///////////////////////////////////////////////////////////////////////// // $Id$ ///////////////////////////////////////////////////////////////////////// // // Copyright (c) 2011-2013 Stanislav Shwartsman // Written by Stanislav Shwartsman [sshwarts at sourceforge net] // // This 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 of the License, or (at your option) any later version. // // This 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 this library; if not, write to the Free Software // Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA B 02110-1301 USA // ///////////////////////////////////////////////////////////////////////// #ifndef BX_SIMD_PFP_FUNCTIONS_H #define BX_SIMD_PFP_FUNCTIONS_H // arithmetic add/sub/mul/div BX_CPP_INLINE void xmm_addps(BxPackedXmmRegister *op1, const BxPackedXmmRegister *op2, float_status_t &status) { for (unsigned n=0;n<4;n++) { op1->xmm32u(n) = float32_add(op1->xmm32u(n), op2->xmm32u(n), status); } } BX_CPP_INLINE void xmm_addps_mask(BxPackedXmmRegister *op1, const BxPackedXmmRegister *op2, float_status_t &status, Bit32u mask) { for (unsigned n=0; n < 4; n++, mask >>= 1) { if (mask & 0x1) op1->xmm32u(n) = float32_add(op1->xmm32u(n), op2->xmm32u(n), status); else op1->xmm32u(n) = 0; } } BX_CPP_INLINE void xmm_addpd(BxPackedXmmRegister *op1, const BxPackedXmmRegister *op2, float_status_t &status) { for (unsigned n=0;n<2;n++) { op1->xmm64u(n) = float64_add(op1->xmm64u(n), op2->xmm64u(n), status); } } BX_CPP_INLINE void xmm_addpd_mask(BxPackedXmmRegister *op1, const BxPackedXmmRegister *op2, float_status_t &status, Bit32u mask) { for (unsigned n=0; n < 2; n++, mask >>= 1) { if (mask & 0x1) op1->xmm64u(n) = float64_add(op1->xmm64u(n), op2->xmm64u(n), status); else op1->xmm64u(n) = 0; } } BX_CPP_INLINE void xmm_subps(BxPackedXmmRegister *op1, const BxPackedXmmRegister *op2, float_status_t &status) { for (unsigned n=0;n<4;n++) { op1->xmm32u(n) = float32_sub(op1->xmm32u(n), op2->xmm32u(n), status); } } BX_CPP_INLINE void xmm_subps_mask(BxPackedXmmRegister *op1, const BxPackedXmmRegister *op2, float_status_t &status, Bit32u mask) { for (unsigned n=0; n < 4; n++, mask >>= 1) { if (mask & 0x1) op1->xmm32u(n) = float32_sub(op1->xmm32u(n), op2->xmm32u(n), status); else op1->xmm32u(n) = 0; } } BX_CPP_INLINE void xmm_subpd(BxPackedXmmRegister *op1, const BxPackedXmmRegister *op2, float_status_t &status) { for (unsigned n=0;n<2;n++) { op1->xmm64u(n) = float64_sub(op1->xmm64u(n), op2->xmm64u(n), status); } } BX_CPP_INLINE void xmm_subpd_mask(BxPackedXmmRegister *op1, const BxPackedXmmRegister *op2, float_status_t &status, Bit32u mask) { for (unsigned n=0; n < 2; n++, mask >>= 1) { if (mask & 0x1) op1->xmm64u(n) = float64_sub(op1->xmm64u(n), op2->xmm64u(n), status); else op1->xmm64u(n) = 0; } } BX_CPP_INLINE void xmm_mulps(BxPackedXmmRegister *op1, const BxPackedXmmRegister *op2, float_status_t &status) { for (unsigned n=0;n<4;n++) { op1->xmm32u(n) = float32_mul(op1->xmm32u(n), op2->xmm32u(n), status); } } BX_CPP_INLINE void xmm_mulps_mask(BxPackedXmmRegister *op1, const BxPackedXmmRegister *op2, float_status_t &status, Bit32u mask) { for (unsigned n=0; n < 4; n++, mask >>= 1) { if (mask & 0x1) op1->xmm32u(n) = float32_mul(op1->xmm32u(n), op2->xmm32u(n), status); else op1->xmm32u(n) = 0; } } BX_CPP_INLINE void xmm_mulpd(BxPackedXmmRegister *op1, const BxPackedXmmRegister *op2, float_status_t &status) { for (unsigned n=0;n<2;n++) { op1->xmm64u(n) = float64_mul(op1->xmm64u(n), op2->xmm64u(n), status); } } BX_CPP_INLINE void xmm_mulpd_mask(BxPackedXmmRegister *op1, const BxPackedXmmRegister *op2, float_status_t &status, Bit32u mask) { for (unsigned n=0; n < 2; n++, mask >>= 1) { if (mask & 0x1) op1->xmm64u(n) = float64_mul(op1->xmm64u(n), op2->xmm64u(n), status); else op1->xmm64u(n) = 0; } } BX_CPP_INLINE void xmm_divps(BxPackedXmmRegister *op1, const BxPackedXmmRegister *op2, float_status_t &status) { for (unsigned n=0;n<4;n++) { op1->xmm32u(n) = float32_div(op1->xmm32u(n), op2->xmm32u(n), status); } } BX_CPP_INLINE void xmm_divps_mask(BxPackedXmmRegister *op1, const BxPackedXmmRegister *op2, float_status_t &status, Bit32u mask) { for (unsigned n=0; n < 4; n++, mask >>= 1) { if (mask & 0x1) op1->xmm32u(n) = float32_div(op1->xmm32u(n), op2->xmm32u(n), status); else op1->xmm32u(n) = 0; } } BX_CPP_INLINE void xmm_divpd(BxPackedXmmRegister *op1, const BxPackedXmmRegister *op2, float_status_t &status) { for (unsigned n=0;n<2;n++) { op1->xmm64u(n) = float64_div(op1->xmm64u(n), op2->xmm64u(n), status); } } BX_CPP_INLINE void xmm_divpd_mask(BxPackedXmmRegister *op1, const BxPackedXmmRegister *op2, float_status_t &status, Bit32u mask) { for (unsigned n=0; n < 2; n++, mask >>= 1) { if (mask & 0x1) op1->xmm64u(n) = float64_div(op1->xmm64u(n), op2->xmm64u(n), status); else op1->xmm64u(n) = 0; } } BX_CPP_INLINE void xmm_addsubps(BxPackedXmmRegister *op1, const BxPackedXmmRegister *op2, float_status_t &status) { op1->xmm32u(0) = float32_sub(op1->xmm32u(0), op2->xmm32u(0), status); op1->xmm32u(1) = float32_add(op1->xmm32u(1), op2->xmm32u(1), status); op1->xmm32u(2) = float32_sub(op1->xmm32u(2), op2->xmm32u(2), status); op1->xmm32u(3) = float32_add(op1->xmm32u(3), op2->xmm32u(3), status); } BX_CPP_INLINE void xmm_addsubps_mask(BxPackedXmmRegister *op1, const BxPackedXmmRegister *op2, float_status_t &status, Bit32u mask) { if (mask & 0x1) op1->xmm32u(0) = float32_sub(op1->xmm32u(0), op2->xmm32u(0), status); else op1->xmm32u(0) = 0; if (mask & 0x2) op1->xmm32u(1) = float32_add(op1->xmm32u(1), op2->xmm32u(1), status); else op1->xmm32u(1) = 0; if (mask & 0x4) op1->xmm32u(2) = float32_sub(op1->xmm32u(2), op2->xmm32u(2), status); else op1->xmm32u(2) = 0; if (mask & 0x8) op1->xmm32u(3) = float32_add(op1->xmm32u(3), op2->xmm32u(3), status); else op1->xmm32u(3) = 0; } BX_CPP_INLINE void xmm_addsubpd(BxPackedXmmRegister *op1, const BxPackedXmmRegister *op2, float_status_t &status) { op1->xmm64u(0) = float64_sub(op1->xmm64u(0), op2->xmm64u(0), status); op1->xmm64u(1) = float64_add(op1->xmm64u(1), op2->xmm64u(1), status); } BX_CPP_INLINE void xmm_addsubpd_mask(BxPackedXmmRegister *op1, const BxPackedXmmRegister *op2, float_status_t &status, Bit32u mask) { if (mask & 0x1) op1->xmm64u(0) = float64_sub(op1->xmm64u(0), op2->xmm64u(0), status); else op1->xmm64u(0) = 0; if (mask & 0x2) op1->xmm64u(1) = float64_add(op1->xmm64u(1), op2->xmm64u(1), status); else op1->xmm64u(1) = 0; } // horizontal arithmetic add/sub BX_CPP_INLINE void xmm_haddps(BxPackedXmmRegister *op1, const BxPackedXmmRegister *op2, float_status_t &status) { op1->xmm32u(0) = float32_add(op1->xmm32u(0), op1->xmm32u(1), status); op1->xmm32u(1) = float32_add(op1->xmm32u(2), op1->xmm32u(3), status); op1->xmm32u(2) = float32_add(op2->xmm32u(0), op2->xmm32u(1), status); op1->xmm32u(3) = float32_add(op2->xmm32u(2), op2->xmm32u(3), status); } BX_CPP_INLINE void xmm_haddps_mask(BxPackedXmmRegister *op1, const BxPackedXmmRegister *op2, float_status_t &status, Bit32u mask) { if (mask & 0x1) op1->xmm32u(0) = float32_add(op1->xmm32u(0), op1->xmm32u(1), status); else op1->xmm32u(0) = 0; if (mask & 0x2) op1->xmm32u(1) = float32_add(op1->xmm32u(2), op1->xmm32u(3), status); else op1->xmm32u(1) = 0; if (mask & 0x4) op1->xmm32u(2) = float32_add(op2->xmm32u(0), op2->xmm32u(1), status); else op1->xmm32u(2) = 0; if (mask & 0x8) op1->xmm32u(3) = float32_add(op2->xmm32u(2), op2->xmm32u(3), status); else op1->xmm32u(3) = 0; } BX_CPP_INLINE void xmm_haddpd(BxPackedXmmRegister *op1, const BxPackedXmmRegister *op2, float_status_t &status) { op1->xmm64u(0) = float64_add(op1->xmm64u(0), op1->xmm64u(1), status); op1->xmm64u(1) = float64_add(op2->xmm64u(0), op2->xmm64u(1), status); } BX_CPP_INLINE void xmm_haddpd_mask(BxPackedXmmRegister *op1, const BxPackedXmmRegister *op2, float_status_t &status, Bit32u mask) { if (mask & 0x1) op1->xmm64u(0) = float64_add(op1->xmm64u(0), op1->xmm64u(1), status); else op1->xmm64u(0) = 0; if (mask & 0x2) op1->xmm64u(1) = float64_add(op2->xmm64u(0), op2->xmm64u(1), status); else op1->xmm64u(1) = 0; } BX_CPP_INLINE void xmm_hsubps(BxPackedXmmRegister *op1, const BxPackedXmmRegister *op2, float_status_t &status) { op1->xmm32u(0) = float32_sub(op1->xmm32u(0), op1->xmm32u(1), status); op1->xmm32u(1) = float32_sub(op1->xmm32u(2), op1->xmm32u(3), status); op1->xmm32u(2) = float32_sub(op2->xmm32u(0), op2->xmm32u(1), status); op1->xmm32u(3) = float32_sub(op2->xmm32u(2), op2->xmm32u(3), status); } BX_CPP_INLINE void xmm_hsubps_mask(BxPackedXmmRegister *op1, const BxPackedXmmRegister *op2, float_status_t &status, Bit32u mask) { if (mask & 0x1) op1->xmm32u(0) = float32_sub(op1->xmm32u(0), op1->xmm32u(1), status); else op1->xmm32u(0) = 0; if (mask & 0x2) op1->xmm32u(1) = float32_sub(op1->xmm32u(2), op1->xmm32u(3), status); else op1->xmm32u(1) = 0; if (mask & 0x4) op1->xmm32u(2) = float32_sub(op2->xmm32u(0), op2->xmm32u(1), status); else op1->xmm32u(2) = 0; if (mask & 0x8) op1->xmm32u(3) = float32_sub(op2->xmm32u(2), op2->xmm32u(3), status); else op1->xmm32u(3) = 0; } BX_CPP_INLINE void xmm_hsubpd(BxPackedXmmRegister *op1, const BxPackedXmmRegister *op2, float_status_t &status) { op1->xmm64u(0) = float64_sub(op1->xmm64u(0), op1->xmm64u(1), status); op1->xmm64u(1) = float64_sub(op2->xmm64u(0), op2->xmm64u(1), status); } BX_CPP_INLINE void xmm_hsubpd_mask(BxPackedXmmRegister *op1, const BxPackedXmmRegister *op2, float_status_t &status, Bit32u mask) { if (mask & 0x1) op1->xmm64u(0) = float64_sub(op1->xmm64u(0), op1->xmm64u(1), status); else op1->xmm64u(0) = 0; if (mask & 0x2) op1->xmm64u(1) = float64_sub(op2->xmm64u(0), op2->xmm64u(1), status); else op1->xmm64u(1) = 0; } // min/max BX_CPP_INLINE void xmm_minps(BxPackedXmmRegister *op1, const BxPackedXmmRegister *op2, float_status_t &status) { for (unsigned n=0;n<4;n++) { op1->xmm32u(n) = float32_min(op1->xmm32u(n), op2->xmm32u(n), status); } } BX_CPP_INLINE void xmm_minps_mask(BxPackedXmmRegister *op1, const BxPackedXmmRegister *op2, float_status_t &status, Bit32u mask) { for (unsigned n=0; n < 4; n++, mask >>= 1) { if (mask & 0x1) op1->xmm32u(n) = float32_min(op1->xmm32u(n), op2->xmm32u(n), status); else op1->xmm32u(n) = 0; } } BX_CPP_INLINE void xmm_minpd(BxPackedXmmRegister *op1, const BxPackedXmmRegister *op2, float_status_t &status) { for (unsigned n=0;n<2;n++) { op1->xmm64u(n) = float64_min(op1->xmm64u(n), op2->xmm64u(n), status); } } BX_CPP_INLINE void xmm_minpd_mask(BxPackedXmmRegister *op1, const BxPackedXmmRegister *op2, float_status_t &status, Bit32u mask) { for (unsigned n=0; n < 2; n++, mask >>= 1) { if (mask & 0x1) op1->xmm64u(n) = float64_min(op1->xmm64u(n), op2->xmm64u(n), status); else op1->xmm64u(n) = 0; } } BX_CPP_INLINE void xmm_maxps(BxPackedXmmRegister *op1, const BxPackedXmmRegister *op2, float_status_t &status) { for (unsigned n=0;n<4;n++) { op1->xmm32u(n) = float32_max(op1->xmm32u(n), op2->xmm32u(n), status); } } BX_CPP_INLINE void xmm_maxps_mask(BxPackedXmmRegister *op1, const BxPackedXmmRegister *op2, float_status_t &status, Bit32u mask) { for (unsigned n=0; n < 4; n++, mask >>= 1) { if (mask & 0x1) op1->xmm32u(n) = float32_max(op1->xmm32u(n), op2->xmm32u(n), status); else op1->xmm32u(n) = 0; } } BX_CPP_INLINE void xmm_maxpd(BxPackedXmmRegister *op1, const BxPackedXmmRegister *op2, float_status_t &status) { for (unsigned n=0;n<2;n++) { op1->xmm64u(n) = float64_max(op1->xmm64u(n), op2->xmm64u(n), status); } } BX_CPP_INLINE void xmm_maxpd_mask(BxPackedXmmRegister *op1, const BxPackedXmmRegister *op2, float_status_t &status, Bit32u mask) { for (unsigned n=0; n < 2; n++, mask >>= 1) { if (mask & 0x1) op1->xmm64u(n) = float64_max(op1->xmm64u(n), op2->xmm64u(n), status); else op1->xmm64u(n) = 0; } } // fma BX_CPP_INLINE void xmm_fmaddps(BxPackedXmmRegister *op1, const BxPackedXmmRegister *op2, const BxPackedXmmRegister *op3, float_status_t &status) { for (unsigned n=0;n<4;n++) { op1->xmm32u(n) = float32_muladd(op1->xmm32u(n), op2->xmm32u(n), op3->xmm32u(n), 0, status); } } BX_CPP_INLINE void xmm_fmaddps_mask(BxPackedXmmRegister *op1, const BxPackedXmmRegister *op2, const BxPackedXmmRegister *op3, float_status_t &status, Bit32u mask) { for (unsigned n=0; n < 4; n++, mask >>= 1) { if (mask & 0x1) op1->xmm32u(n) = float32_muladd(op1->xmm32u(n), op2->xmm32u(n), op3->xmm32u(n), 0, status); else op1->xmm32u(n) = 0; } } BX_CPP_INLINE void xmm_fmaddpd(BxPackedXmmRegister *op1, const BxPackedXmmRegister *op2, const BxPackedXmmRegister *op3, float_status_t &status) { for (unsigned n=0;n<2;n++) { op1->xmm64u(n) = float64_muladd(op1->xmm64u(n), op2->xmm64u(n), op3->xmm64u(n), 0, status); } } BX_CPP_INLINE void xmm_fmaddpd_mask(BxPackedXmmRegister *op1, const BxPackedXmmRegister *op2, const BxPackedXmmRegister *op3, float_status_t &status, Bit32u mask) { for (unsigned n=0; n < 2; n++, mask >>= 1) { if (mask & 0x1) op1->xmm64u(n) = float64_muladd(op1->xmm64u(n), op2->xmm64u(n), op3->xmm64u(n), 0, status); else op1->xmm64u(n) = 0; } } BX_CPP_INLINE void xmm_fmsubps(BxPackedXmmRegister *op1, const BxPackedXmmRegister *op2, const BxPackedXmmRegister *op3, float_status_t &status) { for (unsigned n=0;n<4;n++) { op1->xmm32u(n) = float32_muladd(op1->xmm32u(n), op2->xmm32u(n), op3->xmm32u(n), float_muladd_negate_c, status); } } BX_CPP_INLINE void xmm_fmsubps_mask(BxPackedXmmRegister *op1, const BxPackedXmmRegister *op2, const BxPackedXmmRegister *op3, float_status_t &status, Bit32u mask) { for (unsigned n=0; n < 4; n++, mask >>= 1) { if (mask & 0x1) op1->xmm32u(n) = float32_muladd(op1->xmm32u(n), op2->xmm32u(n), op3->xmm32u(n), float_muladd_negate_c, status); else op1->xmm32u(n) = 0; } } BX_CPP_INLINE void xmm_fmsubpd(BxPackedXmmRegister *op1, const BxPackedXmmRegister *op2, const BxPackedXmmRegister *op3, float_status_t &status) { for (unsigned n=0;n<2;n++) { op1->xmm64u(n) = float64_muladd(op1->xmm64u(n), op2->xmm64u(n), op3->xmm64u(n), float_muladd_negate_c, status); } } BX_CPP_INLINE void xmm_fmsubpd_mask(BxPackedXmmRegister *op1, const BxPackedXmmRegister *op2, const BxPackedXmmRegister *op3, float_status_t &status, Bit32u mask) { for (unsigned n=0; n < 2; n++, mask >>= 1) { if (mask & 0x1) op1->xmm64u(n) = float64_muladd(op1->xmm64u(n), op2->xmm64u(n), op3->xmm64u(n), float_muladd_negate_c, status); else op1->xmm64u(n) = 0; } } BX_CPP_INLINE void xmm_fmaddsubps(BxPackedXmmRegister *op1, const BxPackedXmmRegister *op2, const BxPackedXmmRegister *op3, float_status_t &status) { op1->xmm32u(0) = float32_muladd(op1->xmm32u(0), op2->xmm32u(0), op3->xmm32u(0), float_muladd_negate_c, status); op1->xmm32u(1) = float32_muladd(op1->xmm32u(1), op2->xmm32u(1), op3->xmm32u(1), 0, status); op1->xmm32u(2) = float32_muladd(op1->xmm32u(2), op2->xmm32u(2), op3->xmm32u(2), float_muladd_negate_c, status); op1->xmm32u(3) = float32_muladd(op1->xmm32u(3), op2->xmm32u(3), op3->xmm32u(3), 0, status); } BX_CPP_INLINE void xmm_fmaddsubps_mask(BxPackedXmmRegister *op1, const BxPackedXmmRegister *op2, const BxPackedXmmRegister *op3, float_status_t &status, Bit32u mask) { if (mask & 0x1) op1->xmm32u(0) = float32_muladd(op1->xmm32u(0), op2->xmm32u(0), op3->xmm32u(0), float_muladd_negate_c, status); else op1->xmm32u(0) = 0; if (mask & 0x2) op1->xmm32u(1) = float32_muladd(op1->xmm32u(1), op2->xmm32u(1), op3->xmm32u(1), 0, status); else op1->xmm32u(1) = 0; if (mask & 0x4) op1->xmm32u(2) = float32_muladd(op1->xmm32u(2), op2->xmm32u(2), op3->xmm32u(2), float_muladd_negate_c, status); else op1->xmm32u(2) = 0; if (mask & 0x8) op1->xmm32u(3) = float32_muladd(op1->xmm32u(3), op2->xmm32u(3), op3->xmm32u(3), 0, status); else op1->xmm32u(3) = 0; } BX_CPP_INLINE void xmm_fmaddsubpd(BxPackedXmmRegister *op1, const BxPackedXmmRegister *op2, const BxPackedXmmRegister *op3, float_status_t &status) { op1->xmm64u(0) = float64_muladd(op1->xmm64u(0), op2->xmm64u(0), op3->xmm64u(0), float_muladd_negate_c, status); op1->xmm64u(1) = float64_muladd(op1->xmm64u(1), op2->xmm64u(1), op3->xmm64u(1), 0, status); } BX_CPP_INLINE void xmm_fmaddsubpd_mask(BxPackedXmmRegister *op1, const BxPackedXmmRegister *op2, const BxPackedXmmRegister *op3, float_status_t &status, Bit32u mask) { if (mask & 0x1) op1->xmm64u(0) = float64_muladd(op1->xmm64u(0), op2->xmm64u(0), op3->xmm64u(0), float_muladd_negate_c, status); else op1->xmm64u(0) = 0; if (mask & 0x2) op1->xmm64u(1) = float64_muladd(op1->xmm64u(1), op2->xmm64u(1), op3->xmm64u(1), 0, status); else op1->xmm64u(1) = 0; } BX_CPP_INLINE void xmm_fmsubaddps(BxPackedXmmRegister *op1, const BxPackedXmmRegister *op2, const BxPackedXmmRegister *op3, float_status_t &status) { op1->xmm32u(0) = float32_muladd(op1->xmm32u(0), op2->xmm32u(0), op3->xmm32u(0), 0, status); op1->xmm32u(1) = float32_muladd(op1->xmm32u(1), op2->xmm32u(1), op3->xmm32u(1), float_muladd_negate_c, status); op1->xmm32u(2) = float32_muladd(op1->xmm32u(2), op2->xmm32u(2), op3->xmm32u(2), 0, status); op1->xmm32u(3) = float32_muladd(op1->xmm32u(3), op2->xmm32u(3), op3->xmm32u(3), float_muladd_negate_c, status); } BX_CPP_INLINE void xmm_fmsubaddps_mask(BxPackedXmmRegister *op1, const BxPackedXmmRegister *op2, const BxPackedXmmRegister *op3, float_status_t &status, Bit32u mask) { if (mask & 0x1) op1->xmm32u(0) = float32_muladd(op1->xmm32u(0), op2->xmm32u(0), op3->xmm32u(0), 0, status); else op1->xmm32u(0) = 0; if (mask & 0x2) op1->xmm32u(1) = float32_muladd(op1->xmm32u(1), op2->xmm32u(1), op3->xmm32u(1), float_muladd_negate_c, status); else op1->xmm32u(1) = 0; if (mask & 0x4) op1->xmm32u(2) = float32_muladd(op1->xmm32u(2), op2->xmm32u(2), op3->xmm32u(2), 0, status); else op1->xmm32u(2) = 0; if (mask & 0x8) op1->xmm32u(3) = float32_muladd(op1->xmm32u(3), op2->xmm32u(3), op3->xmm32u(3), float_muladd_negate_c, status); else op1->xmm32u(3) = 0; } BX_CPP_INLINE void xmm_fmsubaddpd(BxPackedXmmRegister *op1, const BxPackedXmmRegister *op2, const BxPackedXmmRegister *op3, float_status_t &status) { op1->xmm64u(0) = float64_muladd(op1->xmm64u(0), op2->xmm64u(0), op3->xmm64u(0), 0, status); op1->xmm64u(1) = float64_muladd(op1->xmm64u(1), op2->xmm64u(1), op3->xmm64u(1), float_muladd_negate_c, status); } BX_CPP_INLINE void xmm_fmsubaddpd_mask(BxPackedXmmRegister *op1, const BxPackedXmmRegister *op2, const BxPackedXmmRegister *op3, float_status_t &status, Bit32u mask) { if (mask & 0x1) op1->xmm64u(0) = float64_muladd(op1->xmm64u(0), op2->xmm64u(0), op3->xmm64u(0), 0, status); else op1->xmm64u(0) = 0; if (mask & 0x2) op1->xmm64u(1) = float64_muladd(op1->xmm64u(1), op2->xmm64u(1), op3->xmm64u(1), float_muladd_negate_c, status); else op1->xmm64u(1) = 0; } BX_CPP_INLINE void xmm_fnmaddps(BxPackedXmmRegister *op1, const BxPackedXmmRegister *op2, const BxPackedXmmRegister *op3, float_status_t &status) { for (unsigned n=0;n<4;n++) { op1->xmm32u(n) = float32_muladd(op1->xmm32u(n), op2->xmm32u(n), op3->xmm32u(n), float_muladd_negate_product, status); } } BX_CPP_INLINE void xmm_fnmaddps_mask(BxPackedXmmRegister *op1, const BxPackedXmmRegister *op2, const BxPackedXmmRegister *op3, float_status_t &status, Bit32u mask) { for (unsigned n=0; n < 4; n++, mask >>= 1) { if (mask & 0x1) op1->xmm32u(n) = float32_muladd(op1->xmm32u(n), op2->xmm32u(n), op3->xmm32u(n), float_muladd_negate_product, status); else op1->xmm32u(n) = 0; } } BX_CPP_INLINE void xmm_fnmaddpd(BxPackedXmmRegister *op1, const BxPackedXmmRegister *op2, const BxPackedXmmRegister *op3, float_status_t &status) { for (unsigned n=0;n<2;n++) { op1->xmm64u(n) = float64_muladd(op1->xmm64u(n), op2->xmm64u(n), op3->xmm64u(n), float_muladd_negate_product, status); } } BX_CPP_INLINE void xmm_fnmaddpd_mask(BxPackedXmmRegister *op1, const BxPackedXmmRegister *op2, const BxPackedXmmRegister *op3, float_status_t &status, Bit32u mask) { for (unsigned n=0; n < 2; n++, mask >>= 1) { if (mask & 0x1) op1->xmm64u(n) = float64_muladd(op1->xmm64u(n), op2->xmm64u(n), op3->xmm64u(n), float_muladd_negate_product, status); else op1->xmm64u(n) = 0; } } BX_CPP_INLINE void xmm_fnmsubps(BxPackedXmmRegister *op1, const BxPackedXmmRegister *op2, const BxPackedXmmRegister *op3, float_status_t &status) { for (unsigned n=0;n<4;n++) { op1->xmm32u(n) = float32_muladd(op1->xmm32u(n), op2->xmm32u(n), op3->xmm32u(n), float_muladd_negate_result, status); } } BX_CPP_INLINE void xmm_fnmsubps_mask(BxPackedXmmRegister *op1, const BxPackedXmmRegister *op2, const BxPackedXmmRegister *op3, float_status_t &status, Bit32u mask) { for (unsigned n=0; n < 4; n++, mask >>= 1) { if (mask & 0x1) op1->xmm32u(n) = float32_muladd(op1->xmm32u(n), op2->xmm32u(n), op3->xmm32u(n), float_muladd_negate_result, status); else op1->xmm32u(n) = 0; } } BX_CPP_INLINE void xmm_fnmsubpd(BxPackedXmmRegister *op1, const BxPackedXmmRegister *op2, const BxPackedXmmRegister *op3, float_status_t &status) { for (unsigned n=0;n<2;n++) { op1->xmm64u(n) = float64_muladd(op1->xmm64u(n), op2->xmm64u(n), op3->xmm64u(n), float_muladd_negate_result, status); } } BX_CPP_INLINE void xmm_fnmsubpd_mask(BxPackedXmmRegister *op1, const BxPackedXmmRegister *op2, const BxPackedXmmRegister *op3, float_status_t &status, Bit32u mask) { for (unsigned n=0; n < 2; n++, mask >>= 1) { if (mask & 0x1) op1->xmm64u(n) = float64_muladd(op1->xmm64u(n), op2->xmm64u(n), op3->xmm64u(n), float_muladd_negate_result, status); else op1->xmm64u(n) = 0; } } // sqrt BX_CPP_INLINE void xmm_sqrtps(BxPackedXmmRegister *op, float_status_t &status) { for (unsigned n=0; n < 4; n++) { op->xmm32u(n) = float32_sqrt(op->xmm32u(n), status); } } BX_CPP_INLINE void xmm_sqrtps_mask(BxPackedXmmRegister *op, float_status_t &status, Bit32u mask) { for (unsigned n=0; n < 4; n++, mask >>= 1) { if (mask & 0x1) op->xmm32u(n) = float32_sqrt(op->xmm32u(n), status); else op->xmm32u(n) = 0; } } BX_CPP_INLINE void xmm_sqrtpd(BxPackedXmmRegister *op, float_status_t &status) { for (unsigned n=0; n < 2; n++) { op->xmm64u(n) = float64_sqrt(op->xmm64u(n), status); } } BX_CPP_INLINE void xmm_sqrtpd_mask(BxPackedXmmRegister *op, float_status_t &status, Bit32u mask) { for (unsigned n=0; n < 2; n++, mask >>= 1) { if (mask & 0x1) op->xmm64u(n) = float64_sqrt(op->xmm64u(n), status); else op->xmm64u(n) = 0; } } // getexp BX_CPP_INLINE void xmm_getexpps(BxPackedXmmRegister *op, float_status_t &status) { for (unsigned n=0; n < 4; n++) { op->xmm32u(n) = float32_getexp(op->xmm32u(n), status); } } BX_CPP_INLINE void xmm_getexpps_mask(BxPackedXmmRegister *op, float_status_t &status, Bit32u mask) { for (unsigned n=0; n < 4; n++, mask >>= 1) { if (mask & 0x1) op->xmm32u(n) = float32_getexp(op->xmm32u(n), status); else op->xmm32u(n) = 0; } } BX_CPP_INLINE void xmm_getexppd(BxPackedXmmRegister *op, float_status_t &status) { for (unsigned n=0; n < 2; n++) { op->xmm64u(n) = float64_getexp(op->xmm64u(n), status); } } BX_CPP_INLINE void xmm_getexppd_mask(BxPackedXmmRegister *op, float_status_t &status, Bit32u mask) { for (unsigned n=0; n < 2; n++, mask >>= 1) { if (mask & 0x1) op->xmm64u(n) = float64_getexp(op->xmm64u(n), status); else op->xmm64u(n) = 0; } } #endif