Re: scheduler 2.99.38 core dumps

[Arnt Gulbrandsen <arnt@gulbrandsen.priv.no>]

Matti Aarnio <mea@nic.funet.fi>
> 	A debugger back-trace from such drop would tell me a lot.

strace always looks like this:

...
    open("13998-5546", O_RDWR)              = 4
    fstat(4, {st_mode=S_IFREG|0460, st_size=2258, ...}) = 0
    read(4, "i 13998-5546\no 1099\nl <28973.1"..., 2258) = 2258
    close(4)                                = 0
    stat("../queue/13998-5546", {st_mode=S_IFREG|0640, st_size=1671, ...}) = 0
    --- SIGSEGV (Segmentation fault) ---
    +++ killed by SIGSEGV +++

Redone using gdb:
dirqueuescan(dir='.') wrksum=31 new=31

    Program received signal SIGSEGV, Segmentation fault.
    0x800d371 in strsave ()
    (gdb) where
    #0  0x800d371 in strsave ()
    #1  0x800ac3c in web_findcreate ()
    #2  0x8004b25 in link_in ()
    #3  0x80045e8 in vtxprep ()
    #4  0x8003480 in schedule ()
    #5  0x800337c in syncweb ()
    #6  0x800297d in main ()
    #7  0x8001dd4 in ___crt_dummy__ ()
    #8  0x801b3b4 in _malloc_statsbuf ()

I'm sorry, but I've no debugging info compiled in.  Can rebuild
tonight, I've been meaning to that anyway.  However, looking at
transport/13998-5546 may tell you enough:

    i 13998-5546
    o 1099
    l <28973.17206.845958957@hagbart.nvg.unit.no>
    e <venaas@nvg.unit.no>
    s smtp nvg.unit.no venaas@nvg.unit.no 2
    r+      local agulbra@troll.no agulbra 99
    m
    Received: from sabre-wulf.nvg.unit.no ([129.241.163.74]) by lupinella.troll.no with ESMTP id <13998-5546>; Tue, 22 Oct 1996 06:36:14 +0200
...
    Content-Length: 592

    s smtp nvg.unit.no venaas@nvg.unit.no 2
    r
    m
    Received: from sabre-wulf.nvg.unit.no ([129.241.163.74]) by lupinella.troll.no with ESMTP id <13998-5546>; Tue, 22 Oct 1996 06:36:14 +0200
...
    Content-Length: 592

That message was successfully delivered to me a few days ago.

--Arnt