[gnu.g++.bug] floating point problems, g++1.35.0

rjk@sawmill.UUCP (Richard Kuhns) (05/20/89)

I've just finished compiling g++ v1.35.0 (which I downloaded from
osu-cis -- thanks guys!) with gcc v1.35 on a Sequent Symmetry running
Dynix 3.0.12.

  When compiling libg++, I was bitten by the -O bug in BitString.cc,
which was correctable by turning off optimization.  However, every
one of the tests which use floating point eventually die from an
assert() error (tRational, tFix, etc.).  Is this a known problem?
If not, I'll happily supply any info requested, or run any other tests
which may be useful.  To the best of my knowledge, gdb++ doesn't run on
a Symmetry (if I'm wrong, someone please tell me!  I'll get it ASAP).

Is there anyone else using g++ on a Symmetry who'd like to compare notes
(bearing in mind that I'm just learning C++ -- I've had 10 years
experience with C, which I understand may be a drawback at first).

Rich Kuhns
{sequent, pur-phy aka newton.physics.purdue.edu}!sawmill!rjk