This is the mail archive of the
glibc-cvs@sourceware.org
mailing list for the glibc project.
GNU C Library master sources branch master updated. glibc-2.22-358-g06d97e5
- From: jsm28 at sourceware dot org
- To: glibc-cvs at sourceware dot org
- Date: 9 Oct 2015 21:02:51 -0000
- Subject: GNU C Library master sources branch master updated. glibc-2.22-358-g06d97e5
This is an automated email from the git hooks/post-receive script. It was
generated because a ref change was pushed to the repository containing
the project "GNU C Library master sources".
The branch, master has been updated
via 06d97e5e6100641df2c379459e41b26bb4d7648b (commit)
from facdd9ea29ab94aac2b188ec3cc41f8733d769e0 (commit)
Those revisions listed above that are new to this repository have
not appeared on any other notification email; so we list those
revisions in full, below.
- Log -----------------------------------------------------------------
http://sourceware.org/git/gitweb.cgi?p=glibc.git;a=commitdiff;h=06d97e5e6100641df2c379459e41b26bb4d7648b
commit 06d97e5e6100641df2c379459e41b26bb4d7648b
Author: Joseph Myers <joseph@codesourcery.com>
Date: Fri Oct 9 21:02:19 2015 +0000
Fix lrint, llrint, lround, llround missing exceptions for MIPS (bug 16399).
For 32-bit MIPS and some other systems, various of the lrint, llrint,
lround, llround functions can be missing exceptions on overflow
because casts do not (in current GCC) result in the proper
exceptions. In the MIPS case there are two problems here: MIPS I code
generation uses an assembler macro that doesn't raise exceptions,
while the libgcc conversions of floating-point values to long long
also do not raise "invalid" on all overflow cases (and can raise
spurious "inexact").
This patch adds support in the generic code (only the functions for
which this problem has actually been seen) for forcing the "invalid"
exception in the problem cases, and enables that support for the
affected MIPS cases.
Tested for MIPS; also tested for x86_64 and x86 that installed
stripped shared libraries are unchanged by this patch.
[BZ #16399]
* sysdeps/generic/fix-fp-int-convert-overflow.h: New file.
* sysdeps/ieee754/dbl-64/s_llrint.c: Include <fenv.h>, <limits.h>
and <fix-fp-int-convert-overflow.h>.
(__llrint) [FE_INVALID]: Force FE_INVALID exception as needed if
FIX_DBL_LLONG_CONVERT_OVERFLOW.
* sysdeps/ieee754/dbl-64/s_llround.c: Include <fenv.h>, <limits.h>
and <fix-fp-int-convert-overflow.h>.
(__llround) [FE_INVALID]: Force FE_INVALID exception as needed if
FIX_DBL_LLONG_CONVERT_OVERFLOW.
* sysdeps/ieee754/dbl-64/s_lrint.c: Include
<fix-fp-int-convert-overflow.h>.
(__lrint) [FE_INVALID]: Force FE_INVALID exception as needed if
FIX_DBL_LLONG_CONVERT_OVERFLOW.
* sysdeps/ieee754/dbl-64/s_lround.c: Include
<fix-fp-int-convert-overflow.h>.
(__lround) [FE_INVALID]: Force FE_INVALID exception as needed if
FIX_DBL_LLONG_CONVERT_OVERFLOW.
* sysdeps/ieee754/flt-32/s_llrintf.c: Include <fenv.h>, <limits.h>
and <fix-fp-int-convert-overflow.h>.
(__llrintf) [FE_INVALID]: Force FE_INVALID exception as needed if
FIX_DBL_LLONG_CONVERT_OVERFLOW.
* sysdeps/ieee754/flt-32/s_llroundf.c: Include <fenv.h>,
<limits.h> and <fix-fp-int-convert-overflow.h>.
(__llroundf) [FE_INVALID]: Force FE_INVALID exception as needed if
FIX_DBL_LLONG_CONVERT_OVERFLOW.
* sysdeps/ieee754/flt-32/s_lrintf.c: Include <fenv.h>, <limits.h>
and <fix-fp-int-convert-overflow.h>.
(__lrintf) [FE_INVALID]: Force FE_INVALID exception as needed if
FIX_DBL_LLONG_CONVERT_OVERFLOW.
* sysdeps/ieee754/flt-32/s_lroundf.c: Include <fenv.h>, <limits.h>
and <fix-fp-int-convert-overflow.h>.
(__lroundf) [FE_INVALID]: Force FE_INVALID exception as needed if
FIX_DBL_LLONG_CONVERT_OVERFLOW.
* sysdeps/mips/mips32/fpu/fix-fp-int-convert-overflow.h: New file.
diff --git a/ChangeLog b/ChangeLog
index c47fd6d..1e9a477 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,3 +1,41 @@
+2015-10-09 Joseph Myers <joseph@codesourcery.com>
+
+ [BZ #16399]
+ * sysdeps/generic/fix-fp-int-convert-overflow.h: New file.
+ * sysdeps/ieee754/dbl-64/s_llrint.c: Include <fenv.h>, <limits.h>
+ and <fix-fp-int-convert-overflow.h>.
+ (__llrint) [FE_INVALID]: Force FE_INVALID exception as needed if
+ FIX_DBL_LLONG_CONVERT_OVERFLOW.
+ * sysdeps/ieee754/dbl-64/s_llround.c: Include <fenv.h>, <limits.h>
+ and <fix-fp-int-convert-overflow.h>.
+ (__llround) [FE_INVALID]: Force FE_INVALID exception as needed if
+ FIX_DBL_LLONG_CONVERT_OVERFLOW.
+ * sysdeps/ieee754/dbl-64/s_lrint.c: Include
+ <fix-fp-int-convert-overflow.h>.
+ (__lrint) [FE_INVALID]: Force FE_INVALID exception as needed if
+ FIX_DBL_LLONG_CONVERT_OVERFLOW.
+ * sysdeps/ieee754/dbl-64/s_lround.c: Include
+ <fix-fp-int-convert-overflow.h>.
+ (__lround) [FE_INVALID]: Force FE_INVALID exception as needed if
+ FIX_DBL_LLONG_CONVERT_OVERFLOW.
+ * sysdeps/ieee754/flt-32/s_llrintf.c: Include <fenv.h>, <limits.h>
+ and <fix-fp-int-convert-overflow.h>.
+ (__llrintf) [FE_INVALID]: Force FE_INVALID exception as needed if
+ FIX_DBL_LLONG_CONVERT_OVERFLOW.
+ * sysdeps/ieee754/flt-32/s_llroundf.c: Include <fenv.h>,
+ <limits.h> and <fix-fp-int-convert-overflow.h>.
+ (__llroundf) [FE_INVALID]: Force FE_INVALID exception as needed if
+ FIX_DBL_LLONG_CONVERT_OVERFLOW.
+ * sysdeps/ieee754/flt-32/s_lrintf.c: Include <fenv.h>, <limits.h>
+ and <fix-fp-int-convert-overflow.h>.
+ (__lrintf) [FE_INVALID]: Force FE_INVALID exception as needed if
+ FIX_DBL_LLONG_CONVERT_OVERFLOW.
+ * sysdeps/ieee754/flt-32/s_lroundf.c: Include <fenv.h>, <limits.h>
+ and <fix-fp-int-convert-overflow.h>.
+ (__lroundf) [FE_INVALID]: Force FE_INVALID exception as needed if
+ FIX_DBL_LLONG_CONVERT_OVERFLOW.
+ * sysdeps/mips/mips32/fpu/fix-fp-int-convert-overflow.h: New file.
+
2015-10-09 Carlos O'Donell <carlos@redhat.com>
[BZ #18589]
diff --git a/NEWS b/NEWS
index 5c239f3..32e4c64 100644
--- a/NEWS
+++ b/NEWS
@@ -10,16 +10,16 @@ Version 2.23
* The following bugs are resolved with this release:
887, 2542, 2543, 2558, 2898, 4404, 6803, 10432, 14341, 14912, 15367,
- 15384, 15786, 15918, 16141, 16296, 16347, 16415, 16517, 16519, 16520,
- 16521, 16620, 16734, 16973, 16985, 17118, 17243, 17244, 17250, 17441,
- 17787, 17886, 17887, 17905, 18084, 18086, 18240, 18265, 18370, 18421,
- 18480, 18525, 18595, 18589, 18610, 18618, 18647, 18661, 18674, 18675,
- 18681, 18724, 18757, 18778, 18781, 18787, 18789, 18790, 18795, 18796,
- 18803, 18820, 18823, 18824, 18825, 18857, 18863, 18870, 18872, 18873,
- 18875, 18887, 18921, 18951, 18952, 18956, 18961, 18966, 18967, 18969,
- 18970, 18977, 18980, 18981, 18985, 19003, 19012, 19016, 19018, 19032,
- 19046, 19049, 19050, 19059, 19071, 19076, 19077, 19078, 19079, 19085,
- 19086, 19088, 19094, 19095.
+ 15384, 15786, 15918, 16141, 16296, 16347, 16399, 16415, 16517, 16519,
+ 16520, 16521, 16620, 16734, 16973, 16985, 17118, 17243, 17244, 17250,
+ 17441, 17787, 17886, 17887, 17905, 18084, 18086, 18240, 18265, 18370,
+ 18421, 18480, 18525, 18595, 18589, 18610, 18618, 18647, 18661, 18674,
+ 18675, 18681, 18724, 18757, 18778, 18781, 18787, 18789, 18790, 18795,
+ 18796, 18803, 18820, 18823, 18824, 18825, 18857, 18863, 18870, 18872,
+ 18873, 18875, 18887, 18921, 18951, 18952, 18956, 18961, 18966, 18967,
+ 18969, 18970, 18977, 18980, 18981, 18985, 19003, 19012, 19016, 19018,
+ 19032, 19046, 19049, 19050, 19059, 19071, 19076, 19077, 19078, 19079,
+ 19085, 19086, 19088, 19094, 19095.
* The obsolete header <regexp.h> has been removed. Programs that require
this header must be updated to use <regex.h> instead.
diff --git a/sysdeps/generic/fix-fp-int-convert-overflow.h b/sysdeps/generic/fix-fp-int-convert-overflow.h
new file mode 100644
index 0000000..d29211b
--- /dev/null
+++ b/sysdeps/generic/fix-fp-int-convert-overflow.h
@@ -0,0 +1,31 @@
+/* Fix for conversion of floating point to integer overflow. Generic version.
+ Copyright (C) 2015 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/>. */
+
+#ifndef FIX_FP_INT_CONVERT_OVERFLOW_H
+#define FIX_FP_INT_CONVERT_OVERFLOW_H 1
+
+/* Define these macros to 1 to workaround conversions of out-of-range
+ floating-point numbers to integer types failing to raise the
+ "invalid" exception, or raising spurious "inexact" or other
+ exceptions. */
+#define FIX_FLT_LONG_CONVERT_OVERFLOW 0
+#define FIX_FLT_LLONG_CONVERT_OVERFLOW 0
+#define FIX_DBL_LONG_CONVERT_OVERFLOW 0
+#define FIX_DBL_LLONG_CONVERT_OVERFLOW 0
+
+#endif /* fix-fp-int-convert-overflow.h */
diff --git a/sysdeps/ieee754/dbl-64/s_llrint.c b/sysdeps/ieee754/dbl-64/s_llrint.c
index 048331e..1673527 100644
--- a/sysdeps/ieee754/dbl-64/s_llrint.c
+++ b/sysdeps/ieee754/dbl-64/s_llrint.c
@@ -18,9 +18,12 @@
License along with the GNU C Library; if not, see
<http://www.gnu.org/licenses/>. */
+#include <fenv.h>
+#include <limits.h>
#include <math.h>
#include <math_private.h>
+#include <fix-fp-int-convert-overflow.h>
static const double two52[2] =
{
@@ -77,8 +80,16 @@ __llrint (double x)
}
else
{
- /* The number is too large. It is left implementation defined
- what happens. */
+#ifdef FE_INVALID
+ /* The number is too large. Unless it rounds to LLONG_MIN,
+ FE_INVALID must be raised and the return value is
+ unspecified. */
+ if (FIX_DBL_LLONG_CONVERT_OVERFLOW && x != (double) LLONG_MIN)
+ {
+ feraiseexcept (FE_INVALID);
+ return sx == 0 ? LLONG_MAX : LLONG_MIN;
+ }
+#endif
return (long long int) x;
}
diff --git a/sysdeps/ieee754/dbl-64/s_llround.c b/sysdeps/ieee754/dbl-64/s_llround.c
index c8b3c19..77ad3c3 100644
--- a/sysdeps/ieee754/dbl-64/s_llround.c
+++ b/sysdeps/ieee754/dbl-64/s_llround.c
@@ -17,9 +17,12 @@
License along with the GNU C Library; if not, see
<http://www.gnu.org/licenses/>. */
+#include <fenv.h>
+#include <limits.h>
#include <math.h>
#include <math_private.h>
+#include <fix-fp-int-convert-overflow.h>
long long int
@@ -65,8 +68,16 @@ __llround (double x)
}
else
{
- /* The number is too large. It is left implementation defined
- what happens. */
+#ifdef FE_INVALID
+ /* The number is too large. Unless it rounds to LLONG_MIN,
+ FE_INVALID must be raised and the return value is
+ unspecified. */
+ if (FIX_DBL_LLONG_CONVERT_OVERFLOW && x != (double) LLONG_MIN)
+ {
+ feraiseexcept (FE_INVALID);
+ return sign == 1 ? LLONG_MAX : LLONG_MIN;
+ }
+#endif
return (long long int) x;
}
diff --git a/sysdeps/ieee754/dbl-64/s_lrint.c b/sysdeps/ieee754/dbl-64/s_lrint.c
index d004594..3fdb05e 100644
--- a/sysdeps/ieee754/dbl-64/s_lrint.c
+++ b/sysdeps/ieee754/dbl-64/s_lrint.c
@@ -23,6 +23,7 @@
#include <math.h>
#include <math_private.h>
+#include <fix-fp-int-convert-overflow.h>
static const double two52[2] =
{
@@ -107,6 +108,11 @@ __lrint (double x)
feraiseexcept (t == LONG_MIN ? FE_INEXACT : FE_INVALID);
return LONG_MIN;
}
+ else if (FIX_DBL_LONG_CONVERT_OVERFLOW && x != (double) LONG_MIN)
+ {
+ feraiseexcept (FE_INVALID);
+ return sx == 0 ? LONG_MAX : LONG_MIN;
+ }
#endif
return (long int) x;
}
diff --git a/sysdeps/ieee754/dbl-64/s_lround.c b/sysdeps/ieee754/dbl-64/s_lround.c
index 91b17b0..699eafb 100644
--- a/sysdeps/ieee754/dbl-64/s_lround.c
+++ b/sysdeps/ieee754/dbl-64/s_lround.c
@@ -22,6 +22,7 @@
#include <math.h>
#include <math_private.h>
+#include <fix-fp-int-convert-overflow.h>
long int
@@ -80,8 +81,18 @@ __lround (double x)
FE_INVALID must be raised and the return value is
unspecified. */
#ifdef FE_INVALID
- if (sizeof (long int) == 4
- && x <= (double) LONG_MIN - 0.5)
+ if (FIX_DBL_LONG_CONVERT_OVERFLOW
+ && !(sign == -1
+ && (sizeof (long int) == 4
+ ? x > (double) LONG_MIN - 0.5
+ : x >= (double) LONG_MIN)))
+ {
+ feraiseexcept (FE_INVALID);
+ return sign == 1 ? LONG_MAX : LONG_MIN;
+ }
+ else if (!FIX_DBL_LONG_CONVERT_OVERFLOW
+ && sizeof (long int) == 4
+ && x <= (double) LONG_MIN - 0.5)
{
/* If truncation produces LONG_MIN, the cast will not raise
the exception, but may raise "inexact". */
diff --git a/sysdeps/ieee754/flt-32/s_llrintf.c b/sysdeps/ieee754/flt-32/s_llrintf.c
index d7588d0..debac8d 100644
--- a/sysdeps/ieee754/flt-32/s_llrintf.c
+++ b/sysdeps/ieee754/flt-32/s_llrintf.c
@@ -18,9 +18,12 @@
License along with the GNU C Library; if not, see
<http://www.gnu.org/licenses/>. */
+#include <fenv.h>
+#include <limits.h>
#include <math.h>
#include <math_private.h>
+#include <fix-fp-int-convert-overflow.h>
static const float two23[2] =
{
@@ -64,8 +67,16 @@ __llrintf (float x)
}
else
{
- /* The number is too large. It is left implementation defined
- what happens. */
+#ifdef FE_INVALID
+ /* The number is too large. Unless it rounds to LLONG_MIN,
+ FE_INVALID must be raised and the return value is
+ unspecified. */
+ if (FIX_FLT_LLONG_CONVERT_OVERFLOW && x != (float) LLONG_MIN)
+ {
+ feraiseexcept (FE_INVALID);
+ return sx == 0 ? LLONG_MAX : LLONG_MIN;
+ }
+#endif
return (long long int) x;
}
diff --git a/sysdeps/ieee754/flt-32/s_llroundf.c b/sysdeps/ieee754/flt-32/s_llroundf.c
index f695c63..0404231 100644
--- a/sysdeps/ieee754/flt-32/s_llroundf.c
+++ b/sysdeps/ieee754/flt-32/s_llroundf.c
@@ -17,9 +17,12 @@
License along with the GNU C Library; if not, see
<http://www.gnu.org/licenses/>. */
+#include <fenv.h>
+#include <limits.h>
#include <math.h>
#include <math_private.h>
+#include <fix-fp-int-convert-overflow.h>
long long int
@@ -51,8 +54,16 @@ __llroundf (float x)
}
else
{
- /* The number is too large. It is left implementation defined
- what happens. */
+#ifdef FE_INVALID
+ /* The number is too large. Unless it rounds to LLONG_MIN,
+ FE_INVALID must be raised and the return value is
+ unspecified. */
+ if (FIX_FLT_LLONG_CONVERT_OVERFLOW && x != (float) LLONG_MIN)
+ {
+ feraiseexcept (FE_INVALID);
+ return sign == 1 ? LLONG_MAX : LLONG_MIN;
+ }
+#endif
return (long long int) x;
}
diff --git a/sysdeps/ieee754/flt-32/s_lrintf.c b/sysdeps/ieee754/flt-32/s_lrintf.c
index bc60fd2..2eb242d 100644
--- a/sysdeps/ieee754/flt-32/s_lrintf.c
+++ b/sysdeps/ieee754/flt-32/s_lrintf.c
@@ -18,9 +18,12 @@
License along with the GNU C Library; if not, see
<http://www.gnu.org/licenses/>. */
+#include <fenv.h>
+#include <limits.h>
#include <math.h>
#include <math_private.h>
+#include <fix-fp-int-convert-overflow.h>
static const float two23[2] =
{
@@ -64,8 +67,16 @@ __lrintf (float x)
}
else
{
- /* The number is too large. It is left implementation defined
- what happens. */
+#ifdef FE_INVALID
+ /* The number is too large. Unless it rounds to LONG_MIN,
+ FE_INVALID must be raised and the return value is
+ unspecified. */
+ if (FIX_FLT_LONG_CONVERT_OVERFLOW && x != (float) LONG_MIN)
+ {
+ feraiseexcept (FE_INVALID);
+ return sx == 0 ? LONG_MAX : LONG_MIN;
+ }
+#endif
return (long int) x;
}
diff --git a/sysdeps/ieee754/flt-32/s_lroundf.c b/sysdeps/ieee754/flt-32/s_lroundf.c
index f20352a..da3551c 100644
--- a/sysdeps/ieee754/flt-32/s_lroundf.c
+++ b/sysdeps/ieee754/flt-32/s_lroundf.c
@@ -17,9 +17,12 @@
License along with the GNU C Library; if not, see
<http://www.gnu.org/licenses/>. */
+#include <fenv.h>
+#include <limits.h>
#include <math.h>
#include <math_private.h>
+#include <fix-fp-int-convert-overflow.h>
long int
@@ -51,8 +54,16 @@ __lroundf (float x)
}
else
{
- /* The number is too large. It is left implementation defined
- what happens. */
+#ifdef FE_INVALID
+ /* The number is too large. Unless it rounds to LONG_MIN,
+ FE_INVALID must be raised and the return value is
+ unspecified. */
+ if (FIX_FLT_LONG_CONVERT_OVERFLOW && x != (float) LONG_MIN)
+ {
+ feraiseexcept (FE_INVALID);
+ return sign == 1 ? LONG_MAX : LONG_MIN;
+ }
+#endif
return (long int) x;
}
diff --git a/sysdeps/mips/mips32/fpu/fix-fp-int-convert-overflow.h b/sysdeps/mips/mips32/fpu/fix-fp-int-convert-overflow.h
new file mode 100644
index 0000000..ee14f7e
--- /dev/null
+++ b/sysdeps/mips/mips32/fpu/fix-fp-int-convert-overflow.h
@@ -0,0 +1,36 @@
+/* Fix for conversion of floating point to integer overflow. MIPS version.
+ Copyright (C) 2015 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/>. */
+
+#ifndef FIX_FP_INT_CONVERT_OVERFLOW_H
+#define FIX_FP_INT_CONVERT_OVERFLOW_H 1
+
+/* As of GCC 5, the generic libgcc2.c conversions from floating point
+ to long long may not raise the correct exceptions on overflow (and
+ may raise spurious "inexact" exceptions even in non-overflow cases,
+ see <https://gcc.gnu.org/bugzilla/show_bug.cgi?id=59412>). */
+#define FIX_FLT_LLONG_CONVERT_OVERFLOW 1
+#define FIX_DBL_LLONG_CONVERT_OVERFLOW 1
+
+/* As of GCC 5 and binutils 2.25, for MIPS I GCC generates calls to
+ assembler macros for conversions from floating point to integer
+ types, and those macros save and restore the whole of FSCR and so
+ lose exceptions. */
+#define FIX_FLT_LONG_CONVERT_OVERFLOW (__mips == 1)
+#define FIX_DBL_LONG_CONVERT_OVERFLOW (__mips == 1)
+
+#endif /* fix-fp-int-convert-overflow.h */
-----------------------------------------------------------------------
Summary of changes:
ChangeLog | 38 ++++++++++++++++++++
NEWS | 20 +++++-----
.../generic/fix-fp-int-convert-overflow.h | 21 ++++++-----
sysdeps/ieee754/dbl-64/s_llrint.c | 15 +++++++-
sysdeps/ieee754/dbl-64/s_llround.c | 15 +++++++-
sysdeps/ieee754/dbl-64/s_lrint.c | 6 +++
sysdeps/ieee754/dbl-64/s_lround.c | 15 +++++++-
sysdeps/ieee754/flt-32/s_llrintf.c | 15 +++++++-
sysdeps/ieee754/flt-32/s_llroundf.c | 15 +++++++-
sysdeps/ieee754/flt-32/s_lrintf.c | 15 +++++++-
sysdeps/ieee754/flt-32/s_lroundf.c | 15 +++++++-
.../mips/mips32/fpu/fix-fp-int-convert-overflow.h | 36 ++++++++++++++++++
12 files changed, 193 insertions(+), 33 deletions(-)
copy benchtests/bench-util.h => sysdeps/generic/fix-fp-int-convert-overflow.h (57%)
create mode 100644 sysdeps/mips/mips32/fpu/fix-fp-int-convert-overflow.h
hooks/post-receive
--
GNU C Library master sources