Commit Graph

10818 Commits

Author SHA1 Message Date
Adhemerval Zanella
3c9399f507 posix: Consolidate Linux msync syscall
This patch consolidates the msync Linux syscall generation on
sysdeps/unix/sysv/linux/msync.c.  It basically removes it from
architectures auto-generation list.

Checked on i686-linux-gnu, x86_64-linux-gnu, x86_64-linux-gnux32,
arch64-linux-gnu, arm-linux-gnueabihf, powerpc64le-linux-gnu,
sparc64-linux-gnu, and sparcv9-linux-gnu.

	* misc/Makefile (CFLAGS-msync.c): New rule.
	* nptl/Makefile (CFLAGS-msync.c): Likewise.
	* sysdeps/unix/syscalls.list: Remove msync from auto-generation list.
	* sysdeps/unix/sysv/linux/msync.c: New file.
2017-05-18 18:06:47 -03:00
Adhemerval Zanella
b3b00aa8b0 posix: Consolidate Linux sigsuspend implementation
This patch consolidates the sigsuspend Linux syscall generation on
sysdeps/unix/sysv/linux/sigsuspend.c.  It basically removes the alpha
assembly version which call the old sigsusped interface using only
the first doubleword from sigset. Current minimum supported kernel
on alpha (3.2) enforces rt_sigsuspend on the architecture
(__ARCH_WANT_SYS_RT_SIGSUSPEND option on kernel), so it is possible
to use the default implementation.

Checked on i686-linux-gnu, x86_64-linux-gnu, x86_64-linux-gnux32,
arch64-linux-gnu, arm-linux-gnueabihf, powerpc64le-linux-gnu,
sparc64-linux-gnu, and sparcv9-linux-gnu.

	* sysdeps/unix/sysv/linux/alpha/sigsuspend.S: Remove file.
	* sysdeps/unix/sysv/linux/sigsuspend.c: Simplify include list.
	* nptl/Makefile (CFLAGS-sigsuspend.c): New rule.
	* sysdeps/unix/sysv/linux/sparc/sparc64/Makefile
	(CFLAGS-sigsuspend.c): Remove rule.
2017-05-18 18:06:47 -03:00
Adhemerval Zanella
2de15b61e9 linux: Consolidate Linux tee implementation
This patch consolidates the tee Linux syscall generation on
sysdeps/unix/sysv/linux/tee.c.  It basically removes it from
architectures auto-generation list.

Checked on i686-linux-gnu, x86_64-linux-gnu, x86_64-linux-gnux32,
arch64-linux-gnu, arm-linux-gnueabihf, powerpc64le-linux-gnu,
sparc64-linux-gnu, and sparcv9-linux-gnu.

	* sysdeps/unix/sysv/linux/Makefile (sysdeps_routines): Add tee.
	(CFLAGS-tee.c): New rule.
	* sysdeps/unix/sysv/linux/syscalls.list: Remove tee from
	auto-generated list.
	* sysdeps/unix/sysv/linux/tee.c: New file.
2017-05-18 18:06:47 -03:00
Adhemerval Zanella
3d0ad5a8df posix: Consolidate Linux nanosleep syscall
This patch consolidates the nanosleep Linux syscall generation on
sysdeps/unix/sysv/linux/nanosleep.c.  It basically removes it from
architectures auto-generation list.

Checked on i686-linux-gnu, x86_64-linux-gnu, x86_64-linux-gnux32,
arch64-linux-gnu, arm-linux-gnueabihf, powerpc64le-linux-gnu,
sparc64-linux-gnu, and sparcv9-linux-gnu.

	* nptl/Makefile (CFLAGS-nanosleep.c): New rule.
	* posix/Makefile (CFLAGS-nanosleep.c): Likewise.
	* sysdeps/unix/sysv/linux/nanosleep.c: New file.
	* sysdeps/unix/sysv/linux/syscalls.list: Remove nanosleep from
	auto-generated list.
2017-05-18 18:06:47 -03:00
Adhemerval Zanella
ebfdb12bf0 posix: Consolidate Linux waitpid syscall
This patch consolidates the waitpid Linux syscall generation on
sysdeps/unix/sysv/linux/waitpid.c.  It basically removes it from
architecture auto-generation list and also remove arch specific
implementations.

Checked on i686-linux-gnu, x86_64-linux-gnu, x86_64-linux-gnux32,
arch64-linux-gnu, arm-linux-gnueabihf, and powerpc64le-linux-gnu.

	* sysdeps/unix/sysv/linux/sh/syscalls.list: Remove waitpid from
	auto-generated list.
	* sysdeps/unix/sysv/linux/m68k/syscalls.list: Likewise.
	* sysdeps/unix/sysv/linux/i386/syscalls.list: Likewise.
	* sysdeps/unix/sysv/linux/sparc/syscalls.list: Remove file.
	* sysdeps/unix/sysv/linux/tile/waitpid.S: Likewise.
	* sysdeps/unix/sysv/linux/powerpc/syscalls.list: Likewise.
2017-05-18 18:06:47 -03:00
Adhemerval Zanella
88499a87ce posix: Consolidate Linux pause syscall
This patch consolidates the pause Linux implementation on
sysdeps/unix/sysv/linux/pause.c.  If defined the pause syscall
(__NR_pause) will be used, other ppoll with 0 arguments will be
used instead.

It has the small advantage of generic pause implementation with
uses rt_sigprocmask plus rt_sigsuspend because it requires only
one syscall and the pause is done atomically regarding signal
handling (for instance, pause may not be interrupted if the
signal arrives between the rt_sigprocmask and rt_sigsuspend
syscall).

Checked on i686-linux-gnu, x86_64-linux-gnu, x86_64-linux-gnux32,
arch64-linux-gnu, arm-linux-gnueabihf, powerpc64le-linux-gnu,
sparc64-linux-gnu, and sparcv9-linux-gnu.

	* sysdeps/unix/sysv/linux/generic/pause.c: Remove file.
	* sysdeps/unix/sysv/linux/sparc/sparc64/pause.c: Likewise.
	* sysdeps/unix/sysv/linux/sparc/kernel-features.h [__arch64__]
	(__NR_pause): Undefine.
	* sysdeps/unix/sysv/linux/pause.c: New file.
	* sysdeps/unix/sysv/linux/syscalls.list: Remove pause from
	auto-generation list.
2017-05-18 18:06:47 -03:00
H.J. Lu
a7fbedff76 Correct comments in x86_64/multiarch/memcmp.S
* sysdeps/x86_64/multiarch/memcmp.S (__GI_memcmp): Correct
	comments.
2017-05-18 14:02:02 -07:00
Rajalakshmi Srinivasaraghavan
dec4a7105e powerpc: Improve memcmp performance for POWER8
Vectorization improves performance over the current implementation.
Tested on powerpc64 and powerpc64le.
2017-05-18 11:21:20 +05:30
Paul Clarke
b2980e3c54 powerpc: Add a POWER8-optimized version of cosf()
This implementation is based on the one already used at
sysdeps/powerpc/powerpc64/fpu/multiarch/s_sinf-power8.S.

	* sysdeps/powerpc/powerpc64/fpu/multiarch/Makefile
	[$(subdir) = math] (libm-sysdep_routines): Add s_cosf-power8 and
	s_cosf-ppc64.
	* sysdeps/powerpc/powerpc64/fpu/multiarch/s_cosf-power8.S: New file.
	* sysdeps/powerpc/powerpc64/fpu/multiarch/s_cosf-ppc64.c: Likewise.
	* sysdeps/powerpc/powerpc64/fpu/multiarch/s_cosf.c: Likewise.
	* sysdeps/powerpc/powerpc64/power8/fpu/s_cosf.S: Likewise.
2017-05-17 18:37:48 -03:00
Gabriel F. T. Gomes
7620dc1235 Convert e_exp2l.c into a template
This patch converts the implementation of exp2l in math/e_exp2l.c into
a template in math/e_exp2_template.c, then adjusts Makefile to use
this template for long double (the implementations for float and
double in sysdeps have higher precedence and are not used).  This
template can also be used for float128, thus reducing the amount of
duplicated code that gets added when adding support the new type.

Tested for powerpc64le and s390x.

	* math/Makefile (libm-calls): Move e_exp2F to gen-libm-calls.
	(gen-libm-calls): Add e_exp2F to use the template.
	* math/e_exp2l.c: Rename to math/e_exp2_template.c.
	* math/e_exp2_template.c: New file, renamed from
	math/e_exp2l.c, and made into a template.
	* sysdeps/generic/math-type-macros.h (M_MIN_EXP): New macro.
2017-05-17 14:44:08 -03:00
Adhemerval Zanella
c79a72aa5c posix: Fix and simplify default p{read,write}v implementation
Currently all architectures but microblaze use wire-up syscall for
p{readv,write}v.  Microblaze still uses the syscall emulation using
sysdep/posix/p{readv,writev}.c and it was reported in some ocasions
[1] [2] that it might have some issues with some linux specific
usage (mainly with O_DIRECT and the alignment requirement).

Although it is not an issue for virtually all the system, this
patch refactors the sysdeps/posix p{read,write}v syscall to avoid
such issue (by using posix_memalign on the buffer used on
p{read,write} call) and by refactoring it common files to avoid
the need check on defines to correct set the alias and internal
symbols.

Checked on microblaze-linux-gnu check with run-built-tests=no and
by using the sysdeps/posix implementation on x86_64-linux-gnu (just
for sanity test where it shown no regression).

	* sysdeps/posix/preadv.c: Use sysdeps/posix/preadv_common.c.
	* sysdeps/posix/preadv64.c: Likewise.
	* sysdeps/unix/sysv/linux/preadv.c: Likewise.
	* sysdeps/unix/sysv/linux/preadv64.c: Likewise.
	* sysdeps/posix/pwritev.c: Use sysdeps/posix/pwritev_common.c.
	* sysdeps/posix/pwritev64.c: Likewise.
	* sysdeps/unix/sysv/linux/pwritev.c: Likewise.
	* sysdeps/unix/sysv/linux/pwritev64.c: Likewise.
	* sysdeps/posix/preadv_common.c: New file.
	* sysdeps/posix/pwritev_common.c: Likewise.

[1] http://www.mail-archive.com/qemu-devel@nongnu.org/msg25282.html
[2] https://bugzilla.redhat.com/show_bug.cgi?id=563103#c8
2017-05-15 16:33:45 -03:00
Gabriel F. T. Gomes
cdd45522b6 float128: Enable use of IEEE wrapper templates
The templates for the IEEE functions wrappers implement wrappers that
do not rely on _LIB_VERSION / matherr / __kernel_standard
functionality to set errno and fix the return value of the functions.
The wrappers are ready to be used by all floating-point types, however
they will first be used by float128, since the old wrappers for float,
double, and long double need to be first deprecated and versioned.

This commits defines __USE_WRAPPER_TEMPLATE to 1 for float128 files,
so that the new wrapper templates are used for this type.

	* sysdeps/generic/math-type-macros-float128.h
	(__USE_WRAPPER_TEMPLATE): Define to 1 to enable use of the
	wrapper templates.
2017-05-15 10:24:14 -03:00
Paul E. Murphy
81f26b53b5 float128: Add private _Float128 declarations for libm.
Add the necessary bits to the private headers to support
building the _Float128 libm functions.

A local override for float.h is provided to include the
missing *FLT128 macros implied by TS 18661-3 for this
type when compiling prior to GCC 7.

	* include/complex.h (__kernel_casinhf128): New declaration.
	* include/float.h: New file.
	* include/math.h (__finitef128): Add a hidden def.
	(__isinff128): Likewise.
	(__isnanf128): Likewise.
	(__fpclassify): Likewise.
	(__issignalling): Likewise.
	(__expf128): Likewise.
	(__expm1f128): Likewise.

	* sysdeps/generic/fix-fp-int-convert-overflow.h:
	(FIX_FLT128_LONG_CONVERT_OVERFLOW): New macro.
	(FIX_FLT128_LLONG_CONVERT_OVERFLOW): Likewise.

	* sysdeps/generic/math-type-macros-float128.h: New file.

	* sysdeps/generic/math_private.h: Include bits/floatn.h and
	math_private_calls.h for _Float128.
	(__isinff128): New inline implementation used when GCC < 7.0,
	since in this case __builtin_isinf_sign is broken.
	(fabsf128): New inline implementation that calls the builtin.
	(__EXPR_FLT128): New macro.
	(min_of_type): Optionally include _Float128 types too.

	* sysdeps/generic/math_private_calls.h (__kernel_sincos):
	Declare for _Float128.
	(__kernel_rem_pio2): Likewise.

	* sysdeps/ieee754/ldbl-opt/s_sin.c:
	(__DECL_SIMD_sincos_disablef128): New macro.
2017-05-15 10:23:28 -03:00
Joseph Myers
e1d6e08dd6 Remove useless SPARC signbit aliases.
The SPARC implementations of __signbit* functions have aliases
signbit, signbitf, signbitl.  These are useless, as they aren't
exported from the shared libraries (only the __signbit* functions are
exported, to be used by the type-generic signbit macro with older
compilers).  This patch removes the useless aliases.

Tested (compilation only) with build-many-glibcs.py for
sparc64-linux-gnu and sparcv9-linux-gnu.

	* sysdeps/sparc/sparc32/fpu/s_signbit.S (signbit): Remove alias.
	(signbitf): Likewise.
	* sysdeps/sparc/sparc64/fpu/multiarch/s_signbit.S (signbit):
	Likewise.
	(signbitl): Likewise.
	* sysdeps/sparc/sparc64/fpu/multiarch/s_signbitf.S (signbitf):
	Likewise.
	* sysdeps/sparc/sparc64/fpu/s_signbit.S (signbit): Likewise.
	(signbitl): Likewise.
	* sysdeps/sparc/sparc64/fpu/s_signbitf.S (signbitf): Likewise.
2017-05-12 16:37:12 +00:00
Florian Weimer
1d2bc2eae9 fork: Remove bogus parent PID assertions [BZ #21386] 2017-05-12 16:04:16 +02:00
Joseph Myers
e8f1225ca4 Remove __ASSUME_STATFS_F_FLAGS.
Now that 3.2 is the minimum Linux kernel version for glibc, this patch
removes __ASSUME_STATFS_F_FLAGS and associated conditional code.

Tested for x86_64.

	* sysdeps/unix/sysv/linux/kernel-features.h
	(__ASSUME_STATFS_F_FLAGS): Remove macro.
	* sysdeps/unix/sysv/linux/internal_statvfs.c
	[!__ASSUME_STATFS_F_FLAGS]: Remove conditional code.
2017-05-12 11:48:37 +00:00
Zack Weinberg
7c3018f9e4 Suppress internal declarations for most of the testsuite.
This patch adds a new build module called 'testsuite'.
IS_IN (testsuite) implies _ISOMAC, as do IS_IN_build and __cplusplus
(which means several ad-hoc tests for __cplusplus can go away).
libc-symbols.h now suppresses almost all of *itself* when _ISOMAC is
defined; in particular, _ISOMAC mode does not get config.h
automatically anymore.

There are still quite a few tests that need to see internal gunk of
one variety or another.  For them, we now have 'tests-internal' and
'test-internal-extras'; files in this category will still be compiled
with MODULE_NAME=nonlib, and everything proceeds as it always has.
The bulk of this patch is moving tests from 'tests' to
'tests-internal'.  There is also 'tests-static-internal', which has
the same effect on files in 'tests-static', and 'modules-names-tests',
which has the *inverse* effect on files in 'modules-names' (it's
inverted because most of the things in modules-names are *not* tests).
For both of these, the file must appear in *both* the new variable and
the old one.

There is also now a special case for when libc-symbols.h is included
without MODULE_NAME being defined at all.  (This happens during the
creation of libc-modules.h, and also when preprocessing Versions
files.)  When this happens, IS_IN is set to be always false and
_ISOMAC is *not* defined, which was the status quo, but now it's
explicit.

The remaining changes to C source files in this patch seemed likely to
cause problems in the absence of the main change.  They should be
relatively self-explanatory.  In a few cases I duplicated a definition
from an internal header rather than move the test to tests-internal;
this was a judgement call each time and I'm happy to change those
however reviewers feel is more appropriate.

	* Makerules: New subdir configuration variables 'tests-internal'
	and 'test-internal-extras'.  Test files in these categories will
	still be compiled with MODULE_NAME=nonlib.  Test files in the
	existing categories (tests, xtests, test-srcs, test-extras) are
	now compiled with MODULE_NAME=testsuite.
	New subdir configuration variable 'modules-names-tests'.  Files
	which are in both 'modules-names' and 'modules-names-tests' will
	be compiled with MODULE_NAME=testsuite instead of
	MODULE_NAME=extramodules.
	(gen-as-const-headers): Move to tests-internal.
	(do-tests-clean, common-mostlyclean): Support tests-internal.
	* Makeconfig (built-modules): Add testsuite.
	* Makefile: Change libof-check-installed-headers-c and
	libof-check-installed-headers-cxx to 'testsuite'.
	* Rules: Likewise.  Support tests-internal.
	* benchtests/strcoll-inputs/filelist#en_US.UTF-8:
	Remove extra-modules.mk.

	* config.h.in: Don't check for __OPTIMIZE__ or __FAST_MATH__ here.
	* include/libc-symbols.h: Move definitions of _GNU_SOURCE,
	PASTE_NAME, PASTE_NAME1, IN_MODULE, IS_IN, and IS_IN_LIB to the
	very top of the file and rationalize their order.
	If MODULE_NAME is not defined at all, define IS_IN to always be
	false, and don't define _ISOMAC.
	If any of IS_IN (testsuite), IS_IN_build, or __cplusplus are
	true, define _ISOMAC and suppress everything else in this file,
	starting with the inclusion of config.h.
	Do check for inappropriate definitions of __OPTIMIZE__ and
	__FAST_MATH__ here, but only if _ISOMAC is not defined.
        Correct some out-of-date commentary.

	* include/math.h: If _ISOMAC is defined, undefine NO_LONG_DOUBLE
	and _Mlong_double_ before including math.h.
	* include/string.h: If _ISOMAC is defined, don't expose
	_STRING_ARCH_unaligned. Move a comment to a more appropriate
	location.

	* include/errno.h, include/stdio.h, include/stdlib.h, include/string.h
	* include/time.h, include/unistd.h, include/wchar.h: No need to
	check __cplusplus nor use __BEGIN_DECLS/__END_DECLS.

	* misc/sys/cdefs.h (__NTHNL): New macro.
	* sysdeps/m68k/m680x0/fpu/bits/mathinline.h
	(__m81_defun): Use __NTHNL to avoid errors with GCC 6.

	* elf/tst-env-setuid-tunables.c: Include config.h with _LIBC
	defined, for HAVE_TUNABLES.
	* inet/tst-checks-posix.c: No need to define _ISOMAC.
	* intl/tst-gettext2.c: Provide own definition of N_.
	* math/test-signgam-finite-c99.c: No need to define _ISOMAC.
	* math/test-signgam-main.c: No need to define _ISOMAC.
	* stdlib/tst-strtod.c: Convert to test-driver. Split locale_test to...
	* stdlib/tst-strtod1i.c: ...this new file.
	* stdlib/tst-strtod5.c: Convert to test-driver and add copyright notice.
        Split tests of __strtod_internal to...
	* stdlib/tst-strtod5i.c: ...this new file.
	* string/test-string.h: Include stdint.h. Duplicate definition of
	inhibit_loop_to_libcall here (from libc-symbols.h).
	* string/test-strstr.c: Provide dummy definition of
	libc_hidden_builtin_def when including strstr.c.
	* sysdeps/ia64/fpu/libm-symbols.h: Suppress entire file in _ISOMAC
	mode; no need to test __STRICT_ANSI__ nor __cplusplus as well.
	* sysdeps/x86_64/fpu/math-tests-arch.h: Include cpu-features.h.
	Don't include init-arch.h.
	* sysdeps/x86_64/multiarch/test-multiarch.h: Include cpu-features.h.
	Don't include init-arch.h.

	* elf/Makefile: Move tst-ptrguard1-static, tst-stackguard1-static,
	tst-tls1-static, tst-tls2-static, tst-tls3-static, loadtest,
	unload, unload2, circleload1, neededtest, neededtest2,
	neededtest3, neededtest4, tst-tls1, tst-tls2, tst-tls3,
	tst-tls6, tst-tls7, tst-tls8, tst-dlmopen2, tst-ptrguard1,
	tst-stackguard1, tst-_dl_addr_inside_object, and all of the
	ifunc tests to tests-internal.
	Don't add $(modules-names) to test-extras.
	* inet/Makefile: Move tst-inet6_scopeid_pton to tests-internal.
	Add tst-deadline to tests-static-internal.
	* malloc/Makefile: Move tst-mallocstate and tst-scratch_buffer to
	tests-internal.
	* misc/Makefile: Move tst-atomic and tst-atomic-long to tests-internal.
	* nptl/Makefile: Move tst-typesizes, tst-rwlock19, tst-sem11,
	tst-sem12, tst-sem13, tst-barrier5, tst-signal7, tst-tls3,
	tst-tls3-malloc, tst-tls5, tst-stackguard1, tst-sem11-static,
	tst-sem12-static, and tst-stackguard1-static to tests-internal.
        Link tests-internal with libpthread also.
	Don't add $(modules-names) to test-extras.
	* nss/Makefile: Move tst-field to tests-internal.
	* posix/Makefile: Move bug-regex5, bug-regex20, bug-regex33,
	tst-rfc3484, tst-rfc3484-2, and tst-rfc3484-3 to tests-internal.
	* stdlib/Makefile: Move tst-strtod1i, tst-strtod3, tst-strtod4,
	tst-strtod5i, tst-tls-atexit, and tst-tls-atexit-nodelete to
	tests-internal.
        * sunrpc/Makefile: Move tst-svc_register to tests-internal.
	* sysdeps/powerpc/Makefile: Move test-get_hwcap and
	test-get_hwcap-static to tests-internal.
	* sysdeps/unix/sysv/linux/Makefile: Move tst-setgetname to
	tests-internal.
	* sysdeps/x86_64/fpu/Makefile: Add all libmvec test modules to
	modules-names-tests.
2017-05-11 19:27:59 -04:00
Zack Weinberg
1711991592 Remove _IO_MTSAFE_IO from public headers.
_IO_MTSAFE_IO controls whether stdio is *built* with support for
multithreading.  In the distant past it might also have worked as a
feature selection macro, allowing library *users* to select
thread-safe or lock-free stdio at application build time, I haven't
done the archaeology.  Nowadays, defining _IO_MTSAFE_IO while using
the installed headers, or in _ISOMAC mode, will cause libio.h to throw
syntax errors.

This patch removes _IO_MTSAFE_IO from the public headers
(specifically, from libio/libio.h).  The most important thing it
controlled in there was whether libio.h defines _IO_lock_t itself or
expects stdio-lock.h to have done it, and we do still need a
inter-header communication macro for that, because stdio-lock.h can
only define _IO_lock_t as a typedef.  I've invented
_IO_lock_t_defined, which is defined by both versions of stdio-lock.h.

_IO_MTSAFE_IO also controlled the definitions of a handful of macros
that _might_ count as part of the public libio.h interface.  They are
now unconditionally given their non-_IO_MTSAFE_IO definition in
libio/libio.h, and include/libio.h redefines them with the
_IO_MTSAFE_IO definition.  This should minimize the odds of breaking
old software that actually uses those macros.

I suspect that this entire mechanism is vestigial, and that glibc
won't build anymore if you *don't* define _IO_MTSAFE_IO, but that's
another patchset.  The bulk of libio.h is internal-use-only stuff that
no longer makes sense to expose (libstdc++ gave up on making a FILE
the same object as a C++ filebuf *decades* ago) but that, too, is
another patchset.

	* libio/libio.h: Condition dummy definition of _IO_lock_t on
	_IO_lock_t_defined, not _IO_MTSAFE_IO. Unconditionally use the
	non-_IO_MTSAFE_IO definitions for _IO_peekc, _IO_flockfile,
	_IO_funlockfile, and _IO_ftrylockfile.  Only define
	_IO_cleanup_region_start and _IO_cleanup_region_end if not
	already defined.
	* include/libio.h: If _IO_MTSAFE_IO is defined, redefine
        _IO_peekc, _IO_flockfile, _IO_funlockfile, and _IO_ftrylockfile
        appropriately.
	* sysdeps/generic/stdio-lock.h, sysdeps/nptl/stdio-lock.h:
	Define _IO_lock_t_defined after defining _IO_lock_t.
2017-05-11 19:14:11 -04:00
Adhemerval Zanella
31073a53d8 powerpc: Fix signal handling in backtrace
Now with read consolidation which uses SYSCALL_CANCEL macro, a frame
pointer is created in the syscall code and this makes the powerpc
backtrace obtain a bogus entry for the signal handling patch.

It is because it does not setup the correct frame pointer register
(r1) based on the saved value from the kernel sigreturn.  It was not
failing because the syscall frame pointer register was the same one
for the next frame (the function that actually called the syscall).

This patch fixes it by setup the next stack frame using the saved
one by the kernel sigreturn.  It fixes tst-backtrace{5,6} from
the read consolidation patch.

Checked on powerpc-linux-gnu and powerpc64le-linux-gnu.

	* sysdeps/powerpc/powerpc32/backtrace.c (is_sigtramp_address): Use
	void* for argument type and use VDSO_SYMBOL macro.
	(is_sigtramp_address_rt): Likewise.
	(__backtrace): Setup expected frame pointer address for signal
	handling.
	* sysdeps/powerpc/powerpc64/backtrace.c (is_sigtramp_address): Use
	void* for argumetn type and use VSDO_SYMBOL macro.
	(__backtrace): Setup expected frame pointer address for signal
	handling.
2017-05-11 17:27:31 -03:00
Adhemerval Zanella
488e08b600 Consolidate Linux writev implementation
This patch consolidates the writev Linux syscall implementation on
sysdeps/unix/sysv/linux/writev.c.

Checked on i686-linux-gnu, x86_64-linux-gnu, x86_64-linux-gnux32,
arch64-linux-gnu, arm-linux-gnueabihf, and powerpc64le-linux-gnu.

	* sysdeps/unix/sysv/linux/writev.c: New file.
2017-05-11 17:27:31 -03:00
Adhemerval Zanella
679e979bf3 Consolidate Linux readv implementation
This patch consolidates the readv Linux syscall implementation on
sysdeps/unix/sysv/linux/readv.c.

Checked on i686-linux-gnu, x86_64-linux-gnu, x86_64-linux-gnux32,
arch64-linux-gnu, arm-linux-gnueabihf, and powerpc64le-linux-gnu.

	* sysdeps/unix/sysv/linux/readv.c: New file.
2017-05-11 17:27:31 -03:00
Adhemerval Zanella
dfaaee33ba Consolidate Linux write syscall
This patch consolidates the write Linux syscall implementation on
sysdeps/unix/sysv/linux/write.c.

Checked on i686-linux-gnu, x86_64-linux-gnu, x86_64-linux-gnux32,
arch64-linux-gnu, arm-linux-gnueabihf, and powerpc64le-linux-gnu.

	* include/unistd.h (write): Add hidden proto.
	* io/Makefile (CFLAGS-write.c): New rule.
	* nptl/Makefile (CFLAGS-write.c): Likewise.
	* sysdeps/unix/sysv/linux/write.c: New file.
2017-05-11 17:27:30 -03:00
Adhemerval Zanella
f6a191a6ee Consolidate Linux read syscall
This patch consolidates the read Linux syscall implementation on
sysdeps/unix/sysv/linux/read.c.  This leads to a different frame
pointer creation on some architectures:

  * It fixes BZ#21428 on aarch64, since now the returned address
    for the read syscall can be correctly found out by
    backtrace_symbols.

  * It makes tst-backtrace{5,6} fails on powerpc due an issue on
    its custom backtrace implementation.  It is fixed on subsequent
    patch from this set.

Checked on i686-linux-gnu, x86_64-linux-gnu, x86_64-linux-gnux32,
arch64-linux-gnu, arm-linux-gnueabihf, and powerpc64le-linux-gnu.

	[BZ #21428]
	* include/unistd.h (read): Add hidden proto.
	* io/Makefile (CFLAGS-read.c): New rule.
	* nptl/Makefile (CFLAGS-read.c): New rule.
	* sysdeps/unix/sysv/linux/read.c: New file.
2017-05-11 17:27:30 -03:00
Adhemerval Zanella
cab6e5af9d Consolidate Linux creat implementation
This patch consolidates the creat Linux syscall implementation on
sysdeps/unix/sysv/linux/creat{64}.c.  The changes are:

  1. Remove creat{64} from auto-generation syscalls.list.
  2. Add a new creat{64}.c implementation.  For architectures that
     define __OFF_T_MATCHES_OFF64_T the default creat64 will create
     alias to required creat symbols.
  3. Use __NR_creat where possible, otherwise use internal open{64}
     call with expected flags.

Checked on i686-linux-gnu, x86_64-linux-gnu, x86_64-linux-gnux32,
arch64-linux-gnu, arm-linux-gnueabihf, and powerpc64le-linux-gnu.

	* io/Makefile (CFLAGS-creat.c): New rule.
	(CFLAGS-creat64.c): Likewise.
	* sysdeps/unix/sysv/linux/alpha/creat.c: Remove file.
	* sysdeps/unix/sysv/linux/generic/creat.c: Likewise.
	* sysdeps/unix/sysv/linux/wordsize-64/creat64.c: Likewise.
	* sysdeps/unix/sysv/linux/creat.c: New file.
	* sysdeps/unix/sysv/linux/creat64.c: Likewise.
	* sysdeps/unix/sysv/linux/syscalls.list: Remove create from
	auto-generated list.
	* sysdeps/unix/sysv/linux/wordsize-64/syscalls.list: Likewise.
2017-05-11 17:27:27 -03:00
Adhemerval Zanella
b41152d716 Consolidate Linux open implementation
This patch consolidates the open Linux syscall implementation on
sysdeps/unix/sysv/linux/open{64}.c.  The changes are:

  1. Remove open{64} from auto-generation syscalls.list.
  2. Add a new open{64}.c implementation.  For architectures that
     define __OFF_T_MATCHES_OFF64_T the default open64 will create
     alias to required open symbols.
  3. Use __NR_openat as default syscall for open{64}.

Checked on i686-linux-gnu, x86_64-linux-gnu, x86_64-linux-gnux32,
arch64-linux-gnu, arm-linux-gnueabihf, and powerpc64le-linux-gnu.

	* sysdeps/unix/sysv/linux/generic/open.c: Remove file.
	* sysdeps/unix/sysv/linux/generic/open64.c: Likewise.
	* sysdeps/unix/sysv/linux/wordsize-64/open64.c: Likewise.
	* sysdeps/unix/sysv/linux/open.c: New file.
	* sysdeps/unix/sysv/linux/open64.c (__libc_open64): Use O_LARGEFILE
	only for __OFF_T_MATCHES_OFF64_T and add alias to open if the case.
	* sysdeps/unix/sysv/linux/wordsize-64/syscalls.list: Remove open
	from auto-generated list.
2017-05-11 15:49:10 -03:00
Adhemerval Zanella
0f01acb340 Consolidate Linux close syscall generation
This patch consolidates the close Linux syscall generation on
sysdeps/unix/sysv/linux/close.c.

Checked on i686-linux-gnu, x86_64-linux-gnu, x86_64-linux-gnux32,
arch64-linux-gnu, arm-linux-gnueabihf, and powerpc64le-linux-gnu.

	* nptl/Makefile (CFLAGS-close.c): New flag.
	* sysdeps/unix/sysv/linux/close.c: New file.
2017-05-11 15:49:10 -03:00
Joseph Myers
5df4854ed2 Condition some sys/ucontext.h contents on __USE_MISC (bug 21457).
Continuing the fixes for namespace issues arising from sys/ucontext.h,
this patch conditions various definitions, that are not needed for
defining mcontext_t / ucontext_t, on __USE_MISC, so they do not appear
in strict POSIX modes.

This patch is non-exhaustive; that is, it only conditions
straightforward cases and there may be more such definitions that can
be conditioned for these and other architectures, to be dealt with
later in separate patches.  Also, using __USE_MISC is the minimum
change for these definitions where they conflict with POSIX; some
headers already have __USE_GNU conditionals on similar definitions of
names for registers.  The patch specifically does not do anything with
definitions in bits/sigcontext.h, and nor does it condition any
inclusions of bits/sigcontext.h even where in fact that is not needed
on some architectures for the definitions of mcontext_t / ucontext_t.

As other namespace issues in these headers remain, this patch does not
fix bug 21457, nor allow any XFAILs to be removed.

Tested with build-many-glibcs.py.

	[BZ #21457]
	* sysdeps/arm/sys/ucontext.h (R0): Condition on [__USE_MISC].
	(R1): Likewise.
	(R2): Likewise.
	(R3): Likewise.
	(R4): Likewise.
	(R5): Likewise.
	(R6): Likewise.
	(R7): Likewise.
	(R8): Likewise.
	(R9): Likewise.
	(R10): Likewise.
	(R11): Likewise.
	(R12): Likewise.
	(R13): Likewise.
	(R14): Likewise.
	(R15): Likewise.
	* sysdeps/i386/sys/ucontext.h (REG_GS): Likewise.
	(REG_FS): Likewise.
	(REG_ES): Likewise.
	(REG_DS): Likewise.
	(REG_EDI): Likewise.
	(REG_ESI): Likewise.
	(REG_EBP): Likewise.
	(REG_ESP): Likewise.
	(REG_EBX): Likewise.
	(REG_EDX): Likewise.
	(REG_ECX): Likewise.
	(REG_EAX): Likewise.
	(REG_TRAPNO): Likewise.
	(REG_ERR): Likewise.
	(REG_EIP): Likewise.
	(REG_CS): Likewise.
	(REG_EFL): Likewise.
	(REG_UESP): Likewise.
	(REG_SS): Likewise.
	* sysdeps/m68k/sys/ucontext.h (R_D0): Likewise.
	(R_D1): Likewise.
	(R_D2): Likewise.
	(R_D3): Likewise.
	(R_D4): Likewise.
	(R_D5): Likewise.
	(R_D6): Likewise.
	(R_D7): Likewise.
	(R_A0): Likewise.
	(R_A1): Likewise.
	(R_A2): Likewise.
	(R_A3): Likewise.
	(R_A4): Likewise.
	(R_A5): Likewise.
	(R_A6): Likewise.
	(R_A7): Likewise.
	(R_SP): Likewise.
	(R_PC): Likewise.
	(R_PS): Likewise.
	(fpregset_t): Likewise.
	(MCONTEXT_VERSION): Likewise.
	* sysdeps/mips/sys/ucontext.h (CTX_R0): Likewise.
	(CTX_AT): Likewise.
	(CTX_V0): Likewise.
	(CTX_V1): Likewise.
	(CTX_A0): Likewise.
	(CTX_A1): Likewise.
	(CTX_A2): Likewise.
	(CTX_A3): Likewise.
	(CTX_T0): Likewise.
	(CTX_T1): Likewise.
	(CTX_T2): Likewise.
	(CTX_T3): Likewise.
	(CTX_T4): Likewise.
	(CTX_T5): Likewise.
	(CTX_T6): Likewise.
	(CTX_T7): Likewise.
	(CTX_S0): Likewise.
	(CTX_S1): Likewise.
	(CTX_S2): Likewise.
	(CTX_S3): Likewise.
	(CTX_S4): Likewise.
	(CTX_S5): Likewise.
	(CTX_S6): Likewise.
	(CTX_S7): Likewise.
	(CTX_T8): Likewise.
	(CTX_T9): Likewise.
	(CTX_K0): Likewise.
	(CTX_K1): Likewise.
	(CTX_GP): Likewise.
	(CTX_SP): Likewise.
	(CTX_S8): Likewise.
	(CTX_RA): Likewise.
	(CTX_MDLO): Likewise.
	(CTX_MDHI): Likewise.
	(CTX_CAUSE): Likewise.
	(CTX_EPC): Likewise.
	* sysdeps/unix/sysv/linux/aarch64/sys/ucontext.h: Condition
	inclusion of <sys/procfs.h> on [__USE_MISC].
	(greg_t): Condition on [__USE_MISC].
	(gregset_t): Likewise.
	(fpregset_t): Likewise.
	* sysdeps/unix/sysv/linux/arm/sys/ucontext.h (greg_t): Likewise.
	(NGREG): Likewise.
	(gregset_t): Likewise.
	(REG_R0): Likewise.
	(REG_R1): Likewise.
	(REG_R2): Likewise.
	(REG_R3): Likewise.
	(REG_R4): Likewise.
	(REG_R5): Likewise.
	(REG_R6): Likewise.
	(REG_R7): Likewise.
	(REG_R8): Likewise.
	(REG_R9): Likewise.
	(REG_R10): Likewise.
	(REG_R11): Likewise.
	(REG_R12): Likewise.
	(REG_R13): Likewise.
	(REG_R14): Likewise.
	(REG_R15): Likewise.
	(struct _libc_fpstate): Likewise.
	(fpregset_t): Likewise.
	* sysdeps/unix/sysv/linux/hppa/sys/ucontext.h (NGREG): Likewise.
	(NFPREG): Likewise.
	(gregset_t): Likewise.
	(fpregset_t): Likewise.
	* sysdeps/unix/sysv/linux/m68k/sys/ucontext.h (R_D0): Likewise.
	(R_D1): Likewise.
	(R_D2): Likewise.
	(R_D3): Likewise.
	(R_D4): Likewise.
	(R_D5): Likewise.
	(R_D6): Likewise.
	(R_D7): Likewise.
	(R_A0): Likewise.
	(R_A1): Likewise.
	(R_A2): Likewise.
	(R_A3): Likewise.
	(R_A4): Likewise.
	(R_A5): Likewise.
	(R_A6): Likewise.
	(R_A7): Likewise.
	(R_SP): Likewise.
	(R_PC): Likewise.
	(R_PS): Likewise.
	(fpregset_t): Likewise.
	(MCONTEXT_VERSION): Likewise.
	* sysdeps/unix/sysv/linux/nios2/sys/ucontext.h (MCONTEXT_VERSION):
	Likewise.
	* sysdeps/unix/sysv/linux/sh/sys/ucontext.h (REG_R0): Likewise.
	(REG_R1): Likewise.
	(REG_R2): Likewise.
	(REG_R3): Likewise.
	(REG_R4): Likewise.
	(REG_R5): Likewise.
	(REG_R6): Likewise.
	(REG_R7): Likewise.
	(REG_R8): Likewise.
	(REG_R9): Likewise.
	(REG_R10): Likewise.
	(REG_R11): Likewise.
	(REG_R12): Likewise.
	(REG_R13): Likewise.
	(REG_R14): Likewise.
	(REG_R15): Likewise.
	* sysdeps/unix/sysv/linux/tile/sys/ucontext.h: Condition inclusion
	of <arch/abi.h> on [__USE_MISC].
	(greg_t): Condition on [__USE_MISC].
	(NGREG): Likewise.
	(gregset_t): Likewise.
2017-05-11 14:15:26 +00:00
Adhemerval Zanella
0a19a91896 Remove wrong definitions from pthread header refactor
This patch removes wrong struct definition from eab380d (Move shared
pthread definitions to common headers) on ARM and hppa.

Checked on arm-linux-gnueabihf.

	* sysdeps/arm/nptl/bits/pthreadtypes-arch.h
	(__pthread_rwlock_arch_t): Remove __data definition.
	* sysdeps/hppa/nptl/bits/pthreadtypes-arch.h
	(__pthread_rwlock_arch_t): Likewise.
2017-05-11 10:46:03 -03:00
Florian Weimer
46ce8881ad getaddrinfo: Unconditionally use malloc for address list
getaddrinfo has to call malloc eventually anyway, so the complexity
of avoiding malloc calls is not worth potential savings.
2017-05-11 10:01:49 +02:00
Joseph Myers
d08a482bc2 Remove MIPS32 accept4, recvmmsg, sendmmsg implementations.
MIPS32 has its own implementations of accept4, recvmmsg and sendmmsg
because at one point it needed to avoid socketcall being used for
those functions (MIPS32 has socketcall, but has never used it in
glibc, and so never had socket.S at the time when socketcall used such
a per-architecture file instead of C code).  The current code no
longer uses socketcall based on __NR_socketcall being defined, and the
syscalls are always present on MIPS for supported kernels so the
socketcall case in the code is dead for MIPS; this patch removes the
implementations that are, as Adhemerval noted, no longer needed.

Tested compilation for mips-linux-gnu with build-many-glibcs.py.

	* sysdeps/unix/sysv/linux/mips/mips32/accept4.c: Remove file.
	* sysdeps/unix/sysv/linux/mips/mips32/recvmmsg.c: Likewise.
	* sysdeps/unix/sysv/linux/mips/mips32/sendmmsg.c: Likewise.
2017-05-09 23:16:19 +00:00
Joseph Myers
e3b0580d0d Simplify accept4, recvmmsg, sendmmsg code.
The accept4, recvmmsg and sendmmsg functions had macros
__ASSUME_*_SYSCALL_WITH_SOCKETCALL.  Before we could assume kernels
with the relevant functionality, these macros represented the
conditions under which, on a socketcall architecture, glibc could just
call the syscall unconditionally and not have to deal with socketcall
at all for those functions, because if the syscall didn't work for
them the socketcall call wouldn't either.

Now we can assume kernels with the relevant functionality, the only
question is whether we can assume the syscall is present; if not, we
are on a socketcall architecture and just use socketcall instead.
Thus, this patch removes the macros that are no longer necessary, and
simplifies the code for accept4, recvmmsg and sendmmsg to use the same
logic as the other C implementations of socket functions that may use
a syscall or socketcall depending on kernel support.

Tested for x86_64 and x86.

	* sysdeps/unix/sysv/linux/accept4.c (accept4): Use syscall if
	[__ASSUME_ACCEPT4_SYSCALL], otherwise socketcall.
	* sysdeps/unix/sysv/linux/recvmmsg.c (recvmmsg): Use syscall if
	[__ASSUME_RECVMMSG_SYSCALL], otherwise socketcall.
	* sysdeps/unix/sysv/linux/sendmmsg.c (__sendmmsg): Use syscall if
	[__ASSUME_SENDMMSG_SYSCALL], otherwise socketcall.
	* sysdeps/unix/sysv/linux/kernel-features.h
	(__ASSUME_ACCEPT4_SYSCALL): Move to general list of macros for
	socket syscalls.
	(__ASSUME_RECVMMSG_SYSCALL): Likewise.
	(__ASSUME_SENDMMSG_SYSCALL): Likewise.
	* sysdeps/unix/sysv/linux/i386/kernel-features.h
	(__ASSUME_RECVMMSG_SYSCALL_WITH_SOCKETCALL): Remove.
	(__ASSUME_SENDMMSG_SYSCALL_WITH_SOCKETCALL): Likewise.
	* sysdeps/unix/sysv/linux/microblaze/kernel-features.h
	(__ASSUME_RECVMMSG_SYSCALL_WITH_SOCKETCALL): Remove.
	* sysdeps/unix/sysv/linux/powerpc/kernel-features.h
	(__ASSUME_SENDMMSG_SYSCALL_WITH_SOCKETCALL): Likewise.
	* sysdeps/unix/sysv/linux/sh/kernel-features.h
	(__ASSUME_SENDMMSG_SYSCALL_WITH_SOCKETCALL): Likewise.
	* sysdeps/unix/sysv/linux/sparc/kernel-features.h
	(__ASSUME_ACCEPT4_SYSCALL_WITH_SOCKETCALL): Likewise.
	(__ASSUME_RECVMMSG_SYSCALL_WITH_SOCKETCALL): Likewise.
	(__ASSUME_SENDMMSG_SYSCALL_WITH_SOCKETCALL): Likewise.
2017-05-09 21:59:36 +00:00
Adhemerval Zanella
eab380d8ec Move shared pthread definitions to common headers
This patch removes all the replicated pthread definition accross the
architectures and consolidates it on shared headers.  The new
organization is as follow:

  * Architecture specific definition (such as pthread types sizes) are
    place in the new pthreadtypes-arch.h header in arch specific path.

  * All shared structure definition are moved to a common NPTL header
    at sysdeps/nptl/bits/pthreadtypes.h (with now includes the arch
    specific one for internal definitions).

  * Also, for C11 future thread support, both mutex and condition
    definition are placed in a common header at
    sysdeps/nptl/bits/thread-shared-types.h.

It is also a refactor patch without expected functional changes.
Checked with a build for all major ABI (aarch64-linux-gnu, alpha-linux-gnu,
arm-linux-gnueabi, i386-linux-gnu, ia64-linux-gnu,
m68k-linux-gnu, microblaze-linux-gnu, mips{64}-linux-gnu, nios2-linux-gnu,
powerpc{64le}-linux-gnu, s390{x}-linux-gnu, sparc{64}-linux-gnu,
tile{pro,gx}-linux-gnu, and x86_64-linux-gnu).

	* posix/Makefile (headers): Add pthreadtypes-arch.h and
	thread-shared-types.h.
	* sysdeps/aarch64/nptl/bits/pthreadtypes-arch.h: New file: arch
	specific thread definition.
	* sysdeps/alpha/nptl/bits/pthreadtypes-arch.h: Likewise.
	* sysdeps/arm/nptl/bits/pthreadtypes-arch.h: Likewise.
	* sysdeps/hppa/nptl/bits/pthreadtypes-arch.h: Likewise.
	* sysdeps/ia64/nptl/bits/pthreadtypes-arch.h: Likewise.
	* sysdeps/m68k/nptl/bits/pthreadtypes-arch.h: Likewise.
	* sysdeps/microblaze/nptl/bits/pthreadtypes-arch.h: Likewise.
	* sysdeps/mips/nptl/bits/pthreadtypes-arch.h: Likewise.
	* sysdeps/nios2/nptl/bits/pthreadtypes-arch.h: Likewise.
	* sysdeps/powerpc/nptl/bits/pthreadtypes-arch.h: Likewise.
	* sysdeps/s390/nptl/bits/pthreadtypes-arch.h: Likewise.
	* sysdeps/sh/nptl/bits/pthreadtypes-arch.h: Likewise.
	* sysdeps/sparc/nptl/bits/pthreadtypes-arch.h: Likewise.
	* sysdeps/tile/nptl/bits/pthreadtypes-arch.h: Likewise.
	* sysdeps/x86/nptl/bits/pthreadtypes-arch.h: Likewise.
	* sysdeps/nptl/bits/thread-shared-types.h: New file: shared
	thread definition between POSIX and C11.
	* sysdeps/aarch64/nptl/bits/pthreadtypes.h.: Remove file.
	* sysdeps/alpha/nptl/bits/pthreadtypes.h: Likewise.
	* sysdeps/arm/nptl/bits/pthreadtypes.h: Likewise.
	* sysdeps/hppa/nptl/bits/pthreadtypes.h: Likewise.
	* sysdeps/m68k/nptl/bits/pthreadtypes.h: Likewise.
	* sysdeps/microblaze/nptl/bits/pthreadtypes.h: Likewise.
	* sysdeps/mips/nptl/bits/pthreadtypes.h: Likewise.
	* sysdeps/nios2/nptl/bits/pthreadtypes.h: Likewise.
	* sysdeps/ia64/nptl/bits/pthreadtypes.h: Likewise.
	* sysdeps/powerpc/nptl/bits/pthreadtypes.h: Likewise.
	* sysdeps/s390/nptl/bits/pthreadtypes.h: Likewise.
	* sysdeps/sh/nptl/bits/pthreadtypes.h: Likewise.
	* sysdeps/sparc/nptl/bits/pthreadtypes.h: Likewise.
	* sysdeps/tile/nptl/bits/pthreadtypes.h: Likewise.
	* sysdeps/x86/nptl/bits/pthreadtypes.h: Likewise.
	* sysdeps/nptl/bits/pthreadtypes.h: New file: common thread
	definitions shared across all architectures.
2017-05-09 17:49:17 -03:00
Joseph Myers
0cb5da5c09 Simplify sendmmsg code.
Now we can assume a kernel with sendmmsg support, this patch
simplifies the implementation to be similar to that for accept4:
either using socketcall or the syscall according to whether the
syscall is known to be available, without further fallback
implementations.  The __ASSUME_SENDMMSG macro is kept (now defined
unconditionally), since it's used in resolv/res_send.c.

Tested for x86_64 and x86.

	* sysdeps/unix/sysv/linux/kernel-features.h
	(__ASSUME_SENDMMSG_SYSCALL): Define unconditionally.
	(__ASSUME_SENDMMSG): Likewise.
	(__ASSUME_SENDMMSG_SOCKETCALL): Remove macro.
	* sysdeps/unix/sysv/linux/sendmmsg.c (__sendmmsg): Define using
	sendmmsg syscall if that can be assumed to be present, socketcall
	otherwise, with no fallback for runtime failure.
2017-05-09 20:01:52 +00:00
Joseph Myers
9a45f54310 Simplify recvmmsg code.
Now we can assume a kernel with recvmmsg support, this patch
simplifies the implementation to be similar to that for accept4:
either using socketcall or the syscall according to whether the
syscall is known to be available, without further fallback
implementations.

(In fact further simplification is possible, getting rid of the
__ASSUME_*_SYSCALL_WITH_SOCKETCALL macros now that the minimum kernel
is guaranteed support for all of accept4, recvmmsg, sendmmsg, whether
through syscalls or through socketcall.  I intend to do that for all
of accept4 / recvmmsg / sendmmsg together - so making their
implementations just like those for older socket functions - once the
basic cleanup for 3.2 minimum kernel is done for sendmmsg as well as
recvmmsg.)

Tested for x86_64 and x86.

	* sysdeps/unix/sysv/linux/kernel-features.h
	(__ASSUME_RECVMMSG_SYSCALL): Define unconditionally.
	(__ASSUME_RECVMMSG_SOCKETCALL): Remove macro.
	(__ASSUME_RECVMMSG): Likewise.
	* sysdeps/unix/sysv/linux/recvmmsg.c (recvmmsg): Define using
	recvmmsg syscall if it can be assumed to be present, socketcall
	otherwise, with no fallback for runtime failure.
2017-05-09 20:01:01 +00:00
Paul E. Murphy
26265c3bce float128: Add _Float128 make bits to libm.
This adds the appropriate common bits for a platform to
enable float128 and expose ABI.

	* math/Makefile:
	(type-float128-suffix): New variable
	(type-float128-routines): Likewise
	(type-float128-yes): Likewise
	(types): Append float128 if supported
	(types-basic): New variable to control the use of templates for
	float, double, and long double, but not for float128 or newer types.
	(type-basic-foreach): Likewise.

	* sysdeps/ieee754/float128/Makeconfig: New file.
	* sysdeps/ieee754/float128/Versions: New file.
2017-05-09 11:40:28 -03:00
Joseph Myers
695d7d138e Assume prlimit64 is available.
This patch makes sysdeps/unix/sysv/linux code assume the prlimit64
syscall is always available, given the minimum of a 3.2 kernel.

__ASSUME_PRLIMIT64, which in fact was no longer used, is removed.
Code conditional on __NR_prlimit64 being defined is made
unconditional.  Fallback code for the case where prlimit64 produces an
ENOSYS error is removed, substantially simplifying some functions.

Tested for x86_64 and x86.

	* sysdeps/unix/sysv/linux/kernel-features.h (__ASSUME_PRLIMIT64):
	Remove macro.
	* sysdeps/unix/sysv/linux/getrlimit64.c (__getrlimit64): Assume
	prlimit64 is always available and does not give an ENOSYS error.
	* sysdeps/unix/sysv/linux/prlimit.c [__NR_prlimit64]: Make code
	unconditional.
	[!__NR_prlimit64]: Remove conditional code.
	* sysdeps/unix/sysv/linux/setrlimit.c (__setrlimit): Assume
	prlimit64 is always available and does not give an ENOSYS error.
	* sysdeps/unix/sysv/linux/setrlimit64.c (__setrlimit64): Likewise.
2017-05-09 14:05:09 +00:00
Zack Weinberg
2bfdaeddaa Rename cppflags-iterator.mk to libof-iterator.mk, remove extra-modules.mk.
cppflags-iterator.mk no longer has anything to do with CPPFLAGS; all
it does is set libof-$(foo) for a list of files.  extra-modules.mk
does the same thing, but with a different input variable, and doesn't
let the caller control the module.  Therefore, this patch gives
cppflags-iterator.mk a better name, removes extra-modules.mk, and
updates all uses of both.

	* extra-modules.mk: Delete file.
	* cppflags-iterator.mk: Rename to ...
	* libof-iterator.mk: ...this.  Adjust comments.

	* Makerules, extra-lib.mk, benchtests/Makefile, elf/Makefile
	* elf/rtld-Rules, iconv/Makefile, locale/Makefile, malloc/Makefile
	* nscd/Makefile, sunrpc/Makefile, sysdeps/s390/Makefile:
	Use libof-iterator.mk instead of cppflags-iterator.mk or
	extra-modules.mk.

	* benchtests/strcoll-inputs/filelist#en_US.UTF-8: Remove
	extra-modules.mk and cppflags-iterator.mk, add libof-iterator.mk.
2017-05-09 07:06:29 -04:00
Stefan Liebler
61f4fa7fd7 S390: Regenerate ULPs
Updated ulps file - Needed if build with GCC 7.

ChangeLog:

	* sysdeps/s390/fpu/libm-test-ulps: Regenerated.
2017-05-09 09:18:09 +02:00
Joseph Myers
1721145f03 Remove __ASSUME_PROC_PID_TASK_COMM.
This patch removes the __ASSUME_PROC_PID_TASK_COMM macro, and
associated conditional code (in a testcase), now that 3.2 is the
global minimum Linux kernel version supported.

Tested for x86_64.

	* sysdeps/unix/sysv/linux/kernel-features.h
	(__ASSUME_PROC_PID_TASK_COMM): Remove macro.
	* sysdeps/unix/sysv/linux/tst-setgetname.c: Do not include
	<kernel-features.h>.
	(do_test) [!__ASSUME_PROC_PID_TASK_COMM]: Remove conditional code.
2017-05-08 16:44:54 +00:00
Joseph Myers
1278ed2c64 Remove __ASSUME_GETCPU_SYSCALL.
This patch removes the definition of __ASSUME_GETCPU_SYSCALL.  In fact
this macro is unused, probably since:

commit dd26c44403
Author: Adhemerval Zanella <adhemerval.zanella@linaro.com>
Date:   Wed Apr 22 14:21:39 2015 -0300

    Consolidate sched_getcpu

so it could have been removed even without the move to 3.2 as minimum
kernel version on x86_64.

Tested for x86_64.

	* sysdeps/unix/sysv/linux/x86_64/kernel-features.h
	(__ASSUME_GETCPU_SYSCALL): Remove macro.
2017-05-08 14:36:42 +00:00
Florian Weimer
cd354a3849 Remove <sys/ultrasound.h>
This header was once used for constants related to the Gravis
Ultrasound sound card.
2017-05-08 14:59:21 +02:00
Joseph Myers
139ace9575 Require Linux kernel 3.2 or later on x86 / x86_64.
As per the recent discussion, this patch implements a requirement for
Linux 3.2 or later for x86 and x86_64.  This is only the initial
change to increase the configured minimum; it's expected that followup
patches would deal with associated removal of conditionals that are no
longer needed.  If we remove the start-up test on the kernel version,
of course the NEWS and README text should then be revised (to reflect
that this version is just one such that glibc does not intend to
include compatibility code for any older kernel version, rather than
older kernels necessarily failing to work or glibc necessarily having
compatibility code for newer interfaces).

The followups would be able to assume presence of getcpu (x86_64),
recvmmsg (not always through its own syscall, sometimes only through
socketcall), sendmmsg (likewise), /proc/$pid/task/$tid/comm, f_flags
from statfs, prlimit64.

Tested for x86_64 and x86.

	* sysdeps/unix/sysv/linux/i386/configure.ac (arch_minimum_kernel):
	Remove.
	* sysdeps/unix/sysv/linux/i386/configure: Regenerated.
	* sysdeps/unix/sysv/linux/x86_64/64/configure.ac
	(arch_minimum_kernel): Remove.
	* sysdeps/unix/sysv/linux/x86_64/64/configure: Regenerated.
	* README: Update statement about Linux kernel requirements.
2017-05-08 10:45:20 +00:00
Joseph Myers
176804300b Fix network headers stdint.h namespace (bug 21455).
conform/ namespace tests of arpa/inet.h, netdb.h and netinet/in.h fail
for UNIX98 and XPG42 because of inclusion of stdint.h, which defines
macros not permitted in those headers for those standards.  UNIX98
allows them to include inttypes.h, but (predating C99) has restricted
inttypes.h contents (not yet tested in the conform/ tests) not
including those macros; XPG4.2 has no such permission and no
inttypes.h / stdint.h at all.

This patch rearranges the headers to avoid this issue.  intN_t
definitions move to bits/stdint-intn.h, and uintN_t definitions to
bits/stdint-uintn.h.  (These are not bits/types/ headers because they
each define four types.  They are separate rather than just a single
header because sys/types.h defines intN_t but u_intN_t rather than
uintN_t - and while sys/types.h could define uintN_t because of the
POSIX reservation of *_t, existing practice there is largely to
condition types on appropriate feature test macros, and indeed there
is at least one open bug report (14553) about a type that's not
so-conditioned, so maybe types there should actually have conditions
added where appropriate.)  The affected network headers are then made
to include bits/stdint-uintn.h instead of stdint.h.  This allows six
XFAILs to be removed.

This doesn't do anything about inttypes.h defining more than it should
for UNIX98, but we don't have conformtest expectations for that case
at present (and my inclination is that a fix for that should be as
local as possible - affecting only inttypes.h, not stdint.h, only for
the case of __USE_UNIX98 && !__USE_ISOC99).

Tested for x86_64.

	[BZ #21455]
	* bits/stdint-intn.h: New file.
	* bits/stdint-uintn.h: Likewise.
	* stdlib/Makefile (headers): Add bits/stdint-intn.h and
	bits/stdint-uintn.h.
	* inet/netinet/in.h: Include <bits/stdint-uintn.h> instead of
	<stdint.h>.
	* posix/sys/types.h: Include <bits/stdint-intn.h>.
	(__int8_t_defined): Do not define here.
	(int8_t): Likewise.
	(int16_t): Likewise.
	(int32_t): Likewise.
	(int64_t): Likewise.
	[__GNUC_PREREQ (2, 7)] (__intN_t): Likewise.
	* resolv/netdb.h: Include <bits/stdint-uintn.h> instead of
	<stdint.h>.
	* include/netdb.h [_ISOMAC]: Do not include <stdint.h>.
	* sysdeps/generic/stdint.h: Include <bits/stdint-intn.h> and
	<bits/stdint-uintn.h>.
	(int8_t): Do not define here.
	(int16_t): Likewise.
	(int32_t): Likewise.
	(int64_t): Likewise.
	(uint8_t): Likewise.
	(uint16_t): Likewise.
	(uint32_t): Likewise.
	(uint64_t): Likewise.
	* conform/Makefile (test-xfail-XPG42/arpa/inet.h/conform): Remove
	variable.
	(test-xfail-XPG42/netdb.h/conform): Likewise.
	(test-xfail-XPG42/netinet/in.h/conform): Likewise.
	(test-xfail-UNIX98/arpa/inet.h/conform): Likewise.
	(test-xfail-UNIX98/netdb.h/conform): Likewise.
	(test-xfail-UNIX98/netinet/in.h/conform): Likewise.
2017-05-04 20:36:42 +00:00
Paul E. Murphy
593bf7189a ldbl-128: Use mathx_hidden_def inplace of hidden_def
This provides a extra macro expansion before invoking
the hidden_def macro.  This is necessary to build the
ldbl-128 files as float128 correctly.

	* sysdeps/generic/math_private.h:
	(mathx_hidden_def): New macro.
	* sysdeps/ieee754/ldbl-128/s_finitel.c: Replace hidden_def with
	the above.
	* sysdeps/ieee754/ldbl-128/s_isinfl.c: Likewise.
	* sysdeps/ieee754/ldbl-128/s_isnanl.c: Likewise.
2017-05-04 14:47:27 -03:00
Joseph Myers
fd9f30f6f9 Use __glibc_reserved convention in mcontext, sigcontext (bug 21457).
This patch implements the most straightforward part of fixing
namespace issues for sys/ucontext.h and related headers: where fields
in sys/ucontext.h or bits/sigcontext.h are named "reserved", "padding"
or similar, they are renamed to use the __glibc_reserved* naming
convention.  It does not change fields with a leading underscore, or
even those with a prefix such as uc_ or sc_.  It only fixes a small
part of bug 21457, so no XFAILs are removed.

Tested for x86_64 and x86, and with build-many-glibcs.py.

	[BZ #21457]
	* sysdeps/unix/sysv/linux/mips/sys/ucontext.h
	[_MIPS_SIM != _ABIO32] (mcontext_t): Rename field reserved to
	__glibc_reserved1.
	* sysdeps/unix/sysv/linux/x86/bits/sigcontext.h
	(struct _fpx_sw_bytes): Rename field padding to __glibc_reserved1.
	(struct _fpxreg): Likewise.
	[!__x86_64__] (struct _fpstate): Rename field reserved to
	__glibc_reserved1.  Rename field padding to __glibc_reserved2.
	[__x86_64__] (struct _fpstate): Rename field padding to
	__glibc_reserved1.
	(struct _xsave_hdr): Rename field reserved1 to __glibc_reserved1.
	Rename field reserved2 to __glibc_reserved2.
	* sysdeps/unix/sysv/linux/x86/sys/ucontext.h
	[__x86_64__] (struct _libc_fpxreg): Rename field padding to
	__glibc_reserved1.
	[__x86_64__] (struct _libc_fpstate): Rename field padding to
	__glibc_reserved1.
2017-05-04 11:26:33 +00:00
Rajalakshmi Srinivasaraghavan
808befbf9e powerpc: Fix strncat ifunc selection
Correct hwcap usage in strncat introduced by commit
249dcdb71b.
Tested on power7 and power8 systems
2017-05-04 12:35:56 +05:30
H.J. Lu
1432d38ea0 x86: Set dl_platform and dl_hwcap from CPU features [BZ #21391]
dl_platform and dl_hwcap are set from AT_PLATFORM and AT_HWCAP very
early during startup.  They are used by dynamic linker to determine
platform and build an array of hardware capability names, which are
added to search path when loading shared object.  dl_platform and
dl_hwcap are unused on x86-64.  On i386, i386, i486, i586 and i686
platforms were supported and only SSE2 capability was used.

On x86, usage of AT_PLATFORM and AT_HWCAP to determine platform and
processor capabilities is obsolete since all information is available
in dl_x86_cpu_features.  This patch sets dl_platform and dl_hwcap from
dl_x86_cpu_features in dynamic linker.  On i386, the available plaforms
are changed to i586 and i686 since i386 has been deprecated.  On x86-64,
the available plaforms are haswell, which is for Haswell class processors
with BMI1, BMI2, LZCNT, MOVBE, POPCNT, AVX2 and FMA, and xeon_phi, which
is for Xeon Phi class processors with AVX512F, AVX512CD, AVX512ER and
AVX512PF.  A capability, avx512_1, is also added to x86-64 for AVX512
ISAs: AVX512F, AVX512CD, AVX512BW, AVX512DQ and AVX512VL.

	[BZ #21391]
	* sysdeps/i386/dl-machine.h (dl_platform_init) [IS_IN (rtld)]:
	Only call init_cpu_features.
	[!IS_IN (rtld)]: Only set GLRO(dl_platform) to NULL if needed.
	* sysdeps/x86_64/dl-machine.h (dl_platform_init): Likewise.
	* sysdeps/i386/dl-procinfo.h: Removed.
	* sysdeps/unix/sysv/linux/i386/dl-procinfo.h: Don't include
	<sysdeps/i386/dl-procinfo.h> nor <ldsodefs.h>.  Include
	<sysdeps/x86/dl-procinfo.h>.
	(_dl_procinfo): Replace _DL_HWCAP_COUNT with 32.
	* sysdeps/unix/sysv/linux/x86_64/dl-procinfo.h [!IS_IN (ldconfig)]:
	Include <sysdeps/x86/dl-procinfo.h> instead of
	 <sysdeps/generic/dl-procinfo.h>.
	* sysdeps/x86/cpu-features.c: Include <dl-hwcap.h>.
	(init_cpu_features): Set dl_platform, dl_hwcap and dl_hwcap_mask.
	* sysdeps/x86/cpu-features.h (bit_cpu_LZCNT): New.
	(bit_cpu_MOVBE): Likewise.
	(bit_cpu_BMI1): Likewise.
	(bit_cpu_BMI2): Likewise.
	(index_cpu_BMI1): Likewise.
	(index_cpu_BMI2): Likewise.
	(index_cpu_LZCNT): Likewise.
	(index_cpu_MOVBE): Likewise.
	(index_cpu_POPCNT): Likewise.
	(reg_BMI1): Likewise.
	(reg_BMI2): Likewise.
	(reg_LZCNT): Likewise.
	(reg_MOVBE): Likewise.
	(reg_POPCNT): Likewise.
	* sysdeps/x86/dl-hwcap.h: New file.
	* sysdeps/x86/dl-procinfo.h: Likewise.
	* sysdeps/x86/dl-procinfo.c (_dl_x86_hwcap_flags): New.
	(_dl_x86_platforms): Likewise.
2017-05-03 13:44:35 -07:00
Adhemerval Zanella
b62c381591 Consolidate Linux epoll_wait syscall
This patch consolidates the epoll_wait Linux syscall generation on
sysdeps/unix/sysv/linux/epoll_wait.c.  The implementation tries to
use __NR_epoll_wait if defined, otherwise calls epoll_pwait.

Checked on i686-linux-gnu, x86_64-linux-gnu, x86_64-linux-gnux32,
arch64-linux-gnu, arm-linux-gnueabihf, and powerpc64le-linux-gnu.

	* sysdeps/unix/sysv/linux/epoll_wait.c: New file.
	* sysdeps/unix/sysv/linux/generic/epoll_wait.c: Remove file.
	* sysdeps/unix/sysv/linux/syscalls.list: Remove epoll_wait from
	auto-generation list.
2017-05-03 10:36:43 -03:00
Adhemerval Zanella
26f28fd73d Consolidate Linux select implementation
This patch consolidates the select Linux syscall implementation on
sysdeps/unix/sysv/linux/select.c.  The changes are:

  1. Remove select from auto-generation syscalls.list on the architecture
     that uses __NR_select.
  2. Remove generic implementation add a default one that handle all
     current cases (with the expection of alpha)
     The new default implementation will either use __NR_select if
     available of fallback to __NR_pselect6 otherwise.
  3. Add a alpha outlier implementation which requires old compatibility
     symbols.

Checked on i686-linux-gnu, x86_64-linux-gnu, x86_64-linux-gnux32,
arch64-linux-gnu, arm-linux-gnueabihf, and powerpc64le-linux-gnu.

	* sysdeps/unix/sysv/linux/alpha/Makefile (sysdep_routines): Add
	osf_select.
	* sysdeps/unix/sysv/linux/alpha/select.c: New file.
	* sysdeps/unix/sysv/linux/alpha/syscalls.list: Remove select and
	osf_select from auto-generation list.
	* sysdeps/unix/sysv/linux/syscalls.list: Likewise.
	* sysdeps/unix/sysv/linux/sparc/sparc64/syscalls.list: Likewise.
	* sysdeps/unix/sysv/linux/generic/select.c: Remove file.
	* sysdeps/unix/sysv/linux/select.c: New file.
2017-05-03 10:36:36 -03:00
Adhemerval Zanella
b5eede9730 Consolidate Linux poll implementation
This patch consolidates the poll Linux syscall implementation on
sysdeps/unix/sysv/linux/poll.c.  It basically removes poll from
auto-generation list and add a default implementation that either
call __NR_poll directly (if the kernel headers defines it) or
ppoll adjusting the timeout argument (as the generic implementation).

Checked on i686-linux-gnu, x86_64-linux-gnu, x86_64-linux-gnux32,
aarch64-linux-gnu, arm-linux-gnueabihf, and powerpc64le-linux-gnu.

	* sysdeps/unix/sysv/linux/generic/poll.c: Remove file.
	* sysdeps/unix/sysv/linux/poll.c: New file.
	* sysdeps/unix/sysv/linux/syscalls.list: Remove poll from
	auto-generation list.
2017-05-03 10:36:01 -03:00