[155] in Athena Bugs
Vax kernel bug
daemon@ATHENA.MIT.EDU (mar@ATHENA.MIT.EDU)
Wed Apr 6 17:50:26 1988
From: <mar@ATHENA.MIT.EDU>
Date: Wed, 6 Apr 88 17:49:32 EDT
To: gmbelaus@ATHENA.MIT.EDU
Cc: henry@ATHENA.MIT.EDU, bugs@ATHENA.MIT.EDU, ilh@ATHENA.MIT.EDU
It has always been the case when I've run across that bug that the
program got the correct answer, but printf displayed it incorrectly.
For example, using xcalc, the sequence:
45 TAN
displays the value 0 instead of the expected 1. But if you then add
1 to the displayed 0, you get 2 instead of 1. When the calculator was
displaying the incorrect 0, it really did have a 1 in it's
accumulator.
Keeping this in mind, you can work around some of the problems with
this kernel bug while still running 5.5T. Also, note that programs
run under Saber will get the correct answer, because saber uses a
different printf that doesn't exercise the bug.
-Mark
