Update to 1.36:

2011-02-26 v1.36 pjacklam (6362 tests)
  * Change bigintpm.inc to reflect recent changes in the Math::BigInt
    distribution (Peter John Acklam).
  * Use a _nok() function more similar to the one in Math::BigInt::Calc
    (Peter John Acklam).

2011-02-08 v1.35 pjacklam (6361 tests)
  * Rename files for testing signature, module loading, and POD so the names
    are within the 8+3 character limit (Peter John Acklam).
  * Rename method _nok_ok() to the correct _nok(). There ought to have been a
    test catching an error like that (Peter John Acklam).
  * Fix _nok() giving wrong output when second input argument is zero
    (Peter John Acklam).
  * Fix _nok() so it doesn't modify its second input arg (Peter John Acklam).
  * Update the included Devel::CheckLib to most recent version as suggested in
    RE #63055 (Peter John Acklam).
  * Apply "chmod 0644" to the few test scripts that don't already have that
    mode (Peter John Acklam).

2011-02-07 v1.34 pjacklam (6361 tests)
  * Rename _num() to _str(). The old _num() did exactly what _str() is supposed
    to do, according to the API documentation (Peter John Acklam).
  * Add a _num() function which (currently) simply numifies the output from
    _str() (Peter John Acklam).
  * Clean up whitespace (Peter John Acklam).
  * Fix POD errors (Peter John Acklam).
  * Add _nok() method. Now the old claim that Math::BigInt::GMP conforms to API
    version 2 is actually true (Peter John Acklam).
  * Edit the test files that were copied from the Math::BigInt distribution, so
    we now test Math::BigInt against Math::BigInt::GMP, not Math::BigInt::Calc.
    I had forgotten this when I copied the test files from the Math::BigInt
    distribution. This reduces the total test count, since some test are not
    executed with Math::BigInt::GMP (Peter John Acklam).
  * Replace morse code in 'README' with proper text (Peter John Acklam).
  * Include '01-load.t' for explicitly testing module loading (Peter John
    Acklam).
  * Use more generic code in 'pod.t' and 'pod_cov.t' (Peter John Acklam).

2011-01-30 v1.33 pjacklam (6411 tests)
  * Fix _modinv() so that it works the same way as _modinv() in other
    Math::BigInt libraries: The output arguments are an object and the
    corresponding sign, not undef (Peter John Acklam).
  * Include most recent versions of the test files from the Math-BigInt
    distribution (bigfltpm.inc, bigfltpm.t, bigintpm.inc, bigintpm.t,
    biglog.t, and bigroot.t) (Peter John Acklam).
  * Include generic SIGNATURE test file (Peter John Acklam).
  * Required version of Math::BigInt is now 1.99_05 (Peter John Acklam).

3 files changed, 14 insertions, 15 deletions

diff --git a/math/p5-Math-BigInt-GMP/Makefile b/math/p5-Math-BigInt-GMP/Makefile
index 10c13fba480..d9b5bc454c2 100644
--- a/math/p5-Math-BigInt-GMP/Makefile
+++ b/math/p5-Math-BigInt-GMP/Makefile
@@ -1,9 +1,8 @@
-# $NetBSD: Makefile,v 1.6 2011/08/14 13:38:22 obache Exp $
+# $NetBSD: Makefile,v 1.7 2011/08/16 22:21:45 wiz Exp $
 #
 
-DISTNAME=		Math-BigInt-GMP-1.32
+DISTNAME=		Math-BigInt-GMP-1.36
 PKGNAME=		p5-${DISTNAME}
-PKGREVISION=		1
 CATEGORIES=		math perl5
 MASTER_SITES=		${MASTER_SITE_PERL_CPAN:=Math/}
 
@@ -12,7 +11,7 @@ HOMEPAGE=		http://search.cpan.org/dist/Math-BigInt-GMP/
 COMMENT=		Perl 5 module for arbitrary size integer math via the GMP library
 LICENSE=		${PERL5_LICENSE}
 
-PERL5_REQD+=		5.12.2nb1 # for Math::BigInt 1.90
+PERL5_REQD+=		5.14.1 # for Math::BigInt 1.994
 
 PKG_DESTDIR_SUPPORT=	user-destdir
 
diff --git a/math/p5-Math-BigInt-GMP/distinfo b/math/p5-Math-BigInt-GMP/distinfo
index e307487f3c0..2547f5f7782 100644
--- a/math/p5-Math-BigInt-GMP/distinfo
+++ b/math/p5-Math-BigInt-GMP/distinfo
@@ -1,6 +1,6 @@
-$NetBSD: distinfo,v 1.3 2010/09/23 21:58:13 sno Exp $
+$NetBSD: distinfo,v 1.4 2011/08/16 22:21:45 wiz Exp $
 
-SHA1 (Math-BigInt-GMP-1.32.tar.gz) = 8807d88ff3d003a5f5192489797f2da8b010cfdc
-RMD160 (Math-BigInt-GMP-1.32.tar.gz) = f10bc70a54596a4945fe5000c643ce70bc4f2829
-Size (Math-BigInt-GMP-1.32.tar.gz) = 49460 bytes
-SHA1 (patch-aa) = 9697876f32b3ea24f43516ed909044e2cf6a795e
+SHA1 (Math-BigInt-GMP-1.36.tar.gz) = e98bf0bd2349b292f892d530d564e1f7dc158c29
+RMD160 (Math-BigInt-GMP-1.36.tar.gz) = affa8eef781d41be40ebc3ea775ccd80dbfa9be3
+Size (Math-BigInt-GMP-1.36.tar.gz) = 53683 bytes
+SHA1 (patch-aa) = 506eaba617f3b5cdf6ef2bff8d46a26452d6e7e7
diff --git a/math/p5-Math-BigInt-GMP/patches/patch-aa b/math/p5-Math-BigInt-GMP/patches/patch-aa
index 422d7e09956..088e8303386 100644
--- a/math/p5-Math-BigInt-GMP/patches/patch-aa
+++ b/math/p5-Math-BigInt-GMP/patches/patch-aa
@@ -1,17 +1,17 @@
-$NetBSD: patch-aa,v 1.2 2010/09/23 21:58:13 sno Exp $
+$NetBSD: patch-aa,v 1.3 2011/08/16 22:21:45 wiz Exp $
 
 From p5-Math-GMP/patches/patch-aa:
 
 Trick perl into emitting use of -lgmp.  Actual library
 location will be handled via buildlink / wrapper.
 
---- Makefile.PL.orig	2010-09-14 16:58:27.000000000 +0000
+--- Makefile.PL.orig	2011-02-26 14:34:32.000000000 +0000
 +++ Makefile.PL
 @@ -53,6 +53,6 @@ WriteMakefile(
- 				Math::BigInt => 1.90,
- 				XSLoader => 0.02,
- 			   },
--    'LIBS'		=> ['-lgmp'],
+                                 Math::BigInt => 1.993,
+                                 XSLoader => 0.02,
+                            },
+-    'LIBS'              => ['-lgmp'],
 +    'LIBS'		=> ["-L$ENV{PREFIX}/lib -lgmp"],
      'SIGN' => 1,
  );