svn commit: r427317 - in head/math/octave-forge-interval: . files

[Stephen Montgomery-Smith]

Author: stephen
Date: Mon Nov 28 13:57:58 2016
New Revision: 427317
URL: https://svnweb.freebsd.org/changeset/ports/427317

Log:
  - A better patch to fix the problem purportedly solved in r427058.
  - Bump portrevision.
  
  Submitted by:	Oliver Heimlich <oheim at posteo.de>

Modified:
  head/math/octave-forge-interval/Makefile
  head/math/octave-forge-interval/files/patch-mpfr__to__string__d.cc

Modified: head/math/octave-forge-interval/Makefile
==============================================================================
--- head/math/octave-forge-interval/Makefile	Mon Nov 28 13:53:22 2016	(r427316)
+++ head/math/octave-forge-interval/Makefile	Mon Nov 28 13:57:58 2016	(r427317)
@@ -3,7 +3,7 @@
 
 PORTNAME=	octave-forge-interval
 PORTVERSION=	2.0.0
-PORTREVISION=	1
+PORTREVISION=	2
 CATEGORIES=	math
 
 MAINTAINER=	stephen at FreeBSD.org

Modified: head/math/octave-forge-interval/files/patch-mpfr__to__string__d.cc
==============================================================================
--- head/math/octave-forge-interval/files/patch-mpfr__to__string__d.cc	Mon Nov 28 13:53:22 2016	(r427316)
+++ head/math/octave-forge-interval/files/patch-mpfr__to__string__d.cc	Mon Nov 28 13:57:58 2016	(r427317)
@@ -1,15 +1,11 @@
---- mpfr_to_string_d.cc.orig	2016-11-24 18:19:39 UTC
+--- mpfr_to_string_d.cc.orig	2016-11-20 13:47:35 UTC
 +++ mpfr_to_string_d.cc
-@@ -201,10 +201,10 @@ DEFUN_DLD (mpfr_to_string_d, args, nargo
+@@ -201,7 +201,7 @@ DEFUN_DLD (mpfr_to_string_d, args, nargo
                // Make subnormal numbers use the exponent -1022
                if (exponent < std::numeric_limits <double>::min_exponent)
                  {
 -                  mantissa /= std::pow (2.0, 
-+                  mantissa /= std::pow (2.0, (int)(
++                  mantissa /= uint64_t (1) << (
                                          std::numeric_limits
                                            <double>::min_exponent - 1
--                                        - exponent);
-+                                        - exponent));
-                   exponent = std::numeric_limits <double>::min_exponent - 1;
-                 }
-             
+                                         - exponent);