/* * Copyright (c) 2013-2018, NVIDIA CORPORATION. All rights reserved. * * Licensed under the Apache License, Version 2.0 (the "License"); * you may not use this file except in compliance with the License. * You may obtain a copy of the License at * * http://www.apache.org/licenses/LICENSE-2.0 * * Unless required by applicable law or agreed to in writing, software * distributed under the License is distributed on an "AS IS" BASIS, * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. * See the License for the specific language governing permissions and * limitations under the License. */ /* * __fvd_cos_vex_256_mask(argument, mask) * __fvd_cos_fma4_256_mask(argument, mask) * * argument: ymm0 * mask: ymm1 * * Compute the cosine of the arguments whose mask is non-zero * */ .text ALN_FUNC .globl ENT(ASM_CONCAT3(__fvd_cos_,TARGET_VEX_OR_FMA,_256_mask)) ENT(ASM_CONCAT3(__fvd_cos_,TARGET_VEX_OR_FMA,_256_mask):) /* RZ_PUSH */ subq $8, %rsp vptest .L_zeromask(%rip), %ymm1 je LBL(.L_fvd_cos_256_done) vandpd %ymm0,%ymm1,%ymm0 CALL(ENT(ASM_CONCAT3(__fvd_cos_,TARGET_VEX_OR_FMA,_256))) LBL(.L_fvd_cos_256_done): /* RZ_POP */ addq $8, %rsp ret ELF_FUNC(ASM_CONCAT3(__fvd_cos_,TARGET_VEX_OR_FMA,_256_mask)) ELF_SIZE(ASM_CONCAT3(__fvd_cos_,TARGET_VEX_OR_FMA,_256_mask)) /* * __fvd_cos_vex_mask(argument, mask) * __fvd_cos_fma4_mask(argument, mask) * * argument: xmm0 * mask: xmm1 * * Compute the cosine of the arguments whose mask is non-zero * */ .text ALN_FUNC .globl ENT(ASM_CONCAT3(__fvd_cos_,TARGET_VEX_OR_FMA,_mask)) ENT(ASM_CONCAT3(__fvd_cos_,TARGET_VEX_OR_FMA,_mask):) subq $8, %rsp vptest .L_zeromask(%rip), %xmm1 je LBL(.L_fvd_cos_done) vandpd %xmm0,%xmm1,%xmm0 CALL(ENT(ASM_CONCAT(__fvd_cos_,TARGET_VEX_OR_FMA))) LBL(.L_fvd_cos_done): addq $8, %rsp ret ELF_FUNC(ASM_CONCAT3(__fvd_cos_,TARGET_VEX_OR_FMA,_mask)) ELF_SIZE(ASM_CONCAT3(__fvd_cos_,TARGET_VEX_OR_FMA,_mask)) /* * __fvs_cos_vex_256_mask(argument, mask) * * argument: ymm0 * mask: ymm1 * * Compute the cosine of the arguments whose mask is non-zero * */ .text ALN_FUNC .globl ENT(ASM_CONCAT3(__fvs_cos_,TARGET_VEX_OR_FMA,_256_mask)) ENT(ASM_CONCAT3(__fvs_cos_,TARGET_VEX_OR_FMA,_256_mask):) subq $8, %rsp vptest .L_s_zeromask(%rip), %ymm1 je LBL(.L_fvs_cos_256_done) vandps %ymm0,%ymm1,%ymm0 CALL(ENT(ASM_CONCAT3(__fvs_cos_,TARGET_VEX_OR_FMA,_256))) LBL(.L_fvs_cos_256_done): addq $8, %rsp ret ELF_FUNC(ASM_CONCAT3(__fvs_cos_,TARGET_VEX_OR_FMA,_256_mask)) ELF_SIZE(ASM_CONCAT3(__fvs_cos_,TARGET_VEX_OR_FMA,_256_mask)) /* * __fvs_cos_vex_mask(argument, mask) * __fvs_cos_fma4_mask(argument, mask) * * argument: xmm0 * mask: xmm1 * * Compute the cosine of the arguments whose mask is non-zero * */ .text ALN_FUNC .globl ENT(ASM_CONCAT3(__fvs_cos_,TARGET_VEX_OR_FMA,_mask)) ENT(ASM_CONCAT3(__fvs_cos_,TARGET_VEX_OR_FMA,_mask):) subq $8, %rsp vptest .L_s_zeromask(%rip), %xmm1 je LBL(.L_fvs_cos_done) vandps %xmm0,%xmm1,%xmm0 CALL(ENT(ASM_CONCAT(__fvs_cos_,TARGET_VEX_OR_FMA))) LBL(.L_fvs_cos_done): addq $8, %rsp ret ELF_FUNC(ASM_CONCAT3(__fvs_cos_,TARGET_VEX_OR_FMA,_mask)) ELF_SIZE(ASM_CONCAT3(__fvs_cos_,TARGET_VEX_OR_FMA,_mask)) /* * __fvd_sin_vex_256_mask(argument, mask) * __fvd_sin_fma4_256_mask(argument, mask) * * argument: ymm0 * mask: ymm1 * * Compute the sine of the arguments whose mask is non-zero * */ .text ALN_FUNC .globl ENT(ASM_CONCAT3(__fvd_sin_,TARGET_VEX_OR_FMA,_256_mask)) ENT(ASM_CONCAT3(__fvd_sin_,TARGET_VEX_OR_FMA,_256_mask):) subq $8, %rsp vptest .L_zeromask(%rip), %ymm1 je LBL(.L_fvd_sin_256_done) vandpd %ymm0,%ymm1,%ymm0 CALL(ENT(ASM_CONCAT3(__fvd_sin_,TARGET_VEX_OR_FMA,_256))) LBL(.L_fvd_sin_256_done): addq $8, %rsp ret ELF_FUNC(ASM_CONCAT3(__fvd_sin_,TARGET_VEX_OR_FMA,_256_mask)) ELF_SIZE(ASM_CONCAT3(__fvd_sin_,TARGET_VEX_OR_FMA,_256_mask)) /* * __fvd_sin_vex_mask(argument, mask) * __fvd_sin_fma4_mask(argument, mask) * * argument: xmm0 * mask: xmm1 * * Compute the sine of the arguments whose mask is non-zero * */ .text ALN_FUNC .globl ENT(ASM_CONCAT3(__fvd_sin_,TARGET_VEX_OR_FMA,_mask)) ENT(ASM_CONCAT3(__fvd_sin_,TARGET_VEX_OR_FMA,_mask):) subq $8, %rsp vptest .L_zeromask(%rip), %xmm1 je LBL(.L_fvd_sin_done) vandpd %xmm0,%xmm1,%xmm0 CALL(ENT(ASM_CONCAT(__fvd_sin_,TARGET_VEX_OR_FMA))) LBL(.L_fvd_sin_done): addq $8, %rsp ret ELF_FUNC(ASM_CONCAT3(__fvd_sin_,TARGET_VEX_OR_FMA,_mask)) ELF_SIZE(ASM_CONCAT3(__fvd_sin_,TARGET_VEX_OR_FMA,_mask)) /* * __fvs_sin_vex_256_mask(argument, mask) * __fvs_sin_fma4_256_mask(argument, mask) * * argument: ymm0 * mask: ymm1 * * Compute the sine of the arguments whose mask is non-zero * */ .text ALN_FUNC .globl ENT(ASM_CONCAT3(__fvs_sin_,TARGET_VEX_OR_FMA,_256_mask)) ENT(ASM_CONCAT3(__fvs_sin_,TARGET_VEX_OR_FMA,_256_mask):) subq $8, %rsp vptest .L_s_zeromask(%rip), %ymm1 je LBL(.L_fvs_sin_256_done) vandps %ymm0,%ymm1,%ymm0 CALL(ENT(ASM_CONCAT3(__fvs_sin_,TARGET_VEX_OR_FMA,_256))) LBL(.L_fvs_sin_256_done): addq $8, %rsp ret ELF_FUNC(ASM_CONCAT3(__fvs_sin_,TARGET_VEX_OR_FMA,_256_mask)) ELF_SIZE(ASM_CONCAT3(__fvs_sin_,TARGET_VEX_OR_FMA,_256_mask)) /* * __fvs_sin_vex_mask(argument, mask) * __fvs_sin_fma4_mask(argument, mask) * * argument: xmm0 * mask: xmm1 * * Compute the sine of the arguments whose mask is non-zero * */ .text ALN_FUNC .globl ENT(ASM_CONCAT3(__fvs_sin_,TARGET_VEX_OR_FMA,_mask)) ENT(ASM_CONCAT3(__fvs_sin_,TARGET_VEX_OR_FMA,_mask):) subq $8, %rsp vptest .L_s_zeromask(%rip), %xmm1 je LBL(.L_fvs_sin_done) vandps %xmm0,%xmm1,%xmm0 CALL(ENT(ASM_CONCAT(__fvs_sin_,TARGET_VEX_OR_FMA))) LBL(.L_fvs_sin_done): addq $8, %rsp ret ELF_FUNC(ASM_CONCAT3(__fvs_sin_,TARGET_VEX_OR_FMA,_mask)) ELF_SIZE(ASM_CONCAT3(__fvs_sin_,TARGET_VEX_OR_FMA,_mask)) /* * __fvd_cosh_vex_256_mask(argument, mask) * __fvd_cosh_fma4_256_mask(argument, mask) * * argument: ymm0 * mask: ymm1 * * Compute the hyperbolic cosine of the arguments whose mask is non-zero * */ .text ALN_FUNC .globl ENT(ASM_CONCAT3(__fvd_cosh_,TARGET_VEX_OR_FMA,_256_mask)) ENT(ASM_CONCAT3(__fvd_cosh_,TARGET_VEX_OR_FMA,_256_mask):) subq $8, %rsp vptest .L_zeromask(%rip), %ymm1 je LBL(.L_fvd_cosh_256_done) vandpd %ymm0,%ymm1,%ymm0 CALL(ENT(ASM_CONCAT3(__fvd_cosh_,TARGET_VEX_OR_FMA,_256))) LBL(.L_fvd_cosh_256_done): addq $8, %rsp ret ELF_FUNC(ASM_CONCAT3(__fvd_cosh_,TARGET_VEX_OR_FMA,_256_mask)) ELF_SIZE(ASM_CONCAT3(__fvd_cosh_,TARGET_VEX_OR_FMA,_256_mask)) /* * __fvd_cosh_vex_mask(argument, mask) * __fvd_cosh_fma4_mask(argument, mask) * * argument: xmm0 * mask: xmm1 * * Compute the hyperbolic cosine of the arguments whose mask is non-zero * */ .text ALN_FUNC .globl ENT(ASM_CONCAT3(__fvd_cosh_,TARGET_VEX_OR_FMA,_mask)) ENT(ASM_CONCAT3(__fvd_cosh_,TARGET_VEX_OR_FMA,_mask):) subq $8, %rsp vptest .L_zeromask(%rip), %xmm1 je LBL(.L_fvd_cosh_done) vandpd %xmm0,%xmm1,%xmm0 CALL(ENT(ASM_CONCAT(__fvd_cosh_,TARGET_VEX_OR_FMA))) LBL(.L_fvd_cosh_done): addq $8, %rsp ret ELF_FUNC(ASM_CONCAT3(__fvd_cosh_,TARGET_VEX_OR_FMA,_mask)) ELF_SIZE(ASM_CONCAT3(__fvd_cosh_,TARGET_VEX_OR_FMA,_mask)) /* * __fvs_cosh_vex_256_mask(argument, mask) * __fvs_cosh_fma4_256_mask(argument, mask) * * argument: ymm0 * mask: ymm1 * * Compute the hyperbolic cosine of the arguments whose mask is non-zero * */ .text ALN_FUNC .globl ENT(ASM_CONCAT3(__fvs_cosh_,TARGET_VEX_OR_FMA,_256_mask)) ENT(ASM_CONCAT3(__fvs_cosh_,TARGET_VEX_OR_FMA,_256_mask):) subq $8, %rsp vptest .L_s_zeromask(%rip), %ymm1 je LBL(.L_fvs_cosh_256_done) vandps %ymm0,%ymm1,%ymm0 CALL(ENT(ASM_CONCAT3(__fvs_cosh_,TARGET_VEX_OR_FMA,_256))) LBL(.L_fvs_cosh_256_done): addq $8, %rsp ret ELF_FUNC(ASM_CONCAT3(__fvs_cosh_,TARGET_VEX_OR_FMA,_256_mask)) ELF_SIZE(ASM_CONCAT3(__fvs_cosh_,TARGET_VEX_OR_FMA,_256_mask)) /* * __fvs_cosh_vex_mask(argument, mask) * __fvs_cosh_fma4_mask(argument, mask) * * argument: xmm0 * mask: xmm1 * * Compute the hyperbolic cosine of the arguments whose mask is non-zero * */ .text ALN_FUNC .globl ENT(ASM_CONCAT3(__fvs_cosh_,TARGET_VEX_OR_FMA,_mask)) ENT(ASM_CONCAT3(__fvs_cosh_,TARGET_VEX_OR_FMA,_mask):) subq $8, %rsp vptest .L_s_zeromask(%rip), %xmm1 je LBL(.L_fvs_cosh_done) vandps %xmm0,%xmm1,%xmm0 CALL(ENT(ASM_CONCAT(__fvs_cosh_,TARGET_VEX_OR_FMA))) LBL(.L_fvs_cosh_done): addq $8, %rsp ret ELF_FUNC(ASM_CONCAT3(__fvs_cosh_,TARGET_VEX_OR_FMA,_mask)) ELF_SIZE(ASM_CONCAT3(__fvs_cosh_,TARGET_VEX_OR_FMA,_mask)) /* * __fvd_sinh_vex_256_mask(argument, mask) * __fvd_sinh_fma4_256_mask(argument, mask) * * argument: ymm0 * mask: ymm1 * * Compute the hypobolic cosine of the arguments whose mask is non-zero * */ .text ALN_FUNC .globl ENT(ASM_CONCAT3(__fvd_sinh_,TARGET_VEX_OR_FMA,_256_mask)) ENT(ASM_CONCAT3(__fvd_sinh_,TARGET_VEX_OR_FMA,_256_mask):) subq $8, %rsp vptest .L_zeromask(%rip), %ymm1 je LBL(.L_fvd_sinh_256_done) vandpd %ymm0,%ymm1,%ymm0 CALL(ENT(ASM_CONCAT3(__fvd_sinh_,TARGET_VEX_OR_FMA,_256))) LBL(.L_fvd_sinh_256_done): addq $8, %rsp ret ELF_FUNC(ASM_CONCAT3(__fvd_sinh_,TARGET_VEX_OR_FMA,_256_mask)) ELF_SIZE(ASM_CONCAT3(__fvd_sinh_,TARGET_VEX_OR_FMA,_256_mask)) /* * __fvd_sinh_vex_mask(argument, mask) * __fvd_sinh_fma4_mask(argument, mask) * * argument: xmm0 * mask: xmm1 * * Compute the hypobolic cosine of the arguments whose mask is non-zero * */ .text ALN_FUNC .globl ENT(ASM_CONCAT3(__fvd_sinh_,TARGET_VEX_OR_FMA,_mask)) ENT(ASM_CONCAT3(__fvd_sinh_,TARGET_VEX_OR_FMA,_mask):) subq $8, %rsp vptest .L_zeromask(%rip), %xmm1 je LBL(.L_fvd_sinh_done) vandpd %xmm0,%xmm1,%xmm0 CALL(ENT(ASM_CONCAT(__fvd_sinh_,TARGET_VEX_OR_FMA))) LBL(.L_fvd_sinh_done): addq $8, %rsp ret ELF_FUNC(ASM_CONCAT3(__fvd_sinh_,TARGET_VEX_OR_FMA,_mask)) ELF_SIZE(ASM_CONCAT3(__fvd_sinh_,TARGET_VEX_OR_FMA,_mask)) /* * __fvs_sinh_vex_256_mask(argument, mask) * __fvs_sinh_fma4_256_mask(argument, mask) * * argument: ymm0 * mask: ymm1 * * Compute the hyperbolic sinine of the arguments whose mask is non-zero * */ .text ALN_FUNC .globl ENT(ASM_CONCAT3(__fvs_sinh_,TARGET_VEX_OR_FMA,_256_mask)) ENT(ASM_CONCAT3(__fvs_sinh_,TARGET_VEX_OR_FMA,_256_mask):) subq $8, %rsp vptest .L_s_zeromask(%rip), %ymm1 je LBL(.L_fvs_sinh_256_done) vandps %ymm0,%ymm1,%ymm0 CALL(ENT(ASM_CONCAT3(__fvs_sinh_,TARGET_VEX_OR_FMA,_256))) LBL(.L_fvs_sinh_256_done): addq $8, %rsp ret ELF_FUNC(ASM_CONCAT3(__fvs_sinh_,TARGET_VEX_OR_FMA,_256_mask)) ELF_SIZE(ASM_CONCAT3(__fvs_sinh_,TARGET_VEX_OR_FMA,_256_mask)) /* * __fvs_sinh_vex_mask(argument, mask) * __fvs_sinh_fma4_mask(argument, mask) * * argument: xmm0 * mask: xmm1 * * Compute the hyperbolic sinine of the arguments whose mask is non-zero * */ .text ALN_FUNC .globl ENT(ASM_CONCAT3(__fvs_sinh_,TARGET_VEX_OR_FMA,_mask)) ENT(ASM_CONCAT3(__fvs_sinh_,TARGET_VEX_OR_FMA,_mask):) subq $8, %rsp vptest .L_s_zeromask(%rip), %xmm1 je LBL(.L_fvs_sinh_done) vandps %xmm0,%xmm1,%xmm0 CALL(ENT(ASM_CONCAT(__fvs_sinh_,TARGET_VEX_OR_FMA))) LBL(.L_fvs_sinh_done): addq $8, %rsp ret ELF_FUNC(ASM_CONCAT3(__fvs_sinh_,TARGET_VEX_OR_FMA,_mask)) ELF_SIZE(ASM_CONCAT3(__fvs_sinh_,TARGET_VEX_OR_FMA,_mask)) /* * __fvs_sincos_vex_256_mask(argument, mask) * __fvs_sincos_fma4_256_mask(argument, mask) * * argument: ymm0 * mask: ymm1 * * Compute the sincos of the arguments whose mask is non-zero * */ .text ALN_FUNC .globl ENT(ASM_CONCAT3(__fvs_sincos_,TARGET_VEX_OR_FMA,_256_mask)) ENT(ASM_CONCAT3(__fvs_sincos_,TARGET_VEX_OR_FMA,_256_mask):) subq $8, %rsp vmovaps %ymm1, %ymm3 vptest .L_s_zeromask(%rip), %ymm1 je LBL(.L_fvs_sincos_256_done) vandps %ymm0,%ymm3,%ymm0 vandps %ymm1,%ymm3,%ymm1 CALL(ENT(ASM_CONCAT3(__fvs_sincos_,TARGET_VEX_OR_FMA,_256))) addq $8, %rsp ret LBL(.L_fvs_sincos_256_done): vmovaps %ymm0, %ymm1 addq $8, %rsp ret ELF_FUNC(ASM_CONCAT3(__fvs_sincos_,TARGET_VEX_OR_FMA,_256_mask)) ELF_SIZE(ASM_CONCAT3(__fvs_sincos_,TARGET_VEX_OR_FMA,_256_mask)) /* * __fvs_sincos_vex_mask(argument, mask) * __fvs_sincos_fma4_mask(argument, mask) * * argument: xmm0 * mask: xmm1 * * Compute the sincos of the arguments whose mask is non-zero * */ .text ALN_FUNC .globl ENT(ASM_CONCAT3(__fvs_sincos_,TARGET_VEX_OR_FMA,_mask)) ENT(ASM_CONCAT3(__fvs_sincos_,TARGET_VEX_OR_FMA,_mask):) subq $8, %rsp vmovaps %xmm1, %xmm3 vptest .L_s_zeromask(%rip), %xmm1 je LBL(.L_fvs_sincos_done) vandps %xmm0,%xmm3,%xmm0 vandps %xmm1,%xmm3,%xmm1 CALL(ENT(ASM_CONCAT(__fvs_sincos_,TARGET_VEX_OR_FMA))) addq $8, %rsp ret LBL(.L_fvs_sincos_done): vmovaps %xmm0, %xmm1 addq $8, %rsp ret ELF_FUNC(ASM_CONCAT3(__fvs_sincos_,TARGET_VEX_OR_FMA,_mask)) ELF_SIZE(ASM_CONCAT3(__fvs_sincos_,TARGET_VEX_OR_FMA,_mask)) /* * __fvd_sincos_vex_256_mask(argument, mask) * __fvd_sincos_fma4_256_mask(argument, mask) * * argument: ymm0 * mask: ymm1 * * Compute the hypobolic cosine of the arguments whose mask is non-zero * */ .text ALN_FUNC .globl ENT(ASM_CONCAT3(__fvd_sincos_,TARGET_VEX_OR_FMA,_256_mask)) ENT(ASM_CONCAT3(__fvd_sincos_,TARGET_VEX_OR_FMA,_256_mask):) subq $8, %rsp vmovapd %ymm1, %ymm3 vptest .L_zeromask(%rip), %ymm1 je LBL(.L_fvd_sincos_256_done) vandpd %ymm0,%ymm3,%ymm0 vandpd %ymm1,%ymm3,%ymm1 CALL(ENT(ASM_CONCAT3(__fvd_sincos_,TARGET_VEX_OR_FMA,_256))) addq $8, %rsp ret LBL(.L_fvd_sincos_256_done): vmovapd %ymm0, %ymm1 addq $8, %rsp ret ELF_FUNC(ASM_CONCAT3(__fvd_sincos_,TARGET_VEX_OR_FMA,_256_mask)) ELF_SIZE(ASM_CONCAT3(__fvd_sincos_,TARGET_VEX_OR_FMA,_256_mask)) /* * __fvd_sincos_vex_mask(argument, mask) * __fvd_sincos_fma4_mask(argument, mask) * * argument: xmm0 * mask: xmm1 * * Compute the hypobolic cosine of the arguments whose mask is non-zero * */ .text ALN_FUNC .globl ENT(ASM_CONCAT3(__fvd_sincos_,TARGET_VEX_OR_FMA,_mask)) ENT(ASM_CONCAT3(__fvd_sincos_,TARGET_VEX_OR_FMA,_mask):) subq $8, %rsp vmovapd %xmm1, %xmm3 vptest .L_zeromask(%rip), %xmm1 je LBL(.L_fvd_sincos_done) vandpd %xmm0,%xmm3,%xmm0 vandpd %xmm1,%xmm3,%xmm1 CALL(ENT(ASM_CONCAT(__fvd_sincos_,TARGET_VEX_OR_FMA))) addq $8, %rsp ret LBL(.L_fvd_sincos_done): vmovapd %xmm0, %xmm1 addq $8, %rsp ret ELF_FUNC(ASM_CONCAT3(__fvd_sincos_,TARGET_VEX_OR_FMA,_mask)) ELF_SIZE(ASM_CONCAT3(__fvd_sincos_,TARGET_VEX_OR_FMA,_mask)) /* * __fvd_exp_vex_256_mask(argument, mask) * __fvd_exp_fma4_256_mask(argument, mask) * * argument: ymm0 * mask: ymm1 * * Compute the exp of the arguments whose mask is non-zero * */ .text ALN_FUNC .globl ENT(ASM_CONCAT3(__fvd_exp_,TARGET_VEX_OR_FMA,_256_mask)) ENT(ASM_CONCAT3(__fvd_exp_,TARGET_VEX_OR_FMA,_256_mask):) subq $8, %rsp vptest .L_zeromask(%rip), %ymm1 je LBL(.L_fvd_exp_256_done) vandpd %ymm0,%ymm1,%ymm0 CALL(ENT(ASM_CONCAT3(__fvd_exp_,TARGET_VEX_OR_FMA,_256))) LBL(.L_fvd_exp_256_done): addq $8, %rsp ret ELF_FUNC(ASM_CONCAT3(__fvd_exp_,TARGET_VEX_OR_FMA,_256_mask)) ELF_SIZE(ASM_CONCAT3(__fvd_exp_,TARGET_VEX_OR_FMA,_256_mask)) /* * __fvd_exp_vex_mask(argument, mask) * __fvd_exp_fma4_mask(argument, mask) * * argument: xmm0 * mask: xmm1 * * Compute the exp of the arguments whose mask is non-zero * */ .text ALN_FUNC .globl ENT(ASM_CONCAT3(__fvd_exp_,TARGET_VEX_OR_FMA,_mask)) ENT(ASM_CONCAT3(__fvd_exp_,TARGET_VEX_OR_FMA,_mask):) subq $8, %rsp vptest .L_zeromask(%rip), %xmm1 je LBL(.L_fvd_exp_done) vandpd %xmm0,%xmm1,%xmm0 CALL(ENT(ASM_CONCAT(__fvd_exp_,TARGET_VEX_OR_FMA))) LBL(.L_fvd_exp_done): addq $8, %rsp ret ELF_FUNC(ASM_CONCAT3(__fvd_exp_,TARGET_VEX_OR_FMA,_mask)) ELF_SIZE(ASM_CONCAT3(__fvd_exp_,TARGET_VEX_OR_FMA,_mask)) /* * __fvs_exp_vex_256_mask(argument, mask) * __fvs_exp_fma4_256_mask(argument, mask) * * argument: ymm0 * mask: ymm1 * * Compute the exp of the arguments whose mask is non-zero * */ .text ALN_FUNC .globl ENT(ASM_CONCAT3(__fvs_exp_,TARGET_VEX_OR_FMA,_256_mask)) ENT(ASM_CONCAT3(__fvs_exp_,TARGET_VEX_OR_FMA,_256_mask):) subq $8, %rsp vptest .L_s_zeromask(%rip), %ymm1 je LBL(.L_fvs_exp_256_done) vandps %ymm0,%ymm1,%ymm0 CALL(ENT(ASM_CONCAT3(__fvs_exp_,TARGET_VEX_OR_FMA,_256))) LBL(.L_fvs_exp_256_done): addq $8, %rsp ret ELF_FUNC(ASM_CONCAT3(__fvs_exp_,TARGET_VEX_OR_FMA,_256_mask)) ELF_SIZE(ASM_CONCAT3(__fvs_exp_,TARGET_VEX_OR_FMA,_256_mask)) /* * __fvs_exp_vex_mask(argument, mask) * __fvs_exp_fma4_mask(argument, mask) * * argument: xmm0 * mask: xmm1 * * Compute the exp of the arguments whose mask is non-zero * */ .text ALN_FUNC .globl ENT(ASM_CONCAT3(__fvs_exp_,TARGET_VEX_OR_FMA,_mask)) ENT(ASM_CONCAT3(__fvs_exp_,TARGET_VEX_OR_FMA,_mask):) subq $8, %rsp vptest .L_s_zeromask(%rip), %xmm1 je LBL(.L_fvs_exp_done) vandps %xmm0,%xmm1,%xmm0 CALL(ENT(ASM_CONCAT(__fvs_exp_,TARGET_VEX_OR_FMA))) LBL(.L_fvs_exp_done): addq $8, %rsp ret ELF_FUNC(ASM_CONCAT3(__fvs_exp_,TARGET_VEX_OR_FMA,_mask)) ELF_SIZE(ASM_CONCAT3(__fvs_exp_,TARGET_VEX_OR_FMA,_mask)) /* * __fvd_pow_vex_256_mask(argument1, argument2, mask) * __fvd_pow_fma4_256_mask(argument1, argument2, mask) * * argument: ymm0, ymm1 * mask: ymm2 * * Compute the power of the arguments whose mask is non-zero * */ .text ALN_FUNC .globl ENT(ASM_CONCAT3(__fvd_pow_,TARGET_VEX_OR_FMA,_256_mask)) ENT(ASM_CONCAT3(__fvd_pow_,TARGET_VEX_OR_FMA,_256_mask):) subq $8, %rsp vptest .L_zeromask(%rip), %ymm2 je LBL(.L_fvd_pow_256_done) vmovupd .L_dpow_mask_two(%rip),%ymm3 vblendvpd %ymm2,%ymm0,%ymm3,%ymm0 vblendvpd %ymm2,%ymm1,%ymm3,%ymm1 CALL(ENT(ASM_CONCAT3(__fvd_pow_,TARGET_VEX_OR_FMA,_256))) LBL(.L_fvd_pow_256_done): addq $8, %rsp ret ELF_FUNC(ASM_CONCAT3(__fvd_pow_,TARGET_VEX_OR_FMA,_256_mask)) ELF_SIZE(ASM_CONCAT3(__fvd_pow_,TARGET_VEX_OR_FMA,_256_mask)) /* * __fvd_pow_vex_mask(argument1, argument2, mask) * __fvd_pow_fma4_mask(argument1, argument2, mask) * * argument: xmm0, xmm1 * mask: xmm2 * * Compute the power of the arguments whose mask is non-zero * */ .text ALN_FUNC .globl ENT(ASM_CONCAT3(__fvd_pow_,TARGET_VEX_OR_FMA,_mask)) ENT(ASM_CONCAT3(__fvd_pow_,TARGET_VEX_OR_FMA,_mask):) subq $8, %rsp vptest .L_zeromask(%rip), %xmm2 je LBL(.L_fvd_pow_done) vmovupd .L_dpow_mask_two(%rip),%xmm3 vblendvpd %xmm2,%xmm0,%xmm3,%xmm0 vblendvpd %xmm2,%xmm1,%xmm3,%xmm1 CALL(ENT(ASM_CONCAT(__fvd_pow_,TARGET_VEX_OR_FMA))) LBL(.L_fvd_pow_done): addq $8, %rsp ret ELF_FUNC(ASM_CONCAT3(__fvd_pow_,TARGET_VEX_OR_FMA,_mask)) ELF_SIZE(ASM_CONCAT3(__fvd_pow_,TARGET_VEX_OR_FMA,_mask)) /* * __fvs_pow_vex_256_mask(argument1, argument2, mask) * __fvs_pow_fma4_256_mask(argument1, argument2, mask) * * argument: ymm0, ymm1 * mask: ymm2 * * Compute the power of the arguments whose mask is non-zero * */ .text ALN_FUNC .globl ENT(ASM_CONCAT3(__fvs_pow_,TARGET_VEX_OR_FMA,_256_mask)) ENT(ASM_CONCAT3(__fvs_pow_,TARGET_VEX_OR_FMA,_256_mask):) subq $8, %rsp vptest .L_s_zeromask(%rip), %ymm2 je LBL(.L_fvs_pow_256_done) vmovups .L_spow_mask_two(%rip),%ymm3 vblendvps %ymm2,%ymm0,%ymm3,%ymm0 vblendvps %ymm2,%ymm1,%ymm3,%ymm1 CALL(ENT(ASM_CONCAT3(__fvs_pow_,TARGET_VEX_OR_FMA,_256))) LBL(.L_fvs_pow_256_done): addq $8, %rsp ret ELF_FUNC(ASM_CONCAT3(__fvs_pow_,TARGET_VEX_OR_FMA,_256_mask)) ELF_SIZE(ASM_CONCAT3(__fvs_pow_,TARGET_VEX_OR_FMA,_256_mask)) /* * __fvs_pow_vex_mask(argument1, argument2, mask) * __fvs_pow_fma4_mask(argument1, argument2, mask) * * argument: xmm0, xmm1 * mask: xmm2 * * Compute the power of the arguments whose mask is non-zero * */ .text ALN_FUNC .globl ENT(ASM_CONCAT3(__fvs_pow_,TARGET_VEX_OR_FMA,_mask)) ENT(ASM_CONCAT3(__fvs_pow_,TARGET_VEX_OR_FMA,_mask):) subq $8, %rsp vptest .L_s_zeromask(%rip), %xmm2 je LBL(.L_fvs_pow_done) vmovups .L_spow_mask_two(%rip),%xmm3 vblendvps %xmm2,%xmm0,%xmm3,%xmm0 vblendvps %xmm2,%xmm1,%xmm3,%xmm1 CALL(ENT(ASM_CONCAT(__fvs_pow_,TARGET_VEX_OR_FMA))) LBL(.L_fvs_pow_done): addq $8, %rsp ret ELF_FUNC(ASM_CONCAT3(__fvs_pow_,TARGET_VEX_OR_FMA,_mask)) ELF_SIZE(ASM_CONCAT3(__fvs_pow_,TARGET_VEX_OR_FMA,_mask)) /* * __fvs_sqrt_fma4_256_mask(argument, mask) * __fvs_sqrt_vex_256_mask(argument, mask) * * argument: ymm0 * mask: ymm1 * * Compute the square root of the arguments whose mask is non-zero * */ .text ALN_FUNC .globl ENT(ASM_CONCAT3(__fvs_sqrt_,TARGET_VEX_OR_FMA,_256_mask)) ENT(ASM_CONCAT3(__fvs_sqrt_,TARGET_VEX_OR_FMA,_256_mask):) subq $8, %rsp vptest .L_zeromask(%rip), %ymm1 je LBL(.L_done_fvs_sqrt_256) vandps %ymm0,%ymm1,%ymm0 vsqrtps %ymm0,%ymm0 LBL(.L_done_fvs_sqrt_256): addq $8, %rsp ret ELF_FUNC(ASM_CONCAT3(__fvs_sqrt_,TARGET_VEX_OR_FMA,_256_mask)) ELF_SIZE(ASM_CONCAT3(__fvs_sqrt_,TARGET_VEX_OR_FMA,_256_mask)) /* * __fvs_sqrt_fma4_mask(argument, mask) * __fvs_sqrt_vex_mask(argument, mask) * * argument: xmm0 * mask: xmm1 * * Compute the square root of the arguments whose mask is non-zero * */ .text ALN_FUNC .globl ENT(ASM_CONCAT3(__fvs_sqrt_,TARGET_VEX_OR_FMA,_mask)) ENT(ASM_CONCAT3(__fvs_sqrt_,TARGET_VEX_OR_FMA,_mask):) subq $8, %rsp vptest .L_zeromask(%rip), %xmm1 je LBL(.L_done_fvs_sqrt) vandps %xmm0,%xmm1,%xmm0 vsqrtps %xmm0,%xmm0 LBL(.L_done_fvs_sqrt): addq $8, %rsp ret ELF_FUNC(ASM_CONCAT3(__fvs_sqrt_,TARGET_VEX_OR_FMA,_mask)) ELF_SIZE(ASM_CONCAT3(__fvs_sqrt_,TARGET_VEX_OR_FMA,_mask)) /* * __fvd_sqrt_fma4_256_mask(argument, mask) * __fvd_sqrt_vex_256_mask(argument, mask) * * argument: ymm0 * mask: ymm1 * * Compute the square root of the arguments whose mask is non-zero * */ .text ALN_FUNC .globl ENT(ASM_CONCAT3(__fvd_sqrt_,TARGET_VEX_OR_FMA,_256_mask)) ENT(ASM_CONCAT3(__fvd_sqrt_,TARGET_VEX_OR_FMA,_256_mask):) subq $8, %rsp vptest .L_zeromask(%rip), %ymm1 je LBL(.L_done_fvd_sqrt_256) vandpd %ymm0,%ymm1,%ymm0 vsqrtpd %ymm0,%ymm0 LBL(.L_done_fvd_sqrt_256): addq $8, %rsp ret ELF_FUNC(ASM_CONCAT3(__fvd_sqrt_,TARGET_VEX_OR_FMA,_256_mask)) ELF_SIZE(ASM_CONCAT3(__fvd_sqrt_,TARGET_VEX_OR_FMA,_256_mask)) /* * __fvd_sqrt_fma4_mask(argument, mask) * __fvd_sqrt_vex_mask(argument, mask) * * argument: xmm0 * mask: xmm1 * * Compute the square root of the arguments whose mask is non-zero * */ .text ALN_FUNC .globl ENT(ASM_CONCAT3(__fvd_sqrt_,TARGET_VEX_OR_FMA,_mask)) ENT(ASM_CONCAT3(__fvd_sqrt_,TARGET_VEX_OR_FMA,_mask):) subq $8, %rsp vptest .L_zeromask(%rip), %xmm1 je LBL(.L_done_fvd_sqrt) vandpd %xmm0,%xmm1,%xmm0 vsqrtpd %xmm0,%xmm0 LBL(.L_done_fvd_sqrt): addq $8, %rsp ret ELF_FUNC(ASM_CONCAT3(__fvd_sqrt_,TARGET_VEX_OR_FMA,_mask)) ELF_SIZE(ASM_CONCAT3(__fvd_sqrt_,TARGET_VEX_OR_FMA,_mask)) /* * __fvs_div_fma4_256_mask(argument1, argument2, mask) * __fvs_div_vex_256_mask(argument1, argument2, mask) * * argument1(dividend): ymm0 * argument2(divisor): ymm1 * mask: ymm2 * * Compute argument1 / argument2 whose mask is non-zero * */ .text ALN_FUNC .globl ENT(ASM_CONCAT3(__fvs_div_,TARGET_VEX_OR_FMA,_256_mask)) ENT(ASM_CONCAT3(__fvs_div_,TARGET_VEX_OR_FMA,_256_mask):) subq $8, %rsp vptest .L_zeromask(%rip), %ymm2 je LBL(.L_done_fvs_div_256) vmovups .L_one_for_mask_fvs(%rip), %ymm3 vblendvps %ymm2,%ymm0,%ymm3,%ymm0 vblendvps %ymm2,%ymm1,%ymm3,%ymm1 vdivps %ymm1,%ymm0,%ymm0 LBL(.L_done_fvs_div_256): addq $8, %rsp ret ELF_FUNC(ASM_CONCAT3(__fvs_div_,TARGET_VEX_OR_FMA,_256_mask)) ELF_SIZE(ASM_CONCAT3(__fvs_div_,TARGET_VEX_OR_FMA,_256_mask)) /* * __fvs_div_fma4_mask(argument1, argument2, mask) * __fvs_div_vex_mask(argument1, argument2, mask) * * argument1(dividend): xmm0 * argument2(divisor): xmm1 * mask: xmm2 * * Compute argument1 / argument2 whose mask is non-zero * */ .text ALN_FUNC .globl ENT(ASM_CONCAT3(__fvs_div_,TARGET_VEX_OR_FMA,_mask)) ENT(ASM_CONCAT3(__fvs_div_,TARGET_VEX_OR_FMA,_mask):) subq $8, %rsp vptest .L_zeromask(%rip), %xmm2 je LBL(.L_done_fvs_div) vmovups .L_one_for_mask_fvs(%rip), %xmm3 vblendvps %xmm2,%xmm0,%xmm3,%xmm0 vblendvps %xmm2,%xmm1,%xmm3,%xmm1 vdivps %xmm1,%xmm0,%xmm0 LBL(.L_done_fvs_div): addq $8, %rsp ret ELF_FUNC(ASM_CONCAT3(__fvs_div_,TARGET_VEX_OR_FMA,_mask)) ELF_SIZE(ASM_CONCAT3(__fvs_div_,TARGET_VEX_OR_FMA,_mask)) /* * __fvd_div_fma4_256_mask(argument1, argument2, mask) * __fvd_div_vex_256_mask(argument1, argument2, mask) * * argument1(dividend): ymm0 * argument2(divisor): ymm1 * mask: ymm2 * * Compute argument1 / argument2 whose mask is non-zero * */ .text ALN_FUNC .globl ENT(ASM_CONCAT3(__fvd_div_,TARGET_VEX_OR_FMA,_256_mask)) ENT(ASM_CONCAT3(__fvd_div_,TARGET_VEX_OR_FMA,_256_mask):) subq $8, %rsp vptest .L_zeromask(%rip), %ymm2 je LBL(.L_done_fvd_div_256) vmovupd .L_one_for_mask_fvd(%rip), %ymm3 vblendvpd %ymm2,%ymm0,%ymm3,%ymm0 vblendvpd %ymm2,%ymm1,%ymm3,%ymm1 vdivpd %ymm1,%ymm0,%ymm0 LBL(.L_done_fvd_div_256): addq $8, %rsp ret ELF_FUNC(ASM_CONCAT3(__fvd_div_,TARGET_VEX_OR_FMA,_256_mask)) ELF_SIZE(ASM_CONCAT3(__fvd_div_,TARGET_VEX_OR_FMA,_256_mask)) /* * __fvd_div_fma4_mask(argument1, argument2, mask) * __fvd_div_vex_mask(argument1, argument2, mask) * * argument1(dividend): xmm0 * argument2(divisor): xmm1 * mask: xmm2 * * Compute argument1 / argument2 whose mask is non-zero * */ .text ALN_FUNC .globl ENT(ASM_CONCAT3(__fvd_div_,TARGET_VEX_OR_FMA,_mask)) ENT(ASM_CONCAT3(__fvd_div_,TARGET_VEX_OR_FMA,_mask):) subq $8, %rsp vptest .L_zeromask(%rip), %xmm2 je LBL(.L_done_fvd_div) vmovupd .L_one_for_mask_fvd(%rip), %xmm3 vblendvpd %xmm2,%xmm0,%xmm3,%xmm0 vblendvpd %xmm2,%xmm1,%xmm3,%xmm1 vdivpd %xmm1,%xmm0,%xmm0 LBL(.L_done_fvd_div): addq $8, %rsp ret ELF_FUNC(ASM_CONCAT3(__fvd_div_,TARGET_VEX_OR_FMA,_mask)) ELF_SIZE(ASM_CONCAT3(__fvd_div_,TARGET_VEX_OR_FMA,_mask)) /* * __fvs_log_fma4_256_mask(argument, mask) * __fvs_log_vex_256_mask(argument, mask) * * argument: ymm0 * mask: ymm1 * * Compute the logarithm of the arguments whose mask is non-zero * */ .text ALN_FUNC .globl ENT(ASM_CONCAT3(__fvs_log_,TARGET_VEX_OR_FMA,_256_mask)) ENT(ASM_CONCAT3(__fvs_log_,TARGET_VEX_OR_FMA,_256_mask):) subq $8, %rsp vptest .L_zeromask(%rip), %ymm1 je LBL(.L_done_fvs_log_256) vmovups .L_one_for_mask_fvs(%rip), %ymm2 vblendvps %ymm1,%ymm0,%ymm2,%ymm0 CALL(ENT(ASM_CONCAT3(__fvs_log_,TARGET_VEX_OR_FMA,_256))) LBL(.L_done_fvs_log_256): addq $8, %rsp ret ELF_FUNC(ASM_CONCAT3(__fvs_log_,TARGET_VEX_OR_FMA,_256_mask)) ELF_SIZE(ASM_CONCAT3(__fvs_log_,TARGET_VEX_OR_FMA,_256_mask)) /* * __fvs_log_fma4_mask(argument, mask) * __fvs_log_vex_mask(argument, mask) * * argument: xmm0 * mask: xmm1 * * Compute the logarithm of the arguments whose mask is non-zero * */ .text ALN_FUNC .globl ENT(ASM_CONCAT3(__fvs_log_,TARGET_VEX_OR_FMA,_mask)) ENT(ASM_CONCAT3(__fvs_log_,TARGET_VEX_OR_FMA,_mask):) subq $8, %rsp vptest .L_zeromask(%rip), %xmm1 je LBL(.L_done_fvs_log) vmovups .L_one_for_mask_fvs(%rip), %xmm2 vblendvps %xmm1,%xmm0,%xmm2,%xmm0 CALL(ENT(ASM_CONCAT(__fvs_log_,TARGET_VEX_OR_FMA))) LBL(.L_done_fvs_log): addq $8, %rsp ret ELF_FUNC(ASM_CONCAT3(__fvs_log_,TARGET_VEX_OR_FMA,_mask)) ELF_SIZE(ASM_CONCAT3(__fvs_log_,TARGET_VEX_OR_FMA,_mask)) /* * __fvd_log_fma4_256_mask(argument, mask) * __fvd_log_vex_256_mask(argument, mask) * * argument: ymm0 * mask: ymm1 * * Compute the logarithm of the arguments whose mask is non-zero * */ .text ALN_FUNC .globl ENT(ASM_CONCAT3(__fvd_log_,TARGET_VEX_OR_FMA,_256_mask)) ENT(ASM_CONCAT3(__fvd_log_,TARGET_VEX_OR_FMA,_256_mask):) subq $8, %rsp vptest .L_zeromask(%rip), %ymm1 je LBL(.L_done_fvd_log_256) vmovupd .L_one_for_mask_fvd(%rip), %ymm2 vblendvpd %ymm1,%ymm0,%ymm2,%ymm0 CALL(ENT(ASM_CONCAT3(__fvd_log_,TARGET_VEX_OR_FMA,_256))) LBL(.L_done_fvd_log_256): addq $8, %rsp ret ELF_FUNC(ASM_CONCAT3(__fvd_log_,TARGET_VEX_OR_FMA,_256_mask)) ELF_SIZE(ASM_CONCAT3(__fvd_log_,TARGET_VEX_OR_FMA,_256_mask)) /* * __fvd_log_fma4_mask(argument, mask) * __fvd_log_vex_mask(argument, mask) * * argument: xmm0 * mask: xmm1 * * Compute the logarithm of the arguments whose mask is non-zero * */ .text ALN_FUNC .globl ENT(ASM_CONCAT3(__fvd_log_,TARGET_VEX_OR_FMA,_mask)) ENT(ASM_CONCAT3(__fvd_log_,TARGET_VEX_OR_FMA,_mask):) subq $8, %rsp vptest .L_zeromask(%rip), %xmm1 je LBL(.L_done_fvd_log) vmovupd .L_one_for_mask_fvd(%rip), %xmm2 vblendvpd %xmm1,%xmm0,%xmm2,%xmm0 CALL(ENT(ASM_CONCAT(__fvd_log_,TARGET_VEX_OR_FMA))) LBL(.L_done_fvd_log): addq $8, %rsp ret ELF_FUNC(ASM_CONCAT3(__fvd_log_,TARGET_VEX_OR_FMA,_mask)) ELF_SIZE(ASM_CONCAT3(__fvd_log_,TARGET_VEX_OR_FMA,_mask)) /* * __fvs_log10_fma4_256_mask(argument, mask) * __fvs_log10_vex_256_mask(argument, mask) * * argument: ymm0 * mask: ymm1 * * Compute the logarithm(base 10) of the arguments whose mask is non-zero * */ .text ALN_FUNC .globl ENT(ASM_CONCAT3(__fvs_log10_,TARGET_VEX_OR_FMA,_256_mask)) ENT(ASM_CONCAT3(__fvs_log10_,TARGET_VEX_OR_FMA,_256_mask):) subq $8, %rsp vptest .L_zeromask(%rip), %ymm1 je LBL(.L_done_fvs_log10_256) vmovups .L_one_for_mask_fvs(%rip), %ymm2 vblendvps %ymm1,%ymm0,%ymm2,%ymm0 CALL(ENT(ASM_CONCAT3(__fvs_log10_,TARGET_VEX_OR_FMA,_256))) LBL(.L_done_fvs_log10_256): addq $8, %rsp ret ELF_FUNC(ASM_CONCAT3(__fvs_log10_,TARGET_VEX_OR_FMA,_256_mask)) ELF_SIZE(ASM_CONCAT3(__fvs_log10_,TARGET_VEX_OR_FMA,_256_mask)) /* * __fvs_log10_fma4_mask(argument, mask) * __fvs_log10_vex_mask(argument, mask) * * argument: xmm0 * mask: xmm1 * * Compute the logarithm(base 10) of the arguments whose mask is non-zero * */ .text ALN_FUNC .globl ENT(ASM_CONCAT3(__fvs_log10_,TARGET_VEX_OR_FMA,_mask)) ENT(ASM_CONCAT3(__fvs_log10_,TARGET_VEX_OR_FMA,_mask):) subq $8, %rsp vptest .L_zeromask(%rip), %xmm1 je LBL(.L_done_fvs_log10) vmovups .L_one_for_mask_fvs(%rip), %xmm2 vblendvps %xmm1,%xmm0,%xmm2,%xmm0 CALL(ENT(ASM_CONCAT(__fvs_log10_,TARGET_VEX_OR_FMA))) LBL(.L_done_fvs_log10): addq $8, %rsp ret ELF_FUNC(ASM_CONCAT3(__fvs_log10_,TARGET_VEX_OR_FMA,_mask)) ELF_SIZE(ASM_CONCAT3(__fvs_log10_,TARGET_VEX_OR_FMA,_mask)) /* * __fvd_log10_fma4_256_mask(argument, mask) * __fvd_log10_vex_256_mask(argument, mask) * * argument: ymm0 * mask: ymm1 * * Compute the logarithm(base 10) of the arguments whose mask is non-zero * */ .text ALN_FUNC .globl ENT(ASM_CONCAT3(__fvd_log10_,TARGET_VEX_OR_FMA,_256_mask)) ENT(ASM_CONCAT3(__fvd_log10_,TARGET_VEX_OR_FMA,_256_mask):) subq $8, %rsp vptest .L_zeromask(%rip), %ymm1 je LBL(.L_done_fvd_log10_256) vmovupd .L_one_for_mask_fvd(%rip), %ymm2 vblendvpd %ymm1,%ymm0,%ymm2,%ymm0 CALL(ENT(ASM_CONCAT3(__fvd_log10_,TARGET_VEX_OR_FMA,_256))) LBL(.L_done_fvd_log10_256): addq $8, %rsp ret ELF_FUNC(ASM_CONCAT3(__fvd_log10_,TARGET_VEX_OR_FMA,_256_mask)) ELF_SIZE(ASM_CONCAT3(__fvd_log10_,TARGET_VEX_OR_FMA,_256_mask)) /* * __fvd_log10_fma4_mask(argument, mask) * __fvd_log10_vex_mask(argument, mask) * * argument: xmm0 * mask: xmm1 * * Compute the logarithm(base 10) of the arguments whose mask is non-zero * */ .text ALN_FUNC .globl ENT(ASM_CONCAT3(__fvd_log10_,TARGET_VEX_OR_FMA,_mask)) ENT(ASM_CONCAT3(__fvd_log10_,TARGET_VEX_OR_FMA,_mask):) subq $8, %rsp vptest .L_zeromask(%rip), %xmm1 je LBL(.L_done_fvd_log10) vmovupd .L_one_for_mask_fvd(%rip), %xmm2 vblendvpd %xmm1,%xmm0,%xmm2,%xmm0 CALL(ENT(ASM_CONCAT(__fvd_log10_,TARGET_VEX_OR_FMA))) LBL(.L_done_fvd_log10): addq $8, %rsp ret ELF_FUNC(ASM_CONCAT3(__fvd_log10_,TARGET_VEX_OR_FMA,_mask)) ELF_SIZE(ASM_CONCAT3(__fvd_log10_,TARGET_VEX_OR_FMA,_mask)) /* * __fvd_tan_vex_256_mask(argument, mask) * __fvd_tan_fma4_256_mask(argument, mask) * * argument: ymm0 * mask: ymm1 * * Compute the tangent of the arguments whose mask is non-zero * */ .text ALN_FUNC .globl ENT(ASM_CONCAT3(__fvd_tan_,TARGET_VEX_OR_FMA,_256_mask)) ENT(ASM_CONCAT3(__fvd_tan_,TARGET_VEX_OR_FMA,_256_mask):) subq $8, %rsp vptest .L_zeromask(%rip), %ymm1 je LBL(.L_fvd_tan_256_done) vandpd %ymm0,%ymm1,%ymm0 CALL(ENT(ASM_CONCAT3(__fvd_tan_,TARGET_VEX_OR_FMA,_256))) LBL(.L_fvd_tan_256_done): addq $8, %rsp ret ELF_FUNC(ASM_CONCAT3(__fvd_tan_,TARGET_VEX_OR_FMA,_256_mask)) ELF_SIZE(ASM_CONCAT3(__fvd_tan_,TARGET_VEX_OR_FMA,_256_mask)) /* * __fvd_tan_vex_mask(argument, mask) * __fvd_tan_fma4_mask(argument, mask) * * argument: xmm0 * mask: xmm1 * * Compute the tangent of the arguments whose mask is non-zero * */ .text ALN_FUNC .globl ENT(ASM_CONCAT3(__fvd_tan_,TARGET_VEX_OR_FMA,_mask)) ENT(ASM_CONCAT3(__fvd_tan_,TARGET_VEX_OR_FMA,_mask):) subq $8, %rsp vptest .L_zeromask(%rip), %xmm1 je LBL(.L_fvd_tan_done) vandpd %xmm0,%xmm1,%xmm0 CALL(ENT(ASM_CONCAT(__fvd_tan_,TARGET_VEX_OR_FMA))) LBL(.L_fvd_tan_done): addq $8, %rsp ret ELF_FUNC(ASM_CONCAT3(__fvd_tan_,TARGET_VEX_OR_FMA,_mask)) ELF_SIZE(ASM_CONCAT3(__fvd_tan_,TARGET_VEX_OR_FMA,_mask)) /* * __fvs_tan_vex_256_mask(argument, mask) * * argument: ymm0 * mask: ymm1 * * Compute the tangent of the arguments whose mask is non-zero * */ .text ALN_FUNC .globl ENT(ASM_CONCAT3(__fvs_tan_,TARGET_VEX_OR_FMA,_256_mask)) ENT(ASM_CONCAT3(__fvs_tan_,TARGET_VEX_OR_FMA,_256_mask):) subq $8, %rsp vptest .L_s_zeromask(%rip), %ymm1 je LBL(.L_fvs_tan_256_done) vandps %ymm0,%ymm1,%ymm0 CALL(ENT(ASM_CONCAT3(__fvs_tan_,TARGET_VEX_OR_FMA,_256))) LBL(.L_fvs_tan_256_done): addq $8, %rsp ret ELF_FUNC(ASM_CONCAT3(__fvs_tan_,TARGET_VEX_OR_FMA,_256_mask)) ELF_SIZE(ASM_CONCAT3(__fvs_tan_,TARGET_VEX_OR_FMA,_256_mask)) /* * __fvs_tan_vex_mask(argument, mask) * __fvs_tan_fma4_mask(argument, mask) * * argument: xmm0 * mask: xmm1 * * Compute the tangent of the arguments whose mask is non-zero * */ .text ALN_FUNC .globl ENT(ASM_CONCAT3(__fvs_tan_,TARGET_VEX_OR_FMA,_mask)) ENT(ASM_CONCAT3(__fvs_tan_,TARGET_VEX_OR_FMA,_mask):) subq $8, %rsp vptest .L_s_zeromask(%rip), %xmm1 je LBL(.L_fvs_tan_done) vandps %xmm0,%xmm1,%xmm0 CALL(ENT(ASM_CONCAT(__fvs_tan_,TARGET_VEX_OR_FMA))) LBL(.L_fvs_tan_done): addq $8, %rsp ret ELF_FUNC(ASM_CONCAT3(__fvs_tan_,TARGET_VEX_OR_FMA,_mask)) ELF_SIZE(ASM_CONCAT3(__fvs_tan_,TARGET_VEX_OR_FMA,_mask))