nb_kernel234_ia32_sse2.intel_syntax.s
来自「最著名最快的分子模拟软件」· S 代码 · 共 2,167 行 · 第 1/5 页
S
2,167 行
mulpd xmm5, xmm6 ;# luB*(3-rsqB*luB*luB) mulpd xmm1, [esp + nb234_half] ;# rinv mulpd xmm5, [esp + nb234_half] ;# rinv movapd [esp + nb234_rinvMH1], xmm1 movapd [esp + nb234_rinvH2M], xmm5 ;# start with OO interaction movapd xmm0, [esp + nb234_rinvOO] movapd xmm4, [esp + nb234_rsqOO] mulpd xmm4, xmm0 ;# xmm4=r mulpd xmm4, [esp + nb234_tsc] cvttpd2pi mm6, xmm4 ;# mm6 = lu idx cvtpi2pd xmm5, mm6 subpd xmm4, xmm5 movapd xmm1, xmm4 ;# xmm1=eps movapd xmm2, xmm1 mulpd xmm2, xmm2 ;# xmm2=eps2 pslld mm6, 3 ;# idx *= 8 movd mm0, eax movd mm1, ebx mov esi, [ebp + nb234_VFtab] movd eax, mm6 psrlq mm6, 32 movd ebx, mm6 ;# dispersion movlpd xmm4, [esi + eax*8] ;# Y1 movlpd xmm3, [esi + ebx*8] ;# Y2 movhpd xmm4, [esi + eax*8 + 8] ;# Y1 F1 movhpd xmm3, [esi + ebx*8 + 8] ;# Y2 F2 movapd xmm5, xmm4 unpcklpd xmm4, xmm3 ;# Y1 Y2 unpckhpd xmm5, xmm3 ;# F1 F2 movlpd xmm6, [esi + eax*8 + 16] ;# G1 movlpd xmm3, [esi + ebx*8 + 16] ;# G2 movhpd xmm6, [esi + eax*8 + 24] ;# G1 H1 movhpd xmm3, [esi + ebx*8 + 24] ;# G2 H2 movapd xmm7, xmm6 unpcklpd xmm6, xmm3 ;# G1 G2 unpckhpd xmm7, xmm3 ;# H1 H2 ;# dispersion table ready, in xmm4-xmm7 mulpd xmm6, xmm1 ;# xmm6=Geps mulpd xmm7, xmm2 ;# xmm7=Heps2 addpd xmm5, xmm6 addpd xmm5, xmm7 ;# xmm5=Fp mulpd xmm7, [esp + nb234_two] ;# two*Heps2 addpd xmm7, xmm6 addpd xmm7, xmm5 ;# xmm7=FF mulpd xmm5, xmm1 ;# xmm5=eps*Fp addpd xmm5, xmm4 ;# xmm5=VV movapd xmm4, [esp + nb234_c6] mulpd xmm7, xmm4 ;# fijD mulpd xmm5, xmm4 ;# Vvdw6 ;# put scalar force on stack Update Vvdwtot directly addpd xmm5, [esp + nb234_Vvdwtot] xorpd xmm3, xmm3 mulpd xmm7, [esp + nb234_tsc] subpd xmm3, xmm7 movapd [esp + nb234_fstmp], xmm3 movapd [esp + nb234_Vvdwtot], xmm5 ;# repulsion movlpd xmm4, [esi + eax*8 + 32] ;# Y1 movlpd xmm3, [esi + ebx*8 + 32] ;# Y2 movhpd xmm4, [esi + eax*8 + 40] ;# Y1 F1 movhpd xmm3, [esi + ebx*8 + 40] ;# Y2 F2 movapd xmm5, xmm4 unpcklpd xmm4, xmm3 ;# Y1 Y2 unpckhpd xmm5, xmm3 ;# F1 F2 movlpd xmm6, [esi + eax*8 + 48] ;# G1 movlpd xmm3, [esi + ebx*8 + 48] ;# G2 movhpd xmm6, [esi + eax*8 + 56] ;# G1 H1 movhpd xmm3, [esi + ebx*8 + 56] ;# G2 H2 movapd xmm7, xmm6 unpcklpd xmm6, xmm3 ;# G1 G2 unpckhpd xmm7, xmm3 ;# H1 H2 ;# table ready, in xmm4-xmm7 mulpd xmm6, xmm1 ;# xmm6=Geps mulpd xmm7, xmm2 ;# xmm7=Heps2 addpd xmm5, xmm6 addpd xmm5, xmm7 ;# xmm5=Fp mulpd xmm7, [esp + nb234_two] ;# two*Heps2 addpd xmm7, xmm6 addpd xmm7, xmm5 ;# xmm7=FF mulpd xmm5, xmm1 ;# xmm5=eps*Fp addpd xmm5, xmm4 ;# xmm5=VV movapd xmm4, [esp + nb234_c12] mulpd xmm7, xmm4 mulpd xmm5, xmm4 addpd xmm5, [esp + nb234_Vvdwtot] movapd xmm3, [esp + nb234_fstmp] mulpd xmm7, [esp + nb234_tsc] subpd xmm3, xmm7 movapd [esp + nb234_Vvdwtot], xmm5 mulpd xmm0, xmm3 movapd xmm1, xmm0 movapd xmm2, xmm0 movd eax, mm0 movd ebx, mm1 xorpd xmm3, xmm3 movapd xmm4, xmm3 movapd xmm5, xmm3 mulpd xmm0, [esp + nb234_dxOO] mulpd xmm1, [esp + nb234_dyOO] mulpd xmm2, [esp + nb234_dzOO] subpd xmm3, xmm0 subpd xmm4, xmm1 subpd xmm5, xmm2 addpd xmm0, [esp + nb234_fixO] addpd xmm1, [esp + nb234_fiyO] addpd xmm2, [esp + nb234_fizO] movapd [esp + nb234_fjxO], xmm3 movapd [esp + nb234_fjyO], xmm4 movapd [esp + nb234_fjzO], xmm5 movapd [esp + nb234_fixO], xmm0 movapd [esp + nb234_fiyO], xmm1 movapd [esp + nb234_fizO], xmm2 ;# H1-H1 interaction movapd xmm0, [esp + nb234_rinvH1H1] movapd xmm7, xmm0 ;# xmm7=rinv movapd xmm5, [esp + nb234_krf] mulpd xmm0, xmm0 ;# xmm0=rinvsq mulpd xmm5, [esp + nb234_rsqH1H1] ;# xmm5=krsq movapd xmm6, xmm5 addpd xmm6, xmm7 ;# xmm6=rinv+ krsq subpd xmm6, [esp + nb234_crf] mulpd xmm6, [esp + nb234_qqHH] ;# xmm6=voul=qq*(rinv+ krsq-crf) mulpd xmm5, [esp + nb234_two] subpd xmm7, xmm5 ;# xmm7=rinv-2*krsq mulpd xmm7, [esp + nb234_qqHH] ;# xmm7 = coul part of fscal addpd xmm6, [esp + nb234_vctot] ;# local vctot summation variable mulpd xmm0, xmm7 movapd [esp + nb234_vctot], xmm6 movapd xmm1, xmm0 movapd xmm2, xmm0 xorpd xmm3, xmm3 movapd xmm4, xmm3 movapd xmm5, xmm3 mulpd xmm0, [esp + nb234_dxH1H1] mulpd xmm1, [esp + nb234_dyH1H1] mulpd xmm2, [esp + nb234_dzH1H1] subpd xmm3, xmm0 subpd xmm4, xmm1 subpd xmm5, xmm2 addpd xmm0, [esp + nb234_fixH1] addpd xmm1, [esp + nb234_fiyH1] addpd xmm2, [esp + nb234_fizH1] movapd [esp + nb234_fjxH1], xmm3 movapd [esp + nb234_fjyH1], xmm4 movapd [esp + nb234_fjzH1], xmm5 movapd [esp + nb234_fixH1], xmm0 movapd [esp + nb234_fiyH1], xmm1 movapd [esp + nb234_fizH1], xmm2 ;# H1-H2 interaction movapd xmm0, [esp + nb234_rinvH1H2] movapd xmm7, xmm0 ;# xmm7=rinv movapd xmm5, [esp + nb234_krf] mulpd xmm0, xmm0 ;# xmm0=rinvsq mulpd xmm5, [esp + nb234_rsqH1H2] ;# xmm5=krsq movapd xmm6, xmm5 addpd xmm6, xmm7 ;# xmm6=rinv+ krsq subpd xmm6, [esp + nb234_crf] mulpd xmm6, [esp + nb234_qqHH] ;# xmm6=voul=qq*(rinv+ krsq-crf) mulpd xmm5, [esp + nb234_two] subpd xmm7, xmm5 ;# xmm7=rinv-2*krsq mulpd xmm7, [esp + nb234_qqHH] ;# xmm7 = coul part of fscal addpd xmm6, [esp + nb234_vctot] ;# local vctot summation variable mulpd xmm0, xmm7 movapd [esp + nb234_vctot], xmm6 movapd xmm1, xmm0 movapd xmm2, xmm0 xorpd xmm3, xmm3 movapd xmm4, xmm3 movapd xmm5, xmm3 mulpd xmm0, [esp + nb234_dxH1H2] mulpd xmm1, [esp + nb234_dyH1H2] mulpd xmm2, [esp + nb234_dzH1H2] subpd xmm3, xmm0 subpd xmm4, xmm1 subpd xmm5, xmm2 addpd xmm0, [esp + nb234_fixH1] addpd xmm1, [esp + nb234_fiyH1] addpd xmm2, [esp + nb234_fizH1] movapd [esp + nb234_fjxH2], xmm3 movapd [esp + nb234_fjyH2], xmm4 movapd [esp + nb234_fjzH2], xmm5 movapd [esp + nb234_fixH1], xmm0 movapd [esp + nb234_fiyH1], xmm1 movapd [esp + nb234_fizH1], xmm2 ;# H1-M interaction movapd xmm0, [esp + nb234_rinvH1M] movapd xmm7, xmm0 ;# xmm7=rinv movapd xmm5, [esp + nb234_krf] mulpd xmm0, xmm0 ;# xmm0=rinvsq mulpd xmm5, [esp + nb234_rsqH1M] ;# xmm5=krsq movapd xmm6, xmm5 addpd xmm6, xmm7 ;# xmm6=rinv+ krsq subpd xmm6, [esp + nb234_crf] mulpd xmm6, [esp + nb234_qqMH] ;# xmm6=voul=qq*(rinv+ krsq-crf) mulpd xmm5, [esp + nb234_two] subpd xmm7, xmm5 ;# xmm7=rinv-2*krsq mulpd xmm7, [esp + nb234_qqMH] ;# xmm7 = coul part of fscal addpd xmm6, [esp + nb234_vctot] ;# local vctot summation variable mulpd xmm0, xmm7 movapd [esp + nb234_vctot], xmm6 movapd xmm1, xmm0 movapd xmm2, xmm0 xorpd xmm3, xmm3 movapd xmm4, xmm3 movapd xmm5, xmm3 mulpd xmm0, [esp + nb234_dxH1M] mulpd xmm1, [esp + nb234_dyH1M] mulpd xmm2, [esp + nb234_dzH1M] subpd xmm3, xmm0 subpd xmm4, xmm1 subpd xmm5, xmm2 addpd xmm0, [esp + nb234_fixH1] addpd xmm1, [esp + nb234_fiyH1] addpd xmm2, [esp + nb234_fizH1] movapd [esp + nb234_fjxM], xmm3 movapd [esp + nb234_fjyM], xmm4 movapd [esp + nb234_fjzM], xmm5 movapd [esp + nb234_fixH1], xmm0 movapd [esp + nb234_fiyH1], xmm1 movapd [esp + nb234_fizH1], xmm2 ;# H2-H1 interaction movapd xmm0, [esp + nb234_rinvH2H1] movapd xmm7, xmm0 ;# xmm7=rinv movapd xmm5, [esp + nb234_krf] mulpd xmm0, xmm0 ;# xmm0=rinvsq mulpd xmm5, [esp + nb234_rsqH2H1] ;# xmm5=krsq movapd xmm6, xmm5 addpd xmm6, xmm7 ;# xmm6=rinv+ krsq subpd xmm6, [esp + nb234_crf] mulpd xmm6, [esp + nb234_qqHH] ;# xmm6=voul=qq*(rinv+ krsq-crf) mulpd xmm5, [esp + nb234_two] subpd xmm7, xmm5 ;# xmm7=rinv-2*krsq mulpd xmm7, [esp + nb234_qqHH] ;# xmm7 = coul part of fscal addpd xmm6, [esp + nb234_vctot] ;# local vctot summation variable mulpd xmm0, xmm7 movapd [esp + nb234_vctot], xmm6 movapd xmm1, xmm0 movapd xmm2, xmm0 movapd xmm3, [esp + nb234_fjxH1] movapd xmm4, [esp + nb234_fjyH1] movapd xmm5, [esp + nb234_fjzH1] mulpd xmm0, [esp + nb234_dxH2H1] mulpd xmm1, [esp + nb234_dyH2H1] mulpd xmm2, [esp + nb234_dzH2H1] subpd xmm3, xmm0 subpd xmm4, xmm1 subpd xmm5, xmm2 addpd xmm0, [esp + nb234_fixH2] addpd xmm1, [esp + nb234_fiyH2] addpd xmm2, [esp + nb234_fizH2] movapd [esp + nb234_fjxH1], xmm3 movapd [esp + nb234_fjyH1], xmm4 movapd [esp + nb234_fjzH1], xmm5 movapd [esp + nb234_fixH2], xmm0 movapd [esp + nb234_fiyH2], xmm1 movapd [esp + nb234_fizH2], xmm2 ;# H2-H2 interaction movapd xmm0, [esp + nb234_rinvH2H2] movapd xmm7, xmm0 ;# xmm7=rinv movapd xmm5, [esp + nb234_krf] mulpd xmm0, xmm0 ;# xmm0=rinvsq mulpd xmm5, [esp + nb234_rsqH2H2] ;# xmm5=krsq movapd xmm6, xmm5 addpd xmm6, xmm7 ;# xmm6=rinv+ krsq subpd xmm6, [esp + nb234_crf] mulpd xmm6, [esp + nb234_qqHH] ;# xmm6=voul=qq*(rinv+ krsq-crf) mulpd xmm5, [esp + nb234_two] subpd xmm7, xmm5 ;# xmm7=rinv-2*krsq mulpd xmm7, [esp + nb234_qqHH] ;# xmm7 = coul part of fscal addpd xmm6, [esp + nb234_vctot] ;# local vctot summation variable mulpd xmm0, xmm7 movapd [esp + nb234_vctot], xmm6 movapd xmm1, xmm0 movapd xmm2, xmm0 movapd xmm3, [esp + nb234_fjxH2] movapd xmm4, [esp + nb234_fjyH2] movapd xmm5, [esp + nb234_fjzH2] mulpd xmm0, [esp + nb234_dxH2H2] mulpd xmm1, [esp + nb234_dyH2H2] mulpd xmm2, [esp + nb234_dzH2H2] subpd xmm3, xmm0 subpd xmm4, xmm1 subpd xmm5, xmm2 addpd xmm0, [esp + nb234_fixH2] addpd xmm1, [esp + nb234_fiyH2] addpd xmm2, [esp + nb234_fizH2] movapd [esp + nb234_fjxH2], xmm3 movapd [esp + nb234_fjyH2], xmm4 movapd [esp + nb234_fjzH2], xmm5 movapd [esp + nb234_fixH2], xmm0 movapd [esp + nb234_fiyH2], xmm1 movapd [esp + nb234_fizH2], xmm2 ;# H2-M interaction movapd xmm0, [esp + nb234_rinvH2M] movapd xmm7, xmm0 ;# xmm7=rinv movapd xmm5, [esp + nb234_krf] mulpd xmm0, xmm0 ;# xmm0=rinvsq mulpd xmm5, [esp + nb234_rsqH2M] ;# xmm5=krsq movapd xmm6, xmm5 addpd xmm6, xmm7 ;# xmm6=rinv+ krsq subpd xmm6, [esp + nb234_crf] mulpd xmm6, [esp + nb234_qqMH] ;# xmm6=voul=qq*(rinv+ krsq-crf) mulpd xmm5, [esp + nb234_two] subpd xmm7, xmm5 ;# xmm7=rinv-2*krsq mulpd xmm7, [esp + nb234_qqMH] ;# xmm7 = coul part of fscal addpd xmm6, [esp + nb234_vctot] ;# local vctot summation variable mulpd xmm0, xmm7 movapd [esp + nb234_vctot], xmm6 movapd xmm1, xmm0 movapd xmm2, xmm0 movapd xmm3, [esp + nb234_fjxM] movapd xmm4, [esp + nb234_fjyM] movapd xmm5, [esp + nb234_fjzM] mulpd xmm0, [esp + nb234_dxH2M] mulpd xmm1, [esp + nb234_dyH2M] mulpd xmm2, [esp + nb234_dzH2M] subpd xmm3, xmm0 subpd xmm4, xmm1 subpd xmm5, xmm2 addpd xmm0, [esp + nb234_fixH2] addpd xmm1, [esp + nb234_fiyH2] addpd xmm2, [esp + nb234_fizH2] movapd [esp + nb234_fjxM], xmm3 movapd [esp + nb234_fjyM], xmm4 movapd [esp + nb234_fjzM], xmm5 movapd [esp + nb234_fixH2], xmm0 movapd [esp + nb234_fiyH2], xmm1 movapd [esp + nb234_fizH2], xmm2 ;# M-H1 interaction movapd xmm0, [esp + nb234_rinvMH1] movapd xmm7, xmm0 ;# xmm7=rinv movapd xmm5, [esp + nb234_krf] mulpd xmm0, xmm0 ;# xmm0=rinvsq mulpd xmm5, [esp + nb234_rsqMH1] ;# xmm5=krsq movapd xmm6, xmm5 addpd xmm6, xmm7 ;# xmm6=rinv+ krsq subpd xmm6, [esp + nb234_crf] mulpd xmm6, [esp + nb234_qqMH] ;# xmm6=voul=qq*(rinv+ krsq-crf) mulpd xmm5, [esp + nb234_two] subpd xmm7, xmm5 ;# xmm7=rinv-2*krsq mulpd xmm7, [esp + nb234_qqMH] ;# xmm7 = coul part of fscal addpd xmm6, [esp + nb234_vctot] ;# local vctot summation variable mulpd xmm0, xmm7 movapd [esp + nb234_vctot], xmm6 movapd xmm1, xmm0 movapd xmm2, xmm0 movapd xmm3, [esp + nb234_fjxH1] movapd xmm4, [esp + nb234_fjyH1] movapd xmm5, [esp + nb234_fjzH1] mulpd xmm0, [esp + nb234_dxMH1] mulpd xmm1, [esp + nb234_dyMH1] mulpd xmm2, [esp + nb234_dzMH1] subpd xmm3, xmm0 subpd xmm4, xmm1 subpd xmm5, xmm2 addpd xmm0, [esp + nb234_fixM] addpd xmm1, [esp + nb234_fiyM] addpd xmm2, [esp + nb234_fizM] movapd [esp + nb234_fjxH1], xmm3 movapd [esp + nb234_fjyH1], xmm4 movapd [esp + nb234_fjzH1], xmm5 movapd [esp + nb234_fixM], xmm0 movapd [esp + nb234_fiyM], xmm1 movapd [esp + nb234_fizM], xmm2 ;# M-H2 interaction movapd xmm0, [esp + nb234_rinvMH2] movapd xmm7, xmm0 ;# xmm7=rinv movapd xmm5, [esp + nb234_krf] mulpd xmm0, xmm0 ;# xmm0=rinvsq mulpd xmm5, [esp + nb234_rsqMH2] ;# xmm5=krsq movapd xmm6, xmm5 addpd xmm6, xmm7 ;# xmm6=rinv+ krsq subpd xmm6, [esp + nb234_crf] mulpd xmm6, [esp + nb234_qqMH] ;# xmm6=voul=qq*(rinv+ krsq-crf) mulpd xmm5, [esp + nb234_two]
⌨️ 快捷键说明
复制代码Ctrl + C
搜索代码Ctrl + F
全屏模式F11
增大字号Ctrl + =
减小字号Ctrl + -
显示快捷键?