mirror of
https://github.com/jart/cosmopolitan.git
synced 2025-01-31 11:37:35 +00:00
455 lines
9.4 KiB
C
455 lines
9.4 KiB
C
/* clang-format off */
|
|
#if defined(__aarch64__) && !(__ASSEMBLER__ + __LINKER__ + 0)
|
|
#ifndef _AARCH64_FP16_H_
|
|
#define _AARCH64_FP16_H_
|
|
#pragma GCC push_options
|
|
#pragma GCC target ("arch=armv8.2-a+fp16")
|
|
typedef __fp16 float16_t;
|
|
__funline float16_t
|
|
vabsh_f16 (float16_t __a)
|
|
{
|
|
return __builtin_aarch64_abshf (__a);
|
|
}
|
|
__funline uint16_t
|
|
vceqzh_f16 (float16_t __a)
|
|
{
|
|
return __builtin_aarch64_cmeqhf_uss (__a, 0.0f);
|
|
}
|
|
__funline uint16_t
|
|
vcgezh_f16 (float16_t __a)
|
|
{
|
|
return __builtin_aarch64_cmgehf_uss (__a, 0.0f);
|
|
}
|
|
__funline uint16_t
|
|
vcgtzh_f16 (float16_t __a)
|
|
{
|
|
return __builtin_aarch64_cmgthf_uss (__a, 0.0f);
|
|
}
|
|
__funline uint16_t
|
|
vclezh_f16 (float16_t __a)
|
|
{
|
|
return __builtin_aarch64_cmlehf_uss (__a, 0.0f);
|
|
}
|
|
__funline uint16_t
|
|
vcltzh_f16 (float16_t __a)
|
|
{
|
|
return __builtin_aarch64_cmlthf_uss (__a, 0.0f);
|
|
}
|
|
__funline float16_t
|
|
vcvth_f16_s16 (int16_t __a)
|
|
{
|
|
return __builtin_aarch64_floathihf (__a);
|
|
}
|
|
__funline float16_t
|
|
vcvth_f16_s32 (int32_t __a)
|
|
{
|
|
return __builtin_aarch64_floatsihf (__a);
|
|
}
|
|
__funline float16_t
|
|
vcvth_f16_s64 (int64_t __a)
|
|
{
|
|
return __builtin_aarch64_floatdihf (__a);
|
|
}
|
|
__funline float16_t
|
|
vcvth_f16_u16 (uint16_t __a)
|
|
{
|
|
return __builtin_aarch64_floatunshihf_us (__a);
|
|
}
|
|
__funline float16_t
|
|
vcvth_f16_u32 (uint32_t __a)
|
|
{
|
|
return __builtin_aarch64_floatunssihf_us (__a);
|
|
}
|
|
__funline float16_t
|
|
vcvth_f16_u64 (uint64_t __a)
|
|
{
|
|
return __builtin_aarch64_floatunsdihf_us (__a);
|
|
}
|
|
__funline int16_t
|
|
vcvth_s16_f16 (float16_t __a)
|
|
{
|
|
return __builtin_aarch64_fix_trunchfhi (__a);
|
|
}
|
|
__funline int32_t
|
|
vcvth_s32_f16 (float16_t __a)
|
|
{
|
|
return __builtin_aarch64_fix_trunchfsi (__a);
|
|
}
|
|
__funline int64_t
|
|
vcvth_s64_f16 (float16_t __a)
|
|
{
|
|
return __builtin_aarch64_fix_trunchfdi (__a);
|
|
}
|
|
__funline uint16_t
|
|
vcvth_u16_f16 (float16_t __a)
|
|
{
|
|
return __builtin_aarch64_fixuns_trunchfhi_us (__a);
|
|
}
|
|
__funline uint32_t
|
|
vcvth_u32_f16 (float16_t __a)
|
|
{
|
|
return __builtin_aarch64_fixuns_trunchfsi_us (__a);
|
|
}
|
|
__funline uint64_t
|
|
vcvth_u64_f16 (float16_t __a)
|
|
{
|
|
return __builtin_aarch64_fixuns_trunchfdi_us (__a);
|
|
}
|
|
__funline int16_t
|
|
vcvtah_s16_f16 (float16_t __a)
|
|
{
|
|
return __builtin_aarch64_lroundhfhi (__a);
|
|
}
|
|
__funline int32_t
|
|
vcvtah_s32_f16 (float16_t __a)
|
|
{
|
|
return __builtin_aarch64_lroundhfsi (__a);
|
|
}
|
|
__funline int64_t
|
|
vcvtah_s64_f16 (float16_t __a)
|
|
{
|
|
return __builtin_aarch64_lroundhfdi (__a);
|
|
}
|
|
__funline uint16_t
|
|
vcvtah_u16_f16 (float16_t __a)
|
|
{
|
|
return __builtin_aarch64_lrounduhfhi_us (__a);
|
|
}
|
|
__funline uint32_t
|
|
vcvtah_u32_f16 (float16_t __a)
|
|
{
|
|
return __builtin_aarch64_lrounduhfsi_us (__a);
|
|
}
|
|
__funline uint64_t
|
|
vcvtah_u64_f16 (float16_t __a)
|
|
{
|
|
return __builtin_aarch64_lrounduhfdi_us (__a);
|
|
}
|
|
__funline int16_t
|
|
vcvtmh_s16_f16 (float16_t __a)
|
|
{
|
|
return __builtin_aarch64_lfloorhfhi (__a);
|
|
}
|
|
__funline int32_t
|
|
vcvtmh_s32_f16 (float16_t __a)
|
|
{
|
|
return __builtin_aarch64_lfloorhfsi (__a);
|
|
}
|
|
__funline int64_t
|
|
vcvtmh_s64_f16 (float16_t __a)
|
|
{
|
|
return __builtin_aarch64_lfloorhfdi (__a);
|
|
}
|
|
__funline uint16_t
|
|
vcvtmh_u16_f16 (float16_t __a)
|
|
{
|
|
return __builtin_aarch64_lflooruhfhi_us (__a);
|
|
}
|
|
__funline uint32_t
|
|
vcvtmh_u32_f16 (float16_t __a)
|
|
{
|
|
return __builtin_aarch64_lflooruhfsi_us (__a);
|
|
}
|
|
__funline uint64_t
|
|
vcvtmh_u64_f16 (float16_t __a)
|
|
{
|
|
return __builtin_aarch64_lflooruhfdi_us (__a);
|
|
}
|
|
__funline int16_t
|
|
vcvtnh_s16_f16 (float16_t __a)
|
|
{
|
|
return __builtin_aarch64_lfrintnhfhi (__a);
|
|
}
|
|
__funline int32_t
|
|
vcvtnh_s32_f16 (float16_t __a)
|
|
{
|
|
return __builtin_aarch64_lfrintnhfsi (__a);
|
|
}
|
|
__funline int64_t
|
|
vcvtnh_s64_f16 (float16_t __a)
|
|
{
|
|
return __builtin_aarch64_lfrintnhfdi (__a);
|
|
}
|
|
__funline uint16_t
|
|
vcvtnh_u16_f16 (float16_t __a)
|
|
{
|
|
return __builtin_aarch64_lfrintnuhfhi_us (__a);
|
|
}
|
|
__funline uint32_t
|
|
vcvtnh_u32_f16 (float16_t __a)
|
|
{
|
|
return __builtin_aarch64_lfrintnuhfsi_us (__a);
|
|
}
|
|
__funline uint64_t
|
|
vcvtnh_u64_f16 (float16_t __a)
|
|
{
|
|
return __builtin_aarch64_lfrintnuhfdi_us (__a);
|
|
}
|
|
__funline int16_t
|
|
vcvtph_s16_f16 (float16_t __a)
|
|
{
|
|
return __builtin_aarch64_lceilhfhi (__a);
|
|
}
|
|
__funline int32_t
|
|
vcvtph_s32_f16 (float16_t __a)
|
|
{
|
|
return __builtin_aarch64_lceilhfsi (__a);
|
|
}
|
|
__funline int64_t
|
|
vcvtph_s64_f16 (float16_t __a)
|
|
{
|
|
return __builtin_aarch64_lceilhfdi (__a);
|
|
}
|
|
__funline uint16_t
|
|
vcvtph_u16_f16 (float16_t __a)
|
|
{
|
|
return __builtin_aarch64_lceiluhfhi_us (__a);
|
|
}
|
|
__funline uint32_t
|
|
vcvtph_u32_f16 (float16_t __a)
|
|
{
|
|
return __builtin_aarch64_lceiluhfsi_us (__a);
|
|
}
|
|
__funline uint64_t
|
|
vcvtph_u64_f16 (float16_t __a)
|
|
{
|
|
return __builtin_aarch64_lceiluhfdi_us (__a);
|
|
}
|
|
__funline float16_t
|
|
vnegh_f16 (float16_t __a)
|
|
{
|
|
return __builtin_aarch64_neghf (__a);
|
|
}
|
|
__funline float16_t
|
|
vrecpeh_f16 (float16_t __a)
|
|
{
|
|
return __builtin_aarch64_frecpehf (__a);
|
|
}
|
|
__funline float16_t
|
|
vrecpxh_f16 (float16_t __a)
|
|
{
|
|
return __builtin_aarch64_frecpxhf (__a);
|
|
}
|
|
__funline float16_t
|
|
vrndh_f16 (float16_t __a)
|
|
{
|
|
return __builtin_aarch64_btrunchf (__a);
|
|
}
|
|
__funline float16_t
|
|
vrndah_f16 (float16_t __a)
|
|
{
|
|
return __builtin_aarch64_roundhf (__a);
|
|
}
|
|
__funline float16_t
|
|
vrndih_f16 (float16_t __a)
|
|
{
|
|
return __builtin_aarch64_nearbyinthf (__a);
|
|
}
|
|
__funline float16_t
|
|
vrndmh_f16 (float16_t __a)
|
|
{
|
|
return __builtin_aarch64_floorhf (__a);
|
|
}
|
|
__funline float16_t
|
|
vrndnh_f16 (float16_t __a)
|
|
{
|
|
return __builtin_aarch64_frintnhf (__a);
|
|
}
|
|
__funline float16_t
|
|
vrndph_f16 (float16_t __a)
|
|
{
|
|
return __builtin_aarch64_ceilhf (__a);
|
|
}
|
|
__funline float16_t
|
|
vrndxh_f16 (float16_t __a)
|
|
{
|
|
return __builtin_aarch64_rinthf (__a);
|
|
}
|
|
__funline float16_t
|
|
vrsqrteh_f16 (float16_t __a)
|
|
{
|
|
return __builtin_aarch64_rsqrtehf (__a);
|
|
}
|
|
__funline float16_t
|
|
vsqrth_f16 (float16_t __a)
|
|
{
|
|
return __builtin_aarch64_sqrthf (__a);
|
|
}
|
|
__funline float16_t
|
|
vaddh_f16 (float16_t __a, float16_t __b)
|
|
{
|
|
return __a + __b;
|
|
}
|
|
__funline float16_t
|
|
vabdh_f16 (float16_t __a, float16_t __b)
|
|
{
|
|
return __builtin_aarch64_fabdhf (__a, __b);
|
|
}
|
|
__funline uint16_t
|
|
vcageh_f16 (float16_t __a, float16_t __b)
|
|
{
|
|
return __builtin_aarch64_facgehf_uss (__a, __b);
|
|
}
|
|
__funline uint16_t
|
|
vcagth_f16 (float16_t __a, float16_t __b)
|
|
{
|
|
return __builtin_aarch64_facgthf_uss (__a, __b);
|
|
}
|
|
__funline uint16_t
|
|
vcaleh_f16 (float16_t __a, float16_t __b)
|
|
{
|
|
return __builtin_aarch64_faclehf_uss (__a, __b);
|
|
}
|
|
__funline uint16_t
|
|
vcalth_f16 (float16_t __a, float16_t __b)
|
|
{
|
|
return __builtin_aarch64_faclthf_uss (__a, __b);
|
|
}
|
|
__funline uint16_t
|
|
vceqh_f16 (float16_t __a, float16_t __b)
|
|
{
|
|
return __builtin_aarch64_cmeqhf_uss (__a, __b);
|
|
}
|
|
__funline uint16_t
|
|
vcgeh_f16 (float16_t __a, float16_t __b)
|
|
{
|
|
return __builtin_aarch64_cmgehf_uss (__a, __b);
|
|
}
|
|
__funline uint16_t
|
|
vcgth_f16 (float16_t __a, float16_t __b)
|
|
{
|
|
return __builtin_aarch64_cmgthf_uss (__a, __b);
|
|
}
|
|
__funline uint16_t
|
|
vcleh_f16 (float16_t __a, float16_t __b)
|
|
{
|
|
return __builtin_aarch64_cmlehf_uss (__a, __b);
|
|
}
|
|
__funline uint16_t
|
|
vclth_f16 (float16_t __a, float16_t __b)
|
|
{
|
|
return __builtin_aarch64_cmlthf_uss (__a, __b);
|
|
}
|
|
__funline float16_t
|
|
vcvth_n_f16_s16 (int16_t __a, const int __b)
|
|
{
|
|
return __builtin_aarch64_scvtfhi (__a, __b);
|
|
}
|
|
__funline float16_t
|
|
vcvth_n_f16_s32 (int32_t __a, const int __b)
|
|
{
|
|
return __builtin_aarch64_scvtfsihf (__a, __b);
|
|
}
|
|
__funline float16_t
|
|
vcvth_n_f16_s64 (int64_t __a, const int __b)
|
|
{
|
|
return __builtin_aarch64_scvtfdihf (__a, __b);
|
|
}
|
|
__funline float16_t
|
|
vcvth_n_f16_u16 (uint16_t __a, const int __b)
|
|
{
|
|
return __builtin_aarch64_ucvtfhi_sus (__a, __b);
|
|
}
|
|
__funline float16_t
|
|
vcvth_n_f16_u32 (uint32_t __a, const int __b)
|
|
{
|
|
return __builtin_aarch64_ucvtfsihf_sus (__a, __b);
|
|
}
|
|
__funline float16_t
|
|
vcvth_n_f16_u64 (uint64_t __a, const int __b)
|
|
{
|
|
return __builtin_aarch64_ucvtfdihf_sus (__a, __b);
|
|
}
|
|
__funline int16_t
|
|
vcvth_n_s16_f16 (float16_t __a, const int __b)
|
|
{
|
|
return __builtin_aarch64_fcvtzshf (__a, __b);
|
|
}
|
|
__funline int32_t
|
|
vcvth_n_s32_f16 (float16_t __a, const int __b)
|
|
{
|
|
return __builtin_aarch64_fcvtzshfsi (__a, __b);
|
|
}
|
|
__funline int64_t
|
|
vcvth_n_s64_f16 (float16_t __a, const int __b)
|
|
{
|
|
return __builtin_aarch64_fcvtzshfdi (__a, __b);
|
|
}
|
|
__funline uint16_t
|
|
vcvth_n_u16_f16 (float16_t __a, const int __b)
|
|
{
|
|
return __builtin_aarch64_fcvtzuhf_uss (__a, __b);
|
|
}
|
|
__funline uint32_t
|
|
vcvth_n_u32_f16 (float16_t __a, const int __b)
|
|
{
|
|
return __builtin_aarch64_fcvtzuhfsi_uss (__a, __b);
|
|
}
|
|
__funline uint64_t
|
|
vcvth_n_u64_f16 (float16_t __a, const int __b)
|
|
{
|
|
return __builtin_aarch64_fcvtzuhfdi_uss (__a, __b);
|
|
}
|
|
__funline float16_t
|
|
vdivh_f16 (float16_t __a, float16_t __b)
|
|
{
|
|
return __a / __b;
|
|
}
|
|
__funline float16_t
|
|
vmaxh_f16 (float16_t __a, float16_t __b)
|
|
{
|
|
return __builtin_aarch64_fmaxhf (__a, __b);
|
|
}
|
|
__funline float16_t
|
|
vmaxnmh_f16 (float16_t __a, float16_t __b)
|
|
{
|
|
return __builtin_aarch64_fmaxhf (__a, __b);
|
|
}
|
|
__funline float16_t
|
|
vminh_f16 (float16_t __a, float16_t __b)
|
|
{
|
|
return __builtin_aarch64_fminhf (__a, __b);
|
|
}
|
|
__funline float16_t
|
|
vminnmh_f16 (float16_t __a, float16_t __b)
|
|
{
|
|
return __builtin_aarch64_fminhf (__a, __b);
|
|
}
|
|
__funline float16_t
|
|
vmulh_f16 (float16_t __a, float16_t __b)
|
|
{
|
|
return __a * __b;
|
|
}
|
|
__funline float16_t
|
|
vmulxh_f16 (float16_t __a, float16_t __b)
|
|
{
|
|
return __builtin_aarch64_fmulxhf (__a, __b);
|
|
}
|
|
__funline float16_t
|
|
vrecpsh_f16 (float16_t __a, float16_t __b)
|
|
{
|
|
return __builtin_aarch64_frecpshf (__a, __b);
|
|
}
|
|
__funline float16_t
|
|
vrsqrtsh_f16 (float16_t __a, float16_t __b)
|
|
{
|
|
return __builtin_aarch64_rsqrtshf (__a, __b);
|
|
}
|
|
__funline float16_t
|
|
vsubh_f16 (float16_t __a, float16_t __b)
|
|
{
|
|
return __a - __b;
|
|
}
|
|
__funline float16_t
|
|
vfmah_f16 (float16_t __a, float16_t __b, float16_t __c)
|
|
{
|
|
return __builtin_aarch64_fmahf (__b, __c, __a);
|
|
}
|
|
__funline float16_t
|
|
vfmsh_f16 (float16_t __a, float16_t __b, float16_t __c)
|
|
{
|
|
return __builtin_aarch64_fnmahf (__b, __c, __a);
|
|
}
|
|
#pragma GCC pop_options
|
|
#endif
|
|
#endif
|