[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[Commits] r17343 - in /fsf/trunk/libc: ./ elf/ math/ resolv/ sysdeps/i386/fpu/ sysdeps/i386/i686/multiarch/ sysdeps/sparc/sparc32/fpu/...
- To: commits@xxxxxxxxxx
- Subject: [Commits] r17343 - in /fsf/trunk/libc: ./ elf/ math/ resolv/ sysdeps/i386/fpu/ sysdeps/i386/i686/multiarch/ sysdeps/sparc/sparc32/fpu/...
- From: eglibc@xxxxxxxxxx
- Date: Thu, 01 Mar 2012 00:01:43 -0000
Author: eglibc
Date: Thu Mar 1 00:01:42 2012
New Revision: 17343
Log:
Import glibc-mainline for 2012-03-01
Added:
fsf/trunk/libc/sysdeps/sparc/sparc32/fpu/w_sqrt.S
fsf/trunk/libc/sysdeps/sparc/sparc32/fpu/w_sqrtf.S
fsf/trunk/libc/sysdeps/sparc/sparc32/sparcv9/fpu/w_sqrt.S
fsf/trunk/libc/sysdeps/sparc/sparc32/sparcv9/fpu/w_sqrtf.S
fsf/trunk/libc/sysdeps/sparc/sparc64/fpu/w_sqrt.S
fsf/trunk/libc/sysdeps/sparc/sparc64/fpu/w_sqrtf.S
fsf/trunk/libc/sysdeps/x86_64/fpu/e_expf.S
Modified:
fsf/trunk/libc/ChangeLog
fsf/trunk/libc/NEWS
fsf/trunk/libc/elf/Makefile
fsf/trunk/libc/elf/rtld.c
fsf/trunk/libc/math/libm-test.inc
fsf/trunk/libc/resolv/res_query.c
fsf/trunk/libc/sysdeps/i386/fpu/libm-test-ulps
fsf/trunk/libc/sysdeps/i386/i686/multiarch/strcasecmp_l-c.c
fsf/trunk/libc/sysdeps/i386/i686/multiarch/strcmp.S
fsf/trunk/libc/sysdeps/i386/i686/multiarch/strncase_l-c.c
fsf/trunk/libc/sysdeps/x86_64/fpu/libm-test-ulps
Modified: fsf/trunk/libc/ChangeLog
==============================================================================
--- fsf/trunk/libc/ChangeLog (original)
+++ fsf/trunk/libc/ChangeLog Thu Mar 1 00:01:42 2012
@@ -1,3 +1,55 @@
+2012-02-29 Joseph Myers <joseph@xxxxxxxxxxxxxxxx>
+
+ [BZ #13786]
+ * sysdeps/i386/i686/multiarch/strcmp.S [USE_AS_STRCASECMP_L]: Do
+ not include ../strcmp.S.
+ [USE_AS_STRNCASECMP_L]: Likewise.
+ * sysdeps/i386/i686/multiarch/strcasecmp_l-c.c
+ (__strcasecmp_l_ia32): Define as alias to __strcasecmp_l_nonascii.
+ * sysdeps/i386/i686/multiarch/strncase_l-c.c
+ (__strncasecmp_l_ia32): Define as alias to
+ __strncasecmp_l_nonascii.
+
+ [BZ #5794]
+ * math/libm-test.inc (expm1_test): Add test for bug 5794.
+ * sysdeps/i386/fpu/libm-test-ulps: Update.
+ * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
+
+ * sysdeps/i386/fpu/libm-test-ulps: Reduce some expected errors.
+ * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
+
+2012-02-29 Jeff Law <law@xxxxxxxxxx>
+
+ * resolv/res_query.c (__libc_res_nquerydomain): Avoid
+ out of bounds read.
+
+2012-02-29 Marek Polacek <polacek@xxxxxxxxxx>
+
+ [BZ #13706]
+ * elf/rtld.c (dl_main): Always set l_used to 1 for vDSO.
+ * elf/Makefile: Add rules to run tst-unused-dep.out.
+
+2012-02-28 David S. Miller <davem@xxxxxxxxxxxxx>
+
+ * sysdeps/sparc/sparc32/fpu/w_sqrt.S: New file.
+ * sysdeps/sparc/sparc32/fpu/w_sqrtf.S: New file.
+ * sysdeps/sparc/sparc32/sparcv9/fpu/w_sqrt.S: New file.
+ * sysdeps/sparc/sparc32/sparcv9/fpu/w_sqrtf.S: New file.
+ * sysdeps/sparc/sparc64/fpu/w_sqrt.S: New file.
+ * sysdeps/sparc/sparc64/fpu/w_sqrtf.S: New file.
+
+2012-02-29 Joseph Myers <joseph@xxxxxxxxxxxxxxxx>
+
+ * math/libm-test.inc (llround_test): Move one test from
+ lround_test. Use TEST_f_L in moved test.
+ (lround_test): Move misplaced test to llround_test. Add testcase
+ from bug 2561.
+
+2012-02-28 Ulrich Drepper <drepper@xxxxxxxxx>
+
+ * sysdeps/x86_64/fpu/e_expf.S: New file.
+ Contributed by Dmitrieva Liubov <liubov.dmitrieva@xxxxxxxxx>.
+
2012-02-28 Stanislav Brabec <sbrabec@xxxxxxx>
[BZ #13637]
Modified: fsf/trunk/libc/NEWS
==============================================================================
--- fsf/trunk/libc/NEWS (original)
+++ fsf/trunk/libc/NEWS Thu Mar 1 00:01:42 2012
@@ -13,7 +13,7 @@
5993, 6884, 6907, 9739, 9902, 10110, 10140, 10210, 11174, 11322, 11365,
11494, 12047, 13058, 13525, 13526, 13527, 13528, 13529, 13530, 13531,
13532, 13533, 13547, 13551, 13552, 13553, 13555, 13559, 13583, 13618,
- 13637, 13695, 13704, 13738
+ 13637, 13695, 13704, 13706, 13738, 13786
* ISO C11 support:
Modified: fsf/trunk/libc/elf/Makefile
==============================================================================
--- fsf/trunk/libc/elf/Makefile (original)
+++ fsf/trunk/libc/elf/Makefile Thu Mar 1 00:01:42 2012
@@ -1203,3 +1203,14 @@
$(objpfx)tst-relsort1mod2.so: $(libm)
$(objpfx)tst-relsort1.out: $(objpfx)tst-relsort1mod1.so \
$(objpfx)tst-relsort1mod2.so
+
+tests: $(objpfx)tst-unused-dep.out
+
+$(objpfx)tst-unused-dep.out: $(objpfx)testobj1.so
+ LD_TRACE_LOADED_OBJECTS=1 \
+ LD_DEBUG=unused \
+ LD_PRELOAD= \
+ $(elf-objpfx)${rtld-installed-name} \
+ --library-path $(rpath-link)$(patsubst %,:%,$(sysdep-library-path)) \
+ $< > $@
+ cmp $@ /dev/null > /dev/null
Modified: fsf/trunk/libc/elf/rtld.c
==============================================================================
--- fsf/trunk/libc/elf/rtld.c (original)
+++ fsf/trunk/libc/elf/rtld.c Thu Mar 1 00:01:42 2012
@@ -1,5 +1,5 @@
/* Run time dynamic linker.
- Copyright (C) 1995-2010, 2011 Free Software Foundation, Inc.
+ Copyright (C) 1995-2012 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
@@ -1375,6 +1375,9 @@
_dl_setup_hash (l);
l->l_relocated = 1;
+ /* The vDSO is always used. */
+ l->l_used = 1;
+
/* Initialize l_local_scope to contain just this map. This allows
the use of dl_lookup_symbol_x to resolve symbols within the vdso.
So we create a single entry list pointing to l_real as its only
Modified: fsf/trunk/libc/math/libm-test.inc
==============================================================================
--- fsf/trunk/libc/math/libm-test.inc (original)
+++ fsf/trunk/libc/math/libm-test.inc Thu Mar 1 00:01:42 2012
@@ -2601,6 +2601,10 @@
TEST_f_f (expm1, 1, M_El - 1.0);
TEST_f_f (expm1, 0.75L, 1.11700001661267466854536981983709561L);
+
+#if defined TEST_LDOUBLE && LDBL_MAX_EXP >= 16384
+ TEST_f_f (expm1, 11356.25L, 9.05128237311923300051376115753226014206e+4931L);
+#endif
errno = 0;
TEST_f_f (expm1, 100000.0, plus_infty);
@@ -4395,7 +4399,7 @@
TEST_f_l (lround, 1073741824.01, 1073741824);
# if LONG_MAX > 281474976710656
TEST_f_l (lround, 281474976710656.025, 281474976710656);
- TEST_f_l (llround, -3.65309740835E17, -365309740835000000);
+ TEST_f_l (lround, 18014398509481974, 18014398509481974);
# endif
TEST_f_l (lround, 2097152.5, 2097153);
TEST_f_l (lround, -2097152.5, -2097153);
@@ -4438,6 +4442,7 @@
TEST_f_L (llround, -2097152.5, -2097153);
TEST_f_L (llround, 34359738368.5, 34359738369ll);
TEST_f_L (llround, -34359738368.5, -34359738369ll);
+ TEST_f_L (llround, -3.65309740835E17, -365309740835000000LL);
#endif
/* Test boundary conditions. */
Modified: fsf/trunk/libc/resolv/res_query.c
==============================================================================
--- fsf/trunk/libc/resolv/res_query.c (original)
+++ fsf/trunk/libc/resolv/res_query.c Thu Mar 1 00:01:42 2012
@@ -556,12 +556,16 @@
* copy without '.' if present.
*/
n = strlen(name);
- if (n >= MAXDNAME) {
+
+ /* Decrement N prior to checking it against MAXDNAME
+ so that we detect a wrap to SIZE_MAX and return
+ a reasonable error. */
+ n--;
+ if (n >= MAXDNAME - 1) {
RES_SET_H_ERRNO(statp, NO_RECOVERY);
return (-1);
}
- n--;
- if (n >= 0 && name[n] == '.') {
+ if (name[n] == '.') {
strncpy(nbuf, name, n);
nbuf[n] = '\0';
} else
Modified: fsf/trunk/libc/sysdeps/i386/fpu/libm-test-ulps
==============================================================================
--- fsf/trunk/libc/sysdeps/i386/fpu/libm-test-ulps (original)
+++ fsf/trunk/libc/sysdeps/i386/fpu/libm-test-ulps Thu Mar 1 00:01:42 2012
@@ -370,8 +370,8 @@
Test "Real part of: ctan (-2 - 3 i) == 0.376402564150424829275122113032269084e-2 - 1.00323862735360980144635859782192726 i":
double: 1
idouble: 1
-ildouble: 439
-ldouble: 439
+ildouble: 1
+ldouble: 1
Test "Imaginary part of: ctan (-2 - 3 i) == 0.376402564150424829275122113032269084e-2 - 1.00323862735360980144635859782192726 i":
float: 1
ifloat: 1
@@ -390,13 +390,13 @@
# ctanh
Test "Real part of: ctanh (-2 - 3 i) == -0.965385879022133124278480269394560686 + 0.988437503832249372031403430350121098e-2 i":
-ildouble: 5
-ldouble: 5
+ildouble: 3
+ldouble: 3
Test "Imaginary part of: ctanh (-2 - 3 i) == -0.965385879022133124278480269394560686 + 0.988437503832249372031403430350121098e-2 i":
float: 1
ifloat: 1
-ildouble: 25
-ldouble: 25
+ildouble: 1
+ldouble: 1
Test "Imaginary part of: ctanh (0 + pi/4 i) == 0.0 + 1.0 i":
float: 1
ifloat: 1
@@ -437,10 +437,6 @@
# exp
Test "exp (0.75) == 2.11700001661267466854536981983709561":
ildouble: 1
-Test "exp (1000.0) == 0.197007111401704699388887935224332313e435":
-ildouble: 754
-Test "exp (50.0) == 5184705528587072464087.45332293348538":
-ildouble: 16
# exp10
Test "exp10 (-1) == 0.1":
@@ -456,6 +452,8 @@
# expm1
Test "expm1 (1) == M_El - 1.0":
ildouble: 1
+Test "expm1 (11356.25) == 9.05128237311923300051376115753226014206e+4931":
+ldouble: 1
# gamma
Test "gamma (-0.5) == log(2*sqrt(pi))":
@@ -942,10 +940,6 @@
ifloat: 1
# Maximal error of functions:
-Function: "acos":
-ildouble: 622
-ldouble: 622
-
Function: "asin":
ildouble: 1
ldouble: 1
@@ -1096,8 +1090,8 @@
float: 4
idouble: 1
ifloat: 4
-ildouble: 763
-ldouble: 763
+ildouble: 6
+ldouble: 6
Function: Imaginary part of "cpow":
double: 2
@@ -1136,8 +1130,8 @@
Function: Real part of "ctan":
double: 1
idouble: 1
-ildouble: 439
-ldouble: 439
+ildouble: 1
+ldouble: 1
Function: Imaginary part of "ctan":
double: 1
@@ -1152,16 +1146,16 @@
float: 1
idouble: 1
ifloat: 1
-ildouble: 5
-ldouble: 5
+ildouble: 3
+ldouble: 3
Function: Imaginary part of "ctanh":
double: 1
float: 1
idouble: 1
ifloat: 1
-ildouble: 25
-ldouble: 25
+ildouble: 1
+ldouble: 1
Function: "erf":
double: 1
@@ -1174,9 +1168,6 @@
ifloat: 1
ildouble: 1
ldouble: 1
-
-Function: "exp":
-ildouble: 754
Function: "exp10":
ildouble: 8
@@ -1184,6 +1175,7 @@
Function: "expm1":
ildouble: 1
+ldouble: 1
Function: "gamma":
double: 1
Modified: fsf/trunk/libc/sysdeps/i386/i686/multiarch/strcasecmp_l-c.c
==============================================================================
--- fsf/trunk/libc/sysdeps/i386/i686/multiarch/strcasecmp_l-c.c (original)
+++ fsf/trunk/libc/sysdeps/i386/i686/multiarch/strcasecmp_l-c.c Thu Mar 1 00:01:42 2012
@@ -6,6 +6,8 @@
#define USE_IN_EXTENDED_LOCALE_MODEL 1
#include <string/strcasecmp.c>
+strong_alias (__strcasecmp_l_nonascii, __strcasecmp_l_ia32)
+
/* The needs of strcasecmp in libc are minimal, no need to go through
the IFUNC. */
strong_alias (__strcasecmp_l_nonascii, __GI___strcasecmp_l)
Modified: fsf/trunk/libc/sysdeps/i386/i686/multiarch/strcmp.S
==============================================================================
--- fsf/trunk/libc/sysdeps/i386/i686/multiarch/strcmp.S (original)
+++ fsf/trunk/libc/sysdeps/i386/i686/multiarch/strcmp.S Thu Mar 1 00:01:42 2012
@@ -111,6 +111,7 @@
# endif
#endif
-#ifndef USE_AS_STRNCMP
+#if !defined USE_AS_STRNCMP && !defined USE_AS_STRCASECMP_L \
+ && !defined USE_AS_STRNCASECMP_L
# include "../strcmp.S"
#endif
Modified: fsf/trunk/libc/sysdeps/i386/i686/multiarch/strncase_l-c.c
==============================================================================
--- fsf/trunk/libc/sysdeps/i386/i686/multiarch/strncase_l-c.c (original)
+++ fsf/trunk/libc/sysdeps/i386/i686/multiarch/strncase_l-c.c Thu Mar 1 00:01:42 2012
@@ -6,6 +6,8 @@
#define USE_IN_EXTENDED_LOCALE_MODEL 1
#include <string/strncase.c>
+strong_alias (__strncasecmp_l_nonascii, __strncasecmp_l_ia32)
+
/* The needs of strcasecmp in libc are minimal, no need to go through
the IFUNC. */
strong_alias (__strncasecmp_l_nonascii, __GI___strncasecmp_l)
Added: fsf/trunk/libc/sysdeps/sparc/sparc32/fpu/w_sqrt.S
==============================================================================
--- fsf/trunk/libc/sysdeps/sparc/sparc32/fpu/w_sqrt.S (added)
+++ fsf/trunk/libc/sysdeps/sparc/sparc32/fpu/w_sqrt.S Thu Mar 1 00:01:42 2012
@@ -1,0 +1,59 @@
+/* sqrt function. sparc32 version.
+ Copyright (C) 2012 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, see
+ <http://www.gnu.org/licenses/>. */
+
+#include <sysdep.h>
+
+ENTRY (__sqrt)
+ clr %g1
+ std %g0, [%sp + 72]
+ std %o0, [%sp + 80]
+ ldd [%sp + 72], %f8
+ ldd [%sp + 80], %f0
+ fcmpd %f0, %f8
+ fbl 1f
+ nop
+8: retl
+ fsqrtd %f0, %f0
+1:
+#ifdef SHARED
+ SETUP_PIC_REG_LEAF(o5, g1)
+#ifdef HAVE_BINUTILS_GOTDATA
+ sethi %gdop_hix22(_LIB_VERSION), %g1
+ xor %g1, %gdop_lox10(_LIB_VERSION), %g1
+ ld [%o5 + %g1], %g1, %gdop(_LIB_VERSION)
+#else
+ sethi %hi(_LIB_VERSION), %g1
+ or %g1, %lo(_LIB_VERSION), %g1
+ ld [%o5 + %g1], %g1
+#endif
+#else
+ sethi %hi(_LIB_VERSION), %g1
+ or %g1, %lo(_LIB_VERSION), %g1
+#endif
+ ld [%g1], %g1
+ cmp %g1, -1
+ be 8b
+ mov %o0, %o2
+ mov %o1, %o3
+ mov 26, %o4
+ mov %o7, %g1
+ call __kernel_standard
+ mov %g1, %o7
+END (__sqrt)
+
+weak_alias (__sqrt, sqrt)
Added: fsf/trunk/libc/sysdeps/sparc/sparc32/fpu/w_sqrtf.S
==============================================================================
--- fsf/trunk/libc/sysdeps/sparc/sparc32/fpu/w_sqrtf.S (added)
+++ fsf/trunk/libc/sysdeps/sparc/sparc32/fpu/w_sqrtf.S Thu Mar 1 00:01:42 2012
@@ -1,0 +1,57 @@
+/* sqrtf function. sparc32 version.
+ Copyright (C) 2012 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, see
+ <http://www.gnu.org/licenses/>. */
+
+#include <sysdep.h>
+
+ENTRY (__sqrtf)
+ st %g0, [%sp + 68]
+ st %o0, [%sp + 72]
+ ld [%sp + 68], %f8
+ ld [%sp + 72], %f0
+ fcmps %f0, %f8
+ fbl 1f
+ nop
+8: retl
+ fsqrts %f0, %f0
+1:
+#ifdef SHARED
+ SETUP_PIC_REG_LEAF(o5, g1)
+#ifdef HAVE_BINUTILS_GOTDATA
+ sethi %gdop_hix22(_LIB_VERSION), %g1
+ xor %g1, %gdop_lox10(_LIB_VERSION), %g1
+ ld [%o5 + %g1], %g1, %gdop(_LIB_VERSION)
+#else
+ sethi %hi(_LIB_VERSION), %g1
+ or %g1, %lo(_LIB_VERSION), %g1
+ ld [%o5 + %g1], %g1
+#endif
+#else
+ sethi %hi(_LIB_VERSION), %g1
+ or %g1, %lo(_LIB_VERSION), %g1
+#endif
+ ld [%g1], %g1
+ cmp %g1, -1
+ be 8b
+ mov %o0, %o1
+ mov 126, %o2
+ mov %o7, %g1
+ call __kernel_standard_f
+ mov %g1, %o7
+END (__sqrtf)
+
+weak_alias (__sqrtf, sqrtf)
Added: fsf/trunk/libc/sysdeps/sparc/sparc32/sparcv9/fpu/w_sqrt.S
==============================================================================
--- fsf/trunk/libc/sysdeps/sparc/sparc32/sparcv9/fpu/w_sqrt.S (added)
+++ fsf/trunk/libc/sysdeps/sparc/sparc32/sparcv9/fpu/w_sqrt.S Thu Mar 1 00:01:42 2012
@@ -1,0 +1,57 @@
+/* sqrt function. sparc32 v9 version.
+ Copyright (C) 2012 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, see
+ <http://www.gnu.org/licenses/>. */
+
+#include <sysdep.h>
+
+ENTRY (__sqrt)
+ std %o0, [%sp + 80]
+ fzero %f8
+ ldd [%sp + 80], %f0
+ fcmpd %f0, %f8
+ fbl 1f
+ nop
+8: retl
+ fsqrtd %f0, %f0
+1:
+#ifdef SHARED
+ SETUP_PIC_REG_LEAF(o5, g1)
+#ifdef HAVE_BINUTILS_GOTDATA
+ sethi %gdop_hix22(_LIB_VERSION), %g1
+ xor %g1, %gdop_lox10(_LIB_VERSION), %g1
+ ld [%o5 + %g1], %g1, %gdop(_LIB_VERSION)
+#else
+ sethi %hi(_LIB_VERSION), %g1
+ or %g1, %lo(_LIB_VERSION), %g1
+ ld [%o5 + %g1], %g1
+#endif
+#else
+ sethi %hi(_LIB_VERSION), %g1
+ or %g1, %lo(_LIB_VERSION), %g1
+#endif
+ ld [%g1], %g1
+ cmp %g1, -1
+ be 8b
+ mov %o0, %o2
+ mov %o1, %o3
+ mov 26, %o4
+ mov %o7, %g1
+ call __kernel_standard
+ mov %g1, %o7
+END (__sqrt)
+
+weak_alias (__sqrt, sqrt)
Added: fsf/trunk/libc/sysdeps/sparc/sparc32/sparcv9/fpu/w_sqrtf.S
==============================================================================
--- fsf/trunk/libc/sysdeps/sparc/sparc32/sparcv9/fpu/w_sqrtf.S (added)
+++ fsf/trunk/libc/sysdeps/sparc/sparc32/sparcv9/fpu/w_sqrtf.S Thu Mar 1 00:01:42 2012
@@ -1,0 +1,56 @@
+/* sqrtf function. sparc32 v9 version.
+ Copyright (C) 2012 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, see
+ <http://www.gnu.org/licenses/>. */
+
+#include <sysdep.h>
+
+ENTRY (__sqrtf)
+ st %o0, [%sp + 72]
+ fzeros %f8
+ ld [%sp + 72], %f0
+ fcmps %f0, %f8
+ fbl 1f
+ nop
+8: retl
+ fsqrts %f0, %f0
+1:
+#ifdef SHARED
+ SETUP_PIC_REG_LEAF(o5, g1)
+#ifdef HAVE_BINUTILS_GOTDATA
+ sethi %gdop_hix22(_LIB_VERSION), %g1
+ xor %g1, %gdop_lox10(_LIB_VERSION), %g1
+ ld [%o5 + %g1], %g1, %gdop(_LIB_VERSION)
+#else
+ sethi %hi(_LIB_VERSION), %g1
+ or %g1, %lo(_LIB_VERSION), %g1
+ ld [%o5 + %g1], %g1
+#endif
+#else
+ sethi %hi(_LIB_VERSION), %g1
+ or %g1, %lo(_LIB_VERSION), %g1
+#endif
+ ld [%g1], %g1
+ cmp %g1, -1
+ be 8b
+ mov %o0, %o1
+ mov 126, %o2
+ mov %o7, %g1
+ call __kernel_standard_f
+ mov %g1, %o7
+END (__sqrtf)
+
+weak_alias (__sqrtf, sqrtf)
Added: fsf/trunk/libc/sysdeps/sparc/sparc64/fpu/w_sqrt.S
==============================================================================
--- fsf/trunk/libc/sysdeps/sparc/sparc64/fpu/w_sqrt.S (added)
+++ fsf/trunk/libc/sysdeps/sparc/sparc64/fpu/w_sqrt.S Thu Mar 1 00:01:42 2012
@@ -1,0 +1,54 @@
+/* sqrt function. sparc64 version.
+ Copyright (C) 2012 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, see
+ <http://www.gnu.org/licenses/>. */
+
+#include <sysdep.h>
+
+ENTRY (__sqrt)
+ fzero %f8
+ fcmpd %fcc2, %f0, %f8
+ fbl,pn %fcc2, 1f
+ nop
+8: retl
+ fsqrtd %f0, %f0
+1:
+#ifdef SHARED
+ SETUP_PIC_REG_LEAF(o5, g1)
+#ifdef HAVE_BINUTILS_GOTDATA
+ sethi %gdop_hix22(_LIB_VERSION), %g1
+ xor %g1, %gdop_lox10(_LIB_VERSION), %g1
+ ldx [%o5 + %g1], %g1, %gdop(_LIB_VERSION)
+#else
+ sethi %hi(_LIB_VERSION), %g1
+ or %g1, %lo(_LIB_VERSION), %g1
+ ldx [%o5 + %g1], %g1
+#endif
+#else
+ sethi %hi(_LIB_VERSION), %g1
+ or %g1, %lo(_LIB_VERSION), %g1
+#endif
+ ld [%g1], %g1
+ cmp %g1, -1
+ be,pt %icc, 8b
+ fmovd %f0, %f2
+ mov 26, %o2
+ mov %o7, %g1
+ call __kernel_standard
+ mov %g1, %o7
+END (__sqrt)
+
+weak_alias (__sqrt, sqrt)
Added: fsf/trunk/libc/sysdeps/sparc/sparc64/fpu/w_sqrtf.S
==============================================================================
--- fsf/trunk/libc/sysdeps/sparc/sparc64/fpu/w_sqrtf.S (added)
+++ fsf/trunk/libc/sysdeps/sparc/sparc64/fpu/w_sqrtf.S Thu Mar 1 00:01:42 2012
@@ -1,0 +1,54 @@
+/* sqrtf function. sparc64 version.
+ Copyright (C) 2012 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, see
+ <http://www.gnu.org/licenses/>. */
+
+#include <sysdep.h>
+
+ENTRY (__sqrtf)
+ fzeros %f8
+ fcmps %fcc2, %f1, %f8
+ fbl,pn %fcc2, 1f
+ nop
+8: retl
+ fsqrts %f1, %f0
+1:
+#ifdef SHARED
+ SETUP_PIC_REG_LEAF(o5, g1)
+#ifdef HAVE_BINUTILS_GOTDATA
+ sethi %gdop_hix22(_LIB_VERSION), %g1
+ xor %g1, %gdop_lox10(_LIB_VERSION), %g1
+ ldx [%o5 + %g1], %g1, %gdop(_LIB_VERSION)
+#else
+ sethi %hi(_LIB_VERSION), %g1
+ or %g1, %lo(_LIB_VERSION), %g1
+ ldx [%o5 + %g1], %g1
+#endif
+#else
+ sethi %hi(_LIB_VERSION), %g1
+ or %g1, %lo(_LIB_VERSION), %g1
+#endif
+ ld [%g1], %g1
+ cmp %g1, -1
+ be,pt %icc, 8b
+ fmovs %f1, %f3
+ mov 126, %o2
+ mov %o7, %g1
+ call __kernel_standard_f
+ mov %g1, %o7
+END (__sqrtf)
+
+weak_alias (__sqrtf, sqrtf)
Added: fsf/trunk/libc/sysdeps/x86_64/fpu/e_expf.S
==============================================================================
--- fsf/trunk/libc/sysdeps/x86_64/fpu/e_expf.S (added)
+++ fsf/trunk/libc/sysdeps/x86_64/fpu/e_expf.S Thu Mar 1 00:01:42 2012
@@ -1,0 +1,340 @@
+/* Optimized __ieee754_expf function.
+ Copyright (C) 2012 Free Software Foundation, Inc.
+ Contributed by Intel Corporation.
+ 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. */
+
+#include <sysdep.h>
+
+/* Short algorithm description:
+ *
+ * Let K = 64 (table size).
+ * e^x = 2^(x/log(2)) = 2^n * T[j] * (1 + P(y))
+ * where
+ * x = m*log(2)/K + y, y in [0.0..log(2)/K]
+ * m = n*K + j, m,n,j - signed integer, j in [0..K-1]
+ * values of 2^(j/K) are tabulated as T[j].
+ *
+ * P(y) is a minimax polynomial approximation of expf(x)-1
+ * on small interval [0.0..log(2)/K].
+ *
+ * P(y) = P3*y*y*y*y + P2*y*y*y + P1*y*y + P0*y, calculated as
+ * z = y*y; P(y) = (P3*z + P1)*z + (P2*z + P0)*y
+ *
+ * Special cases:
+ * expf(NaN) = NaN
+ * expf(+INF) = +INF
+ * expf(-INF) = 0
+ * expf(x) = 1 for subnormals
+ * for finite argument, only expf(0)=1 is exact
+ * expf(x) overflows if x>88.7228317260742190
+ * expf(x) underflows if x<-103.972076416015620
+ */
+
+ .text
+ENTRY(__ieee754_expf)
+ /* Input: single precision x in %xmm0 */
+ cvtss2sd %xmm0, %xmm1 /* Convert x to double precision */
+ movd %xmm0, %ecx /* Copy x */
+ movsd L(DP_KLN2)(%rip), %xmm2 /* DP K/log(2) */
+ movsd L(DP_P2)(%rip), %xmm3 /* DP P2 */
+ movl %ecx, %eax /* x */
+ mulsd %xmm1, %xmm2 /* DP x*K/log(2) */
+ andl $0x7fffffff, %ecx /* |x| */
+ lea L(DP_T)(%rip), %rsi /* address of table T[j] */
+ cmpl $0x42ad496b, %ecx /* |x|<125*log(2) ? */
+ movsd L(DP_P3)(%rip), %xmm4 /* DP P3 */
+ addsd L(DP_RS)(%rip), %xmm2 /* DP x*K/log(2)+RS */
+ jae L(special_paths)
+
+ /* Here if |x|<125*log(2) */
+ cmpl $0x31800000, %ecx /* |x|<2^(-28) ? */
+ jb L(small_arg)
+
+ /* Main path: here if 2^(-28)<=|x|<125*log(2) */
+ cvtsd2ss %xmm2, %xmm2 /* SP x*K/log(2)+RS */
+ movd %xmm2, %eax /* bits of n*K+j with trash */
+ subss L(SP_RS)(%rip), %xmm2 /* SP t=round(x*K/log(2)) */
+ movl %eax, %edx /* n*K+j with trash */
+ cvtss2sd %xmm2, %xmm2 /* DP t */
+ andl $0x3f, %eax /* bits of j */
+ mulsd L(DP_NLN2K)(%rip), %xmm2/* DP -t*log(2)/K */
+ andl $0xffffffc0, %edx /* bits of n */
+#ifdef __AVX__
+ vaddsd %xmm1, %xmm2, %xmm0 /* DP y=x-t*log(2)/K */
+ vmulsd %xmm0, %xmm0, %xmm2 /* DP z=y*y */
+#else
+ addsd %xmm1, %xmm2 /* DP y=x-t*log(2)/K */
+ movaps %xmm2, %xmm0 /* DP y */
+ mulsd %xmm2, %xmm2 /* DP z=y*y */
+#endif
+ mulsd %xmm2, %xmm4 /* DP P3*z */
+ addl $0x1fc0, %edx /* bits of n + SP exponent bias */
+ mulsd %xmm2, %xmm3 /* DP P2*z */
+ shll $17, %edx /* SP 2^n */
+ addsd L(DP_P1)(%rip), %xmm4 /* DP P3*z+P1 */
+ addsd L(DP_P0)(%rip), %xmm3 /* DP P2*z+P0 */
+ movd %edx, %xmm1 /* SP 2^n */
+ mulsd %xmm2, %xmm4 /* DP (P3*z+P1)*z */
+ mulsd %xmm3, %xmm0 /* DP (P2*z+P0)*y */
+ addsd %xmm4, %xmm0 /* DP P(y) */
+ mulsd (%rsi,%rax,8), %xmm0 /* DP P(y)*T[j] */
+ addsd (%rsi,%rax,8), %xmm0 /* DP T[j]*(P(y)+1) */
+ cvtsd2ss %xmm0, %xmm0 /* SP T[j]*(P(y)+1) */
+ mulss %xmm1, %xmm0 /* SP result=2^n*(T[j]*(P(y)+1)) */
+ ret
+
+ .p2align 4
+L(small_arg):
+ /* Here if 0<=|x|<2^(-28) */
+ addss L(SP_ONE)(%rip), %xmm0 /* 1.0 + x */
+ /* Return 1.0 with inexact raised, except for x==0 */
+ ret
+
+ .p2align 4
+L(special_paths):
+ /* Here if 125*log(2)<=|x| */
+ shrl $31, %eax /* Get sign bit of x, and depending on it: */
+ lea L(SP_RANGE)(%rip), %rdx /* load over/underflow bound */
+ cmpl (%rdx,%rax,4), %ecx /* |x|<under/overflow bound ? */
+ jbe L(near_under_or_overflow)
+
+ /* Here if |x|>under/overflow bound */
+ cmpl $0x7f800000, %ecx /* |x| is finite ? */
+ jae L(arg_inf_or_nan)
+
+ /* Here if |x|>under/overflow bound, and x is finite */
+ testq %rax, %rax /* sign of x nonzero ? */
+ je L(res_overflow)
+
+ /* Here if -inf<x<underflow bound (x<0) */
+ movss L(SP_SMALL)(%rip), %xmm0/* load small value 2^(-100) */
+ mulss %xmm0, %xmm0 /* Return underflowed result (zero or subnormal) */
+ ret
+
+ .p2align 4
+L(res_overflow):
+ /* Here if overflow bound<x<inf (x>0) */
+ movss L(SP_LARGE)(%rip), %xmm0/* load large value 2^100 */
+ mulss %xmm0, %xmm0 /* Return overflowed result (Inf or max normal) */
+ ret
+
+ .p2align 4
+L(arg_inf_or_nan):
+ /* Here if |x| is Inf or NAN */
+ jne L(arg_nan) /* |x| is Inf ? */
+
+ /* Here if |x| is Inf */
+ lea L(SP_INF_0)(%rip), %rdx /* depending on sign of x: */
+ movss (%rdx,%rax,4), %xmm0 /* return zero or Inf */
+ ret
+
+ .p2align 4
+L(arg_nan):
+ /* Here if |x| is NaN */
+ addss %xmm0, %xmm0 /* Return x+x (raise invalid) */
+ ret
+
+ .p2align 4
+L(near_under_or_overflow):
+ /* Here if 125*log(2)<=|x|<under/overflow bound */
+ cvtsd2ss %xmm2, %xmm2 /* SP x*K/log(2)+RS */
+ movd %xmm2, %eax /* bits of n*K+j with trash */
+ subss L(SP_RS)(%rip), %xmm2 /* SP t=round(x*K/log(2)) */
+ movl %eax, %edx /* n*K+j with trash */
+ cvtss2sd %xmm2, %xmm2 /* DP t */
+ andl $0x3f, %eax /* bits of j */
+ mulsd L(DP_NLN2K)(%rip), %xmm2/* DP -t*log(2)/K */
+ andl $0xffffffc0, %edx /* bits of n */
+#ifdef __AVX__
+ vaddsd %xmm1, %xmm2, %xmm0 /* DP y=x-t*log(2)/K */
+ vmulsd %xmm0, %xmm0, %xmm2 /* DP z=y*y */
+#else
+ addsd %xmm1, %xmm2 /* DP y=x-t*log(2)/K */
+ movaps %xmm2, %xmm0 /* DP y */
+ mulsd %xmm2, %xmm2 /* DP z=y*y */
+#endif
+ mulsd %xmm2, %xmm4 /* DP P3*z */
+ addl $0xffc0, %edx /* bits of n + DP exponent bias */
+ mulsd %xmm2, %xmm3 /* DP P2*z */
+ shlq $46, %rdx /* DP 2^n */
+ addsd L(DP_P1)(%rip), %xmm4 /* DP P3*z+P1 */
+ addsd L(DP_P0)(%rip), %xmm3 /* DP P2*z+P0 */
+ movd %rdx, %xmm1 /* DP 2^n */
+ mulsd %xmm2, %xmm4 /* DP (P3*z+P1)*z */
+ mulsd %xmm3, %xmm0 /* DP (P2*z+P0)*y */
+ addsd %xmm4, %xmm0 /* DP P(y) */
+ mulsd (%rsi,%rax,8), %xmm0 /* DP P(y)*T[j] */
+ addsd (%rsi,%rax,8), %xmm0 /* DP T[j]*(P(y)+1) */
+ mulsd %xmm1, %xmm0 /* DP result=2^n*(T[j]*(P(y)+1)) */
+ cvtsd2ss %xmm0, %xmm0 /* convert result to single precision */
+ ret
+END(__ieee754_expf)
+
+ .section .rodata, "a"
+ .p2align 3
+L(DP_T): /* table of double precision values 2^(j/K) for j=[0..K-1] */
+ .long 0x00000000, 0x3ff00000
+ .long 0x3e778061, 0x3ff02c9a
+ .long 0xd3158574, 0x3ff059b0
+ .long 0x18759bc8, 0x3ff08745
+ .long 0x6cf9890f, 0x3ff0b558
+ .long 0x32d3d1a2, 0x3ff0e3ec
+ .long 0xd0125b51, 0x3ff11301
+ .long 0xaea92de0, 0x3ff1429a
+ .long 0x3c7d517b, 0x3ff172b8
+ .long 0xeb6fcb75, 0x3ff1a35b
+ .long 0x3168b9aa, 0x3ff1d487
+ .long 0x88628cd6, 0x3ff2063b
+ .long 0x6e756238, 0x3ff2387a
+ .long 0x65e27cdd, 0x3ff26b45
+ .long 0xf51fdee1, 0x3ff29e9d
+ .long 0xa6e4030b, 0x3ff2d285
+ .long 0x0a31b715, 0x3ff306fe
+ .long 0xb26416ff, 0x3ff33c08
+ .long 0x373aa9cb, 0x3ff371a7
+ .long 0x34e59ff7, 0x3ff3a7db
+ .long 0x4c123422, 0x3ff3dea6
+ .long 0x21f72e2a, 0x3ff4160a
+ .long 0x6061892d, 0x3ff44e08
+ .long 0xb5c13cd0, 0x3ff486a2
+ .long 0xd5362a27, 0x3ff4bfda
+ .long 0x769d2ca7, 0x3ff4f9b2
+ .long 0x569d4f82, 0x3ff5342b
+ .long 0x36b527da, 0x3ff56f47
+ .long 0xdd485429, 0x3ff5ab07
+ .long 0x15ad2148, 0x3ff5e76f
+ .long 0xb03a5585, 0x3ff6247e
+ .long 0x82552225, 0x3ff66238
+ .long 0x667f3bcd, 0x3ff6a09e
+ .long 0x3c651a2f, 0x3ff6dfb2
+ .long 0xe8ec5f74, 0x3ff71f75
+ .long 0x564267c9, 0x3ff75feb
+ .long 0x73eb0187, 0x3ff7a114
+ .long 0x36cf4e62, 0x3ff7e2f3
+ .long 0x994cce13, 0x3ff82589
+ .long 0x9b4492ed, 0x3ff868d9
+ .long 0x422aa0db, 0x3ff8ace5
+ .long 0x99157736, 0x3ff8f1ae
+ .long 0xb0cdc5e5, 0x3ff93737
+ .long 0x9fde4e50, 0x3ff97d82
+ .long 0x82a3f090, 0x3ff9c491
+ .long 0x7b5de565, 0x3ffa0c66
+ .long 0xb23e255d, 0x3ffa5503
+ .long 0x5579fdbf, 0x3ffa9e6b
+ .long 0x995ad3ad, 0x3ffae89f
+ .long 0xb84f15fb, 0x3ffb33a2
+ .long 0xf2fb5e47, 0x3ffb7f76
+ .long 0x904bc1d2, 0x3ffbcc1e
+ .long 0xdd85529c, 0x3ffc199b
+ .long 0x2e57d14b, 0x3ffc67f1
+ .long 0xdcef9069, 0x3ffcb720
+ .long 0x4a07897c, 0x3ffd072d
+ .long 0xdcfba487, 0x3ffd5818
+ .long 0x03db3285, 0x3ffda9e6
+ .long 0x337b9b5f, 0x3ffdfc97
+ .long 0xe78b3ff6, 0x3ffe502e
+ .long 0xa2a490da, 0x3ffea4af
+ .long 0xee615a27, 0x3ffefa1b
+ .long 0x5b6e4540, 0x3fff5076
+ .long 0x819e90d8, 0x3fffa7c1
+ ASM_TYPE_DIRECTIVE(L(DP_T), @object)
+ ASM_SIZE_DIRECTIVE(L(DP_T))
+
+ .section .rodata.cst8,"aM",@progbits,8
+ .p2align 3
+L(DP_KLN2): /* double precision K/log(2) */
+ .long 0x652b82fe, 0x40571547
+ ASM_TYPE_DIRECTIVE(L(DP_KLN2), @object)
+ ASM_SIZE_DIRECTIVE(L(DP_KLN2))
+
+ .p2align 3
+L(DP_NLN2K): /* double precision -log(2)/K */
+ .long 0xfefa39ef, 0xbf862e42
+ ASM_TYPE_DIRECTIVE(L(DP_NLN2K), @object)
+ ASM_SIZE_DIRECTIVE(L(DP_NLN2K))
+
+ .p2align 3
+L(DP_RS): /* double precision 2^23+2^22 */
+ .long 0x00000000, 0x41680000
+ ASM_TYPE_DIRECTIVE(L(DP_RS), @object)
+ ASM_SIZE_DIRECTIVE(L(DP_RS))
+
+ .p2align 3
+L(DP_P3): /* double precision polynomial coefficient P3 */
+ .long 0xeb78fa85, 0x3fa56420
+ ASM_TYPE_DIRECTIVE(L(DP_P3), @object)
+ ASM_SIZE_DIRECTIVE(L(DP_P3))
+
+ .p2align 3
+L(DP_P1): /* double precision polynomial coefficient P1 */
+ .long 0x008d6118, 0x3fe00000
+ ASM_TYPE_DIRECTIVE(L(DP_P1), @object)
+ ASM_SIZE_DIRECTIVE(L(DP_P1))
+
+ .p2align 3
+L(DP_P2): /* double precision polynomial coefficient P2 */
+ .long 0xda752d4f, 0x3fc55550
+ ASM_TYPE_DIRECTIVE(L(DP_P2), @object)
+ ASM_SIZE_DIRECTIVE(L(DP_P2))
+
+ .p2align 3
+L(DP_P0): /* double precision polynomial coefficient P0 */
+ .long 0xffffe7c6, 0x3fefffff
+ ASM_TYPE_DIRECTIVE(L(DP_P0), @object)
+ ASM_SIZE_DIRECTIVE(L(DP_P0))
+
+ .p2align 2
+L(SP_RANGE): /* single precision overflow/underflow bounds */
+ .long 0x42b17217 /* if x>this bound, then result overflows */
+ .long 0x42cff1b4 /* if x<this bound, then result underflows */
+ ASM_TYPE_DIRECTIVE(L(SP_RANGE), @object)
+ ASM_SIZE_DIRECTIVE(L(SP_RANGE))
+
+ .p2align 2
+L(SP_INF_0):
+ .long 0x7f800000 /* single precision Inf */
+ .long 0 /* single precision zero */
+ ASM_TYPE_DIRECTIVE(L(SP_INF_0), @object)
+ ASM_SIZE_DIRECTIVE(L(SP_INF_0))
+
+ .section .rodata.cst4,"aM",@progbits,4
+ .p2align 2
+L(SP_RS): /* single precision 2^23+2^22 */
+ .long 0x4b400000
+ ASM_TYPE_DIRECTIVE(L(SP_RS), @object)
+ ASM_SIZE_DIRECTIVE(L(SP_RS))
+
+ .p2align 2
+L(SP_SMALL): /* single precision small value 2^(-100) */
+ .long 0x0d800000
+ ASM_TYPE_DIRECTIVE(L(SP_SMALL), @object)
+ ASM_SIZE_DIRECTIVE(L(SP_SMALL))
+
+ .p2align 2
+L(SP_LARGE): /* single precision large value 2^100 */
+ .long 0x71800000
+ ASM_TYPE_DIRECTIVE(L(SP_LARGE), @object)
+ ASM_SIZE_DIRECTIVE(L(SP_LARGE))
+
+ .p2align 2
+L(SP_ONE): /* single precision 1.0 */
+ .long 0x3f800000
+ ASM_TYPE_DIRECTIVE(L(SP_ONE), @object)
+ ASM_SIZE_DIRECTIVE(L(SP_ONE))
+
+strong_alias (__ieee754_expf, __expf_finite)
Modified: fsf/trunk/libc/sysdeps/x86_64/fpu/libm-test-ulps
==============================================================================
--- fsf/trunk/libc/sysdeps/x86_64/fpu/libm-test-ulps (original)
+++ fsf/trunk/libc/sysdeps/x86_64/fpu/libm-test-ulps Thu Mar 1 00:01:42 2012
@@ -420,8 +420,8 @@
Test "Real part of: ctan (-2 - 3 i) == 0.376402564150424829275122113032269084e-2 - 1.00323862735360980144635859782192726 i":
double: 1
idouble: 1
-ildouble: 439
-ldouble: 439
+ildouble: 1
+ldouble: 1
Test "Imaginary part of: ctan (-2 - 3 i) == 0.376402564150424829275122113032269084e-2 - 1.00323862735360980144635859782192726 i":
float: 1
ifloat: 1
@@ -444,11 +444,11 @@
float: 2
idouble: 1
ifloat: 2
-ildouble: 5
-ldouble: 5
+ildouble: 3
+ldouble: 3
Test "Imaginary part of: ctanh (-2 - 3 i) == -0.965385879022133124278480269394560686 + 0.988437503832249372031403430350121098e-2 i":
-ildouble: 25
-ldouble: 25
+ildouble: 1
+ldouble: 1
Test "Imaginary part of: ctanh (0 + pi/4 i) == 0.0 + 1.0 i":
float: 1
ifloat: 1
@@ -511,6 +511,9 @@
float: 1
idouble: 1
ifloat: 1
+Test "expm1 (11356.25) == 9.05128237311923300051376115753226014206e+4931":
+ildouble: 1
+ldouble: 1
# gamma
Test "gamma (-0.5) == log(2*sqrt(pi))":
@@ -1215,8 +1218,8 @@
Function: Real part of "ctan":
double: 1
idouble: 1
-ildouble: 439
-ldouble: 439
+ildouble: 1
+ldouble: 1
Function: Imaginary part of "ctan":
double: 1
@@ -1231,16 +1234,16 @@
float: 2
idouble: 1
ifloat: 2
-ildouble: 5
-ldouble: 5
+ildouble: 3
+ldouble: 3
Function: Imaginary part of "ctanh":
double: 1
float: 1
idouble: 1
ifloat: 1
-ildouble: 25
-ldouble: 25
+ildouble: 1
+ldouble: 1
Function: "erf":
double: 1
@@ -1265,6 +1268,8 @@
float: 1
idouble: 1
ifloat: 1
+ildouble: 1
+ldouble: 1
Function: "gamma":
ildouble: 1
_______________________________________________
Commits mailing list
Commits@xxxxxxxxxx
http://eglibc.org/cgi-bin/mailman/listinfo/commits