nb_kernel214_ia32_sse.s

来自「最著名最快的分子模拟软件」· S 代码 · 共 1,984 行 · 第 1/5 页

S
1,984 行
字号
12345
        movaps %xmm5,%xmm6        addps  %xmm7,%xmm6      ## xmm6=rinv+ krsq         subps  nb214_crf(%esp),%xmm6        mulps  nb214_qqHH(%esp),%xmm6   ## xmm6=voul=qq*(rinv+ krsq-crf)         mulps nb214_two(%esp),%xmm5        subps  %xmm5,%xmm7      ## xmm7=rinv-2*krsq         mulps  nb214_qqHH(%esp),%xmm7   ## xmm7 = coul part of fscal         addps  nb214_vctot(%esp),%xmm6   ## local vctot summation variable         mulps  %xmm7,%xmm0        movaps %xmm0,%xmm1        movaps %xmm0,%xmm2        xorps %xmm3,%xmm3        movaps %xmm3,%xmm4        movaps %xmm3,%xmm5        mulps nb214_dxH1H1(%esp),%xmm0        mulps nb214_dyH1H1(%esp),%xmm1        mulps nb214_dzH1H1(%esp),%xmm2        subps %xmm0,%xmm3        subps %xmm1,%xmm4        subps %xmm2,%xmm5        addps nb214_fixH1(%esp),%xmm0        addps nb214_fiyH1(%esp),%xmm1        addps nb214_fizH1(%esp),%xmm2        movaps %xmm3,nb214_fjxH1(%esp)        movaps %xmm4,nb214_fjyH1(%esp)        movaps %xmm5,nb214_fjzH1(%esp)        movaps %xmm0,nb214_fixH1(%esp)        movaps %xmm1,nb214_fiyH1(%esp)        movaps %xmm2,nb214_fizH1(%esp)        ## H1-H2 interaction         movaps nb214_rinvH1H2(%esp),%xmm0        movaps %xmm0,%xmm7      ## xmm7=rinv         movaps nb214_krf(%esp),%xmm5        movaps %xmm0,%xmm1        mulps  nb214_rsqH1H2(%esp),%xmm5   ## xmm5=krsq         movaps %xmm5,%xmm4        addps  %xmm7,%xmm4      ## xmm4=r inv+ krsq         subps  nb214_crf(%esp),%xmm4        mulps  %xmm0,%xmm0        mulps  nb214_qqHH(%esp),%xmm4   ## xmm4=voul=qq*(rinv+ krsq-crf)         mulps  nb214_two(%esp),%xmm5        subps  %xmm5,%xmm7      ## xmm7=rinv-2*krsq         mulps  nb214_qqHH(%esp),%xmm7   ## xmm7 = coul part of fscal         addps  %xmm4,%xmm6      ## add to local vctot         mulps %xmm7,%xmm0       ## fsOH1          movaps %xmm0,%xmm1        movaps %xmm0,%xmm2        xorps %xmm3,%xmm3        movaps %xmm3,%xmm4        movaps %xmm3,%xmm5        mulps nb214_dxH1H2(%esp),%xmm0        mulps nb214_dyH1H2(%esp),%xmm1        mulps nb214_dzH1H2(%esp),%xmm2        subps %xmm0,%xmm3        subps %xmm1,%xmm4        subps %xmm2,%xmm5        addps nb214_fixH1(%esp),%xmm0        addps nb214_fiyH1(%esp),%xmm1        addps nb214_fizH1(%esp),%xmm2        movaps %xmm3,nb214_fjxH2(%esp)        movaps %xmm4,nb214_fjyH2(%esp)        movaps %xmm5,nb214_fjzH2(%esp)        movaps %xmm0,nb214_fixH1(%esp)        movaps %xmm1,nb214_fiyH1(%esp)        movaps %xmm2,nb214_fizH1(%esp)        ## H1-M interaction          movaps nb214_rinvH1M(%esp),%xmm0        movaps %xmm0,%xmm7      ## xmm7=Rinv         movaps nb214_krf(%esp),%xmm5        movaps %xmm0,%xmm1        mulps  nb214_rsqH1M(%esp),%xmm5   ## xmm5=krsq         movaps %xmm5,%xmm4        addps  %xmm7,%xmm4      ## xmm4=r inv+ krsq         subps  nb214_crf(%esp),%xmm4        mulps %xmm0,%xmm0        mulps  nb214_qqMH(%esp),%xmm4   ## xmm4=voul=qq*(rinv+ krsq-crf)         mulps  nb214_two(%esp),%xmm5        subps  %xmm5,%xmm7      ## xmm7=rinv-2*krsq         mulps  nb214_qqMH(%esp),%xmm7   ## xmm7 = coul part of fscal         addps  %xmm4,%xmm6      ## add to local vctot         mulps %xmm7,%xmm0       ## fsOH2         movaps %xmm0,%xmm1        movaps %xmm0,%xmm2        xorps %xmm3,%xmm3        movaps %xmm3,%xmm4        movaps %xmm3,%xmm5        mulps nb214_dxH1M(%esp),%xmm0        mulps nb214_dyH1M(%esp),%xmm1        mulps nb214_dzH1M(%esp),%xmm2        subps %xmm0,%xmm3        subps %xmm1,%xmm4        subps %xmm2,%xmm5        addps nb214_fixH1(%esp),%xmm0        addps nb214_fiyH1(%esp),%xmm1        addps nb214_fizH1(%esp),%xmm2        movaps %xmm3,nb214_fjxM(%esp)        movaps %xmm4,nb214_fjyM(%esp)        movaps %xmm5,nb214_fjzM(%esp)        movaps %xmm0,nb214_fixH1(%esp)        movaps %xmm1,nb214_fiyH1(%esp)        movaps %xmm2,nb214_fizH1(%esp)        ## H2-H1 interaction         movaps nb214_rinvH2H1(%esp),%xmm0        movaps %xmm0,%xmm7      ## xmm7=rinv         movaps nb214_krf(%esp),%xmm5        movaps %xmm0,%xmm1        mulps  nb214_rsqH2H1(%esp),%xmm5   ## xmm5=krsq         movaps %xmm5,%xmm4        addps  %xmm7,%xmm4      ## xmm4=r inv+ krsq         subps  nb214_crf(%esp),%xmm4        mulps %xmm0,%xmm0        mulps  nb214_qqHH(%esp),%xmm4   ## xmm4=voul=qq*(rinv+ krsq-crf)         mulps  nb214_two(%esp),%xmm5        subps  %xmm5,%xmm7      ## xmm7=rinv-2*krsq         mulps  nb214_qqHH(%esp),%xmm7   ## xmm7 = coul part of fscal         addps  %xmm4,%xmm6      ## add to local vctot         mulps %xmm7,%xmm0       ## fsOH2         movaps %xmm0,%xmm1        movaps %xmm0,%xmm2        movaps nb214_fjxH1(%esp),%xmm3        movaps nb214_fjyH1(%esp),%xmm4        movaps nb214_fjzH1(%esp),%xmm5        mulps nb214_dxH2H1(%esp),%xmm0        mulps nb214_dyH2H1(%esp),%xmm1        mulps nb214_dzH2H1(%esp),%xmm2        subps %xmm0,%xmm3        subps %xmm1,%xmm4        subps %xmm2,%xmm5        addps nb214_fixH2(%esp),%xmm0        addps nb214_fiyH2(%esp),%xmm1        addps nb214_fizH2(%esp),%xmm2        movaps %xmm3,nb214_fjxH1(%esp)        movaps %xmm4,nb214_fjyH1(%esp)        movaps %xmm5,nb214_fjzH1(%esp)        movaps %xmm0,nb214_fixH2(%esp)        movaps %xmm1,nb214_fiyH2(%esp)        movaps %xmm2,nb214_fizH2(%esp)        ## H2-H2 interaction         movaps nb214_rinvH2H2(%esp),%xmm0        movaps %xmm0,%xmm7      ## xmm7=rinv         movaps nb214_krf(%esp),%xmm5        movaps %xmm0,%xmm1        mulps  nb214_rsqH2H2(%esp),%xmm5   ## xmm5=krsq         movaps %xmm5,%xmm4        addps  %xmm7,%xmm4      ## xmm4=r inv+ krsq         subps  nb214_crf(%esp),%xmm4        mulps %xmm0,%xmm0        mulps  nb214_qqHH(%esp),%xmm4   ## xmm4=voul=qq*(rinv+ krsq-crf)         mulps  nb214_two(%esp),%xmm5        subps  %xmm5,%xmm7      ## xmm7=rinv-2*krsq         mulps  nb214_qqHH(%esp),%xmm7   ## xmm7 = coul part of fscal         addps  %xmm4,%xmm6      ## add to local vctot         mulps %xmm7,%xmm0       ## fsOH2         movaps %xmm0,%xmm1        movaps %xmm0,%xmm2        movaps nb214_fjxH2(%esp),%xmm3        movaps nb214_fjyH2(%esp),%xmm4        movaps nb214_fjzH2(%esp),%xmm5        mulps nb214_dxH2H2(%esp),%xmm0        mulps nb214_dyH2H2(%esp),%xmm1        mulps nb214_dzH2H2(%esp),%xmm2        subps %xmm0,%xmm3        subps %xmm1,%xmm4        subps %xmm2,%xmm5        addps nb214_fixH2(%esp),%xmm0        addps nb214_fiyH2(%esp),%xmm1        addps nb214_fizH2(%esp),%xmm2        movaps %xmm3,nb214_fjxH2(%esp)        movaps %xmm4,nb214_fjyH2(%esp)        movaps %xmm5,nb214_fjzH2(%esp)        movaps %xmm0,nb214_fixH2(%esp)        movaps %xmm1,nb214_fiyH2(%esp)        movaps %xmm2,nb214_fizH2(%esp)        ## H2-M interaction         movaps nb214_rinvH2M(%esp),%xmm0        movaps %xmm0,%xmm7      ## xmm7=rinv         movaps nb214_krf(%esp),%xmm5        movaps %xmm0,%xmm1        mulps  nb214_rsqH2M(%esp),%xmm5   ## xmm5=krsq         movaps %xmm5,%xmm4        addps  %xmm7,%xmm4      ## xmm4=r inv+ krsq         subps  nb214_crf(%esp),%xmm4        mulps %xmm0,%xmm0        mulps  nb214_qqMH(%esp),%xmm4   ## xmm4=voul=qq*(rinv+ krsq-crf)         mulps  nb214_two(%esp),%xmm5        subps  %xmm5,%xmm7      ## xmm7=rinv-2*krsq         mulps  nb214_qqMH(%esp),%xmm7   ## xmm7 = coul part of fscal         addps  %xmm4,%xmm6      ## add to local vctot         mulps %xmm7,%xmm0       ## fsOH2         movaps %xmm0,%xmm1        movaps %xmm0,%xmm2        movaps nb214_fjxM(%esp),%xmm3        movaps nb214_fjyM(%esp),%xmm4        movaps nb214_fjzM(%esp),%xmm5        mulps nb214_dxH2M(%esp),%xmm0        mulps nb214_dyH2M(%esp),%xmm1        mulps nb214_dzH2M(%esp),%xmm2        subps %xmm0,%xmm3        subps %xmm1,%xmm4        subps %xmm2,%xmm5        addps nb214_fixH2(%esp),%xmm0        addps nb214_fiyH2(%esp),%xmm1        addps nb214_fizH2(%esp),%xmm2        movaps %xmm3,nb214_fjxM(%esp)        movaps %xmm4,nb214_fjyM(%esp)        movaps %xmm5,nb214_fjzM(%esp)        movaps %xmm0,nb214_fixH2(%esp)        movaps %xmm1,nb214_fiyH2(%esp)        movaps %xmm2,nb214_fizH2(%esp)        ## M-H1 interaction         movaps nb214_rinvMH1(%esp),%xmm0        movaps %xmm0,%xmm7      ## xmm7=rinv         movaps nb214_krf(%esp),%xmm5        movaps %xmm0,%xmm1        mulps  nb214_rsqMH1(%esp),%xmm5   ## xmm5=krsq         movaps %xmm5,%xmm4        addps  %xmm7,%xmm4      ## xmm4=r inv+ krsq         subps  nb214_crf(%esp),%xmm4        mulps %xmm0,%xmm0        mulps  nb214_qqMH(%esp),%xmm4   ## xmm4=voul=qq*(rinv+ krsq-crf)         mulps  nb214_two(%esp),%xmm5        subps  %xmm5,%xmm7      ## xmm7=rinv-2*krsq         mulps  nb214_qqMH(%esp),%xmm7   ## xmm7 = coul part of fscal         addps  %xmm4,%xmm6      ## add to local vctot         mulps %xmm7,%xmm0       ## fsOH2         movaps %xmm0,%xmm1        movaps %xmm0,%xmm2        movaps nb214_fjxH1(%esp),%xmm3        movaps nb214_fjyH1(%esp),%xmm4        movaps nb214_fjzH1(%esp),%xmm5        mulps nb214_dxMH1(%esp),%xmm0        mulps nb214_dyMH1(%esp),%xmm1        mulps nb214_dzMH1(%esp),%xmm2        subps %xmm0,%xmm3        subps %xmm1,%xmm4        subps %xmm2,%xmm5        addps nb214_fixM(%esp),%xmm0        addps nb214_fiyM(%esp),%xmm1        addps nb214_fizM(%esp),%xmm2        movaps %xmm3,nb214_fjxH1(%esp)        movaps %xmm4,nb214_fjyH1(%esp)        movaps %xmm5,nb214_fjzH1(%esp)        movaps %xmm0,nb214_fixM(%esp)        movaps %xmm1,nb214_fiyM(%esp)        movaps %xmm2,nb214_fizM(%esp)        ## M-H2 interaction         movaps nb214_rinvMH2(%esp),%xmm0        movaps %xmm0,%xmm7      ## xmm7=rinv         movaps nb214_krf(%esp),%xmm5        movaps %xmm0,%xmm1        mulps  nb214_rsqMH2(%esp),%xmm5   ## xmm5=krsq         movaps %xmm5,%xmm4        addps  %xmm7,%xmm4      ## xmm4=r inv+ krsq         subps  nb214_crf(%esp),%xmm4        mulps %xmm0,%xmm0        mulps  nb214_qqMH(%esp),%xmm4   ## xmm4=voul=qq*(rinv+ krsq-crf)         mulps  nb214_two(%esp),%xmm5        subps  %xmm5,%xmm7      ## xmm7=rinv-2*krsq         mulps  nb214_qqMH(%esp),%xmm7   ## xmm7 = coul part of fscal         addps  %xmm4,%xmm6      ## add to local vctot         mulps %xmm7,%xmm0       ## fsOH2         movaps %xmm0,%xmm1        movaps %xmm0,%xmm2        movaps nb214_fjxH2(%esp),%xmm3        movaps nb214_fjyH2(%esp),%xmm4        movaps nb214_fjzH2(%esp),%xmm5        mulps nb214_dxMH2(%esp),%xmm0        mulps nb214_dyMH2(%esp),%xmm1        mulps nb214_dzMH2(%esp),%xmm2        subps %xmm0,%xmm3        subps %xmm1,%xmm4        subps %xmm2,%xmm5        addps nb214_fixM(%esp),%xmm0        addps nb214_fiyM(%esp),%xmm1        addps nb214_fizM(%esp),%xmm2        movaps %xmm3,nb214_fjxH2(%esp)        movaps %xmm4,nb214_fjyH2(%esp)        movaps %xmm5,nb214_fjzH2(%esp)        movaps %xmm0,nb214_fixM(%esp)        movaps %xmm1,nb214_fiyM(%esp)        movaps %xmm2,nb214_fizM(%esp)        ## M-M interaction         movaps nb214_rinvMM(%esp),%xmm0        movaps %xmm0,%xmm7      ## xmm7=rinv         movaps nb214_krf(%esp),%xmm5        movaps %xmm0,%xmm1        mulps  nb214_rsqMM(%esp),%xmm5   ## xmm5=krsq         movaps %xmm5,%xmm4        addps  %xmm7,%xmm4      ## xmm4=r inv+ krsq         subps  nb214_crf(%esp),%xmm4        mulps %xmm0,%xmm0        mulps  nb214_qqMM(%esp),%xmm4   ## xmm4=voul=qq*(rinv+ krsq-crf)         mulps  nb214_two(%esp),%xmm5        subps  %xmm5,%xmm7      ## xmm7=rinv-2*krsq         mulps  nb214_qqMM(%esp),%xmm7   ## xmm7 = coul part of fscal         addps  %xmm4,%xmm6      ## add to local vctot         mulps %xmm7,%xmm0       ## fsOH2         movaps %xmm0,%xmm1        movaps %xmm0,%xmm2        movaps %xmm0,%xmm1        movaps %xmm6,nb214_vctot(%esp)        movaps %xmm0,%xmm2        movaps nb214_fjxM(%esp),%xmm3        movaps nb214_fjyM(%esp),%xmm4        movaps nb214_fjzM(%esp),%xmm5        mulps nb214_dxMM(%esp),%xmm0        mulps nb214_dyMM(%esp),%xmm1        mulps nb214_dzMM(%esp),%xmm2        subps %xmm0,%xmm3        subps %xmm1,%xmm4        subps %xmm2,%xmm5        addps nb214_fixM(%esp),%xmm0        addps nb214_fiyM(%esp),%xmm1        addps nb214_fizM(%esp),%xmm2        movaps %xmm3,nb214_fjxM(%esp)        movaps %xmm4,nb214_fjyM(%esp)        movaps %xmm5,nb214_fjzM(%esp)        movaps %xmm0,nb214_fixM(%esp)        movaps %xmm1,nb214_fiyM(%esp)        movaps %xmm2,nb214_fizM(%esp)        movl nb214_faction(%ebp),%edi        ## update j forces         ## 4 j waters with four atoms each.        ## step 1 : transpose fjxO, fjyO, fjzO, fjxH1        movaps nb214_fjxO(%esp),%xmm0        movaps nb214_fjyO(%esp),%xmm1        movaps nb214_fjzO(%esp),%xmm2        movaps nb214_fjxH1(%esp),%xmm3        movaps %xmm0,%xmm4        movaps %xmm1,%xmm5        unpcklps %xmm2,%xmm4        unpcklps %xmm3,%xmm5        unpckhps %xmm2,%xmm0        unpckhps %xmm3,%xmm1        movaps %xmm4,%xmm2        movaps %xmm0,%xmm3        unpcklps %xmm5,%xmm4        unpckhps %xmm5,%xmm2        unpcklps %xmm1,%xmm0        unpckhps %xmm1,%xmm3        ## results are now in xmm4, xmm2, xmm0, xmm3        ## load the corresponding j forces from memory        movlps   (%edi,%eax,4),%xmm1        movlps   (%edi,%ebx,4),%xmm5        movlps   (%edi,%ecx,4),%xmm6        movlps   (%edi,%edx,4),%xmm7        movhps   8(%edi,%eax,4),%xmm1        movhps   8(%edi,%ebx,4),%xmm5        movhps   8(%edi,%ecx,4),%xmm6        movhps   8(%edi,%edx,4),%xmm7        ## add        addps    %xmm4,%xmm1        addps    %xmm2,%xmm5        addps    %xmm0,%xmm6        addps    %xmm3,%xmm7        ## store back        movlps   %xmm1,(%edi,%eax,4)        movlps   %xmm5,(%edi,%ebx,4)        movlps   %xmm6,(%edi,%ecx,4)        movlps   %xmm7,(%edi,%edx,4)        movhps   %xmm1,8(%edi,%eax,4)        movhps   %xmm5,8(%edi,%ebx,4)        movhps   %xmm6,8(%edi,%ecx,4)        movhps   %xmm7,8(%edi,%edx,4)        ## step 2 : transpose fjyH1, fjzH1, fjxH2, fjyH2        movaps nb214_fjyH1(%esp),%xmm0        movaps nb214_fjzH1(%esp),%xmm1        movaps nb214_fjxH2(%esp),%xmm2        movaps nb214_fjyH2(%esp),%xmm3        movaps %xmm0,%xmm4        movaps %xmm1,%xmm5        unpcklps %xmm2,%xmm4        unpcklps %xmm3,%xmm5        unpckhps %xmm2,%xmm0
12345

nb_kernel214_ia32_sse.s - 源码说明

本页面展示了「最著名最快的分子模拟软件」中的 nb_kernel214_ia32_sse.s 源码文件，采用 S 编程语言编写，共 1,984 行代码。您可以在线阅读完整代码内容，也可以返回资源详情页下载完整源码包进行本地学习和开发。

虫虫开发者社区收录了大量与分子模拟相关的技术资源，包括源代码、技术文档、电路图等，是电子工程师和嵌入式开发者的专业学习平台。

⌨️ 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?