[154] in Athena Bugs
Vax kernel bug
daemon@ATHENA.MIT.EDU (Gregory M. Belaus)
Wed Apr 6 17:35:00 1988
From: Gregory M. Belaus <gmbelaus@ATHENA.MIT.EDU>
To: geer@ATHENA.MIT.EDU, treese@ATHENA.MIT.EDU, henry@ATHENA.MIT.EDU
Cc: bugs@ATHENA.MIT.EDU, ilh@ATHENA.MIT.EDU
Date: Wed, 06 Apr 88 16:33:39 EST
Hi,
A friend of mine (ilh) keeps tripping over the Vax kernel
floating point bug. He is doing matrix work with c++, and keeps getting
(inconsistenly) 0 as the result of multiplying two non-0 doubles that
should multiply to be 1.
Is there a way around this bug for 55T. I heard that it is
fixed in the 6.0 kernel, but this won't help him for a while. Is it
possible to copy the 6.0 kernel onto a 55t machine without side effects?
We could copy it onto a machine at TDC (non-public).
Greg
