From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 2644 invoked by alias); 13 Jun 2005 15:08:40 -0000 Mailing-List: contact gcc-bugs-help@gcc.gnu.org; run by ezmlm Precedence: bulk List-Archive: List-Post: List-Help: Sender: gcc-bugs-owner@gcc.gnu.org Received: (qmail 1996 invoked by alias); 13 Jun 2005 15:08:25 -0000 Date: Mon, 13 Jun 2005 15:08:00 -0000 Message-ID: <20050613150824.1995.qmail@sourceware.org> From: "ro at techfak dot uni-bielefeld dot de" To: gcc-bugs@gcc.gnu.org In-Reply-To: <20050607162103.21950.ro@techfak.uni-bielefeld.de> References: <20050607162103.21950.ro@techfak.uni-bielefeld.de> Reply-To: gcc-bugzilla@gcc.gnu.org Subject: [Bug libfortran/21950] All gfortran execution tests fail on Tru64 UNIX V4.0F: scalbn unresolved X-Bugzilla-Reason: CC X-SW-Source: 2005-06/txt/msg01764.txt.bz2 List-Id: ------- Additional Comments From ro at techfak dot uni-bielefeld dot de 2005-06-13 15:08 ------- Subject: Re: All gfortran execution tests fail on Tru64 UNIX V4.0F: scalbn unresolved > is very portable). What about the naïve implementation: > > double scalbn (double x, int n) { return x * pow(FLT_RADIX,n); } I haven't checked if this handles the corner cases (overflow, NaN or Inf as input) correctly. > Is FLT_RADIX available on Tru64 UNIX V4.0F and IRIX? That would at least allow > gfortran to build, even if we could enhance this code afterwards. It is, on all of Tru64 UNIX V4.0F, IRIX 5.3 and 6.5. > The fix is: add a check for scalbn in the configure.ac, add the fallback > function to c99_functions.c and the prototype to c99_protos.h. I can write that > and submit the patch for review if someone can test it on Tru64 UNIX and IRIX. I can easily perform the testing. Rainer -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=21950