nb_kernel214_ia32_sse2.intel_syntax.s

来自「最著名最快的分子模拟软件」· S 代码 · 共 2,137 行 · 第 1/5 页

	movapd xmm0, [esp + nb214_rsqH2H1]
	movapd xmm4, [esp + nb214_rsqH2H2]	
	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 + nb214_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 + nb214_half] ;# iter1a 
	mulsd   xmm7, [esp + nb214_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 + nb214_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 + nb214_half] ;# rinv 
	mulsd   xmm5, [esp + nb214_half] ;# rinv 
	movapd [esp + nb214_rinvH2H1], xmm1
	movapd [esp + nb214_rinvH2H2], xmm5

	movapd xmm0, [esp + nb214_rsqMH1]
	movapd xmm4, [esp + nb214_rsqH2M]	
	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 + nb214_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 + nb214_half] ;# iter1a 
	mulsd   xmm7, [esp + nb214_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 + nb214_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 + nb214_half] ;# rinv 
	mulsd   xmm5, [esp + nb214_half] ;# rinv 
	movapd [esp + nb214_rinvMH1], xmm1
	movapd [esp + nb214_rinvH2M], xmm5

	;# start with OO interaction 
	movsd xmm0, [esp + nb214_rinvsqOO] ;# xmm0=rinvsq
	movapd  xmm1, xmm0	
	mulsd   xmm1, xmm1 ;# rinv4
	mulsd   xmm1, xmm0 ;#rinvsix
	movsd  xmm2, xmm1
	mulsd	xmm2, xmm2 ;# rinvtwelve
	mulsd  xmm1, [esp + nb214_c6]
	mulsd  xmm2, [esp + nb214_c12]
	movsd xmm3, xmm2
	subsd  xmm3, xmm1	;# Vvdw=Vvdw12-Vvdw6 		
	addsd  xmm3, [esp + nb214_Vvdwtot]
	mulsd  xmm1, [esp + nb214_six]
	mulsd  xmm2, [esp + nb214_twelve]
	subsd  xmm2, xmm1
	mulsd  xmm2, xmm0
	movsd [esp + nb214_Vvdwtot], xmm3

	movapd xmm0, xmm2
	movapd xmm1, xmm2

	
	xorpd xmm3, xmm3
	movapd xmm4, xmm3
	movapd xmm5, xmm3
	mulsd xmm0, [esp + nb214_dxOO]
	mulsd xmm1, [esp + nb214_dyOO]
	mulsd xmm2, [esp + nb214_dzOO]
	subsd xmm3, xmm0
	subsd xmm4, xmm1
	subsd xmm5, xmm2
	addsd xmm0, [esp + nb214_fixO]
	addsd xmm1, [esp + nb214_fiyO]
	addsd xmm2, [esp + nb214_fizO]
	movsd [esp + nb214_fjxO], xmm3
	movsd [esp + nb214_fjyO], xmm4
	movsd [esp + nb214_fjzO], xmm5
	movsd [esp + nb214_fixO], xmm0
	movsd [esp + nb214_fiyO], xmm1
	movsd [esp + nb214_fizO], xmm2

	;# H1-H1 interaction 
	movsd xmm0, [esp + nb214_rinvH1H1]
	movsd xmm7, xmm0	;# xmm7=rinv 
	movsd xmm5, [esp + nb214_krf]
	mulsd  xmm0, xmm0	;# xmm0=rinvsq 

	mulsd  xmm5, [esp + nb214_rsqH1H1] ;# xmm5=krsq 
	movsd xmm6, xmm5
	addsd  xmm6, xmm7	;# xmm6=rinv+ krsq 
	subsd  xmm6, [esp + nb214_crf]
	mulsd  xmm6, [esp + nb214_qqHH] ;# xmm6=voul=qq*(rinv+ krsq-crf) 
	mulsd  xmm5, [esp + nb214_two]
	subsd  xmm7, xmm5	;# xmm7=rinv-2*krsq 
	mulsd  xmm7, [esp + nb214_qqHH] ;# xmm7 = coul part of fscal 
	
	addsd  xmm6, [esp + nb214_vctot] ;# local vctot summation variable 
	mulsd  xmm0, xmm7
	movsd [esp + nb214_vctot], xmm6
	
	movapd xmm1, xmm0
	movapd xmm2, xmm0
	
	xorpd xmm3, xmm3
	movapd xmm4, xmm3
	movapd xmm5, xmm3
	mulsd xmm0, [esp + nb214_dxH1H1]
	mulsd xmm1, [esp + nb214_dyH1H1]
	mulsd xmm2, [esp + nb214_dzH1H1]
	subsd xmm3, xmm0
	subsd xmm4, xmm1
	subsd xmm5, xmm2
	addsd xmm0, [esp + nb214_fixH1]
	addsd xmm1, [esp + nb214_fiyH1]
	addsd xmm2, [esp + nb214_fizH1]
	movsd [esp + nb214_fjxH1], xmm3
	movsd [esp + nb214_fjyH1], xmm4
	movsd [esp + nb214_fjzH1], xmm5
	movsd [esp + nb214_fixH1], xmm0
	movsd [esp + nb214_fiyH1], xmm1
	movsd [esp + nb214_fizH1], xmm2

	;# H1-H2 interaction  
	movsd xmm0, [esp + nb214_rinvH1H2]
	movsd xmm7, xmm0	;# xmm7=rinv 
	movsd xmm5, [esp + nb214_krf]
	mulsd  xmm0, xmm0	;# xmm0=rinvsq 

	mulsd  xmm5, [esp + nb214_rsqH1H2] ;# xmm5=krsq 
	movsd xmm6, xmm5
	addsd  xmm6, xmm7	;# xmm6=rinv+ krsq 
	subsd  xmm6, [esp + nb214_crf]
	mulsd  xmm6, [esp + nb214_qqHH] ;# xmm6=voul=qq*(rinv+ krsq-crf) 
	mulsd  xmm5, [esp + nb214_two]
	subsd  xmm7, xmm5	;# xmm7=rinv-2*krsq 
	mulsd  xmm7, [esp + nb214_qqHH] ;# xmm7 = coul part of fscal 
	
	addsd  xmm6, [esp + nb214_vctot] ;# local vctot summation variable 
	mulsd  xmm0, xmm7
	movsd [esp + nb214_vctot], xmm6
	
	movapd xmm1, xmm0
	movapd xmm2, xmm0
	
	xorpd xmm3, xmm3
	movapd xmm4, xmm3
	movapd xmm5, xmm3
	mulsd xmm0, [esp + nb214_dxH1H2]
	mulsd xmm1, [esp + nb214_dyH1H2]
	mulsd xmm2, [esp + nb214_dzH1H2]
	subsd xmm3, xmm0
	subsd xmm4, xmm1
	subsd xmm5, xmm2
	addsd xmm0, [esp + nb214_fixH1]
	addsd xmm1, [esp + nb214_fiyH1]
	addsd xmm2, [esp + nb214_fizH1]
	movsd [esp + nb214_fjxH2], xmm3
	movsd [esp + nb214_fjyH2], xmm4
	movsd [esp + nb214_fjzH2], xmm5
	movsd [esp + nb214_fixH1], xmm0
	movsd [esp + nb214_fiyH1], xmm1
	movsd [esp + nb214_fizH1], xmm2

	;# H1-M interaction 
	movsd xmm0, [esp + nb214_rinvH1M]
	movsd xmm7, xmm0	;# xmm7=rinv 
	movsd xmm5, [esp + nb214_krf]
	mulsd  xmm0, xmm0	;# xmm0=rinvsq 

	mulsd  xmm5, [esp + nb214_rsqH1M] ;# xmm5=krsq 
	movsd xmm6, xmm5
	addsd  xmm6, xmm7	;# xmm6=rinv+ krsq 
	subsd  xmm6, [esp + nb214_crf]
	mulsd  xmm6, [esp + nb214_qqMH] ;# xmm6=voul=qq*(rinv+ krsq-crf) 
	mulsd  xmm5, [esp + nb214_two]
	subsd  xmm7, xmm5	;# xmm7=rinv-2*krsq 
	mulsd  xmm7, [esp + nb214_qqMH] ;# xmm7 = coul part of fscal 
	
	addsd  xmm6, [esp + nb214_vctot] ;# local vctot summation variable 
	mulsd  xmm0, xmm7
	movsd [esp + nb214_vctot], xmm6

	movapd xmm1, xmm0
	movapd xmm2, xmm0
	
	xorpd xmm3, xmm3
	movapd xmm4, xmm3
	movapd xmm5, xmm3
	mulsd xmm0, [esp + nb214_dxH1M]
	mulsd xmm1, [esp + nb214_dyH1M]
	mulsd xmm2, [esp + nb214_dzH1M]
	subsd xmm3, xmm0
	subsd xmm4, xmm1
	subsd xmm5, xmm2
	addsd xmm0, [esp + nb214_fixH1]
	addsd xmm1, [esp + nb214_fiyH1]
	addsd xmm2, [esp + nb214_fizH1]
	movsd [esp + nb214_fjxM], xmm3
	movsd [esp + nb214_fjyM], xmm4
	movsd [esp + nb214_fjzM], xmm5
	movsd [esp + nb214_fixH1], xmm0
	movsd [esp + nb214_fiyH1], xmm1
	movsd [esp + nb214_fizH1], xmm2

	;# H2-H1 interaction 
	movsd xmm0, [esp + nb214_rinvH2H1]
	movsd xmm7, xmm0	;# xmm7=rinv 
	movsd xmm5, [esp + nb214_krf]
	mulsd  xmm0, xmm0	;# xmm0=rinvsq 

	mulsd  xmm5, [esp + nb214_rsqH2H1] ;# xmm5=krsq 
	movsd xmm6, xmm5
	addsd  xmm6, xmm7	;# xmm6=rinv+ krsq 
	subsd  xmm6, [esp + nb214_crf]
	mulsd  xmm6, [esp + nb214_qqHH] ;# xmm6=voul=qq*(rinv+ krsq-crf) 
	mulsd  xmm5, [esp + nb214_two]
	subsd  xmm7, xmm5	;# xmm7=rinv-2*krsq 
	mulsd  xmm7, [esp + nb214_qqHH] ;# xmm7 = coul part of fscal 
	
	addsd  xmm6, [esp + nb214_vctot] ;# local vctot summation variable 
	mulsd  xmm0, xmm7
	movsd [esp + nb214_vctot], xmm6
	
	movapd xmm1, xmm0
	movapd xmm2, xmm0
	
	movapd xmm3, [esp + nb214_fjxH1]
	movapd xmm4, [esp + nb214_fjyH1]
	movapd xmm5, [esp + nb214_fjzH1]
	mulsd xmm0, [esp + nb214_dxH2H1]
	mulsd xmm1, [esp + nb214_dyH2H1]
	mulsd xmm2, [esp + nb214_dzH2H1]
	subsd xmm3, xmm0
	subsd xmm4, xmm1
	subsd xmm5, xmm2
	addsd xmm0, [esp + nb214_fixH2]
	addsd xmm1, [esp + nb214_fiyH2]
	addsd xmm2, [esp + nb214_fizH2]
	movsd [esp + nb214_fjxH1], xmm3
	movsd [esp + nb214_fjyH1], xmm4
	movsd [esp + nb214_fjzH1], xmm5
	movsd [esp + nb214_fixH2], xmm0
	movsd [esp + nb214_fiyH2], xmm1
	movsd [esp + nb214_fizH2], xmm2

	;# H2-H2 interaction 
	movsd xmm0, [esp + nb214_rinvH2H2]
	movsd xmm7, xmm0	;# xmm7=rinv 
	movsd xmm5, [esp + nb214_krf]
	mulsd  xmm0, xmm0	;# xmm0=rinvsq 

	mulsd  xmm5, [esp + nb214_rsqH2H2] ;# xmm5=krsq 
	movsd xmm6, xmm5
	addsd  xmm6, xmm7	;# xmm6=rinv+ krsq 
	subsd  xmm6, [esp + nb214_crf]
	mulsd  xmm6, [esp + nb214_qqHH] ;# xmm6=voul=qq*(rinv+ krsq-crf) 
	mulsd  xmm5, [esp + nb214_two]
	subsd  xmm7, xmm5	;# xmm7=rinv-2*krsq 
	mulsd  xmm7, [esp + nb214_qqHH] ;# xmm7 = coul part of fscal 
	
	addsd  xmm6, [esp + nb214_vctot] ;# local vctot summation variable 
	mulsd  xmm0, xmm7
	movsd [esp + nb214_vctot], xmm6
	
	movsd xmm1, xmm0
	movsd xmm2, xmm0

	movsd xmm3, [esp + nb214_fjxH2]
	movsd xmm4, [esp + nb214_fjyH2]
	movsd xmm5, [esp + nb214_fjzH2]
	mulsd xmm0, [esp + nb214_dxH2H2]
	mulsd xmm1, [esp + nb214_dyH2H2]
	mulsd xmm2, [esp + nb214_dzH2H2]
	subsd xmm3, xmm0
	subsd xmm4, xmm1
	subsd xmm5, xmm2
	addsd xmm0, [esp + nb214_fixH2]
	addsd xmm1, [esp + nb214_fiyH2]
	addsd xmm2, [esp + nb214_fizH2]
	movsd [esp + nb214_fjxH2], xmm3
	movsd [esp + nb214_fjyH2], xmm4
	movsd [esp + nb214_fjzH2], xmm5
	movsd [esp + nb214_fixH2], xmm0
	movsd [esp + nb214_fiyH2], xmm1
	movsd [esp + nb214_fizH2], xmm2

	;# H2-M interaction 
	movsd xmm0, [esp + nb214_rinvH2M]
	movsd xmm7, xmm0	;# xmm7=rinv 
	movsd xmm5, [esp + nb214_krf]
	mulsd  xmm0, xmm0	;# xmm0=rinvsq 

	mulsd  xmm5, [esp + nb214_rsqH2M] ;# xmm5=krsq 
	movsd xmm6, xmm5
	addsd  xmm6, xmm7	;# xmm6=rinv+ krsq 
	subsd  xmm6, [esp + nb214_crf]
	mulsd  xmm6, [esp + nb214_qqMH] ;# xmm6=voul=qq*(rinv+ krsq-crf) 
	mulsd  xmm5, [esp + nb214_two]
	subsd  xmm7, xmm5	;# xmm7=rinv-2*krsq 
	mulsd  xmm7, [esp + nb214_qqMH] ;# xmm7 = coul part of fscal 
	
	addsd  xmm6, [esp + nb214_vctot] ;# local vctot summation variable 
	mulsd  xmm0, xmm7
	movsd [esp + nb214_vctot], xmm6
	
	movapd xmm1, xmm0
	movapd xmm2, xmm0

	movapd xmm3, [esp + nb214_fjxM]
	movapd xmm4, [esp + nb214_fjyM]
	movapd xmm5, [esp + nb214_fjzM]
	mulsd xmm0, [esp + nb214_dxH2M]
	mulsd xmm1, [esp + nb214_dyH2M]
	mulsd xmm2, [esp + nb214_dzH2M]
	subsd xmm3, xmm0
	subsd xmm4, xmm1
	subsd xmm5, xmm2
	addsd xmm0, [esp + nb214_fixH2]
	addsd xmm1, [esp + nb214_fiyH2]
	addsd xmm2, [esp + nb214_fizH2]
	movsd [esp + nb214_fjxM], xmm3
	movsd [esp + nb214_fjyM], xmm4
	movsd [esp + nb214_fjzM], xmm5
	movsd [esp + nb214_fixH2], xmm0
	movsd [esp + nb214_fiyH2], xmm1
	movsd [esp + nb214_fizH2], xmm2

	;# M-H1 interaction 
	movsd xmm0, [esp + nb214_rinvMH1]
	movsd xmm7, xmm0	;# xmm7=rinv 
	movsd xmm5, [esp + nb214_krf]
	mulsd  xmm0, xmm0	;# xmm0=rinvsq 

	mulsd  xmm5, [esp + nb214_rsqMH1] ;# xmm5=krsq 
	movsd xmm6, xmm5
	addsd  xmm6, xmm7	;# xmm6=rinv+ krsq 
	subsd  xmm6, [esp + nb214_crf]
	mulsd  xmm6, [esp + nb214_qqMH] ;# xmm6=voul=qq*(rinv+ krsq-crf) 
	mulsd  xmm5, [esp + nb214_two]
	subsd  xmm7, xmm5	;# xmm7=rinv-2*krsq 
	mulsd  xmm7, [esp + nb214_qqMH] ;# xmm7 = coul part of fscal 
	
	addsd  xmm6, [esp + nb214_vctot] ;# local vctot summation variable 
	mulsd  xmm0, xmm7
	movsd [esp + nb214_vctot], xmm6
	
	movapd xmm1, xmm0
	movapd xmm2, xmm0

	movapd xmm3, [esp + nb214_fjxH1]
	movapd xmm4, [esp + nb214_fjyH1]
	movapd xmm5, [esp + nb214_fjzH1]
	mulsd xmm0, [esp + nb214_dxMH1]
	mulsd xmm1, [esp + nb214_dyMH1]
	mulsd xmm2, [esp + nb214_dzMH1]
	subsd xmm3, xmm0
	subsd xmm4, xmm1
	subsd xmm5, xmm2
	addsd xmm0, [esp + nb214_fixM]
	addsd xmm1, [esp + nb214_fiyM]
	addsd xmm2, [esp + nb214_fizM]
	movsd [esp + nb214_fjxH1], xmm3
	movsd [esp + nb214_fjyH1], xmm4
	movsd [esp + nb214_fjzH1], xmm5
	movsd [esp + nb214_fixM], xmm0
	movsd [esp + nb214_fiyM], xmm1
	movsd [esp + nb214_fizM], xmm2

	;# M-H2 interaction 
	movsd xmm0, [esp + nb214_rinvMH2]
	movsd xmm7, xmm0	;# xmm7=rinv 
	movsd xmm5, [esp + nb214_krf]
	mulsd  xmm0, xmm0	;# xmm0=rinvsq 

	mulsd  xmm5, [esp + nb214_rsqMH2] ;# xmm5=krsq 
	movsd xmm6, xmm5
	addsd  xmm6, xmm7	;# xmm6=rinv+ krsq 
	subsd  xmm6, [esp + nb214_crf]
	mulsd  xmm6, [esp + nb214_qqMH] ;# xmm6=voul=qq*(rinv+ krsq-crf) 
	mulsd  xmm5, [esp + nb214_two]
	subsd  xmm7, xmm5	;# xmm7=rinv-2*krsq 
	mulsd  xmm7, [esp + nb214_qqMH] ;# xmm7 = coul part of fscal