MPFR issue on sun4v machines has been solved.

69 views
Skip to first unread message

Dr. David Kirkby

unread,
Oct 22, 2009, 9:53:04 AM10/22/09
to mp...@loria.fr, sage-...@googlegroups.com
There were 20 test failures of mpfr due to a bug in the memset implementation on
Solaris 10 on sun4m machines (those based on the T1, T2, or T2+ processors).


http://sagetrac.org/sage_trac/ticket/6453


The T2+ is used in the Sun T5240 at t2.math.washington.edu.

Sun have sent me a 'T' patch, which I have installed on t2.math.washington.edu.
This fixes the bug. This patch will soon be available on Sunsolve soon as
142542-01, though currently it is not generally available.

Dr. David Kirkby

unread,
Oct 22, 2009, 12:31:45 PM10/22/09
to sage-...@googlegroups.com, mp...@loria.fr
Dr. David Kirkby wrote:
> There were 20 test failures of mpfr due to a bug in the memset implementation on
> Solaris 10 on sun4m machines (those based on the T1, T2, or T2+ processors).


Oops, sun4m are quite old. It is sun4v which is effected by this bug.

Reply all
Reply to author
Forward
0 new messages