Roland McGrath [Thu, 12 Feb 2015 21:38:47 +0000 (13:38 -0800)]
Support after-link variable to run a final step on binaries.
RĂ¼diger Sonderfeld [Thu, 12 Feb 2015 19:51:31 +0000 (14:51 -0500)]
Document tv_sec is of type time_t:
The tv_sec is of type time_t in both struct timeval and struct timespec.
This matches the implementation and also the relevant standard (checked
C11 for timespec and opengroup for timeval).
Joseph Myers [Thu, 12 Feb 2015 19:02:45 +0000 (19:02 +0000)]
Fix exp2 spurious underflows (bug 16560).
This patch fixes the remaining part of bug 16560, spurious underflows
from exp2 of arguments close to 0 (when the result is close to 1, so
should not underflow), by just using 1+x instead of a more complicated
calculation when the argument is sufficiently small.
Tested for x86_64, x86 and mips64.
[BZ #16560]
* math/e_exp2l.c [LDBL_MANT_DIG == 106] (LDBL_EPSILON): Undefine
and redefine.
(__ieee754_exp2l): Do not multiply small fractional parts by
M_LN2l.
* sysdeps/i386/fpu/e_exp2l.S (__ieee754_exp2l): Just add 1 to
small argument.
* sysdeps/ieee754/dbl-64/e_exp2.c (__ieee754_exp2): Likewise.
* sysdeps/ieee754/flt-32/e_exp2f.c (__ieee754_exp2f): Likewise.
* sysdeps/x86_64/fpu/e_exp2l.S (__ieee754_exp2l): Likewise.
* math/auto-libm-test-in: Add more tests of exp2.
* math/auto-libm-test-out: Regenerated.
Rajalakshmi Srinivasaraghavan [Wed, 28 Jan 2015 13:43:29 +0000 (08:43 -0500)]
powerpc: POWER7 strncpy optimization for unaligned string
This patch optimizes strncpy for power7 for unaligned source or
destination address. The source or destination address is aligned
to doubleword and data is shifted based on the alignment and
added with the previous loaded data to be written as a doubleword.
For each load, cmpb instruction is used for faster null check.
The new optimization shows 10 to 70% of performance improvement
for longer string though it does not show big difference on string
size less than 16 due to additional checks.Hence this new algorithm
is restricted to string greater than 16.
Adhemerval Zanella [Thu, 12 Feb 2015 11:34:16 +0000 (05:34 -0600)]
powerpc: Fix TABORT encoding for little endian
This patch fix the TABORT encoding for toolchains with no support for
HTM builtins.
Andreas Schwab [Mon, 11 Aug 2014 09:18:26 +0000 (11:18 +0200)]
Filter out PTHREAD_MUTEX_NO_ELISION_NP bit in pthread_mutexattr_gettype (BZ #15790)
pthread_mutexattr_settype adds PTHREAD_MUTEX_NO_ELISION_NP to kind,
which is an internal flag that pthread_mutexattr_gettype shouldn't
expose, since pthread_mutexattr_settype wouldn't accept it.
Samuel Thibault [Wed, 11 Feb 2015 23:27:50 +0000 (23:27 +0000)]
Revert "hurd: Fix abi-tag, following ba90e05"
and revert the corresponding part of ba90e05 which was making the fix
necessary.
* abi-tags: Revert ae20c9a: rename back gnu into gnu-gnu.
* configure.ac, configure: Revert ba90e05: modify gnu-* host_os back
into gnu-gnu, and update comment to refer to abi-tags.
Joseph Myers [Wed, 11 Feb 2015 23:17:25 +0000 (23:17 +0000)]
Fix sincos errno setting (bug 15467).
This patch makes sincos set errno to EDOM when passed an infinity,
similarly to sin and cos.
Tested for x86_64, x86, powerpc and mips64. I don't know if the
architecture-specific implementations for ia64 and m68k might need
corresponding fixes.
2015-02-11 Joseph Myers <joseph@codesourcery.com>
[BZ #15467]
* sysdeps/ieee754/dbl-64/s_sincos.c: Include <errno.h>.
(__sincos): Set errno to EDOM for infinite argument.
* sysdeps/ieee754/flt-32/s_sincosf.c: Include <errno.h>.
(SINCOSF_FUNC): Set errno to EDOM for infinite argument.
* sysdeps/ieee754/ldbl-128/s_sincosl.c: Include <errno.h>.
(__sincosl): Set errno to EDOM for infinite argument.
* sysdeps/ieee754/ldbl-128ibm/s_sincosl.c: Include <errno.h>.
(__sincosl): Set errno to EDOM for infinite argument.
* sysdeps/ieee754/ldbl-96/s_sincosl.c: Include <errno.h>.
(__sincosl): Set errno to EDOM for infinite argument.
* math/libm-test.inc (sincos_test_data): Test errno setting.
Roland McGrath [Wed, 11 Feb 2015 22:39:30 +0000 (14:39 -0800)]
Another pointless update in README.
Roland McGrath [Wed, 11 Feb 2015 21:42:15 +0000 (13:42 -0800)]
Pointless update in README.
Leonhard Holz [Wed, 11 Feb 2015 11:49:58 +0000 (12:49 +0100)]
Remove unused definitions
Joseph Myers [Tue, 10 Feb 2015 22:01:39 +0000 (22:01 +0000)]
Remove sysdeps/mips soft-fp subdirectories.
As noted in
<https://sourceware.org/ml/libc-alpha/2014-10/msg00369.html>, soft-fp
sysdeps subdirectories (and more generally, subdirectories where
sysdeps/foo/Implies contains foo/bar) are unnecessary and should be
eliminated. This patch does so for MIPS.
Tested for MIPS64 (all three ABIs, soft-float) that installed stripped
shared libraries are unchanged by this patch.
* sysdeps/mips/soft-fp/sfp-machine.h: Move to ....
* sysdeps/mips/mips32/sfp-machine.h: ... here.
* sysdeps/mips/mips64/soft-fp/Makefile: Move to ....
* sysdeps/mips/mips64/Makefile: ... here.
* sysdeps/mips/mips64/soft-fp/e_sqrtl.c: Move to ....
* sysdeps/mips/mips64/e_sqrtl.c: ... here.
* sysdeps/mips/mips64/soft-fp/sfp-machine.h: Move to ....
* sysdeps/mips/mips64/sfp-machine.h: ... here.
* sysdeps/mips/mips32/Implies: Remove mips/soft-fp.
* sysdeps/mips/mips64/n32/Implies: Remove mips/mips64/soft-fp.
* sysdeps/mips/mips64/n64/Implies: Likewise.
Roland McGrath [Tue, 10 Feb 2015 21:55:29 +0000 (13:55 -0800)]
Clean up math/test-snan.
Evangelos Foutras [Tue, 10 Feb 2015 03:22:58 +0000 (03:22 +0000)]
Fix __memcpy_chk on non-SSE2 CPUs
In commit 8b4416d, the 1: jump label in __mempcpy_chk was accidentally
moved. This resulted in failures of mempcpy on CPU without SSE2.
Adhemerval Zanella [Tue, 10 Feb 2015 11:09:22 +0000 (06:09 -0500)]
powerpc: drop R_PPC_REL16 check
Current minimum support binutils (2.22) supports R_PPC_REL16 as default,
so no need to extra configure checks.
Adhemerval Zanella [Tue, 3 Feb 2015 11:48:24 +0000 (06:48 -0500)]
powerpc: sysdeps/powerpc configure cleanup
Current minimum binutils supported (2.22) has ".machine altivec" support
as default, so there is no need to add a configure check for such
functionality. This patches removes the configure checks for it.
Samuel Thibault [Mon, 9 Feb 2015 16:25:42 +0000 (17:25 +0100)]
Add fixed bug numbers to NEWS
Adhemerval Zanella [Wed, 21 Jan 2015 12:57:14 +0000 (07:57 -0500)]
powerpc: wordcopy/memmove cleanup for ppc32
This patch cleanup some multiarch code related to memmmove
optimization. Initial IFUNC support added specialized wordcopy
symbols which turned in local IFUNC calls used by memmove default
implementation. The patch removes the internal IFUNC for wordcopy
symbols and uses local branches in the memmmove optimization instead.
Adhemerval Zanella [Wed, 21 Jan 2015 12:38:28 +0000 (07:38 -0500)]
powerpc: multiarch Makefile cleanup for powerpc32
This patch cleanups the multiarch Makefile by putting the wide chars
implementation to correct wcsmbs rule.
Adhemerval Zanella [Wed, 21 Jan 2015 12:41:46 +0000 (07:41 -0500)]
powerpc: wordcopy/memmove cleanup for ppc64
This patch cleanup some multiarch code related to memmmove
optimization. Initial IFUNC support added specialized wordcopy
symbols which turned in local IFUNC calls used by memmove default
implementation.
This change by removing then and used the optimized memmove instead
for supported chips.
Adhemerval Zanella [Wed, 21 Jan 2015 03:08:21 +0000 (22:08 -0500)]
powerpc: Remove POWER7 wordcopy ifunc
This patch remove the POWER7 ifunc wordcopy function
(_wordcopy_*_power7), since now GLIBC provides a optimized memmove/bcopy
for POWER7.
Adhemerval Zanella [Wed, 21 Jan 2015 03:02:03 +0000 (22:02 -0500)]
powerpc: Simplify bcopy default implementation
This patch simplify the default bcopy symbol for powerpc64 by just using
memmove instead of implementing using the default bcopy. Since the
symbol is deprecated, it trades speed by code size.
Adhemerval Zanella [Tue, 20 Jan 2015 20:41:38 +0000 (15:41 -0500)]
powerpc: multiarch Makefile cleanup for powerpc64
This patch cleanups the multiarch Makefile by putting the wide chars
implementation to correct wcsmbs rule.
Andreas Schwab [Sun, 1 Feb 2015 13:04:15 +0000 (14:04 +0100)]
Fix value of O_TMPFILE for architectures with non-default O_DIRECTORY (bug 17912)
Samuel Thibault [Sun, 8 Feb 2015 17:46:00 +0000 (18:46 +0100)]
hurd: fix unwind-resume.c build
Samuel Thibault [Sun, 8 Feb 2015 17:35:54 +0000 (18:35 +0100)]
hurd: fix unwind-resume.c build
Samuel Thibault [Sun, 8 Feb 2015 16:49:16 +0000 (17:49 +0100)]
add hurd/hurdsocket.h file missing from a5eb23d
Samuel Thibault [Sun, 8 Feb 2015 16:45:25 +0000 (17:45 +0100)]
Fix time/getdate.c build.
Samuel Thibault [Sun, 8 Feb 2015 04:38:03 +0000 (05:38 +0100)]
hurd: Fix abi-tag, following ba90e05
Samuel Thibault [Sun, 8 Feb 2015 04:06:17 +0000 (05:06 +0100)]
hurd: fix tls.h build
Samuel Thibault [Sun, 8 Feb 2015 03:25:12 +0000 (04:25 +0100)]
hurd: Ignore bytes beyond sockaddr length for AF_UNIX
Paul Eggert [Thu, 5 Feb 2015 23:52:20 +0000 (15:52 -0800)]
Add ersatz _Static_assert on older C hosts
* misc/sys/cdefs.h (_Static_assert): Define a substitute, if on a
pre-C11 C platform that is not known to support _Static_assert.
Samuel Thibault [Sat, 7 Feb 2015 21:39:38 +0000 (22:39 +0100)]
hurd: fix compilation of signal.h in C++
Samuel Thibault [Sat, 7 Feb 2015 21:35:35 +0000 (22:35 +0100)]
hurd: fix compilation of signal.h in C++
Samuel Thibault [Sat, 7 Feb 2015 21:26:09 +0000 (22:26 +0100)]
hurd: add basic types for ioctls
Samuel Thibault [Sat, 7 Feb 2015 21:22:40 +0000 (22:22 +0100)]
hurd: support mmap with PROT_NONE
Samuel Thibault [Sat, 7 Feb 2015 21:18:03 +0000 (22:18 +0100)]
hurdselect: remove dead code.
This removes code which actually never happens, and is already taken
care of in the function.
This is in the second part of select, when the __mach_msg() function
over the portset has returned something else than MACH_MSG_SUCCESS. I
guess in the past the value returned by __mach_msg() was stored in err,
so this code was necessary to set back err to 0, but now it is stored in
msgerr, so err is already still 0 by default. It can thus never contain
MACH_RCV_TIMED_OUT, i.e. the code is dead. The first case mentioned in
the comment is already handled: on time out with no message, err is
already still the default 0. On time out due to poll, err would still be
0, unless some of the io_select RPCs has returned EINTR, in which case
it contains EINTR. If any other io_select RPCs had returned a proper
answer, got!=0, and thus err is set to 0 just below. The code is thus
indeed not useful any more.
Samuel Thibault [Sat, 7 Feb 2015 21:16:19 +0000 (22:16 +0100)]
hurd: fix sigstate locking
It looks like _hurd_thread_sigstate used to return with the sigstate
lock held long ago, but since that's no longer the case, don't unlock
something that isn't locked.
Note that it's unlikely this change fixes anything in practice since
its current implementation (on i386) makes this call a nop.
Samuel Thibault [Sat, 7 Feb 2015 21:13:46 +0000 (22:13 +0100)]
hurdselect: Let select get interrupted by signals
We need to set MACH_RCV_INTERRUPT to avoid __mach_msg just looping on signals,
and then we can handle the interruption.
Samuel Thibault [Sat, 7 Feb 2015 21:09:09 +0000 (22:09 +0100)]
hurd: map nice levels 1-to-1 with Mach prio levels
Samuel Thibault [Sat, 7 Feb 2015 21:07:38 +0000 (22:07 +0100)]
hurd: allow poll() array bigger than FD_SETSIZE
Samuel Thibault [Sat, 7 Feb 2015 20:56:15 +0000 (21:56 +0100)]
hurd: fix f?chflags prototypes, declare them and their flags
Samuel Thibault [Sat, 7 Feb 2015 20:45:59 +0000 (21:45 +0100)]
hurd: fix build with pthread aio
Roland McGrath [Fri, 6 Feb 2015 22:42:18 +0000 (14:42 -0800)]
ARM: Add missing sfi_breg in LDR_GLOBAL macro.
Roland McGrath [Fri, 6 Feb 2015 20:31:11 +0000 (12:31 -0800)]
NPTL: Conditionalize some sanity tests for SIGCANCEL/SIGSETXID.
Roland McGrath [Fri, 6 Feb 2015 20:30:58 +0000 (12:30 -0800)]
Conditionalize use of SIGRTMIN in nptl/tst-locale1.c.
Roland McGrath [Fri, 6 Feb 2015 20:30:38 +0000 (12:30 -0800)]
Don't set unused field in rt/tst-timer2.
Roland McGrath [Fri, 6 Feb 2015 20:30:23 +0000 (12:30 -0800)]
NPTL: Build tests using clone directly only for Linux.
Roland McGrath [Fri, 6 Feb 2015 20:30:05 +0000 (12:30 -0800)]
Use signal rather than sigaction in nptl/tst-cleanup2.
Roland McGrath [Fri, 6 Feb 2015 20:29:50 +0000 (12:29 -0800)]
Conditionalize some tests' use of SA_SIGINFO.
Roland McGrath [Fri, 6 Feb 2015 20:29:39 +0000 (12:29 -0800)]
Do not use SA_NOCLDWAIT in tst-pselect.
Roland McGrath [Fri, 6 Feb 2015 20:29:27 +0000 (12:29 -0800)]
Move tst-getlogin to login/ subdirectory.
Roland McGrath [Fri, 6 Feb 2015 20:24:16 +0000 (12:24 -0800)]
Fix libio/tst-atime not to presume ST_NOATIME exists.
Roland McGrath [Fri, 6 Feb 2015 20:23:35 +0000 (12:23 -0800)]
Fix dirent/tst-fdopendir not to presume O_NOATIME exists.
Roland McGrath [Fri, 6 Feb 2015 20:22:23 +0000 (12:22 -0800)]
Fix nptl/tst-kill5 not to presume SIGRTMAX exists.
Roland McGrath [Fri, 6 Feb 2015 20:21:38 +0000 (12:21 -0800)]
Clean up nptl/tst-join5 use of nanosleep.
Roland McGrath [Fri, 6 Feb 2015 18:53:38 +0000 (10:53 -0800)]
Clean up NPTL system to be compat-only.
Roland McGrath [Fri, 6 Feb 2015 18:53:27 +0000 (10:53 -0800)]
Clean up NPTL longjmp to be compat-only.
Roland McGrath [Fri, 6 Feb 2015 18:53:07 +0000 (10:53 -0800)]
Clean up NPTL fork to be compat-only.
Roland McGrath [Fri, 6 Feb 2015 18:44:11 +0000 (10:44 -0800)]
x86: Clean up __vdso_clock_gettime variable.
Roland McGrath [Fri, 6 Feb 2015 18:43:19 +0000 (10:43 -0800)]
Exclude rpcent functions and NSS backends for rpc, key when excluding sunrpc.
Roland McGrath [Fri, 6 Feb 2015 18:42:08 +0000 (10:42 -0800)]
Clean up sysdep-dl-routines variable.
Carlos O'Donell [Fri, 6 Feb 2015 17:59:37 +0000 (12:59 -0500)]
NEWS: Fix spelling.
Florian Weimer [Fri, 6 Feb 2015 15:28:24 +0000 (16:28 +0100)]
NEWS: Also mention CVE-2015-1473
Joseph Myers [Fri, 6 Feb 2015 15:44:07 +0000 (15:44 +0000)]
soft-fp: Fix _FP_FMA when product is zero and third argument is finite (bug 17932).
soft-fp's _FP_FMA fails to set the result's exponent for cases where
the result of the multiplication is 0, yielding incorrect (arbitrary,
depending on uninitialized values) results for those cases. This
affects libm for architectures using soft-fp to implement fma. This
patch adds the exponent setting and tests for this case.
Tested for ARM soft-float (which uses soft-fp fma), x86_64 and x86 (to
verify not introducing new libm test failures there).
(This bug showed up in testing my patch to move the Linux kernel to
current soft-fp. math/Makefile has "override CFLAGS +=
-Wno-uninitialized" which would have stopped compiler warnings from
showing up this problem, although I wouldn't be surprised if removing
that shows spurious warnings from this code, if the compiler fails to
follow that various cases where the exponent is uninitialized don't
need it initialized because the class is set to a value meaning the
uninitialized exponent isn't used.)
[BZ #17932]
* soft-fp/op-common.h (_FP_FMA): Set exponent of result in case
where multiplication results in zero and third argument is finite
and nonzero.
* math/auto-libm-test-in: Add more tests of fma.
* math/auto-libm-test-out: Regenerated.
Joseph Myers [Fri, 6 Feb 2015 15:41:49 +0000 (15:41 +0000)]
soft-fp: Refine FP_EX_DENORM handling for comparisons.
In <https://sourceware.org/ml/libc-alpha/2014-09/msg00488.html>, I
noted that comparisons in soft-fp did not set FP_EX_DENORM unless
denormal operands were flushed to zero.
This patch fixes soft-fp to check for denormal operands for
comparisons and set that exception whenever FP_EX_DENORM is not zero.
In particular, for the one architecture for which the Linux kernel
defines FP_EX_DENORM (alpha), this corresponds to the existing logic
for comparisons and so allows that logic to be replaced by a simple
call to FP_CMP_D when soft-fp is updated in the kernel.
Tested for powerpc (e500) that installed stripped shared libraries are
unchanged by this patch.
* soft-fp/op-common.h (_FP_CMP_CHECK_DENORM): New macro.
(_FP_CMP_CHECK_FLUSH_ZERO): Likewise.
(_FP_CMP): Use_FP_CMP_CHECK_DENORM and _FP_CMP_CHECK_FLUSH_ZERO.
(_FP_CMP_EQ): Likewise.
(_FP_CMP_UNORD): Use _FP_CMP_CHECK_DENORM.
Joseph Myers [Fri, 6 Feb 2015 15:39:17 +0000 (15:39 +0000)]
soft-fp: Support floating-point extensions without quieting sNaNs.
One special case needed in soft-fp to replace the old version in the
Linux kernel is extending from a narrower floating-point format to a
wider one without quieting signaling NaNs. (This is for
arch/powerpc/math-emu/lfs.c, where previously it used the old FP_CONV
which didn't do anything special for NaNs, then handled packing
specially for NaNs to avoid quieting at packing time, and discarded
the exceptions from unpacking.)
This patch accordingly refactors FP_EXTEND, creating a separate
_FP_EXTEND_CNAN that offers a choice of how NaNs are handled, with
FP_EXTEND reimplemented as a wrapper that provides the common case of
the IEEE operation that does quiet signaling NaNs and raise exceptions
for them.
Tested for powerpc (e500) that installed stripped shared libraries are
unchanged by this patch.
* soft-fp/op-common.h (FP_EXTEND): Rename to _FP_EXTEND_CNAN with
extra argument CHECK_NAN. Redefine as wrapper around
_FP_EXTEND_CNAN.
Carlos O'Donell [Fri, 6 Feb 2015 07:04:37 +0000 (02:04 -0500)]
Fix missing ChangeLog attribution.
Carlos O'Donell [Fri, 6 Feb 2015 07:03:41 +0000 (02:03 -0500)]
Open development for 2.22.
Carlos O'Donell [Fri, 6 Feb 2015 06:19:42 +0000 (01:19 -0500)]
Update version.h and include/features.h for 2.21 release
Carlos O'Donell [Fri, 6 Feb 2015 06:56:35 +0000 (01:56 -0500)]
hppa: Sync with pthread.h.
This reverts part of the previous commit to refactor pthread.h.
The refactoring must be done by having pthread.h include arch
bits headers, not the other way around. Then hppa provides the
arch bits header. For now we synchronzie again with pthread.h
and include the entire contents in the hppa copy.
Paul Pluzhnikov [Fri, 6 Feb 2015 05:30:42 +0000 (00:30 -0500)]
CVE-2015-1472: wscanf allocates too little memory
BZ #16618
Under certain conditions wscanf can allocate too little memory for the
to-be-scanned arguments and overflow the allocated buffer. The
implementation now correctly computes the required buffer size when
using malloc.
A regression test was added to tst-sscanf.
Carlos O'Donell [Fri, 6 Feb 2015 03:56:00 +0000 (22:56 -0500)]
glibc 2.21 pre-release update.
Update all translations.
Update contributions in the manual.
Update installation notes with information about newest working tools.
Reconfigure using exactly autoconf 2.69.
Regenerate INSTALL.
Carlos O'Donell [Fri, 6 Feb 2015 02:33:03 +0000 (21:33 -0500)]
hppa: Remove warnings and fix conformance errors.
(1) Fix warnings.
This is a bulk update to fix all the warnings that were causing
build failures with -Werror on hppa.
The most egregious problems are in dl-fptr.c which needs to be
entirely rewritten, thus I've used -Wno-error for that.
(2) Fix conformance errors.
The sysdep.c file had __syscall_error and syscall in one file
which caused conformance issues by including syscall when
__syscall_error was linked to. The fix is obviously to split
the file and use syscall.c to implement syscall.
Chung-Lin Tang [Thu, 5 Feb 2015 14:19:36 +0000 (06:19 -0800)]
Function name typo error in non-PIC case, fixed in this patch.
David S. Miller [Sun, 1 Feb 2015 03:07:28 +0000 (19:07 -0800)]
Fix two bugs in sparc atomics.
* sysdeps/sparc/sparc32/bits/atomic.h
(__sparc32_atomic_do_unlock24): Put the memory barrier before the
unlock not after it.
(__v9_compare_and_exchange_val_32_acq): Use unions to avoid getting
volatile register usage warnings from the compiler.
David S. Miller [Sun, 1 Feb 2015 02:59:40 +0000 (18:59 -0800)]
Fix sparc semaphore implementation after recent changes.
* sysdeps/sparc/nptl/sem_init.c: Delete.
* sysdeps/sparc/nptl/sem_post.c: Delete.
* sysdeps/sparc/nptl/sem_timedwait.c: Delete.
* sysdeps/sparc/nptl/sem_wait.c: Delete.
* sysdeps/sparc/sparc32/sem_init.c: New file.
* sysdeps/sparc/sparc32/sem_waitcommon.c: New file.
* sysdeps/sparc/sparc32/sem_open.c: Generic nptl version with
padding explicitly initialized.
* sysdeps/sparc/sparc32/sem_post.c: Generic nptl version using
padding for in-semaphore spinlock.
* sysdeps/sparc/sparc32/sem_wait.c: Likewise.
* sysdeps/sparc/sparc32/sem_trywait.c: Delete.
* sysdeps/sparc/sparc32/sem_timedwait.c: Delete.
* sysdeps/sparc/sparc32/sparcv9/sem_init.c: New file.
* sysdeps/sparc/sparc32/sparcv9/sem_open.c: New file.
* sysdeps/sparc/sparc32/sparcv9/sem_post.c: New file.
* sysdeps/sparc/sparc32/sparcv9/sem_waitcommon.c: New file.
* sysdeps/sparc/sparc32/sparcv9/sem_wait.c: Redirect to nptl
version.
* sysdeps/sparc/sparc32/sparcv9/sem_timedwait.c: Delete.
* sysdeps/sparc/sparc32/sparcv9/sem_trywait.c: Delete.
H.J. Lu [Fri, 30 Jan 2015 14:50:20 +0000 (06:50 -0800)]
Use AVX unaligned memcpy only if AVX2 is available
memcpy with unaligned 256-bit AVX register loads/stores are slow on older
processorsl like Sandy Bridge. This patch adds bit_AVX_Fast_Unaligned_Load
and sets it only when AVX2 is available.
[BZ #17801]
* sysdeps/x86_64/multiarch/init-arch.c (__init_cpu_features):
Set the bit_AVX_Fast_Unaligned_Load bit for AVX2.
* sysdeps/x86_64/multiarch/init-arch.h (bit_AVX_Fast_Unaligned_Load):
New.
(index_AVX_Fast_Unaligned_Load): Likewise.
(HAS_AVX_FAST_UNALIGNED_LOAD): Likewise.
* sysdeps/x86_64/multiarch/memcpy.S (__new_memcpy): Check the
bit_AVX_Fast_Unaligned_Load bit instead of the bit_AVX_Usable bit.
* sysdeps/x86_64/multiarch/memcpy_chk.S (__memcpy_chk): Likewise.
* sysdeps/x86_64/multiarch/mempcpy.S (__mempcpy): Likewise.
* sysdeps/x86_64/multiarch/mempcpy_chk.S (__mempcpy_chk): Likewise.
* sysdeps/x86_64/multiarch/memmove.c (__libc_memmove): Replace
HAS_AVX with HAS_AVX_FAST_UNALIGNED_LOAD.
* sysdeps/x86_64/multiarch/memmove_chk.c (__memmove_chk): Likewise.
Andreas Schwab [Sun, 25 Jan 2015 10:05:53 +0000 (11:05 +0100)]
Include <signal.h> in sysdeps/nptl/allocrtsig.c
Architectures which don't use hp-timing-common.h don't include <signal.h>
via <sys/param.h>.
Siddhesh Poyarekar [Thu, 29 Jan 2015 05:01:10 +0000 (10:31 +0530)]
Fix up ChangeLog formatting
Siddhesh Poyarekar [Thu, 29 Jan 2015 05:00:09 +0000 (10:30 +0530)]
Initialize nscd stats data [BZ #17892]
The padding bytes in the statsdata struct are not initialized, due to
which valgrind throws a warning:
==11384== Memcheck, a memory error detector
==11384== Copyright (C) 2002-2012, and GNU GPL'd, by Julian Seward et al.
==11384== Using Valgrind-3.8.1 and LibVEX; rerun with -h for copyright info
==11384== Command: nscd -d
==11384==
Fri 25 Apr 2014 10:34:53 AM CEST - 11384: handle_request: request received (Version = 2) from PID 11396
Fri 25 Apr 2014 10:34:53 AM CEST - 11384: GETSTAT
==11384== Thread 6:
==11384== Syscall param socketcall.sendto(msg) points to uninitialised byte(s)
==11384== at 0x4E4ACDC: send (in /lib64/libpthread-2.12.so)
==11384== by 0x11AF6B: send_stats (in /usr/sbin/nscd)
==11384== by 0x112F75: nscd_run_worker (in /usr/sbin/nscd)
==11384== by 0x4E439D0: start_thread (in /lib64/libpthread-2.12.so)
==11384== by 0x599AB6C: clone (in /lib64/libc-2.12.so)
==11384== Address 0x15708395 is on thread 6's stack
Fix the warning by initializing the structure.
Martin Sebor [Thu, 29 Jan 2015 04:07:01 +0000 (21:07 -0700)]
Clarify math/README.libm-test. Add "How to read the test output."
Chris Metcalf [Wed, 28 Jan 2015 19:51:21 +0000 (14:51 -0500)]
tilegx32: set __HAVE_64B_ATOMICS to 0
This is because of alignment issues in the sem_t support.
tilegx32 does in fact support 64-bit atomics and we will need
to revisit this after the 2.21 freeze.
Joseph Myers [Wed, 28 Jan 2015 18:40:35 +0000 (18:40 +0000)]
Disable 64-bit atomics for MIPS n32.
This patch disables use of 64-bit atomics for MIPS n32 to fix the
problems with unaligned semaphores.
Before 64-bit atomics are used for anything for which such alignment
issues do not arise, and before the addition of any new ILP32 ports
with 64-bit semaphores for which the ABI can be set to have the
greater alignment (AARCH64?), a better approach will need to be
established that allows architectures to declare their 64-bit atomics
availability accurately, without doing so causing inappropriate use of
such atomics on unaligned semaphores.
Tested for MIPS n32 that this fixes the nptl/tst-sem3 failure.
* sysdeps/mips/bits/atomic.h [_MIPS_SIM == _ABIN32]
(__HAVE_64B_ATOMICS): Define to 0.
Adhemerval Zanella [Tue, 27 Jan 2015 12:37:10 +0000 (06:37 -0600)]
powerpc: Fix fesetexceptflag [BZ#17885]
This patch fixes a bug introduced by
18f2945ae9216cfc, where it optimizes
the FPSCR set by just issuing a mtfs instruction if new flag is different
from older one. The issue is a typo, where the new flag should the the
new value, instead of the old one.
It fixes BZ#17885.
Adhemerval Zanella [Tue, 27 Jan 2015 18:16:39 +0000 (13:16 -0500)]
powerpc: Fix fsqrt build in libm [BZ#16576]
Some powerpc64 processors (e5500 core for instance) does not provide the
fsqrt instruction, however current check to use in math_private.h is
__WORDSIZE and _ARCH_PWR4 (ISA 2.02). This is patch change it to use
the compiler flag _ARCH_PPCSQ (which is the same condition GCC uses to
decide whether to generate fsqrt instruction).
It fixes BZ#16576.
Andreas Krebbel [Tue, 27 Jan 2015 08:37:04 +0000 (09:37 +0100)]
iconv: Suppress array out of bounds warning.
Andreas Schwab [Sun, 25 Jan 2015 21:31:29 +0000 (22:31 +0100)]
ia64: avoid set-but-not-used warning
Andreas Schwab [Sun, 25 Jan 2015 21:09:05 +0000 (22:09 +0100)]
m68k/coldfire: avoid warning about volatile register variables
Andreas Schwab [Sun, 25 Jan 2015 10:49:50 +0000 (11:49 +0100)]
m68k: fix missing definition of __feraiseexcept
Andreas Schwab [Sun, 25 Jan 2015 09:08:04 +0000 (10:08 +0100)]
m68k: force inlining bswap functions
Bram [Thu, 18 Apr 2013 14:50:49 +0000 (16:50 +0200)]
Fix segmentation fault when LD_LIBRARY_PATH contains only non-existings paths
Adhemerval Zanella [Fri, 23 Jan 2015 16:31:49 +0000 (11:31 -0500)]
powerpc: Fix powerpc64 build failure with binutils 2.22
GLIBC memset optimization for POWER8 uses the '.machine power8'
directive, which is only supported officially on binutils 2.24+. This
causes a build failure on older binutils.
Since the requirement of .machine power8 is to correctly assembly the
'mtvsrd' instruction and it is already handled by the MTVSRD_V1_R4
macro, there is no really needed of using it.
The patch replaces the power8 with power7 for .machine directive.
It fixes BZ#17869.
Adhemerval Zanella [Fri, 23 Jan 2015 16:16:49 +0000 (11:16 -0500)]
powerpc: Fix ifuncmain6pie failure with GCC 4.9
This patch fix the elf/ifuncmain6pie failure when building with GCC
4.9+. For some reason, the compiler removes the branch taken code at
resolve_ifunc (sysdeps/powerpc/powerpc64/dl-machine.h) as dead-code
and thus the testcase fails because the ifunc resolves branches to an
invalid memory location. It fixes by explicit adding a dependency of
value based on odp variable to avoid compiler optimization.
It fixes BZ#17868.
H.J. Lu [Sat, 24 Jan 2015 02:52:45 +0000 (18:52 -0800)]
Also treat model numbers 0x5a/0x5d as Silvermont
H.J. Lu [Sat, 24 Jan 2015 01:27:09 +0000 (17:27 -0800)]
Treat model numbers 0x4a/0x4d as Silvermont
* sysdeps/x86_64/multiarch/init-arch.c (__init_cpu_features):
Treat model numbers 0x4a/0x4d as Intel Silvermont architecture.
H.J. Lu [Sat, 24 Jan 2015 00:21:07 +0000 (16:21 -0800)]
Also use uint64_t in __new_sem_wait_fast
H.J. Lu [Fri, 23 Jan 2015 22:48:40 +0000 (14:48 -0800)]
Use uint64_t and (uint64_t) 1 for 64-bit int
This patch replaces unsigned long int and 1UL with uint64_t and
(uint64_t) 1 to support ILP32 targets like x32.
[BZ #17870]
* nptl/sem_post.c (__new_sem_post): Replace unsigned long int
with uint64_t.
* nptl/sem_waitcommon.c (__sem_wait_cleanup): Replace 1UL with
(uint64_t) 1.
(__new_sem_wait_slow): Replace unsigned long int with uint64_t.
Replace 1UL with (uint64_t) 1.
* sysdeps/nptl/internaltypes.h (new_sem): Replace unsigned long
int with uint64_t.
Roland McGrath [Fri, 23 Jan 2015 19:29:02 +0000 (11:29 -0800)]
Add missing libc_hidden_weak to stub if_nameindex, if_freenameindex.
Roland McGrath [Fri, 23 Jan 2015 18:41:37 +0000 (10:41 -0800)]
Add missing libc_hidden_def to stub getrlimit64.