[396] in athena10
Re: Larvnet design discussion
daemon@ATHENA.MIT.EDU (Greg Hudson)
Thu Aug 7 00:43:38 2008
Message-ID: <489A7D62.5050801@mit.edu>
Date: Thu, 07 Aug 2008 00:43:14 -0400
From: Greg Hudson <ghudson@MIT.EDU>
MIME-Version: 1.0
To: Greg Hudson <ghudson@mit.edu>
CC: Timothy G Abbott <tabbott@mit.edu>, athena10@mit.edu
In-Reply-To: <1218040449.12433.127.camel@error-messages.mit.edu>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
Greg Hudson wrote:
> I'll check this in after testing it.
>
Small snag in the code to test if a gdm session is active.
Checking for someone logged into tty7 does not seem to be very robust. On my office machine, gdm is running on tty9. I'm guessing this has to do with experimenting with user switching in the past. Even if that won't be an issue for cluster machines, I'm no longer comfortable with that test.
Checking for the existence of /var/lib/gdm/*.Xservers appears to be pretty robust, but can only be performed as root.
I think I will just configure inetd to run busyd as root. It's a very short perl script which doesn't touch the input it receives from the network, so I don't think it's very security-sensitive. If someone can come up with a robust and simple check for an active gdm session which can be performed as nobody, I'll revisit that.
