[5285] in Athena Bugs
pmax: moira core dump
daemon@ATHENA.MIT.EDU (Ken Raeburn)
Tue Jun 26 02:04:46 1990
Date: Tue, 26 Jun 90 02:04:26 EDT
From: Ken Raeburn <Raeburn@MIT.Edu>
To: bugs@ATHENA.MIT.EDU
Because of a mail message talking about a "frame" file system (it
should have been "evaluate", we determined later), I tried looking up
info in moira on file system "fram*". From the "filesystem" menu:
...
Command: 1
Get Filesystem Name Information
Name of Filesystem: fram*
Segmentation violation (core dumped)
[podge.MIT.EDU!raeburn]
{/afs/athena.mit.edu/astaff/project/gccdev/build}
<1119 1:34am> %
The stack trace:
> 0 _doprnt(0x7fffc6fc, 0x0, 0x0, 0x0, 0x417a50) ["doprnt.c":203, 0x428920]
1 menu_com_err_hook(0x2d9ed07, 0x0, 0x7fffc6fc, 0x100012fc, 0x417ac8) ["menu.c":95, 0x405c3c]
2 com_err_va(0x40a738, 0x7fffc728, 0x10030042, 0x19, 0x4002d4) ["com_err.c":94, 0x417a4c]
3 com_err.com_err(0x2d9ed07, 0x0, 0x1001ddb8, 0x7fffc728, 0x50) ["com_err.c":121, 0x417ac4]
I'm not sure what sources were used to build moira for the pmax, but
the code for menu_com_err_hook in menu.c in the moiradev file system
is not it.
The executable has a call to _doprnt, which is correct for my latest
version of com_err; the moiradev version calls sprintf, which is
incorrect. I know that Mark has made changes to com_err, but if it's
still going through com_err_va, I doubt he's changed the calling
sequence such that sprintf can be used.
