nb_kernel200_ia32_sse2.intel_syntax.s
来自「最著名最快的分子模拟软件」· S 代码 · 共 1,049 行 · 第 1/2 页
S
1,049 行
movapd xmm0, [esp + nb200_fix] movapd xmm1, [esp + nb200_fiy] movapd xmm2, [esp + nb200_fiz] movhlps xmm3, xmm0 movhlps xmm4, xmm1 movhlps xmm5, xmm2 addsd xmm0, xmm3 addsd xmm1, xmm4 addsd xmm2, xmm5 ;# sum is in low xmm0-xmm2 ;# increment i force movsd xmm3, [edi + ecx*8] movsd xmm4, [edi + ecx*8 + 8] movsd xmm5, [edi + ecx*8 + 16] addsd xmm3, xmm0 addsd xmm4, xmm1 addsd xmm5, xmm2 movsd [edi + ecx*8], xmm3 movsd [edi + ecx*8 + 8], xmm4 movsd [edi + ecx*8 + 16], xmm5 ;# increment fshift force movsd xmm3, [esi + edx*8] movsd xmm4, [esi + edx*8 + 8] movsd xmm5, [esi + edx*8 + 16] addsd xmm3, xmm0 addsd xmm4, xmm1 addsd xmm5, xmm2 movsd [esi + edx*8], xmm3 movsd [esi + edx*8 + 8], xmm4 movsd [esi + edx*8 + 16], xmm5 ;# get n from stack mov esi, [esp + nb200_n] ;# get group index for i particle mov edx, [ebp + nb200_gid] ;# base of gid[] mov edx, [edx + esi*4] ;# ggid=gid[n] ;# accumulate total potential energy and update it movapd xmm7, [esp + nb200_vctot] ;# accumulate movhlps xmm6, xmm7 addsd xmm7, xmm6 ;# low xmm7 has the sum now ;# add earlier value from mem mov eax, [ebp + nb200_Vc] addsd xmm7, [eax + edx*8] ;# move back to mem movsd [eax + edx*8], xmm7 ;# finish if last mov ecx, [esp + nb200_nn1] ;# esi already loaded with n inc esi sub ecx, esi jz .nb200_outerend ;# not last, iterate outer loop once more! mov [esp + nb200_n], esi jmp .nb200_outer.nb200_outerend: ;# check if more outer neighborlists remain mov ecx, [esp + nb200_nri] ;# esi already loaded with n above sub ecx, esi jz .nb200_end ;# non-zero, do one more workunit jmp .nb200_threadloop.nb200_end: emms mov eax, [esp + nb200_nouter] mov ebx, [esp + nb200_ninner] mov ecx, [ebp + nb200_outeriter] mov edx, [ebp + nb200_inneriter] mov [ecx], eax mov [edx], ebx mov eax, [esp + nb200_salign] add esp, eax add esp, 304 pop edi pop esi pop edx pop ecx pop ebx pop eax leave ret.globl nb_kernel200nf_ia32_sse2.globl _nb_kernel200nf_ia32_sse2nb_kernel200nf_ia32_sse2: _nb_kernel200nf_ia32_sse2: .equiv nb200nf_p_nri, 8.equiv nb200nf_iinr, 12.equiv nb200nf_jindex, 16.equiv nb200nf_jjnr, 20.equiv nb200nf_shift, 24.equiv nb200nf_shiftvec, 28.equiv nb200nf_fshift, 32.equiv nb200nf_gid, 36.equiv nb200nf_pos, 40.equiv nb200nf_faction, 44.equiv nb200nf_charge, 48.equiv nb200nf_p_facel, 52.equiv nb200nf_argkrf, 56.equiv nb200nf_argcrf, 60.equiv nb200nf_Vc, 64.equiv nb200nf_type, 68.equiv nb200nf_p_ntype, 72.equiv nb200nf_vdwparam, 76.equiv nb200nf_Vvdw, 80.equiv nb200nf_p_tabscale, 84.equiv nb200nf_VFtab, 88.equiv nb200nf_invsqrta, 92.equiv nb200nf_dvda, 96.equiv nb200nf_p_gbtabscale, 100.equiv nb200nf_GBtab, 104.equiv nb200nf_p_nthreads, 108.equiv nb200nf_count, 112.equiv nb200nf_mtx, 116.equiv nb200nf_outeriter, 120.equiv nb200nf_inneriter, 124.equiv nb200nf_work, 128 ;# stack offsets for local variables ;# bottom of stack is cache-aligned for sse use .equiv nb200nf_ix, 0.equiv nb200nf_iy, 16.equiv nb200nf_iz, 32.equiv nb200nf_iq, 48.equiv nb200nf_vctot, 64.equiv nb200nf_half, 80.equiv nb200nf_three, 96.equiv nb200nf_krf, 112.equiv nb200nf_crf, 128.equiv nb200nf_is3, 144.equiv nb200nf_ii3, 148.equiv nb200nf_innerjjnr, 152.equiv nb200nf_innerk, 156.equiv nb200nf_n, 160.equiv nb200nf_nn1, 164.equiv nb200nf_nri, 168.equiv nb200nf_facel, 176 ;# uses 8 bytes.equiv nb200nf_nouter, 184.equiv nb200nf_ninner, 188.equiv nb200nf_salign, 192 push ebp mov ebp,esp push eax push ebx push ecx push edx push esi push edi sub esp, 172 ;# local stack space mov eax, esp and eax, 0xf sub esp, eax mov [esp + nb200nf_salign], eax emms ;# Move args passed by reference to stack mov ecx, [ebp + nb200nf_p_nri] mov esi, [ebp + nb200nf_p_facel] mov ecx, [ecx] movsd xmm7, [esi] mov [esp + nb200nf_nri], ecx movsd [esp + nb200nf_facel], xmm7 ;# zero iteration counters mov eax, 0 mov [esp + nb200nf_nouter], eax mov [esp + nb200nf_ninner], eax ;# create constant floating-point factors on stack mov eax, 0x00000000 ;# lower half of double 0.5 IEEE (hex) mov ebx, 0x3fe00000 mov [esp + nb200nf_half], eax mov [esp + nb200nf_half+4], ebx movsd xmm1, [esp + nb200nf_half] shufpd xmm1, xmm1, 0 ;# splat to all elements movapd xmm3, xmm1 addpd xmm3, xmm3 ;# 1.0 movapd xmm2, xmm3 addpd xmm2, xmm2 ;# 2.0 addpd xmm3, xmm2 ;# 3.0 movapd [esp + nb200nf_half], xmm1 movapd [esp + nb200nf_three], xmm3 mov esi, [ebp + nb200nf_argkrf] mov edi, [ebp + nb200nf_argcrf] movsd xmm5, [esi] movsd xmm6, [edi] shufpd xmm5, xmm5, 0 movapd [esp + nb200nf_krf], xmm5 shufpd xmm6, xmm6, 0 movapd [esp + nb200nf_crf], xmm6.nb200nf_threadloop: mov esi, [ebp + nb200nf_count] ;# pointer to sync counter mov eax, [esi].nb200nf_spinlock: mov ebx, eax ;# ebx=*count=nn0 add ebx, 1 ;# ebx=nn1=nn0+10 lock cmpxchg [esi], ebx ;# write nn1 to *counter, ;# if it hasnt changed. ;# or reread *counter to eax. pause ;# -> better p4 performance jnz .nb200nf_spinlock ;# if(nn1>nri) nn1=nri mov ecx, [esp + nb200nf_nri] mov edx, ecx sub ecx, ebx cmovle ebx, edx ;# if(nn1>nri) nn1=nri ;# Cleared the spinlock if we got here. ;# eax contains nn0, ebx contains nn1. mov [esp + nb200nf_n], eax mov [esp + nb200nf_nn1], ebx sub ebx, eax ;# calc number of outer lists mov esi, eax ;# copy n to esi jg .nb200nf_outerstart jmp .nb200nf_end.nb200nf_outerstart: ;# ebx contains number of outer iterations add ebx, [esp + nb200nf_nouter] mov [esp + nb200nf_nouter], ebx.nb200nf_outer: mov eax, [ebp + nb200nf_shift] ;# eax = pointer into shift[] mov ebx, [eax+esi*4] ;# ebx=shift[n] lea ebx, [ebx + ebx*2] ;# ebx=3*is mov eax, [ebp + nb200nf_shiftvec] ;# eax = base of shiftvec[] movsd xmm0, [eax + ebx*8] movsd xmm1, [eax + ebx*8 + 8] movsd xmm2, [eax + ebx*8 + 16] mov ecx, [ebp + nb200nf_iinr] ;# ecx = pointer into iinr[] mov ebx, [ecx+esi*4] ;# ebx =ii mov edx, [ebp + nb200nf_charge] movsd xmm3, [edx + ebx*8] mulsd xmm3, [esp + nb200nf_facel] shufpd xmm3, xmm3, 0 lea ebx, [ebx + ebx*2] ;# ebx = 3*ii=ii3 mov eax, [ebp + nb200nf_pos] ;# eax = base of pos[] addsd xmm0, [eax + ebx*8] addsd xmm1, [eax + ebx*8 + 8] addsd xmm2, [eax + ebx*8 + 16] movapd [esp + nb200nf_iq], xmm3 shufpd xmm0, xmm0, 0 shufpd xmm1, xmm1, 0 shufpd xmm2, xmm2, 0 movapd [esp + nb200nf_ix], xmm0 movapd [esp + nb200nf_iy], xmm1 movapd [esp + nb200nf_iz], xmm2 mov [esp + nb200nf_ii3], ebx ;# clear vctot xorpd xmm4, xmm4 movapd [esp + nb200nf_vctot], xmm4 mov eax, [ebp + nb200nf_jindex] mov ecx, [eax + esi*4] ;# jindex[n] mov edx, [eax + esi*4 + 4] ;# jindex[n+1] sub edx, ecx ;# number of innerloop atoms mov esi, [ebp + nb200nf_pos] mov eax, [ebp + nb200nf_jjnr] shl ecx, 2 add eax, ecx mov [esp + nb200nf_innerjjnr], eax ;# pointer to jjnr[nj0] mov ecx, edx sub edx, 2 add ecx, [esp + nb200nf_ninner] mov [esp + nb200nf_ninner], ecx add edx, 0 mov [esp + nb200nf_innerk], edx ;# number of innerloop atoms jge .nb200nf_unroll_loop jmp .nb200nf_checksingle.nb200nf_unroll_loop: ;# twice unrolled innerloop here mov edx, [esp + nb200nf_innerjjnr] ;# pointer to jjnr[k] mov eax, [edx] mov ebx, [edx + 4] add dword ptr [esp + nb200nf_innerjjnr], 8 ;# advance pointer (unrolled 2) mov esi, [ebp + nb200nf_charge] ;# base of charge[] movlpd xmm3, [esi + eax*8] movhpd xmm3, [esi + ebx*8] movapd xmm5, [esp + nb200nf_iq] mulpd xmm3, xmm5 ;# qq lea eax, [eax + eax*2] ;# replace jnr with j3 lea ebx, [ebx + ebx*2] mov esi, [ebp + nb200nf_pos] ;# base of pos[] ;# move two coordinates to xmm0-xmm2 movlpd xmm0, [esi + eax*8] movlpd xmm1, [esi + eax*8 + 8] movlpd xmm2, [esi + eax*8 + 16] movhpd xmm0, [esi + ebx*8] movhpd xmm1, [esi + ebx*8 + 8] movhpd xmm2, [esi + ebx*8 + 16] ;# move ix-iz to xmm4-xmm6 movapd xmm4, [esp + nb200nf_ix] movapd xmm5, [esp + nb200nf_iy] movapd xmm6, [esp + nb200nf_iz] ;# calc dr subpd xmm4, xmm0 subpd xmm5, xmm1 subpd xmm6, xmm2 ;# square it mulpd xmm4,xmm4 mulpd xmm5,xmm5 mulpd xmm6,xmm6 addpd xmm4, xmm5 addpd xmm4, xmm6 ;# rsq in xmm4 cvtpd2ps xmm5, xmm4 rsqrtps xmm5, xmm5 cvtps2pd xmm2, xmm5 ;# lu in low xmm2 movapd xmm7, [esp + nb200nf_krf] ;# lookup seed in xmm2 movapd xmm5, xmm2 ;# copy of lu mulpd xmm2, xmm2 ;# lu*lu movapd xmm1, [esp + nb200nf_three] mulpd xmm7, xmm4 ;# krsq mulpd xmm2, xmm4 ;# rsq*lu*lu movapd xmm0, [esp + nb200nf_half] subpd xmm1, xmm2 ;# 30-rsq*lu*lu mulpd xmm1, xmm5 mulpd xmm1, xmm0 ;# xmm0=iter1 of rinv (new lu) movapd xmm5, xmm1 ;# copy of lu mulpd xmm1, xmm1 ;# lu*lu movapd xmm2, [esp + nb200nf_three] mulpd xmm1, xmm4 ;# rsq*lu*lu movapd xmm0, [esp + nb200nf_half] subpd xmm2, xmm1 ;# 30-rsq*lu*lu mulpd xmm2, xmm5 mulpd xmm0, xmm2 ;# xmm0=rinv movapd xmm4, xmm0 mulpd xmm4, xmm4 ;# xmm4=rinvsq movapd xmm6, xmm0 addpd xmm6, xmm7 ;# xmm6=rinv+ krsq movapd xmm1, xmm4 subpd xmm6, [esp + nb200nf_crf] mulpd xmm6, xmm3 ;# xmm6=vcoul=qq*(rinv+ krsq) addpd xmm6, [esp + nb200nf_vctot] movapd [esp + nb200nf_vctot], xmm6 ;# should we do one more iteration? sub dword ptr [esp + nb200nf_innerk], 2 jl .nb200nf_checksingle jmp .nb200nf_unroll_loop.nb200nf_checksingle: mov edx, [esp + nb200nf_innerk] and edx, 1 jnz .nb200nf_dosingle jmp .nb200nf_updateouterdata.nb200nf_dosingle: mov esi, [ebp + nb200nf_charge] mov edi, [ebp + nb200nf_pos] mov ecx, [esp + nb200nf_innerjjnr] xorpd xmm3, xmm3 mov eax, [ecx] movlpd xmm3, [esi + eax*8] movapd xmm5, [esp + nb200nf_iq] mulpd xmm3, xmm5 ;# qq mov esi, [ebp + nb200nf_pos] ;# base of pos[] lea eax, [eax + eax*2] ;# replace jnr with j3 ;# move two coordinates to xmm0-xmm2 movlpd xmm0, [esi + eax*8] movlpd xmm1, [esi + eax*8 + 8] movlpd xmm2, [esi + eax*8 + 16] ;# move ix-iz to xmm4-xmm6 movapd xmm4, [esp + nb200nf_ix] movapd xmm5, [esp + nb200nf_iy] movapd xmm6, [esp + nb200nf_iz] ;# calc dr subsd xmm4, xmm0 subsd xmm5, xmm1 subsd xmm6, xmm2 ;# square it mulsd xmm4,xmm4 mulsd xmm5,xmm5 mulsd xmm6,xmm6 addsd xmm4, xmm5 addsd xmm4, xmm6 ;# rsq in xmm4 cvtsd2ss xmm5, xmm4 rsqrtss xmm5, xmm5 cvtss2sd xmm2, xmm5 ;# lu in low xmm2 movapd xmm7, [esp + nb200nf_krf] ;# lookup seed in xmm2 movapd xmm5, xmm2 ;# copy of lu mulsd xmm2, xmm2 ;# lu*lu movapd xmm1, [esp + nb200nf_three] mulsd xmm7, xmm4 ;# krsq mulsd xmm2, xmm4 ;# rsq*lu*lu movapd xmm0, [esp + nb200nf_half] subsd xmm1, xmm2 ;# 30-rsq*lu*lu mulsd xmm1, xmm5 mulsd xmm1, xmm0 ;# xmm0=iter1 of rinv (new lu) movapd xmm5, xmm1 ;# copy of lu mulsd xmm1, xmm1 ;# lu*lu movapd xmm2, [esp + nb200nf_three] mulsd xmm1, xmm4 ;# rsq*lu*lu movapd xmm0, [esp + nb200nf_half] subsd xmm2, xmm1 ;# 30-rsq*lu*lu mulsd xmm2, xmm5 mulsd xmm0, xmm2 ;# xmm0=rinv movapd xmm4, xmm0 mulsd xmm4, xmm4 ;# xmm4=rinvsq movapd xmm6, xmm0 addsd xmm6, xmm7 ;# xmm6=rinv+ krsq movapd xmm1, xmm4 subsd xmm6, [esp + nb200nf_crf] mulsd xmm1, xmm4 mulsd xmm1, xmm4 ;# xmm1=rinvsix movapd xmm2, xmm1 mulsd xmm2, xmm2 ;# xmm2=rinvtwelve mulsd xmm6, xmm3 ;# xmm6=vcoul=qq*(rinv+ krsq) addsd xmm6, [esp + nb200nf_vctot] movlpd [esp + nb200nf_vctot], xmm6.nb200nf_updateouterdata: ;# get n from stack mov esi, [esp + nb200nf_n] ;# get group index for i particle mov edx, [ebp + nb200nf_gid] ;# base of gid[] mov edx, [edx + esi*4] ;# ggid=gid[n] ;# accumulate total potential energy and update it movapd xmm7, [esp + nb200nf_vctot] ;# accumulate movhlps xmm6, xmm7 addsd xmm7, xmm6 ;# low xmm7 has the sum now ;# add earlier value from mem mov eax, [ebp + nb200nf_Vc] addsd xmm7, [eax + edx*8] ;# move back to mem movsd [eax + edx*8], xmm7 ;# finish if last mov ecx, [esp + nb200nf_nn1] ;# esi already loaded with n inc esi sub ecx, esi jz .nb200nf_outerend ;# not last, iterate outer loop once more! mov [esp + nb200nf_n], esi jmp .nb200nf_outer.nb200nf_outerend: ;# check if more outer neighborlists remain mov ecx, [esp + nb200nf_nri] ;# esi already loaded with n above sub ecx, esi jz .nb200nf_end ;# non-zero, do one more workunit jmp .nb200nf_threadloop.nb200nf_end: emms mov eax, [esp + nb200nf_nouter] mov ebx, [esp + nb200nf_ninner] mov ecx, [ebp + nb200nf_outeriter] mov edx, [ebp + nb200nf_inneriter] mov [ecx], eax mov [edx], ebx mov eax, [esp + nb200nf_salign] add esp, eax add esp, 172 pop edi pop esi pop edx pop ecx pop ebx pop eax leave ret⌨️ 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?