[22] in 6.033-lab
Re: Another stub bug
daemon@ATHENA.MIT.EDU (Kevin 'Frostbyte' McCormick)
Wed Mar 26 02:13:56 1997
Date: Wed, 26 Mar 1997 02:13:31 -0500 (EST)
From: Kevin 'Frostbyte' McCormick <fbyte@sub-zero.mit.edu>
To: Benjie Chen <benjie@cyber.mit.edu>
Cc: Jeffrey Hu <jhu@MIT.EDU>, 6.033-lab@MIT.EDU, csapuntz@MIT.EDU
In-Reply-To: <m0w9mPM-000YqbC@cyber>
On Wed, 26 Mar 1997, Benjie Chen wrote:
> One problem that I ran into and decided to ignore it for now is
> real time expiration... since when I am doing procedure themselves,
> I don't check timers, so if timers expire then, I wait til later
> to take care of them... there is ways to do active timers, with
> signal handlers, but then we lose on not being able to pass info
> to the handlers and instead have to use global variables (yacks).
Plus, if the real procedure that's getting executed isn't expecting
signals to come along, an errant signal from your timers might interrupt a
system call in the procedure, totally screwing things up. Also, it would
be bad if the procedure decided to use alarm(), sleep() or any itimers...
> I decided that this was something not important enough to worry
> about, unless I do multithreading on the server side.
That would be the best way to do it, but that's a little excessive for
this project...
-------------------------------------------------------------------------------
Kevin "Frostbyte" McCormick fbyte@sub-zero.mit.edu
Massachusetts Institute of Technology kmccorm@mit.edu
Linux: The choice of a GNU generation.
