Re: [PATCH] New division operators, and optimization for fractions

[Mark H Weaver <mhw@netris.org>]

[-- Attachment #1: Type: text/plain, Size: 694 bytes --]

ludo@gnu.org (Ludovic Courtès) writes:
> This patch series seems to break compilation on i686-linux-gnu:
>
>   http://hydra.nixos.org/build/913079
>
> The relevant part of the log is:
>
>   Running numbers.test
>   FAIL: numbers.test: Number-theoretic division: truncate/: mixed types: (130.0 10/7)
>   FAIL: numbers.test: Number-theoretic division: truncate/: mixed types: (130.0 -10/7)
>   FAIL: numbers.test: Number-theoretic division: truncate/: mixed types: (-130.0 10/7)
>   FAIL: numbers.test: Number-theoretic division: truncate/: mixed types: (-130.0 -10/7)

The following patch should fix it.

     Best,
      Mark

PS: Thanks for importing log1p so quickly!



[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #2: Use trunc in scm_i_inexact_truncate_divide --]
[-- Type: text/x-diff, Size: 1115 bytes --]

From 4a9c388d3c837244e358916beda59d4634121723 Mon Sep 17 00:00:00 2001
From: Mark H Weaver <mhw@netris.org>
Date: Tue, 15 Feb 2011 06:10:06 -0500
Subject: [PATCH] Use trunc in scm_i_inexact_truncate_divide

* libguile/numbers.c (scm_i_inexact_truncate_divide): Use trunc instead
  of floor and ceil.  Important for consistency with
  scm_truncate_quotient and scm_truncate_remainder.
---
 libguile/numbers.c |    7 ++-----
 1 files changed, 2 insertions(+), 5 deletions(-)

diff --git a/libguile/numbers.c b/libguile/numbers.c
index 59d8e74..7c4ea1b 100644
--- a/libguile/numbers.c
+++ b/libguile/numbers.c
@@ -2450,11 +2450,8 @@ scm_i_inexact_truncate_divide (double x, double y, SCM *qp, SCM *rp)
     scm_num_overflow (s_scm_truncate_divide);  /* or return a NaN? */
   else
     {
-      double q, r, q1;
-      /* FIXME: Use trunc, after it has been imported from gnulib */
-      q1 = x / y;
-      q = (q1 >= 0) ? floor (q1) : ceil (q1);
-      r = x - q * y;
+      double q = trunc (x / y);
+      double r = x - q * y;
       *qp = scm_from_double (q);
       *rp = scm_from_double (r);
     }
-- 
1.7.1