On some architectures, the parts of math_private.h relating to the
floating-point environment are in a separate file fenv_private.h
included from math_private.h. As this is purely an
architecture-specific convention used by several architectures,
however, all such architectures still need their own math_private.h,
even if it has nothing to do beyond #include <fenv_private.h> and
peculiarity of including the i386 file directly instead of having a
shared file in sysdeps/x86.
This patch makes the fenv_private.h name an architecture-independent
convention in glibc. The include of fenv_private.h from
math_private.h becomes architecture-independent (until callers are
updated to include fenv_private.h directly so the include from
math_private.h is no longer needed). Some architecture math_private.h
headers are removed if no longer needed, or renamed to fenv_private.h
if all they define belongs in that header; architecture fenv_private.h
headers now do require #include_next <fenv_private.h>. The i386
fenv_private.h file moves to sysdeps/x86/fpu/ to reflect how it is
actually shared with x86_64. The generic math_private.h gets a new
include of <stdbool.h>, as needed for bool in some prototypes in that
header (previously that was indirectly included via include/fenv.h,
which now only gets included too late in math_private.h, after those
prototypes).
Tested for x86_64 and x86, and tested with build-many-glibcs.py that
installed stripped shared libraries are unchanged by the patch.
* sysdeps/aarch64/fpu/fenv_private.h: New file. Based on ....
* sysdeps/aarch64/fpu/math_private.h: ... this file. All contents
moved to fenv_private.h except for ...
(TOINT_INTRINSICS): Kept in math_private.h.
(roundtoint): Likewise.
(converttoint): Likewise.
* sysdeps/arm/fenv_private.h: Change multiple-include guard to
[ARM_FENV_PRIVATE_H]. Include next <fenv_private.h>.
* sysdeps/arm/math_private.h: Remove.
* sysdeps/generic/fenv_private.h: New file. Contents moved from
....
* sysdeps/generic/math_private.h: ... this file. Include
<stdbool.h>. Do not include <fenv.h> or <get-rounding-mode.h>.
Include <fenv_private.h>. Remove functions and macros moved to
fenv_private.h.
* sysdeps/i386/fpu/math_private.h: Remove.
* sysdeps/mips/math_private.h: Move to ....
* sysdeps/mips/fpu/fenv_private.h: ... here. Change
multiple-include guard to [MIPS_FENV_PRIVATE_H]. Remove
[__mips_hard_float] conditional. Include next <fenv_private.h>.
* sysdeps/powerpc/fpu/fenv_private.h: Change multiple-include
guard to [POWERPC_FENV_PRIVATE_H]. Include next <fenv_private.h>.
* sysdeps/powerpc/fpu/math_private.h: Do not include
<fenv_private.h>.
* sysdeps/riscv/rvf/math_private.h: Move to ....
* sysdeps/riscv/rvf/fenv_private.h: ... here. Change
multiple-include guard to [RISCV_FENV_PRIVATE_H]. Include next
<fenv_private.h>.
* sysdeps/sparc/fpu/fenv_private.h: Change multiple-include guard
to [SPARC_FENV_PRIVATE_H]. Include next <fenv_private.h>.
* sysdeps/sparc/fpu/math_private.h: Remove.
* sysdeps/i386/fpu/fenv_private.h: Move to ....
* sysdeps/x86/fpu/fenv_private.h: ... here. Change
multiple-include guard to [X86_FENV_PRIVATE_H]. Include next
<fenv_private.h>.
* sysdeps/x86_64/fpu/math_private.h: Do not include
<sysdeps/i386/fpu/fenv_private.h>.
This patch currently only affects aarch64.
The roundtoint and converttoint internal functions are only called with small
values, so 32 bit result is enough for converttoint and it is a signed int
conversion so the return type is changed to int32_t.
The original idea was to help the compiler keeping the result in uint64_t,
then it's clear that no sign extension is needed and there is no accidental
undefined or implementation defined signed int arithmetics.
But it turns out gcc does a good job with inlining so changing the type has
no overhead and the semantics of the conversion is less surprising this way.
Since we want to allow the asuint64 (x + 0x1.8p52) style conversion, the top
bits were never usable and the existing code ensures that only the bottom
32 bits of the conversion result are used.
On aarch64 the neon intrinsics (which round ties to even) are changed to
round and lround (which round ties away from zero) this does not affect the
results in a significant way, but more portable (relies on round and lround
being inlined which works with -fno-math-errno).
The TOINT_SHIFT and TOINT_RINT macros were removed, only keep separate code
paths for TOINT_INTRINSICS and !TOINT_INTRINSICS.
* sysdeps/aarch64/fpu/math_private.h (roundtoint): Use round.
(converttoint): Use lround.
* sysdeps/ieee754/flt-32/math_config.h (roundtoint): Declare and
document the semantics when TOINT_INTRINSICS is set.
(converttoint): Likewise.
(TOINT_RINT): Remove.
(TOINT_SHIFT): Remove.
* sysdeps/ieee754/flt-32/e_expf.c (__expf): Remove the TOINT_RINT code
path.
This patch continues cleaning up math_private.h by moving the
math_opt_barrier and math_force_eval macros to a separate header
math-barriers.h.
At present, those macros are inside a "#ifndef math_opt_barrier" in
math_private.h to allow architectures to override them and then use
a separate math-barriers.h header, no such #ifndef or #include_next is
needed; architectures just have their own alternative version of
math-barriers.h when providing their own optimized versions that avoid
going through memory unnecessarily. The generic math-barriers.h has a
comment added to document these two macros.
In this patch, math_private.h is made to #include <math-barriers.h>,
so files using these macros do not need updating yet. That is because
of uses of math_force_eval in math_check_force_underflow and
math_check_force_underflow_nonneg, which are still defined in
math_private.h. Once those are moved out to a separate header, that
separate header can be made to include <math-barriers.h>, as can the
other files directly using these barrier macros, and then the include
of <math-barriers.h> from math_private.h can be removed.
Tested for x86_64 and x86. Also tested with build-many-glibcs.py that
installed stripped shared libraries are unchanged by this patch.
* sysdeps/generic/math-barriers.h: New file.
* sysdeps/generic/math_private.h [!math_opt_barrier]
(math_opt_barrier): Move to math-barriers.h.
[!math_opt_barrier] (math_force_eval): Likewise.
* sysdeps/aarch64/fpu/math-barriers.h: New file.
* sysdeps/aarch64/fpu/math_private.h (math_opt_barrier): Move to
math-barriers.h.
(math_force_eval): Likewise.
* sysdeps/alpha/fpu/math-barriers.h: New file.
* sysdeps/alpha/fpu/math_private.h (math_opt_barrier): Move to
math-barriers.h.
(math_force_eval): Likewise.
* sysdeps/x86/fpu/math-barriers.h: New file.
* sysdeps/i386/fpu/fenv_private.h (math_opt_barrier): Move to
math-barriers.h.
(math_force_eval): Likewise.
* sysdeps/m68k/m680x0/fpu/math_private.h: Move to....
* sysdeps/m68k/m680x0/fpu/math-barriers.h: ... here. Adjust
multiple-include guard for rename.
* sysdeps/powerpc/fpu/math-barriers.h: New file.
* sysdeps/powerpc/fpu/math_private.h (math_opt_barrier): Move to
math-barriers.h.
(math_force_eval): Likewise.
Remove the now unused target specific__ieee754_sqrt(f/l) inlines.
Also remove inlines of sqrt which are for really old GCC versions.
Removing these is desirable, under the general principle of leaving
such inlining to the compiler rather than trying to do it in installed
headers, especially when only very old compilers are affected.
Note that removing inlines for __ieee754_sqrt disables inlining in the
sqrt wrapper functions. Given the sqrt function will typically only be
called for negative arguments, it doesn't matter whether the inlining
happens or not.
* sysdeps/aarch64/fpu/math_private.h (__ieee754_sqrt): Remove.
(__ieee754_sqrtf): Remove.
* sysdeps/alpha/fpu/math_private.h (__ieee754_sqrt): Remove.
(__ieee754_sqrtf): Remove.
* sysdeps/generic/math-type-macros.h (M_SQRT): Use sqrt.
* sysdeps/m68k/m680x0/fpu/mathimpl.h (__ieee754_sqrt): Remove.
* sysdeps/powerpc/fpu/math_private.h (__ieee754_sqrt): Remove.
(__ieee754_sqrtf): Remove.
* sysdeps/s390/fpu/bits/mathinline.h: Remove file.
* sysdeps/sparc/fpu/bits/mathinline.h (sqrt) Remove.
(sqrtf): Remove.
(sqrtl): Remove.
(__ieee754_sqrt): Remove.
(__ieee754_sqrtf): Remove.
(__ieee754_sqrtl): Remove.
* sysdeps/m68k/m680x0/fpu/mathimpl.h (__ieee754_sqrt): Remove.
* sysdeps/x86/fpu/math_private.h (__ieee754_sqrt): Remove.
* sysdeps/x86_64/fpu/math_private.h (__ieee754_sqrt): Remove.
(__ieee754_sqrtf): Remove.
(__ieee754_sqrtl): Remove.
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.
Various C90 and UNIX98 libm functions call feraiseexcept, which is not
in those standards. This causes linknamespace test failures - except
on x86 / x86_64, where feraiseexcept is inline (for the relevant
constant arguments) in bits/fenv.h.
This patch fixes this by making those functions call __feraiseexcept
instead. All changes are applied to all architectures rather than
considering the possibility that some might not be needed in some
cases (e.g. x86) as it seems most maintainable to keep architectures
consistent.
Where __feraiseexcept does not exist, it is added, with feraiseexcept
made a weak alias; where it is a strong alias, it is made weak.
libm_hidden_def / libm_hidden_proto are used with __feraiseexcept
(this might in some cases improve code generation for existing calls
to __feraiseexcept in some code on some architectures). Where there
are dummy feraiseexcept macros (on architectures without
floating-point exceptions support, to avoid compile errors from
references to undefined FE_* macros), corresponding dummy
__feraiseexcept macros are added. And on x86, to ensure
__feraiseexcept calls still get inlined, the inline function in
bits/fenv.h is refactored so that most of it can be reused in an
inline __feraiseexcept in a separate include/bits/fenv.h.
Calls are changed in C90/UNIX98 functions, but generally not in
functions missing from those standards. They are also changed in
libc_fe* functions (on the basis that those might be used in any libm
function), and in feupdateenv (on the same basis - may be used, via
default libc_*, in any libm function - of course feupdateenv will need
changing to __feupdateenv in a subsequent patch to make that fully
namespace-clean).
No __feraiseexcept is added corresponding to the feraiseexcept in
powerpc bits/fenvinline.h, because that macro definition is
conditional on !defined __NO_MATH_INLINES, and glibc libm is built
with -D__NO_MATH_INLINES, so changing internal calls to use
__feraiseexcept should make no difference.
Tested for x86_64 (testsuite; the only change in disassembly of
installed shared libraries is a slight code reordering in clog10, of
no apparent significance). Also tested for MIPS, where (in the
configuration tested) it eliminates math.h linknamespace failures for
n32 and n64 (some for o32 remain because of other issues).
[BZ #17723]
* include/fenv.h (__feraiseexcept): Use libm_hidden_proto.
* math/fraiseexcpt.c (__feraiseexcept): Use libm_hidden_def.
* sysdeps/aarch64/fpu/fraiseexcpt.c (feraiseexcept): Rename to
__feraiseexcept and define as weak alias of __feraiseexcept. Use
libm_hidden_weak.
* sysdeps/arm/fraiseexcpt.c (feraiseexcept): Likewise.
* sysdeps/hppa/fpu/fraiseexcpt.c (feraiseexcept): Likewise.
* sysdeps/i386/fpu/fraiseexcpt.c (__feraiseexcept): Use
libm_hidden_def.
* sysdeps/ia64/fpu/fraiseexcpt.c (feraiseexcept): Rename to
__feraiseexcept and define as weak alias of __feraiseexcept. Use
libm_hidden_weak.
* sysdeps/m68k/coldfire/fpu/fraiseexcpt.c (feraiseexcept):
Likewise.
* sysdeps/microblaze/math_private.h (__feraiseexcept): New macro.
* sysdeps/mips/fpu/fraiseexcpt.c (feraiseexcept): Rename to
__feraiseexcept and define as weak alias of __feraiseexcept. Use
libm_hidden_weak.
* sysdeps/powerpc/fpu/fraiseexcpt.c (__feraiseexcept): Use
libm_hidden_def.
* sysdeps/powerpc/nofpu/fraiseexcpt.c (__feraiseexcept): Likewise.
* sysdeps/powerpc/powerpc32/e500/nofpu/fraiseexcpt.c
(__feraiseexcept): Likewise.
* sysdeps/s390/fpu/fraiseexcpt.c (feraiseexcept): Rename to
__feraiseexcept and define as weak alias of __feraiseexcept. Use
libm_hidden_weak.
* sysdeps/sh/sh4/fpu/fraiseexcpt.c (feraiseexcept): Likewise.
* sysdeps/sparc/fpu/fraiseexcpt.c (__feraiseexcept): Use
libm_hidden_def.
* sysdeps/tile/math_private.h (__feraiseexcept): New macro.
* sysdeps/unix/sysv/linux/alpha/fraiseexcpt.S (__feraiseexcept):
Use libm_hidden_def.
* sysdeps/x86_64/fpu/fraiseexcpt.c (__feraiseexcept): Use
libm_hidden_def.
(feraiseexcept): Define as weak not strong alias. Use
libm_hidden_weak.
* sysdeps/x86/fpu/bits/fenv.h (__feraiseexcept_invalid_divbyzero):
New inline function. Factored out of ...
(feraiseexcept): ... here. Use __feraiseexcept_invalid_divbyzero.
* sysdeps/x86/fpu/include/bits/fenv.h: New file.
* math/e_scalb.c (invalid_fn): Call __feraiseexcept instead of
feraiseexcept.
* math/w_acos.c (__acos): Likewise.
* math/w_asin.c (__asin): Likewise.
* math/w_ilogb.c (__ilogb): Likewise.
* math/w_j0.c (y0): Likewise.
* math/w_j1.c (y1): Likewise.
* math/w_jn.c (yn): Likewise.
* math/w_log.c (__log): Likewise.
* math/w_log10.c (__log10): Likewise.
* sysdeps/aarch64/fpu/feupdateenv.c (feupdateenv): Likewise.
* sysdeps/aarch64/fpu/math_private.h
(libc_feupdateenv_test_aarch64): Likewise.
* sysdeps/alpha/fpu/feupdateenv.c (__feupdateenv): Likewise.
* sysdeps/arm/fenv_private.h (libc_feupdateenv_test_vfp): Likewise.
* sysdeps/arm/feupdateenv.c (feupdateenv): Likewise.
* sysdeps/ia64/fpu/feupdateenv.c (feupdateenv): Likewise.
* sysdeps/m68k/fpu/feupdateenv.c (__feupdateenv): Likewise.
* sysdeps/mips/fpu/feupdateenv.c (feupdateenv): Likewise.
* sysdeps/powerpc/fpu/e_sqrt.c (__slow_ieee754_sqrt): Likewise.
* sysdeps/s390/fpu/feupdateenv.c (feupdateenv): Likewise.
* sysdeps/sh/sh4/fpu/feupdateenv.c (feupdateenv): Likewise.
* sysdeps/sparc/fpu/feupdateenv.c (__feupdateenv): Likewise.
further optimization. libc_feholdsetround_aarch64_ctx now only needs to
read the FPCR in the typical case, avoiding a redundant FPSR read.
Performance results show a good improvement (5-10% on sin()) on cores with
expensive FPCR/FPSR instructions.
