Szabolcs Nagy [Mon, 4 Sep 2017 16:55:33 +0000 (17:55 +0100)]
New generic powf
without wrapper on aarch64:
powf reciprocal-throughput: 4.2x faster
powf latency: 2.6x faster
old worst-case error: 1.11 ulp
new worst-case error: 0.82 ulp
aarch64 .text size: -780 bytes
aarch64 .rodata size: +144 bytes
powf(x,y) is implemented as exp2(y*log2(x)) with the same algorithms
that are used in exp2f and log2f, except that the log2f polynomial is
larger for extra precision and its output (and exp2f input) may be
scaled by a power of 2 (POWF_SCALE) to simplify the argument reduction
step of exp2 (possible when efficient round and convert toint operation
is available).
The special case handling tries to minimize the checks in the hot path.
When the input of exp2_inline is checked, int arithmetics is used as
that was faster on the tested aarch64 cores.
2017-09-19 Szabolcs Nagy <szabolcs.nagy@arm.com>
* math/Makefile (type-float-routines): Add e_powf_log2_data.
* sysdeps/ieee754/flt-32/e_powf.c: New implementation.
* sysdeps/ieee754/flt-32/e_powf_log2_data.c: New file.
* sysdeps/ieee754/flt-32/math_config.h (__powf_data): Define.
(issignalingf_inline): Likewise.
(POWF_LOG2_TABLE_BITS): Likewise.
(POWF_LOG2_POLY_ORDER): Likewise.
(POWF_SCALE_BITS): Likewise.
(POWF_SCALE): Likewise.
* sysdeps/i386/fpu/e_powf_log2_data.c: New file.
* sysdeps/ia64/fpu/e_powf_log2_data.c: New file.
* sysdeps/m68k/m680x0/fpu/e_powf_log2_data.c: New file.
H.J. Lu [Thu, 28 Sep 2017 22:28:00 +0000 (15:28 -0700)]
x86: Allow undefined _DYNAMIC in static executable
When --enable-static-pie is used to build static PIE, _DYNAMIC is used
to compute the load address of static PIE. But _DYNAMIC is undefined
when creating static executable. This patch makes _DYNAMIC weak in PIE
libc.a so that it can be undefined.
* sysdeps/i386/dl-machine.h (elf_machine_load_address): Allow
undefined _DYNAMIC in PIE libc.a.
* sysdeps/x86_64/dl-machine.h (elf_machine_load_address):
Likewse.
Remove __signbit inlines from mathinline.h. Math.h already uses
the builtin when supported, so additional inlines are only used
on pre 4.0 GCCs. Similarly remove ancient copysign and fabs
inlines.
Simplify the C99 isgreater macros. Although some support was added
in GCC 2.97, not all targets added support until GCC 3.1. Therefore
only use the builtins in math.h from GCC 3.1 onwards, and defer to
generic macros otherwise. Improve the generic isunordered macro
to use compares rather than call fpclassify twice - this is not only
faster but also correct for signaling NaNs.
Joseph Myers [Thu, 28 Sep 2017 01:59:02 +0000 (01:59 +0000)]
Fix nearbyint arithmetic moved before feholdexcept (bug 22225).
In <https://sourceware.org/ml/libc-alpha/2013-05/msg00722.html> I
remarked on the possibility of arithmetic in various nearbyint
implementations being scheduled before feholdexcept calls, resulting
in spurious "inexact" exceptions.
I'm now actually observing this occurring in glibc built for ARM with
GCC 7 (in fact, both copies of the same addition/subtraction sequence
being combined and moved out before the conditionals and
feholdexcept/fesetenv pairs), resulting in test failures.
This patch makes the nearbyint implementations with this particular
feholdexcept / arithmetic / fesetenv pattern consistently use
math_opt_barrier on the function argument when first used in
arithmetic, and also consistently use math_force_eval before fesetenv
(the latter was generally already done, but the dbl-64/wordsize-64
implementation used math_opt_barrier instead, and as
math_opt_barrier's intended effect is through its output value being
used, such a use that doesn't use the return value is suspect).
Tested for x86_64 (--disable-multi-arch so more of these
implementations get used), and for ARM in a configuration where I saw
the problem scheduling.
[BZ #22225]
* sysdeps/ieee754/dbl-64/s_nearbyint.c (__nearbyint): Use
math_opt_barrier on argument when doing arithmetic on it.
* sysdeps/ieee754/dbl-64/wordsize-64/s_nearbyint.c (__nearbyint):
Likewise. Use math_force_eval not math_opt_barrier after
arithmetic.
* sysdeps/ieee754/flt-32/s_nearbyintf.c (__nearbyintf): Use
math_opt_barrier on argument when doing arithmetic on it.
* sysdeps/ieee754/ldbl-128/s_nearbyintl.c (__nearbyintl):
Likewise.
Samuel Thibault [Wed, 27 Sep 2017 22:19:18 +0000 (00:19 +0200)]
hurd: Fix `getifaddrs' and `freeifaddrs' symbol exposition
from `freeaddrinfo'.
`getifaddrs' and `freeifaddrs' are not in POSIX, they should not be
exposed along `freeaddrinfo' (through `__check_pf') which is POSIX.
* include/ifaddrs.h (__getifaddrs, __freeifaddrs): New declarations,
and use libc_hidden_def on them.
* inet/ifaddrs.c (__getifaddrs, __freeifaddrs): Use libc_hidden_def on
them.
* sysdeps/gnu/ifaddrs.c (__getifaddrs, __freeifaddrs): Likewise.
* inet/check_pf.c (__check_pf): Use __getifaddrs and __freeifaddrs
instead of getifaddrs and freeifaddrs.
Samuel Thibault [Wed, 27 Sep 2017 22:41:02 +0000 (00:41 +0200)]
hurd: Fix `seekdir' symbol exposition from `rewinddir'
`seekdir' is MISC || XOPEN, it should not be exposed along `rewinddir' which
is POSIX.
* include/dirent.h (__seekdir): New declaration.
* sysdeps/mach/hurd/seekdir.c (seekdir): Rename to __seekdir and
redefine as weak alias.
* sysdeps/mach/hurd/rewinddir.c (__rewinddir): Use __seekdir instead
of seekdir.
Samuel Thibault [Wed, 27 Sep 2017 22:01:40 +0000 (00:01 +0200)]
hurd: Fix `revoke' symbol exposition from `unlockpt'
`revoke' is MISC only, it should not be exposed along `unlockpt' which is
XOPEN.
* include/unistd.h (__revoke): New declaration.
* misc/revoke.c (revoke): Rename to __revoke, and redefine as weak
alias.
* sysdeps/mach/hurd/revoke.c (revoke): Likewise.
* sysdeps/unix/bsd/unlockpt.c (unlockpt): Use __revoke instead of
revoke.
Joseph Myers [Tue, 26 Sep 2017 21:21:01 +0000 (21:21 +0000)]
Fix make-syscalls.sh VDSO support for GCC 8.
sysdeps/unix/make-syscalls.sh has support, used only by x32, for
generating IFUNCs for kernel VDSO symbols. This support creates
IFUNCs by setting symbol types manually, which is bad for debug info
and does not work with current GCC mainline because it results in
errors from the checks on types of function aliases.
This patch fixes it to use the common __ifunc macro, which uses the
ifunc attribute when available and so works with GCC mainline. Note
however that the original error resulted from an indirect inclusion of
a header declaring __gettimeofday from the generated sources, and
using __ifunc now relies on such an indirect inclusion remaining as it
means use of __typeof to determine the correct types. If glibc's
headers change in such a way as to remove that indirect inclusion, it
will become necessary to change the syscalls.list syntax for VDSO
syscalls so the name of the header to include can be specified.
Tested (compilation only) with build-many-glibcs.py that this fixes
the build for x32 with GCC mainline.
* sysdeps/unix/make-syscalls.sh: Use __ifunc to define symbols
using VDSO.
Current implementation of tunables does not set arena_max and arena_test
values. Any value provided by glibc.malloc.arena_max and
glibc.malloc.arena_test parameters is ignored.
These tunables have minval value set to 1 (see elf/dl-tunables.list file)
and undefined maxval value. In that case default value (which is 0. see
scripts/gen-tunables.awk) is being used to set maxval.
For instance, generated tunable_list[] entry for arena_max is:
(gdb) p *cur
$1 = {name = 0x7ffff7df6217 "glibc.malloc.arena_max",
type = {type_code = TUNABLE_TYPE_SIZE_T, min = 1, max = 0},
val = {numval = 0, strval = 0x0}, initialized = false,
security_level = TUNABLE_SECLEVEL_SXID_IGNORE,
env_alias = 0x7ffff7df622e "MALLOC_ARENA_MAX"}
As a result, any value of glibc.malloc.arena_max is ignored by
TUNABLE_SET_VAL_IF_VALID_RANGE macro
__type min = (__cur)->type.min; <- initialized to 1
__type max = (__cur)->type.max; <- initialized to 0!
if (min == max) <- false
{
min = __default_min;
max = __default_max;
}
if ((__type) (__val) >= min && (__type) (val) <= max) <- false
{
(__cur)->val.numval = val;
(__cur)->initialized = true;
}
Assigning correct min/max values at a build time fixes a problem.
Plus, a bit of optimization: Setting of default min/max values for the
given type at a run time might be eliminated.
* elf/dl-tunables.c (do_tunable_update_val): Range checking fix.
* scripts/gen-tunables.awk: Set unspecified minval and/or maxval
values to correct default value for given type.
with 0 file size. ld.so should skip such PT_DYNAMIC segments.
Without a PT_DYNAMIC segment the loading of the shared object will
fail, and therefore ldd on such objects will also fail instead of
crashing. This provides better diagnostics for tooling that is
attempting to inspect the invalid shared objects which may just
contain debug information.
Joseph Myers [Tue, 26 Sep 2017 19:49:33 +0000 (19:49 +0000)]
Use generic __ifunc for SPARC.
glibc fails to build with GCC mainline for SPARC because of the use of
manually-created IFUNCs, which fail the tests of compatibility of
function alias types. This patch changes sparc-ifunc.h to use the
generic __ifunc in defining sparc_libm_ifunc. The generic __ifunc can
use the GCC ifunc attribute when available, so ensuring
type-correctness as well as better debug info than when setting symbol
types in asm statements.
Note that for this to fix the build with GCC mainline the GCC patch
<https://gcc.gnu.org/ml/gcc-patches/2017-09/msg01779.html>, or
building GCC with --enable-gnu-indirect-function, is also needed.
Tested (compilation only) with build-many-glibcs.py (sparc64-linux-gnu
and sparcv9-linux-gnu, with GCC 8 with the above patch, and also with
GCC 7).
* sysdeps/sparc/sparc-ifunc.h [!__ASSEMBLER__] (sparc_libm_ifunc):
Define using __ifunc.
Joseph Myers [Tue, 26 Sep 2017 16:30:46 +0000 (16:30 +0000)]
Fix ia64 executable stack default (bug 22156).
As per https://gcc.gnu.org/ml/gcc-patches/2017-09/msg01220.html ia64
defaults to non-executable stacks in the Linux kernel (furthermore,
the use of function descriptors means that trampolines for nested
function pointers never need an executable stack). glibc however
defines DEFAULT_STACK_PERMS to include PF_X for that architecture,
meaning (a) elf/check-execstack fails and (b) (from code inspection,
not tested, but this is why I think this is a user-visible bug) thread
stacks are unnecessarily mapped with execute permission. This patch
fixes the DEFAULT_STACK_PERMS definition in question.
Tested (compilation only) with build-many-glibcs.py for ia64. This
fixes the check-execstack failure.
Problem reported by Florian Weimer [1] and solution suggested by
Andreas Schwab [2]. It also set the same buffer size independent
of architecture max_align_t size.
Checked on x86_64-linux-gnu and i686-linux-gnu.
* lib/malloc/scratch_buffer.h (struct scratch_buffer):
Use an union instead of a max_align_t array for __space,
so that __space is the same size on all platforms.
* malloc/scratch_buffer_grow_preserve.c
(__libc_scratch_buffer_grow_preserve): Likewise.
posix: Add compat glob symbol to not follow dangling symbols
This patch follows commit 5554304f0 (posix: Allow glob to match dangling
symlinks [BZ #866]) by adding a compat symbol that follow previous
semantic of not following dangling symlinks and thus avoiding call
gl_lstat with GLOB_ALTDIRFUNC.
It avoids failure with old binaries that not set the alternate function
pointer for lstat (GNUmake for instance). The following scenario, for
instance, fails with current GNUmake because glibc will access unitialized
memory when calling gl_lstat:
obj/%.o: src/%.c
$(CC) $(CFLAGS) -c $< -o $@
$ make
This works as expected with the patch applied. Since it is for generic
ABI, default compat symbols are added with override for Linux due LFS.
Now we have two compat symbols for glob on Linux:
1. sysdeps/unix/sysv/linux/oldglob.c which implements glob64 with
the old dirent layout. For this implementation I also set it to
not follow dangling symlinks (which is the safest path).
2. sysdeps/unix/sysv/linux/glob{64}-lstat-compat.c which implements
the compat symbol for dangling symlinks. As for generic glob,
the implementation uses XSTAT_IS_XSTAT64 to define whether
both __glob_lstat_compat and __glob64_lstat_compat should be
different implementations. For archictures that define
XSTAT_IS_XSTAT64, __glob_lstat_compat is aliased to
__glob64_lstat_compat.
3. sysdeps/unix/sysv/linux/alpha/oldglob.c with a different glob_t
layout. As for 1. this patch changes it to not follow dangling
symlinks.
The patch also bumps _GNU_GLOB_INTERFACE_VERSION to 2 to advertise the
new semantic. On GNUmake, for instance, it will force to it use its
internal glob implementation instead and avoiding triggering the same
failure on builds against newer GLIBCs.
Checked on x86_64-linux-gnu and i686-linux-gnu. I also checked
with a build against the major ABIs required to check for the abilist.
The changes should also work on gnulib (I run gnulib-tool.py check glob
and it shown no regressions).
Paul Pluzhnikov [Tue, 26 Sep 2017 00:51:34 +0000 (17:51 -0700)]
Reduce total memory required to create all threads to 128MiB. This fixes
intermittent failure in stdlib/test-{atexit,at_quick_exit,...} tests
(Bug 22207).
Reviewed-by: Jonathan Nieder <jrnieder@gmail.com> Reviewed-by: Andreas Schwab <schwab@suse.de>
A lock is held by mempool_allocate() when CACHEABLE is true; we
must release this lock if we exit early.
Szabolcs Nagy [Tue, 12 Sep 2017 11:20:50 +0000 (12:20 +0100)]
New expf and exp2f version without SVID compat wrapper
This patch changes the expf and exp2f error handling semantics to only
set errno accoring to POSIX rules. New symbol version is introduced at
GLIBC_2.27.
The old wrappers are kept for compat symbols.
Internal calls to __expf now get the new error semantics, this seems to
only affect sysdeps/i386/fpu/s_expm1f.S where the errno-only behaviour
should be correct.
ia64 needed assembly change to have the new and compat versioned symbol
map to the same function.
Szabolcs Nagy [Wed, 6 Sep 2017 16:42:00 +0000 (17:42 +0100)]
Optimized generic expf and exp2f with wrappers
Based on new expf and exp2f code from
https://github.com/ARM-software/optimized-routines/
with wrapper on aarch64:
expf reciprocal-throughput: 2.3x faster
expf latency: 1.7x faster
without wrapper on aarch64:
expf reciprocal-throughput: 3.3x faster
expf latency: 1.7x faster
without wrapper on aarch64:
exp2f reciprocal-throughput: 2.8x faster
exp2f latency: 1.3x faster
libm.so size on aarch64:
.text size: -152 bytes
.rodata size: -1740 bytes
expf/exp2f worst case nearest rounding error: 0.502 ulp
worst case non-nearest rounding error: 1 ulp
Error checks are inline and errno setting is in separate tail called
functions, but the wrappers are kept in this patch to handle the
_LIB_VERSION==_SVID_ case. (So e.g. errno is set twice for expf calls
and once for __expf_finite calls on targets where the new code is used.)
Double precision arithmetics is used which is expected to be faster on
most targets (including soft-float) than using single precision and it
is easier to get good precision result with it.
Const data is kept in a separate translation unit which complicates
maintenance a bit, but is expected to give good code for literal loads
on most targets and allows sharing data across expf, exp2f and powf.
(This data is disabled on i386, m68k and ia64 which have their own
expf, exp2f and powf code.)
Some details may need target specific tweaks:
- best convert and round to int operation in the arg reduction may be
different across targets.
- code was optimized on fma target, optimal polynomial eval may be
different without fma.
- gcc does not always generate good code for fp bit representation
access via unions or it may be inherently slow on some targets.
The libm-test-ulps will need adjustment because..
- The argument reduction ideally uses nearest rounded rint, but that is
not efficient on most targets, so the polynomial can get evaluated on a
wider interval in non-nearest rounding mode making 1 ulp errors common
in that case.
- The polynomial is evaluated such that it may have 1 ulp error on
negative tiny inputs with upward rounding.
* math/Makefile (type-float-routines): Add math_errf and e_exp2f_data.
* sysdeps/aarch64/fpu/math_private.h (TOINT_INTRINSICS): Define.
(roundtoint, converttoint): Likewise.
* sysdeps/ieee754/flt-32/e_expf.c: New implementation.
* sysdeps/ieee754/flt-32/e_exp2f.c: New implementation.
* sysdeps/ieee754/flt-32/e_exp2f_data.c: New file.
* sysdeps/ieee754/flt-32/math_config.h: New file.
* sysdeps/ieee754/flt-32/math_errf.c: New file.
* sysdeps/ieee754/flt-32/t_exp2f.h: Remove.
* sysdeps/i386/fpu/e_exp2f_data.c: New file.
* sysdeps/i386/fpu/math_errf.c: New file.
* sysdeps/ia64/fpu/e_exp2f_data.c: New file.
* sysdeps/ia64/fpu/math_errf.c: New file.
* sysdeps/m68k/m680x0/fpu/e_exp2f_data.c: New file.
* sysdeps/m68k/m680x0/fpu/math_errf.c: New file.
Samuel Thibault [Sun, 24 Sep 2017 23:55:02 +0000 (01:55 +0200)]
hurd: Fix exposition of s/gettimeofday through timespec_s/get
conform/ISO11/time.h/linknamespace complains that using timespec_get exposes
gettimeofday.
conform/POSIX/time.h/linknamespace complains that using clock_settime
exposes settimeofday.
* sysdeps/unix/clock_gettime.c (realtime_gettime, __clock_gettime): Use
__gettimeofday instead of gettimeofday.
* sysdeps/unix/clock_settime.c (__clock_settime): Use __settimeofday
instead of settimeofday.
Samuel Thibault [Sun, 10 Sep 2017 15:41:03 +0000 (17:41 +0200)]
resolv_test.c: also cope with CONNREFUSED errors returned by recvfrom
server_thread_udp_process_one already takes care of calling sendto()
instead of xsendto to be able to ignore the case where the client has
closed the socket. Depending on the TCP/IP stack behavior, this error
could be notified later through recvfrom(), so we need to ignore it
there too.
* support/resolv_test.c (server_thread_udp_process_one): Call recvfrom
instead of xrecvfrom, and ignore ECONNREFUSED errors.
Samuel Thibault [Sun, 24 Sep 2017 20:20:10 +0000 (22:20 +0200)]
hurd: Fix bits/socket.h conformity
* sysdeps/mach/hurd/bits/socket.h: Include <bits/wordsize.h> instead
of <limits.h>
(__need_NULL): Do not define.
(__ss_aligntype): Use __WORDSIZE instead of ULONG_MAX to determine
alignment.
[!__USE_MISC] (pseudo_AF_XTP, pseudo_AF_RTIP, pseudo_AF_PIP,
CMGROUP_MAX, cmsgcred): Do not define.
(CMSG_FIRSTHDR, __cmsg_nxthdr): Use (struct cmsghdr *) 0 instead of
NULL.
* bits/socket.h: Likewise.
Samuel Thibault [Sun, 24 Sep 2017 15:54:02 +0000 (17:54 +0200)]
hurd: Make sure dl-sysdep.c defines proper symbol names
* sysdeps/mach/hurd/dl-sysdep.c (check_no_hidden): New macro.
(__open, __close, __libc_read, __libc_write, __writev, __libc_lseek64,
__mmap, __fxstat64, __xstat64, __access, __access_noerrno, __getpid,
__getcwd, __sbrk, __strtoul_internal, _exit, abort): Use check_no_hidden
to make sure that these symbols are defined.
Joseph Myers [Fri, 22 Sep 2017 20:24:12 +0000 (20:24 +0000)]
Use libm_alias_float in flt-32.
This patch makes flt-32 libm functions use libm_alias_float to define
public interfaces (in cases where _Float32 aliases of those interfaces
would be appropriate, so not for finitef / isinff / isnanf).
Tested for x86_64. Also tested with build-many-glibcs.py that
installed stripped shared libraries are unchanged by the patch.
* sysdeps/ieee754/flt-32/s_asinhf.c: Include <libm-alias-float.h>.
(asinhf): Define using libm_alias_float.
* sysdeps/ieee754/flt-32/s_atanf.c: Include <libm-alias-float.h>.
(atanf): Define using libm_alias_float.
* sysdeps/ieee754/flt-32/s_cbrtf.c: Include <libm-alias-float.h>.
(cbrtf): Define using libm_alias_float.
* sysdeps/ieee754/flt-32/s_ceilf.c: Include <libm-alias-float.h>.
(ceilf): Define using libm_alias_float.
* sysdeps/ieee754/flt-32/s_copysignf.c: Include
<libm-alias-float.h>.
(copysignf): Define using libm_alias_float.
* sysdeps/ieee754/flt-32/s_cosf.c: Include <libm-alias-float.h>.
(cosf): Define using libm_alias_float.
* sysdeps/ieee754/flt-32/s_erff.c: Include <libm-alias-float.h>.
(erff): Define using libm_alias_float.
(erfcf): Likewise.
* sysdeps/ieee754/flt-32/s_expm1f.c: Include <libm-alias-float.h>.
(expm1f): Define using libm_alias_float.
* sysdeps/ieee754/flt-32/s_fabsf.c: Include <libm-alias-float.h>.
(fabsf): Define using libm_alias_float.
* sysdeps/ieee754/flt-32/s_floorf.c: Include <libm-alias-float.h>.
(floorf): Define using libm_alias_float.
* sysdeps/ieee754/flt-32/s_frexpf.c: Include <libm-alias-float.h>.
(frexpf): Define using libm_alias_float.
* sysdeps/ieee754/flt-32/s_fromfpf.c (fromfpf): Define using
libm_alias_float.
* sysdeps/ieee754/flt-32/s_fromfpf_main.c: Include
<libm-alias-float.h>.
* sysdeps/ieee754/flt-32/s_fromfpxf.c (fromfpxf): Define using
libm_alias_float.
* sysdeps/ieee754/flt-32/s_getpayloadf.c: Include
<libm-alias-float.h>.
(getpayloadf): Define using libm_alias_float.
* sysdeps/ieee754/flt-32/s_llrintf.c: Include
<libm-alias-float.h>.
(llrintf): Define using libm_alias_float.
* sysdeps/ieee754/flt-32/s_llroundf.c: Include
<libm-alias-float.h>.
(llroundf): Define using libm_alias_float.
* sysdeps/ieee754/flt-32/s_logbf.c: Include <libm-alias-float.h>.
(logbf): Define using libm_alias_float.
* sysdeps/ieee754/flt-32/s_lrintf.c: Include <libm-alias-float.h>.
(lrintf): Define using libm_alias_float.
* sysdeps/ieee754/flt-32/s_lroundf.c: Include <libm-alias-float.h>.
(lroundf): Define using libm_alias_float.
* sysdeps/ieee754/flt-32/s_modff.c: Include <libm-alias-float.h>.
(modff): Define using libm_alias_float.
* sysdeps/ieee754/flt-32/s_nearbyintf.c: Include
<libm-alias-float.h>.
(nearbyintf): Define using libm_alias_float.
* sysdeps/ieee754/flt-32/s_nextafterf.c: Include
<libm-alias-float.h>.
(nextafterf): Define using libm_alias_float.
* sysdeps/ieee754/flt-32/s_nextupf.c: Include
<libm-alias-float.h>.
(nextupf): Define using libm_alias_float.
* sysdeps/ieee754/flt-32/s_remquof.c: Include
<libm-alias-float.h>.
(remquof): Define using libm_alias_float.
* sysdeps/ieee754/flt-32/s_rintf.c: Include <libm-alias-float.h>.
(rintf): Define using libm_alias_float.
* sysdeps/ieee754/flt-32/s_roundevenf.c: Include
<libm-alias-float.h>.
(roundevenf): Define using libm_alias_float.
* sysdeps/ieee754/flt-32/s_roundf.c: Include <libm-alias-float.h>.
(roundf): Define using libm_alias_float.
* sysdeps/ieee754/flt-32/s_setpayloadf.c (setpayloadf): Define
using libm_alias_float.
* sysdeps/ieee754/flt-32/s_setpayloadf_main.c: Include
<libm-alias-float.h>.
* sysdeps/ieee754/flt-32/s_setpayloadsigf.c (setpayloadsigf):
Define using libm_alias_float.
* sysdeps/ieee754/flt-32/s_sincosf.c: Include
<libm-alias-float.h>.
(sincosf): Define using libm_alias_float.
* sysdeps/ieee754/flt-32/s_sinf.c: Include <libm-alias-float.h>.
(sinf): Define using libm_alias_float.
* sysdeps/ieee754/flt-32/s_tanf.c: Include <libm-alias-float.h>.
(tanf): Define using libm_alias_float.
* sysdeps/ieee754/flt-32/s_tanhf.c: Include <libm-alias-float.h>.
(tanhf): Define using libm_alias_float.
* sysdeps/ieee754/flt-32/s_totalorderf.c: Include
<libm-alias-float.h>.
(totalorderf): Define using libm_alias_float.
* sysdeps/ieee754/flt-32/s_totalordermagf.c: Include
<libm-alias-float.h>.
(totalordermagf): Define using libm_alias_float.
* sysdeps/ieee754/flt-32/s_truncf.c: Include <libm-alias-float.h>.
(truncf): Define using libm_alias_float.
* sysdeps/ieee754/flt-32/s_ufromfpf.c (ufromfpf): Define using
libm_alias_float.
* sysdeps/ieee754/flt-32/s_ufromfpxf.c (ufromfpxf): Define using
libm_alias_float.
Let fpclassify use the builtin when optimizing for size in C++ mode (bug 22146)
When optimization for size is on (-Os), fpclassify does not use the
type-generic __builtin_fpclassify builtin, instead it uses __MATH_TG.
However, when library support for float128 is available, __MATH_TG uses
__builtin_types_compatible_p, which is not available in C++ mode.
On the other hand, libstdc++ undefines (in cmath) many macros from
math.h, including fpclassify, so that it can provide its own functions.
However, during its configure tests, libstdc++ just tests for the
availability of the macros (it does not undefine them, nor does it
provide its own functions).
Finally, when libstdc++ is configured with optimization for size
enabled, its configure tests include math.h and get the definition of
fpclassify that uses __MATH_TG (and __builtin_types_compatible_p).
Since libstdc++ does not undefine the macros during its configure tests,
they fail.
This patch lets fpclassify use the builtin in C++ mode, even when
optimization for size is on. This allows the configure test in
libstdc++ to work.
Tested for powerpc64le and x86_64.
[BZ #22146]
math/math.h: Let fpclassify use the builtin in C++ mode, even
when optimazing for size.
Joseph Myers [Fri, 22 Sep 2017 17:43:42 +0000 (17:43 +0000)]
Remove Banner mechanism.
This patch removes the Banner mechanism, with the last remaining
Banner file replaced by a contrib.texi entry. This accords with the
principle that the output of running libc.so.6 is not the place to
credit particular contributions (the manual is), and with all other
configuration options not mentioned there, it doesn't seem appropriate
to focus there on the one question of whether the one remaining piece
configured as an add-on was enabled or not.
Tested for x86_64.
* csu/Makefile (generated): Do not add version-info.h.
(before-compile): Likewise.
(all-Banner-files): Remove variable.
($(objpfx)version-info.h): Remove rule.
* csu/version.c (banner): Do not include "version-info.h".
* libidn/Banner: Remove.
* manual/contrib.texi (Simon Josefsson): New entry.
Remove conditional on LDBL_MANT_DIG from e_lgammal_r.c
The IEEE 754 implementation of lgammal in sysdeps/ieee754/ldbl-128/ used
to be shared by IBM's implementation in sysdeps/ieee754/ldbl-128ibm/ (by
an inclusion of the source file). In order for the algorithm to work
for IBM's implementation, a check for LDBL_MANT_DIG was required. Since
the source file is no longer shared, the requirement for the check is
gone. This patch removes the conditionals.
ldbl-128ibm: Automatic replacing of _Float128 and L()
The ldbl-128ibm implementation of j0l, j1l, lgammal_r, and cbrtl, as
well as the tables used by expl were copied from ldbl-128. However, the
original files used _Float128 for the type and L() for the literal
suffix. This patch uses the following sed command to rewrite _Float128
as long double and L(x) as xL (for e_expl.c, e_j0l.c, e_j1l.c,
e_lgammal_r.c, and t_expl.h):
sed -i <filename> \
-e "/^#define _Float128 long double/d" \
-e "/^#define L(x) x ## L/d" \
-e "/L(/s/)/L/" \
-e "/L(/s/L(//" \
-e "s/_Float128/long double/g"
For sysdeps/ieee754/ldbl-128ibm/s_cbrtl.c, this sed command incorrectly
replaces a few occurrences of L(), so the following command is used
instead:
* sysdeps/ieee754/ldbl-128ibm/e_expl.c: Remove definitions of
_Float128 and L().
* sysdeps/ieee754/ldbl-128ibm/e_j0l.c: Remove definitions of
_Float128 and L(). Replace _Float128 with long double and L(x)
with xL, throughout the file.
* sysdeps/ieee754/ldbl-128ibm/e_j1l.c: Likewise.
* sysdeps/ieee754/ldbl-128ibm/e_lgammal_r.c: Likewise.
* sysdeps/ieee754/ldbl-128ibm/s_cbrtl.c: Likewise.
* sysdeps/ieee754/ldbl-128ibm/t_expl.h: Likewise.
ldbl-128ibm: Copy implementations from ldbl-128 instead of including them
Some files under sysdeps/ieee754/ldbl-128ibm/ are able to reuse the
implementation in sysdeps/ieee754/ldbl-128/ by defining _Float128 to
long double. This relied on compiler support for _Float128 being
disabled. On powerpc, such support was disabled by default, however, it
got enabled by default [1] in GCC 8.
This patch copies the implementations from ldbl-128 to ldbl-128ibm. The
uses of _Float128 and L() are kept intact in this patch and are replaced
with a script in a subsequent patch.
* sysdeps/ieee754/ldbl-128ibm/e_expl.c: Include tables from
sysdeps/ieee754/ldbl-128ibm.
* sysdeps/ieee754/ldbl-128ibm/e_j0l.c: Copy contents from the
equivalent implementation in sysdeps/ieee754/ldbl-128/ instead
of including it. Keep _Float128 and L() intact. These will be
reviewed by a separate patch.
* sysdeps/ieee754/ldbl-128ibm/e_j1l.c: Likewise.
* sysdeps/ieee754/ldbl-128ibm/e_lgammal_r.c: Likewise.
* sysdeps/ieee754/ldbl-128ibm/s_cbrtl.c: Likewise.
* sysdeps/ieee754/ldbl-128ibm/t_expl.h: Likewise.
powerpc: Add redirection for finitef128, isinf128, and isnanf128
On powerpc64le, compiler support for float128 is not enabled by default
on gcc. To enable it, the flag -mfloat128 must be passed as a command
line option to the compiler. This means that only the few files that
actively have -mfloat128 passed as an argument get compiler support for
float128, whereas all other files don't.
When -mfloat128 becomes enabled by default on powerpc [1], all the files
that do not currently have compiler support for float128 enabled during
their compilation, will start to have it. This will lead to build
errors in s_finite.c, s_isinf.c, and s_isnan.c.
The errors are due to the unintended macro expansion of __finitef128 to
__redirect_finitef128 in math/bits/mathcalls-helper-functions.h. In
that header, __MATHDECL_1 takes '__finite' and 'f128' as arguments and
concatenates them. However, since '__finite' has been redefined in
s_finite.c, the function declaration becomes __redirect_finitef128:
extern int __redirect___finitef128 (_Float128 __value) __attribute__ ((__nothrow__ )) __attribute__ ((__const__));
This declaration itself is OK. The problem arises when include/math.h
creates the hidden prototype ('hidden_proto (__finitef128)'), which
expands to:
Since __finitef128 is not declared, __typeof fails. This effect was
already true for the 'float' and 'long double' versions and is now true
for float128. Likewise for isinsff128 and isnanf128.
This patch defines __finitef128 as __redirect___finitef128 in
sysdeps/powerpc/powerpc64/fpu/multiarch/s_finite.c, similarly to what's
done for the float and long double versions of these functions, to get
rid of the build error. Likewise for isinff128 and isnanf128.
powerpc64le: Add -mfloat128 to tst-strtod-nan-locale testcase
On powerpc64le, not all files can have the flag -mfloat128 passed as an
option on the compile command, since that could conflict with other
flags, such as -mno-vsx. Each file that needs the flag, gets it through
a CFLAGS-filename variable on sysdeps/powerpc/powerpc64le/Makefile.
The test cases tst-strtod-nan-locale and tst-wcstod-nan-locale are
missing this flag.
Tested for powerpc64le.
* sysdeps/powerpc/powerpc64le/Makefile
(CFLAGS-tst-strtod-nan-locale.c): New variable.
(CFLAGS-tst-wcstod-nan-locale.c): New variable.
Paul Pluzhnikov [Thu, 21 Sep 2017 19:14:41 +0000 (12:14 -0700)]
Fix BZ# 22180.
POSIX requires that dlclose() and exit() be thread safe, therefore
you can have one thread in the middle of dlclose() and another thread
executing exit() without causing any undefined behaviour on the part
of the implementation.
The existing implementation had a flaw that exit() exit handler processing
did not consider a concurrent dlclose() and would not mark already run
exit handlers using the ef_free flavour. The consequence of this is that
a concurrent exit() with dlclose() will run all the exit handlers that
dlclose() had not yet run, but then will block on the loader lock. The
concurrent dlclose() will continue to run all the exit handlers again
(twice) in violation of the Itanium C++ ABI requirements for __cxa_atexit().
This commit fixes this by having exit() mark all handlers with ef_free to
ensure that concurrent dlclose() won't re-run registered exit handlers that
have already run.
Joseph Myers [Thu, 21 Sep 2017 17:49:51 +0000 (17:49 +0000)]
Remove non-add-on Banner files.
Various subdirectories of glibc include Banner files to put some text
in the output of executing libc.so.6, under "Available extensions".
Some of those subdirectories (e.g. crypt) may originally have been
add-ons (and so optional, so a particular glibc build might or might
not have included them), but except for libidn they aren't now (or if
only included in some builds, in the case of soft-fp, the inclusion
depends on the architecture for which glibc is configured rather than
having any glibc configuration for which it's an optional feature),
and it doesn't seem useful for the libc.so.6 output to call out a few
features like that.
This patch removes the non-add-on Banner files, updating contrib.texi
where they noted contributions not otherwise mentioned there.
Tested for x86_64.
* crypt/Banner: Remove file.
* nptl/Banner: Likewise.
* resolv/Banner: Likewise.
* soft-fp/Banner: Likewise.
* nptl/Makefile ($(objpfx)banner.h): Remove rule.
($(objpfx)version.d): Remove dependency on banner.h.
($(objpfx)version.os): Likewise.
* nptl/version.c (banner): Do not include banner.h.
* manual/contrib.texi: Update entries for Richard Henderson, Jakub
Jelinek and BIND code.
Paul Pluzhnikov [Thu, 21 Sep 2017 02:37:45 +0000 (19:37 -0700)]
Implement xdlopen, xdlsym and xdlclose routines which terminate test
program with appropriate message if the corresponding dlfcn.h routine
returns an error.
Joseph Myers [Wed, 20 Sep 2017 16:54:05 +0000 (16:54 +0000)]
Add SSE4.1 trunc, truncf (bug 20142).
This patch adds SSE4.1 versions of trunc and truncf, using the roundsd
/ roundss instructions, similar to the versions of ceil, floor, rint
and nearbyint functions we already have. In my testing with the glibc
benchtests these are about 30% faster than the C versions for double,
20% faster for float.
Joseph Myers [Tue, 19 Sep 2017 15:50:38 +0000 (15:50 +0000)]
Fix fexecve build where syscall macros call sizeof.
The recent fexecve changes broke the build on (at least) alpha (maybe
other configurations, that was the first breakage I saw in my
build-many-glibcs.py run):
In file included from ../sysdeps/unix/sysv/linux/alpha/sysdep.h:29:0,
from ../sysdeps/alpha/nptl/tls.h:31,
from ../include/errno.h:25,
from ../sysdeps/unix/sysv/linux/fexecve.c:18:
../sysdeps/unix/sysv/linux/fexecve.c: In function 'fexecve':
../sysdeps/unix/alpha/sysdep.h:203:10: error: 'sizeof' on array function parameter 'argv' will return size of 'char * const*' [-Werror=sizeof-array-argument]
(sizeof(arg) == 4 ? (long)(int)(long)(arg) : (long)(arg))
^
../sysdeps/unix/alpha/sysdep.h:302:26: note: in expansion of macro 'syscall_promote'
register long _tmp_18 = syscall_promote (arg3); \
^~~~~~~~~~~~~~~
../sysdeps/unix/alpha/sysdep.h:173:2: note: in expansion of macro 'inline_syscall5'
inline_syscall##nr(__NR_##name, args); \
^~~~~~~~~~~~~~
../sysdeps/unix/sysv/linux/alpha/sysdep.h:85:2: note: in expansion of macro 'INLINE_SYSCALL1'
INLINE_SYSCALL1(name, nr, args); \
^~~~~~~~~~~~~~~
../sysdeps/unix/sysv/linux/fexecve.c:42:3: note: in expansion of macro 'INLINE_SYSCALL'
INLINE_SYSCALL (execveat, 5, fd, "", argv, envp, AT_EMPTY_PATH);
^~~~~~~~~~~~~~
../sysdeps/unix/sysv/linux/fexecve.c:33:30: note: declared here
fexecve (int fd, char *const argv[], char *const envp[])
^~~~
This patch fixes this similarly to previous fixes for such issues: use
&argv[0] and &envp[0] as the syscall macro arguments. Tested
(compilation only) for alpha-linux-gnu with build-many-glibcs.py.
* sysdeps/unix/sysv/linux/fexecve.c (fexecve) [__NR_execveat]:
Explicitly take address of first element of array arguments in
call to INLINE_SYSCALL.
Enable unwind info in libc-start.c and backtrace.c
Add unwind info to __libc_start_main so that unwinding continues one
extra level to _start. Similarly add unwind info to backtrace.
Given many targets require this, do this in a general way.
Joseph Myers [Tue, 19 Sep 2017 12:59:01 +0000 (12:59 +0000)]
Add benchtests for trunc and truncf.
This patch adds benchtests for the trunc and truncf functions. The
inputs listed are fairly arbitrary; I do not assert they are
representative of any particular application.
* benchtests/Makefile (bench-math): Add trunc and truncf.
(CFLAGS-bench-trunc.c): New variable.
(CFLAGS-bench-truncf.c): Likewise.
* benchtests/trunc-inputs: New file.
* benchtests/truncf-inputs: Likewise.
As per the section "3.1.4.2 Alignment Interrupts" of the "POWER8 Processor
User's Manual for the Single-Chip Module", alignment interrupt is reported
for misaligned stores in Caching-inhibited storage. As memset is used in
some drivers for DMA (like xorg), this patch avoids misaligned stores for
sizes less than 8 in memset.
Joseph Myers [Mon, 18 Sep 2017 23:26:35 +0000 (23:26 +0000)]
Fix powerpc64le problem from last ldbl-opt patch.
This patch fixes a problem on powerpc64le that I missed in initial
testing of my last patch to ldbl-opt. In the specific case of
powerpc64le, the weak aliases for exp10l and remainderl do not get
defined in the generic wrappers because of how those wrappers
undefine and redefine weak_alias. This patch restores those aliases
in the ldbl-opt code.
Tested (compilation only) for powerpc64le with build-many-glibcs.py.
Joseph Myers [Mon, 18 Sep 2017 17:51:33 +0000 (17:51 +0000)]
Use libm_alias_ldouble in math/.
This patch converts libm function implementations in math/ from using
weak_alias to using libm_alias_ldouble to define public function
names, in cases where it would be appropriate to define _Float128 /
_Float64x aliases for those functions as well (in cases where either
or both of those types exist and have the same ABI as long double).
This eliminates many ldbl-opt wrappers round these function
implementations.
Tested for x86_64, and with build-many-glibcs.py. All installed
stripped shared libraries are unchanged except for libm.so on
powerpc64le. As noted for a previous patch, powerpc64le's use of
ldbl-opt means various long double functions get defined using
long_double_symbol which gives them an explicit symbol version in the
object files, and this patch results in some such functions using
weak_alias instead (because powerpc64le never had a previous version
of these functions for long double = double); both produce a valid
libm.so with the same public symbols at the same versions, but macros
expanding to call weak_alias is cleaner in this case.
* math/s_fmal.c: Include <libm-alias-ldouble.h>.
(fmal): Define using libm_alias_ldouble.
* math/w_acoshl_compat.c: Include <libm-alias-ldouble.h>.
(acoshl): Define using libm_alias_ldouble.
* math/w_acosl_compat.c: Include <libm-alias-ldouble.h>.
(acosl): Define using libm_alias_ldouble.
* math/w_asinl_compat.c: Include <libm-alias-ldouble.h>.
(asinl): Define using libm_alias_ldouble.
* math/w_atan2l_compat.c: Include <libm-alias-ldouble.h>.
(atan2l): Define using libm_alias_ldouble.
* math/w_atanhl_compat.c: Include <libm-alias-ldouble.h>.
(atanhl): Define using libm_alias_ldouble.
* math/w_coshl_compat.c: Include <libm-alias-ldouble.h>.
(coshl): Define using libm_alias_ldouble.
* math/w_exp10l_compat.c: Include <libm-alias-ldouble.h>.
(exp10l): Define using libm_alias_ldouble.
* math/w_exp2l_compat.c: Include <libm-alias-ldouble.h>.
(exp2l): Define using libm_alias_ldouble.
* math/w_expl_compat.c: Include <libm-alias-ldouble.h>.
(expl): Define using libm_alias_ldouble.
* math/w_fmodl_compat.c: Include <libm-alias-ldouble.h>.
(fmodl): Define using libm_alias_ldouble.
* math/w_hypotl_compat.c: Include <libm-alias-ldouble.h>.
(hypotl): Define using libm_alias_ldouble.
* math/w_j0l_compat.c: Include <libm-alias-ldouble.h>.
(j0l): Define using libm_alias_ldouble.
(y0l): Likewise.
* math/w_j1l_compat.c: Include <libm-alias-ldouble.h>.
(j1l): Define using libm_alias_ldouble.
(y1l): Likewise.
* math/w_jnl_compat.c: Include <libm-alias-ldouble.h>.
(jnl): Define using libm_alias_ldouble.
(ynl): Likewise.
* math/w_log10l_compat.c: Include <libm-alias-ldouble.h>.
(log10l): Define using libm_alias_ldouble.
* math/w_log2l_compat.c: Include <libm-alias-ldouble.h>.
(log2l): Define using libm_alias_ldouble.
* math/w_logl_compat.c: Include <libm-alias-ldouble.h>.
(logl): Define using libm_alias_ldouble.
* math/w_powl_compat.c: Include <libm-alias-ldouble.h>.
(powl): Define using libm_alias_ldouble.
* math/w_remainderl_compat.c: Include <libm-alias-ldouble.h>.
(remainderl): Define using libm_alias_ldouble.
* math/w_sinhl_compat.c: Include <libm-alias-ldouble.h>.
(sinhl): Define using libm_alias_ldouble.
* math/w_sqrtl_compat.c: Include <libm-alias-ldouble.h>.
(sqrtl): Define using libm_alias_ldouble.
* math/w_tgammal_compat.c: Include <libm-alias-ldouble.h>.
(tgammal): Define using libm_alias_ldouble.
* sysdeps/ieee754/ldbl-opt/w_exp10l_compat.c [LIBM_SVID_COMPAT]
(exp10l): Do not use long_double_symbol here.
* sysdeps/ieee754/ldbl-opt/w_remainderl_compat.c
[LIBM_SVID_COMPAT] (remainderl): Likewise.
* sysdeps/ieee754/ldbl-opt/s_fmal.c: Remove.
* sysdeps/ieee754/ldbl-opt/w_acoshl_compat.c: Likewise.
* sysdeps/ieee754/ldbl-opt/w_acosl_compat.c: Likewise.
* sysdeps/ieee754/ldbl-opt/w_asinl_compat.c: Likewise.
* sysdeps/ieee754/ldbl-opt/w_atan2l_compat.c: Likewise.
* sysdeps/ieee754/ldbl-opt/w_atanhl_compat.c: Likewise.
* sysdeps/ieee754/ldbl-opt/w_coshl_compat.c: Likewise.
* sysdeps/ieee754/ldbl-opt/w_expl_compat.c: Likewise.
* sysdeps/ieee754/ldbl-opt/w_fmodl_compat.c: Likewise.
* sysdeps/ieee754/ldbl-opt/w_hypotl_compat.c: Likewise.
* sysdeps/ieee754/ldbl-opt/w_j0l_compat.c: Likewise.
* sysdeps/ieee754/ldbl-opt/w_j1l_compat.c: Likewise.
* sysdeps/ieee754/ldbl-opt/w_jnl_compat.c: Likewise.
* sysdeps/ieee754/ldbl-opt/w_log10l_compat.c: Likewise.
* sysdeps/ieee754/ldbl-opt/w_log2l_compat.c: Likewise.
* sysdeps/ieee754/ldbl-opt/w_logl_compat.c: Likewise.
* sysdeps/ieee754/ldbl-opt/w_powl_compat.c: Likewise.
* sysdeps/ieee754/ldbl-opt/w_sinhl_compat.c: Likewise.
* sysdeps/ieee754/ldbl-opt/w_sqrtl_compat.c: Likewise.
* sysdeps/ieee754/ldbl-opt/w_tgammal_compat.c: Likewise.
Wang Boshi [Fri, 15 Sep 2017 16:53:23 +0000 (17:53 +0100)]
AArch64: use movz/movk instead of literal pools in start.S
eXecute-Only Memory (XOM) is a protection mechanism against some ROP
attacks. XOM sets the code as executable and unreadable, so the access
to any data, like literal pools, in the code section causes the fault
with XOM. The compiler can disable literal pools for C source files,
but not for assembly files, so I use movz/movk instead of literal pools
in start.S for XOM.
I add MOVL macro with movz/movk instructions like movl pseudo-instruction
in armasm, and use the macro instead of literal pools.
* sysdeps/aarch64/start.S: Use MOVL instead of literal pools.
* sysdeps/aarch64/sysdep.h (MOVL): Add MOVL macro.
benchtests: New -g option to generate graphs in compare_strings.py
The compare_strings.py option unconditionally generates a graph PNG
image of the input data, which can be unnecessary and slow. Put this
behind an optional flag -g.
* benchtests/scripts/compare_strings.py: New option -g.
(draw_graph): Print a message that a graph is being generated.
(process_results): Generate graph only if -g is passed.
(main): Process option -g.
benchtests: Make compare_strings.py output a bit prettier
Make the column widths for the outputs fixed so that they look a
little less messy. They will still look bad with lots of IFUNCs (like
on x86) but it's still a step forward.
* benchtests/scripts/compare_strings.py (process_results):
Better spacing for output.
Make the script more usable by adding proper command line options
along with a way to query the options. The script is capable of doing
a bunch of things right now like choosing a base for comparison,
choosing to generate graphs, etc. and they should be accessible via
command line switches.
* benchtests/scripts/compare_strings.py: Use argparse.
* benchtests/README: Document existence of compare_strings.py.
Joseph Myers [Fri, 15 Sep 2017 23:10:02 +0000 (23:10 +0000)]
Use libm_alias_double in math/.
This patch converts libm function implementations in math/ from using
weak_alias to using libm_alias_double to define public function names,
in cases where it would be appropriate to define _Float64 / _Float32x
aliases for those functions as well. This eliminates many
NO_LONG_DOUBLE conditionals and ldbl-opt wrappers round these function
implementations.
Tested for x86_64. Also tested with build-many-glibcs.py. Binary
differences seen are that the different order in which remainder and
drem symbols get defined as a result of this patch (the same source
file defines the same aliases, but in a different order of definition)
changes the order of symbols in the final libm.so when long double =
double, and for ldbl-opt configurations, the compat symbols for Bessel
functions were previously defined by e.g. "compat_symbol (libm, j0,
j0l, GLIBC_2_0)", which declares j0l as a compat symbol based on j0
and so makes j0l weak because j0 is weak, and are now defined
(indirectly via the relevant macros) based on e.g. __j0, so are no
longer weak because __j0 isn't weak.
* math/s_fma.c: Include <libm-alias-double.h>.
(fma): Define using libm_alias_double.
* math/s_nextafter.c: Include <libm-alias-double.h>.
(nextafter): Define using libm_alias_double.
* math/w_acos_compat.c: Include <libm-alias-double.h>.
(acos): Define using libm_alias_double.
* math/w_acosh_compat.c: Include <libm-alias-double.h>.
(aocsh): Define using libm_alias_double.
* math/w_asin_compat.c: Include <libm-alias-double.h>.
(asin): Define using libm_alias_double.
* math/w_atan2_compat.c: Include <libm-alias-double.h>.
(atan2): Define using libm_alias_double.
* math/w_atanh_compat.c: Include <libm-alias-double.h>.
(atanh): Define using libm_alias_double.
* math/w_cosh_compat.c: Include <libm-alias-double.h>.
(cosh): Define using libm_alias_double.
* math/w_exp10_compat.c: Include <libm-alias-double.h>.
(exp10): Define using libm_alias_double.
* math/w_exp2_compat.c: Include <libm-alias-double.h>.
(exp2): Define using libm_alias_double.
* math/w_exp_compat.c: Include <libm-alias-double.h>.
(exp): Define using libm_alias_double.
* math/w_fmod_compat.c: Include <libm-alias-double.h>.
(fmod): Define using libm_alias_double.
* math/w_hypot_compat.c: Include <libm-alias-double.h>.
(hypot): Define using libm_alias_double.
* math/w_j0_compat.c: Include <libm-alias-double.h>.
(j0): Define using libm_alias_double.
(y0): Likewise.
* math/w_j1_compat.c: Include <libm-alias-double.h>.
(j1): Define using libm_alias_double.
(y1): Likewise.
* math/w_jn_compat.c: Include <libm-alias-double.h>.
(jn): Define using libm_alias_double.
(yn): Likewise.
* math/w_log10_compat.c: Include <libm-alias-double.h>.
(log10): Define using libm_alias_double.
* math/w_log2_compat.c: Include <libm-alias-double.h>.
(log2): Define using libm_alias_double.
* math/w_log_compat.c: Include <libm-alias-double.h>.
(log): Define using libm_alias_double.
* math/w_pow_compat.c: Include <libm-alias-double.h>.
(pow): Define using libm_alias_double.
* math/w_remainder_compat.c: Include <libm-alias-double.h>.
(remainder): Define using libm_alias_double.
* math/w_sinh_compat.c: Include <libm-alias-double.h>.
(sinh): Define using libm_alias_double.
* math/w_sqrt_compat.c: Include <libm-alias-double.h>.
(sqrt): Define using libm_alias_double.
* math/w_tgamma_compat.c: Include <libm-alias-double.h>.
(tgamma): Define using libm_alias_double.
* sysdeps/ieee754/ldbl-opt/s_nextafter.c [LONG_DOUBLE_COMPAT(libm,
GLIBC_2_0)] (nextafterl): Do not define compat symbol here.
* sysdeps/ieee754/ldbl-opt/w_exp10_compat.c
[LONG_DOUBLE_COMPAT(libm, GLIBC_2_1)] (exp10l): Likewise.
* sysdeps/ieee754/ldbl-opt/w_remainder_compat.c
[LONG_DOUBLE_COMPAT(libm, GLIBC_2_0)] (remainderl): Likewise.
* sysdeps/ieee754/ldbl-opt/w_acos_compat.c: Remove.
* sysdeps/ieee754/ldbl-opt/w_acosh_compat.c: Likewise.
* sysdeps/ieee754/ldbl-opt/w_asin_compat.c: Likewise.
* sysdeps/ieee754/ldbl-opt/w_atan2_compat.c: Likewise.
* sysdeps/ieee754/ldbl-opt/w_atanh_compat.c: Likewise.
* sysdeps/ieee754/ldbl-opt/w_cosh_compat.c: Likewise.
* sysdeps/ieee754/ldbl-opt/w_exp_compat.c: Likewise.
* sysdeps/ieee754/ldbl-opt/w_fmod_compat.c: Likewise.
* sysdeps/ieee754/ldbl-opt/w_hypot_compat.c: Likewise.
* sysdeps/ieee754/ldbl-opt/w_j0_compat.c: Likewise.
* sysdeps/ieee754/ldbl-opt/w_j1_compat.c: Likewise.
* sysdeps/ieee754/ldbl-opt/w_jn_compat.c: Likewise.
* sysdeps/ieee754/ldbl-opt/w_log10_compat.c: Likewise.
* sysdeps/ieee754/ldbl-opt/w_log2_compat.c: Likewise.
* sysdeps/ieee754/ldbl-opt/w_log_compat.c: Likewise.
* sysdeps/ieee754/ldbl-opt/w_pow_compat.c: Likewise.
* sysdeps/ieee754/ldbl-opt/w_sinh_compat.c: Likewise.
* sysdeps/ieee754/ldbl-opt/w_sqrt_compat.c: Likewise.
* sysdeps/ieee754/ldbl-opt/w_tgamma_compat.c: Likewise.
Joseph Myers [Fri, 15 Sep 2017 20:05:09 +0000 (20:05 +0000)]
Remove ENOSYS math/ function stubs.
The math/ directory has various stub implementations of functions
(generally long double functions, plus sqrtf128) that set errno to
ENOSYS and print a message.
These are a relic of when the long double libm implementation in glibc
was originally incomplete. They are only present for a limited set of
libm functions, not all those that would need implementing for any
hypothetical new long double format, and only for long double (plus
sqrtf128), not for other floating-point types. I removed the
corresponding ENOSYS tests in the testsuite in
<https://sourceware.org/ml/libc-alpha/2013-05/msg00599.html>.
I think these stubs are an inherently different case from stubs for
functionality needing OS-specific implementations, since
floating-point formats are generally standardized between modern
processors (a new long double format does not seem likely - I'd expect
new ports all to have long double = binary128 or long double =
binary64 - though _Float16 functions *f16 are entirely possible in
future) and libm functions are generally only expected to fail for
domain / range / pole errors, not for other unknown system-specific
conditions the way the functions needing OS-specific implementations
may fail.
This patch removes these stub implementations. The effect (which, as
noted above, was actually mostly already the case, given the
incompleteness of the set of stubs, and was certainly the case if you
wanted your libm port to *work* rather than just build) is that any
new long double format needs to provide a full set of functions for
the build to succeed; any port using ldbl-128 needs to provide sqrtl
(possibly using the soft-fp version from another port); any port using
ldbl-96 needs to provide all those functions that all of
x86_64/i386/ia64/m68k provide rather than getting from the ldbl-96
directory; any port with _Float128 as a type with a distinct format
needs to provide sqrtf128.
Tested for x86_64. Also tested with build-many-glibcs.py that
installed stripped shared libraries are unchanged by the patch.
[BZ #21745] powerpc: build some IFUNC math functions for libc and libm
Some math functions have to be distributed in libc because they're
required by printf.
libc and libm require their own builds of these functions, e.g. libc
functions have to call __stack_chk_fail_local in order to bypass the
PLT, while libm functions have to call __stack_chk_fail.
While math/Makefile treat the generic cases, i.e. s_isinff, the
multiarch Makefile has to treat its own files, i.e. s_isinff-ppc64.
[BZ #21745]
* sysdeps/powerpc/powerpc64/fpu/multiarch/Makefile:
[$(subdir) = math] (sysdep_calls): New variable. Has the
previous contents of sysdep_routines, but re-sorted..
[$(subdir) = math] (sysdep_routines): Re-use the contents from
sysdep_calls.
[$(subdir) = math] (libm-sysdep_routines): Remove the functions
defined in sysdep_calls and replace by the respective m_* names.
* sysdeps/powerpc/powerpc64/fpu/multiarch/s_isnan-ppc64.S:
(compat_symbol): Undefine to avoid duplicated compat symbols in
libc.
Joseph Myers [Fri, 15 Sep 2017 17:12:02 +0000 (17:12 +0000)]
Use libm_alias_float in math/.
This patch converts libm function implementations in math/ from using
weak_alias to using libm_alias_float to define public function names,
in cases where it would be appropriate to define _Float32 aliases for
those functions as well. expf and exp2f are omitted from this patch,
given the in-progress patches that would change their symbol
versioning arrangements (at a later stage it will be necessary to add
macros that can be used for functions with such symbol versioning
arrangements - which will apply to lgammaf as well - but for the
initial patches in this area I'm just dealing with easy cases, and any
symbol versioning changes to these functions while the work is in
progress can effectively just undo the libm_alias_* changes as regards
those functions).
Tested for x86_64. Also tested with build-many-glibcs.py that
installed stripped shared libraries are unchanged by the patch.
* math/s_fmaf.c: Include <libm-alias-float.h>.
(fmaf): Define using libm_alias_float.
* math/w_acosf_compat.c: Include <libm-alias-float.h>.
(acosf): Define using libm_alias_float.
* math/w_acoshf_compat.c: Include <libm-alias-float.h>.
(acoshf): Define using libm_alias_float.
* math/w_asinf_compat.c: Include <libm-alias-float.h>.
(asinf): Define using libm_alias_float.
* math/w_atan2f_compat.c: Include <libm-alias-float.h>.
(atan2f): Define using libm_alias_float.
* math/w_atanhf_compat.c: Include <libm-alias-float.h>.
(atanhf): Define using libm_alias_float.
* math/w_coshf_compat.c: Include <libm-alias-float.h>.
(coshf): Define using libm_alias_float.
* math/w_exp10f_compat.c: Include <libm-alias-float.h>.
(exp10f): Define using libm_alias_float.
* math/w_fmodf_compat.c: Include <libm-alias-float.h>.
(fmodf): Define using libm_alias_float.
* math/w_hypotf_compat.c: Include <libm-alias-float.h>.
(hypotf): Define using libm_alias_float.
* math/w_j0f_compat.c: Include <libm-alias-float.h>.
(j0f): Define using libm_alias_float.
(y0f): Likewise.
* math/w_j1f_compat.c: Include <libm-alias-float.h>.
(j1f): Define using libm_alias_float.
(y1f): Likewise.
* math/w_jnf_compat.c: Include <libm-alias-float.h>.
(jnf): Define using libm_alias_float.
(ynf): Likewise.
* math/w_log10f_compat.c: Include <libm-alias-float.h>.
(log10f): Define using libm_alias_float.
* math/w_log2f_compat.c: Include <libm-alias-float.h>.
(log2f): Define using libm_alias_float.
* math/w_logf_compat.c: Include <libm-alias-float.h>.
(logf): Define using libm_alias_float.
* math/w_powf_compat.c: Include <libm-alias-float.h>.
(powf): Define using libm_alias_float.
* math/w_remainderf_compat.c: Include <libm-alias-float.h>.
(remainderf): Define using libm_alias_float.
* math/w_sinhf_compat.c: Include <libm-alias-float.h>.
(sinhf): Define using libm_alias_float.
* math/w_sqrtf_compat.c: Include <libm-alias-float.h>.
(sqrtf): Define using libm_alias_float.
* math/w_tgammaf_compat.c: Include <libm-alias-float.h>.
(tgammaf): Define using libm_alias_float.
Joseph Myers [Thu, 14 Sep 2017 22:28:53 +0000 (22:28 +0000)]
Make more libm functions into weak aliases.
Many libm functions define the function as __<func> and then define
<func> as a weak alias. This is not at all limited to cases where
there is an internal call that has namespace reasons to need to call
__<func> instead of <func>.
The common macros for creating libm function aliases work on the basis
of public function names all being aliases; that is, they define
aliases for functions using the above pattern. Thus, where a function
just defines the public name <func> directly, changing that to be a
weak alias enables a subsequent conversion to the common macros to
retain the exact existing symbols (and so be testable by comparison of
stripped binaries).
This patch converts many existing functions to use the weak alias
pattern, as preparation for subsequent conversions to common macros.
I do expect that _FloatN/_FloatNx function aliases will end up needing
new variants of the common macros that do *not* create the original
float / double / long double name of a function - for cases where that
name is created specially to give it a particular symbol version, for
example - but for functions that can use the most common macros to
create all the public names as aliases, it makes sense for them to do
so.
Regarding the Bessel function wrappers in this patch: only float and
double wrappers are changed because the long double wrappers already
used the weak alias pattern.
Tested for x86_64, and with build-many-glibcs.py.
* include/math.h (roundeven): Change hidden_proto call to
__roundeven.
* math/w_j0_compat.c (j0): Rename to __j0 and define as weak
alias.
[NO_LONG_DOUBLE] (__j0l): New strong alias.
(y0): Rename to __y0 and define as weak alias.
[NO_LONG_DOUBLE] (__y0l): New strong alias.
* math/w_j0f_compat.c (j0f): Rename to __j0f and define as weak
alias.
(y0f): Rename to __y0f and define as weak alias.
* math/w_j1_compat.c (j1): Rename to __j1 and define as weak
alias.
[NO_LONG_DOUBLE] (__j1l): New strong alias.
(y1): Rename to __y1 and define as weak alias.
[NO_LONG_DOUBLE] (__y1l): New strong alias.
* math/w_j1f_compat.c (j1f): Rename to __j1f and define as weak
alias.
(y1f): Rename to __y1f and define as weak alias.
* math/w_jn_compat.c (jn): Rename to __jn and define as weak
alias.
[NO_LONG_DOUBLE] (__jnl): New strong alias.
(yn): Rename to __yn and define as weak alias.
[NO_LONG_DOUBLE] (__ynl): New strong alias.
* math/w_jnf_compat.c (jnf): Rename to __jnf and define as weak
alias.
(ynf): Rename to __ynf and define as weak alias.
* sysdeps/ieee754/dbl-64/s_fromfp.c (FUNC): Define to __fromfp.
(fromfp): Define as weak alias.
[NO_LONG_DOUBLE] (__fromfpl): New strong alias.
* sysdeps/ieee754/dbl-64/s_fromfpx.c (FUNC): Define to __fromfpx.
(fromfpx): Define as weak alias.
[NO_LONG_DOUBLE] (__fromfpxl): New strong alias.
* sysdeps/ieee754/dbl-64/s_getpayload.c (getpayload): Rename to
__getpayload and define as weak alias.
[NO_LONG_DOUBLE] (__getpayloadl): New strong alias.
* sysdeps/ieee754/dbl-64/s_roundeven.c (roundeven): Rename to
__roundeven and define as weak alias.
[NO_LONG_DOUBLE] (__roundevenl): New strong alias.
* sysdeps/ieee754/dbl-64/s_setpayload.c (FUNC): Define to
__setpayload.
(setpayload): Define as weak alias.
[NO_LONG_DOUBLE] (__setpayloadl): New strong alias.
* sysdeps/ieee754/dbl-64/s_setpayloadsig.c (FUNC): Define to
__setpayloadsig.
(setpayloadsig): Define as weak alias.
[NO_LONG_DOUBLE] (__setpayloadsigl): New strong alias.
* sysdeps/ieee754/dbl-64/s_totalorder.c (totalorder): Rename to
__totalorder and define as weak alias.
[NO_LONG_DOUBLE] (__totalorderl): New strong alias.
* sysdeps/ieee754/dbl-64/s_totalordermag.c (totalordermag): Rename
to __totalordermag and define as weak alias.
[NO_LONG_DOUBLE] (__totalordermagl): New strong alias.
* sysdeps/ieee754/dbl-64/s_ufromfp.c (FUNC): Define to __ufromfp.
(ufromfp): Define as weak alias.
[NO_LONG_DOUBLE] (__ufromfpl): New strong alias.
* sysdeps/ieee754/dbl-64/s_ufromfpx.c (FUNC): Define to
__ufromfpx.
(ufromfpx): Define as weak alias.
[NO_LONG_DOUBLE] (__ufromfpxl): New strong alias.
* sysdeps/ieee754/dbl-64/wordsize-64/s_getpayload.c (getpayload):
Rename to __getpayload and define as weak alias.
[NO_LONG_DOUBLE] (__getpayloadl): New strong alias.
* sysdeps/ieee754/dbl-64/wordsize-64/s_roundeven.c (roundeven):
Rename to __roundeven and define as weak alias.
[NO_LONG_DOUBLE] (__roundevenl): New strong alias.
* sysdeps/ieee754/dbl-64/wordsize-64/s_totalorder.c (totalorder):
Rename to __totalorder and define as weak alias.
[NO_LONG_DOUBLE] (__totalorderl): New strong alias.
* sysdeps/ieee754/dbl-64/wordsize-64/s_totalordermag.c
(totalordermag): Rename to __totalordermag and define as weak
alias.
[NO_LONG_DOUBLE] (__totalordermagl): New strong alias.
* sysdeps/ieee754/float128/float128_private.h (__getpayloadl): New
macro.
(__roundevenl): Likewise.
(__totalorderl): Likewise.
(__totalordermagl): Likewise
* sysdeps/ieee754/float128/s_fromfpf128.c (FUNC): Define to
__fromfpf128.
(fromfpf128): Define as weak alias.
* sysdeps/ieee754/float128/s_fromfpxf128.c (FUNC): Define to
__fromfpxf128.
(fromfpxf128): Define as weak alias.
* sysdeps/ieee754/float128/s_setpayloadf128.c (FUNC): Define to
__setpayloadf128.
(setpayloadf128): Define as weak alias.
* sysdeps/ieee754/float128/s_setpayloadsigf128.c (FUNC): Define to
__setpayloadsigf128.
(setpayloadsigf128): Define as weak alias.
* sysdeps/ieee754/float128/s_ufromfpf128.c (FUNC): Define to
__ufromfpf128.
(ufromfpf128): Define as weak alias.
* sysdeps/ieee754/float128/s_ufromfpxf128.c (FUNC): Define to
__ufromfpxf128.
(ufromfpxf128): Define as weak alias.
* sysdeps/ieee754/flt-32/s_fromfpf.c (FUNC): Define to __fromfpf.
(fromfpf): Define as weak alias.
* sysdeps/ieee754/flt-32/s_fromfpxf.c (FUNC): Define to
__fromfpxf.
(fromfpxf): Define as weak alias.
* sysdeps/ieee754/flt-32/s_getpayloadf.c (getpayloadf): Rename to
__getpayloadf and define as weak alias.
* sysdeps/ieee754/flt-32/s_roundevenf.c (roundevenf): Rename to
__roundevenf and define as weak alias.
* sysdeps/ieee754/flt-32/s_setpayloadf.c (FUNC): Define to
__setpayloadf.
(setpayloadf): Define as weak alias.
* sysdeps/ieee754/flt-32/s_setpayloadsigf.c (FUNC): Define to
__setpayloadsigf.
(setpayloadsigf): Define as weak alias.
* sysdeps/ieee754/flt-32/s_totalorderf.c (totalorderf): Rename to
__totalorderf and define as weak alias.
* sysdeps/ieee754/flt-32/s_totalordermagf.c (totalordermagf):
Rename to __totalordermagf and define as weak alias.
* sysdeps/ieee754/flt-32/s_ufromfpf.c (FUNC): Define to
__ufromfpf.
(ufromfpf): Define as weak alias.
* sysdeps/ieee754/flt-32/s_ufromfpxf.c (FUNC): Define to
__ufromfpxf.
(ufromfpxf): Define as weak alias.
* sysdeps/ieee754/ldbl-128/s_fromfpl.c (FUNC): Define to
__fromfpl.
(fromfpl): Define as weak alias.
* sysdeps/ieee754/ldbl-128/s_fromfpxl.c (FUNC): Define to
__fromfpxl.
(fromfpxl): Define as weak alias.
* sysdeps/ieee754/ldbl-128/s_getpayloadl.c (getpayloadl): Rename
to __getpayloadl and define as weak alias.
* sysdeps/ieee754/ldbl-128/s_roundevenl.c (roundevenl): Rename to
__roundevenl and define as weak alias.
* sysdeps/ieee754/ldbl-128/s_setpayloadl.c (FUNC): Define to
__setpayloadl.
(setpayloadl): Define as weak alias.
* sysdeps/ieee754/ldbl-128/s_setpayloadsigl.c (FUNC): Define to
__setpayloadsigl.
(setpayloadsigl): Define as weak alias.
* sysdeps/ieee754/ldbl-128/s_totalorderl.c (totalorderl): Rename
to __totalorderl and define as weak alias.
* sysdeps/ieee754/ldbl-128/s_totalordermagl.c (totalordermagl):
Rename to __totalordermagl and define as weak alias.
* sysdeps/ieee754/ldbl-128/s_ufromfpl.c (FUNC): Define to
__ufromfpl.
(ufromfpl): Define as weak alias.
* sysdeps/ieee754/ldbl-128/s_ufromfpxl.c (FUNC): Define to
__ufromfpxl.
(ufromfpxl): Define as weak alias.
* sysdeps/ieee754/ldbl-128ibm/s_fromfpl.c (FUNC): Define to
__fromfpl.
(fromfpl): Define as weak alias.
* sysdeps/ieee754/ldbl-128ibm/s_fromfpxl.c (FUNC): Define to
__fromfpxl.
(fromfpxl): Define as weak alias.
* sysdeps/ieee754/ldbl-128ibm/s_getpayloadl.c (getpayloadl):
Rename to __getpayloadl and define as weak alias.
* sysdeps/ieee754/ldbl-128ibm/s_roundevenl.c (roundevenl): Rename
to __roundevenl and define as weak alias. Call __roundeven
instead of roundeven.
* sysdeps/ieee754/ldbl-128ibm/s_setpayloadl.c (FUNC): Define to
__setpayloadl.
(setpayloadl): Define as weak alias.
* sysdeps/ieee754/ldbl-128ibm/s_setpayloadsigl.c (FUNC): Define to
__setpayloadsigl.
(setpayloadsigl): Define as weak alias.
* sysdeps/ieee754/ldbl-128ibm/s_totalorderl.c (totalorderl):
Rename to __totalorderl and define as weak alias.
* sysdeps/ieee754/ldbl-128ibm/s_totalordermagl.c (totalordermagl):
Rename to __totalordermagl and define as weak alias.
* sysdeps/ieee754/ldbl-128ibm/s_ufromfpl.c (FUNC): Define to
__ufromfpl.
(ufromfpl): Define as weak alias.
* sysdeps/ieee754/ldbl-128ibm/s_ufromfpxl.c (FUNC): Define to
__ufromfpxl.
(ufromfpxl): Define as weak alias.
* sysdeps/ieee754/ldbl-96/s_fromfpl.c (FUNC): Define to
__fromfpl.
(fromfpl): Define as weak alias.
* sysdeps/ieee754/ldbl-96/s_fromfpxl.c (FUNC): Define to
__fromfpxl.
(fromfpxl): Define as weak alias.
* sysdeps/ieee754/ldbl-96/s_getpayloadl.c (getpayloadl): Rename to
__getpayloadl and define as weak alias.
* sysdeps/ieee754/ldbl-96/s_roundevenl.c (roundevenl): Rename to
__roundevenl and define as weak alias.
* sysdeps/ieee754/ldbl-96/s_setpayloadl.c (FUNC): Define to
__setpayloadl.
(setpayloadl): Define as weak alias.
* sysdeps/ieee754/ldbl-96/s_setpayloadsigl.c (FUNC): Define to
__setpayloadsigl.
(setpayloadsigl): Define as weak alias.
* sysdeps/ieee754/ldbl-96/s_totalorderl.c (totalorderl): Rename to
__totalorderl and define as weak alias.
* sysdeps/ieee754/ldbl-96/s_totalordermagl.c (totalordermagl):
Rename to __totalordermagl and define as weak alias.
* sysdeps/ieee754/ldbl-96/s_ufromfpl.c (FUNC): Define to
__ufromfpl.
(ufromfpl): Define as weak alias.
* sysdeps/ieee754/ldbl-96/s_ufromfpxl.c (FUNC): Define to
__ufromfpxl.
(ufromfpxl): Define as weak alias.
Keeping the same buffers along with copying the same size of data into
the same location means that the first routine is typically the
slowest since it has to bear the cost of fetching data into to cache.
Reallocating buffers stabilizes numbers by a bit.
* benchtests/bench-string.h (realloc_bufs): New function.
(test_init): Call it.
* benchtests/bench-memset-large.c (do_test): Likewise.
* benchtests/bench-memset.c (do_test): Likewise.
Make the memset benchmarks (bench-memset and bench-memset-large) print
their output in JSON so that they can be evaluated using the
compare_strings.py script.
* benchtests/bench-memset-large.c: Print output in JSON
format.
* benchtests/bench-memset.c: Likewise.
Mike FABIAN [Thu, 7 Sep 2017 13:28:28 +0000 (15:28 +0200)]
Add new codepage charmaps/IBM858 [BZ #21084]
This code page is identical to code page 850 except that X'D5'
has been changed from LI61 (dotless i) to SC20 (euro symbol).
The code points from /x01 to /x1f in the /localedata/charmaps/IBM858
file have the same mapping as those in localedata/charmaps/ANSI_X3.4-1968.
That means they disagree with with
in that range.
For example, localedata/charmaps/IBM858 and localedata/charmaps/ANSI_X3.4-1968 have:
“<U0001> /x01 START OF HEADING (SOH)”
whereas CP00858.txt has:
“01 SS000000 Smiling Face”
That means that CP00858.txt is not really ASCII-compatible and to make
it ASCII-compatible we deviate fro CP00858.txt in the code points from /x01
to /x1f.
[BZ #21084]
* benchtests/strcoll-inputs/filelist#en_US.UTF-8: Add IBM858 and ibm858.c.
* iconvdata/Makefile: Add IBM858.
* iconvdata/gconv-modules: Add IBM858.
* iconvdata/ibm858.c: New file.
* iconvdata/tst-tables.sh: Add IBM858
* localedata/charmaps/IBM858: New file.
Joseph Myers [Thu, 14 Sep 2017 01:11:46 +0000 (01:11 +0000)]
Define and use libm_alias_float128.
Continuing the process of setting up common macros for libm function
aliases, with a view to using them to define _FloatN / _FloatNx
aliases in future, this patch adds a libm_alias_float128 macro and
uses it in the type-generic templates. (_Float128 functions will end
up with _Float64x aliases on powerpc64le, but not on x86_64/x86/ia64
because _Float64x has long double format there, and the macro will
provide a single place for the conditionals for that choice, as well
as for once ldbl-128 functions always build *f128 and need
conditionals for whether to have *l aliases.)
Tested for x86_64. Also tested with build-many-glibcs.py that
installed stripped shared libraries are unchanged by the patch.
* sysdeps/generic/libm-alias-float128.h: New file.
* sysdeps/generic/math-type-macros-float128.h: Include
<libm-alias-float128.h>.
[!declare_mgen_alias] (declare_mgen_alias): Define macro.
Joseph Myers [Wed, 13 Sep 2017 22:17:23 +0000 (22:17 +0000)]
Define and use libm_alias_ldouble.
Continuing the process of setting up common macros for libm function
aliases, with a view to using them to define _FloatN / _FloatNx
aliases in future, this patch adds a libm_alias_ldouble macro and uses
it in the type-generic templates.
Since math-type-macros-ldouble.h already did the appropriate thing for
each symbol (weak_alias or long_double_symbol), this is just a
straightforward rearrangement of code, to make the required logic
available in a place that can also be used outside of the type-generic
templates in future (in particular, to eliminate various wrappers for
functions in ldbl-opt and ldbl-64-128).
Tested for x86_64. Also tested with build-many-glibcs.py that
installed stripped shared libraries are unchanged by the patch.
* sysdeps/generic/libm-alias-ldouble.h: New file.
* sysdeps/ieee754/ldbl-opt/libm-alias-ldouble.h: Likewise.
* sysdeps/ieee754/ldbl-opt/math-type-macros-ldouble.h: Remove.
* sysdeps/generic/math-type-macros-ldouble.h: Include
<libm-alias-ldouble.h>.
[!declare_mgen_alias] (declare_mgen_alias): Define to use
libm_alias_ldouble.
Joseph Myers [Wed, 13 Sep 2017 15:47:26 +0000 (15:47 +0000)]
Clear up log1p, ldexp, scalbn, scalbln compat handling.
This patch cleans up how compat symbols / long double versioning are
handled for log1p, ldexp, scalbn and scalbln functions.
The general principle is to do as much as possible through the
type-generic templates. Previously, when errno-setting wrappers were
added the compat long double symbols were left pointing directly to
the underlying implementations; they are moved to point to the
errno-setting wrappers. For the functions also present in libc,
compat symbol handling for the libc copies needs to go in ldbl-opt
wrappers, but the type-generic templates can handle it for the libm
copies. There is no need for w_scalbln_template.c to disable the
creation of an unused internal alias (such code made sense in the
context of patches trying to avoid any changes to generated code for
ease of comparison, but can be removed in a change that specifically
does intend to change details of where symbols point).
Tested for x86_64, and with build-many-glibcs.py.
* math/w_scalbln_template.c (strong_alias): Do not undefine and
redefine.
* sysdeps/ieee754/ldbl-opt/s_ldexp.c (declare_mgen_alias): Remove
macro.
(ldexpl): Only define as compat symbol for libc, not libm.
(scalbnl): Define as compat symbol for libc here.
* sysdeps/ieee754/ldbl-opt/s_ldexpl.c (declare_mgen_alias): Only
define for [IS_IN (libc)].
(__ldexpl_2): Remove alias.
(ldexpl): Only define with long_double_symbol if [IS_IN (libc)].
(scalbnl): Likewise. Use __wrap_scalbnl not __ldexpl_2 as base
name in long_double_symbol call.
* sysdeps/ieee754/ldbl-opt/s_log1p.c: Remove file.
* sysdeps/ieee754/ldbl-opt/s_scalbln.c: Likewise.
* sysdeps/ieee754/ldbl-opt/s_scalbn.c: Likewise.
* sysdeps/ieee754/ldbl-opt/w_log1p.c: Likewise.
* sysdeps/ieee754/ldbl-opt/w_scalbln.c (declare_mgen_alias):
Remove macro.
[IS_IN (libc) && LONG_DOUBLE_COMPAT (libc, GLIBC_2_1)] (scalblnl):
Define as compat symbol.
Joseph Myers [Wed, 13 Sep 2017 01:13:30 +0000 (01:13 +0000)]
Define and use libm_alias_double.
Continuing the process of setting up common macros for libm function
aliases, with a view to using them to define _FloatN / _FloatNx
aliases in future, this patch adds a libm_alias_double macro and uses
it in the type-generic templates.
This macro handles defining aliases for double, and for long double in
the NO_LONG_DOUBLE case. It also handles defining compat symbols for
long double = double for architectures that changed their long double
format. By so doing, it eliminates the need for the
M_LIBM_NEED_COMPAT and declare_mgen_libm_compat macros; the single
declare_mgen_alias call in each template now suffices to define all
required compat symbols. When used for more double functions (not
based on type-generic templates), I expect it will eliminate the need
for most ldbl-opt wrappers for such functions.
A few special cases are needed. __clog10l is a public symbol (for
historical reasons) so needs to be given appropriate compat versions
for architectures that changed their long double format, but is not
defined as an alias using the normal macros since __clog10* are *not*
public symbols for _FloatN / _FloatNx types. For scalbn, scalbln and
log1p, the changes adding errno setting support for those functions
left compat symbols pointing directly to the non-errno-setting
implementations. There is no requirement for the compat symbols not
to set errno; that just made for the simplest patches at that time.
Now, with these common macros, it's natural to redirect the compat
symbols to the errno-setting wrappers, which I intend to do in a
separate patch.
Tested for x86_64, and with build-many-glibcs.py. For ldbl-opt
platforms the stripped libm.so binaries are changed (disassembly
unchanged) because the details of how the clog10l compat symbol is
created mean it ceases to be weak as it was before; for other
platforms, stripped libm.so binaries are unchanged.
Joseph Myers [Tue, 12 Sep 2017 20:00:00 +0000 (20:00 +0000)]
Remove declare_mgen_alias_2.
The libm template machinery includes a macro declare_mgen_alias_2, to
define two function aliases rather than one. This macro is only used
in one place, to define ldexp and scalbn, and only has one nondefault
definition, for double in the case where long double has the same
format. That definition is because declare_mgen_alias for double, in
that case, defines <internal-func>l as an alias of <internal-func>, so
cannot be called twice for aliases of the same function.
Now, I suspect the <internal-func>l aliases are generally not needed
(with maybe a few exceptions such as __clog10l, which is an exported
function). But even in the presence of them, there is no need for a
special declare_mgen_alias_2 macro for this case. This patch
eliminates the need for such a macro by defining __wrap_scalbn<suffix>
as an alias of __ldexp<suffix>, and then using that when defining the
scalbn public aliases. This is similar to how such internal aliases
are created for functions with multiple symbol versions, for example.
Tested for x86_64, and with build-many-glibcs.py. (There *are* some
cases where installed stripped shared libraries change - not in the
generated code but because such changes to static symbols on input to
ld, even nonexported symbols that don't affect the code or dynamic
symbols, can affect the particular representation in the output of
string tables, hash tables etc.)
H.J. Lu [Tue, 12 Sep 2017 14:46:11 +0000 (07:46 -0700)]
x86: Add MathVec_Prefer_No_AVX512 to cpu-features [BZ #21967]
AVX512 functions in mathvec are used on machines with AVX512. An AVX2
wrapper is also provided and it can be used when the AVX512 version
isn't profitable. MathVec_Prefer_No_AVX512 is addded to cpu-features.
If glibc.tune.hwcaps=MathVec_Prefer_No_AVX512 is set in GLIBC_TUNABLES
environment variable, the AVX2 wrapper will be used.
Tested on x86-64 machines with and without AVX512. Also verified
glibc.tune.hwcaps=MathVec_Prefer_No_AVX512 on AVX512 machine.
[BZ #21967]
* sysdeps/x86/cpu-features.h (bit_arch_MathVec_Prefer_No_AVX512):
New.
(index_arch_MathVec_Prefer_No_AVX512): Likewise.
* sysdeps/x86/cpu-tunables.c (TUNABLE_CALLBACK (set_hwcaps)):
Handle MathVec_Prefer_No_AVX512.
* sysdeps/x86_64/fpu/multiarch/ifunc-mathvec-avx512.h
(IFUNC_SELECTOR): Return AVX2 version if MathVec_Prefer_No_AVX512
is set.
Joseph Myers [Mon, 11 Sep 2017 23:21:25 +0000 (23:21 +0000)]
Define and use a libm_alias_float macro.
Fully supporting TS 18661-3 _FloatN / _FloatNx types in the cases
where they have the same format as other supported types (in line with
the principles described at
<https://sourceware.org/ml/libc-alpha/2017-01/msg00333.html>) means
adding a lot of function aliases to libm (and a few to libc). float
functions will have *f32 aliases, double functions will have *f32x and
*f64 aliases, long double functions may have *f64x, *f128 or both
aliases depending on the configuration, float128 functions have have
*f64x aliases depending on the configuration.
At present, most individual libm functions have their own weak_alias
calls to define the public names for those functions. For TS 18661-3
support, it is desirable that functions not all need to duplicate the
logic for which alias names to define.
Thus, common macros for defining the public aliases to a libm function
make sense. In the double and long double cases, such macros will
also help simplify existing code (with LONG_DOUBLE_COMPAT
etc. conditionals), by eliminating existing conditionals and ldbl-opt
/ ldbl-64-128 wrappers (using the generated ldbl-compat-choose.h to
allow a single macro definition to expand appropriately for each
symbol depending on LONG_DOUBLE_COMPAT for that symbol).
This patch starts the process of adding such macros with a
straightforward case: a libm_alias_float macro, initially only used in
the case of type-generic templates, to define aliases for float
functions (currently just the *f public names, in future also *f32).
Future patches are intended to add such macros for other types and to
extend the cases in which they are used, with a view to as many places
as possible using them before support for _FloatN / _FloatNx aliases
is enabled. (I think it's inevitable that some places doing
architecture-specific things with aliases and symbol versioning may
end up needing to replicate logic for the new aliases, but hopefully
the number of such places can be kept to a minimum.)
The libm_alias_float macro takes unsuffixed names for both the
internal and public function names. The need for unsuffixed public
names is obvious, since such macros will end up defining multiple
public names with different suffixes. Unsuffixed internal names are
because I expect the ldbl-128 functions to end up in a form that
always defines *f128 names and sometimes also defines *l names - with
the main internal names being e.g. __ieee754_<func>f128 (so many
macros in float128_private.h can go away). But __ieee754_<func>l
aliases will still be needed for e.g. use from math/ complex
functions, meaning the alias macro needs to see just __ieee754_<func>
as internal name so it can create an alias based on that name. Since
libm_alias_float128 will thus need the unsuffixed internal name, it
seems to make sense for all such macros to receive the unsuffixed
name.
Tested for x86_64. Also tested with build-many-glibcs.py that
installed stripped shared libraries are unchanged by the patch.
* sysdeps/generic/libm-alias-float.h: New file.
* sysdeps/generic/math-type-macros-float.h: Include
<libm-alias-float.h>.
[!declare_mgen_alias] (declare_mgen_alias): Define macro.
H.J. Lu [Mon, 11 Sep 2017 15:18:11 +0000 (08:18 -0700)]
x86: Add x86_64 to x86-64 HWCAP [BZ #22093]
Before glibc 2.26, ld.so set dl_platform to "x86_64" and searched the
"x86_64" subdirectory when loading a shared library. ld.so in glibc
2.26 was changed to set dl_platform to "haswell" or "xeon_phi", based
on supported ISAs. This led to shared library loading failure for
shared libraries placed under the "x86_64" subdirectory.
This patch adds "x86_64" to x86-64 dl_hwcap so that ld.so will always
search the "x86_64" subdirectory when loading a shared library.
NB: We can't set x86-64 dl_platform to "x86-64" since ld.so will skip
the "haswell" and "xeon_phi" subdirectories on "haswell" and "xeon_phi"
machines.
Mike FABIAN [Sat, 9 Sep 2017 10:01:06 +0000 (12:01 +0200)]
Finish change from “Bengali” to “Bangla”
“Bengali” still remained in some comments in the bn_BD locale file,
in iso-639.def and in a test input file. Change it there as well.
“Bangla” is now used as the English name for this language in CLDR.
[BZ #14925]
* libio/tst-widetext.input: Change “Bengali” to “Bangla”.
* locale/iso-639.def: Change “Bengali” to “Bangla”.
* localedata/locales/bn_BD: “Bengali” was still used in some
comments. Change it to “Bangla”.
Steve Ellcey [Fri, 8 Sep 2017 18:02:09 +0000 (11:02 -0700)]
Add thunderx2t99 and thunderx2t99p1 CPU names to tunables list
* manual/tunables.texi (glibc.tune.cpu): Add thunderx2t99 and
thunderx2t99p1 to list of cpu names.
* sysdeps/unix/sysv/linux/aarch64/cpu-features.c (cpu_list):
Add thunderx2t99 and thunderx2t99p1 entries to cpu_list.
Joseph Myers [Fri, 8 Sep 2017 16:20:23 +0000 (16:20 +0000)]
Add Linux 4.13 constants to bits/fcntl-linux.h.
This patch adds new interfaces (F_GET_RW_HINT etc., and associated
RW[FH]_WRITE_LIFE_*) from Linux 4.13 to bits/fcntl-linux.h
(conditional on __USE_GNU).
Joseph Myers [Fri, 8 Sep 2017 16:19:21 +0000 (16:19 +0000)]
Add fcntl sealing interfaces from Linux 3.17 to bits/fcntl-linux.h.
While reviewing Linux 4.13 for glibc header changes needed, I noticed
that bits/fcntl-linux.h was missing F_ADD_SEALS etc. from Linux 3.17.
I didn't find any discussion indicating this omission is deliberate.
Now, these interfaces can only be used with file descriptors created
with memfd_create, and we don't have a memfd_create wrapper in glibc
(a patch was submitted in October 2014, albeit without documentation /
tests, and discussions continued over the next few months, but without
consensus on whether to add the interface - and we still lack any
general consensus on syscall wrappers), but I don't think that's a
reason to exclude the constants from bits/fcntl-linux.h (especially as
the header does not look compatible with simultaneously including
linux/fcntl.h).
(Some of those 2014/2015 discussions raised concerns about difficulty
using the memfd_create / sealing interface, but those seem to me more
like a question of whether it should be part of the OS-independent GNU
API - in my view, even fairly specialized syscalls ought to have
wrappers added to glibc if not obsolescent, but there may be cases
where we only want to include them in the Linux-specific API and
anything in the OS-independent GNU API should be different - rather
than being relevant to whether constants for use with fcntl should
appear in headers.)
git://sourceware.org / glibc.git / log