diff -r b150d6f590fc -r d3e603e7c17c patches/glibc/2.3.6/210-fp-byteorder.patch --- a/patches/glibc/2.3.6/210-fp-byteorder.patch Mon Jul 28 21:08:01 2008 +0000 +++ /dev/null Thu Jan 01 00:00:00 1970 +0000 @@ -1,205 +0,0 @@ -Taken from http://sources.redhat.com/ml/crossgcc/2004-02/msg00104.html -Author: addsub@eyou.com -Target: ARM - -Fixes http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/testsuite/gcc.c-torture/execute/920501-8.c -and makes printf("%f", 1.0) work. - -Lennert Buytenhek wrote in http://sources.redhat.com/ml/crossgcc/2004-09/msg00115.html : - It ... fixes the 'printf("%f\n", 0.5); prints 0.000000' and general 'floating point - is broken' on my big-endian hardfloat FPA ARM platform. ... - It's definitely needed for hardfloat. So I'd think it's needed for - big-endian systems in any case, and for VFP on little-endian systems - too. Someone would have to verify that though. - -Lennert Buytenhek wrote in http://sources.redhat.com/ml/crossgcc/2004-09/msg00123.html - I just had a look at glibc-20040830, and [this patch] is still needed and useful - for this version. glibc-20040830 out-of-the-box still contains the - following wrong assumptions: - - sysdeps/arm/bits/endian.h: float word order is big endian (which it is - not on vfp systems) - - sysdeps/arm/gmp-mparam.h: IEEE doubles are mixed endian (which they - are not on big endian systems, neither on vfp systems) - - sysdeps/arm/ieee754.h: IEEE doubles are in little endian byte order - (which they are not on big endian systems) - [This patch] seems the right solution for all of these issues. - -Dimitry Andric wrote in http://sources.redhat.com/ml/crossgcc/2004-09/msg00132.html : - It's even needed for glibc CVS, AFAICS. - The patch hunk which modifies glibc.new/sysdeps/arm/bits/endian.h - (currently at version 1.4) is only needed for proper VFP operation. - But the hunk which modifies sysdeps/arm/gmp-mparam.h, and the hunk - that deletes sysdeps/arm/ieee754.h (yes, this IS correct), are needed - for proper operation of *any* FP model on big endian ARM. - -See also discussion in followups to -http://sources.redhat.com/ml/crossgcc/2004-05/msg00245.html) - -Message-ID: <276985760.37584@eyou.com> -Received: from unknown (HELO eyou.com) (172.16.2.2) - by 0.0.0.0 with SMTP; Tue, 17 Feb 2004 10:42:40 +0800 -Received: (qmail 8238 invoked by uid 65534); 17 Feb 2004 10:42:38 +0800 -Date: 17 Feb 2004 10:42:38 +0800 -Message-ID: <20040217104238.8237.qmail@eyou.com> -From: "add" -To: dank@kegel.com -Reply-To: "add" -Subject: Re:   problem while building arm vfp softfloat gcc ` - -Hi, Dan, This is a patch I applied to my glibc-2.3.2, then my softfloat -toolchain can printf("%f\n",1.0). So you may have a try of this - - -diff -uNrp glibc.old/sysdeps/arm/bits/endian.h glibc.new/sysdeps/arm/bits/endian.h ---- glibc.old/sysdeps/arm/bits/endian.h 1999-04-12 11:59:13.000000000 -0400 -+++ glibc.new/sysdeps/arm/bits/endian.h 2004-02-12 09:15:13.000000000 -0500 -@@ -9,4 +9,9 @@ - #else - #define __BYTE_ORDER __LITTLE_ENDIAN - #endif -+ -+#ifdef __VFP_FP__ -+#define __FLOAT_WORD_ORDER __BYTE_ORDER -+#else - #define __FLOAT_WORD_ORDER __BIG_ENDIAN -+#endif -diff -uNrp glibc.old/sysdeps/arm/gmp-mparam.h glibc.new/sysdeps/arm/gmp-mparam.h ---- glibc.old/sysdeps/arm/gmp-mparam.h 2001-07-07 15:21:19.000000000 -0400 -+++ glibc.new/sysdeps/arm/gmp-mparam.h 2004-02-12 09:15:13.000000000 -0500 -@@ -26,5 +26,13 @@ MA 02111-1307, USA. */ - #define BITS_PER_SHORTINT 16 - #define BITS_PER_CHAR 8 - --#define IEEE_DOUBLE_BIG_ENDIAN 0 --#define IEEE_DOUBLE_MIXED_ENDIAN 1 -+#if defined(__ARMEB__) -+# define IEEE_DOUBLE_MIXED_ENDIAN 0 -+# define IEEE_DOUBLE_BIG_ENDIAN 1 -+#elif defined(__VFP_FP__) -+# define IEEE_DOUBLE_MIXED_ENDIAN 0 -+# define IEEE_DOUBLE_BIG_ENDIAN 0 -+#else -+# define IEEE_DOUBLE_BIG_ENDIAN 0 -+# define IEEE_DOUBLE_MIXED_ENDIAN 1 -+#endif -diff -uNrp glibc.old/sysdeps/arm/ieee754.h glibc.new/sysdeps/arm/ieee754.h ---- glibc.old/sysdeps/arm/ieee754.h 2001-07-07 15:21:19.000000000 -0400 -+++ glibc.new/sysdeps/arm/ieee754.h 1969-12-31 19:00:00.000000000 -0500 -@@ -1,115 +0,0 @@ --/* Copyright (C) 1992, 1995, 1996, 1998 Free Software Foundation, Inc. -- This file is part of the GNU C Library. -- -- The GNU C Library is free software; you can redistribute it and/or -- modify it under the terms of the GNU Lesser General Public -- License as published by the Free Software Foundation; either -- version 2.1 of the License, or (at your option) any later version. -- -- The GNU C Library is distributed in the hope that it will be useful, -- but WITHOUT ANY WARRANTY; without even the implied warranty of -- MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU -- Lesser General Public License for more details. -- -- You should have received a copy of the GNU Lesser General Public -- License along with the GNU C Library; if not, write to the Free -- Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA -- 02111-1307 USA. */ -- --#ifndef _IEEE754_H -- --#define _IEEE754_H 1 --#include -- --#include -- --__BEGIN_DECLS -- --union ieee754_float -- { -- float f; -- -- /* This is the IEEE 754 single-precision format. */ -- struct -- { -- unsigned int mantissa:23; -- unsigned int exponent:8; -- unsigned int negative:1; -- } ieee; -- -- /* This format makes it easier to see if a NaN is a signalling NaN. */ -- struct -- { -- unsigned int mantissa:22; -- unsigned int quiet_nan:1; -- unsigned int exponent:8; -- unsigned int negative:1; -- } ieee_nan; -- }; -- --#define IEEE754_FLOAT_BIAS 0x7f /* Added to exponent. */ -- -- --union ieee754_double -- { -- double d; -- -- /* This is the IEEE 754 double-precision format. */ -- struct -- { -- unsigned int mantissa0:20; -- unsigned int exponent:11; -- unsigned int negative:1; -- unsigned int mantissa1:32; -- } ieee; -- -- /* This format makes it easier to see if a NaN is a signalling NaN. */ -- struct -- { -- unsigned int mantissa0:19; -- unsigned int quiet_nan:1; -- unsigned int exponent:11; -- unsigned int negative:1; -- unsigned int mantissa1:32; -- } ieee_nan; -- }; -- --#define IEEE754_DOUBLE_BIAS 0x3ff /* Added to exponent. */ -- -- --/* The following two structures are correct for `new' floating point systems but -- wrong for the old FPPC. The only solution seems to be to avoid their use on -- old hardware. */ -- --union ieee854_long_double -- { -- long double d; -- -- /* This is the IEEE 854 double-extended-precision format. */ -- struct -- { -- unsigned int exponent:15; -- unsigned int empty:16; -- unsigned int negative:1; -- unsigned int mantissa1:32; -- unsigned int mantissa0:32; -- } ieee; -- -- /* This is for NaNs in the IEEE 854 double-extended-precision format. */ -- struct -- { -- unsigned int exponent:15; -- unsigned int empty:16; -- unsigned int negative:1; -- unsigned int mantissa1:32; -- unsigned int mantissa0:30; -- unsigned int quiet_nan:1; -- unsigned int one:1; -- } ieee_nan; -- }; -- --#define IEEE854_LONG_DOUBLE_BIAS 0x3fff -- --__END_DECLS -- --#endif /* ieee754.h */ - -Signed-off-by: Robert P. J. Day