Subject: bin/410: /bin/sh coredump
To: None <gnats-admin>
From: der Mouse <mouse@Collatz.McRCIM.McGill.EDU>
List: netbsd-bugs
Date: 08/11/1994 17:05:05
>Number: 410
>Category: bin
>Synopsis: /bin/sh coredump
>Confidential: no
>Severity: serious
>Priority: high
>Responsible: gnats-admin (Utility Bug People)
>State: open
>Class: sw-bug
>Submitter-Id: net
>Arrival-Date: Thu Aug 11 17:05:03 1994
>Originator: der Mouse
>Organization:
‰¼ïòçáîéúáôéïî ïæ ÐÒ áõôèïò ¨íõìôéðìå ìéîåó©¾
>Release:
>Environment:
SPARC IPC
Kernel built from sources a couple of days old
sh built from sources checked against sun-lamp with sup as of
minutes ago (23:12 UTC, Aug 11 1994)
rest of user-land from the Aug 1 tarballs
>Description:
sh coredumps on certain (incorrect) lines. I don't know
precisely what it's getting upset over, but I have a repeatable
example.
>How-To-Repeat:
/bin/sh -c '( echo foo | > /tmp/$$ ; cat /tmp/$$ )'
>Fix:
Sorry. I was going to investigate further, but gdb is
completely useless, it turns out, not even stopping at the
segfault. (I notice the GDB option is turned off, but it's
turned off in deraadt's SPARC config as well, so I haven't
quite dared turn it on.)
>Audit-Trail:
>Unformatted:
------------------------------------------------------------------------------