[3451] in testers
Re: bugs in 8.2 Sparc?
daemon@ATHENA.MIT.EDU (Greg Hudson)
Thu May 14 14:40:04 1998
To: Brett David Rosen <bdrosen@MIT.EDU>
Cc: testers@MIT.EDU
In-Reply-To: Your message of "Thu, 14 May 1998 12:23:33 EDT."
<199805141623.MAA03991@cogsworth.mit.edu>
Date: Thu, 14 May 1998 14:39:58 EDT
From: Greg Hudson <ghudson@MIT.EDU>
> I think if you change X_START_WAIT to 45 instead of 30 it will make
> things better. (dm.c)
Both empirically and theoretically, no. What's happening is that dm
is trying three times to start the X server, and each time the X
server dies, presumably because of some inability to get at the frame
buffer. There is certainly no 30-second delay involved.
In SIPB-Athena, we added a sleep(1) before each try at starting up X
to address this problem. I'd like to take a slightly more complicated
approach this time around (to also deal with the ^P problems); I'll
submit a patch shortly.
