atl_smm4x4x16_us.c
来自「基于Blas CLapck的.用过的人知道是干啥的」· C语言 代码 · 共 1,551 行 · 第 1/3 页
C
1,551 行
fadds rC11, m1, rC11 fmuls ra1, rb0, m1 fadds rC21, m2, rC21 fmuls ra2, rb0, m2 fadds rC31, m3, rC31 fmuls ra3, rb0, m3 ldd [pB2+24], rB0 fadds rC00, m0, rC00 fmuls ra0, rb1, m0 fadds rC10, m1, rC10 fmuls ra1, rb1, m1 fadds rC20, m2, rC20 fmuls ra2, rb1, m2 fadds rC30, m3, rC30 fmuls ra3, rb1, m3 ldd [pB3+24], rB1 fadds rC01, m0, rC01 fmuls rA0, rB0, m0 fadds rC11, m1, rC11 fmuls rA1, rB0, m1 fadds rC21, m2, rC21 fmuls rA2, rB0, m2 fadds rC31, m3, rC31 fmuls rA3, rB0, m3 fadds rC02, m0, rC02 fmuls rA0, rB1, m0 fadds rC12, m1, rC12 fmuls rA1, rB1, m1 fadds rC22, m2, rC22 fmuls rA2, rB1, m2 fadds rC32, m3, rC32 fmuls rA3, rB1, m3 fadds rC03, m0, rC03 fmuls ra0, rb0, m0 fadds rC13, m1, rC13 fmuls ra1, rb0, m1 fadds rC23, m2, rC23 fmuls ra2, rb0, m2 fadds rC33, m3, rC33 fmuls ra3, rb0, m3 ldd [pB0+32], rB0 fadds rC02, m0, rC02 fmuls ra0, rb1, m0 ldd [pA0+32], rA0 fadds rC12, m1, rC12 fmuls ra1, rb1, m1 ldd [pA1+32], rA1 fadds rC22, m2, rC22 fmuls ra2, rb1, m2 ldd [pA2+32], rA2 fadds rC32, m3, rC32 fmuls ra3, rb1, m3 ldd [pA3+32], rA3 fadds rC03, m0, rC03 fmuls rA0, rB0, m0 ldd [pB1+32], rB1 fadds rC13, m1, rC13 fmuls rA1, rB0, m1 fadds rC23, m2, rC23 fmuls rA2, rB0, m2 fadds rC33, m3, rC33 fmuls rA3, rB0, m3!! K=8 iteration! fadds rC00, m0, rC00 fmuls rA0, rB1, m0 fadds rC10, m1, rC10 fmuls rA1, rB1, m1 fadds rC20, m2, rC20 fmuls rA2, rB1, m2 fadds rC30, m3, rC30 fmuls rA3, rB1, m3 fadds rC01, m0, rC01 fmuls ra0, rb0, m0 fadds rC11, m1, rC11 fmuls ra1, rb0, m1 fadds rC21, m2, rC21 fmuls ra2, rb0, m2 fadds rC31, m3, rC31 fmuls ra3, rb0, m3 ldd [pB2+32], rB0 fadds rC00, m0, rC00 fmuls ra0, rb1, m0 fadds rC10, m1, rC10 fmuls ra1, rb1, m1 fadds rC20, m2, rC20 fmuls ra2, rb1, m2 fadds rC30, m3, rC30 fmuls ra3, rb1, m3 ldd [pB3+32], rB1 fadds rC01, m0, rC01 fmuls rA0, rB0, m0 fadds rC11, m1, rC11 fmuls rA1, rB0, m1 fadds rC21, m2, rC21 fmuls rA2, rB0, m2 fadds rC31, m3, rC31 fmuls rA3, rB0, m3 fadds rC02, m0, rC02 fmuls rA0, rB1, m0 fadds rC12, m1, rC12 fmuls rA1, rB1, m1 fadds rC22, m2, rC22 fmuls rA2, rB1, m2 fadds rC32, m3, rC32 fmuls rA3, rB1, m3 fadds rC03, m0, rC03 fmuls ra0, rb0, m0 fadds rC13, m1, rC13 fmuls ra1, rb0, m1 fadds rC23, m2, rC23 fmuls ra2, rb0, m2 fadds rC33, m3, rC33 fmuls ra3, rb0, m3 ldd [pB0+40], rB0 fadds rC02, m0, rC02 fmuls ra0, rb1, m0 ldd [pA0+40], rA0 fadds rC12, m1, rC12 fmuls ra1, rb1, m1 ldd [pA1+40], rA1 fadds rC22, m2, rC22 fmuls ra2, rb1, m2 ldd [pA2+40], rA2 fadds rC32, m3, rC32 fmuls ra3, rb1, m3 ldd [pA3+40], rA3 fadds rC03, m0, rC03 fmuls rA0, rB0, m0 ldd [pB1+40], rB1 fadds rC13, m1, rC13 fmuls rA1, rB0, m1 fadds rC23, m2, rC23 fmuls rA2, rB0, m2 fadds rC33, m3, rC33 fmuls rA3, rB0, m3!! K=10 iteration! fadds rC00, m0, rC00 fmuls rA0, rB1, m0 fadds rC10, m1, rC10 fmuls rA1, rB1, m1 fadds rC20, m2, rC20 fmuls rA2, rB1, m2 fadds rC30, m3, rC30 fmuls rA3, rB1, m3 fadds rC01, m0, rC01 fmuls ra0, rb0, m0 fadds rC11, m1, rC11 fmuls ra1, rb0, m1 fadds rC21, m2, rC21 fmuls ra2, rb0, m2 fadds rC31, m3, rC31 fmuls ra3, rb0, m3 ldd [pB2+40], rB0 fadds rC00, m0, rC00 fmuls ra0, rb1, m0 fadds rC10, m1, rC10 fmuls ra1, rb1, m1 fadds rC20, m2, rC20 fmuls ra2, rb1, m2 fadds rC30, m3, rC30 fmuls ra3, rb1, m3 ldd [pB3+40], rB1 fadds rC01, m0, rC01 fmuls rA0, rB0, m0 fadds rC11, m1, rC11 fmuls rA1, rB0, m1 fadds rC21, m2, rC21 fmuls rA2, rB0, m2 fadds rC31, m3, rC31 fmuls rA3, rB0, m3 fadds rC02, m0, rC02 fmuls rA0, rB1, m0 fadds rC12, m1, rC12 fmuls rA1, rB1, m1 fadds rC22, m2, rC22 fmuls rA2, rB1, m2 fadds rC32, m3, rC32 fmuls rA3, rB1, m3 fadds rC03, m0, rC03 fmuls ra0, rb0, m0 fadds rC13, m1, rC13 fmuls ra1, rb0, m1 fadds rC23, m2, rC23 fmuls ra2, rb0, m2 fadds rC33, m3, rC33 fmuls ra3, rb0, m3 ldd [pB0+48], rB0 fadds rC02, m0, rC02 fmuls ra0, rb1, m0 ldd [pA0+48], rA0 fadds rC12, m1, rC12 fmuls ra1, rb1, m1 ldd [pA1+48], rA1 fadds rC22, m2, rC22 fmuls ra2, rb1, m2 ldd [pA2+48], rA2 fadds rC32, m3, rC32 fmuls ra3, rb1, m3 ldd [pA3+48], rA3 fadds rC03, m0, rC03 fmuls rA0, rB0, m0 ldd [pB1+48], rB1 fadds rC13, m1, rC13 fmuls rA1, rB0, m1 fadds rC23, m2, rC23 fmuls rA2, rB0, m2 fadds rC33, m3, rC33 fmuls rA3, rB0, m3!! K=12 iteration! fadds rC00, m0, rC00 fmuls rA0, rB1, m0 fadds rC10, m1, rC10 fmuls rA1, rB1, m1 fadds rC20, m2, rC20 fmuls rA2, rB1, m2 fadds rC30, m3, rC30 fmuls rA3, rB1, m3 fadds rC01, m0, rC01 fmuls ra0, rb0, m0 fadds rC11, m1, rC11 fmuls ra1, rb0, m1 fadds rC21, m2, rC21 fmuls ra2, rb0, m2 fadds rC31, m3, rC31 fmuls ra3, rb0, m3 ldd [pB2+48], rB0 fadds rC00, m0, rC00 fmuls ra0, rb1, m0 fadds rC10, m1, rC10 fmuls ra1, rb1, m1 fadds rC20, m2, rC20 fmuls ra2, rb1, m2 fadds rC30, m3, rC30 fmuls ra3, rb1, m3 ldd [pB3+48], rB1 fadds rC01, m0, rC01 fmuls rA0, rB0, m0 fadds rC11, m1, rC11 fmuls rA1, rB0, m1 fadds rC21, m2, rC21 fmuls rA2, rB0, m2 fadds rC31, m3, rC31 fmuls rA3, rB0, m3 fadds rC02, m0, rC02 fmuls rA0, rB1, m0 fadds rC12, m1, rC12 fmuls rA1, rB1, m1 fadds rC22, m2, rC22 fmuls rA2, rB1, m2 fadds rC32, m3, rC32 fmuls rA3, rB1, m3 fadds rC03, m0, rC03 fmuls ra0, rb0, m0 fadds rC13, m1, rC13 fmuls ra1, rb0, m1 fadds rC23, m2, rC23 fmuls ra2, rb0, m2 fadds rC33, m3, rC33 fmuls ra3, rb0, m3 ldd [pB0+56], rB0 fadds rC02, m0, rC02 fmuls ra0, rb1, m0 ldd [pA0+56], rA0 fadds rC12, m1, rC12 fmuls ra1, rb1, m1 ldd [pA1+56], rA1 fadds rC22, m2, rC22 fmuls ra2, rb1, m2 ldd [pA2+56], rA2 fadds rC32, m3, rC32 fmuls ra3, rb1, m3 ldd [pA3+56], rA3 fadds rC03, m0, rC03 fmuls rA0, rB0, m0 ldd [pB1+56], rB1 fadds rC13, m1, rC13 fmuls rA1, rB0, m1 fadds rC23, m2, rC23 fmuls rA2, rB0, m2 fadds rC33, m3, rC33 fmuls rA3, rB0, m3!! K=14 iteration! fadds rC00, m0, rC00 fmuls rA0, rB1, m0 fadds rC10, m1, rC10 fmuls rA1, rB1, m1 fadds rC20, m2, rC20 fmuls rA2, rB1, m2 fadds rC30, m3, rC30 fmuls rA3, rB1, m3 fadds rC01, m0, rC01 fmuls ra0, rb0, m0 fadds rC11, m1, rC11 fmuls ra1, rb0, m1 fadds rC21, m2, rC21 fmuls ra2, rb0, m2 fadds rC31, m3, rC31 fmuls ra3, rb0, m3 ldd [pB2+56], rB0 fadds rC00, m0, rC00 fmuls ra0, rb1, m0 fadds rC10, m1, rC10 fmuls ra1, rb1, m1 fadds rC20, m2, rC20 fmuls ra2, rb1, m2 fadds rC30, m3, rC30 fmuls ra3, rb1, m3 ldd [pB3+56], rB1 fadds rC01, m0, rC01 fmuls rA0, rB0, m0 add pA0, 64, pA0 fadds rC11, m1, rC11 fmuls rA1, rB0, m1 add pA1, 64, pA1 fadds rC21, m2, rC21 fmuls rA2, rB0, m2 add pA2, 64, pA2 fadds rC31, m3, rC31 fmuls rA3, rB0, m3 add pA3, 64, pA3 fadds rC02, m0, rC02 fmuls rA0, rB1, m0 add pB0, 64, pB0 fadds rC12, m1, rC12 fmuls rA1, rB1, m1 add pB1, 64, pB1 fadds rC22, m2, rC22 fmuls rA2, rB1, m2 add pB2, 64, pB2 fadds rC32, m3, rC32 fmuls rA3, rB1, m3 add pB3, 64, pB3 fadds rC03, m0, rC03 fmuls ra0, rb0, m0 fadds rC13, m1, rC13 fmuls ra1, rb0, m1 fadds rC23, m2, rC23 fmuls ra2, rb0, m2 fadds rC33, m3, rC33 fmuls ra3, rb0, m3 ldd [pB0], rB0 fadds rC02, m0, rC02 fmuls ra0, rb1, m0 ldd [pA0], rA0 fadds rC12, m1, rC12 fmuls ra1, rb1, m1 ldd [pA1], rA1 fadds rC22, m2, rC22 fmuls ra2, rb1, m2 ldd [pA2], rA2 fadds rC32, m3, rC32 fmuls ra3, rb1, m3 ldd [pA3], rA3 fadds rC03, m0, rC03 fmuls rA0, rB0, m0 ldd [pB1], rB1 fadds rC13, m1, rC13 fmuls rA1, rB0, m1 subcc KK, 1, KK fadds rC23, m2, rC23 fmuls rA2, rB0, m2 fadds rC33, m3, rC33 fmuls rA3, rB0, m3 bnz KLOOP nop#endif!! Drain mul pipe on last iteration of K-loop!#if K == 0KDRAIN:#endif!! K=0 iteration! fadds rC00, m0, rC00 fmuls rA0, rB1, m0 fadds rC10, m1, rC10 fmuls rA1, rB1, m1 fadds rC20, m2, rC20 fmuls rA2, rB1, m2 fadds rC30, m3, rC30 fmuls rA3, rB1, m3 fadds rC01, m0, rC01 fmuls ra0, rb0, m0 fadds rC11, m1, rC11 fmuls ra1, rb0, m1 fadds rC21, m2, rC21 fmuls ra2, rb0, m2 fadds rC31, m3, rC31 fmuls ra3, rb0, m3 ldd [pB2+0], rB0 fadds rC00, m0, rC00 fmuls ra0, rb1, m0 fadds rC10, m1, rC10 fmuls ra1, rb1, m1 fadds rC20, m2, rC20 fmuls ra2, rb1, m2 fadds rC30, m3, rC30 fmuls ra3, rb1, m3 ldd [pB3+0], rB1 fadds rC01, m0, rC01 fmuls rA0, rB0, m0 fadds rC11, m1, rC11 fmuls rA1, rB0, m1 fadds rC21, m2, rC21 fmuls rA2, rB0, m2 fadds rC31, m3, rC31 fmuls rA3, rB0, m3 fadds rC02, m0, rC02 fmuls rA0, rB1, m0 fadds rC12, m1, rC12 fmuls rA1, rB1, m1 fadds rC22, m2, rC22 fmuls rA2, rB1, m2 fadds rC32, m3, rC32 fmuls rA3, rB1, m3 fadds rC03, m0, rC03 fmuls ra0, rb0, m0 fadds rC13, m1, rC13 fmuls ra1, rb0, m1 fadds rC23, m2, rC23 fmuls ra2, rb0, m2 fadds rC33, m3, rC33 fmuls ra3, rb0, m3 ldd [pB0+8], rB0 fadds rC02, m0, rC02 fmuls ra0, rb1, m0 ldd [pA0+8], rA0 fadds rC12, m1, rC12 fmuls ra1, rb1, m1 ldd [pA1+8], rA1 fadds rC22, m2, rC22 fmuls ra2, rb1, m2 ldd [pA2+8], rA2 fadds rC32, m3, rC32 fmuls ra3, rb1, m3 ldd [pA3+8], rA3 fadds rC03, m0, rC03 fmuls rA0, rB0, m0 ldd [pB1+8], rB1 fadds rC13, m1, rC13 fmuls rA1, rB0, m1 fadds rC23, m2, rC23 fmuls rA2, rB0, m2 fadds rC33, m3, rC33 fmuls rA3, rB0, m3!! K=2 iteration! fadds rC00, m0, rC00 fmuls rA0, rB1, m0 fadds rC10, m1, rC10 fmuls rA1, rB1, m1 fadds rC20, m2, rC20 fmuls rA2, rB1, m2 fadds rC30, m3, rC30 fmuls rA3, rB1, m3 fadds rC01, m0, rC01 fmuls ra0, rb0, m0 fadds rC11, m1, rC11 fmuls ra1, rb0, m1 fadds rC21, m2, rC21 fmuls ra2, rb0, m2 fadds rC31, m3, rC31 fmuls ra3, rb0, m3 ldd [pB2+8], rB0 fadds rC00, m0, rC00 fmuls ra0, rb1, m0 fadds rC10, m1, rC10 fmuls ra1, rb1, m1 fadds rC20, m2, rC20 fmuls ra2, rb1, m2 fadds rC30, m3, rC30 fmuls ra3, rb1, m3 ldd [pB3+8], rB1 fadds rC01, m0, rC01 fmuls rA0, rB0, m0 fadds rC11, m1, rC11 fmuls rA1, rB0, m1 fadds rC21, m2, rC21 fmuls rA2, rB0, m2 fadds rC31, m3, rC31 fmuls rA3, rB0, m3 fadds rC02, m0, rC02 fmuls rA0, rB1, m0 fadds rC12, m1, rC12 fmuls rA1, rB1, m1 fadds rC22, m2, rC22 fmuls rA2, rB1, m2 fadds rC32, m3, rC32 fmuls rA3, rB1, m3 fadds rC03, m0, rC03 fmuls ra0, rb0, m0 fadds rC13, m1, rC13 fmuls ra1, rb0, m1 fadds rC23, m2, rC23 fmuls ra2, rb0, m2 fadds rC33, m3, rC33 fmuls ra3, rb0, m3 ldd [pB0+16], rB0 fadds rC02, m0, rC02 fmuls ra0, rb1, m0 ldd [pA0+16], rA0 fadds rC12, m1, rC12 fmuls ra1, rb1, m1 ldd [pA1+16], rA1 fadds rC22, m2, rC22 fmuls ra2, rb1, m2
⌨️ 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?