C99 wrongly specified a divide-by-zero exception for pow(+/- 0, -Inf);
C11 made it optional after this was pointed out, and the permission
for this exception has been removed in the current C2x draft. This
patch makes the glibc pow tests reflect the stricter requirement
(which follows the normal IEEE rules that a divide-by-zero exception
is for the case of exact infinite results from *finite* operands, not
for such results when any operand is infinite).
Tested for x86_64 and x86. (If any other pow implementation in glibc,
not exercised on those architectures, turns out to fail the stricter
test, it should be fixed to avoid the exception in this case.)
* math/libm-test-pow.inc (pow_test_data): Do not allow
divide-by-zero exception for pow(+/- 0, -Inf).
In my review
<https://sourceware.org/ml/libc-alpha/2018-06/msg00375.html> of a
patch for bug 23584, I expressed concern that the proposed changes
didn't deal with certain cases similar to the ones in the bug but
where test coverage was missing.
This patch adds such tests of fma (Inf, finite, finite) and fma
(finite, Inf, finite) to libm-test-fma.inc. It does *not* do anything
to fix the bug, simply adds test coverage to provide stronger evidence
of whether any proposed revised fix does address the cases I was
concerned with.
Tested for x86_64 and x86.
* math/libm-test-fma.inc (fma_test_data): Add more tests.
After my changes to move various macros, inlines and other content
from math_private.h to more specific headers, many files including
math_private.h no longer need to do so. Furthermore, since the
optimized inlines of various functions have been moved to
include/fenv.h or replaced by use of function names GCC inlines
automatically, a missing math_private.h include where one is
appropriate will reliably cause a build failure rather than possibly
causing code to be less well optimized while still building
successfully. Thus, this patch removes includes of math_private.h
that are now unnecessary. In the case of two RISC-V files, the
include is replaced by one of stdbool.h because the files in question
were relying on math_private.h to get a definition of bool.
Tested for x86_64 and x86, and with build-many-glibcs.py.
* math/fromfp.h: Do not include <math_private.h>.
* math/s_cacosh_template.c: Likewise.
* math/s_casin_template.c: Likewise.
* math/s_casinh_template.c: Likewise.
* math/s_ccos_template.c: Likewise.
* math/s_cproj_template.c: Likewise.
* math/s_fdim_template.c: Likewise.
* math/s_fmaxmag_template.c: Likewise.
* math/s_fminmag_template.c: Likewise.
* math/s_iseqsig_template.c: Likewise.
* math/s_ldexp_template.c: Likewise.
* math/s_nextdown_template.c: Likewise.
* math/w_log1p_template.c: Likewise.
* math/w_scalbln_template.c: Likewise.
* sysdeps/aarch64/fpu/feholdexcpt.c: Likewise.
* sysdeps/aarch64/fpu/fesetround.c: Likewise.
* sysdeps/aarch64/fpu/fgetexcptflg.c: Likewise.
* sysdeps/aarch64/fpu/ftestexcept.c: Likewise.
* sysdeps/aarch64/fpu/s_llrint.c: Likewise.
* sysdeps/aarch64/fpu/s_llrintf.c: Likewise.
* sysdeps/aarch64/fpu/s_lrint.c: Likewise.
* sysdeps/aarch64/fpu/s_lrintf.c: Likewise.
* sysdeps/i386/fpu/s_atanl.c: Likewise.
* sysdeps/i386/fpu/s_f32xaddf64.c: Likewise.
* sysdeps/i386/fpu/s_f32xsubf64.c: Likewise.
* sysdeps/i386/fpu/s_fdim.c: Likewise.
* sysdeps/i386/fpu/s_logbl.c: Likewise.
* sysdeps/i386/fpu/s_rintl.c: Likewise.
* sysdeps/i386/fpu/s_significandl.c: Likewise.
* sysdeps/ia64/fpu/s_matherrf.c: Likewise.
* sysdeps/ia64/fpu/s_matherrl.c: Likewise.
* sysdeps/ieee754/dbl-64/s_atan.c: Likewise.
* sysdeps/ieee754/dbl-64/s_cbrt.c: Likewise.
* sysdeps/ieee754/dbl-64/s_fma.c: Likewise.
* sysdeps/ieee754/dbl-64/s_fmaf.c: Likewise.
* sysdeps/ieee754/flt-32/s_cbrtf.c: Likewise.
* sysdeps/ieee754/k_standardf.c: Likewise.
* sysdeps/ieee754/k_standardl.c: Likewise.
* sysdeps/ieee754/ldbl-128ibm/s_copysignl.c: Likewise.
* sysdeps/ieee754/ldbl-64-128/s_finitel.c: Likewise.
* sysdeps/ieee754/ldbl-64-128/s_fpclassifyl.c: Likewise.
* sysdeps/ieee754/ldbl-64-128/s_isinfl.c: Likewise.
* sysdeps/ieee754/ldbl-64-128/s_isnanl.c: Likewise.
* sysdeps/ieee754/ldbl-64-128/s_signbitl.c: Likewise.
* sysdeps/ieee754/ldbl-96/s_cbrtl.c: Likewise.
* sysdeps/ieee754/ldbl-96/s_fma.c: Likewise.
* sysdeps/ieee754/ldbl-96/s_fmal.c: Likewise.
* sysdeps/ieee754/s_signgam.c: Likewise.
* sysdeps/powerpc/power5+/fpu/s_modf.c: Likewise.
* sysdeps/powerpc/power5+/fpu/s_modff.c: Likewise.
* sysdeps/powerpc/power7/fpu/s_logbf.c: Likewise.
* sysdeps/riscv/rv64/rvd/s_ceil.c: Likewise.
* sysdeps/riscv/rv64/rvd/s_floor.c: Likewise.
* sysdeps/riscv/rv64/rvd/s_nearbyint.c: Likewise.
* sysdeps/riscv/rv64/rvd/s_round.c: Likewise.
* sysdeps/riscv/rv64/rvd/s_roundeven.c: Likewise.
* sysdeps/riscv/rv64/rvd/s_trunc.c: Likewise.
* sysdeps/riscv/rvd/s_finite.c: Likewise.
* sysdeps/riscv/rvd/s_fmax.c: Likewise.
* sysdeps/riscv/rvd/s_fmin.c: Likewise.
* sysdeps/riscv/rvd/s_fpclassify.c: Likewise.
* sysdeps/riscv/rvd/s_isinf.c: Likewise.
* sysdeps/riscv/rvd/s_isnan.c: Likewise.
* sysdeps/riscv/rvd/s_issignaling.c: Likewise.
* sysdeps/riscv/rvf/fegetround.c: Likewise.
* sysdeps/riscv/rvf/feholdexcpt.c: Likewise.
* sysdeps/riscv/rvf/fesetenv.c: Likewise.
* sysdeps/riscv/rvf/fesetround.c: Likewise.
* sysdeps/riscv/rvf/feupdateenv.c: Likewise.
* sysdeps/riscv/rvf/fgetexcptflg.c: Likewise.
* sysdeps/riscv/rvf/ftestexcept.c: Likewise.
* sysdeps/riscv/rvf/s_ceilf.c: Likewise.
* sysdeps/riscv/rvf/s_finitef.c: Likewise.
* sysdeps/riscv/rvf/s_floorf.c: Likewise.
* sysdeps/riscv/rvf/s_fmaxf.c: Likewise.
* sysdeps/riscv/rvf/s_fminf.c: Likewise.
* sysdeps/riscv/rvf/s_fpclassifyf.c: Likewise.
* sysdeps/riscv/rvf/s_isinff.c: Likewise.
* sysdeps/riscv/rvf/s_isnanf.c: Likewise.
* sysdeps/riscv/rvf/s_issignalingf.c: Likewise.
* sysdeps/riscv/rvf/s_nearbyintf.c: Likewise.
* sysdeps/riscv/rvf/s_roundevenf.c: Likewise.
* sysdeps/riscv/rvf/s_roundf.c: Likewise.
* sysdeps/riscv/rvf/s_truncf.c: Likewise.
* sysdeps/riscv/rv64/rvd/s_rint.c: Include <stdbool.h> instead of
<math_private.h>.
* sysdeps/riscv/rvf/s_rintf.c: Likewise.
The algorithm is exp(y * log(x)), where log(x) is computed with about
1.3*2^-68 relative error (1.5*2^-68 without fma), returning the result
in two doubles, and the exp part uses the same algorithm (and lookup
tables) as exp, but takes the input as two doubles and a sign (to handle
negative bases with odd integer exponent). The __exp1 internal symbol
is no longer necessary.
There is separate code path when fma is not available but the worst case
error is about 0.54 ULP in both cases. The lookup table and consts for
log are 4168 bytes. The .rodata+.text is decreased by 37908 bytes on
aarch64. The non-nearest rounding error is less than 1 ULP.
Improvements on Cortex-A72 compared to current glibc master:
pow thruput: 2.40x in [0.01 11.1]x[0.01 11.1]
pow latency: 1.84x in [0.01 11.1]x[0.01 11.1]
Tested on
aarch64-linux-gnu (defined __FP_FAST_FMA, TOINT_INTRINSICS) and
arm-linux-gnueabihf (!defined __FP_FAST_FMA, !TOINT_INTRINSICS) and
x86_64-linux-gnu (!defined __FP_FAST_FMA, !TOINT_INTRINSICS) and
powerpc64le-linux-gnu (defined __FP_FAST_FMA, !TOINT_INTRINSICS) targets.
* NEWS: Mention pow improvements.
* math/Makefile (type-double-routines): Add e_pow_log_data.
* sysdeps/generic/math_private.h (__exp1): Remove.
* sysdeps/i386/fpu/e_pow_log_data.c: New file.
* sysdeps/ia64/fpu/e_pow_log_data.c: New file.
* sysdeps/ieee754/dbl-64/Makefile (CFLAGS-e_pow.c): Allow fma
contraction.
* sysdeps/ieee754/dbl-64/e_exp.c (__exp1): Remove.
(exp_inline): Remove.
(__ieee754_exp): Only single double input is handled.
* sysdeps/ieee754/dbl-64/e_pow.c: Rewrite.
* sysdeps/ieee754/dbl-64/e_pow_log_data.c: New file.
* sysdeps/ieee754/dbl-64/math_config.h (issignaling_inline): Define.
(__pow_log_data): Define.
* sysdeps/ieee754/dbl-64/upow.h: Remove.
* sysdeps/ieee754/dbl-64/upow.tbl: Remove.
* sysdeps/m68k/m680x0/fpu/e_pow_log_data.c: New file.
* sysdeps/x86_64/fpu/multiarch/Makefile (CFLAGS-e_pow-fma.c): Allow fma
contraction.
(CFLAGS-e_pow-fma4.c): Likewise.
Continuing the move to use, within libm, public names for libm
functions that can be inlined as built-in functions on many
architectures, this patch moves calls to __rint functions to call the
corresponding rint names instead, with asm redirection to __rint when
the calls are not inlined. The x86_64 math_private.h is removed as no
longer useful after this patch.
This patch is relative to a tree with my floor patch
<https://sourceware.org/ml/libc-alpha/2018-09/msg00148.html> applied,
and much the same considerations arise regarding possibly replacing an
IFUNC call with a direct inline expansion.
Tested for x86_64, and with build-many-glibcs.py.
* include/math.h [!_ISOMAC && !(__FINITE_MATH_ONLY__ &&
__FINITE_MATH_ONLY__ > 0) && !NO_MATH_REDIRECT] (rint): Redirect
using MATH_REDIRECT.
* sysdeps/aarch64/fpu/s_rint.c: Define NO_MATH_REDIRECT before
header inclusion.
* sysdeps/aarch64/fpu/s_rintf.c: Likewise.
* sysdeps/alpha/fpu/s_rint.c: Likewise.
* sysdeps/alpha/fpu/s_rintf.c: Likewise.
* sysdeps/i386/fpu/s_rintl.c: Likewise.
* sysdeps/ieee754/dbl-64/s_rint.c: Likewise.
* sysdeps/ieee754/dbl-64/wordsize-64/s_rint.c: Likewise.
* sysdeps/ieee754/float128/s_rintf128.c: Likewise.
* sysdeps/ieee754/flt-32/s_rintf.c: Likewise.
* sysdeps/ieee754/ldbl-128/s_rintl.c: Likewise.
* sysdeps/ieee754/ldbl-128ibm/s_rintl.c: Likewise.
* sysdeps/m68k/coldfire/fpu/s_rint.c: Likewise.
* sysdeps/m68k/coldfire/fpu/s_rintf.c: Likewise.
* sysdeps/m68k/m680x0/fpu/s_rint.c: Likewise.
* sysdeps/m68k/m680x0/fpu/s_rintf.c: Likewise.
* sysdeps/m68k/m680x0/fpu/s_rintl.c: Likewise.
* sysdeps/powerpc/fpu/s_rint.c: Likewise.
* sysdeps/powerpc/fpu/s_rintf.c: Likewise.
* sysdeps/riscv/rv64/rvd/s_rint.c: Likewise.
* sysdeps/riscv/rvf/s_rintf.c: Likewise.
* sysdeps/sparc/sparc32/sparcv9/fpu/multiarch/s_rint.c: Likewise.
* sysdeps/sparc/sparc32/sparcv9/fpu/multiarch/s_rintf.c: Likewise.
* sysdeps/sparc/sparc64/fpu/multiarch/s_rint.c: Likewise.
* sysdeps/sparc/sparc64/fpu/multiarch/s_rintf.c: Likewise.
* sysdeps/x86_64/fpu/multiarch/s_rint.c: Likewise.
* sysdeps/x86_64/fpu/multiarch/s_rintf.c: Likewise.
* sysdeps/x86_64/fpu/math_private.h: Remove file.
* math/e_scalb.c (invalid_fn): Use rint functions instead of
__rint variants.
* math/e_scalbf.c (invalid_fn): Likewise.
* math/e_scalbl.c (invalid_fn): Likewise.
* sysdeps/ieee754/dbl-64/e_gamma_r.c (__ieee754_gamma_r):
Likewise.
* sysdeps/ieee754/flt-32/e_gammaf_r.c (__ieee754_gammaf_r):
Likewise.
* sysdeps/ieee754/k_standard.c (__kernel_standard): Likewise.
* sysdeps/ieee754/k_standardl.c (__kernel_standard_l): Likewise.
* sysdeps/ieee754/ldbl-128/e_gammal_r.c (__ieee754_gammal_r):
Likewise.
* sysdeps/ieee754/ldbl-128ibm/e_gammal_r.c (__ieee754_gammal_r):
Likewise.
* sysdeps/ieee754/ldbl-96/e_gammal_r.c (__ieee754_gammal_r):
Likewise.
* sysdeps/powerpc/powerpc32/fpu/s_llrint.c (__llrint): Likewise.
* sysdeps/powerpc/powerpc32/fpu/s_llrintf.c (__llrintf): Likewise.
Similar to the changes that were made to call sqrt functions directly
in glibc, instead of __ieee754_sqrt variants, so that the compiler
could inline them automatically without needing special inline
definitions in lots of math_private.h headers, this patch makes libm
code call floor functions directly instead of __floor variants,
removing the inlines / macros for x86_64 (SSE4.1) and powerpc
(POWER5).
The redirection used to ensure that __ieee754_sqrt does still get
called when the compiler doesn't inline a built-in function expansion
is refactored so it can be applied to other functions; the refactoring
is arranged so it's not limited to unary functions either (it would be
reasonable to use this mechanism for copysign - removing the inline in
math_private_calls.h but also eliminating unnecessary local PLT entry
use in the cases (powerpc soft-float and e500v1, for IBM long double)
where copysign calls don't get inlined).
The point of this change is that more architectures can get floor
calls inlined where they weren't previously (AArch64, for example),
without needing special inline definitions in their math_private.h,
and existing such definitions in math_private.h headers can be
removed.
Note that it's possible that in some cases an inline may be used where
an IFUNC call was previously used - this is the case on x86_64, for
example. I think the direct calls to floor are still appropriate; if
there's any significant performance cost from inline SSE2 floor
instead of an IFUNC call ending up with SSE4.1 floor, that indicates
that either the function should be doing something else that's faster
than using floor at all, or it should itself have IFUNC variants, or
that the compiler choice of inlining for generic tuning should change
to allow for the possibility that, by not inlining, an SSE4.1 IFUNC
might be called at runtime - but not that glibc should avoid calling
floor internally. (After all, all the same considerations would apply
to any user program calling floor, where it might either be inlined or
left as an out-of-line call allowing for a possible IFUNC.)
Tested for x86_64, and with build-many-glibcs.py.
* include/math.h [!_ISOMAC && !(__FINITE_MATH_ONLY__ &&
__FINITE_MATH_ONLY__ > 0) && !NO_MATH_REDIRECT] (MATH_REDIRECT):
New macro.
[!_ISOMAC && !(__FINITE_MATH_ONLY__ && __FINITE_MATH_ONLY__ > 0)
&& !NO_MATH_REDIRECT] (MATH_REDIRECT_LDBL): Likewise.
[!_ISOMAC && !(__FINITE_MATH_ONLY__ && __FINITE_MATH_ONLY__ > 0)
&& !NO_MATH_REDIRECT] (MATH_REDIRECT_F128): Likewise.
[!_ISOMAC && !(__FINITE_MATH_ONLY__ && __FINITE_MATH_ONLY__ > 0)
&& !NO_MATH_REDIRECT] (MATH_REDIRECT_UNARY_ARGS): Likewise.
[!_ISOMAC && !(__FINITE_MATH_ONLY__ && __FINITE_MATH_ONLY__ > 0)
&& !NO_MATH_REDIRECT] (sqrt): Redirect using MATH_REDIRECT.
[!_ISOMAC && !(__FINITE_MATH_ONLY__ && __FINITE_MATH_ONLY__ > 0)
&& !NO_MATH_REDIRECT] (floor): Likewise.
* sysdeps/aarch64/fpu/s_floor.c: Define NO_MATH_REDIRECT before
header inclusion.
* sysdeps/aarch64/fpu/s_floorf.c: Likewise.
* sysdeps/ieee754/dbl-64/s_floor.c: Likewise.
* sysdeps/ieee754/dbl-64/wordsize-64/s_floor.c: Likewise.
* sysdeps/ieee754/float128/s_floorf128.c: Likewise.
* sysdeps/ieee754/flt-32/s_floorf.c: Likewise.
* sysdeps/ieee754/ldbl-128/s_floorl.c: Likewise.
* sysdeps/ieee754/ldbl-128ibm/s_floorl.c: Likewise.
* sysdeps/m68k/m680x0/fpu/s_floor_template.c: Likewise.
* sysdeps/powerpc/powerpc32/power4/fpu/multiarch/s_floor.c: Likewise.
* sysdeps/powerpc/powerpc32/power4/fpu/multiarch/s_floorf.c: Likewise.
* sysdeps/powerpc/powerpc64/fpu/multiarch/s_floor.c: Likewise.
* sysdeps/powerpc/powerpc64/fpu/multiarch/s_floorf.c: Likewise.
* sysdeps/riscv/rv64/rvd/s_floor.c: Likewise.
* sysdeps/riscv/rvf/s_floorf.c: Likewise.
* sysdeps/sparc/sparc64/fpu/multiarch/s_floor.c: Likewise.
* sysdeps/sparc/sparc64/fpu/multiarch/s_floorf.c: Likewise.
* sysdeps/x86_64/fpu/multiarch/s_floor.c: Likewise.
* sysdeps/x86_64/fpu/multiarch/s_floorf.c: Likewise.
* sysdeps/powerpc/fpu/math_private.h [_ARCH_PWR5X] (__floor):
Remove macro.
[_ARCH_PWR5X] (__floorf): Likewise.
* sysdeps/x86_64/fpu/math_private.h [__SSE4_1__] (__floor): Remove
inline function.
[__SSE4_1__] (__floorf): Likewise.
* math/w_lgamma_main.c (LGFUNC (__lgamma)): Use floor functions
instead of __floor variants.
* math/w_lgamma_r_compat.c (__lgamma_r): Likewise.
* math/w_lgammaf_main.c (LGFUNC (__lgammaf)): Likewise.
* math/w_lgammaf_r_compat.c (__lgammaf_r): Likewise.
* math/w_lgammal_main.c (LGFUNC (__lgammal)): Likewise.
* math/w_lgammal_r_compat.c (__lgammal_r): Likewise.
* math/w_tgamma_compat.c (__tgamma): Likewise.
* math/w_tgamma_template.c (M_DECL_FUNC (__tgamma)): Likewise.
* math/w_tgammaf_compat.c (__tgammaf): Likewise.
* math/w_tgammal_compat.c (__tgammal): Likewise.
* sysdeps/ieee754/dbl-64/e_lgamma_r.c (sin_pi): Likewise.
* sysdeps/ieee754/dbl-64/k_rem_pio2.c (__kernel_rem_pio2):
Likewise.
* sysdeps/ieee754/dbl-64/lgamma_neg.c (__lgamma_neg): Likewise.
* sysdeps/ieee754/flt-32/e_lgammaf_r.c (sin_pif): Likewise.
* sysdeps/ieee754/flt-32/lgamma_negf.c (__lgamma_negf): Likewise.
* sysdeps/ieee754/ldbl-128/e_lgammal_r.c (__ieee754_lgammal_r):
Likewise.
* sysdeps/ieee754/ldbl-128/e_powl.c (__ieee754_powl): Likewise.
* sysdeps/ieee754/ldbl-128/lgamma_negl.c (__lgamma_negl):
Likewise.
* sysdeps/ieee754/ldbl-128/s_expm1l.c (__expm1l): Likewise.
* sysdeps/ieee754/ldbl-128ibm/e_lgammal_r.c (__ieee754_lgammal_r):
Likewise.
* sysdeps/ieee754/ldbl-128ibm/e_powl.c (__ieee754_powl): Likewise.
* sysdeps/ieee754/ldbl-128ibm/lgamma_negl.c (__lgamma_negl):
Likewise.
* sysdeps/ieee754/ldbl-128ibm/s_expm1l.c (__expm1l): Likewise.
* sysdeps/ieee754/ldbl-128ibm/s_truncl.c (__truncl): Likewise.
* sysdeps/ieee754/ldbl-96/e_lgammal_r.c (sin_pi): Likewise.
* sysdeps/ieee754/ldbl-96/lgamma_negl.c (__lgamma_negl): Likewise.
* sysdeps/powerpc/power5+/fpu/s_modf.c (__modf): Likewise.
* sysdeps/powerpc/power5+/fpu/s_modff.c (__modff): Likewise.
Similar algorithm is used as in log: log2(2^k x) = k + log2(c) + log2(x/c)
where the last term is approximated by a polynomial of x/c - 1, the first
order coefficient is about 1/ln2 in this case.
There is separate code path when fma instruction is not available for
computing x/c - 1 precisely, for which the table size is doubled.
The worst case error is 0.547 ULP (0.55 without fma), the read only
global data size is 1168 bytes (2192 without fma) on aarch64. The
non-nearest rounding error is less than 1 ULP.
Improvements on Cortex-A72 compared to current glibc master:
log2 thruput: 2.00x in [0.01 11.1]
log2 latency: 2.04x in [0.01 11.1]
log2 thruput: 2.17x in [0.999 1.001]
log2 latency: 2.88x in [0.999 1.001]
Tested on
aarch64-linux-gnu (defined __FP_FAST_FMA)
arm-linux-gnueabihf (!defined __FP_FAST_FMA)
x86_64-linux-gnu (!defined __FP_FAST_FMA)
powerpc64le-linxu-gnu (defined __FP_FAST_FMA)
targets.
* NEWS: Mention log2 improvements.
* math/Makefile (type-double-routines): Add e_log2_data.
* sysdeps/i386/fpu/e_log2_data.c: New file.
* sysdeps/ia64/fpu/e_log2_data.c: New file.
* sysdeps/ieee754/dbl-64/e_log2.c: Rewrite.
* sysdeps/ieee754/dbl-64/e_log2_data.c: New file.
* sysdeps/ieee754/dbl-64/math_config.h (__log2_data): Add.
* sysdeps/ieee754/dbl-64/wordsize-64/e_log2.c: Remove.
* sysdeps/m68k/m680x0/fpu/e_log2_data.c: New file.
Optimized log using carefully generated lookup table with 1/c and log(c)
values for small intervalls around 1. The log(c) is very near a double
precision value, it has about 62 bits precision. The algorithm is
log(2^k x) = k log(2) + log(c) + log(x/c), where the last term is
approximated by a polynomial of x/c - 1. Near 1 a single polynomial of
x - 1 is used.
There is separate code path when fma instruction is not available for
computing x/c - 1 precisely, in which case the table size is doubled.
The code uses __builtin_fma under __FP_FAST_FMA to ensure it is inlined
as an instruction.
With the default configuration settings the worst case error is 0.519 ULP
(and 0.520 without fma), the rodata size is 2192 bytes (4240 without fma).
The non-nearest rounding error is less than 1 ULP.
Improvements on Cortex-A72 compared to current glibc master:
log thruput: 3.28x in [0.01 11.1]
log latency: 2.23x in [0.01 11.1]
log thruput: 1.56x in [0.999 1.001]
log latency: 1.57x in [0.999 1.001]
Tested on
aarch64-linux-gnu (defined __FP_FAST_FMA)
arm-linux-gnueabihf (!defined __FP_FAST_FMA)
x86_64-linux-gnu (!defined __FP_FAST_FMA)
powerpc64le-linux-gnu (defined __FP_FAST_FMA)
targets.
* NEWS: Mention log improvement.
* math/Makefile (type-double-routines): Add e_log_data.
* sysdeps/i386/fpu/e_log_data.c: New file.
* sysdeps/ia64/fpu/e_log_data.c: New file.
* sysdeps/ieee754/dbl-64/e_log.c: Rewrite.
* sysdeps/ieee754/dbl-64/e_log_data.c: New file.
* sysdeps/ieee754/dbl-64/math_config.h (__log_data): Add.
* sysdeps/ieee754/dbl-64/ulog.h: Remove.
* sysdeps/ieee754/dbl-64/ulog.tbl: Remove.
* sysdeps/m68k/m680x0/fpu/e_log_data.c: New file.
Optimized exp and exp2 implementations using a lookup table for
fractional powers of 2. There are several variants, see e_exp_data.c,
they can be selected by modifying math_config.h allowing different
tradeoffs.
The default selection should be acceptable as generic libm code.
Worst case error is 0.509 ULP for exp and 0.507 ULP for exp2, on
aarch64 the rodata size is 2160 bytes, shared between exp and exp2.
On aarch64 .text + .rodata size decreased by 24912 bytes.
The non-nearest rounding error is less than 1 ULP even on targets
without efficient round implementation (although the error rate is
higher in that case). Targets with single instruction, rounding mode
independent, to nearest integer rounding and conversion can use them
by setting TOINT_INTRINSICS and adding the necessary code to their
math_private.h.
The __exp1 code uses the same algorithm, so the error bound of pow
increased a bit.
New double precision error handling code was added following the
style of the single precision error handling code.
Improvements on Cortex-A72 compared to current glibc master:
exp thruput: 1.61x in [-9.9 9.9]
exp latency: 1.53x in [-9.9 9.9]
exp thruput: 1.13x in [0.5 1]
exp latency: 1.30x in [0.5 1]
exp2 thruput: 2.03x in [-9.9 9.9]
exp2 latency: 1.64x in [-9.9 9.9]
For small (< 1) inputs the current exp code uses a separate algorithm
so the speed up there is less.
Was tested on
aarch64-linux-gnu (TOINT_INTRINSICS, fma contraction) and
arm-linux-gnueabihf (!TOINT_INTRINSICS, no fma contraction) and
x86_64-linux-gnu (!TOINT_INTRINSICS, no fma contraction) and
powerpc64le-linux-gnu (!TOINT_INTRINSICS, fma contraction) targets,
only non-nearest rounding ulp errors increase and they are within
acceptable bounds (ulp updates are in separate patches).
* NEWS: Mention exp and exp2 improvements.
* math/Makefile (libm-support): Remove t_exp.
(type-double-routines): Add math_err and e_exp_data.
* sysdeps/aarch64/libm-test-ulps: Update.
* sysdeps/arm/libm-test-ulps: Update.
* sysdeps/i386/fpu/e_exp_data.c: New file.
* sysdeps/i386/fpu/math_err.c: New file.
* sysdeps/i386/fpu/t_exp.c: Remove.
* sysdeps/ia64/fpu/e_exp_data.c: New file.
* sysdeps/ia64/fpu/math_err.c: New file.
* sysdeps/ia64/fpu/t_exp.c: Remove.
* sysdeps/ieee754/dbl-64/e_exp.c: Rewrite.
* sysdeps/ieee754/dbl-64/e_exp2.c: Rewrite.
* sysdeps/ieee754/dbl-64/e_exp_data.c: New file.
* sysdeps/ieee754/dbl-64/e_pow.c (__ieee754_pow): Update error bound.
* sysdeps/ieee754/dbl-64/eexp.tbl: Remove.
* sysdeps/ieee754/dbl-64/math_config.h: New file.
* sysdeps/ieee754/dbl-64/math_err.c: New file.
* sysdeps/ieee754/dbl-64/t_exp.c: Remove.
* sysdeps/ieee754/dbl-64/t_exp2.h: Remove.
* sysdeps/ieee754/dbl-64/uexp.h: Remove.
* sysdeps/ieee754/dbl-64/uexp.tbl: Remove.
* sysdeps/m68k/m680x0/fpu/e_exp_data.c: New file.
* sysdeps/m68k/m680x0/fpu/math_err.c: New file.
* sysdeps/m68k/m680x0/fpu/t_exp.c: Remove.
* sysdeps/powerpc/fpu/libm-test-ulps: Update.
* sysdeps/x86_64/fpu/libm-test-ulps: Update.
<fenv_private.h> has inline versions of various <fenv.h> functions,
and their __fe* variants, for systems (generally soft-float) without
support for floating-point exceptions, rounding modes or both.
Having these inlines in a separate header introduces a risk of a
source file including <fenv.h> and compiling OK on x86_64, but failing
to compile (because the feraiseexcept inline is actually a macro that
discards its argument, to avoid the need for #ifdef FE_INVALID
conditionals), or not being properly optimized, on systems without the
exceptions and rounding modes support (when these inlines were in
math_private.h, we had a few cases where this broke the build because
there was no obvious reason for a file to need math_private.h and it
didn't need that header on x86_64). By moving those inlines to
include/fenv.h, this risk can be avoided, and fenv_private.h becomes
more clearly defined as specifically the header for the internal
libc_fe* and SET_RESTORE_ROUND* interfaces.
This patch makes that move, removing fenv_private.h includes that are
no longer needed (or replacing them by fenv.h includes in a few cases
that didn't already have such an include).
Tested for x86_64 and x86, and tested with build-many-glibcs.py that
installed stripped shared libraries are unchanged by the patch.
* sysdeps/generic/fenv_private.h [FE_ALL_EXCEPT == 0]: Move this
code ....
[!FE_HAVE_ROUNDING_MODES]: And this code ....
* include/fenv.h [!_ISOMAC]: ... to here.
* math/fraiseexcpt.c (__feraiseexcept): Undefine as macro.
(feraiseexcept): Likewise.
* math/fromfp.h: Do not include <fenv_private.h>.
* math/s_cexp_template.c: Likewise.
* math/s_csin_template.c: Likewise.
* math/s_csinh_template.c: Likewise.
* math/s_ctan_template.c: Likewise.
* math/s_ctanh_template.c: Likewise.
* math/s_iseqsig_template.c: Likewise.
* math/w_acos_compat.c: Likewise.
* math/w_acosf_compat.c: Likewise.
* math/w_acosl_compat.c: Likewise.
* math/w_asin_compat.c: Likewise.
* math/w_asinf_compat.c: Likewise.
* math/w_asinl_compat.c: Likewise.
* math/w_j0_compat.c: Likewise.
* math/w_j0f_compat.c: Likewise.
* math/w_j0l_compat.c: Likewise.
* math/w_j1_compat.c: Likewise.
* math/w_j1f_compat.c: Likewise.
* math/w_j1l_compat.c: Likewise.
* math/w_jn_compat.c: Likewise.
* math/w_jnf_compat.c: Likewise.
* math/w_log10_compat.c: Likewise.
* math/w_log10f_compat.c: Likewise.
* math/w_log10l_compat.c: Likewise.
* math/w_log2_compat.c: Likewise.
* math/w_log2f_compat.c: Likewise.
* math/w_log2l_compat.c: Likewise.
* math/w_log_compat.c: Likewise.
* math/w_logf_compat.c: Likewise.
* math/w_logl_compat.c: Likewise.
* sysdeps/ieee754/dbl-64/s_llrint.c: Likewise.
* sysdeps/ieee754/dbl-64/s_llround.c: Likewise.
* sysdeps/ieee754/dbl-64/s_lrint.c: Likewise.
* sysdeps/ieee754/dbl-64/s_lround.c: Likewise.
* sysdeps/ieee754/dbl-64/wordsize-64/s_lround.c: Likewise.
* sysdeps/ieee754/flt-32/s_llrintf.c: Likewise.
* sysdeps/ieee754/flt-32/s_llroundf.c: Likewise.
* sysdeps/ieee754/flt-32/s_lrintf.c: Likewise.
* sysdeps/ieee754/flt-32/s_lroundf.c: Likewise.
* sysdeps/ieee754/k_standardl.c: Likewise.
* sysdeps/ieee754/ldbl-128/e_expl.c: Likewise.
* sysdeps/ieee754/ldbl-128/s_fmal.c: Likewise.
* sysdeps/ieee754/ldbl-128/s_llrintl.c: Likewise.
* sysdeps/ieee754/ldbl-128/s_llroundl.c: Likewise.
* sysdeps/ieee754/ldbl-128/s_lrintl.c: Likewise.
* sysdeps/ieee754/ldbl-128/s_lroundl.c: Likewise.
* sysdeps/ieee754/ldbl-128/s_nearbyintl.c: Likewise.
* sysdeps/ieee754/ldbl-128ibm/s_llrintl.c: Likewise.
* sysdeps/ieee754/ldbl-128ibm/s_llroundl.c: Likewise.
* sysdeps/ieee754/ldbl-128ibm/s_lrintl.c: Likewise.
* sysdeps/ieee754/ldbl-128ibm/s_lroundl.c: Likewise.
* sysdeps/ieee754/ldbl-96/s_fma.c: Likewise.
* sysdeps/ieee754/ldbl-96/s_fmal.c: Likewise.
* sysdeps/ieee754/ldbl-96/s_llrintl.c: Likewise.
* sysdeps/ieee754/ldbl-96/s_llroundl.c: Likewise.
* sysdeps/ieee754/ldbl-96/s_lrintl.c: Likewise.
* sysdeps/ieee754/ldbl-96/s_lroundl.c: Likewise.
* math/w_ilogb_template.c: Include <fenv.h> instead of
<fenv_private.h>.
* math/w_llogb_template.c: Likewise.
* sysdeps/powerpc/fpu/e_sqrt.c: Likewise.
* sysdeps/powerpc/fpu/e_sqrtf.c: Likewise.
Continuing the clean-up related to the catch-all math_private.h
header, this patch stops math_private.h from including fenv_private.h.
Instead, fenv_private.h is included directly from those users of
math_private.h that also used interfaces from fenv_private.h. No
attempt is made to remove unused includes of math_private.h, but that
is a natural followup.
(However, since math_private.h sometimes defines optimized versions of
math.h interfaces or __* variants thereof, as well as defining its own
interfaces, I think it might make sense to get all those optimized
versions included from include/math.h, not requiring a separate header
at all, before eliminating unused math_private.h includes - that
avoids a file quietly becoming less-optimized if someone adds a call
to one of those interfaces without restoring a math_private.h include
to that file.)
There is still a pitfall that if code uses plain fe* and __fe*
interfaces, but only includes fenv.h and not fenv_private.h or (before
this patch) math_private.h, it will compile on platforms with
exceptions and rounding modes but not get the optimized versions (and
possibly not compile) on platforms without exception and rounding mode
support, so making it easy to break the build for such platforms
accidentally.
I think it would be most natural to move the inlines / macros for fe*
and __fe* in the case of no exceptions and rounding modes into
include/fenv.h, so that all code including fenv.h with _ISOMAC not
defined automatically gets them. Then fenv_private.h would be purely
the header for the libc_fe*, SET_RESTORE_ROUND etc. internal
interfaces and the risk of breaking the build on other platforms than
the one you tested on because of a missing fenv_private.h include
would be much reduced (and there would be some unused fenv_private.h
includes to remove along with unused math_private.h includes).
Tested for x86_64 and x86, and tested with build-many-glibcs.py that
installed stripped shared libraries are unchanged by this patch.
* sysdeps/generic/math_private.h: Do not include <fenv_private.h>.
* math/fromfp.h: Include <fenv_private.h>.
* math/math-narrow.h: Likewise.
* math/s_cexp_template.c: Likewise.
* math/s_csin_template.c: Likewise.
* math/s_csinh_template.c: Likewise.
* math/s_ctan_template.c: Likewise.
* math/s_ctanh_template.c: Likewise.
* math/s_iseqsig_template.c: Likewise.
* math/w_acos_compat.c: Likewise.
* math/w_acosf_compat.c: Likewise.
* math/w_acosl_compat.c: Likewise.
* math/w_asin_compat.c: Likewise.
* math/w_asinf_compat.c: Likewise.
* math/w_asinl_compat.c: Likewise.
* math/w_ilogb_template.c: Likewise.
* math/w_j0_compat.c: Likewise.
* math/w_j0f_compat.c: Likewise.
* math/w_j0l_compat.c: Likewise.
* math/w_j1_compat.c: Likewise.
* math/w_j1f_compat.c: Likewise.
* math/w_j1l_compat.c: Likewise.
* math/w_jn_compat.c: Likewise.
* math/w_jnf_compat.c: Likewise.
* math/w_llogb_template.c: Likewise.
* math/w_log10_compat.c: Likewise.
* math/w_log10f_compat.c: Likewise.
* math/w_log10l_compat.c: Likewise.
* math/w_log2_compat.c: Likewise.
* math/w_log2f_compat.c: Likewise.
* math/w_log2l_compat.c: Likewise.
* math/w_log_compat.c: Likewise.
* math/w_logf_compat.c: Likewise.
* math/w_logl_compat.c: Likewise.
* sysdeps/aarch64/fpu/feholdexcpt.c: Likewise.
* sysdeps/aarch64/fpu/fesetround.c: Likewise.
* sysdeps/aarch64/fpu/fgetexcptflg.c: Likewise.
* sysdeps/aarch64/fpu/ftestexcept.c: Likewise.
* sysdeps/ieee754/dbl-64/e_atan2.c: Likewise.
* sysdeps/ieee754/dbl-64/e_exp.c: Likewise.
* sysdeps/ieee754/dbl-64/e_exp2.c: Likewise.
* sysdeps/ieee754/dbl-64/e_gamma_r.c: Likewise.
* sysdeps/ieee754/dbl-64/e_jn.c: Likewise.
* sysdeps/ieee754/dbl-64/e_pow.c: Likewise.
* sysdeps/ieee754/dbl-64/e_remainder.c: Likewise.
* sysdeps/ieee754/dbl-64/e_sqrt.c: Likewise.
* sysdeps/ieee754/dbl-64/gamma_product.c: Likewise.
* sysdeps/ieee754/dbl-64/lgamma_neg.c: Likewise.
* sysdeps/ieee754/dbl-64/s_atan.c: Likewise.
* sysdeps/ieee754/dbl-64/s_fma.c: Likewise.
* sysdeps/ieee754/dbl-64/s_fmaf.c: Likewise.
* sysdeps/ieee754/dbl-64/s_llrint.c: Likewise.
* sysdeps/ieee754/dbl-64/s_llround.c: Likewise.
* sysdeps/ieee754/dbl-64/s_lrint.c: Likewise.
* sysdeps/ieee754/dbl-64/s_lround.c: Likewise.
* sysdeps/ieee754/dbl-64/s_nearbyint.c: Likewise.
* sysdeps/ieee754/dbl-64/s_sin.c: Likewise.
* sysdeps/ieee754/dbl-64/s_sincos.c: Likewise.
* sysdeps/ieee754/dbl-64/s_tan.c: Likewise.
* sysdeps/ieee754/dbl-64/wordsize-64/s_lround.c: Likewise.
* sysdeps/ieee754/dbl-64/wordsize-64/s_nearbyint.c: Likewise.
* sysdeps/ieee754/dbl-64/x2y2m1.c: Likewise.
* sysdeps/ieee754/float128/float128_private.h: Likewise.
* sysdeps/ieee754/flt-32/e_gammaf_r.c: Likewise.
* sysdeps/ieee754/flt-32/e_j1f.c: Likewise.
* sysdeps/ieee754/flt-32/e_jnf.c: Likewise.
* sysdeps/ieee754/flt-32/lgamma_negf.c: Likewise.
* sysdeps/ieee754/flt-32/s_llrintf.c: Likewise.
* sysdeps/ieee754/flt-32/s_llroundf.c: Likewise.
* sysdeps/ieee754/flt-32/s_lrintf.c: Likewise.
* sysdeps/ieee754/flt-32/s_lroundf.c: Likewise.
* sysdeps/ieee754/flt-32/s_nearbyintf.c: Likewise.
* sysdeps/ieee754/k_standardl.c: Likewise.
* sysdeps/ieee754/ldbl-128/e_expl.c: Likewise.
* sysdeps/ieee754/ldbl-128/e_gammal_r.c: Likewise.
* sysdeps/ieee754/ldbl-128/e_j1l.c: Likewise.
* sysdeps/ieee754/ldbl-128/e_jnl.c: Likewise.
* sysdeps/ieee754/ldbl-128/gamma_productl.c: Likewise.
* sysdeps/ieee754/ldbl-128/lgamma_negl.c: Likewise.
* sysdeps/ieee754/ldbl-128/s_fmal.c: Likewise.
* sysdeps/ieee754/ldbl-128/s_llrintl.c: Likewise.
* sysdeps/ieee754/ldbl-128/s_llroundl.c: Likewise.
* sysdeps/ieee754/ldbl-128/s_lrintl.c: Likewise.
* sysdeps/ieee754/ldbl-128/s_lroundl.c: Likewise.
* sysdeps/ieee754/ldbl-128/s_nearbyintl.c: Likewise.
* sysdeps/ieee754/ldbl-128/x2y2m1l.c: Likewise.
* sysdeps/ieee754/ldbl-128ibm/e_expl.c: Likewise.
* sysdeps/ieee754/ldbl-128ibm/e_gammal_r.c: Likewise.
* sysdeps/ieee754/ldbl-128ibm/e_j1l.c: Likewise.
* sysdeps/ieee754/ldbl-128ibm/e_jnl.c: Likewise.
* sysdeps/ieee754/ldbl-128ibm/lgamma_negl.c: Likewise.
* sysdeps/ieee754/ldbl-128ibm/s_fmal.c: Likewise.
* sysdeps/ieee754/ldbl-128ibm/s_llrintl.c: Likewise.
* sysdeps/ieee754/ldbl-128ibm/s_llroundl.c: Likewise.
* sysdeps/ieee754/ldbl-128ibm/s_lrintl.c: Likewise.
* sysdeps/ieee754/ldbl-128ibm/s_lroundl.c: Likewise.
* sysdeps/ieee754/ldbl-128ibm/s_rintl.c: Likewise.
* sysdeps/ieee754/ldbl-128ibm/x2y2m1l.c: Likewise.
* sysdeps/ieee754/ldbl-96/e_gammal_r.c: Likewise.
* sysdeps/ieee754/ldbl-96/e_jnl.c: Likewise.
* sysdeps/ieee754/ldbl-96/gamma_productl.c: Likewise.
* sysdeps/ieee754/ldbl-96/lgamma_negl.c: Likewise.
* sysdeps/ieee754/ldbl-96/s_fma.c: Likewise.
* sysdeps/ieee754/ldbl-96/s_fmal.c: Likewise.
* sysdeps/ieee754/ldbl-96/s_llrintl.c: Likewise.
* sysdeps/ieee754/ldbl-96/s_llroundl.c: Likewise.
* sysdeps/ieee754/ldbl-96/s_lrintl.c: Likewise.
* sysdeps/ieee754/ldbl-96/s_lroundl.c: Likewise.
* sysdeps/ieee754/ldbl-96/x2y2m1l.c: Likewise.
* sysdeps/powerpc/fpu/e_sqrt.c: Likewise.
* sysdeps/powerpc/fpu/e_sqrtf.c: Likewise.
* sysdeps/riscv/rv64/rvd/s_ceil.c: Likewise.
* sysdeps/riscv/rv64/rvd/s_floor.c: Likewise.
* sysdeps/riscv/rv64/rvd/s_nearbyint.c: Likewise.
* sysdeps/riscv/rv64/rvd/s_round.c: Likewise.
* sysdeps/riscv/rv64/rvd/s_roundeven.c: Likewise.
* sysdeps/riscv/rv64/rvd/s_trunc.c: Likewise.
* sysdeps/riscv/rvd/s_finite.c: Likewise.
* sysdeps/riscv/rvd/s_fmax.c: Likewise.
* sysdeps/riscv/rvd/s_fmin.c: Likewise.
* sysdeps/riscv/rvd/s_fpclassify.c: Likewise.
* sysdeps/riscv/rvd/s_isinf.c: Likewise.
* sysdeps/riscv/rvd/s_isnan.c: Likewise.
* sysdeps/riscv/rvd/s_issignaling.c: Likewise.
* sysdeps/riscv/rvf/fegetround.c: Likewise.
* sysdeps/riscv/rvf/feholdexcpt.c: Likewise.
* sysdeps/riscv/rvf/fesetenv.c: Likewise.
* sysdeps/riscv/rvf/fesetround.c: Likewise.
* sysdeps/riscv/rvf/feupdateenv.c: Likewise.
* sysdeps/riscv/rvf/fgetexcptflg.c: Likewise.
* sysdeps/riscv/rvf/ftestexcept.c: Likewise.
* sysdeps/riscv/rvf/s_ceilf.c: Likewise.
* sysdeps/riscv/rvf/s_finitef.c: Likewise.
* sysdeps/riscv/rvf/s_floorf.c: Likewise.
* sysdeps/riscv/rvf/s_fmaxf.c: Likewise.
* sysdeps/riscv/rvf/s_fminf.c: Likewise.
* sysdeps/riscv/rvf/s_fpclassifyf.c: Likewise.
* sysdeps/riscv/rvf/s_isinff.c: Likewise.
* sysdeps/riscv/rvf/s_isnanf.c: Likewise.
* sysdeps/riscv/rvf/s_issignalingf.c: Likewise.
* sysdeps/riscv/rvf/s_nearbyintf.c: Likewise.
* sysdeps/riscv/rvf/s_roundevenf.c: Likewise.
* sysdeps/riscv/rvf/s_roundf.c: Likewise.
* sysdeps/riscv/rvf/s_truncf.c: Likewise.
When converting gen-libm-test to Python, in one place I noted a bug in
the old Perl version that I preserved in the Python version so that
the generated output files were the same with both versions, as such
comparisons help give confidence in the correctness of such a rewrite
of a script. Now that the conversion has been done, this patch fixes
that bug, by arranging for tests with plus_oflow or minus_oflow
results (manually written tests in libm-test-*.inc that have
overflowing results that thus depend on the rounding mode) to be
properly treated as having non-finite results, and thus not run for
the __FINITE_MATH_ONLY__ tests. (As the affected tests in fact did
pass for __FINITE_MATH_ONLY__ testing, this is just a matter of
logical correctness in the choice of which tests run for that case,
rather than fixing any actual test failures.)
Tested for x86_64.
* math/gen-libm-test.py (gen_test_args_res): Also treat plus_oflow
and minus_oflow as non-finite.
This patch completes the move of ROUNDING_TESTS_* macros to typo-proof
conventions by stopping redefining them in test-*-vlen*.h. Instead,
libm-test-driver.c is made to check TEST_MATHVEC when setting
non-to-nearest rounding modes.
Tested for x86_64.
* math/test-double-vlen2.h: Don't include <math-tests-rounding.h>.
(ROUNDING_TESTS_double): Remove.
* math/test-double-vlen4.h: Don't include <math-tests-rounding.h>.
(ROUNDING_TESTS_double): Remove.
* math/test-double-vlen8.h: Don't include <math-tests-rounding.h>.
(ROUNDING_TESTS_double): Remove.
* math/test-float-vlen16.h: Don't include <math-tests-rounding.h>.
(ROUNDING_TESTS_float): Remove.
* math/test-float-vlen4.h: Don't include <math-tests-rounding.h>.
(ROUNDING_TESTS_float): Remove.
* math/test-float-vlen8.h: Don't include <math-tests-rounding.h>.
(ROUNDING_TESTS_float): Remove.
* math/libm-test-driver.c (IF_ROUND_INIT_FE_DOWNWARD): Check
!TEST_MATHVEC here.
(IF_ROUND_INIT_FE_TOWARDZERO): Likewise.
(IF_ROUND_INIT_FE_UPWARD): Likewise.
Continuing moving macros out of math-tests.h to smaller headers
following typo-proof conventions instead of using #ifndef, this patch
moves the ROUNDING_TESTS_* macros for individual types out to their
own sysdeps header.
In the soft-float case where FE_TONEAREST is the only rounding mode
macro defined, there is no need to define ROUNDING_TESTS_*; it is only
necessary when rounding modes macros are defined that may not be
supported at runtime. Thus, the ROUNDING_TESTS_* definitions for some
configurations are just removed, not moved to new
math-tests-rounding.h headers; the only architectures needing
math-tests-rounding.h are those where the macros are defined in
bits/fenv.h because of the possibility of a soft-float compilation
using a hard-float glibc with the same ABI (i.e., ARM and RISC-V).
The test-*-vlen*.h headers, by using #undef, do not yet follow
typo-proof conventions (but they no longer implicitly rely on being
included before math-tests.h, and this area can always be cleaned up
further in future).
Tested with build-many-glibcs.py.
* sysdeps/generic/math-tests-rounding.h: New file.
* sysdeps/generic/math-tests.h: Include <math-tests-rounding.h>.
(ROUNDING_TESTS_float): Do not define here.
(ROUNDING_TESTS_double): Likewise.
(ROUNDING_TESTS_long_double): Likewise.
(ROUNDING_TESTS_float128): Likewise.
* math/test-double-vlen2.h: Include <math-tests-rounding.h>.
(ROUNDING_TESTS_double): Undefine before defining.
* math/test-double-vlen4.h: Include <math-tests-rounding.h>.
(ROUNDING_TESTS_double): Undefine before defining.
* math/test-double-vlen8.h: Include <math-tests-rounding.h>.
(ROUNDING_TESTS_double): Undefine before defining.
* math/test-float-vlen16.h: Include <math-tests-rounding.h>.
(ROUNDING_TESTS_float): Undefine before defining.
* math/test-float-vlen4.h: Include <math-tests-rounding.h>.
(ROUNDING_TESTS_float): Undefine before defining.
* math/test-float-vlen8.h: Include <math-tests-rounding.h>.
(ROUNDING_TESTS_float): Undefine before defining.
* sysdeps/arm/nofpu/math-tests-rounding.h: New file.
* sysdeps/arm/math-tests.h [__SOFTFP__] (ROUNDING_TESTS_float): Do
not define here.
[__SOFTFP__] (ROUNDING_TESTS_double): Likewise.
[__SOFTFP__] (ROUNDING_TESTS_long_double): Likewise.
* sysdeps/riscv/nofpu/math-tests-rounding.h: New file.
* sysdeps/riscv/math-tests.h [!__riscv_flen]
(ROUNDING_TESTS_float): Do not define here.
[!__riscv_flen] (ROUNDING_TESTS_double): Likewise.
[!__risv_flen] (ROUNDING_TESTS_long_double): Likewise.
* sysdeps/m68k/coldfire/math-tests.h [!__mcffpu__]
(ROUNDING_TESTS_float): Likewise.
[!__mcffpu__] (ROUNDING_TESTS_double): Likewise.
[!__mcffpu__] (ROUNDING_TESTS_long_double): Likewise.
* sysdeps/mips/math-tests.h [__mips_soft_float]
(ROUNDING_TESTS_float): Likewise.
[__mips_soft_float] (ROUNDING_TESTS_double): Likewise.
[__mips_soft_float] (ROUNDING_TESTS_long_double): Likewise.
* sysdeps/nios2/math-tests.h (ROUNDING_TESTS_float): Likewise.
(ROUNDING_TESTS_double): Likewise.
(ROUNDING_TESTS_long_double): Likewise.
This patch is a complete rewrite of sincosf. The new version is
significantly faster, as well as simple and accurate.
The worst-case ULP is 0.5607, maximum relative error is 0.5303 * 2^-23 over
all 4 billion inputs. In non-nearest rounding modes the error is 1ULP.
The algorithm uses 3 main cases: small inputs which don't need argument
reduction, small inputs which need a simple range reduction and large inputs
requiring complex range reduction. The code uses approximate integer
comparisons to quickly decide between these cases.
The small range reducer uses a single reduction step to handle values up to
120.0. It is fastest on targets which support inlined round instructions.
The large range reducer uses integer arithmetic for simplicity. It does a
32x96 bit multiply to compute a 64-bit modulo result. This is more than
accurate enough to handle the worst-case cancellation for values close to
an integer multiple of PI/4. It could be further optimized, however it is
already much faster than necessary.
sincosf throughput gains on Cortex-A72:
* |x| < 0x1p-12 : 1.6x
* |x| < M_PI_4 : 1.7x
* |x| < 2 * M_PI: 1.5x
* |x| < 120.0 : 1.8x
* |x| < Inf : 2.3x
* math/Makefile: Add s_sincosf_data.c.
* sysdeps/ia64/fpu/s_sincosf_data.c: New file.
* sysdeps/ieee754/flt-32/s_sincosf.h (abstop12): Add new function.
(sincosf_poly): Likewise.
(reduce_small): Likewise.
(reduce_large): Likewise.
* sysdeps/ieee754/flt-32/s_sincosf.c (sincosf): Rewrite.
* sysdeps/ieee754/flt-32/s_sincosf_data.c: New file with sincosf data.
* sysdeps/m68k/m680x0/fpu/s_sincosf_data.c: New file.
* sysdeps/x86_64/fpu/s_sincosf_data.c: New file.
Following the recent discussion of using Python instead of Perl and
Awk for glibc build / test, this patch replaces gen-libm-test.pl with
a new gen-libm-test.py script. This script should work with all
Python versions supported by glibc (tested by hand with Python 2.7,
tested in the build system with Python 3.5; configure prefers Python 3
if available).
This script is designed to give identical output to gen-libm-test.pl
for ease of verification of the change, except for generated comments
referring to .py instead of .pl. (That is, identical for actual
inputs passed to the script, not necessarily for all possible input;
for example, this version more precisely follows the C standard syntax
for floating-point constants when deciding when to add LIT macro
calls.) In one place a comment notes that the generation of
NON_FINITE flags is replicating a bug in the Perl script to assist in
such comparisons (with the expectation that this bug can then be
separately fixed in the Python script later).
Tested for x86_64, including comparison of generated files (and hand
testing of the case of generating a sorted libm-test-ulps file, which
isn't covered by normal "make check").
I'd expect to follow this up by extending the new script to produce
the ulps tables for the manual as well (replacing
manual/libm-err-tab.pl, so that then we just have one ulps file
parser) - at which point the manual build would depend on both Perl
and Python (eliminating the Perl dependency would require someone to
rewrite summary.pl in Python, and that would only eliminate the
*direct* Perl dependency; current makeinfo is written in Perl so there
would still be an indirect dependency).
I think install.texi is more or less equally out-of-date regarding
Perl and Python uses before and after this patch, so I don't think
this patch depends on my patch
<https://sourceware.org/ml/libc-alpha/2018-08/msg00133.html> to update
install.texi regarding such uses (pending review).
* math/gen-libm-test.py: New file.
* math/gen-libm-test.pl: Remove.
* math/Makefile [$(PERL) != no]: Change condition to [PYTHON].
($(objpfx)libm-test-ulps.h): Use gen-libm-test.py instead of
gen-libm-test.pl.
($(libm-test-c-noauto-obj)): Likewise.
($(libm-test-c-auto-obj)): Likewise.
($(libm-test-c-narrow-obj)): Likewise.
(regen-ulps): Likewise.
* math/README.libm-test: Update references to gen-libm-test.pl.
* math/libm-test-driver.c (struct test_fj_f_data): Update comment
referencing gen-libm-test.pl.
* math/libm-test-nexttoward.inc (nexttoward_test_data): Likewise.
* math/libm-test-support.c: Likewise.
* math/libm-test-support.h: Likewise.
* sysdeps/generic/libm-test-ulps: Likewise.
When libm tests were split into separate per-function .inc files, a
comment relating to the nexttoward tests ended up at the end of
libm-test-nextdown.inc (because the split was based on starting each
function's tests with the <function>_test_data definition, which
failed to allow for comments before such definitions). This patch
moves that comment to the correct location.
Tested for x86_64.
* math/libm-test-nextdown.inc (do_test): Move comment to ....
* math/libm-test-nexttoward.inc (nexttoward_test_data): ... here.
Some TEST_* lines in libm-test-*.inc end with semicolons not commas.
This works at present because gen-libm-test.pl ignores whatever comes
after the TEST_* call, but is logically wrong (since the TEST_* calls
generate array elements, not statements) and the Python replacement
for gen-libm-test.pl that I'm working on is stricter about the syntax
here. This patch fixes the lines in question to use commas like most
such lines already do.
Tested for x86_64.
* math/libm-test-ilogb.inc (ilogb_test_data): Use ',' not ';'
after TEST_* calls.
* math/libm-test-llogb.inc (llogb_test_data): Likewise.
* math/libm-test-logb.inc (logb_test_data): Likewise.
math/test-misc.c contains some code that uses fenv.h macros
FE_UNDERFLOW, FE_OVERFLOW and FE_UPWARD without being conditional on
those macros being defined.
That would normally break the build for configurations (typically
soft-float) not defining those macros. However, the code in question
is inside LDBL_MANT_DIG > DBL_MANT_DIG conditionals. And, while we
have configurations lacking rounding mode and exception support where
LDBL_MANT_DIG > DBL_MANT_DIG (soft-float MIPS64 and RISC-V), those
configurations currently define the fenv.h macros in question even for
soft-float.
There may be some case for defining those macros in cases where a
soft-float compilation could use a hard-float libm (where both
soft-float and hard-float can use the same ABI, as on ARM and RISC-V,
for example). But MIPS is not such a case - the hard-float and
soft-float ABIs are incompatible - and thus I am testing a patch to
stop defining those macros for soft-float MIPS (motivated by reducing
the extent to which architectures need their own definitions of
math-tests.h macros - if lack of rounding mode / exception support can
be determined by the lack of macros in fenv.h, that avoids the need
for math-tests.h to declare that lack as well). Introducing a case of
LDBL_MANT_DIG > DBL_MANT_DIG without these macros defined shows up the
problem with math/test-misc.c. This patch then fixes that problem by
adding appropriate conditionals.
Tested for MIPS64 in conjunction with changes to stop defining the
macros in question in bits/fenv.h for soft-float.
* math/test-misc.c (do_test) [LDBL_MANT_DIG > DBL_MANT_DIG]: Make
code using FE_UNDERFLOW conditional on [FE_UNDERFLOW], code using
FE_OVERFLOW conditional on [FE_OVERFLOW] and code using FE_UPWARD
conditional on [FE_UPWARD].
Move declare_mgen_finite_alias, declare_mgen_finite_alias_s and
declare_mgen_finite_alias_x to a shared place in order to reuse them in
other files that also declare _finite aliases.
* math/e_exp2_template.c (declare_mgen_finite_alias,
declare_mgen_finite_alias_s, declare_mgen_finite_alias_x): Move to...
* sysdeps/generic/math-type-macros.h (declare_mgen_finite_alias,
declare_mgen_finite_alias_s, declare_mgen_finite_alias_x): ... here.
Signed-off-by: Tulio Magno Quites Machado Filho <tuliom@linux.ibm.com>
As in https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86113 for
__builtin_nan, bits/mathcalls.h wrongly declares the nan function with
the __const__ attribute. Because the function reads memory pointed to
by an argument, it's only pure, not const. This patch removes the
incorrect attribute and adds a testcase for the bug. No __pure__
attribute is added to replace the incorrect __const__ one, since that
would introduce problems when using GCC versions that have the
incorrect built-in __const__ attribute and warn for the combination of
those two attributes.
Tested for x86_64.
[BZ #23277]
* math/bits/mathcalls.h [__USE_ISOC99] (nan): Do not use __const__
attribute.
* math/test-nan-const.c: New file.
* math/Makefile (tests): Add test-nan-const.
(CFLAGS-test-nan-const.c): New variable.
As reported in bug 23272, the ldbl-96 implementation of fma (fma for
double, in terms of ldbl-96 as the internal arithmetic type, as used
on 32-bit x86) is missing some of the special-case handling for
non-finite arguments, resulting in incorrect NaN results when the
first two arguments are infinities, the third is finite and so the
infinities go through the logic for finite arguments. This patch
fixes it by handling all cases of non-finite arguments up front, with
additional fma tests for the problem cases being added to the
testsuite.
Tested for x86_64 and x86.
[BZ #23272]
* sysdeps/ieee754/ldbl-96/s_fma.c (__fma): Start by handling all
cases of non-finite arguments.
* math/libm-test-fma.inc (fma_test_data): Add more tests.
The commit
commit c85e54ac6c
Author: Gabriel F. T. Gomes <gabriel@inconstante.eti.br>
Date: Fri Nov 3 10:44:36 2017 -0200
Provide a C++ version of iseqsig (bug 22377)
mistakenly used double parameters in the long double version of iseqsig,
thus causing spurious conversions to double, as reported on bug 23171.
Tested for powerpc64le and x86_64.
It has been noted that test-tgmath3 is slow to compile, and to link on
some systems
<https://sourceware.org/ml/libc-alpha/2018-02/msg00477.html>, because
of the size of the test.
I'm working on tgmath.h support for the TS 18661-1 / 18661-3 functions
that round their results to a narrower type. For the functions
already present in glibc, this wouldn't make test-tgmath3 much bigger,
because those functions only have two arguments. For the narrowing
versions of fma (for which I've not yet added the functions to glibc),
however, it would result in many configurations building tests of the
type-generic macros f32fma, f64fma, f32xfma, f64xfma, each with 21
possible types for each of three arguments (float, double, long double
aren't valid argument types for these macros when they return a
_FloatN / _FloatNx type), so substantially increasing the size of the
testcase.
To avoid further increasing the size of a single test when adding the
type-generic narrowing fma macros, this patch arranges for the
test-tgmath3 tests to be run separately for each function tested. The
fma tests are still by far the largest (next is pow, as that has two
arguments that can be real or complex; after that, the two-argument
real-only functions), but each type-generic fma macro for a different
return type would end up with its tests being run separately, rather
than increasing the size of a single test.
To avoid accidentally missing testing a macro because
gen-tgmath-tests.py supports testing it but the makefile fails to call
it for that function, a test is also added that verifies that the
lists of macros in the makefile and gen-tgmath-tests.py agree.
Tested for x86_64.
* math/gen-tgmath-tests.py: Import sys.
(Tests.__init__): Initialize macros_seen.
(Tests.add_tests): Add macro to macros_seen. Only generate tests
if requested to do so for this macro.
(Tests.add_all_tests): Take argument for macro for which to
generate tests.
(Tests.check_macro_list): New function.
(main): Handle check-list argument and argument specifying macro
for which to generate tests.
* math/Makefile [PYTHON] (tgmath3-macros): New variable.
[PYTHON] (tgmath3-macro-tests): Likewise.
[PYTHON] (tests): Add $(tgmath3-macro-tests) not test-tgmath3.
[PYTHON] (generated): Add $(addsuffix .c,$(tgmath3-macro-tests))
not test-tgmath3.c.
[PYTHON] (CFLAGS-test-tgmath3.c): Remove.
[PYTHON] ($(tgmath3-macro-tests:%=$(objpfx)%.o): Add -fno-builtin
to CFLAGS.
[PYTHON] ($(objpfx)test-tgmath3.c): Replace rule by....
[PYTHON] ($(foreach
m,$(tgmath3-macros),$(objpfx)test-tgmath3-$(m).c): ... this. New
rule.
[PYTHON] (tests-special): Add
$(objpfx)test-tgmath3-macro-list.out.
[PYTHON] ($(objpfx)test-tgmath3-macro-list.out): New rule.
When compiling C++ code with -mabi=ieeelongdouble, KCtype is
unavailable and the long double type should be used instead.
This is also providing macro __HAVE_FLOAT128_UNLIKE_LDBL in order to
identify the kind of long double type is being used in the current
compilation unit.
Notice that bits/floatn.h cannot benefit from the new macro due to order
of header inclusion.
* bits/floatn-common.h: Define __HAVE_FLOAT128_UNLIKE_LDBL.
* math/math.h: Restrict the prototype definition for the functions
issignaling(_Float128) and iszero(_Float128); and template
__iseqsig_type<_Float128>, from __HAVE_DISTINCT_FLOAT128 to
__HAVE_FLOAT128_UNLIKE_LDBL.
* sysdeps/powerpc/bits/floatn.h [__HAVE_FLOAT128
&& (!__GNUC_PREREQ (7, 0) || defined __cplusplus)
&& __LDBL_MANT_DIG__ == 113]: Use long double suffix for
__f128() constants; define the type _Float128 as long double;
and reuse long double in __CFLOAT128.
Signed-off-by: Tulio Magno Quites Machado Filho <tuliom@linux.ibm.com>
This patch continues the math_private.h cleanup by stopping
math_private.h from including math-barriers.h and making the users of
the barrier macros include the latter header directly. No attempt is
made to remove any math_private.h includes that are now unused, except
in strtod_l.c where that is done to avoid line number changes in
assertions, so that installed stripped shared libraries can be
compared before and after the patch. (I think the floating-point
environment support in math_private.h should also move out - some
architectures already have fenv_private.h as an architecture-internal
header included from their math_private.h - and after moving that out
might be a better time to identify unused math_private.h includes.)
Tested for x86_64 and x86, and tested with build-many-glibcs.py that
installed stripped shared libraries are unchanged by the patch.
* sysdeps/generic/math_private.h: Do not include
<math-barriers.h>.
* stdlib/strtod_l.c: Include <math-barriers.h> instead of
<math_private.h>.
* math/fromfp.h: Include <math-barriers.h>.
* math/math-narrow.h: Likewise.
* math/s_nextafter.c: Likewise.
* math/s_nexttowardf.c: Likewise.
* sysdeps/aarch64/fpu/s_llrint.c: Likewise.
* sysdeps/aarch64/fpu/s_llrintf.c: Likewise.
* sysdeps/aarch64/fpu/s_lrint.c: Likewise.
* sysdeps/aarch64/fpu/s_lrintf.c: Likewise.
* sysdeps/i386/fpu/s_nextafterl.c: Likewise.
* sysdeps/i386/fpu/s_nexttoward.c: Likewise.
* sysdeps/i386/fpu/s_nexttowardf.c: Likewise.
* sysdeps/ieee754/dbl-64/e_atan2.c: Likewise.
* sysdeps/ieee754/dbl-64/e_atanh.c: Likewise.
* sysdeps/ieee754/dbl-64/e_exp.c: Likewise.
* sysdeps/ieee754/dbl-64/e_exp2.c: Likewise.
* sysdeps/ieee754/dbl-64/e_j0.c: Likewise.
* sysdeps/ieee754/dbl-64/e_sqrt.c: Likewise.
* sysdeps/ieee754/dbl-64/s_expm1.c: Likewise.
* sysdeps/ieee754/dbl-64/s_fma.c: Likewise.
* sysdeps/ieee754/dbl-64/s_fmaf.c: Likewise.
* sysdeps/ieee754/dbl-64/s_log1p.c: Likewise.
* sysdeps/ieee754/dbl-64/s_nearbyint.c: Likewise.
* sysdeps/ieee754/dbl-64/wordsize-64/s_nearbyint.c: Likewise.
* sysdeps/ieee754/flt-32/e_atanhf.c: Likewise.
* sysdeps/ieee754/flt-32/e_j0f.c: Likewise.
* sysdeps/ieee754/flt-32/s_expm1f.c: Likewise.
* sysdeps/ieee754/flt-32/s_log1pf.c: Likewise.
* sysdeps/ieee754/flt-32/s_nearbyintf.c: Likewise.
* sysdeps/ieee754/flt-32/s_nextafterf.c: Likewise.
* sysdeps/ieee754/k_standardl.c: Likewise.
* sysdeps/ieee754/ldbl-128/e_asinl.c: Likewise.
* sysdeps/ieee754/ldbl-128/e_expl.c: Likewise.
* sysdeps/ieee754/ldbl-128/e_powl.c: Likewise.
* sysdeps/ieee754/ldbl-128/s_fmal.c: Likewise.
* sysdeps/ieee754/ldbl-128/s_nearbyintl.c: Likewise.
* sysdeps/ieee754/ldbl-128/s_nextafterl.c: Likewise.
* sysdeps/ieee754/ldbl-128/s_nexttoward.c: Likewise.
* sysdeps/ieee754/ldbl-128/s_nexttowardf.c: Likewise.
* sysdeps/ieee754/ldbl-128ibm/e_asinl.c: Likewise.
* sysdeps/ieee754/ldbl-128ibm/s_fmal.c: Likewise.
* sysdeps/ieee754/ldbl-128ibm/s_nextafterl.c: Likewise.
* sysdeps/ieee754/ldbl-128ibm/s_nexttoward.c: Likewise.
* sysdeps/ieee754/ldbl-128ibm/s_nexttowardf.c: Likewise.
* sysdeps/ieee754/ldbl-128ibm/s_rintl.c: Likewise.
* sysdeps/ieee754/ldbl-96/e_atanhl.c: Likewise.
* sysdeps/ieee754/ldbl-96/e_j0l.c: Likewise.
* sysdeps/ieee754/ldbl-96/s_fma.c: Likewise.
* sysdeps/ieee754/ldbl-96/s_fmal.c: Likewise.
* sysdeps/ieee754/ldbl-96/s_nexttoward.c: Likewise.
* sysdeps/ieee754/ldbl-96/s_nexttowardf.c: Likewise.
* sysdeps/ieee754/ldbl-opt/s_nexttowardfd.c: Likewise.
* sysdeps/m68k/m680x0/fpu/s_nextafterl.c: Likewise.
This patch continues cleaning up the math_private.h header, which
contains lots of different definitions many of which are only needed
by a limited subset of files using that header (and some of which are
overridden by architectures that only want to override selected parts
of the header), by moving the math_narrow_eval macro out to a separate
math-narrow-eval.h header, only included by those files that need it.
That header is placed in include/ (since it's used in stdlib/, not
just files built in math/, but no sysdeps variants are needed at
present).
Tested for x86_64, and with build-many-glibcs.py. (Installed stripped
shared libraries change because of line numbers in assertions in
strtod_l.c.)
* include/math-narrow-eval.h: New file. Contents moved from ....
* sysdeps/generic/math_private.h: ... here.
(math_narrow_eval): Remove macro. Moved to math-narrow-eval.h.
[FLT_EVAL_METHOD != 0] (excess_precision): Likewise.
* math/s_fdim_template.c: Include <math-narrow-eval.h>.
* stdlib/strtod_l.c: Likewise.
* sysdeps/i386/fpu/s_f32xaddf64.c: Likewise.
* sysdeps/i386/fpu/s_f32xsubf64.c: Likewise.
* sysdeps/i386/fpu/s_fdim.c: Likewise.
* sysdeps/ieee754/dbl-64/e_cosh.c: Likewise.
* sysdeps/ieee754/dbl-64/e_gamma_r.c: Likewise.
* sysdeps/ieee754/dbl-64/e_j1.c: Likewise.
* sysdeps/ieee754/dbl-64/e_jn.c: Likewise.
* sysdeps/ieee754/dbl-64/e_lgamma_r.c: Likewise.
* sysdeps/ieee754/dbl-64/e_sinh.c: Likewise.
* sysdeps/ieee754/dbl-64/gamma_productf.c: Likewise.
* sysdeps/ieee754/dbl-64/k_rem_pio2.c: Likewise.
* sysdeps/ieee754/dbl-64/lgamma_neg.c: Likewise.
* sysdeps/ieee754/dbl-64/s_erf.c: Likewise.
* sysdeps/ieee754/dbl-64/s_llrint.c: Likewise.
* sysdeps/ieee754/dbl-64/s_lrint.c: Likewise.
* sysdeps/ieee754/flt-32/e_coshf.c: Likewise.
* sysdeps/ieee754/flt-32/e_exp2f.c: Likewise.
* sysdeps/ieee754/flt-32/e_expf.c: Likewise.
* sysdeps/ieee754/flt-32/e_gammaf_r.c: Likewise.
* sysdeps/ieee754/flt-32/e_j1f.c: Likewise.
* sysdeps/ieee754/flt-32/e_jnf.c: Likewise.
* sysdeps/ieee754/flt-32/e_lgammaf_r.c: Likewise.
* sysdeps/ieee754/flt-32/e_sinhf.c: Likewise.
* sysdeps/ieee754/flt-32/k_rem_pio2f.c: Likewise.
* sysdeps/ieee754/flt-32/lgamma_negf.c: Likewise.
* sysdeps/ieee754/flt-32/s_erff.c: Likewise.
* sysdeps/ieee754/flt-32/s_llrintf.c: Likewise.
* sysdeps/ieee754/flt-32/s_lrintf.c: Likewise.
* sysdeps/ieee754/ldbl-96/gamma_product.c: Likewise.
libm_hidden_def expand the parameters and do not require an extra layer
of macros.
These were the last 3 files in math/ still using hidden_def().
* math/w_exp_compat.c: Replace hidden_def with libm_hidden_def..
* math/w_expl_compat.c: Likewise.
* math/w_exp_template.c: Likewise. Remove hidden_def_x.
Signed-off-by: Tulio Magno Quites Machado Filho <tuliom@linux.ibm.com>
According to math-type-macros.h, M_SUF should be used to paste the
suffix used by functions, while M_MLIT is used with macro constants.
* math/e_exp2_template.c: Replace M_SUF (M_LN2) with M_MLIT (M_LN2).
Signed-off-by: Tulio Magno Quites Machado Filho <tuliom@linux.ibm.com>
math-type-macros.h provides the macro M_FABS in order to reference
the correct fabs function for a specific type.
In most of the cases, M_FABS is identical to M_SUF (fabs), but that
may change in the future.
* math/w_acos_template.c: Replace M_SUF (fabs) with M_FABS.
* math/w_asin_template.c: Likewise.
* math/w_atanh_template.c: Likewise.
Signed-off-by: Tulio Magno Quites Machado Filho <tuliom@linux.ibm.com>
If build with -Os on s390x, the test-tgmath fails with:
float functions not called often enough (-10000)
Within compile_testf(), the counter (count_float) is saved
before the complex functions are called.
Afterwards the saved counter differs to the current-counter.
But the tests with the complex functions do not increment count_float!
Instead count_float is saved to a register before calling totalorder
and totalordermag which both increment count_float.
The compiler is allowed to do that as totalorderf and totalordermagf
is declared with __attribute__ ((__const__)) in math/bits/mathcalls.h.
Thus this patch adjusts the global counters to be volatile.
Then count_float is saved after totalordermag.
ChangeLog:
* math/test-tgmath.c (count_double, count_float,
count_ldouble, count_cdouble, count_cfloat,
count_cldouble): Use volatile int.
This patch series cleans up the many uses of __ieee754_sqrt(f/l) in GLIBC.
The goal is to enable GCC to do the inlining, and if this fails call the
__ieee754_sqrt function. This is done by internally declaring sqrt with asm
redirects. The compat symbols and sqrt wrappers need to disable the redirect.
The redirect is also disabled if there are already redirects defined when
using -ffinite-math-only.
All math functions (but not math tests, non-library code and libnldbl) are
built with -fno-math-errno which means GCC will typically inline sqrt as a
single instruction. This means targets are no longer forced to add a special
inline for sqrt.
* include/math.h (sqrt): Declare with asm redirect.
(sqrtf): Likewise.
(sqrtl): Likewise.
(sqrtf128): Likewise.
* Makeconfig: Add -fno-math-errno for libc/libm, but build testsuite,
nonlib and libnldbl with -fmath-errno.
* math/w_sqrt_compat.c: Define NO_MATH_REDIRECT.
* math/w_sqrt_template.c: Likewise.
* math/w_sqrtf_compat.c: Likewise.
* math/w_sqrtl_compat.c: Likewise.
* sysdeps/i386/fpu/w_sqrt.c: Likewise.
* sysdeps/i386/fpu/w_sqrt_compat.c: Likewise.
* sysdeps/generic/math-type-macros-float128.h: Remove math.h and
complex.h.
Remove the slow paths from pow. Like several other double precision math
functions, pow is exactly rounded. This is not required from math functions
and causes major overheads as it requires multiple fallbacks using higher
precision arithmetic if a result is close to 0.5ULP. Ridiculous slowdowns
of up to 100000x have been reported when the highest precision path triggers.
All GLIBC math tests pass on AArch64 and x64 (with ULP of pow set to 1).
The worst case error is ~0.506ULP. A simple test over a few hundred million
values shows pow is 10% faster on average. This fixes BZ #13932.
[BZ #13932]
* sysdeps/ieee754/dbl-64/uexp.h (err_1): Remove.
* benchtests/pow-inputs: Update comment for slow path cases.
* manual/probes.texi (slowpow_p10): Delete removed probe.
(slowpow_p10): Likewise.
* math/Makefile: Remove halfulp.c and slowpow.c.
* sysdeps/aarch64/libm-test-ulps: Set ULP of pow to 1.
* sysdeps/generic/math_private.h (__exp1): Remove error argument.
(__halfulp): Remove.
(__slowpow): Remove.
* sysdeps/i386/fpu/halfulp.c: Delete file.
* sysdeps/i386/fpu/slowpow.c: Likewise.
* sysdeps/ia64/fpu/halfulp.c: Likewise.
* sysdeps/ia64/fpu/slowpow.c: Likewise.
* sysdeps/ieee754/dbl-64/e_exp.c (__exp1): Remove error argument,
improve comments and add error analysis.
* sysdeps/ieee754/dbl-64/e_pow.c (__ieee754_pow): Add error analysis.
(power1): Remove function:
(log1): Remove error argument, add error analysis.
(my_log2): Remove function.
* sysdeps/ieee754/dbl-64/halfulp.c: Delete file.
* sysdeps/ieee754/dbl-64/slowpow.c: Likewise.
* sysdeps/m68k/m680x0/fpu/halfulp.c: Likewise.
* sysdeps/m68k/m680x0/fpu/slowpow.c: Likewise.
* sysdeps/powerpc/power4/fpu/Makefile: Remove CPPFLAGS-slowpow.c.
* sysdeps/x86_64/fpu/libm-test-ulps: Set ULP of pow to 1.
* sysdeps/x86_64/fpu/multiarch/Makefile: Remove slowpow-fma.c,
slowpow-fma4.c, halfulp-fma.c, halfulp-fma4.c.
* sysdeps/x86_64/fpu/multiarch/e_pow-fma.c (__slowpow): Remove define.
* sysdeps/x86_64/fpu/multiarch/e_pow-fma4.c (__slowpow): Likewise.
* sysdeps/x86_64/fpu/multiarch/halfulp-fma.c: Delete file.
* sysdeps/x86_64/fpu/multiarch/halfulp-fma4.c: Likewise.
* sysdeps/x86_64/fpu/multiarch/slowpow-fma.c: Likewise.
* sysdeps/x86_64/fpu/multiarch/slowpow-fma4.c: Likewise.
This patch adds the narrowing add functions from TS 18661-1 to glibc's
libm: fadd, faddl, daddl, f32addf64, f32addf32x, f32xaddf64 for all
configurations; f32addf64x, f32addf128, f64addf64x, f64addf128,
f32xaddf64x, f32xaddf128, f64xaddf128 for configurations with
_Float64x and _Float128; __nldbl_daddl for ldbl-opt. As discussed for
the build infrastructure patch, tgmath.h support is deliberately
deferred, and FP_FAST_* macros are not applicable without optimized
function implementations.
Function implementations are added for all relevant pairs of formats
(including certain cases of a format and itself where more than one
type has that format). The main implementations use round-to-odd, or
a trivial computation in the case where both formats are the same or
where the wider format is IBM long double (in which case we don't
attempt to be correctly rounding). The sysdeps/ieee754/soft-fp
implementations use soft-fp, and are used automatically for
configurations without exceptions and rounding modes by virtue of
existing Implies files. As previously discussed, optimized versions
for particular architectures are possible, but not included.
i386 gets a special version of f32xaddf64 to avoid problems with
double rounding (similar to the existing fdim version), since this
function must round just once without an intermediate rounding to long
double. (No such special version is needed for any other function,
because the nontrivial functions use round-to-odd, which does the
intermediate computation with the rounding mode set to round-to-zero,
and double rounding is OK except in round-to-nearest mode, so is OK
for that intermediate round-to-zero computation.) mul and div will
need slightly different special versions for i386 (using round-to-odd
on long double instead of precision control) because of the
possibility of inexact intermediate results in the subnormal range for
double.
To reduce duplication among the different function implementations,
math-narrow.h gets macros CHECK_NARROW_ADD, NARROW_ADD_ROUND_TO_ODD
and NARROW_ADD_TRIVIAL.
In the trivial cases and for any architecture-specific optimized
implementations, the overhead of the errno setting might be
significant, but I think that's best handled through compiler built-in
functions rather than providing separate no-errno versions in glibc
(and likewise there are no __*_finite entry points for these function
provided, __*_finite effectively being no-errno versions at present in
most cases).
Tested for x86_64 and x86, with both GCC 6 and GCC 7. Tested for
mips64 (all three ABIs, both hard and soft float) and powerpc with GCC
7. Tested with build-many-glibcs.py with both GCC 6 and GCC 7.
* math/Makefile (libm-narrow-fns): Add add.
(libm-test-funcs-narrow): Likewise.
* math/Versions (GLIBC_2.28): Add narrowing add functions.
* math/bits/mathcalls-narrow.h (add): Use __MATHCALL_NARROW .
* math/gen-auto-libm-tests.c (test_functions): Add add.
* math/math-narrow.h (CHECK_NARROW_ADD): New macro.
(NARROW_ADD_ROUND_TO_ODD): Likewise.
(NARROW_ADD_TRIVIAL): Likewise.
* sysdeps/ieee754/float128/float128_private.h (__faddl): New
macro.
(__daddl): Likewise.
* sysdeps/ieee754/ldbl-opt/Makefile (libnldbl-calls): Add fadd and
dadd.
(CFLAGS-nldbl-dadd.c): New variable.
(CFLAGS-nldbl-fadd.c): Likewise.
* sysdeps/ieee754/ldbl-opt/Versions (GLIBC_2.28): Add
__nldbl_daddl.
* sysdeps/ieee754/ldbl-opt/nldbl-compat.h (__nldbl_daddl): New
prototype.
* manual/arith.texi (Misc FP Arithmetic): Document fadd, faddl,
daddl, fMaddfN, fMaddfNx, fMxaddfN and fMxaddfNx.
* math/auto-libm-test-in: Add tests of add.
* math/auto-libm-test-out-narrow-add: New generated file.
* math/libm-test-narrow-add.inc: New file.
* sysdeps/i386/fpu/s_f32xaddf64.c: Likewise.
* sysdeps/ieee754/dbl-64/s_f32xaddf64.c: Likewise.
* sysdeps/ieee754/dbl-64/s_fadd.c: Likewise.
* sysdeps/ieee754/float128/s_f32addf128.c: Likewise.
* sysdeps/ieee754/float128/s_f64addf128.c: Likewise.
* sysdeps/ieee754/float128/s_f64xaddf128.c: Likewise.
* sysdeps/ieee754/ldbl-128/s_daddl.c: Likewise.
* sysdeps/ieee754/ldbl-128/s_f64xaddf128.c: Likewise.
* sysdeps/ieee754/ldbl-128/s_faddl.c: Likewise.
* sysdeps/ieee754/ldbl-128ibm/s_daddl.c: Likewise.
* sysdeps/ieee754/ldbl-128ibm/s_faddl.c: Likewise.
* sysdeps/ieee754/ldbl-96/s_daddl.c: Likewise.
* sysdeps/ieee754/ldbl-96/s_faddl.c: Likewise.
* sysdeps/ieee754/ldbl-opt/nldbl-dadd.c: Likewise.
* sysdeps/ieee754/ldbl-opt/nldbl-fadd.c: Likewise.
* sysdeps/ieee754/soft-fp/s_daddl.c: Likewise.
* sysdeps/ieee754/soft-fp/s_fadd.c: Likewise.
* sysdeps/ieee754/soft-fp/s_faddl.c: Likewise.
* sysdeps/powerpc/fpu/libm-test-ulps: Update.
* sysdeps/mach/hurd/i386/libm.abilist: Likewise.
* sysdeps/unix/sysv/linux/aarch64/libm.abilist: Likewise.
* sysdeps/unix/sysv/linux/alpha/libm.abilist: Likewise.
* sysdeps/unix/sysv/linux/arm/libm.abilist: Likewise.
* sysdeps/unix/sysv/linux/hppa/libm.abilist: Likewise.
* sysdeps/unix/sysv/linux/i386/libm.abilist: Likewise.
* sysdeps/unix/sysv/linux/ia64/libm.abilist: Likewise.
* sysdeps/unix/sysv/linux/m68k/coldfire/libm.abilist: Likewise.
* sysdeps/unix/sysv/linux/m68k/m680x0/libm.abilist: Likewise.
* sysdeps/unix/sysv/linux/microblaze/libm.abilist: Likewise.
* sysdeps/unix/sysv/linux/mips/mips32/libm.abilist: Likewise.
* sysdeps/unix/sysv/linux/mips/mips64/libm.abilist: Likewise.
* sysdeps/unix/sysv/linux/nios2/libm.abilist: Likewise.
* sysdeps/unix/sysv/linux/powerpc/powerpc32/fpu/libm.abilist: Likewise.
* sysdeps/unix/sysv/linux/powerpc/powerpc32/nofpu/libm.abilist: Likewise.
* sysdeps/unix/sysv/linux/powerpc/powerpc64/libm-le.abilist: Likewise.
* sysdeps/unix/sysv/linux/powerpc/powerpc64/libm.abilist: Likewise.
* sysdeps/unix/sysv/linux/riscv/rv64/libm.abilist: Likewise.
* sysdeps/unix/sysv/linux/s390/s390-32/libm.abilist: Likewise.
* sysdeps/unix/sysv/linux/s390/s390-64/libm.abilist: Likewise.
* sysdeps/unix/sysv/linux/sh/libm.abilist: Likewise.
* sysdeps/unix/sysv/linux/sparc/sparc32/libm.abilist: Likewise.
* sysdeps/unix/sysv/linux/sparc/sparc64/libm.abilist: Likewise.
* sysdeps/unix/sysv/linux/tile/tilegx32/libm.abilist: Likewise.
* sysdeps/unix/sysv/linux/tile/tilegx64/libm.abilist: Likewise.
* sysdeps/unix/sysv/linux/x86_64/64/libm.abilist: Likewise.
* sysdeps/unix/sysv/linux/x86_64/x32/libm.abilist: Likewise.
Testing narrowing functions for x86_64 with GCC 6 showed up a further
testsuite fix needed: there is no _Float128 sNaN support before GCC 7
on x86_64 / x86, and the existing tests of SNAN_TESTS only checked it
for the return type, not for the argument type. This patch fixes the
code to check SNAN_TESTS (ARG_FLOAT) as well (in a variable set in
libm-test-driver.c, since libm-test-support.c is compiled only once
for each choice of FLOAT).
Tested for x86_64 and x86 with GCC 6 in conjunction with the main
patch adding narrowing add functions.
* math/libm-test-driver.c (snan_tests_arg): New variable.
* math/libm-test-support.h (snan_tests_arg): New declaration.
* math/libm-test-support.c (enable_test): Check snan_tests_arg.
This patch continues preparations for adding TS 18661-1 narrowing libm
functions by adding the required testsuite infrastructure to test such
functions through the libm-test infrastructure.
That infrastructure is based around testing for a single type, FLOAT.
For the narrowing functions, FLOAT, the "main" type for testing, is
the function return type; the argument type is ARG_FLOAT. This is
consistent with how the code built once for each type,
libm-test-support.c, depends on FLOAT for such things as calculating
ulps errors in results but can already handle different argument types
(pointers, integers, long double for nexttoward).
Makefile machinery is added to handle building tests for all pairs of
types for which there are narrowing functions (as with non-narrowing
functions, aliases are tested just the same as the functions they
alias). gen-auto-libm-tests gains a --narrow option for building
outputs for narrowing functions (so narrowing sqrt and fma will share
the same inputs as non-narrowing, but gen-auto-libm-tests will be run
with and without that option to generate different output files). In
the narrowing case, the auto-libm-test-out-narrow-* files include
annotations for each test about what properties ARG_FLOAT must have to
be able to represent all the inputs for that test; those annotations
result in calls to the TEST_COND_arg_fmt macro.
gen-libm-test.pl has some minor updates to handle narrowing tests (for
example, arguments in such tests must be surrounded by ARG_LIT calls
instead of LIT calls). Various new macros are added to the C test
support code (for example, sNaN initializers need to be properly
typed, so arg_snan_value is added; other such arg_* macros are added
as it seems cleanest to do so, though some are not strictly required).
Special-casing of the ibm128 format to allow for its limitations is
adjusted to handle it as the argument format as well as as the result
format; thus, the tests of the new functions allow nonzero ulps only
in the case where ibm128 is the argument format, as otherwise the
functions correspond to fully-defined IEEE operations. The ulps in
question appear as e.g. 'Function: "add_ldouble"' in libm-test-ulps
(with 1ulp errors then listed for double and float for that function
in powerpc); no support is added to generate corresponding faddl /
daddl ulps listings in the ulps table in the manual.
For the previous patch, I noted the need to avoid spurious macro
expansions of identifiers such as "add". A test test-narrow-macros.c
is added to verify such macro expansions are successfully avoided, and
there is also a -mlong-double-64 version of that test for ldbl-opt.
This test is set up to cover the full set of relevant identifiers from
the start rather than adding functions one at a time as each function
group is added.
Tested for x86_64 (this patch in isolation, as well as testing for
various configurations in conjunction with the actual addition of
"add" functions).
* math/Makefile (test-type-pairs): New variable.
(test-type-pairs-f64xf128-yes): Likewise.
(tests): Add test-narrow-macros.
(libm-test-funcs-narrow): New variable.
(libm-test-c-narrow): Likewise.
(generated): Add $(libm-test-c-narrow).
(libm-tests-base-narrow): New variable.
(libm-tests-narrow): Likewise.
(libm-tests): Add $(libm-tests-narrow).
(libm-tests-for-type): Handle $(libm-tests-narrow).
(libm-test-c-narrow-obj): New variable.
($(libm-test-c-narrow-obj)): New rule.
($(foreach t,$(libm-tests-narrow),$(objpfx)$(t).c)): Likewise.
($(foreach f,$(libm-test-funcs-narrow),$(objpfx)$(o)-$(f).o)): Use
$(o-iterator) to set dependencies and CFLAGS.
* math/gen-auto-libm-tests.c: Document use for narrowing
functions.
(output_for_one_input_case): Take argument NARROW.
(generate_output): Likewise. Update call to
output_for_one_input_case.
(main): Take --narrow option. Update call to generate_output.
* math/gen-libm-test.pl (_apply_lit): Take macro name as argument.
(apply_lit): Update call to _apply_lit.
(apply_arglit): New function.
(parse_args): Handle "a" arguments.
(parse_auto_input): Handle format names using ":".
* math/README.libm-test: Document "a" parameter type.
* math/libm-test-support.h (ARG_TYPE_MIN): New macro.
(ARG_TYPE_TRUE_MIN): Likewise.
(ARG_TYPE_MAX): Likwise.
(ARG_MIN_EXP): Likewise.
(ARG_MAX_EXP): Likewise.
(ARG_MANT_DIG): Likewise.
(TEST_COND_arg_ibm128): Likewise.
(TEST_COND_ibm128_libgcc): Define conditional on [ARG_FLOAT].
(TEST_COND_arg_fmt): New macro.
(init_max_error): Update prototype.
* math/libm-test-support.c (test_ibm128): New variable.
(init_max_error): Take argument testing_ibm128 and set test_ibm128
instead of using [TEST_COND_ibm128] conditional.
(test_exceptions): Use test_ibm128 instead of TEST_COND_ibm128.
* math/libm-test-driver.c (STR_ARG_FLOAT): New macro.
[TEST_NARROW] (TEST_MSG): New definition.
(arg_plus_zero): New macro.
(arg_minus_zero): Likewise.
(arg_plus_infty): Likewise.
(arg_minus_infty): Likewise.
(arg_qnan_value_pl): Likewise.
(arg_qnan_value): Likewise.
(arg_snan_value_pl): Likewise.
(arg_snan_value): Likewise.
(arg_max_value): Likewise.
(arg_min_value): Likewise.
(arg_min_subnorm_value): Likewise.
[ARG_FLOAT] (struct test_aa_f_data): New struct type.
(RUN_TEST_LOOP_aa_f): New macro.
(TEST_SUFF): New macro.
(TEST_SUFF_STR): Likewise.
[!TEST_MATHVEC] (VEC_SUFF): Don't define.
(TEST_COND_any_ibm128): New macro.
(START): Use TEST_SUFF and TEST_SUFF_STR in initializer for
this_func. Update call to init_max_error.
* math/test-double.h (FUNC_NARROW_PREFIX): New macro.
* math/test-float.h (FUNC_NARROW_PREFIX): Likewise.
* math/test-float128.h (FUNC_NARROW_PREFIX): Likewise.
* math/test-float32.h (FUNC_NARROW_PREFIX): Likewise.
* math/test-float32x.h (FUNC_NARROW_PREFIX): Likewise.
* math/test-float64.h (FUNC_NARROW_PREFIX): Likewise.
* math/test-float64x.h (FUNC_NARROW_PREFIX): Likewise.
* math/test-math-scalar.h (TEST_NARROW): Likewise.
* math/test-math-vector.h (TEST_NARROW): Likewise.
* math/test-arg-double.h: New file.
* math/test-arg-float128.h: Likewise.
* math/test-arg-float32x.h: Likewise.
* math/test-arg-float64.h: Likewise.
* math/test-arg-float64x.h: Likewise.
* math/test-arg-ldouble.h: Likewise.
* math/test-math-narrow.h: Likewise.
* math/test-narrow-macros.c: Likewise.
* sysdeps/ieee754/ldbl-opt/test-narrow-macros-ldbl-64.c: Likewise.
* sysdeps/ieee754/ldbl-opt/Makefile (tests): Add
test-narrow-macros-ldbl-64.
(CFLAGS-test-narrow-macros-ldbl-64.c): New variable.
TS 18661-1 defines libm functions that carry out an operation (+ - * /
sqrt fma) on their arguments and return a result rounded to a
(usually) narrower type, as if the original result were computed to
infinite precision and then rounded directly to the result type
without any intermediate rounding to the argument type. For example,
fadd, faddl and daddl for addition. These are the last remaining TS
18661-1 functions left to be added to glibc. TS 18661-3 extends this
to corresponding functions for _FloatN and _FloatNx types.
As functions parametrized by two rather than one varying
floating-point types, these functions require infrastructure in glibc
that was not required for previous libm functions. This patch
provides such infrastructure - excluding test support, and actual
function implementations, which will be in subsequent patches.
Declaring the functions uses a header bits/mathcalls-narrow.h, which
is included many times, for each relevant pair of types. This will
end up containing macro calls of the form
__MATHCALL_NARROW (__MATHCALL_NAME (add), __MATHCALL_REDIR_NAME (add), 2);
for each family of narrowing functions. (The structure of this macro
call, with the calls to __MATHCALL_NAME and __MATHCALL_REDIR_NAME
there rather than in the definition of __MATHCALL_NARROW, arises from
the names such as "add" *not* themselves being reserved identifiers -
meaning it's necessary to avoid any indirection that would result in a
user-defined "add" macro being expanded.) Whereas for existing
functions declaring long double functions is disabled if _LIBC in the
case where they alias double functions, to facilitate defining the
long double functions as aliases of the double ones, there is no such
logic for the narrowing functions in this patch. Rather, the files
defining such functions are expected to use #define to hide the
original declarations of the alias names, to avoid errors about
defining aliases with incompatible types.
math/Makefile support is added for building the functions (listed in
libm-narrow-fns, currently empty) for all relevant pairs of types. An
internal header math-narrow.h is added for macros shared between
multiple function implementations - currently a ROUND_TO_ODD macro to
facilitate writing functions using the round-to-odd implementation
approach, and alias macros to create all the required function
aliases. libc_feholdexcept_setroundf128 and libc_feupdateenv_testf128
are added for use when required (only for x86_64). float128_private.h
support is added for ldbl-128 narrowing functions to be used for
_Float128.
Certain things are specifically omitted from this patch and the
immediate followups. tgmath.h support is deferred; there remain
unresolved questions about how the type-generic macros for these
functions are supposed to work, especially in the case of arguments of
integer type. The math.h / bits/mathcalls-narrow.h logic, and the
logic for determining what functions / aliases to define, will need
some adjustments to support the sqrt and fma functions, where
e.g. f32xsqrtf64 can just be an alias for sqrt rather than a separate
function. TS 18661-1 defines FP_FAST_* macros but no support is
included for defining them (they won't in general be true without
architecture-specific optimized function versions).
For each of the function groups (add sub mul div sqrt fma) there are
always six functions present (e.g. fadd, faddl, daddl, f32addf64,
f32addf32x, f32xaddf64). When _Float64x and _Float128 are supported,
there are seven more (e.g. f32addf64x, f32addf128, f64addf64x,
f64addf128, f32xaddf64x, f32xaddf128, f64xaddf128). In addition, in
the ldbl-opt case there are function names such as __nldbl_daddl (an
alias for f32xaddf64, which is not a reserved name in TS 18661-1, only
in TS 18661-3), for calls to daddl to be mapped to in the
-mlong-double-64 case. (Calls to faddl just get mapped to fadd, and
for sqrt and fma there won't be __nldbl_* functions because dsqrtl and
dfmal can just be mapped to sqrt and fma with -mlong-double-64.)
While there are six or thirteen functions present in each group (plus
__nldbl_* names only as an ABI, not an API), not all are distinct;
they fall in various groups of aliases. There are two distinct
versions built if long double has the same format as double; four if
they have distinct formats but there is no _Float64x or _Float128
support; five if long double has binary128 format; seven when
_Float128 is distinct from long double.
Architecture-specific optimized versions are possible, but not
included in my patches. For example, IA64 generally supports
narrowing the result of most floating-point instructions; Power ISA
2.07 (POWER8) supports double values as arguments to float
instructions, with the results narrowed as expected; Power ISA 3
(POWER9) supports round-to-odd for float128 instructions, so meaning
that approach can be used without needing to set and restore the
rounding mode and test "inexact". I intend to leave any such
optimized versions to the architecture maintainers. Generally in such
cases it would also make sense for calls to these functions to be
expanded inline (given -fno-math-errno); I put a suggestion for TS
18661-1 built-in functions at <https://gcc.gnu.org/wiki/SummerOfCode>.
Tested for x86_64 (this patch in isolation, as well as testing for
various configurations in conjunction with further patches).
* math/bits/mathcalls-narrow.h: New file.
* include/bits/mathcalls-narrow.h: Likewise.
* math/math-narrow.h: Likewise.
* math/math.h (__MATHCALL_NARROW_ARGS_1): New macro.
(__MATHCALL_NARROW_ARGS_2): Likewise.
(__MATHCALL_NARROW_ARGS_3): Likewise.
(__MATHCALL_NARROW_NORMAL): Likewise.
(__MATHCALL_NARROW_REDIR): Likewise.
(__MATHCALL_NARROW): Likewise.
[__GLIBC_USE (IEC_60559_BFP_EXT)]: Repeatedly include
<bits/mathcalls-narrow.h> with _Mret_, _Marg_ and __MATHCALL_NAME
defined.
[__GLIBC_USE (IEC_60559_TYPES_EXT)]: Likewise.
* math/Makefile (headers): Add bits/mathcalls-narrow.h.
(libm-narrow-fns): New variable.
(libm-narrow-types-basic): Likewise.
(libm-narrow-types-ldouble-yes): Likewise.
(libm-narrow-types-float128-yes): Likewise.
(libm-narrow-types-float128-alias-yes): Likewise.
(libm-narrow-types): Likewise.
(libm-routines): Add narrowing functions.
* sysdeps/i386/fpu/fenv_private.h [__x86_64__]
(libc_feholdexcept_setroundf128): New macro.
[__x86_64__] (libc_feupdateenv_testf128): Likewise.
* sysdeps/ieee754/float128/float128_private.h: Include
<math/math-narrow.h>.
[libc_feholdexcept_setroundf128] (libc_feholdexcept_setroundl):
Undefine and redefine.
[libc_feupdateenv_testf128] (libc_feupdateenv_testl): Likewise.
(libm_alias_float_ldouble): Undefine and redefine.
(libm_alias_double_ldouble): Likewise.
The math/Makefile variable libm-test-incs was formerly used, but no
longer is. This patch removes it.
Tested for x86_64.
* math/Makefile [$(PERL) != no] (libm-test-incs): Remove variable.
Currently math_errhandling is always set to MATH_ERRNO | MATH_ERREXCEPT
even if -fno-math-errno is used. It is not defined at all when fast-math
is used. Set it to 0 with fast-math - this is noncomforming but more
useful than not define math_errhandling at all. Also take __NO_MATH_ERRNO__
into account and update comment.
* math/math.h (math_errhandling): Set to 0 with __FAST_MATH__.
Add __NO_MATH_ERRNO__ check.
I found that "make regen-ulps" failed when building with unmodified
GNU make 4.1, and an objdir /some/where/math/ longer than about 37
characters, because the list of tests in the "for run in $^" loop
exceeded the Linux kernel's MAX_ARG_STRLEN limit (131072 bytes) on the
length of a single argument passed to a command.
Some GNU/Linux distributions have a patch to make to work around this
limit (see e.g. Debian bug 688601), but clearly this ought to work
without needing such a patch. This patch arranges for the shell loop
to be over the test names without a $(objdir) prefix, which reduces
the space used to less than half MAX_ARG_STRLEN.
(I think we ought to aim to get rid of bits/mathinline.h completely -
filing GCC bugs for any optimizations GCC can't currently do with
-ffast-math - which would mean we could halve the number of libm tests
run because separate inline function tests would no longer be needed.
However, with a long directory name even half the number of tests
could make this command exceed MAX_ARG_STRLEN without my patch.)
Tested regen-ulps on a system where it failed before this patch.
* math/Makefile (run-regen-ulps): Add $(objpfx) to test name here.
(regen-ulps): Use $(libm-tests) not $^ in shell loop.
Joseph Myers
Szabolcs Nagy
Wilco Dijkstra
Tulio Magno Quites Machado Filho
Gabriel F. T. Gomes
Stefan Liebler