From: Jarkko Hietaniemi <jhi@iki.fi>
Date: Wed, 15 Aug 2001 12:48:32 +0000 (+0000)
Subject: Re-establish the fp overflow detection for VAX VMS; there
X-Git-Url: http://git.shadowcat.co.uk/gitweb/gitweb.cgi?a=commitdiff_plain;h=67597c89125e7e144f9ba60f5b1fe23b951286d8;p=p5sagit%2Fp5-mst-13.2.git

Re-establish the fp overflow detection for VAX VMS; there
is no easy way to have the IEEE fp silent overflow semantics.
(in Alpha VMS we still will use IEEE fp by default-- but it
is still possible to configure Perl to use G_FLOAT)

p4raw-id: //depot/perl@11679
---

diff --git a/numeric.c b/numeric.c
index 68c1671..c601943 100644
--- a/numeric.c
+++ b/numeric.c
@@ -572,6 +572,25 @@ S_mulexp10(NV value, I32 exponent)
 	negative = 1;
 	exponent = -exponent;
     }
+    /* Avoid %SYSTEM-F-FLTOVF_F sans VAXC$ESTABLISH.
+     * In VAX VMS we by default use the D_FLOAT double format,
+     * and that format does not have *easy* capabilities [1] for
+     * overflowing doubles 'silently' as IEEE fp does.  Therefore we
+     * need to detect early whether we would overflow (this is
+     * the behaviour of the native string-to-float conversion routines,
+     * and therefore the behaviour of native applications, too.)
+     *
+     * [1] VAXC$EXTABLISH is the capability but it is basically a signal
+     * handler setup routine, and one cannot return from a fp exception
+     * handler and except much anything useful. */
+#if defined(VMS) && !defined(__IEEE_FP)
+#  if defined(__DECC_VER) && __DECC_VER <= 50390006
+    /* __F_FLT_MAX_10_EXP - 5 == 33 */
+    if (!negative &&
+          (log10(value) + exponent) >= (__F_FLT_MAX_10_EXP - 5))
+        return NV_MAX;
+#  endif
+#endif
     for (bit = 1; exponent; bit <<= 1) {
 	if (exponent & bit) {
 	    exponent ^= bit;