INN 2.4.0 dumps core on Solaris 9

[Russ Allbery]

Nathan Coraor <nate at cse.psu.edu> writes:

>   Hmm.  Perhaps a little more background:

>   13953/1:        open("/var/news/spool/overview/a/v/u/m/alt.visa.us.marriage-based.IDX", O_RDWR|O_CREAT, 0664) = 57
>   13953/1:        fstat(57, 0xFFBFF0A8)                           = 0
>   13953/1:        fcntl(57, F_GETFD, 0x00000000)                  = 0
>   13953/1:        fcntl(57, F_SETFD, 0x00000001)                  = 0
>   13953/1:        open("/var/news/spool/overview/a/v/u/m/alt.visa.us.marriage-based.DAT", O_RDWR|O_APPEND|O_CREAT, 0664) = 58
>   13953/1:        fcntl(58, F_GETFD, 0x00000000)                  = 0
>   13953/1:        fcntl(58, F_SETFD, 0x00000001)                  = 0
>   13953/1:        time()                                          = 1067372524
>   13953/1:        fcntl(10, F_SETLKW, 0xFFBFF188)                 = 0
>   13953/1:        write(58, " 2 2 6 7 3 5\t R e :   E".., 308)    = 308
>   13953/1:        lseek(58, 0, SEEK_CUR)                          = 308
>   13953/1:        pwrite(57, "\0\0\0\0\0\001 4 ?9ECFEC".., 36, 4608) = 36
>   13953/1:        memcntl(0xFEE84000, 8192, MC_SYNC, MS_ASYNC, 0, 0) = 0
>   13953/1:        fcntl(10, F_SETLKW, 0xFFBFF188)                 = 0
>   13953/1:            Incurred fault #6, FLTBOUNDS  %pc = 0xFEFB3464
>   13953/1:              siginfo: SIGSEGV SEGV_MAPERR addr=0x00000000
>   13953/1:            Received signal #11, SIGSEGV [default]
>   13953/1:              siginfo: SIGSEGV SEGV_MAPERR addr=0x00000000
>   13955:  read(0, 0x0025A69C, 8192)                       = 0
>   13955:  llseek(4, 0, SEEK_CUR)                          Err#29 ESPIPE
>   13955:  close(4)                                        = 0

>   So it looks like it was trying to write the overview when it crashed...

It just finished writing overview and was trying to write history.  What
I'm wondering is if that overview entry contains the message ID.  If so,
that narrows the search for what's munging the message ID quite a bit.  If
not, that still narrows it, but not quite as much.

-- 
Russ Allbery (rra at stanford.edu)             <http://www.eyrie.org/~eagle/>

    Please send questions to the list rather than mailing me directly.
     <http://www.eyrie.org/~eagle/faqs/questions.html> explains why.