[4209] in testers
Re: one more...
daemon@ATHENA.MIT.EDU (Robert A Basch)
Mon Jun 7 20:21:11 1999
Message-Id: <199906080021.UAA54367@boom.mit.edu>
To: Alex T Prengel <alexp@MIT.EDU>
Cc: rbasch@MIT.EDU, testers@MIT.EDU
In-Reply-To: Your message of "Mon, 07 Jun 1999 17:07:43 EDT."
<199906072107.RAA01168@andrew.mit.edu>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Date: Mon, 07 Jun 1999 20:21:04 -0400
From: Robert A Basch <rbasch@MIT.EDU>
> ... Athena 8.3 SGI glitch. When I type "sysinfo" on an 8.3 Indy (but not O2)
> I get:
>
> ~ athena% sysinfo
>
>
> G E N E R A L I N F O R M A T I O N
>
> Host Name is andrew.mit.edu
> Host Aliases is andrew
> Host Address(es) is 18.152.0.97
> Host ID is 12980061
> Serial Number is 1762135215
> Manufacturer is sgi (Silicon Graphics, Inc.)
> System Model is Indy
> Main Memory is 64 MB
> Bus error
I figured out the problem. It dies after calling swapctl() to obtain the
amount of virtual memory. It passes in a pointer to a 32-bit int, but the
system returns a 64-bit value.
On IRIX, SC_GETSWAPVIRT returns the amount of virtual swap, which is non-0
here only on older Indy's which have an additional swap file; apparently,
a 0 value allows the program to survive.
I'll work on a patch to address this; since sysinfo 1) misuses the swapctl()
SC_GETSWAPVIRT function, and 2) passes the result in the wrong unit to
a conversion function expecting a 32-bit value, maybe the easiest thing
will be to just skip the virtual memory query on sgi...
Bob
