[PATCH v7 05/13] LoongArch: Generic <math.h> and soft-fp Routines
Adhemerval Zanella Netto
adhemerval.zanella@linaro.org
Wed Jul 20 13:04:07 GMT 2022
LGTM, thanks.
Reviewed-by: Adhemerval Zanella <adhemerval.zanella@linaro.org>
On 18/07/22 22:20, caiyinyu wrote:
> ---
> sysdeps/loongarch/bits/fenv.h | 90 ++++++++++++++++++++++++++++
> sysdeps/loongarch/e_sqrtl.c | 39 ++++++++++++
> sysdeps/loongarch/fpu_control.h | 90 ++++++++++++++++++++++++++++
> sysdeps/loongarch/sfp-machine.h | 102 ++++++++++++++++++++++++++++++++
> sysdeps/loongarch/tininess.h | 1 +
> 5 files changed, 322 insertions(+)
> create mode 100644 sysdeps/loongarch/bits/fenv.h
> create mode 100644 sysdeps/loongarch/e_sqrtl.c
> create mode 100644 sysdeps/loongarch/fpu_control.h
> create mode 100644 sysdeps/loongarch/sfp-machine.h
> create mode 100644 sysdeps/loongarch/tininess.h
>
> diff --git a/sysdeps/loongarch/bits/fenv.h b/sysdeps/loongarch/bits/fenv.h
> new file mode 100644
> index 0000000000..7e223b402b
> --- /dev/null
> +++ b/sysdeps/loongarch/bits/fenv.h
> @@ -0,0 +1,90 @@
> +/* Floating point environment.
> + Copyright (C) 2022 Free Software Foundation, Inc.
> + This file is part of the GNU C Library.
> +
> + The GNU C Library is free software; you can redistribute it and/or
> + modify it under the terms of the GNU Lesser General Public
> + License as published by the Free Software Foundation; either
> + version 2.1 of the License, or (at your option) any later version.
> +
> + The GNU C Library is distributed in the hope that it will be useful,
> + but WITHOUT ANY WARRANTY; without even the implied warranty of
> + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
> + Lesser General Public License for more details.
> +
> + You should have received a copy of the GNU Lesser General Public
> + License along with the GNU C Library. If not, see
> + <https://www.gnu.org/licenses/>. */
> +
> +#ifndef _FENV_H
> +#error "Never use <bits/fenv.h> directly; include <fenv.h> instead."
> +#endif
> +
> +/* Define bits representing the exception. We use the bit positions
> + of the appropriate bits in the FPU control word. */
> +enum
> +{
> + FE_INEXACT =
> +#define FE_INEXACT 0x010000
> + FE_INEXACT,
> + FE_UNDERFLOW =
> +#define FE_UNDERFLOW 0x020000
> + FE_UNDERFLOW,
> + FE_OVERFLOW =
> +#define FE_OVERFLOW 0x040000
> + FE_OVERFLOW,
> + FE_DIVBYZERO =
> +#define FE_DIVBYZERO 0x080000
> + FE_DIVBYZERO,
> + FE_INVALID =
> +#define FE_INVALID 0x100000
> + FE_INVALID,
> +};
> +
> +#define FE_ALL_EXCEPT \
> + (FE_INEXACT | FE_DIVBYZERO | FE_UNDERFLOW | FE_OVERFLOW | FE_INVALID)
> +
> +/* The LoongArch FPU supports all of the four defined rounding modes. We
> + use again the bit positions in the FPU control word as the values
> + for the appropriate macros. */
> +enum
> +{
> + FE_TONEAREST =
> +#define FE_TONEAREST 0x000
> + FE_TONEAREST,
> + FE_TOWARDZERO =
> +#define FE_TOWARDZERO 0x100
> + FE_TOWARDZERO,
> + FE_UPWARD =
> +#define FE_UPWARD 0x200
> + FE_UPWARD,
> + FE_DOWNWARD =
> +#define FE_DOWNWARD 0x300
> + FE_DOWNWARD
> +};
> +
> +/* Type representing exception flags. */
> +typedef unsigned int fexcept_t;
> +
> +/* Type representing floating-point environment. This function corresponds
> + to the layout of the block written by the `fstenv'. */
> +typedef struct
> +{
> + unsigned int __fp_control_register;
> +} fenv_t;
> +
> +/* If the default argument is used we use this value. */
> +#define FE_DFL_ENV ((const fenv_t *) -1)
> +
> +#ifdef __USE_GNU
> +/* Floating-point environment where none of the exception is masked. */
> +#define FE_NOMASK_ENV ((const fenv_t *) -257)
> +#endif
> +
> +#if __GLIBC_USE (IEC_60559_BFP_EXT_C2X)
> +/* Type representing floating-point control modes. */
> +typedef unsigned int femode_t;
> +
> +/* Default floating-point control modes. */
> +#define FE_DFL_MODE ((const femode_t *) -1L)
> +#endif
Ok.
> diff --git a/sysdeps/loongarch/e_sqrtl.c b/sysdeps/loongarch/e_sqrtl.c
> new file mode 100644
> index 0000000000..020c8911e6
> --- /dev/null
> +++ b/sysdeps/loongarch/e_sqrtl.c
> @@ -0,0 +1,39 @@
> +/* long double square root in software floating-point emulation.
> + Copyright (C) 2022 Free Software Foundation, Inc.
> + This file is part of the GNU C Library.
> +
> + The GNU C Library is free software; you can redistribute it and/or
> + modify it under the terms of the GNU Lesser General Public
> + License as published by the Free Software Foundation; either
> + version 2.1 of the License, or (at your option) any later version.
> +
> + The GNU C Library is distributed in the hope that it will be useful,
> + but WITHOUT ANY WARRANTY; without even the implied warranty of
> + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
> + Lesser General Public License for more details.
> +
> + You should have received a copy of the GNU Lesser General Public
> + License along with the GNU C Library. If not, see
> + <https://www.gnu.org/licenses/>. */
> +
> +#include <stdlib.h>
> +#include <soft-fp/soft-fp.h>
> +#include <soft-fp/quad.h>
> +#include <libm-alias-finite.h>
> +
> +long double
> +__ieee754_sqrtl (const long double a)
> +{
> + FP_DECL_EX;
> + FP_DECL_Q (A);
> + FP_DECL_Q (C);
> + long double c;
> +
> + FP_INIT_ROUNDMODE;
> + FP_UNPACK_Q (A, a);
> + FP_SQRT_Q (C, A);
> + FP_PACK_Q (c, C);
> + FP_HANDLE_EXCEPTIONS;
> + return c;
> +}
> +libm_alias_finite (__ieee754_sqrtl, __sqrtl)
Ok.
> diff --git a/sysdeps/loongarch/fpu_control.h b/sysdeps/loongarch/fpu_control.h
> new file mode 100644
> index 0000000000..d34936c62b
> --- /dev/null
> +++ b/sysdeps/loongarch/fpu_control.h
> @@ -0,0 +1,90 @@
> +/* FPU control word bits.
> + Copyright (C) 2022 Free Software Foundation, Inc.
> + This file is part of the GNU C Library.
> +
> + The GNU C Library is free software; you can redistribute it and/or
> + modify it under the terms of the GNU Lesser General Public
> + License as published by the Free Software Foundation; either
> + version 2.1 of the License, or (at your option) any later version.
> +
> + The GNU C Library is distributed in the hope that it will be useful,
> + but WITHOUT ANY WARRANTY; without even the implied warranty of
> + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
> + Lesser General Public License for more details.
> +
> + You should have received a copy of the GNU Lesser General Public
> + License along with the GNU C Library. If not, see
> + <https://www.gnu.org/licenses/>. */
> +
> +#ifndef _FPU_CONTROL_H
> +#define _FPU_CONTROL_H
> +
> +/* LoongArch FPU floating point control register bits.
> + *
> + * 31-29 -> reserved (read as 0, can not changed by software)
> + * 28 -> cause bit for invalid exception
> + * 27 -> cause bit for division by zero exception
> + * 26 -> cause bit for overflow exception
> + * 25 -> cause bit for underflow exception
> + * 24 -> cause bit for inexact exception
> + * 23-21 -> reserved (read as 0, can not changed by software)
> + * 20 -> flag invalid exception
> + * 19 -> flag division by zero exception
> + * 18 -> flag overflow exception
> + * 17 -> flag underflow exception
> + * 16 -> flag inexact exception
> + * 9-8 -> rounding control
> + * 7-5 -> reserved (read as 0, can not changed by software)
> + * 4 -> enable exception for invalid exception
> + * 3 -> enable exception for division by zero exception
> + * 2 -> enable exception for overflow exception
> + * 1 -> enable exception for underflow exception
> + * 0 -> enable exception for inexact exception
> + *
> + *
> + * Rounding Control:
> + * 00 - rounding ties to even (RNE)
> + * 01 - rounding toward zero (RZ)
> + * 10 - rounding (up) toward plus infinity (RP)
> + * 11 - rounding (down) toward minus infinity (RM)
> + */
> +
> +#include <features.h>
> +
> +/* Masks for interrupts. */
> +#define _FPU_MASK_V 0x10 /* Invalid operation */
> +#define _FPU_MASK_Z 0x08 /* Division by zero */
> +#define _FPU_MASK_O 0x04 /* Overflow */
> +#define _FPU_MASK_U 0x02 /* Underflow */
> +#define _FPU_MASK_I 0x01 /* Inexact operation */
> +
> +/* Flush denormalized numbers to zero. */
> +#define _FPU_FLUSH_TZ 0x1000000
> +
> +/* Rounding control. */
> +#define _FPU_RC_NEAREST 0x000 /* RECOMMENDED */
> +#define _FPU_RC_ZERO 0x100
> +#define _FPU_RC_UP 0x200
> +#define _FPU_RC_DOWN 0x300
> +/* Mask for rounding control. */
> +#define _FPU_RC_MASK 0x300
> +
> +#define _FPU_RESERVED 0x0
> +
> +#define _FPU_DEFAULT 0x0
> +#define _FPU_IEEE 0x1F
> +
> +/* Type of the control word. */
> +typedef unsigned int fpu_control_t __attribute__ ((__mode__ (__SI__)));
> +
> +/* Macros for accessing the hardware control word. */
> +extern fpu_control_t __loongarch_fpu_getcw (void) __THROW;
> +extern void __loongarch_fpu_setcw (fpu_control_t) __THROW;
> +#define _FPU_GETCW(cw) __asm__ volatile ("movfcsr2gr %0,$r0" : "=r"(cw))
> +#define _FPU_SETCW(cw) __asm__ volatile ("movgr2fcsr $r0,%0" : : "r"(cw))
> +
> +/* Default control word set at startup. */
> +extern fpu_control_t __fpu_control;
> +
> +#endif /* fpu_control.h */
> +
Ok.
> diff --git a/sysdeps/loongarch/sfp-machine.h b/sysdeps/loongarch/sfp-machine.h
> new file mode 100644
> index 0000000000..5b92ac4ba2
> --- /dev/null
> +++ b/sysdeps/loongarch/sfp-machine.h
> @@ -0,0 +1,102 @@
> +/* LoongArch softfloat definitions
> + Copyright (C) 2022 Free Software Foundation, Inc.
> + This file is part of the GNU C Library.
> +
> +
> + The GNU C Library is free software; you can redistribute it and/or
> + modify it under the terms of the GNU Lesser General Public
> + License as published by the Free Software Foundation; either
> + version 2.1 of the License, or (at your option) any later version.
> +
> + The GNU C Library is distributed in the hope that it will be useful,
> + but WITHOUT ANY WARRANTY; without even the implied warranty of
> + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
> + Lesser General Public License for more details.
> +
> + You should have received a copy of the GNU Lesser General Public
> + License along with the GNU C Library. If not, see
> + <https://www.gnu.org/licenses/>. */
> +
> +#include <fenv.h>
> +#include <fpu_control.h>
> +
> +#define _FP_W_TYPE_SIZE 64
> +#define _FP_W_TYPE unsigned long long
> +#define _FP_WS_TYPE signed long long
> +#define _FP_I_TYPE long long
> +
> +#define _FP_MUL_MEAT_S(R, X, Y) _FP_MUL_MEAT_1_imm (_FP_WFRACBITS_S, R, X, Y)
> +#define _FP_MUL_MEAT_D(R, X, Y) \
> + _FP_MUL_MEAT_1_wide (_FP_WFRACBITS_D, R, X, Y, umul_ppmm)
> +#define _FP_MUL_MEAT_Q(R, X, Y) \
> + _FP_MUL_MEAT_2_wide_3mul (_FP_WFRACBITS_Q, R, X, Y, umul_ppmm)
> +
> +#define _FP_MUL_MEAT_DW_S(R, X, Y) \
> + _FP_MUL_MEAT_DW_1_imm (_FP_WFRACBITS_S, R, X, Y)
> +#define _FP_MUL_MEAT_DW_D(R, X, Y) \
> + _FP_MUL_MEAT_DW_1_wide (_FP_WFRACBITS_D, R, X, Y, umul_ppmm)
> +#define _FP_MUL_MEAT_DW_Q(R, X, Y) \
> + _FP_MUL_MEAT_DW_2_wide_3mul (_FP_WFRACBITS_Q, R, X, Y, umul_ppmm)
> +
> +#define _FP_DIV_MEAT_S(R, X, Y) \
> + _FP_DIV_MEAT_1_imm (S, R, X, Y, _FP_DIV_HELP_imm)
> +#define _FP_DIV_MEAT_D(R, X, Y) _FP_DIV_MEAT_1_udiv_norm (D, R, X, Y)
> +#define _FP_DIV_MEAT_Q(R, X, Y) _FP_DIV_MEAT_2_udiv (Q, R, X, Y)
> +
> +#define _FP_NANFRAC_S _FP_QNANBIT_S
> +#define _FP_NANFRAC_D _FP_QNANBIT_D
> +#define _FP_NANFRAC_Q _FP_QNANBIT_Q, 0
> +
> +#define _FP_NANSIGN_S 0
> +#define _FP_NANSIGN_D 0
> +#define _FP_NANSIGN_Q 0
> +
> +#define _FP_KEEPNANFRACP 1
> +#define _FP_QNANNEGATEDP 0
> +
> +/* NaN payloads should be preserved for NAN2008. */
> +#define _FP_CHOOSENAN(fs, wc, R, X, Y, OP) \
> + do \
> + { \
> + R##_s = X##_s; \
> + _FP_FRAC_COPY_##wc (R, X); \
> + R##_c = FP_CLS_NAN; \
> + } \
> + while (0)
> +
> +#define _FP_DECL_EX fpu_control_t _fcw
> +
> +#define FP_ROUNDMODE (_fcw & 0x300)
> +
> +#define FP_RND_NEAREST FE_TONEAREST
> +#define FP_RND_ZERO FE_TOWARDZERO
> +#define FP_RND_PINF FE_UPWARD
> +#define FP_RND_MINF FE_DOWNWARD
> +
> +#define FP_EX_INVALID FE_INVALID
> +#define FP_EX_OVERFLOW FE_OVERFLOW
> +#define FP_EX_UNDERFLOW FE_UNDERFLOW
> +#define FP_EX_DIVZERO FE_DIVBYZERO
> +#define FP_EX_INEXACT FE_INEXACT
> +
> +#define _FP_TININESS_AFTER_ROUNDING 1
> +
> +#ifdef __loongarch_hard_float
> +#define FP_INIT_ROUNDMODE \
> + do \
> + { \
> + _FPU_GETCW (_fcw); \
> + } \
> + while (0)
> +
> +#define FP_HANDLE_EXCEPTIONS \
> + do \
> + { \
> + if (__glibc_unlikely (_fex)) \
> + _FPU_SETCW (_fcw | _fex | (_fex << 8)); \
> + } \
> + while (0)
> +#define FP_TRAPPING_EXCEPTIONS ((_fcw << 16) & 0x1f0000)
> +#else
> +#define FP_INIT_ROUNDMODE _fcw = FP_RND_NEAREST
> +#endif
> diff --git a/sysdeps/loongarch/tininess.h b/sysdeps/loongarch/tininess.h
> new file mode 100644
> index 0000000000..90956c35f7
> --- /dev/null
> +++ b/sysdeps/loongarch/tininess.h
> @@ -0,0 +1 @@
> +#define TININESS_AFTER_ROUNDING 1
Ok.
More information about the Libc-alpha
mailing list