Since there are no multiarch versions of memmove_chk and memset_chk,
test multiarch versions of memmove_chk and memset_chk only in libc.so.
[BZ #21741]
* sysdeps/x86_64/multiarch/ifunc-impl-list.c
(__libc_ifunc_impl_list): Test memmove_chk and memset_chk only
in libc.so.
This patch increases the timeouts for some tests that I've seen timing
out on slow systems in my 2.26 release testing. (In the case of
tst-tsearch.c, increasing the timeout means removing a setting of 10
that was put there before the default timeout was increased to 20
seconds, so putting the default into effect.)
* iconvdata/tst-loading.c (TIMEOUT): Define to 30.
* misc/tst-tsearch.c (TIMEOUT): Remove.
* nptl/tst-create-detached.c (TIMEOUT): Define to 100.
* nptl/tst-robust-fork.c (TIMEOUT): Likewise.
* nptl/tst-rwlock19.c (TIMEOUT): Likewise.
* string/tst-cmp.c (TIMEOUT): Define to 600.
This patch fixes some build issues when including types/sigevent_t.h
along with bits/pthreadtypes.h.
Checked on x86_64-linux-gnu and on a build on supported major ABIs.
[BZ #21715]
* sysdeps/nptl/bits/pthreadtypes.h (__have_pthread_attr_t): Fix typo
on definition.
This change forces realignment of the stack pointer in __tls_get_addr, so
that binaries compiled by GCCs older than GCC 4.9:
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=58066
continue to work even if vector instructions are used in glibc which
require the ABI stack realignment.
__tls_get_addr_slow is added to handle the slow paths in the default
implementation of__tls_get_addr in elf/dl-tls.c. The new __tls_get_addr
calls __tls_get_addr_slow after realigning the stack. Internal calls
within ld.so go directly to the default implementation of __tls_get_addr
because they do not need stack realignment.
[BZ #21609]
* sysdeps/x86_64/Makefile (sysdep-dl-routines): Add tls_get_addr.
(gen-as-const-headers): Add rtld-offsets.sym.
* sysdeps/x86_64/dl-tls.c: New file.
* sysdeps/x86_64/rtld-offsets.sym: Likwise.
* sysdeps/x86_64/tls_get_addr.S: Likewise.
* sysdeps/x86_64/dl-tls.h: Add multiple inclusion guards.
* sysdeps/x86_64/tlsdesc.sym (TI_MODULE_OFFSET): New.
(TI_OFFSET_OFFSET): Likwise.
This patch fix the return value for error conditions for default
posix_spawn (where the errno is expected). It also avoid clobber
errno on fork call.
Checked on x86_64 (with Linux implementation removed).
[BZ# 21697]
* sysdeps/posix/spawni.c (__spawni_child): Fix return value.
(__spawnix): Do not clober errno.
Locking overhead can be significant in some stdio operations
that are common in single threaded applications.
This patch adds the _IO_FLAGS2_NEED_LOCK flag to indicate if
an _IO_FILE object needs to be locked and some of the stdio
functions just jump to their _unlocked variant when not. The
flag is set on all _IO_FILE objects when the first thread is
created. A new GLIBC_PRIVATE libc symbol, _IO_enable_locks,
was added to do this from libpthread.
The optimization can be applied to more stdio functions,
currently it is only applied to single flag check or single
non-wide-char standard operations. The flag should probably
be never set for files with _IO_USER_LOCK, but that's just a
further optimization, not a correctness requirement.
The optimization is valid in a single thread because stdio
operations are non-as-safe (so lock state is not observable
from a signal handler) and stdio locks are recursive (so lock
state is not observable via deadlock). The optimization is not
valid if a thread may be created while an stdio lock is taken
and thus it should be disabled if any user code may run during
an stdio operation (interposed malloc, printf hooks, etc).
This makes the optimization more complicated for some stdio
operations (e.g. printf), but those are bigger and thus less
important to optimize so this patch does not try to do that.
* libio/libio.h (_IO_FLAGS2_NEED_LOCK, _IO_need_lock): Define.
* libio/libioP.h (_IO_enable_locks): Declare.
* libio/Versions (_IO_enable_locks): New symbol.
* libio/genops.c (_IO_enable_locks): Define.
(_IO_old_init): Initialize flags2.
* libio/feof.c.c (_IO_feof): Avoid locking when not needed.
* libio/ferror.c (_IO_ferror): Likewise.
* libio/fputc.c (fputc): Likewise.
* libio/putc.c (_IO_putc): Likewise.
* libio/getc.c (_IO_getc): Likewise.
* libio/getchar.c (getchar): Likewise.
* libio/ioungetc.c (_IO_ungetc): Likewise.
* nptl/pthread_create.c (__pthread_create_2_1): Enable stdio locks.
* libio/iofopncook.c (_IO_fopencookie): Enable locking for the file.
* sysdeps/pthread/flockfile.c (__flockfile): Likewise.
A dot-less host name without an /etc/resolv.conf file caused an
assertion failure in update_from_conf because the function would not
deal correctly with the empty search list case.
Thanks to Andreas Schwab for debugging assistence.
This patch updates build-many-glibcs.py to use the current release
branch of binutils and current releases of GMP and the Linux kernel.
* scripts/build-many-glibcs.py (Context.checkout): Default
binutils version to 2.29 branch, GMP version to 6.1.2 and Linux
kernel version to 4.12.
This commit enhances the stub resolver to reload the configuration
in the per-thread _res object if the /etc/resolv.conf file has
changed. The resolver checks whether the application has modified
_res and will not overwrite the _res object in that case.
The struct resolv_context mechanism is used to check the
configuration file only once per name lookup.
This commit adds the remaining unchanging members (which are loaded
from /etc/resolv.conf) to struct resolv_conf.
The extended name server list is currently not used by the stub
resolver. The switch depends on a cleanup: The _u._ext.nssocks
array stores just a single socket, and needs to be replaced with
a single socket value.
(The compatibility gethostname implementation does not use the
extended addres sort list, either. Updating the compat code is
not worthwhile.)
This change uses the extended resolver state in struct resolv_conf to
store the search list. If applications have not patched the _res
object directly, this extended search list will be used by the stub
resolver during name resolution.
This change provides additional resolver configuration state which
is not exposed through the _res ABI. It reuses the existing
initstamp field in the supposedly-private part of _res. Some effort
is undertaken to avoid memory safety issues introduced by applications
which directly patch the _res object.
With this commit, only the initstamp field is moved into struct
resolv_conf. Additional members will be added later, eventually
migrating the entire resolver configuration.
struct resolv_context objects provide a temporary resolver context
which does not change during a name lookup operation. Only when the
outmost context is created, the stub resolver configuration is
verified to be current (at present, only against previous res_init
calls). Subsequent attempts to obtain the context will reuse the
result of the initial verification operation.
struct resolv_context can also be extended in the future to store
data which needs to be deallocated during thread cancellation.
posix/sched_cpucount.c assumes that size of __cpu_mask == size of long,
which is incorrect for x32. This patch uses __builtin_popcount, which
is availabe in GCC 4.9, in posix/sched_cpucount.c.
Tested on i686, x86-64 and x32 with multi-arch disabled.
[BZ #21696]
* posix/sched_cpucount.c: Don't include <limits.h>.
(__sched_cpucount): Use __builtin_popcount.
In math/math.h, __MATH_TG will expand signbit to __builtin_signbit*,
e.g.: __builtin_signbitf128, before GCC 6. However, there has never
been a __builtin_signbitf128 in GCC and the type-generic builtin is
only available since GCC 6. For older GCC, this patch defines
__builtin_signbitf128 to __signbitf128, so that the internal function
is used instead of the non-existent builtin.
This patch also changes the implementation of __signbitf128, because
it was reusing the implementation of __signbitl from ldbl-128, which
calls __builtin_signbitl. Using the long double version of the
builtin is not correct on machines where _Float128 is ABI-distinct
from long double (i.e.: ia64, powerpc64le, x86, x86_84). The new
implementation does not rely on builtins when being built with GCC
versions older than 6.0.
The new code does not currently affect powerpc64le builds, because
only GCC 6.2 fulfills the requirements from configure. It might
affect powerpc64le builds if those requirements are backported to
older versions of the compiler. The new code affects x86_64 builds,
since glibc is supposed to build correctly with older versions of GCC.
Tested for powerpc64le and x86_64.
* include/math.h (__signbitf128): Define as hidden.
* sysdeps/ieee754/float128/s_signbitf128.c (__signbitf128):
Reimplement without builtins.
* sysdeps/ia64/bits/floatn.h [!__GNUC_PREREQ (6, 0)]
(__builtin_signbitf128): Define to __signbitf128.
* sysdeps/powerpc/bits/floatn.h: Likewise.
* sysdeps/x86/bits/floatn.h: Likewise.
Add a new tunable (glibc.tune.cpu) to override CPU identification on
aarch64. This is useful in two cases: one where it is desirable to
pretend to be another CPU for purposes of testing or because routines
written for that CPU are beneficial for specific workloads and second
where the underlying kernel does not support emulation of MRS to get
the MIDR of the CPU.
* elf/dl-tunables.h (tunable_is_name): Move from...
* elf/dl-tunables.c (is_name): ... here.
(parse_tunables, __tunables_init): Adjust.
* manual/tunables.texi: Document glibc.tune.cpu.
* sysdeps/aarch64/dl-tunables.list: New file.
* sysdeps/unix/sysv/linux/aarch64/cpu-features.c (struct
cpu_list): New type.
(cpu_list): New list of CPU names and their MIDR.
(get_midr_from_mcpu): New function.
(init_cpu_features): Override MIDR if necessary.
The string function implementations implemented so far do not use any
instructions that may deviate from standard aarch64, so it is possible
for all routines to run on all armv8 hardware. Select all
implementations in the benchmarks and tests.
* sysdeps/aarch64/multiarch/ifunc-impl-list.c
(__libc_ifunc_impl_list): Unconditionally select thunderx
routine for testing.
GCC 7 changed the definition of max_align_t on i386:
https://gcc.gnu.org/git/?p=gcc.git;a=commitdiff;h=9b5c49ef97e63cc63f1ffa13baf771368105ebe2
As a result, glibc malloc no longer returns memory blocks which are as
aligned as max_align_t requires.
This causes malloc/tst-malloc-thread-fail to fail with an error like this
one:
error: allocation function 0, size 144 not aligned to 16
This patch moves the MALLOC_ALIGNMENT definition to <malloc-alignment.h>
and increases the malloc alignment to 16 for i386.
[BZ #21120]
* malloc/malloc-internal.h (MALLOC_ALIGNMENT): Moved to ...
* sysdeps/generic/malloc-alignment.h: Here. New file.
* sysdeps/i386/malloc-alignment.h: Likewise.
* sysdeps/generic/malloc-machine.h: Include <malloc-alignment.h>.
This patch improves the default posix implementation of posix_spawn{p}
and align with Linux one. The main idea is to fix some issues already
fixed in Linux code, and deprecated vfork internal usage (source of
various bug reports). In a short:
- It moves POSIX_SPAWN_USEVFORK usage and sets it a no-op. Since
the process that actually spawn the new process do not share
memory with parent (with vfork), it fixes BZ#14750 for this
implementation.
- It uses a pipe to correctly obtain the return upon failure
of execution (BZ#18433).
- It correctly enable/disable asynchronous cancellation (checked
on ptl/tst-exec5.c).
- It correctly disable/enable signal handling.
Using this version instead of Linux shows only one regression,
posix/tst-spawn3, because of pipe2 usage which increase total
number of file descriptor.
* sysdeps/posix/spawni.c (__spawni_child): New function.
(__spawni): Rename to __spawnix.
This patch adds tgmath.h support for _Float128, so eliminating the
awkward caveat in NEWS about the type not being supported there. This
does inevitably increase the size of macro expansions (which grows
particularly fast when you have nested calls to tgmath.h macros), but
only when _Float128 is supported and the declarations of _Float128
interfaces are visible; otherwise the expansions are unchanged.
Tested for x86_64 and arm.
* math/tgmath.h: Include <bits/libc-header-start.h> and
<bits/floatn.h>.
(__TGMATH_F128): New macro.
(__TGMATH_CF128): Likewise.
(__TGMATH_UNARY_REAL_ONLY): Use __TGMATH_F128.
(__TGMATH_UNARY_REAL_RET_ONLY): Likewise.
(__TGMATH_BINARY_FIRST_REAL_ONLY): Likewise.
(__TGMATH_BINARY_FIRST_REAL_STD_ONLY): New macro.
(__TGMATH_BINARY_REAL_ONLY): Use __TGMATH_F128.
(__TGMATH_BINARY_REAL_STD_ONLY): New macro.
(__TGMATH_BINARY_REAL_RET_ONLY): Use __TGMATH_F128.
(__TGMATH_TERNARY_FIRST_SECOND_REAL_ONLY): Likewise.
(__TGMATH_TERNARY_REAL_ONLY): Likewise.
(__TGMATH_TERNARY_FIRST_REAL_RET_ONLY): Likewise.
(__TGMATH_UNARY_REAL_IMAG): Use __TGMATH_CF128.
(__TGMATH_UNARY_IMAG): Use __TGMATH_F128.
(__TGMATH_UNARY_REAL_IMAG_RET_REAL): Use __TGMATH_CF128.
(__TGMATH_BINARY_REAL_IMAG): Likewise.
(nexttoward): Use __TGMATH_BINARY_FIRST_REAL_STD_ONLY.
[__USE_MISC] (scalb): Use __TGMATH_BINARY_REAL_STD_ONLY.
* math/gen-tgmath-tests.py (Type.init_types): Enable _FloatN and
_FloatNx types if the corresponding HUGE_VAL macros are defined.
As a GNU extension, for _GNU_SOURCE glibc's complex.h provides a
clog10 function and tgmath.h supports complex arguments to the log10
macro. However, tgmath.h uses __clog10 not clog10 in defining the
macro.
There is no namespace reason (ignoring the block-scope namespace
issues that would apply equally to *every* function called by tgmath.h
macros) for using __clog10 here, since this is only for _GNU_SOURCE so
clog10 is always visible when this macro definition is used.
Furthermore, __clog10f128 is not exported, so supporting _Float128 in
tgmath.h implies using clog10 not __clog10 there. (__clog10 and
clog10 aren't used in libstdc++ either, although that library would
have a good case for using the __clog10 reserved-namespace export: the
standard C++ library includes log10 of a complex number.) This patch
duly changes the header to use clog10, and enables tests of the macro
for complex arguments.
Tested for x86_64.
* math/tgmath.h [__USE_GNU] (log10): Use clog10 not __clog10.
* math/gen-tgmath-tests.py (Tests.add_all_tests): Test log10 for
complex arguments.
The tgmath.h totalorder and totalordermag macros wrongly return a
floating-point type. They should return int, like the underlying
functions. This patch fixes them accordingly, updating tests
including enabling tests of those functions from gen-tgmath-tests.py.
Tested for x86_64.
[BZ #21687]
* math/tgmath.h (__TGMATH_BINARY_REAL_RET_ONLY): New macro.
(totalorder): Use it.
(totalordermag): Likewise.
* math/gen-tgmath-tests.py (Tests.add_all_tests): Enable tests of
totalorder and totalordermag.
* math/test-tgmath.c (F(compile_test)): Do not call totalorder or
totalordermag in arguments of calls to those functions.
(NCALLS): Change to 134.
The tgmath.h macros for function with integer return types generate
unnecessary casts to the return type. Since in those cases the return
type does not depend on the argument type, all the cases in the
conditional expressions already have the right type, and no casts are
needed; this patch removes them.
Tested for x86_64.
* math/tgmath.h (__TGMATH_UNARY_REAL_RET_ONLY): Do not take or
cast to return type argument.
(__TGMATH_TERNARY_FIRST_REAL_RET_ONLY): Likewise.
(lrint): Update call to __TGMATH_UNARY_REAL_RET_ONLY.
(llrint): Likewise.
(lround): Likewise.
(llround): Likewise.
(ilogb): Likewise.
(llogb): Likewise.
(fromfp): Update call to __TGMATH_TERNARY_FIRST_REAL_RET_ONLY.
(ufromfp): Likewise.
(fromfpx): Likewise.
(ufromfpx): Likewise.
As noted in bug 21607, NO_LONG_DOUBLE conditionals in libm tests are
no longer effective. For most this is harmless - they were only
present because of long double functions not being declared with _LIBC
defined, and _LIBC is no longer defined for building most tests. For
the few where this is actually relevant to the test, testing
LDBL_MANT_DIG > DBL_MANT_DIG is more appropriate as that limits the
test to public APIs. This patch fixes the tests accordingly.
Tested for x86_64 and arm.
[BZ #21607]
* math/basic-test.c [!NO_LONG_DOUBLE]: Change conditionals to
[LDBL_MANT_DIG > DBL_MANT_DIG].
* math/bug-nextafter.c [!NO_LONG_DOUBLE]: Remove conditionals.
* math/bug-nexttoward.c [!NO_LONG_DOUBLE]: Likewise.
* math/test-math-isinff.cc [!NO_LONG_DOUBLE]: Likewise.
* math/test-math-iszero.cc [!NO_LONG_DOUBLE]: Likewise.
* math/test-nan-overflow.c [!NO_LONG_DOUBLE]: Likewise.
* math/test-nan-payload.c [!NO_LONG_DOUBLE]: Likewise.
* math/test-nearbyint-except-2.c [!NO_LONG_DOUBLE]: Likewise.
* math/test-nearbyint-except.c [!NO_LONG_DOUBLE]: Likewise.
* math/test-powl.c [!NO_LONG_DOUBLE]: Likewise.
* math/test-signgam-finite-c99.c [!NO_LONG_DOUBLE]: Likewise.
* math/test-signgam-finite.c [!NO_LONG_DOUBLE]: Likewise.
* math/test-signgam-main.c [!NO_LONG_DOUBLE]: Likewise.
* math/test-snan.c [!NO_LONG_DOUBLE]: Likewise.
* math/test-tgmath-ret.c [!NO_LONG_DOUBLE]: Likewise.
* math/test-tgmath.c: Include <float.h>.
[!NO_LONG_DOUBLE]: Change conditionals to [LDBL_MANT_DIG >
DBL_MANT_DIG].
* math/test-tgmath2.c: Include <float.h>.
[!NO_LONG_DOUBLE]: Change conditionals to [LDBL_MANT_DIG >
DBL_MANT_DIG].
This patch adds a more thorough test of tgmath.h macros, verifying
both the return type and the function called for all the cases of
valid argument types. (Cases with current problems - I've just filed
four bugs - are disabled or omitted pending fixing those problems.)
The test uses a Python generator (works with both Python 2 and 3) to
generate a C file which is then built and run as a test in the usual
way (and that C file includes its own dummy definitions of libm
functions similar to existing tgmath.h tests). The motivation is to
make it easier to add tests of tgmath.h for _Float128 when adding
tgmath.h support for that type; the _FloatN / _FloatNx support is
present in the script, but disabled until the tgmath.h support is
written.
Tested for x86_64, and for arm to check things in the long double =
double case. (In that case, it's OK to call either double or long
double functions when the selected type is double or long double, as
long as the return type of the macro is exactly correct.)
* math/gen-tgmath-tests.py: New file.
* math/Makefile [PYTHON] (tests): Add test-tgmath3.
[PYTHON] (generated): Add test-tgmath3.c.
[PYTHON] (CFLAGS-test-tgmath3.c): New variable.
[PYTHON] ($(objpfx)test-tgmath3.c): New rule.
This patch implements a requirement of binutils >= 2.25 (up from 2.22)
to build glibc. Tests for 2.24 or later on x86_64 and s390 are
removed. It was already the case, as indicated by buildbot results,
that 2.24 was too old for building tests for 32-bit x86 (produced
internal linker errors linking elf/tst-gnu2-tls1mod.so). I don't know
if any configure tests for binutils features are obsolete given the
increased version requirement.
Tested for x86_64.
* configure.ac (AS): Require binutils 2.25 or later.
(LD): Likewise.
* configure: Regenerated.
* sysdeps/s390/configure.ac (AS): Remove version check.
* sysdeps/s390/configure: Regenerated.
* sysdeps/x86_64/configure.ac (AS): Remove version check.
* sysdeps/x86_64/configure: Regenerated.
* manual/install.texi (Tools for Compilation): Document
requirement for binutils 2.25 or later.
* INSTALL: Regenerated.
This patch fixes various miscellaneous namespace issues in
sys/ucontext.h headers.
Some struct tags are removed where the structs also have *_t typedef
names, while other struct tags without such names are renamed to start
__; the changes are noted in NEWS as they can affect C++ name mangling
(although there seems to be little if any external use of these types,
at least based on checking codesearch.debian.net). For powerpc,
pointers to struct pt_regs (not defined in this header) are changed to
point to struct __ctx(pt_regs), so in the __USE_MISC case those struct
fields continue to point to the existing struct pt_regs type for
maximum compatibility, while when that's a namespace issue they point
to a struct __pt_regs type which is always an incomplete struct.
Tested for affected architectures with build-many-glibcs.py.
[BZ #21457]
* sysdeps/unix/sysv/linux/m68k/sys/ucontext.h (fpregset_t): Remove
struct tag.
* sysdeps/unix/sysv/linux/mips/sys/ucontext.h (fpregset_t):
Likewise.
* sysdeps/unix/sysv/linux/nios2/sys/ucontext.h (mcontext_t):
Likewise.
* sysdeps/unix/sysv/linux/powerpc/sys/ucontext.h (pt_regs):
Declare struct type with __ctx.
[__WORDSIZE != 32] (mcontext_t): Use __ctx with pt_regs struct
tag.
(ucontext_t) [__WORDSIZE == 32]: Use __ctx with pt_regs struct tag
and regs field name.
Building the testsuite with current GCC mainline fails with:
loadtest.c: In function 'main':
loadtest.c:76:3: error: macro expands to multiple statements [-Werror=multistatement-macros]
for (map = MAPS; map != NULL; map = map->l_next) \
^
loadtest.c:165:2: note: in expansion of macro 'OUT'
OUT;
^~~
loadtest.c:164:7: note: some parts of macro expansion are not guarded by this 'if' clause
if (debug)
^~
This seems like a genuine bug, although fairly harmless; it means the
fflush call in the OUT macro is unconditional instead of being inside
the conditional as presumably intended. This patch makes this macro
use do { } while (0) to avoid the problem.
Tested for x86_64 (testsuite), and with build-many-glibcs.py for
aarch64-linux-gnu with GCC mainline.
* elf/loadtest.c (OUT): Define using do { } while (0).
Building with current GCC mainline fails with:
strftime_l.c: In function '__strftime_internal':
strftime_l.c:719:4: error: macro expands to multiple statements [-Werror=multistatement-macros]
digits = d > width ? d : width; \
^
strftime_l.c:1260:6: note: in expansion of macro 'DO_NUMBER'
DO_NUMBER (1, tp->tm_year + TM_YEAR_BASE);
^~~~~~~~~
strftime_l.c:1259:4: note: some parts of macro expansion are not guarded by this 'else' clause
else
^~~~
In fact this particular instance is harmless; the code looks like:
if (modifier == L_('O'))
goto bad_format;
else
DO_NUMBER (1, tp->tm_year + TM_YEAR_BASE);
and because of the goto, it doesn't matter that part of the expansion
isn't under the "else" conditional. But it's also clearly bad style
to rely on that. This patch changes DO_NUMBER and DO_NUMBER_SPACEPAD
to use do { } while (0) to avoid such problems.
Tested (full testsuite) for x86_64 (GCC 6), and with
build-many-glibcs.py with GCC mainline, in conjunction with my libgcc
patch <https://gcc.gnu.org/ml/gcc-patches/2017-06/msg02032.html>.
* time/strftime_l.c (DO_NUMBER): Define using do { } while (0).
(DO_NUMBER_SPACEPAD): Likewise.
This patch provides an optimised implementation of memchr using NEON
instructions to improve its performance, especially with longer search regions.
This gave an improvement in performance against the Thumb2+DSP optimised code,
with more significant gains for larger inputs. The NEON code also wins in cases
where the input is small (less than 8 bytes) by defaulting to a simple
byte-by-byte search. This avoids the overhead imposed by filling two quadword
registers from memory.
* sysdeps/arm/armv7/multiarch/Makefile: Add memchr_neon to
sysdep_routines.
* sysdeps/arm/armv7/multiarch/ifunc-impl-list.c: Add define for
__memchr_neon.
Add ifunc definitions for __memchr_neon and __memchr_noneon.
* sysdeps/arm/armv7/multiarch/memchr.S: New file.
* sysdeps/arm/armv7/multiarch/memchr_impl.S: Likewise.
* sysdeps/arm/armv7/multiarch/memchr_neon.S: Likewise.
Testing done: Ran regression tests for arm-none-linux-gnueabihf as well as a
full toolchain bootstrap. Benchmark tests were ran on ARMv7-A and ARMv8-A
hardware targets.
This patch adds an ifunc variant to use the cu instruction on arch12 CPUs.
This new ifunc variant can be built if binutils support z13 vector
instructions. At runtime, HWCAP_S390_VXE decides if we can use the
cu21 instruction.
ChangeLog:
* sysdeps/s390/utf8-utf16-z9.c (__to_utf8_loop_vx_cu):
Use vector and cu21 instruction.
* sysdeps/s390/multiarch/utf8-utf16-z9.c:
Add __to_utf8_loop_vx_cu in ifunc resolver.
This patch adds an ifunc variant to use the cu instruction on arch12 CPUs.
This new ifunc variant can be built if binutils support z13 vector
instructions. At runtime, HWCAP_S390_VXE decides if we can use the
cu24 instruction.
ChangeLog:
* sysdeps/s390/utf16-utf32-z9.c (__from_utf16_loop_vx_cu):
Use vector and cu24 instruction.
* sysdeps/s390/multiarch/utf16-utf32-z9.c:
Add __from_utf16_loop_vx_cu in ifunc resolver.
This patch adds an ifunc variant to use the cu instruction on arch12 CPUs.
This new ifunc variant can be built if binutils support z13 vector
instructions. At runtime, HWCAP_S390_VXE decides if we can use the
cu42 instruction.
ChangeLog:
* sysdeps/s390/utf16-utf32-z9.c (__to_utf16_loop_vx_cu):
Use vector and cu42 instruction.
* sysdeps/s390/multiarch/utf16-utf32-z9.c:
Add __to_utf16_loop_vx_cu in ifunc resolver.
This patch adds an ifunc variant to use the cu instruction on arch12 CPUs.
This new ifunc variant can be built if binutils support z13 vector
instructions. At runtime, HWCAP_S390_VXE decides if we can use the
cu41 instruction.
ChangeLog:
* sysdeps/s390/utf8-utf32-z9.c (__to_utf8_loop_vx_cu):
Use vector and cu41 instruction.
* sysdeps/s390/multiarch/utf8-utf32-z9.c: Add __to_utf8_loop_vx_cu
in ifunc resolver.
The new hwcap values indicate support for:
- Vector packed decimal facility
- Vector enhancements facility 1
- Guarded storage facility
Note: arch12 is NOT the official name of the new CPU.
It refers to the edition number of the Principle of Operations manual.
ChangeLog:
* sysdeps/s390/dl-procinfo.c (_dl_s390_cap_flags):
Add vxd, vxe, gs flag.
* sysdeps/s390/dl-procinfo.h: Add HWCAP_S390_VXD, HWCAP_S390_VXE,
HWCAP_S390_GS capability.
* sysdeps/unix/sysv/linux/s390/bits/hwcap.h
(HWCAP_S390_VXD, HWCAP_S390_VXE, HWCAP_S390_GS): Define.
Check the first 32 bytes before checking size when size >= 32 bytes
to avoid unnecessary branch if the difference is in the first 32 bytes.
Replace vpmovmskb/subl/jnz with vptest/jnc.
On Haswell, the new version is as fast as the previous one. On Skylake,
the new version is a little bit faster.
* sysdeps/x86_64/multiarch/memcmp-avx2-movbe.S (MEMCMP): Check
the first 32 bytes before checking size when size >= 32 bytes.
Replace vpmovmskb/subl/jnz with vptest/jnc.
The testcases elf/tst-tls1-static and elf/tst-tls2-static
are failing on s390 if gcc is configured with --enable-default-pie.
According to binutils, there are only four valid cases for IE -> LE transition:
lg %rx,(0,%ry) -> sllg %rx,%ry,0
lg %rx,(%ry,0) -> sllg %rx,%ry,0
lg %rx,(%ry,%r12) -> sllg %rx,%ry,0
lg %rx,(%r12,%ry) -> sllg %rx,%ry,0
Thus I've adjusted the TLS_IE macro in s390 tls-macros.h files in PIC case
in order to use r12 for the address of global offset table.
The adjusted macro definition is not used if gcc is not configured with
--enable-default-pie.
ChangeLog:
* sysdeps/s390/s390-32/tls-macros.h (TLS_IE): Use r12 for GOT address.
* sysdeps/s390/s390-64/tls-macros.h (TLS_IE): Likewise.
Some Linux kernels have very aggressive ICMP rate limiting on the
loopback interface. This commit introduces a minimal echoing DNS server
inside the network namespace, so that there is no need for ICMP error
messages anymore.
Many callers of __res_maybe_init also call _res_hconf_init.
Additional calls to the latter do not hurt because the function
does its work only once. (/etc/hosts.conf is not reloaded or
even checked for changes.) This means that we can simplify the
code by calling _res_hconf_init directly from __res_vinit.
This patch enables float128 support for ia64, so that all the
configurations where GCC supports _Float128 / __float128 as an
ABI-distinct type now have glibc support as well. bits/floatn.h
declares the support to be available for GCC 4.4 and later, which is
when the libgcc support was added. The removal of
sysdeps/ia64/fpu/k_rem_pio2.c is because the generic k_rem_pio2.c
defines a function required by the float128 code.
Tested (compilation only) with build-many-glibcs.py for ia64 (GCC 6
and GCC 7).
Given how long it is since libm-test-ulps has been updated for ia64, I
think truncating the file and regenerating it from scratch would be a
good idea when doing a regeneration to add float128 ulps. I expect
various ia64 libm issues (at least some already filed in Bugzilla) to
result in test failures even after ulps regeneration, but hopefully
the float128 code will pass tests as it's the same as used on other
architectures.
* sysdeps/ia64/Implies: Add ieee754/float128.
* sysdeps/ia64/bits/floatn.h: New file.
* sysdeps/ia64/float128-abi.h: Likewise.
* manual/math.texi (Mathematics): Document support for _Float128
on ia64.
* sysdeps/ia64/Makefile [$(subdir) = math] (CPPFLAGS): Append to
Makefile variable.
* sysdeps/ia64/fpu/e_sqrtf128.c: New file.
* sysdeps/ia64/fpu/k_rem_pio2.c: Remove file.
* sysdeps/ia64/fpu/sfp-machine.h: New file. Based on libgcc.
* sysdeps/ia64/math-tests.h: New file.
* math/libm-test-support.h (XFAIL_FLOAT128_PAYLOAD): Also define
based on TEST_COND_binary128 for [__ia64__].
* sysdeps/unix/sysv/linux/ia64/libc.abilist: Update.
* sysdeps/unix/sysv/linux/ia64/libm.abilist: Likewise,
The ucontext_t type has a tag struct ucontext. As with previous such
issues for siginfo_t and stack_t, this tag is not permitted by POSIX
(is not in a reserved namespace), and so namespace conformance means
breaking C++ name mangling for this type.
In this case, the type does need to have some tag rather than just a
typedef name, because it includes a pointer to itself. This patch
uses struct ucontext_t as the new tag, so the type is mangled as
ucontext_t (the POSIX *_t reservation applies in all namespaces, not
just the namespace of ordinary identifiers). Another reserved name
such as struct __ucontext could of course be used.
Because of other namespace issues, this patch does not by itself fix
bug 21457 or allow any XFAILs to be removed.
Tested for x86_64, and with build-many-glibcs.py.
[BZ #21457]
* sysdeps/arm/sys/ucontext.h (struct ucontext): Rename to struct
ucontext_t.
* sysdeps/generic/sys/ucontext.h (struct ucontext): Likewise.
* sysdeps/i386/sys/ucontext.h (struct ucontext): Likewise.
* sysdeps/m68k/sys/ucontext.h (struct ucontext): Likewise.
* sysdeps/mips/sys/ucontext.h (struct ucontext): Likewise.
* sysdeps/unix/sysv/linux/aarch64/sys/ucontext.h (struct
ucontext): Likewise.
* sysdeps/unix/sysv/linux/alpha/sys/ucontext.h (struct ucontext):
Likewise.
* sysdeps/unix/sysv/linux/arm/sys/ucontext.h (struct ucontext):
Likewise.
* sysdeps/unix/sysv/linux/hppa/sys/ucontext.h (struct ucontext):
Likewise.
* sysdeps/unix/sysv/linux/ia64/sys/ucontext.h (struct ucontext):
Likewise.
* sysdeps/unix/sysv/linux/m68k/sys/ucontext.h (struct ucontext):
Likewise.
* sysdeps/unix/sysv/linux/mips/sys/ucontext.h (struct ucontext):
Likewise.
* sysdeps/unix/sysv/linux/nios2/sys/ucontext.h (struct ucontext):
Likewise.
* sysdeps/unix/sysv/linux/powerpc/sys/ucontext.h (struct
ucontext): Likewise.
* sysdeps/unix/sysv/linux/s390/sys/ucontext.h (struct ucontext):
Likewise.
* sysdeps/unix/sysv/linux/sh/sys/ucontext.h (struct ucontext):
Likewise.
* sysdeps/unix/sysv/linux/sparc/sys/ucontext.h (struct ucontext):
Likewise.
* sysdeps/unix/sysv/linux/tile/sys/ucontext.h (struct ucontext):
Likewise.
* sysdeps/unix/sysv/linux/x86/sys/ucontext.h (struct ucontext):
Likewise.
* sysdeps/powerpc/powerpc32/backtrace.c (struct
rt_signal_frame_32): Likewise.
* sysdeps/powerpc/powerpc64/backtrace.c (struct signal_frame_64):
Likewise.
* sysdeps/unix/sysv/linux/aarch64/kernel_rt_sigframe.h (struct
kernel_rt_sigframe): Likewise.
* sysdeps/unix/sysv/linux/aarch64/sigcontextinfo.h (SIGCONTEXT):
Likewise.
* sysdeps/unix/sysv/linux/arm/register-dump.h (register_dump):
Likewise.
* sysdeps/unix/sysv/linux/arm/sigcontextinfo.h (SIGCONTEXT):
Likewise.
* sysdeps/unix/sysv/linux/hppa/profil-counter.h
(__profil_counter): Likewise.
* sysdeps/unix/sysv/linux/microblaze/sigcontextinfo.h
(SIGCONTEXT): Likewise.
* sysdeps/unix/sysv/linux/mips/kernel_rt_sigframe.h (struct
kernel_rt_sigframe): Likewise.
* sysdeps/unix/sysv/linux/nios2/kernel_rt_sigframe.h (struct
kernel_rt_sigframe): Likewise.
* sysdeps/unix/sysv/linux/nios2/sigcontextinfo.h (SIGCONTEXT):
Likewise.
* sysdeps/unix/sysv/linux/sh/makecontext.S (__makecontext):
Likewise.
* sysdeps/unix/sysv/linux/sparc/sparc64/makecontext.c
(__start_context): Likewise.
* sysdeps/unix/sysv/linux/tile/sigcontextinfo.h (SIGCONTEXT):
Likewise.
* sysdeps/unix/sysv/linux/x86_64/register-dump.h (register_dump):
Likewise.
* sysdeps/unix/sysv/linux/x86_64/sigcontextinfo.h (SIGCONTEXT):
Likewise.
This patch enables float128 support for x86_64 and x86. All GCC
versions that can build glibc provide the required support, but since
GCC 6 and before don't provide __builtin_nanq / __builtin_nansq, sNaN
tests and some tests of NaN payloads need to be disabled with such
compilers (this does not affect the generated glibc binaries at all,
just the tests). bits/floatn.h declares float128 support to be
available for GCC versions that provide the required libgcc support
(4.3 for x86_64, 4.4 for i386 GNU/Linux, 4.5 for i386 GNU/Hurd);
compilation-only support was present some time before then, but not
really useful without the libgcc functions.
fenv_private.h needed updating to avoid trying to put _Float128 values
in registers. I make no assertion of optimality of the
math_opt_barrier / math_force_eval definitions for this case; they are
simply intended to be sufficient to work correctly.
Tested for x86_64 and x86, with GCC 7 and GCC 6. (Testing for x32 was
compilation tests only with build-many-glibcs.py to verify the ABI
baseline updates. I have not done any testing for Hurd, although the
float128 support is enabled there as for GNU/Linux.)
* sysdeps/i386/Implies: Add ieee754/float128.
* sysdeps/x86_64/Implies: Likewise.
* sysdeps/x86/bits/floatn.h: New file.
* sysdeps/x86/float128-abi.h: Likewise.
* manual/math.texi (Mathematics): Document support for _Float128
on x86_64 and x86.
* sysdeps/i386/fpu/fenv_private.h: Include <bits/floatn.h>.
(math_opt_barrier): Do not put _Float128 values in floating-point
registers.
(math_force_eval): Likewise.
[__x86_64__] (SET_RESTORE_ROUNDF128): New macro.
* sysdeps/x86/fpu/Makefile [$(subdir) = math] (CPPFLAGS): Append
to Makefile variable.
* sysdeps/x86/fpu/e_sqrtf128.c: New file.
* sysdeps/x86/fpu/sfp-machine.h: Likewise. Based on libgcc.
* sysdeps/x86/math-tests.h: New file.
* math/libm-test-support.h (XFAIL_FLOAT128_PAYLOAD): New macro.
* math/libm-test-getpayload.inc (getpayload_test_data): Use
XFAIL_FLOAT128_PAYLOAD.
* math/libm-test-setpayload.inc (setpayload_test_data): Likewise.
* math/libm-test-totalorder.inc (totalorder_test_data): Likewise.
* math/libm-test-totalordermag.inc (totalordermag_test_data):
Likewise.
* sysdeps/unix/sysv/linux/i386/libc.abilist: Update.
* sysdeps/unix/sysv/linux/i386/libm.abilist: Likewise.
* sysdeps/unix/sysv/linux/x86_64/64/libc.abilist: Likewise.
* sysdeps/unix/sysv/linux/x86_64/64/libm.abilist: Likewise.
* sysdeps/unix/sysv/linux/x86_64/x32/libc.abilist: Likewise.
* sysdeps/unix/sysv/linux/x86_64/x32/libm.abilist: Likewise.
* sysdeps/i386/fpu/libm-test-ulps: Likewise.
* sysdeps/i386/i686/fpu/multiarch/libm-test-ulps: Likewise.
* sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
On aarch64, alpha, arm, hppa, mips, nios2, powerpc, sh, sparc, tile,
and x86_64 the clone syscall jumps to _exit after the child execution
and the function ends the process execution by calling exit_group.
This behavior have a small issue where threads created with
CLONE_THREAD using clone syscall directly will eventually exit the
whole group altogether instead of just the thread created. Also,
s390, microblaze, ia64, i386, and m68k differs by calling exit
syscall directly.
This patch changes all architectures to call the exit syscall
directly, as for s390, microblaze, ia64, i386, and m68k. This do not
have change glibc internal behavior in any sort, since the only
usage of clone implementation in posix_spawn calls _exit directly
in the created child (fork uses a direct call to clone).
Checked on x86_64-linux-gnu, i686-linux-gnu, aarch64-linux-gnu,
powerpc-linux-gnu, powerpc64le-linux-gnu, sparc64-linux-gnu,
and sparcv9-linux-gnu.
[BZ #21512]
* sysdeps/unix/sysv/linux/aarch64/clone.S (__clone): Call exit
syscall instead of jump to _exit.
(CLONE_VM_BIT): Remove unused define.
(CLONE_VM): Likewise.
(CLONE_THREAD_BIT): Likewise.
(CLONE_THREAD): Likewise.
* sysdeps/unix/sysv/linux/alpha/clone.S (__clone): Likewise.
(CLONE_VM): Remove unused define.
* sysdeps/unix/sysv/linux/arm/clone.S (__clone): Likewise.
(CLONE_VM): Remove unused define.
(CLONE_THREAD): Likewise.
* sysdeps/unix/sysv/linux/i386/clone.S (CLONE_VM): Likewise.
* sysdeps/unix/sysv/linux/ia64/clone2.S (__clone2): Call exit
syscall instead of jump to _exit.
* sysdeps/unix/sysv/linux/hppa/clone.S (__clone): Likewise.
* sysdeps/unix/sysv/linux/mips/clone.S (__clone): Likewise.
(CLONE_VM): Remove unused define.
(CLONE_THREAD): Likewise.
* sysdeps/unix/sysv/linux/nios2/clone.S (__clone): Likewise.
(CLONE_VM): Remove unused define.
* sysdeps/unix/sysv/linux/powerpc/powerpc32/clone.S (__clone):
Likewise.
(CLONE_VM): Remove unused define.
(CLONE_THREAD): Likewise.
* sysdeps/unix/sysv/linux/powerpc/powerpc64/clone.S (__clone):
Likewise.
(CLONE_VM): Remove unused define.
(CLONE_THREAD): Likewise.
* sysdeps/unix/sysv/linux/sh/clone.S (__clone): Likewise.
* sysdeps/unix/sysv/linux/sparc/sparc32/clone.S (__clone): Likewise.
(CLONE_VM): Remove unused define.
* sysdeps/unix/sysv/linux/sparc/sparc64/clone.S (__clone): Likewise.
(CLONE_VM): Remove unused define.
* sysdeps/unix/sysv/linux/tile/clone.S (__clone): Likewise.
* sysdeps/unix/sysv/linux/x86_64/clone.S (__clone): Likewise.
(CLONE_VM): Remove unused define.
* sysdeps/unix/sysv/linux/Makefile (tests): Add tst-clone3.
* sysdeps/unix/sysv/linux/tst-clone3.c: New file.
Fix
This patch adds ULPs for the float128 type, updates the abilist for libc
and libm, and adds the files bits/floatn.h and float128-abi.h, in order to
enable the new type for powerpc64le.
This patch also adds the implementation of sqrtf128 for powerpc64le, since
it is not implemented in libgcc. The sfp-machine.h header is taken from
libgcc.
Tested for powerpc64le (GCC 6.2 and GCC 7.1), powerpc64 and s390x.
* manual/math.texi (Mathematics): Mention the enabling of float128
for powerpc64le.
* sysdeps/powerpc/bits/floatn.h: New file.
* sysdeps/powerpc/fpu/libm-test-ulps: Regenerated.
* sysdeps/powerpc/fpu/math_private.h:
(__ieee754_sqrtf128): New inline override.
* sysdeps/powerpc/powerpc64le/Implies-before: New file.
* sysdeps/powerpc/powerpc64le/Makefile: New file.
* sysdeps/powerpc/powerpc64le/fpu/e_sqrtf128.c: New file.
* sysdeps/powerpc/powerpc64le/fpu/sfp-machine.h: New file.
* sysdeps/powerpc/powerpc64le/power9/fpu/e_sqrtf128.c: New file.
* sysdeps/unix/sysv/linux/powerpc/powerpc64/libc-le.abilist:
Updated.
* sysdeps/unix/sysv/linux/powerpc/powerpc64/libm-le.abilist:
Likewise.
* sysdeps/unix/sysv/linux/powerpc/powerpc64le/float128-abi.h:
New file.
Support for powerpc64le requires POWER8 or newer processors. Builds for
older processors are not tested. Require at least POWER8 to avoid
unintentional builds.
* sysdeps/powerpc/powerpc64le/configure.ac: Check for POWER8.
* sysdeps/powerpc/powerpc64le/configure: Update.
On powerpc64le, support for float128 will be enabled, which requires some
compiler features to be present. This patch adds a configure test to check
for such features, which are provided for powerpc64le since GCC 6.2.
Tested for powerpc64 and powerpc64le.
* INSTALL: Regenerate.
* manual/install.texi (Recommended Tools for Compilation): Mention
the powerpc64le-specific requirement in the manual.
* sysdeps/powerpc/powerpc64le/configure.ac: New file with checks
for the compiler features required for building float128.
* sysdeps/powerpc/powerpc64le/configure: New, auto-generated file.
According to ISO C11, section 6.5.3.3 "Unary arithmetic operators", the
result of the ~ operator is the bitwise complement of its (promoted)
operand.
This can lead to a comparison of a char with another integer type.
Tested on powerpc, powerpc64 and powerpc64le.
* malloc/tst-alloc_buffer.c (test_misaligned): Cast to char
before comparing with another char.
The .symver directive on common symbol just creates a new common symbol,
not an alias and the newer assembler with the bug fix for
https://sourceware.org/bugzilla/show_bug.cgi?id=21661
will issue an error. Before the fix, we got
$ readelf -sW libc.so | grep "loc[12s]"
5109: 00000000003a0608 8 OBJECT LOCAL DEFAULT 36 loc1
5188: 00000000003a0610 8 OBJECT LOCAL DEFAULT 36 loc2
5455: 00000000003a0618 8 OBJECT LOCAL DEFAULT 36 locs
6575: 00000000003a05f0 8 OBJECT GLOBAL DEFAULT 36 locs@GLIBC_2.2.5
7156: 00000000003a05f8 8 OBJECT GLOBAL DEFAULT 36 loc1@GLIBC_2.2.5
7312: 00000000003a0600 8 OBJECT GLOBAL DEFAULT 36 loc2@GLIBC_2.2.5
in libc.so. The versioned loc1, loc2 and locs have the wrong addresses.
After the fix, we got
$ readelf -sW libc.so | grep "loc[12s]"
6570: 000000000039e3b8 8 OBJECT GLOBAL DEFAULT 34 locs@GLIBC_2.2.5
7151: 000000000039e3c8 8 OBJECT GLOBAL DEFAULT 34 loc1@GLIBC_2.2.5
7307: 000000000039e3c0 8 OBJECT GLOBAL DEFAULT 34 loc2@GLIBC_2.2.5
[BZ #21666]
* misc/regexp.c (loc1): Add __attribute__ ((nocommon));
(loc2): Likewise.
(locs): Likewise.
Testing with GCC 7 for 32-bit x86 showed some ulps differences,
presumably from variation in when values with excess precision get
spilled to the stack and so lose that precision. This patch updates
the libm-test-ulps files accordingly.
* sysdeps/i386/fpu/libm-test-ulps: Update.
* sysdeps/i386/i686/fpu/multiarch/libm-test-ulps: Likewise.
The libm-test-support code clears exceptions at the end of
test_exceptions. However, it's possible that exceptions get raised as
part of the subsequent checks on the results of the function being
tested, and such exceptions then carry through to the next test run
and can cause that to fail spuriously. In particular, it's possible
for the ulps calculation to raise exceptions in some cases. This
patch puts exception clearing at the end of each of the check_*
functions, alongside errno clearing, to avoid one test affecting the
next in that way.
Tested for x86_64 (in conjunction with float128 patches).
* libm-test-support.c (check_float_internal): Clear exceptions at
end of function.
(check_int): Likewise.
(check_long): Likewise.
(check_bool): Likewise.
(check_longlong): Likewise.
(check_intmax_t): Likewise.
(check_uintmax_t): Likewise.
Building for x86_64 with float128 support, I get a localplt test
failure from lrintf128 calling feraiseexcept.
The problem is that an inline optimized version of feraiseexcept calls
__feraiseexcept_renamed in cases where it doesn't completely expand
inline, and that in turn is redirected to feraiseexcept for a library
call, so meaning the redirection of feraiseexcept to
__GI_feraiseexcept inside libm is lost for that call.
This patch fixes the problem by moving the redirect to an internal
header in the _LIBC case, with the internal header using
__GI_feraiseexcept where appropriate.
Tested for x86_64 (in conjunction with float128 patches).
* sysdeps/x86/fpu/bits/fenv.h [_LIBC] (__feraiseexcept_renamed):
Do not declare.
* sysdeps/x86/fpu/include/bits/fenv.h [_LIBC &&
__USE_EXTERN_INLINES] (__feraiseexcept_renamed): Declare here,
redirected to __GI_feraiseexcept if [SHARED && IS_IN (libm)].
This code:
L(between_2_3):
/* Load as big endian with overlapping loads and bswap to avoid
branches. */
movzwl -2(%rdi, %rdx), %eax
movzwl -2(%rsi, %rdx), %ecx
shll $16, %eax
shll $16, %ecx
movzwl (%rdi), %edi
movzwl (%rsi), %esi
orl %edi, %eax
orl %esi, %ecx
bswap %eax
bswap %ecx
subl %ecx, %eax
ret
needs a saturating subtract because the full register is used.
With this commit, only the lower 24 bits of the register are used,
so a regular subtraction suffices.
The test case change adds coverage for these kinds of bugs.
Similar to the other functions in the strfrom class, strfromf128 calls
__printf_fp in order to convert the floating-point value to characters.
This requires the value of IO_MTSAFE_IO.
Tested for powerpc64le and s390x.
* sysdeps/ieee754/float128/Makefile (CFLAGS-strfromf128.c): Add
$(libio-mtsafe) to get the value of IO_MTSAFE_IO.
When float128 support gets enabled for powerpc64le, the ULP errors for
float128 functions need to be presented in the manual. This patch adds
support for displaying them.
Tested for powerpc64le and s390x.
* manual/libm-err-tab.pl (@all_floats, %suffices, parse_ulps):
Enable generation of float128 entries on the error table.
The file include/float.h uses the macro __GLIBC_USE to test for TS 18661-3
support. Such macro is provided by bits/libc-header-start.h, so include it
to get the definition.
Tested for powerpc64le and s390x.
* include/float.h: Include libc-header-start.h to get the
definition of __GLIBC_USE.
Most significant changes are code simplification and use of doubles for
intermediate values. Also, some rearrangement to move early
non-dependent code later, out of the faster paths.
* sysdeps/ieee754/flt-32/e_powf.c: Optimized implementation utilizing
rearranged code and doubles float types.
sysdeps/powerpc/powerpc64/fpu/multiarch/s_sinf-ppc64.c should fall back to
sysdeps/powerpc/fpu/s_sinf.c not to sysdeps/ieee754/flt-32/s_sinf.c.
* sysdeps/powerpc/powerpc64/fpu/multiarch/s_sinf-ppc64.c: Change
s_sinf.c from sysdeps/ieee754/flt-32/ to sysdeps/powerpc/fpu/.
Commit cdfbe5037f added sysconf support
for cache geometries on powerpc, but mishandled errno. For valid input
parameters, sysconf() should not set errno.
* sysdeps/unix/sysv/linux/powerpc/sysconf.c: Remove references
to errno, and simplify remaining related code.
Linux commit ID a4700a26107241cc7b9ac8528b2c6714ff99983d reserved 2 more
bits for the instructions darn (Deliver a Random Number) and scv (System
Call Vectored).
Linux commit ID 6997e57d693b07289694239e52a10d2f02c3a46f reserved
another bit for internal usage.
* sysdeps/powerpc/bits/hwcap.h: Add PPC_FEATURE2_DARN and
PPC_FEATURE2_SCV.
* sysdeps/powerpc/dl-procinfo.c (_dl_powerpc_cap_flags): Add scv
and darn.
This patch adds _Float128 support to the ldbl-96 bits/iscanonical.h,
as needed for x86_64 / x86 / ia64 support of _Float128.
Tested for x86_64 (in conjunction with float128 patches).
* sysdeps/ieee754/ldbl-96/bits/iscanonical.h
[__HAVE_DISTINCT_FLOAT128] (__iscanonicalf128): New macro.
This patch makes math-tests.h, as used to describe support of given
floating-point types for sNaNs, rounding modes and exceptions, handle
distinguishing _Float128 from long double. This is needed for x86_64,
where if building with GCC 6 or earlier there is no __builtin_nansq,
so no way to get a signaling NaN of _Float128 type, so associated
tests cannot be run (although glibc itself works fine, as there is
never any need to create such an sNaN with a built-in function inside
glibc).
Tested for x86_64 (in conjunction with float128 patches).
* sysdeps/generic/math-tests.h: Include <bits/floatn.h>.
(MATH_TESTS_TG): New macro.
(SNAN_TESTS_float128): Likewise.
(ROUNDING_TESTS_float128): Likewise.
(EXCEPTION_TESTS_float128): Likewise.
(SNAN_TESTS): Define using MATH_TESTS_TG.
(ROUNDING_TESTS): Likewise.
(EXCEPTION_TESTS): Likewise.
As with other long double identifiers, float128_private.h has a
redefinition of SET_RESTORE_ROUNDL. However, that redefinition is
broken, since this is a macro with one argument being defined to take
no arguments. This patch fixes the redefinition. (x86_64 needs the
redefinition because SET_RESTORE_ROUNDL only changes the x87 rounding
mode, whereas _Float128 arithmetic uses the SSE rounding mode instead
on x86_64.)
Tested for x86_64 (in conjunction with float128 patches).
* sysdeps/ieee754/float128/float128_private.h
[SET_RESTORE_ROUNDF128] (SET_RESTORE_ROUNDL): Take an argument and
pass it to SET_RESTORE_ROUNDF128.
float128_private.h redefines ieee754.h identifiers ieee854_long_double
and IEEE854_LONG_DOUBLE_BIAS to map them to identifiers from
ieee754_float128.h.
This causes problems when ieee754.h is included after
float128_private.h and it's a version of ieee754.h that also defines
those identifiers; specifically, sysdeps/ieee754/ieee754.h, which
defines those identifiers for the x86 extended format. This patch
fixes this by ensuring an include of ieee754.h from float128_private.h
before the redefinitions.
Tested for x86_64 (in conjunction with float128 patches).
* sysdeps/ieee754/float128/float128_private.h: Include
<ieee754.h>.
The math_private.h macro min_of_type has broken _Float128 handling:
instead of passing its type argument to the key __EXPR_FLT128 macro,
it passes x, which is not a macro argument but whatever variable
called x happens to be visible in the calling function. If that
variable has the wrong type, the wrong one of long double and
_Float128 can get chosen. In particular, this applies to some
_Complex long double functions (where x happens to have type _Complex
long double, resulting in min_of_type returning a _Float128 value when
it should return a long double value). For some reason, this only
caused test failures for me on x86_64 with GCC 6 but not GCC 7 (I
suspect it triggers known bugs in conversions from x86 long double to
_Float128 that are present in GCC 6's soft-fp).
Tested for x86_64 (in conjunction with float128 patches).
* sysdeps/generic/math_private.h (__EXPR_FLT128): Do not apply
typeof to argument passed to __builtin_types_compatible_p.
(min_of_type): Pass type argument, not x, to __EXPR_FLT128.
Various type-generic libm wrapper templates, as used for float128, set
errno but do not include errno.h. I presume they must get an implicit
include from some internal header on powerpc64le; they don't get such
an implicit include on x86_64. This patch adds the missing includes
of errno.h to each such wrapper.
Tested for x86_64 (in conjunction with float128 patches).
* math/w_acos_template.c [__USE_WRAPPER_TEMPLATE]: Include
<errno.h>.
* math/w_acosh_template.c [__USE_WRAPPER_TEMPLATE]: Likewise.
* math/w_asin_template.c [__USE_WRAPPER_TEMPLATE]: Likewise.
* math/w_atanh_template.c [__USE_WRAPPER_TEMPLATE]: Likewise.
* math/w_cosh_template.c [__USE_WRAPPER_TEMPLATE]: Likewise.
* math/w_exp10_template.c [__USE_WRAPPER_TEMPLATE]: Likewise.
* math/w_exp2_template.c [__USE_WRAPPER_TEMPLATE]: Likewise.
* math/w_exp_template.c [__USE_WRAPPER_TEMPLATE]: Likewise.
* math/w_fmod_template.c [__USE_WRAPPER_TEMPLATE]: Likewise.
* math/w_hypot_template.c [__USE_WRAPPER_TEMPLATE]: Likewise.
* math/w_j0_template.c [__USE_WRAPPER_TEMPLATE]: Likewise.
* math/w_j1_template.c [__USE_WRAPPER_TEMPLATE]: Likewise.
* math/w_jn_template.c [__USE_WRAPPER_TEMPLATE]: Likewise.
* math/w_lgamma_r_template.c [__USE_WRAPPER_TEMPLATE]: Likewise.
* math/w_lgamma_template.c [__USE_WRAPPER_TEMPLATE]: Likewise.
* math/w_log10_template.c [__USE_WRAPPER_TEMPLATE]: Likewise.
* math/w_log2_template.c [__USE_WRAPPER_TEMPLATE]: Likewise.
* math/w_log_template.c [__USE_WRAPPER_TEMPLATE]: Likewise.
* math/w_pow_template.c [__USE_WRAPPER_TEMPLATE]: Likewise.
* math/w_remainder_template.c [__USE_WRAPPER_TEMPLATE]: Likewise.
* math/w_sinh_template.c [__USE_WRAPPER_TEMPLATE]: Likewise.
* math/w_sqrt_template.c [__USE_WRAPPER_TEMPLATE]: Likewise.
* math/w_tgamma_template.c [__USE_WRAPPER_TEMPLATE]: Likewise.
Three float128 files still include xlocale.h after it was removed. I
don't know why this didn't cause problems for powerpc64le float128
testing; it did cause problems for my x86_64 float128 testing. This
patch changes the includes to use bits/types/locale_t.h.
Tested for x86_64 (in conjunction with float128 patches).
* sysdeps/ieee754/float128/strtof128_l.c: Include
<bits/types/locale_t.h> instead of <xlocale.h>.
* sysdeps/ieee754/float128/wcstof128.c: Likewise.
* sysdeps/ieee754/float128/wcstof128_l.c: Likewise.
In preparation for the documentation of _FloatN and _FloatNx variants of
the remainder function, this patch changes the descriptions of remainder
and drem, so that remainder is described as primary and drem as an
alternative name for the same functionality.
* manual/arith.texi (Remainder Functions): Describe remainder as
primary and drem as an alternative name. Change the comment on
remainder to ISO, since it is defined in ISO C99.
The macro F128 in stdlib/tst-strtod.h is defined to provide the literal
suffix for _Float128 constants. It uses the macro __f128 (), which is
defined in bits/floatn.h to provide the correct literal suffix depending on
what is provided by the compiler.
However, F128 was not being expanded and only worked correctly, when
compiling with GCC 7 (or greater), since F128 is the literal suffix itself.
This patch adds an additional macro expansion so that the macro F128
expands to the correct literal suffix on older compilers.
* stdlib/tst-strtod.h (MMFUNC): New macro to provide an addition
macro expansion.
(GEN_TEST_STRTOD_FOREACH): Use MMFUNC for _Float128.
* Unicode 10.0.0 Support: Character encoding, character type info, and
transliteration tables are all updated to Unicode 10.0.0, using
generator scripts contributed by Mike FABIAN (Red Hat).
This commit adds fixed-size allocation buffers. The primary use
case is in NSS modules, where dynamically sized data is stored
in a fixed-size buffer provided by the caller.
Other uses include a replacement of mempcpy cascades (which is
safer due to the size checking inherent to allocation buffers).
Implement strcmp family IFUNC selectors in C.
All internal calls within libc.so can use IFUNC on x86-64 since unlike
x86, x86-64 supports PC-relative addressing to access the GOT entry so
that it can call via PLT without using an extra register. For libc.a,
we can't use IFUNC for functions which are called before IFUNC has been
initialized. Use IFUNC internally reduces the icache footprint since
libc.so and other codes in the process use the same implementations.
This patch uses IFUNC for strcmp family functions within libc.
* sysdeps/x86_64/multiarch/Makefile (sysdep_routines): Add
strcmp-sse2, strcmp-sse4_2, strncmp-sse2, strncmp-sse4_2,
strcasecmp_l-sse2, strcasecmp_l-sse4_2, strcasecmp_l-avx,
strncase_l-sse2, strncase_l-sse4_2 and strncase_l-avx.
* sysdeps/x86_64/multiarch/ifunc-strcasecmp.h: New file.
* sysdeps/x86_64/multiarch/strcasecmp.c: Likewise.
* sysdeps/x86_64/multiarch/strcasecmp_l-avx.S: Likewise.
* sysdeps/x86_64/multiarch/strcasecmp_l-sse2.S: Likewise.
* sysdeps/x86_64/multiarch/strcasecmp_l-sse4_2.S: Likewise.
* sysdeps/x86_64/multiarch/strcasecmp_l.c: Likewise.
* sysdeps/x86_64/multiarch/strcmp-sse2.S: Likewise.
* sysdeps/x86_64/multiarch/strcmp-sse4_2.S: Likewise.
* sysdeps/x86_64/multiarch/strcmp.c: Likewise.
* sysdeps/x86_64/multiarch/strncase.c: Likewise.
* sysdeps/x86_64/multiarch/strncase_l-avx.S : Likewise.
* sysdeps/x86_64/multiarch/strncase_l-sse2.S: Likewise.
* sysdeps/x86_64/multiarch/strncase_l-sse4_2.S: Likewise.
* sysdeps/x86_64/multiarch/strncase_l.c: Likewise.
* sysdeps/x86_64/multiarch/strncmp-sse2.S: Likewise.
* sysdeps/x86_64/multiarch/strncmp-sse4_2.S: Likewise.
* sysdeps/x86_64/multiarch/strncmp.c: Likewise.
* sysdeps/x86_64/multiarch/strcasecmp_l.S: Removed.
* sysdeps/x86_64/multiarch/strcmp.S: Likewise.
* sysdeps/x86_64/multiarch/strncase_l.S: Likewise.
* sysdeps/x86_64/multiarch/strncmp.S: Likewise.
* sysdeps/x86_64/multiarch/strcmp-sse42.S: Include <sysdep.h>.
(STRCMP_SSE42): New. Defined to __strcmp_sse42 if not defined.
[USE_AS_STRCASECMP_L || USE_AS_STRNCASECMP_L]: Include
"locale-defines.h".
(UPDATE_STRNCMP_COUNTER): New.
(SECTION): Likewise.
(GLABEL): Likewise.
(LABEL): Likewise.
* sysdeps/x86_64/multiarch/strncmp-ssse3.S: Rewrite and enable
for libc.a.
As shown by conform/ tests once the remaining namespace issues are
fixed, the tile bits/sigaction.h fails to declare SA_RESETHAND,
SA_RESTART and SA_NODEFER for non-XSI POSIX.1:2008 as other versions
do. Those constants were moved from XSI to Base in the 2008 edition
of POSIX. This patch fixes the conditions to match other versions of
this header.
Tested (compilation only) for tilegx-linux-gnu with
build-many-glibcs.py.
[BZ #21622]
* sysdeps/unix/sysv/linux/tile/bits/sigaction.h (SA_RESTART):
Define for [__USE_UNIX98 || __USE_XOPEN2K8], not [__USE_UNIX98 ||
__USE_MISC].
(SA_NODEFER): Likewise.
(SA_RESETHAND): Likewise.
Rename glibc.tune.ifunc to glibc.tune.hwcaps and move it to
sysdeps/x86/dl-tunables.list since it is x86 specicifc. Also
change type of data_cache_size, data_cache_size and
non_temporal_threshold to unsigned long int to match size_t.
Remove usage DEFAULT_STRLEN from cpu-tunables.c.
* elf/dl-tunables.list (glibc.tune.ifunc): Removed.
* sysdeps/x86/dl-tunables.list (glibc.tune.hwcaps): New.
Remove security_level on all fields.
* manual/tunables.texi: Replace ifunc with hwcaps.
* sysdeps/x86/cpu-features.c (TUNABLE_CALLBACK (set_ifunc)):
Renamed to ..
(TUNABLE_CALLBACK (set_hwcaps)): This.
(init_cpu_features): Updated.
* sysdeps/x86/cpu-features.h (cpu_features): Change type of
data_cache_size, data_cache_size and non_temporal_threshold to
unsigned long int.
* sysdeps/x86/cpu-tunables.c (DEFAULT_STRLEN): Removed.
(TUNABLE_CALLBACK (set_ifunc)): Renamed to ...
(TUNABLE_CALLBACK (set_hwcaps)): This. Update comments. Don't
use DEFAULT_STRLEN.
Backtrace through _dl_tlsdesc_resolve_rela was broken because the offset
of x30 from cfa was not in the debug info.
Add enough annotation so backtracing from the dynamic linker through
tlsdesc entry points works and the debugger shows registers correctly.
This patch add an extra test for passing invalid flags and check its
expected failure. It shows an invalid LO_HI_LONG macro definition for
x86_64 with leads to passing invalid flags on some configurations.
The new tests fails on i686-linux-gnu and potentially on other 32 bits
architecture that uses the compat syscall definition due a kernel bug.
It is intended to be fixed upstream.
Checked on x86_64-linux-gnu
* misc/tst-preadvwritev2-common.c: New file.
* misc/tst-preadvwritev2.c (do_test): Add test for invalid flag.
* misc/tst-preadvwritev64v2.c (do_test): Likewise.
Many of the things defined by bits/signum.h are invariant across all
supported operating systems. This patch factors out all of them to a
new header bits/signum-generic.h, which each bits/signum.h will include
and then override whichever things need adjustment. Normally that will
mean, at most, adding or changing a few signal numbers.
A user-visible side effect is that the obsolete signal constant SIGUNUSED
(which is an alias for SIGSYS on all platforms that define it) is no
longer exposed by any version of bits/signum.h.
A side effect only relevant to glibc hackers is that _NSIG is now defined
in terms of __SIGRTMAX, instead of the other way around. This is because
__SIGRTMAX varies from platform to platform, but _NSIG==__SIGRTMAX+1 is
true universally. If your platform doesn't support realtime signals,
leave __SIGRTMAX equal to __SIGRTMIN.
I also added a Linux-specific test to make sure that our signal constants
match the ones in <asm/signal.h>, since we can't use that header (it's
not even vaguely namespace-clean).
* bits/signum-generic.h: Renamed from bits/signum.h.
Add proper multiple include guard and misuse check.
Define __SIGRTMIN = __SIGRTMAX = 32, and define _NSIG = __SIGRTMAX+1.
Move definition of SIGIO to "archaic names for compatibility" section.
* bits/signum.h: New file which just includes bits/signum-generic.h.
* sysdeps/unix/bsd/bits/signum.h
* sysdeps/unix/sysv/linux/bits/signum.h
* sysdeps/unix/sysv/linux/alpha/bits/signum.h
* sysdeps/unix/sysv/linux/hppa/bits/signum.h
* sysdeps/unix/sysv/linux/mips/bits/signum.h
* sysdeps/unix/sysv/linux/sparc/bits/signum.h
Just include <bits/signum-generic.h> and then add or adjust
signal constants. Do not define SIGUNUSED, SIGRTMIN, or SIGRTMAX.
* signal/Makefile: Install bits/signum-generic.h.
* signal/signal.h: Define SIGRTMIN and SIGRTMAX here.
* sysdeps/generic/siglist.h: SIGSYS and SIGWINCH are
universal. Prefer SIGPOLL to SIGIO. Simplify #ifdeffage.
* sysdeps/unix/sysv/linux/tst-signal-numbers.sh: New test.
* sysdeps/unix/sysv/linux/Makefile: Run it.
<locale.h> is specified to define locale_t in POSIX.1-2008, and so are
all of the headers that define functions that take locale_t arguments.
Under _GNU_SOURCE, the additional headers that define such functions
have also always defined locale_t. Therefore, there is no need to use
__locale_t in public function prototypes, nor in any internal code.
* ctype/ctype-c99_l.c, ctype/ctype.h, ctype/ctype_l.c
* include/monetary.h, include/stdlib.h, include/time.h
* include/wchar.h, locale/duplocale.c, locale/freelocale.c
* locale/global-locale.c, locale/langinfo.h, locale/locale.h
* locale/localeinfo.h, locale/newlocale.c
* locale/nl_langinfo_l.c, locale/uselocale.c
* localedata/bug-usesetlocale.c, localedata/tst-xlocale2.c
* stdio-common/vfscanf.c, stdlib/monetary.h, stdlib/stdlib.h
* stdlib/strfmon_l.c, stdlib/strtod_l.c, stdlib/strtof_l.c
* stdlib/strtol.c, stdlib/strtol_l.c, stdlib/strtold_l.c
* stdlib/strtoll_l.c, stdlib/strtoul_l.c, stdlib/strtoull_l.c
* string/strcasecmp.c, string/strcoll_l.c, string/string.h
* string/strings.h, string/strncase.c, string/strxfrm_l.c
* sysdeps/ieee754/float128/strtof128_l.c
* sysdeps/ieee754/float128/wcstof128.c
* sysdeps/ieee754/float128/wcstof128_l.c
* sysdeps/ieee754/ldbl-128ibm/strtold_l.c
* sysdeps/ieee754/ldbl-64-128/strtold_l.c
* sysdeps/ieee754/ldbl-opt/nldbl-compat.c
* sysdeps/ieee754/ldbl-opt/nldbl-strfmon_l.c
* sysdeps/ieee754/ldbl-opt/nldbl-strtold_l.c
* sysdeps/ieee754/ldbl-opt/nldbl-wcstold_l.c
* sysdeps/powerpc/powerpc32/power7/strcasecmp.S
* sysdeps/powerpc/powerpc64/power7/strcasecmp.S
* sysdeps/x86_64/strcasecmp_l-nonascii.c
* sysdeps/x86_64/strncase_l-nonascii.c, time/strftime_l.c
* time/strptime_l.c, time/time.h, wcsmbs/mbsrtowcs_l.c
* wcsmbs/wchar.h, wcsmbs/wcscasecmp.c, wcsmbs/wcsncase.c
* wcsmbs/wcstod.c, wcsmbs/wcstod_l.c, wcsmbs/wcstof.c
* wcsmbs/wcstof_l.c, wcsmbs/wcstol_l.c, wcsmbs/wcstold.c
* wcsmbs/wcstold_l.c, wcsmbs/wcstoll_l.c, wcsmbs/wcstoul_l.c
* wcsmbs/wcstoull_l.c, wctype/iswctype_l.c
* wctype/towctrans_l.c, wctype/wcfuncs_l.c
* wctype/wctrans_l.c, wctype/wctype.h, wctype/wctype_l.c:
Change all uses of __locale_t to locale_t.
xlocale.h is already a single-type micro-header, defining struct
__locale_struct and the typedefs __locale_t and locale_t. This patch
brings it into the bits/types/ scheme: there are now
bits/types/__locale_t.h which defines only __locale_struct and
__locale_t, and bits/types/locale_t.h which defines locale_t as well
as the other two. None of *our* headers need __locale_t.h, but it
appears to me that libstdc++ could make use of it.
There are a lot of external uses of xlocale.h, but all the uses I
checked had an autoconf test or equivalent for its existence. It has
never been available from other C libraries, and it has always
contained a comment reading "This file is not standardized, don't rely
on it, it can go away without warning" so I think dropping it is
pretty safe.
I also took the opportunity to clean up comments in various public
header files that still talk about the *_l interfaces as though they
were completely nonstandard. There are a few of them, notably the
strtoX_l and wcstoX_l families, that haven't been standardized, but
the bulk are in POSIX.1-2008.
* locale/xlocale.h: Rename to...
* locale/bits/types/__locale_t.h: ...here. Adjust commentary.
Only define struct __locale_struct and __locale_t, not locale_t.
* locale/bits/types/locale_t.h: New file; define locale_t here.
* locale/Makefile (headers): Update to match.
* include/xlocale.h: Delete wrapper.
* include/bits/types/__locale_t.h: New wrapper.
* include/bits/types/locale_t.h: New wrapper.
* ctype/ctype.h, include/printf.h, include/time.h
* locale/langinfo.h, locale/locale.h, stdlib/monetary.h
* stdlib/stdlib.h, string/string.h, string/strings.h, time/time.h
* wcsmbs/wchar.h, wctype/wctype.h: Use bits/types/locale_t.h.
Correct outdated comments regarding the standardization status of
the functions that take locale_t arguments.
* stdlib/strtod_l.c, stdlib/strtof_l.c, stdlib/strtol_l.c
* stdlib/strtold_l.c, stdlib/strtoul_l.c, stdlib/strtoull_l.c
* sysdeps/ieee754/ldbl-128ibm/strtold_l.c
* sysdeps/ieee754/ldbl-64-128/strtold_l.c
* wcsmbs/wcstod.c, wcsmbs/wcstod_l.c, wcsmbs/wcstof.c
* wcsmbs/wcstof_l.c, wcsmbs/wcstold.c, wcsmbs/wcstold_l.c:
Don't include xlocale.h. If necessary, include locale.h instead.
* stdlib/strtold_l.c: Unconditionally include wchar.h.
This patch consolidates the open Linux syscall implementation on
sysdeps/unix/sysv/linux/open{64}.c. The changes are:
1. Remove wordsize-64 openat{64}.
2. For architetures that define __OFF_T_MATCHES_OFF64_T openat64
will be default one with alias to required symbols. Otherwise
openat64 will pass the required O_LARGEFILE flag on syscall.
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/openat.c (__libc_openat): Build only
for !__OFF_T_MATCHES_OFF64_T.
* sysdeps/unix/sysv/linux/openat64.c (__libc_openat64): New
implementation based on open64.
* sysdeps/unix/sysv/linux/wordsize-64/openat.c: Remove file.
* sysdeps/unix/sysv/linux/wordsize-64/openat64.c: Likewise.
This patch XFAILs one test where the powerpc32 ucontext_t has the
wrong type of a field, to allow the conform/ tests as a whole to pass
once the namespace issues are fixed.
Tested with build-many-glibcs.py.
[BZ #21635]
* sysdeps/unix/sysv/linux/powerpc/powerpc32/Makefile
[$(subdir) = conform] (conformtest-xfail-conds): New variable.
* conform/data/signal.h-data (uc_mcontext): XFAIL for
powerpc32-linux.
* conform/data/ucontext.h-data (uc_mcontext): Likewise.
This patch XFAILs one test where the ia64 ucontext_t has the wrong
type of a field, to allow the conform/ tests as a whole to pass once
the namespace issues are fixed.
Tested with build-many-glibcs.py.
[BZ #21634]
* sysdeps/unix/sysv/linux/ia64/Makefile [$(subdir) = conform]
(conformtest-xfail-conds): New variable.
* conform/data/signal.h-data (uc_sigmask): XFAIL for ia64-linux.
* conform/data/ucontext.h-data (uc_sigmask): Likewise.
Add a workload for powf. This is a reduced trace based on 2.3 billion
samples extracted from wrf. The distribution of values, in particular
frequency of commonly used operands is the same as in the full trace.
* benchtests/powf-inputs: Add reduced trace from wrf.
The current IFUNC selection is based on microbenchmarks in glibc. It
should give the best performance for most workloads. But other choices
may have better performance for a particular workload or on the hardware
which wasn't available at the selection was made. The environment
variable, GLIBC_TUNABLES=glibc.tune.ifunc=-xxx,yyy,-zzz...., can be used
to enable CPU/ARCH feature yyy, disable CPU/ARCH feature yyy and zzz,
where the feature name is case-sensitive and has to match the ones in
cpu-features.h. It can be used by glibc developers to override the
IFUNC selection to tune for a new processor or improve performance for
a particular workload. It isn't intended for normal end users.
NOTE: the IFUNC selection may change over time. Please check all
multiarch implementations when experimenting.
Also, GLIBC_TUNABLES=glibc.tune.x86_non_temporal_threshold=NUMBER is
provided to set threshold to use non temporal store to NUMBER,
GLIBC_TUNABLES=glibc.tune.x86_data_cache_size=NUMBER to set data cache
size, GLIBC_TUNABLES=glibc.tune.x86_shared_cache_size=NUMBER to set
shared cache size.
* elf/dl-tunables.list (tune): Add ifunc,
x86_non_temporal_threshold,
x86_data_cache_size and x86_shared_cache_size.
* manual/tunables.texi: Document glibc.tune.ifunc,
glibc.tune.x86_data_cache_size, glibc.tune.x86_shared_cache_size
and glibc.tune.x86_non_temporal_threshold.
* sysdeps/unix/sysv/linux/x86/dl-sysdep.c: New file.
* sysdeps/x86/cpu-tunables.c: Likewise.
* sysdeps/x86/cacheinfo.c
(init_cacheinfo): Check and get data cache size, shared cache
size and non temporal threshold from cpu_features.
* sysdeps/x86/cpu-features.c [HAVE_TUNABLES] (TUNABLE_NAMESPACE):
New.
[HAVE_TUNABLES] Include <unistd.h>.
[HAVE_TUNABLES] Include <elf/dl-tunables.h>.
[HAVE_TUNABLES] (TUNABLE_CALLBACK (set_ifunc)): Likewise.
[HAVE_TUNABLES] (init_cpu_features): Use TUNABLE_GET to set
IFUNC selection, data cache size, shared cache size and non
temporal threshold.
* sysdeps/x86/cpu-features.h (cpu_features): Add data_cache_size,
shared_cache_size and non_temporal_threshold.
Improve support for math function benchmarking. This patch adds
a feature that allows accurate benchmarking of traces extracted
from real workloads. This is done by iterating over all samples
rather than repeating each sample many times (which completely
ignores branch prediction and cache effects). A trace can be
added to existing math function inputs via
"## name: workload-<name>", followed by the trace.
* benchtests/README: Describe workload feature.
* benchtests/bench-skeleton.c (main): Add support for
benchmarking traces from workloads.
Remove one more string inline that was defined directly in string.h;
in the absence of the rest of the inlines, it broke the build.
Like other ifunc shims for these functions,
x86_64/multiarch/{mem,st}pcpy.c need to define __NO_STRING_INLINES and
NO_MEMPCPY_STPCPY_REDIRECT.
* string/string.h (__mempcpy_inline): Delete.
* sysdeps/x86_64/multiarch/mempcpy.c
* sysdeps/x86_64/multiarch/stpcpy.c:
Define NO_MEMPCPY_STPCPY_REDIRECT and __NO_STRING_INLINES
before including string.h.
Add powf() bench test with input which covers these cases:
- positive base to positive exponent
- exponent 0
- negative base to even exponent
- exponent 1
- exponent -1
- squared
- squareroot
- 1 to negative exponent
- -1 to negative exponent
- base 0
- -1 to even exponent
- small base
- small exponent
* benchtests/Makefile (bench-math): Add powf.
* benchtests/powf-inputs: New file.
These machine-dependent inline string functions have never been on by
default, and even if they were a good idea at the time they were
introduced, they haven't really been touched in ten to fifteen years
and probably aren't a good idea on current-gen processors. Current
thinking is that this class of optimization is best left to the
compiler.
* bits/string.h, string/bits/string.h
* sysdeps/aarch64/bits/string.h
* sysdeps/m68k/m680x0/m68020/bits/string.h
* sysdeps/s390/bits/string.h, sysdeps/sparc/bits/string.h
* sysdeps/x86/bits/string.h: Delete file.
* string/string.h: Don't include bits/string.h.
* string/bits/string3.h: Rename to bits/string_fortified.h.
No need to undef various symbols that the removed headers
might have defined as macros.
* string/Makefile (headers): Remove bits/string.h, change
bits/string3.h to bits/string_fortified.h.
* string/string-inlines.c: Update commentary. Remove definitions
of various macros that nothing looks at anymore. Don't directly
include bits/string.h. Set _STRING_INLINE_unaligned here, based on
compiler-predefined macros.
* string/strncat.c: If STRNCAT is not defined, or STRNCAT_PRIMARY
_is_ defined, provide internal hidden alias __strncat.
* include/string.h: Declare internal hidden alias __strncat.
Only forward __stpcpy to __builtin_stpcpy if __NO_STRING_INLINES is
not defined.
* include/bits/string3.h: Rename to bits/string_fortified.h,
update to match above.
* sysdeps/i386/string-inlines.c: Define compat symbols for
everything formerly defined by sysdeps/x86/bits/string.h.
Make existing definitions into compat symbols as well.
Remove some no-longer-necessary messing around with macros.
* sysdeps/powerpc/powerpc32/power4/multiarch/mempcpy.c
* sysdeps/powerpc/powerpc64/multiarch/mempcpy.c
* sysdeps/powerpc/powerpc64/multiarch/stpcpy.c
* sysdeps/s390/multiarch/mempcpy.c
No need to define _HAVE_STRING_ARCH_mempcpy.
Do define __NO_STRING_INLINES and NO_MEMPCPY_STPCPY_REDIRECT.
* sysdeps/i386/i686/multiarch/strncat-c.c
* sysdeps/s390/multiarch/strncat-c.c
* sysdeps/x86_64/multiarch/strncat-c.c
Define STRNCAT_PRIMARY. Don't change definition of libc_hidden_def.
This patch removes some MIPS code in glibc that was conditional on old
GCC versions no longer supported for building glibc.
Tested with build-many-glibcs.py.
* sysdeps/mips/atomic-machine.h (R10K_BEQZ_INSN): Remove.
[__GNUC_PREREQ (4, 8) || __mips16]: Make code unconditional.
[!__GNUC_PREREQ (4, 8) && !__mips16]: Remove conditional code.
* sysdeps/mips/math-tests.h
[_MIPS_SIM != _ABIO32 && !__GNUC_PREREQ (4, 9)]: Remove
conditional code.
The wait3 function was removed in the 2001 edition of POSIX.
sys/wait.h wrongly declares it for the 2001 and 2008 editions of POSIX
when XSI features are enabled. This patch fixes the conditionals.
Tested for x86_64.
[BZ #21625]
* posix/sys/wait.h (strust rusage forward declaration): Change
[__USE_XOPEN_EXTENDED] conditional to [__USE_XOPEN_EXTENDED &&
!__USE_XOPEN2K].
(wait3): Likewise.
This patch makes configure require GCC 4.9 or later for building
glibc, and documents that requirement. Requiring GCC 4.9 or later
allows use of _Generic (as in tzcode). It would allow <stdatomic.h>
and _Atomic to be used as well if desired, although we need to avoid
any libatomic dependencies on any platforms. This patch is explicitly
the minimum to implement a new version requirement, with any
consequent cleanups of conditional code (not in installed headers or
files shared with gnulib etc.) to be done separately.
Tested for x86_64.
* configure.ac (libc_cv_compiler_ok): Require GCC 4.9 or later.
* configure: Regenerated.
* manual/install.texi (Tools for Compilation): Document
requirement for GCC 4.9 or later.
* INSTALL: Regenerated.
LD_LIBRARY_PATH can only be used to reorder system search paths, which
is not useful functionality.
This makes an exploitable unbounded alloca in _dl_init_paths unreachable
for AT_SECURE=1 programs.
This patch removes PTRACE_GETREGS, PTRACE_SETREGS, PTRACE_GETFPREGS
and PTRACE_SETFPREGS as these requests does not exist on s390 kernel.
But the kernel has support for PTRACE_SINGLEBLOCK,
PTRACE_SECCOMP_GET_FILTER, PTRACE_PEEKUSR_AREA, PTRACE_POKEUSR_AREA,
PTRACE_GET_LAST_BREAK, PTRACE_ENABLE_TE, PTRACE_DISABLE_TE and
PTRACE_TE_ABORT_RAND. Thus those are defined now.
The current kernel s390 specific ptrace.h file also defines
PTRACE_PEEKTEXT_AREA, PTRACE_PEEKDATA_AREA, PTRACE_POKETEXT_AREA,
PTRACE_POKEDATA_AREA, PTRACE_PEEK_SYSTEM_CALL, PTRACE_POKE_SYSTEM_CALL
and PTRACE_PROT, but those requests are not supported.
Thus those defines are skipped in glibc ptrace.h.
There were old includes of ptrace.h in sysdeps/s390/fpu/fesetenv.c.
The ptrace feature isn't used there anymore, thus I removed the includes.
Before this patch, <glibc>/sysdeps/unix/sysv/linux/s390/sys/ptrace.h
uses ptrace-request 12 for PTRACE_GETREGS,
but <kernel>/include/uapi/linux/ptrace.h uses 12 for PTRACE_SINGLEBLOCK.
The s390 kernel has never had support for PTRACE_GETREGS!
Thus glibc ptrace.h is adjusted to match kernel ptrace.h.
The new s390 specific test ensures, that PTRACE_SINGLEBLOCK defined
in glibc works as expected. If the kernel would interpret it as
PTRACE_GETREGS, then the testcase will not make any progress
and will time out.
ChangeLog:
[BZ #21539]
* NEWS: Mention s390 ptrace request changes.
* sysdeps/unix/sysv/linux/s390/sys/ptrace.h
(PTRACE_GETREGS, PTRACE_SETREGS, PTRACE_GETFPREGS,
PTRACE_SETFPREGS): Remove enum constant.
(PT_GETREGS, PT_SETREGS, PT_GETFPREGS, T_SETFPREGS):
Remove defines.
(PTRACE_SINGLEBLOCK): New enum constant.
(PT_STEPBLOCK): New define.
(PTRACE_PEEKUSR_AREA, PTRACE_POKEUSR_AREA,
PTRACE_GET_LAST_BREAK, PTRACE_ENABLE_TE, PTRACE_DISABLE_TE,
PTRACE_TE_ABORT_RAND): New enum constant and define.
* sysdeps/s390/fpu/fesetenv.c: Remove ptrace.h includes.
* sysdeps/unix/sysv/linux/s390/tst-ptrace-singleblock.c:
New file.
* sysdeps/unix/sysv/linux/s390/Makefile: Add test.
This patch fixes a namespace issue for one more field in the x86
sys/ucontext.h that I missed in my previous changes.
Tested for x86_64.
[BZ #21457]
* sysdeps/unix/sysv/linux/x86/sys/ucontext.h [__x86_64__] (struct
_libc_xmmreg): Use __ctx in defining field.
The siginterrupt function is XSI-shaded in POSIX. glibc wrongly
declares it in <signal.h> for non-XSI POSIX.1:2008. This patch fixes
the conditionals (moving it into the __USE_XOPEN_EXTENDED ||
__USE_MISC block as the minimum conservative change, although it's not
clear this obsolescent function should be declared by default).
Tested for x86_64.
[BZ #21597]
* signal/signal.h (siginterrupt): Make declaration conditional on
[__USE_XOPEN_EXTENDED || __USE_MISC].
In POSIX.1:2008, various signal stack symbols - sigaltstack, SS_*,
SIGSTKSZ, MINSIGSTKSZ - are all XSI-shaded. glibc wrongly makes them
visible for non-XSI POSIX. This patch fixes the conditionals, leaving
the symbols available for __USE_MISC as it seems likely some of them
are widely used and should be visible by default. (Note that stack_t
is *not* XSI-shaded and so the conditionals on that are correctly
unchanged.)
Tested for x86_64.
[BZ #21584]
* signal/signal.h: Make includes of <bits/sigstack.h> and
<bits/ss_flags.h> conditional on [__USE_XOPEN_EXTENDED ||
__USE_MISC].
(sigaltstack): Make declaration conditional on
[__USE_XOPEN_EXTENDED || __USE_MISC].
Some older standards (XPG4.2 through POSIX.1:2001, XSI only) require
sys/wait.h to include the definition of struct rusage. This is
missing in glibc.
This patch adds the required definition. struct rusage is moved to a
new header bits/types/struct_rusage.h to avoid bringing in the whole
of sys/resource.h (although the standards in question do allow the
whole of sys/resource.h to be brought in). In the five
bits/resource.h headers, the only variation between the definitions of
struct rusage is that the sysdeps/unix/sysv/linux version is prepared
for x32 (by having anonymous unions with __syscall_slong_t fields) and
the others are not. Thus, this version is suitable for use
generically (everything other than x32 simply has __syscall_slong_t
the same as long int, so there are no API or ABI changes involved, and
anonymous unions are already a required language feature for glibc
headers elsewhere), and this patch uses it as a base for the single
implementation of bits/types/struct_rusage.h.
Tested for x86_64, and with build-many-glibcs.py.
[BZ #21575]
* resource/bits/types/struct_rusage.h: New file.
* include/bits/types/struct_rusage.h: Likewise.
* bits/resource.h (struct rusage): Include
<bits/types/struct_rusage.h> instead of defining here.
* sysdeps/unix/sysv/linux/bits/resource.h (struct rusage):
Likewise.
* sysdeps/unix/sysv/linux/alpha/bits/resource.h (struct rusage):
Likewise.
* sysdeps/unix/sysv/linux/mips/bits/resource.h (struct rusage):
Likewise.
* sysdeps/unix/sysv/linux/sparc/bits/resource.h (struct rusage):
Likewise.
* resource/Makefile (headers): Add bits/types/struct_rusage.h.
* posix/sys/wait.h [__USE_XOPEN_EXTENDED && !__USE_XOPEN2K8]:
Include <bits/types/struct_rusage.h>
Awk 3.1.x does not support multi-dimensional arrays, so fix up to make
sure that gen-tunables.awk works on it.
* scripts/gen-tunables.awk: Avoid multi-dimensional arrays.
Building glibc with gcc configured with --enable-default-pie failed on s390
due to assembler messages:
../sysdeps/unix/sysv/linux/s390/s390-32/__makecontext_ret.S:44:
Error: junk at end of line, first unrecognized character is `@'
HIDDEN_JUMPTARGET was expanded to exit@PLT@GOTOFF.
If SHARED is not defined, HIDDEN_JUMPTARGET is defined to JUMPTARGET
in sysdeps/s390/s390-32/sysdep.h. There it expanded to exit@PLT
in non SHARED case as PIC is defined if gcc is configured with
--enable-default-pie. Thus I've changed the "ifdef PIC" to "ifdef SHARED"
as we do not want PLTs in the static obj files. I've also changed this
in sysdeps/s390/s390-64/sysdep.h.
I've also adjusted sysdeps/unix/sysv/linux/s390/s390-32/__makecontext_ret.S.
If glibc is configured with --disable-hidden-plt, then NO_HIDDEN is defined.
In SHARED case HIDDEN_JUMPTARGET would be expanded to exit@PLT@GOTOFF
instead of __GI_exit@GOTOFF.
Now we jump to:
- __GI_exit if SHARED is defined
- exit@PLT if SHARED and NO_HIDDEN is defined
- exit if both are not defined.
On s390 31bit we have to setup GOT pointer in r12 if we use a PLT stub.
Therefore I use SYSCALL_PIC_SETUP from sysdep.h and added the missing semicolons.
ChangeLog:
[BZ #21537]
* sysdeps/s390/s390-32/sysdep.h (JUMPTARGET, SYSCALL_PIC_SETUP):
Check SHARED instead of PIC.
(SYSCALL_PIC_SETUP): Add missing semicolons.
* sysdeps/s390/s390-64/sysdep.h (JUMPTARGET, SYSCALL_PIC_SETUP):
Check SHARED instead of PIC.
* sysdeps/unix/sysv/linux/s390/s390-32/__makecontext_ret.S
(__makecontext_ret): Adjust code to jump to exit.
Since kernel 2.6.0 all Linux version accept the system call number
in register 1 for svc 0. There is no need to have special handling
that uses EX for system calls < 256. This will simplify and speed
up that code.
A microbenchmark doing "syscall(__NR_getpid);" in a loops gets faster
by ~12%.
* sysdeps/unix/sysv/linux/s390/s390-32/syscall.S: Simplify
code by always using SVC 0 instead of EX.
* sysdeps/unix/sysv/linux/s390/s390-64/syscall.S: Likewise.
Since _dl_out_of_memory is static in elf/dl-error-skeleton.c:
static const char _dl_out_of_memory[] = "out of memory";
remove _dl_out_of_memory from elf/Versions.
* elf/Versions (ld): Remove _dl_out_of_memory.
This patch updates files coming from tzcode to the versions in tzcode
2017b. A couple of changes to other glibc code are needed.
time/tzset.c was using the SECSPERDAY macro from tzfile.h, which no
longer defines that macro, so a local definition is added to tzset.c.
Because timezone/private.h now defines the _ macro whenever
HAVE_GETTEXT is true, even if it was previously defined, it is also
necessary to avoid a conflict with the definition in
include/libintl.h. Defining _ISOMAC is the obvious way to avoid such
internal definitions being visible, together with defining TZ_DOMAIN
so that zic and zdump continue to get the messages from the libc
domain as desired. However, zic and zdump rely on PKGVERSION and
REPORT_BUGS_TO from config.h, which is not included by default with
_ISOMAC, so -include config.h needs adding to the options for these
programs as well. Together those changes allow unmodified tzcode
2017b sources to work in glibc.
Tested for x86_64.
* timezone/private.h: Update from tzcode 2017b.
* timezone/tzfile.h: Likewise.
* timezone/tzselect.ksh: Likewise.
* timezone/zdump.c: Likewise.
* timezone/zic.c: Likewise.
* timezone/Makefile (tz-cflags): Add -D_ISOMAC
-DTZ_DOMAIN='"libc"' -include $(common-objpfx)config.h.
* time/tzset.c (SECSPERDAY): New macro.
The Summary is now generated from @standards, and syntax-checking is
performed. If invalid @standards syntax is detected, summary.pl will
fail, reporting all errors. Failure and error reporting is disabled
for now, however, since much of the manual is still incomplete
wrt. header and standards annotations.
Note that the sorting order of the Summary has changed; summary.pl
respects the locale, like summary.awk did, but the use of LC_ALL=C is
introduced in the Makefile. Other notable deviations are improved
detection of the annotated elements' names, which are used for
sorting, and improved detection of the @node used to reference into
the manual. The most noticeable difference in the rendered Summary is
that entries may now contain multiple lines, one for each header and
standard combination.
summary.pl accepts a `--help' option, which details the expected
syntax of @standards. If errors are reported, the user is directed to
this feature for further information.
* manual/Makefile: Generate summary.texi with summary.pl.
Force use of the C locale. Update Perl dependency comment.
* manual/header.texi: Update reference to summary.awk.
* manual/macros.texi: Refer authors to `summary.pl --help'.
* manual/summary.awk: Remove file.
* manual/summary.pl: New file. Generate summary.texi, and
check for @standards-related syntax errors.
* manual/argp.texi: Convert header and standards @comments to
@standards.
* manual/arith.texi: Likewise.
* manual/charset.texi: Likewise.
* manual/conf.texi: Likewise.
* manual/creature.texi: Likewise.
* manual/crypt.texi: Likewise.
* manual/ctype.texi: Likewise.
* manual/debug.texi: Likewise.
* manual/errno.texi: Likewise.
* manual/filesys.texi: Likewise.
* manual/getopt.texi: Likewise.
* manual/job.texi: Likewise.
* manual/lang.texi: Likewise.
* manual/llio.texi: Likewise.
* manual/locale.texi: Likewise.
* manual/math.texi: Likewise.
* manual/memory.texi: Likewise.
* manual/message.texi: Likewise.
* manual/pattern.texi: Likewise.
* manual/pipe.texi: Likewise.
* manual/process.texi: Likewise.
* manual/resource.texi: Likewise.
* manual/search.texi: Likewise.
* manual/setjmp.texi: Likewise.
* manual/signal.texi: Likewise.
* manual/socket.texi: Likewise.
* manual/startup.texi: Likewise.
* manual/stdio.texi: Likewise.
* manual/string.texi: Likewise.
* manual/sysinfo.texi: Likewise.
* manual/syslog.texi: Likewise.
* manual/terminal.texi: Likewise.
* manual/threads.texi: Likewise.
* manual/time.texi: Likewise.
* manual/users.texi: Likewise.
Header and standards annotations are slated for standardization,
including being rendered in the description of functions, variables,
etc. (elements), and eventually required. This commit adds @standards
dummy macros so we can convert all existing annotations to the new
framework while maintaining the rendered status quo.
There needs to be a way to disambiguate annotations in lists of @*x
elements, where a common description is shared but some elements may
have different headers or standards. The @standardsx macro fills this
role by accepting an additional parameter: the name of the annotated
element.
* manual/macros.texi (@standards): New macro. Provide
placeholder for header and standards annotations.
(@standardsx): New macro. Likewise, for lists of @*x
elements.
This patch consolidates Linux sync_file_range at default
sysdeps/unix/sysv/linux/sync_file_range.c implementation. It also
moves the rules flags from generic io/Makefile to Linux one due the
fact it is a Linux-only symbol.
Checked on i686-linux-gnu and x86_64-linux-gnu.
* io/Makefile (CFLAGS-sync_file_range.c): Remove rule.
* sysdeps/unix/sysv/linux/Makefile (CFLAGS-sync_file_range.c): New
rule.
* sysdeps/unix/sysv/linux/wordsize-64/syscalls.list: Remove
sync_file_range.
This patch adds an option to test to add small command line option
through CMDLINE_OPTSTRING define. For instance:
#define CMDLINE_OPTSTRING "vd"
static void
cmdline_process_function (int c)
{
switch (c):
'v':
/* process '-v' option. */
break;
'd':
/* process '-d' option. */
break;
}
#define CMDLINE_PROCESS cmdline_process_function
It will add both '-v' and '-d' along with already default long options.
* support/support_test_main.c (support_test_main): Use optstring
member for option string in getopt_long.
* support/test-driver.c: Add comment about CMDLINE_OPTSTRING.
(CMDLINE_OPTSTRING): New define.
* support/test-driver.h (test_config): Add optstring member.
Implement strcspn/strpbrk/strspn IFUNC selectors in C
All internal calls within libc.so can use IFUNC on x86-64 since unlike
x86, x86-64 supports PC-relative addressing to access the GOT entry so
that it can call via PLT without using an extra register. For libc.a,
we can't use IFUNC for functions which are called before IFUNC has been
initialized. Use IFUNC internally reduces the icache footprint since
libc.so and other codes in the process use the same implementations.
This patch uses IFUNC for strcspn/strpbrk/strspn functions within libc.
* sysdeps/x86_64/multiarch/Makefile (sysdep_routines): Add
strcspn-sse2, strpbrk-sse2 and strspn-sse2.
* sysdeps/x86_64/strcspn.S (STRPBRK_P): Removed.
Check USE_AS_STRPBRK instead of STRPBRK_P.
* sysdeps/x86_64/strpbrk.S (USE_AS_STRPBRK): New.
* sysdeps/x86_64/multiarch/ifunc-sse4_2.h: New file.
* sysdeps/x86_64/multiarch/strcspn-sse2.S: Likewise.
* sysdeps/x86_64/multiarch/strcspn.c: Likewise.
* sysdeps/x86_64/multiarch/strpbrk-sse2.S: Likewise.
* sysdeps/x86_64/multiarch/strpbrk.c: Likewise.
* sysdeps/x86_64/multiarch/strspn-sse2.S: Likewise.
* sysdeps/x86_64/multiarch/strspn.c: Likewise.
* sysdeps/x86_64/multiarch/strcspn.S: Removed.
* sysdeps/x86_64/multiarch/strpbrk.S: Likewise.
* sysdeps/x86_64/multiarch/strspn.S: Likewise.
* sysdeps/x86_64/multiarch/strpbrk-c.c: Remove "#ifdef SHARED"
and "#endif".
Implement strcat family IFUNC selectors in C.
All internal calls within libc.so can use IFUNC on x86-64 since unlike
x86, x86-64 supports PC-relative addressing to access the GOT entry so
that it can call via PLT without using an extra register. For libc.a,
we can't use IFUNC for functions which are called before IFUNC has been
initialized. Use IFUNC internally reduces the icache footprint since
libc.so and other codes in the process use the same implementations.
This patch uses IFUNC for strcat family functions within libc.
* sysdeps/x86_64/multiarch/Makefile (sysdep_routines): Add
strcat-sse2.
* sysdeps/x86_64/multiarch/strcat-sse2.S: New file.
* sysdeps/x86_64/multiarch/strcat.c: Likewise.
* sysdeps/x86_64/multiarch/strncat.c: Likewise.
* sysdeps/x86_64/multiarch/strcat.S: Removed.
* sysdeps/x86_64/multiarch/strncat.S: Likewise.
Implement memcmp family IFUNC selectors in C.
All internal calls within libc.so can use IFUNC on x86-64 since unlike
x86, x86-64 supports PC-relative addressing to access the GOT entry so
that it can call via PLT without using an extra register. For libc.a,
we can't use IFUNC for functions which are called before IFUNC has been
initialized. Use IFUNC internally reduces the icache footprint since
libc.so and other codes in the process use the same implementations.
This patch uses IFUNC for memcmp family functions within libc.
* sysdeps/x86_64/multiarch/Makefile (sysdep_routines): Add
memcmp-sse2.
* sysdeps/x86_64/multiarch/ifunc-memcmp.h: New file.
* sysdeps/x86_64/multiarch/memcmp-sse2.S: Likewise.
* sysdeps/x86_64/multiarch/memcmp.c: Likewise.
* sysdeps/x86_64/multiarch/wmemcmp.c: Likewise.
* sysdeps/x86_64/multiarch/memcmp.S: Removed.
* sysdeps/x86_64/multiarch/wmemcmp.S: Likewise.
Implement memset family IFUNC selectors in C.
All internal calls within libc.so can use IFUNC on x86-64 since unlike
x86, x86-64 supports PC-relative addressing to access the GOT entry so
that it can call via PLT without using an extra register. For libc.a,
we can't use IFUNC for functions which are called before IFUNC has been
initialized. Use IFUNC internally reduces the icache footprint since
libc.so and other codes in the process use the same implementations.
This patch uses IFUNC for memset functions within libc.
2017-06-07 H.J. Lu <hongjiu.lu@intel.com>
Erich Elsen <eriche@google.com>
* sysdeps/x86_64/multiarch/Makefile (sysdep_routines): Add
memset-sse2-unaligned-erms, and memset_chk-nonshared.
* sysdeps/x86_64/multiarch/ifunc-impl-list.c
(__libc_ifunc_impl_list): Add test for __memset_chk_erms.
Update comments.
* sysdeps/x86_64/multiarch/ifunc-memset.h: New file.
* sysdeps/x86_64/multiarch/memset-sse2-unaligned-erms.S: Likewise.
* sysdeps/x86_64/multiarch/memset.c: Likewise.
* sysdeps/x86_64/multiarch/memset_chk-nonshared.S: Likewise.
* sysdeps/x86_64/multiarch/memset_chk.c: Likewise.
* sysdeps/x86_64/multiarch/memset.S: Removed.
* sysdeps/x86_64/multiarch/memset_chk.S: Likewise.
* sysdeps/x86_64/multiarch/memset-vec-unaligned-erms.S
(__memset_chk_erms): New function.
All of the major architectures are adopting tunables as a way to add
tuning to the library, from hwcap_mask for aarch64 to HLE for s390 and
ifunc and cache geometry for x86. Given this adoption and the fact
that we don't want additional tuning knobs to be added outside of
tunables, it makes sense to enable tunables by default using this
trivial patch.
Smoke tested on x86 to ensure that tunables code was built without
specifying it as a configure flag. I have kept it as --enabled and
not changed it to --disable since we want to still keep the option of
different kinds of front-ends for tunables.
* configure.ac(--enable-tunables): Enable by default.
* configure: Regenerate.
* NEWS: Mention change.
* manual/install.texi (enable-tunables): Adjust documentation.
* INSTALL: Regenerate.
Current allocate_stack logic for create stacks is to first mmap all
the required memory with the desirable memory and then mprotect the
guard area with PROT_NONE if required. Although it works as expected,
it pessimizes the allocation because it requires the kernel to actually
increase commit charge (it counts against the available physical/swap
memory available for the system).
The only issue is to actually check this change since side-effects are
really Linux specific and to actually account them it would require a
kernel specific tests to parse the system wide information. On the kernel
I checked /proc/self/statm does not show any meaningful difference for
vmm and/or rss before and after thread creation. I could only see
really meaningful information checking on system wide /proc/meminfo
between thread creation: MemFree, MemAvailable, and Committed_AS shows
large difference without the patch. I think trying to use these
kind of information on a testcase is fragile.
The BZ#18988 reports shows that the commit pages are easily seen with
mlockall (MCL_FUTURE) (with lock all pages that become mapped in the
process) however a more straighfoward testcase shows that pthread_create
could be faster using this patch:
--
static const int inner_count = 256;
static const int outer_count = 128;
static
void *thread1(void *arg)
{
return NULL;
}
static
void *sleeper(void *arg)
{
pthread_t ts[inner_count];
for (int i = 0; i < inner_count; i++)
pthread_create (&ts[i], &a, thread1, NULL);
for (int i = 0; i < inner_count; i++)
pthread_join (ts[i], NULL);
return NULL;
}
int main(void)
{
pthread_attr_init(&a);
pthread_attr_setguardsize(&a, 1<<20);
pthread_attr_setstacksize(&a, 1134592);
pthread_t ts[outer_count];
for (int i = 0; i < outer_count; i++)
pthread_create(&ts[i], &a, sleeper, NULL);
for (int i = 0; i < outer_count; i++)
pthread_join(ts[i], NULL);
assert(r == 0);
}
return 0;
}
--
On x86_64 (4.4.0-45-generic, gcc 5.4.0) running the small benchtests
I see:
$ time ./test
real 0m3.647s
user 0m0.080s
sys 0m11.836s
While with the patch I see:
$ time ./test
real 0m0.696s
user 0m0.040s
sys 0m1.152s
So I added a pthread_create benchtest (thread_create) which check
the thread creation latency. As for the simple benchtests, I saw
improvements in thread creation on all architectures I tested the
change.
Checked on x86_64-linux-gnu, i686-linux-gnu, aarch64-linux-gnu,
arm-linux-gnueabihf, powerpc64le-linux-gnu, sparc64-linux-gnu,
and sparcv9-linux-gnu.
[BZ #18988]
* benchtests/thread_create-inputs: New file.
* benchtests/thread_create-source.c: Likewise.
* support/xpthread_attr_setguardsize.c: Likewise.
* support/Makefile (libsupport-routines): Add
xpthread_attr_setguardsize object.
* support/xthread.h: Add xpthread_attr_setguardsize prototype.
* benchtests/Makefile (bench-pthread): Add thread_create.
* nptl/allocatestack.c (allocate_stack): Call mmap with PROT_NONE and
then mprotect the required area.
Implement memmove family IFUNC selectors in C.
All internal calls within libc.so can use IFUNC on x86-64 since unlike
x86, x86-64 supports PC-relative addressing to access the GOT entry so
that it can call via PLT without using an extra register. For libc.a,
we can't use IFUNC for functions which are called before IFUNC has been
initialized. Use IFUNC internally reduces the icache footprint since
libc.so and other codes in the process use the same implementations.
This patch uses IFUNC for memmove family functions within libc.
* sysdeps/x86_64/multiarch/Makefile (sysdep_routines): Add
memmove-sse2-unaligned-erms, memcpy_chk-nonshared,
mempcpy_chk-nonshared and memmove_chk-nonshared.
* sysdeps/x86_64/multiarch/ifunc-impl-list.c
(__libc_ifunc_impl_list): Add tests for __memmove_chk_erms,
__memcpy_chk_erms and __mempcpy_chk_erms. Update comments.
* sysdeps/x86_64/multiarch/ifunc-memmove.h: New file.
* sysdeps/x86_64/multiarch/memcpy.c: Likewise.
* sysdeps/x86_64/multiarch/memcpy_chk-nonshared.S: Likewise.
* sysdeps/x86_64/multiarch/memcpy_chk.c: Likewise.
* sysdeps/x86_64/multiarch/memmove-sse2-unaligned-erms.S: Likewise.
* sysdeps/x86_64/multiarch/memmove.c: Likewise.
* sysdeps/x86_64/multiarch/memmove_chk-nonshared.S: Likewise.
* sysdeps/x86_64/multiarch/memmove_chk.c: Likewise.
* sysdeps/x86_64/multiarch/mempcpy.c: Likewise.
* sysdeps/x86_64/multiarch/mempcpy_chk-nonshared.S: Likewise.
* sysdeps/x86_64/multiarch/mempcpy_chk.c: Likewise.
* sysdeps/x86_64/multiarch/memcpy.S: Removed.
* sysdeps/x86_64/multiarch/memcpy_chk.S: Likewise.
* sysdeps/x86_64/multiarch/memmove.S: Likewise.
* sysdeps/x86_64/multiarch/memmove_chk.S: Likewise.
* sysdeps/x86_64/multiarch/mempcpy.S: Likewise.
* sysdeps/x86_64/multiarch/mempcpy_chk.S: Likewise.
* sysdeps/x86_64/multiarch/memmove-vec-unaligned-erms.S
(__mempcpy_chk_erms): New function.
(__memmove_chk_erms): Likewise.
(__memcpy_chk_erms): New alias.
Since commit d957c4d3fa (i386: Compile
rtld-*.os with -mno-sse -mno-mmx -mfpmath=387), vector intrinsics can
no longer be used in ld.so, even if the compiled code never makes it
into the final ld.so link. This commit adds the missing IS_IN (libc)
guard to the SSE 4.2 strcspn implementation, so that it can be used from
ld.so in the future.
This is fairly complicated, not because the users of __need_Emath and
__need_error_t have complicated requirements, but because the core
changes had a lot of fallout.
__need_error_t exists for gnulib compatibility in argz.h and argp.h.
error_t itself is a Hurdism, an enum containing all the E-constants,
so you can do 'p (error_t) errno' in gdb and get a symbolic value.
argz.h and argp.h use it for function return values, and they want to
fall back to 'int' when that's not available. There is no reason why
these nonstandard headers cannot just go ahead and include all of
errno.h; so we do that.
__need_Emath is defined only by .S files; what they _really_ need is
for errno.h to avoid declaring anything other than the E-constants
(e.g. 'extern int __errno_location(void);' is a syntax error in
assembly language). This is replaced with a check for __ASSEMBLER__ in
errno.h, plus a carefully documented requirement for bits/errno.h not
to define anything other than macros. That in turn has the
consequence that bits/errno.h must not define errno - fortunately, all
live ports use the same definition of errno, so I've moved it to
errno.h. The Hurd bits/errno.h must also take care not to define
error_t when __ASSEMBLER__ is defined, which involves repeating all of
the definitions twice, but it's a generated file so that's okay.
* stdlib/errno.h: Remove __need_Emath and __need_error_t logic.
Reorganize file. Declare errno here. When __ASSEMBLER__ is
defined, don't declare anything other than the E-constants.
* include/errno.h: Change conditional for exposing internal
declarations to (not _ISOMAC and not __ASSEMBLER__).
* bits/errno.h: Remove logic for __need_Emath. Document
requirements for a port-specific bits/errno.h.
* sysdeps/unix/sysv/linux/bits/errno.h
* sysdeps/unix/sysv/linux/alpha/bits/errno.h
* sysdeps/unix/sysv/linux/hppa/bits/errno.h
* sysdeps/unix/sysv/linux/mips/bits/errno.h
* sysdeps/unix/sysv/linux/sparc/bits/errno.h:
Add multiple-include guard and check against improper inclusion.
Remove __need_Emath logic. Don't declare errno here. Ensure all
constants are defined as simple integer literals. Consistent
formatting.
* sysdeps/mach/hurd/errnos.awk: Likewise. Only define error_t and
enum __error_t_codes if __ASSEMBLER__ is not defined.
* sysdeps/mach/hurd/bits/errno.h: Regenerate.
* argp/argp.h, string/argz.h: Don't define __need_error_t before
including errno.h.
* sysdeps/i386/i686/fpu/multiarch/s_cosf-sse2.S
* sysdeps/i386/i686/fpu/multiarch/s_sincosf-sse2.S
* sysdeps/i386/i686/fpu/multiarch/s_sinf-sse2.S
* sysdeps/x86_64/fpu/s_cosf.S
* sysdeps/x86_64/fpu/s_sincosf.S
* sysdeps/x86_64/fpu/s_sinf.S:
Just include errno.h; don't define __need_Emath or include
bits/errno.h directly.
