On Tue, May 01, 2001 at 02:47:44PM -0600, Charles Cazabon wrote:
> Yes -- but the event which caused the core-dump appeared to happen in the
> child, so we don't know exactly what caused it. There are options to strace
> (and I assume truss) to make it also trace any children -- try adding that
> option to your truss invocation, and capture another trace.
The bastard child is ezmlm-send... I can run it on its own, I get the
following from truss:
open("mailinglist", O_RDONLY|O_NDELAY) = 5
read(5, " c o n t a c t t e s t".., 128) = 52
read(5, 0x0002E484, 128) = 0
close(5) = 0
open("listid", O_RDONLY|O_NDELAY) Err#2 ENOENT
open("headeradd", O_RDONLY|O_NDELAY) = 5
read(5, " P r e c e d e n c e : ".., 256) = 215
read(5, 0x0002E748, 256) = 0
close(5) = 0
read(0, " a s d f\n", 256) = 5
read(0, 0x0002E888, 256) = 0
write(4, " R e t u r n - P a t h :".., 139) = 139
fdsync(4, O_RDONLY|O_SYNC) = 0
fchmod(4, 0744) = 0
close(4) = 0
open("key", O_RDONLY|O_NDELAY) = 4
read(4, "80 3\0\098 2\0\0 d\0\0\0".., 32) = 32
read(4, "03 7\v\0 " =\v\0E6 C\v\0".., 32) = 32
read(4, " - k\v\0 b |\v\0D183\v\0".., 32) = 32
read(4, "B49F\v\0D6AD\v\0EEB5\v\0".., 32) = 32
read(4, "8EED\v\0A0F1\v\0 2FA\v\0".., 32) = 32
read(4, " i &\f\0E7 1\f\0A6 ;\f\0".., 32) = 32
read(4, "E2 |\f\0 t85\f\0 h93\f\0".., 32) = 32
read(4, "F1CF\f\084D8\f\01FE1\f\0".., 32) = 32
read(4, " * !\r\0 p /\r\005 8\r\0".., 32) = 16
read(4, 0x0002DB20, 32) = 0
close(4) = 0
Incurred fault #6, FLTBOUNDS %pc = 0x0001AF28
siginfo: SIGSEGV SEGV_MAPERR addr=0x00000000
Received signal #11, SIGSEGV [default]
siginfo: SIGSEGV SEGV_MAPERR addr=0x00000000
*** process killed ***
> As a side note, your mail to this list is a good example of how to properly
> submit a problem report.
Well, considering I've been reading the list for around 4 years now, I should
hope so :)
--Adam
