input(c6ba8800,c6ba8800,800,c4e4fc00,800,...) at ip_input+0x6a0
> netisr_dispatch(2,c6ba8800,154,3,0,...) at netisr_dispatch+0x52
> ether_demux(c4e4fc00,c6ba8800,3,0,3,...) at ether_demux+0x1c1
> ether_input(c4e4fc00,c6ba8800,1,c5018bb4,c4e4fc00,...) at ether_input+0x3c3
> fwe_as_input(c5018b84,0,c09cb9f9,8ca,2a7f3318,...) at fwe_as_input+0x134
> fwohci_intr_body(c4e80d80,f88bfcb0,c04da23f,c5013000,1,...) at fwohci_intr_body+0x2cc
> fwohci_complete(c5013000,1,f88bfc94,c067afc7,18bfcbc,...) at fwohci_complete+0x1c
> taskqueue_run(c4e80d80,f88bfcf4,c048d1bb,0,0,...) at taskqueue_run+0x17f
> taskqueue_swi_run(0,0,0,0,0,...) at taskqueue_swi_run+0x13
> ithread_loop(c4e80a80,f88bfd38,0,0,0,...) at ithread_loop+0x1cb
> fork_exit(c048cff0,c4e80a80,f88bfd38) at fork_exit+0xa9
> fork_trampoline() at fork_trampoline+0x8
> --- trap 0, eip = 0, esp = 0xf88bfd70, ebp = 0 ---
> KDB: enter: panic
> [thread pid 16 tid 100010 ]
> Stopped at      kdb_enter+0x32: leave
> db> reset
> - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
> 
> 
> - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
> --- sys/netinet/tcp_subr.c.orig Sat May 19 18:08:13 2007
> +++ sys/netinet/tcp_subr.c      Sun May 20 13:47:16 2007
> _at__at_ -2148,6 +2148,6 _at__at_
>         if (th)
>                 sprintf(sp, " tcpflags 0x%b", th->th_flags, PRINT_TH_FLAGS);
>         if (s[size] != '\0')
> -               panic("%s: string too long", __func__);
> +               panic("%s: string too long:size=%u, s=%*s", __func__, size, size, s);
>         return (s);
>  }
> - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
> 
>