nb_kernel212_x86_64_sse.intel_syntax.s
来自「最著名最快的分子模拟软件」· S 代码 · 共 2,217 行 · 第 1/5 页
S
2,217 行
shufps xmm2, xmm3, 136 ;# 10001000 => jzOa jzOb jzOc jzOd ;# xmm0 = Ox ;# xmm1 = Oy ;# xmm2 = Oz movaps xmm3, xmm0 movaps xmm4, xmm1 movaps xmm5, xmm2 movaps xmm6, xmm0 movaps xmm7, xmm1 movaps xmm8, xmm2 subps xmm0, [rsp + nb212_ixO] subps xmm1, [rsp + nb212_iyO] subps xmm2, [rsp + nb212_izO] subps xmm3, [rsp + nb212_ixH1] subps xmm4, [rsp + nb212_iyH1] subps xmm5, [rsp + nb212_izH1] subps xmm6, [rsp + nb212_ixH2] subps xmm7, [rsp + nb212_iyH2] subps xmm8, [rsp + nb212_izH2] movaps [rsp + nb212_dxOO], xmm0 movaps [rsp + nb212_dyOO], xmm1 movaps [rsp + nb212_dzOO], xmm2 mulps xmm0, xmm0 mulps xmm1, xmm1 mulps xmm2, xmm2 movaps [rsp + nb212_dxH1O], xmm3 movaps [rsp + nb212_dyH1O], xmm4 movaps [rsp + nb212_dzH1O], xmm5 mulps xmm3, xmm3 mulps xmm4, xmm4 mulps xmm5, xmm5 movaps [rsp + nb212_dxH2O], xmm6 movaps [rsp + nb212_dyH2O], xmm7 movaps [rsp + nb212_dzH2O], xmm8 mulps xmm6, xmm6 mulps xmm7, xmm7 mulps xmm8, xmm8 addps xmm0, xmm1 addps xmm0, xmm2 addps xmm3, xmm4 addps xmm3, xmm5 addps xmm6, xmm7 addps xmm6, xmm8 ;# start doing invsqrt for jO atoms rsqrtps xmm1, xmm0 rsqrtps xmm4, xmm3 rsqrtps xmm7, xmm6 movaps xmm2, xmm1 movaps xmm5, xmm4 movaps xmm8, xmm7 mulps xmm1, xmm1 ;# lu*lu mulps xmm4, xmm4 ;# lu*lu mulps xmm7, xmm7 ;# lu*lu movaps xmm9, [rsp + nb212_three] movaps xmm10, xmm9 movaps xmm11, xmm9 mulps xmm1, xmm0 ;# rsq*lu*lu mulps xmm4, xmm3 ;# rsq*lu*lu mulps xmm7, xmm6 ;# rsq*lu*lu subps xmm9, xmm1 subps xmm10, xmm4 subps xmm11, xmm7 ;# 3-rsq*lu*lu mulps xmm9, xmm2 mulps xmm10, xmm5 mulps xmm11, xmm8 ;# lu*(3-rsq*lu*lu) movaps xmm4, [rsp + nb212_half] mulps xmm9, xmm4 ;# rinvOO mulps xmm10, xmm4 ;# rinvH1O mulps xmm11, xmm4 ;# rinvH2O ;# O interactions ;# rsq in xmm0,xmm3,xmm6 ;# rinv in xmm9, xmm10, xmm11 movaps xmm1, xmm9 ;# copy of rinv movaps xmm4, xmm10 movaps xmm7, xmm11 movaps xmm2, [rsp + nb212_krf] mulps xmm9, xmm9 ;# rinvsq mulps xmm10, xmm10 mulps xmm11, xmm11 mulps xmm0, xmm2 ;# k*rsq mulps xmm3, xmm2 mulps xmm6, xmm2 movaps xmm2, xmm0 ;# copy of k*rsq movaps xmm5, xmm3 movaps xmm8, xmm6 addps xmm2, xmm1 ;# rinv+krsq addps xmm5, xmm4 addps xmm8, xmm7 movaps xmm12, xmm9 mulps xmm12, xmm12 ;# rinv4 mulps xmm12, xmm9 ;# rinv6 subps xmm2, [rsp + nb212_crf] ;# rinv+krsq-crf subps xmm5, [rsp + nb212_crf] subps xmm8, [rsp + nb212_crf] mulps xmm2, [rsp + nb212_qqOO] ;# voul=qq*(rinv+ krsq-crf) mulps xmm5, [rsp + nb212_qqOH] ;# voul=qq*(rinv+ krsq-crf) mulps xmm8, [rsp + nb212_qqOH] ;# voul=qq*(rinv+ krsq-crf) addps xmm0, xmm0 ;# 2*krsq addps xmm3, xmm3 addps xmm6, xmm6 subps xmm1, xmm0 ;# rinv-2*krsq subps xmm4, xmm3 subps xmm7, xmm6 movaps xmm13, xmm12 ;# rinv6 mulps xmm12, xmm12 ;# rinv12 mulps xmm13, [rsp + nb212_c6] mulps xmm12, [rsp + nb212_c12] movaps xmm14, xmm12 subps xmm14, xmm13 mulps xmm1, [rsp + nb212_qqOO] ;# (rinv-2*krsq)*qq mulps xmm4, [rsp + nb212_qqOH] mulps xmm7, [rsp + nb212_qqOH] addps xmm2, [rsp + nb212_vctot] addps xmm5, xmm8 addps xmm2, xmm5 movaps xmm15, xmm2 addps xmm14, [rsp + nb212_Vvdwtot] mulps xmm13, [rsp + nb212_six] mulps xmm12, [rsp + nb212_twelve] movaps [rsp + nb212_Vvdwtot], xmm14 subps xmm12, xmm13 ;# LJ fscal addps xmm1, xmm12 mulps xmm1, xmm9 ;# fscal mulps xmm4, xmm10 mulps xmm7, xmm11 ;# move j O forces to local temp variables movlps xmm9, [rdi + rax*4] ;# jxOa jyOa - - movlps xmm10, [rdi + rcx*4] ;# jxOc jyOc - - movhps xmm9, [rdi + rbx*4] ;# jxOa jyOa jxOb jyOb movhps xmm10, [rdi + rdx*4] ;# jxOc jyOc jxOd jyOd movss xmm11, [rdi + rax*4 + 8] ;# jzOa - - - movss xmm12, [rdi + rcx*4 + 8] ;# jzOc - - - movhps xmm11, [rdi + rbx*4 + 8] ;# jzOa - jzOb - movhps xmm12, [rdi + rdx*4 + 8] ;# jzOc - jzOd - shufps xmm11, xmm12, 136 ;# 10001000 => jzOa jzOb jzOc jzOd ;# xmm9: jxOa jyOa jxOb jyOb ;# xmm10: jxOc jyOc jxOd jyOd ;# xmm11: jzOa jzOb jzOc jzOd movaps xmm0, xmm1 movaps xmm2, xmm1 movaps xmm3, xmm4 movaps xmm5, xmm4 movaps xmm6, xmm7 movaps xmm8, xmm7 mulps xmm0, [rsp + nb212_dxOO] mulps xmm1, [rsp + nb212_dyOO] mulps xmm2, [rsp + nb212_dzOO] mulps xmm3, [rsp + nb212_dxH1O] mulps xmm4, [rsp + nb212_dyH1O] mulps xmm5, [rsp + nb212_dzH1O] mulps xmm6, [rsp + nb212_dxH2O] mulps xmm7, [rsp + nb212_dyH2O] mulps xmm8, [rsp + nb212_dzH2O] movaps xmm13, xmm0 movaps xmm14, xmm1 addps xmm11, xmm2 addps xmm0, [rsp + nb212_fixO] addps xmm1, [rsp + nb212_fiyO] addps xmm2, [rsp + nb212_fizO] addps xmm13, xmm3 addps xmm14, xmm4 addps xmm11, xmm5 addps xmm3, [rsp + nb212_fixH1] addps xmm4, [rsp + nb212_fiyH1] addps xmm5, [rsp + nb212_fizH1] addps xmm13, xmm6 addps xmm14, xmm7 addps xmm11, xmm8 addps xmm6, [rsp + nb212_fixH2] addps xmm7, [rsp + nb212_fiyH2] addps xmm8, [rsp + nb212_fizH2] movaps [rsp + nb212_fixO], xmm0 movaps [rsp + nb212_fiyO], xmm1 movaps [rsp + nb212_fizO], xmm2 movaps [rsp + nb212_fixH1], xmm3 movaps [rsp + nb212_fiyH1], xmm4 movaps [rsp + nb212_fizH1], xmm5 movaps [rsp + nb212_fixH2], xmm6 movaps [rsp + nb212_fiyH2], xmm7 movaps [rsp + nb212_fizH2], xmm8 ;# xmm9 = fOx ;# xmm10 = fOy ;# xmm11 = fOz movaps xmm0, xmm13 unpcklps xmm13, xmm14 unpckhps xmm0, xmm14 addps xmm9, xmm13 addps xmm10, xmm0 movhlps xmm12, xmm11 ;# fOzc fOzd movlps [rdi + rax*4], xmm9 movhps [rdi + rbx*4], xmm9 movlps [rdi + rcx*4], xmm10 movhps [rdi + rdx*4], xmm10 movss [rdi + rax*4 + 8], xmm11 movss [rdi + rcx*4 + 8], xmm12 shufps xmm11, xmm11, 1 shufps xmm12, xmm12, 1 movss [rdi + rbx*4 + 8], xmm11 movss [rdi + rdx*4 + 8], xmm12 ;# move j H1 coordinates to local temp variables movlps xmm0, [rsi + rax*4 + 12] ;# jxH1a jyH1a - - movlps xmm1, [rsi + rcx*4 + 12] ;# jxH1c jyH1c - - movhps xmm0, [rsi + rbx*4 + 12] ;# jxH1a jyH1a jxH1b jyH1b movhps xmm1, [rsi + rdx*4 + 12] ;# jxH1c jyH1c jxH1d jyH1d movss xmm2, [rsi + rax*4 + 20] ;# jzH1a - - - movss xmm3, [rsi + rcx*4 + 20] ;# jzH1c - - - movhps xmm2, [rsi + rbx*4 + 20] ;# jzH1a - jzH1b - movhps xmm3, [rsi + rdx*4 + 20] ;# jzH1c - jzH1d - movaps xmm4, xmm0 unpcklps xmm0, xmm1 ;# jxH1a jxH1c jyH1a jyH1c unpckhps xmm4, xmm1 ;# jxH1b jxH1d jyH1b jyH1d movaps xmm1, xmm0 unpcklps xmm0, xmm4 ;# x unpckhps xmm1, xmm4 ;# y shufps xmm2, xmm3, 136 ;# 10001000 => jzH1a jzH1b jzH1c jzH1d ;# xmm0 = H1x ;# xmm1 = H1y ;# xmm2 = H1z movaps xmm3, xmm0 movaps xmm4, xmm1 movaps xmm5, xmm2 movaps xmm6, xmm0 movaps xmm7, xmm1 movaps xmm8, xmm2 subps xmm0, [rsp + nb212_ixO] subps xmm1, [rsp + nb212_iyO] subps xmm2, [rsp + nb212_izO] subps xmm3, [rsp + nb212_ixH1] subps xmm4, [rsp + nb212_iyH1] subps xmm5, [rsp + nb212_izH1] subps xmm6, [rsp + nb212_ixH2] subps xmm7, [rsp + nb212_iyH2] subps xmm8, [rsp + nb212_izH2] movaps [rsp + nb212_dxOH1], xmm0 movaps [rsp + nb212_dyOH1], xmm1 movaps [rsp + nb212_dzOH1], xmm2 mulps xmm0, xmm0 mulps xmm1, xmm1 mulps xmm2, xmm2 movaps [rsp + nb212_dxH1H1], xmm3 movaps [rsp + nb212_dyH1H1], xmm4 movaps [rsp + nb212_dzH1H1], xmm5 mulps xmm3, xmm3 mulps xmm4, xmm4 mulps xmm5, xmm5 movaps [rsp + nb212_dxH2H1], xmm6 movaps [rsp + nb212_dyH2H1], xmm7 movaps [rsp + nb212_dzH2H1], xmm8 mulps xmm6, xmm6 mulps xmm7, xmm7 mulps xmm8, xmm8 addps xmm0, xmm1 addps xmm0, xmm2 addps xmm3, xmm4 addps xmm3, xmm5 addps xmm6, xmm7 addps xmm6, xmm8 ;# start doing invsqrt for jH1 atoms rsqrtps xmm1, xmm0 rsqrtps xmm4, xmm3 rsqrtps xmm7, xmm6 movaps xmm2, xmm1 movaps xmm5, xmm4 movaps xmm8, xmm7 mulps xmm1, xmm1 ;# lu*lu mulps xmm4, xmm4 ;# lu*lu mulps xmm7, xmm7 ;# lu*lu movaps xmm9, [rsp + nb212_three] movaps xmm10, xmm9 movaps xmm11, xmm9 mulps xmm1, xmm0 ;# rsq*lu*lu mulps xmm4, xmm3 ;# rsq*lu*lu mulps xmm7, xmm6 ;# rsq*lu*lu subps xmm9, xmm1 subps xmm10, xmm4 subps xmm11, xmm7 ;# 3-rsq*lu*lu mulps xmm9, xmm2 mulps xmm10, xmm5 mulps xmm11, xmm8 ;# lu*(3-rsq*lu*lu) movaps xmm4, [rsp + nb212_half] mulps xmm9, xmm4 ;# rinvOH1 mulps xmm10, xmm4 ;# rinvH1H1 mulps xmm11, xmm4 ;# rinvH2H1 ;# H1 interactions ;# rsq in xmm0,xmm3,xmm6 ;# rinv in xmm9, xmm10, xmm11 movaps xmm1, xmm9 ;# copy of rinv movaps xmm4, xmm10 movaps xmm7, xmm11 movaps xmm2, [rsp + nb212_krf] mulps xmm9, xmm9 ;# rinvsq mulps xmm10, xmm10 mulps xmm11, xmm11 mulps xmm0, xmm2 ;# k*rsq mulps xmm3, xmm2 mulps xmm6, xmm2 movaps xmm2, xmm0 ;# copy of k*rsq movaps xmm5, xmm3 movaps xmm8, xmm6 addps xmm2, xmm1 ;# rinv+krsq addps xmm5, xmm4 addps xmm8, xmm7 movaps xmm14, [rsp + nb212_crf] subps xmm2, xmm14 ;# rinv+krsq-crf subps xmm5, xmm14 subps xmm8, xmm14 movaps xmm12, [rsp + nb212_qqOH] movaps xmm13, [rsp + nb212_qqHH] mulps xmm2, xmm12 ;# xmm6=voul=qq*(rinv+ krsq-crf) mulps xmm5, xmm13 ;# xmm6=voul=qq*(rinv+ krsq-crf) mulps xmm8, xmm13 ;# xmm6=voul=qq*(rinv+ krsq-crf) addps xmm0, xmm0 ;# 2*krsq addps xmm3, xmm3 addps xmm6, xmm6 subps xmm1, xmm0 ;# rinv-2*krsq subps xmm4, xmm3 subps xmm7, xmm6 mulps xmm1, xmm12 ;# (rinv-2*krsq)*qq mulps xmm4, xmm13 mulps xmm7, xmm13 addps xmm15, xmm2 addps xmm5, xmm8 addps xmm15, xmm5 mulps xmm1, xmm9 ;# fscal mulps xmm4, xmm10 mulps xmm7, xmm11 ;# move j H1 forces to local temp variables movlps xmm9, [rdi + rax*4 + 12] ;# jxH1a jyH1a - - movlps xmm10, [rdi + rcx*4 + 12] ;# jxH1c jyH1c - - movhps xmm9, [rdi + rbx*4 + 12] ;# jxH1a jyH1a jxH1b jyH1b movhps xmm10, [rdi + rdx*4 + 12] ;# jxH1c jyH1c jxH1d jyH1d movss xmm11, [rdi + rax*4 + 20] ;# jzH1a - - - movss xmm12, [rdi + rcx*4 + 20] ;# jzH1c - - - movhps xmm11, [rdi + rbx*4 + 20] ;# jzH1a - jzH1b - movhps xmm12, [rdi + rdx*4 + 20] ;# jzH1c - jzH1d - shufps xmm11, xmm12, 136 ;# 10001000 => jzH1a jzH1b jzH1c jzH1d ;# xmm9: jxH1a jyH1a jxH1b jyH1b ;# xmm10: jxH1c jyH1c jxH1d jyH1d ;# xmm11: jzH1a jzH1b jzH1c jzH1d movaps xmm0, xmm1 movaps xmm2, xmm1 movaps xmm3, xmm4 movaps xmm5, xmm4 movaps xmm6, xmm7 movaps xmm8, xmm7 mulps xmm0, [rsp + nb212_dxOH1] mulps xmm1, [rsp + nb212_dyOH1] mulps xmm2, [rsp + nb212_dzOH1] mulps xmm3, [rsp + nb212_dxH1H1] mulps xmm4, [rsp + nb212_dyH1H1] mulps xmm5, [rsp + nb212_dzH1H1] mulps xmm6, [rsp + nb212_dxH2H1] mulps xmm7, [rsp + nb212_dyH2H1] mulps xmm8, [rsp + nb212_dzH2H1] movaps xmm13, xmm0 movaps xmm14, xmm1 addps xmm11, xmm2 addps xmm0, [rsp + nb212_fixO] addps xmm1, [rsp + nb212_fiyO] addps xmm2, [rsp + nb212_fizO] addps xmm13, xmm3 addps xmm14, xmm4 addps xmm11, xmm5 addps xmm3, [rsp + nb212_fixH1] addps xmm4, [rsp + nb212_fiyH1] addps xmm5, [rsp + nb212_fizH1] addps xmm13, xmm6 addps xmm14, xmm7 addps xmm11, xmm8 addps xmm6, [rsp + nb212_fixH2] addps xmm7, [rsp + nb212_fiyH2] addps xmm8, [rsp + nb212_fizH2] movaps [rsp + nb212_fixO], xmm0 movaps [rsp + nb212_fiyO], xmm1 movaps [rsp + nb212_fizO], xmm2 movaps [rsp + nb212_fixH1], xmm3 movaps [rsp + nb212_fiyH1], xmm4 movaps [rsp + nb212_fizH1], xmm5 movaps [rsp + nb212_fixH2], xmm6 movaps [rsp + nb212_fiyH2], xmm7 movaps [rsp + nb212_fizH2], xmm8 ⌨️ 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?