nb_kernel132_ia32_sse2.intel_syntax.s
来自「最著名最快的分子模拟软件」· S 代码 · 共 2,224 行 · 第 1/5 页
S
2,224 行
subsd xmm4, [esp + nb132_jyH1] subsd xmm5, [esp + nb132_jzH1] movapd [esp + nb132_dxH2O], xmm0 movapd [esp + nb132_dyH2O], xmm1 movapd [esp + nb132_dzH2O], xmm2 mulsd xmm0, xmm0 mulsd xmm1, xmm1 mulsd xmm2, xmm2 movapd [esp + nb132_dxH2H1], xmm3 movapd [esp + nb132_dyH2H1], xmm4 movapd [esp + nb132_dzH2H1], xmm5 mulsd xmm3, xmm3 mulsd xmm4, xmm4 mulsd xmm5, xmm5 addsd xmm0, xmm1 addsd xmm0, xmm2 addsd xmm4, xmm3 addsd xmm4, xmm5 movapd [esp + nb132_rsqH2O], xmm0 movapd [esp + nb132_rsqH2H1], xmm4 movapd xmm0, [esp + nb132_ixH2] movapd xmm1, [esp + nb132_iyH2] movapd xmm2, [esp + nb132_izH2] subsd xmm0, [esp + nb132_jxH2] subsd xmm1, [esp + nb132_jyH2] subsd xmm2, [esp + nb132_jzH2] movapd [esp + nb132_dxH2H2], xmm0 movapd [esp + nb132_dyH2H2], xmm1 movapd [esp + nb132_dzH2H2], xmm2 mulsd xmm0, xmm0 mulsd xmm1, xmm1 mulsd xmm2, xmm2 addsd xmm0, xmm1 addsd xmm0, xmm2 movapd [esp + nb132_rsqH2H2], xmm0 ;# start doing invsqrt use rsq values in xmm0, xmm4 cvtsd2ss xmm1, xmm0 cvtsd2ss xmm5, xmm4 rsqrtss xmm1, xmm1 rsqrtss xmm5, xmm5 cvtss2sd xmm1, xmm1 cvtss2sd xmm5, xmm5 movapd xmm2, xmm1 ;# copy of luA movapd xmm6, xmm5 ;# copy of luB mulsd xmm1, xmm1 ;# luA*luA mulsd xmm5, xmm5 ;# luB*luB movapd xmm3, [esp + nb132_three] mulsd xmm1, xmm0 ;# rsqA*luA*luA mulsd xmm5, xmm4 ;# rsqB*luB*luB movapd xmm7, xmm3 subsd xmm3, xmm1 ;# 3-rsqA*luA*luA subsd xmm7, xmm5 ;# 3-rsqB*luB*luB mulsd xmm3, xmm2 ;# luA*(3-rsqA*luA*luA) mulsd xmm7, xmm6 ;# luB*(3-rsqB*luB*luB) mulsd xmm3, [esp + nb132_half] ;# iter1 mulsd xmm7, [esp + nb132_half] ;# iter1 movapd xmm2, xmm3 ;# copy of luA movapd xmm6, xmm7 ;# copy of luB mulsd xmm3, xmm3 ;# luA*luA mulsd xmm7, xmm7 ;# luB*luB movapd xmm1, [esp + nb132_three] mulsd xmm3, xmm0 ;# rsqA*luA*luA mulsd xmm7, xmm4 ;# rsqB*luB*luB movapd xmm5, xmm1 subsd xmm1, xmm3 ;# 3-rsqA*luA*luA subsd xmm5, xmm7 ;# 3-rsqB*luB*luB mulsd xmm1, xmm2 ;# luA*(3-rsqA*luA*luA) mulsd xmm5, xmm6 ;# luB*(3-rsqB*luB*luB) mulsd xmm1, [esp + nb132_half] ;# rinv mulsd xmm5, [esp + nb132_half] ;# rinv movapd [esp + nb132_rinvH2H2], xmm1 movapd [esp + nb132_rinvH2H1], xmm5 movapd xmm0, [esp + nb132_rsqOO] movapd xmm4, [esp + nb132_rsqOH1] cvtsd2ss xmm1, xmm0 cvtsd2ss xmm5, xmm4 rsqrtss xmm1, xmm1 rsqrtss xmm5, xmm5 cvtss2sd xmm1, xmm1 cvtss2sd xmm5, xmm5 movapd xmm2, xmm1 ;# copy of luA movapd xmm6, xmm5 ;# copy of luB mulsd xmm1, xmm1 ;# luA*luA mulsd xmm5, xmm5 ;# luB*luB movapd xmm3, [esp + nb132_three] mulsd xmm1, xmm0 ;# rsqA*luA*luA mulsd xmm5, xmm4 ;# rsqB*luB*luB movapd xmm7, xmm3 subsd xmm3, xmm1 ;# 3-rsqA*luA*luA subsd xmm7, xmm5 ;# 3-rsqB*luB*luB mulsd xmm3, xmm2 ;# luA*(3-rsqA*luA*luA) mulsd xmm7, xmm6 ;# luB*(3-rsqB*luB*luB) mulsd xmm3, [esp + nb132_half] ;# iter1 of mulsd xmm7, [esp + nb132_half] ;# iter1 of movapd xmm2, xmm3 ;# copy of luA movapd xmm6, xmm7 ;# copy of luB mulsd xmm3, xmm3 ;# luA*luA mulsd xmm7, xmm7 ;# luB*luB movapd xmm1, [esp + nb132_three] mulsd xmm3, xmm0 ;# rsqA*luA*luA mulsd xmm7, xmm4 ;# rsqB*luB*luB movapd xmm5, xmm1 subsd xmm1, xmm3 ;# 3-rsqA*luA*luA subsd xmm5, xmm7 ;# 3-rsqB*luB*luB mulsd xmm1, xmm2 ;# luA*(3-rsqA*luA*luA) mulsd xmm5, xmm6 ;# luB*(3-rsqB*luB*luB) mulsd xmm1, [esp + nb132_half] ;# rinv mulsd xmm5, [esp + nb132_half] ;# rinv movapd [esp + nb132_rinvOO], xmm1 movapd [esp + nb132_rinvOH1], xmm5 movapd xmm0, [esp + nb132_rsqOH2] movapd xmm4, [esp + nb132_rsqH1O] cvtsd2ss xmm1, xmm0 cvtsd2ss xmm5, xmm4 rsqrtss xmm1, xmm1 rsqrtss xmm5, xmm5 cvtss2sd xmm1, xmm1 cvtss2sd xmm5, xmm5 movapd xmm2, xmm1 ;# copy of luA movapd xmm6, xmm5 ;# copy of luB mulsd xmm1, xmm1 ;# luA*luA mulsd xmm5, xmm5 ;# luB*luB movapd xmm3, [esp + nb132_three] mulsd xmm1, xmm0 ;# rsqA*luA*luA mulsd xmm5, xmm4 ;# rsqB*luB*luB movapd xmm7, xmm3 subsd xmm3, xmm1 ;# 3-rsqA*luA*luA subsd xmm7, xmm5 ;# 3-rsqB*luB*luB mulsd xmm3, xmm2 ;# luA*(3-rsqA*luA*luA) mulsd xmm7, xmm6 ;# luB*(3-rsqB*luB*luB) mulsd xmm3, [esp + nb132_half] ;# iter1 mulsd xmm7, [esp + nb132_half] ;# iter1 movapd xmm2, xmm3 ;# copy of luA movapd xmm6, xmm7 ;# copy of luB mulsd xmm3, xmm3 ;# luA*luA mulsd xmm7, xmm7 ;# luB*luB movapd xmm1, [esp + nb132_three] mulsd xmm3, xmm0 ;# rsqA*luA*luA mulsd xmm7, xmm4 ;# rsqB*luB*luB movapd xmm5, xmm1 subsd xmm1, xmm3 ;# 3-rsqA*luA*luA subsd xmm5, xmm7 ;# 3-rsqB*luB*luB mulsd xmm1, xmm2 ;# luA*(3-rsqA*luA*luA) mulsd xmm5, xmm6 ;# luB*(3-rsqB*luB*luB) mulsd xmm1, [esp + nb132_half] ;# rinv mulsd xmm5, [esp + nb132_half] ;# rinv movapd [esp + nb132_rinvOH2], xmm1 movapd [esp + nb132_rinvH1O], xmm5 movapd xmm0, [esp + nb132_rsqH1H1] movapd xmm4, [esp + nb132_rsqH1H2] cvtsd2ss xmm1, xmm0 cvtsd2ss xmm5, xmm4 rsqrtss xmm1, xmm1 rsqrtss xmm5, xmm5 cvtss2sd xmm1, xmm1 cvtss2sd xmm5, xmm5 movapd xmm2, xmm1 ;# copy of luA movapd xmm6, xmm5 ;# copy of luB mulsd xmm1, xmm1 ;# luA*luA mulsd xmm5, xmm5 ;# luB*luB movapd xmm3, [esp + nb132_three] mulsd xmm1, xmm0 ;# rsqA*luA*luA mulsd xmm5, xmm4 ;# rsqB*luB*luB movapd xmm7, xmm3 subsd xmm3, xmm1 ;# 3-rsqA*luA*luA subsd xmm7, xmm5 ;# 3-rsqB*luB*luB mulsd xmm3, xmm2 ;# luA*(3-rsqA*luA*luA) mulsd xmm7, xmm6 ;# luB*(3-rsqB*luB*luB) mulsd xmm3, [esp + nb132_half] ;# iter1a mulsd xmm7, [esp + nb132_half] ;# iter1b movapd xmm2, xmm3 ;# copy of luA movapd xmm6, xmm7 ;# copy of luB mulsd xmm3, xmm3 ;# luA*luA mulsd xmm7, xmm7 ;# luB*luB movapd xmm1, [esp + nb132_three] mulsd xmm3, xmm0 ;# rsqA*luA*luA mulsd xmm7, xmm4 ;# rsqB*luB*luB movapd xmm5, xmm1 subsd xmm1, xmm3 ;# 3-rsqA*luA*luA subsd xmm5, xmm7 ;# 3-rsqB*luB*luB mulsd xmm1, xmm2 ;# luA*(3-rsqA*luA*luA) mulsd xmm5, xmm6 ;# luB*(3-rsqB*luB*luB) mulsd xmm1, [esp + nb132_half] ;# rinv mulsd xmm5, [esp + nb132_half] ;# rinv movapd [esp + nb132_rinvH1H1], xmm1 movapd [esp + nb132_rinvH1H2], xmm5 movapd xmm0, [esp + nb132_rsqH2O] cvtsd2ss xmm1, xmm0 rsqrtss xmm1, xmm1 cvtss2sd xmm1, xmm1 movapd xmm2, xmm1 ;# copy of luA mulsd xmm1, xmm1 ;# luA*luA movapd xmm3, [esp + nb132_three] mulsd xmm1, xmm0 ;# rsqA*luA*luA subsd xmm3, xmm1 ;# 3-rsqA*luA*luA mulsd xmm3, xmm2 ;# luA*(3-rsqA*luA*luA) mulsd xmm3, [esp + nb132_half] ;# iter1 movapd xmm2, xmm3 ;# copy of luA mulsd xmm3, xmm3 ;# luA*luA movapd xmm1, [esp + nb132_three] mulsd xmm3, xmm0 ;# rsqA*luA*luA subsd xmm1, xmm3 ;# 3-rsqA*luA*luA mulsd xmm1, xmm2 ;# luA*(3-rsqA*luA*luA) mulsd xmm1, [esp + nb132_half] ;# rinv movapd [esp + nb132_rinvH2O], xmm1 ;# start with OO interaction movsd xmm0, [esp + nb132_rinvOO] movsd xmm7, xmm0 ;# xmm7=rinv mulsd xmm7, [esp + nb132_qqOO] ;# xmm7 = qq*rinv movsd xmm6, xmm7 mulsd xmm7, xmm0 addsd xmm6, [esp + nb132_vctot] movsd [esp + nb132_vctot], xmm6 movsd [esp + nb132_fstmp], xmm7 ;# LJ table interaction movsd xmm4, [esp + nb132_rsqOO] mulsd xmm4, xmm0 ;# xmm4=r mulsd xmm4, [esp + nb132_tsc] cvttsd2si ebx, xmm4 ;# mm6 = lu idx cvtsi2sd xmm5, ebx subsd xmm4, xmm5 movsd xmm1, xmm4 ;# xmm1=eps movsd xmm2, xmm1 mulsd xmm2, xmm2 ;# xmm2=eps2 shl ebx, 3 mov esi, [ebp + nb132_VFtab] ;# dispersion movlpd xmm4, [esi + ebx*8] ;# Y1 movhpd xmm4, [esi + ebx*8 + 8] ;# Y1 F1 movapd xmm5, xmm4 unpcklpd xmm4, xmm3 ;# Y1 Y2 unpckhpd xmm5, xmm3 ;# F1 F2 movlpd xmm6, [esi + ebx*8 + 16] ;# G1 movhpd xmm6, [esi + ebx*8 + 24] ;# G1 H1 movapd xmm7, xmm6 unpcklpd xmm6, xmm3 ;# G1 G2 unpckhpd xmm7, xmm3 ;# H1 H2 ;# dispersion table ready, in xmm4-xmm7 mulsd xmm6, xmm1 ;# xmm6=Geps mulsd xmm7, xmm2 ;# xmm7=Heps2 addsd xmm5, xmm6 addsd xmm5, xmm7 ;# xmm5=Fp mulsd xmm7, [esp + nb132_two] ;# two*Heps2 addsd xmm7, xmm6 addsd xmm7, xmm5 ;# xmm7=FF mulsd xmm5, xmm1 ;# xmm5=eps*Fp addsd xmm5, xmm4 ;# xmm5=VV movsd xmm4, [esp + nb132_c6] mulsd xmm7, xmm4 ;# fijD mulsd xmm5, xmm4 ;# Vvdw6 ;# put scalar force on stack Update Vvdwtot directly addsd xmm5, [esp + nb132_Vvdwtot] movsd xmm3, [esp + nb132_fstmp] mulsd xmm7, [esp + nb132_tsc] subsd xmm3, xmm7 movsd [esp + nb132_fstmp], xmm3 movsd [esp + nb132_Vvdwtot], xmm5 ;# repulsion movlpd xmm4, [esi + ebx*8 + 32] ;# Y1 movhpd xmm4, [esi + ebx*8 + 40] ;# Y1 F1 movapd xmm5, xmm4 unpcklpd xmm4, xmm3 ;# Y1 Y2 unpckhpd xmm5, xmm3 ;# F1 F2 movlpd xmm6, [esi + ebx*8 + 48] ;# G1 movhpd xmm6, [esi + ebx*8 + 56] ;# G1 H1 movapd xmm7, xmm6 unpcklpd xmm6, xmm3 ;# G1 G2 unpckhpd xmm7, xmm3 ;# H1 H2 ;# table ready, in xmm4-xmm7 mulsd xmm6, xmm1 ;# xmm6=Geps mulsd xmm7, xmm2 ;# xmm7=Heps2 addsd xmm5, xmm6 addsd xmm5, xmm7 ;# xmm5=Fp mulsd xmm7, [esp + nb132_two] ;# two*Heps2 addsd xmm7, xmm6 addsd xmm7, xmm5 ;# xmm7=FF mulsd xmm5, xmm1 ;# xmm5=eps*Fp addsd xmm5, xmm4 ;# xmm5=VV movsd xmm4, [esp + nb132_c12] mulsd xmm7, xmm4 mulsd xmm5, xmm4 addsd xmm5, [esp + nb132_Vvdwtot] movsd xmm3, [esp + nb132_fstmp] mulsd xmm7, [esp + nb132_tsc] subsd xmm3, xmm7 movsd [esp + nb132_Vvdwtot], xmm5 mulsd xmm0, xmm3 movapd xmm1, xmm0 movapd xmm2, xmm0 mov edi, [ebp + nb132_faction] ;# update forces xorpd xmm3, xmm3 movapd xmm4, xmm3 movapd xmm5, xmm3 mulsd xmm0, [esp + nb132_dxOO] mulsd xmm1, [esp + nb132_dyOO] mulsd xmm2, [esp + nb132_dzOO] subsd xmm3, xmm0 subsd xmm4, xmm1 subsd xmm5, xmm2 addsd xmm0, [esp + nb132_fixO] addsd xmm1, [esp + nb132_fiyO] addsd xmm2, [esp + nb132_fizO] movlpd [esp + nb132_fjxO], xmm3 movlpd [esp + nb132_fjyO], xmm4 movlpd [esp + nb132_fjzO], xmm5 movlpd [esp + nb132_fixO], xmm0 movlpd [esp + nb132_fiyO], xmm1 movlpd [esp + nb132_fizO], xmm2 ;# O-H1 interaction movsd xmm0, [esp + nb132_rinvOH1] movsd xmm6, xmm0 ;# xmm6=rinv mulsd xmm6, [esp + nb132_qqOH] mulsd xmm0, xmm0 mulsd xmm0, xmm6 addsd xmm6, [esp + nb132_vctot] movsd xmm1, xmm0 movsd xmm2, xmm0 xorpd xmm3, xmm3 movapd xmm4, xmm3 movapd xmm5, xmm3 mulsd xmm0, [esp + nb132_dxOH1] mulsd xmm1, [esp + nb132_dyOH1] mulsd xmm2, [esp + nb132_dzOH1] subsd xmm3, xmm0 subsd xmm4, xmm1 subsd xmm5, xmm2 addsd xmm0, [esp + nb132_fixO] addsd xmm1, [esp + nb132_fiyO] addsd xmm2, [esp + nb132_fizO] movlpd [esp + nb132_fjxH1], xmm3 movlpd [esp + nb132_fjyH1], xmm4 movlpd [esp + nb132_fjzH1], xmm5 movlpd [esp + nb132_fixO], xmm0 movlpd [esp + nb132_fiyO], xmm1 movlpd [esp + nb132_fizO], xmm2 ;# O-H2 interaction movsd xmm0, [esp + nb132_rinvOH2] movsd xmm4, xmm0 ;# xmm4=rinv mulsd xmm4, [esp + nb132_qqOH] mulsd xmm0, xmm0 mulsd xmm0, xmm4 addsd xmm6, xmm4 movsd xmm1, xmm0 movsd xmm2, xmm0 xorpd xmm3, xmm3 movapd xmm4, xmm3 movapd xmm5, xmm3 mulsd xmm0, [esp + nb132_dxOH2] mulsd xmm1, [esp + nb132_dyOH2] mulsd xmm2, [esp + nb132_dzOH2] subsd xmm3, xmm0 subsd xmm4, xmm1 subsd xmm5, xmm2 addsd xmm0, [esp + nb132_fixO] addsd xmm1, [esp + nb132_fiyO] addsd xmm2, [esp + nb132_fizO] movlpd [esp + nb132_fjxH2], xmm3 movlpd [esp + nb132_fjyH2], xmm4 movlpd [esp + nb132_fjzH2], xmm5 movlpd [esp + nb132_fixO], xmm0 movlpd [esp + nb132_fiyO], xmm1 movlpd [esp + nb132_fizO], xmm2 ;# H1-O interaction movsd xmm0, [esp + nb132_rinvH1O] movsd xmm4, xmm0 ;# xmm4=rinv mulsd xmm4, [esp + nb132_qqOH] mulsd xmm0, xmm0 mulsd xmm0, xmm4 addsd xmm6, xmm4 movsd xmm1, xmm0 movsd xmm2, xmm0 movapd xmm3, [esp + nb132_fjxO] movapd xmm4, [esp + nb132_fjyO] movapd xmm5, [esp + nb132_fjzO] mulsd xmm0, [esp + nb132_dxH1O] mulsd xmm1, [esp + nb132_dyH1O] mulsd xmm2, [esp + nb132_dzH1O] subsd xmm3, xmm0 subsd xmm4, xmm1 subsd xmm5, xmm2 addsd xmm0, [esp + nb132_fixH1] addsd xmm1, [esp + nb132_fiyH1] addsd xmm2, [esp + nb132_fizH1] movlpd [esp + nb132_fjxO], xmm3 movlpd [esp + nb132_fjyO], xmm4 movlpd [esp + nb132_fjzO], xmm5 movlpd [esp + nb132_fixH1], xmm0 movlpd [esp + nb132_fiyH1], xmm1 movlpd [esp + nb132_fizH1], xmm2 ;# H1-H1 interaction movsd xmm0, [esp + nb132_rinvH1H1] movsd xmm4, xmm0 ;# xmm4=rinv mulsd xmm4, [esp + nb132_qqHH] mulsd xmm0, xmm0 mulsd xmm0, xmm4 addsd xmm6, xmm4⌨️ 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?