/dports/math/libflame/libflame-5.2.0/src/lapack/dec/bsvd/v/flamec/ |
H A D | FLA_Bsvd_iteracc_v_opt_var1.c | 26 int k_iter = 0; in FLA_Bsvd_iteracc_v_ops_var1() local 53 k_iter += k; in FLA_Bsvd_iteracc_v_ops_var1() 58 *n_iter_perf = k_iter; in FLA_Bsvd_iteracc_v_ops_var1() 64 if ( k_iter == n_GH ) in FLA_Bsvd_iteracc_v_ops_var1() 165 k_iter += 1; in FLA_Bsvd_iteracc_v_ops_var1() 170 *n_iter_perf = k_iter; in FLA_Bsvd_iteracc_v_ops_var1() 189 int k_iter = 0; in FLA_Bsvd_iteracc_v_opd_var1() local 216 k_iter += k; in FLA_Bsvd_iteracc_v_opd_var1() 237 if ( k_iter == n_GH ) in FLA_Bsvd_iteracc_v_opd_var1() 354 k_iter += 1; in FLA_Bsvd_iteracc_v_opd_var1() [all …]
|
/dports/math/libflame/libflame-5.2.0/src/lapack/dec/tevd/v/flamec/ |
H A D | FLA_Tevd_iteracc_v_opt_var1.c | 36 int k_iter = 0; in FLA_Tevd_iteracc_v_opd_var1() local 42 dcomplex* G1 = buff_G + (k_iter)*cs_G; in FLA_Tevd_iteracc_v_opd_var1() 44 int k_left = n_G - k_iter; in FLA_Tevd_iteracc_v_opd_var1() 62 k_iter += k; in FLA_Tevd_iteracc_v_opd_var1() 71 *n_iter_perf = k_iter; in FLA_Tevd_iteracc_v_opd_var1() 84 if ( k_iter == n_G ) in FLA_Tevd_iteracc_v_opd_var1() 86 *n_iter_perf = k_iter; in FLA_Tevd_iteracc_v_opd_var1() 99 int n_Gr = n_G - k_iter; in FLA_Tevd_iteracc_v_opd_var1() 152 dcomplex* g1 = buff_G + (k_iter)*cs_G; in FLA_Tevd_iteracc_v_opd_var1() 186 k_iter += 1; in FLA_Tevd_iteracc_v_opd_var1() [all …]
|
H A D | FLA_Tevd_iteracc_v_opt_var3.c | 44 int k_iter = 0; in FLA_Tevd_iteracc_v_opd_var3() local 53 dcomplex* G1 = buff_G + (k_iter)*cs_G; in FLA_Tevd_iteracc_v_opd_var3() 55 int k_left = n_G - k_iter; in FLA_Tevd_iteracc_v_opd_var3() 133 if ( r_val == FLA_FAILURE && k_iter == n_G ) in FLA_Tevd_iteracc_v_opd_var3() 138 *n_iter_perf = k_iter; in FLA_Tevd_iteracc_v_opd_var3() 144 k_iter += k; in FLA_Tevd_iteracc_v_opd_var3() 157 if ( k_iter == n_G ) in FLA_Tevd_iteracc_v_opd_var3() 159 *n_iter_perf = k_iter; in FLA_Tevd_iteracc_v_opd_var3() 172 int n_Gr = n_G - k_iter; in FLA_Tevd_iteracc_v_opd_var3() 269 k_iter += 1; in FLA_Tevd_iteracc_v_opd_var3() [all …]
|
/dports/math/libflame/libflame-5.2.0/src/lapack/dec/tevd/n/flamec/ |
H A D | FLA_Tevd_iteracc_n_opt_var1.c | 34 int k_iter = 0; in FLA_Tevd_iteracc_n_opd_var1() local 41 int k_left = n_G - k_iter; in FLA_Tevd_iteracc_n_opd_var1() 58 k_iter += k; in FLA_Tevd_iteracc_n_opd_var1() 67 *n_iter_perf = k_iter; in FLA_Tevd_iteracc_n_opd_var1() 80 if ( k_iter == n_G ) in FLA_Tevd_iteracc_n_opd_var1() 82 *n_iter_perf = k_iter; in FLA_Tevd_iteracc_n_opd_var1() 95 int n_Gr = n_G - k_iter; in FLA_Tevd_iteracc_n_opd_var1() 114 printf( "FLA_Tevd_iteracc_n_opd_var1: GB(0,0) i,j: %d %d\n", ijTL + m_TLr+1, k_iter ); in FLA_Tevd_iteracc_n_opd_var1() 129 *n_iter_perf = k_iter + max( n_iter_perfTL, n_iter_perfBR ); in FLA_Tevd_iteracc_n_opd_var1() 166 k_iter += 1; in FLA_Tevd_iteracc_n_opd_var1() [all …]
|
/dports/math/tblis/tblis-1.2.0/src/configs/sandybridge/ |
H A D | bli_gemm_asm_d8x4.c | 55 uint64_t k_iter = k / 4; in bli_sgemm_asm_8x8() local 1018 "m" (k_iter), // 0 in bli_sgemm_asm_8x8() 1055 uint64_t k_iter = k / 4; in bli_dgemm_asm_8x4() local 1707 "m" (k_iter), // 0 in bli_dgemm_asm_8x4() 1744 uint64_t k_iter = k / 4; in bli_cgemm_asm_8x4() local 2681 "m" (k_iter), // 0 in bli_cgemm_asm_8x4() 2720 uint64_t k_iter = k / 4; in bli_zgemm_asm_4x4() local 3547 "m" (k_iter), // 0 in bli_zgemm_asm_4x4()
|
/dports/math/tblis/tblis-1.2.0/src/configs/piledriver/ |
H A D | bli_gemm_asm_d8x3.c | 55 uint64_t k_iter = k / 8; in bli_sgemm_asm_16x3() local 885 "m" (k_iter), // 0 in bli_sgemm_asm_16x3() 922 dim_t k_iter = k / 8; in bli_dgemm_asm_8x3() local 1600 "m" (k_iter), // 0 in bli_dgemm_asm_8x3() 1637 dim_t k_iter = k / 8; in bli_cgemm_asm_4x2() local 2144 "m" (k_iter), // 0 in bli_cgemm_asm_4x2() 2181 dim_t k_iter = k / 8; in bli_zgemm_asm_2x2() local 2674 "m" (k_iter), // 0 in bli_zgemm_asm_2x2()
|
/dports/math/tblis/tblis-1.2.0/src/configs/bulldozer/ |
H A D | bli_gemm_asm_d4x6_fma4.c | 100 uint64_t k_iter = k / 4; in bli_sgemm_asm_8x8_fma4() local 738 "m" (k_iter), // 0 in bli_sgemm_asm_8x8_fma4() 868 dim_t k_iter = k / 12; in bli_dgemm_asm_4x6_fma4() local 1028 "r" (k_iter), in bli_dgemm_asm_4x6_fma4() 1079 dim_t k_iter = k / 4; in bli_cgemm_asm_8x4_fma4() local 1825 "m" (k_iter), // 0 in bli_cgemm_asm_8x4_fma4() 1886 dim_t k_iter = k / 4; in bli_zgemm_asm_4x4_fma4() local 2494 "m" (k_iter), // 0 in bli_zgemm_asm_4x4_fma4()
|
/dports/math/blis/blis-0.8.1/kernels/haswell/3/old/ |
H A D | bli_gemm_haswell_asm_d6x8.c | 94 uint64_t k_iter = k0 / 4; in bli_sgemm_haswell_asm_6x16() local 617 "m" (k_iter), // 0 in bli_sgemm_haswell_asm_6x16() 683 uint64_t k_iter = k0 / 4; in bli_dgemm_haswell_asm_6x8() local 1201 "m" (k_iter), // 0 in bli_dgemm_haswell_asm_6x8() 1274 uint64_t k_iter = k0 / 4; in bli_cgemm_haswell_asm_3x8() local 1734 "m" (k_iter), // 0 in bli_cgemm_haswell_asm_3x8() 1803 uint64_t k_iter = k0 / 4; in bli_zgemm_haswell_asm_3x4() local 2263 "m" (k_iter), // 0 in bli_zgemm_haswell_asm_3x4()
|
H A D | bli_gemm_haswell_asm_d8x6.c | 94 uint64_t k_iter = k0 / 4; in bli_sgemm_haswell_asm_16x6() local 618 "m" (k_iter), // 0 in bli_sgemm_haswell_asm_16x6() 681 uint64_t k_iter = k0 / 4; in bli_dgemm_haswell_asm_8x6() local 1204 "m" (k_iter), // 0 in bli_dgemm_haswell_asm_8x6() 1277 uint64_t k_iter = k0 / 4; in bli_cgemm_haswell_asm_8x3() local 1737 "m" (k_iter), // 0 in bli_cgemm_haswell_asm_8x3() 1806 uint64_t k_iter = k0 / 4; in bli_zgemm_haswell_asm_4x3() local 2266 "m" (k_iter), // 0 in bli_zgemm_haswell_asm_4x3()
|
H A D | bli_gemm_haswell_asm_d12x4.c | 94 uint64_t k_iter = k0 / 4; in bli_sgemm_haswell_asm_24x4() local 656 "m" (k_iter), // 0 in bli_sgemm_haswell_asm_24x4() 719 uint64_t k_iter = k0 / 4; in bli_dgemm_haswell_asm_12x4() local 1281 "m" (k_iter), // 0 in bli_dgemm_haswell_asm_12x4() 1322 uint64_t k_iter = k0 / 4; 1348 uint64_t k_iter = k0 / 4;
|
H A D | bli_gemm_haswell_asm_d4x12.c | 94 uint64_t k_iter = k0 / 4; in bli_sgemm_haswell_asm_4x24() local 622 "m" (k_iter), // 0 in bli_sgemm_haswell_asm_4x24() 685 uint64_t k_iter = k0 / 4; in bli_dgemm_haswell_asm_4x12() local 1214 "m" (k_iter), // 0 in bli_dgemm_haswell_asm_4x12() 1255 uint64_t k_iter = k0 / 4; 1281 uint64_t k_iter = k0 / 4;
|
/dports/math/blis/blis-0.8.1/kernels/armv8a/3/ |
H A D | bli_gemm_armv8a_asm_d6x8.c | 70 uint64_t k_iter = k0 / 4; in bli_sgemm_armv8a_asm_8x12() local 1060 [k_iter] "m" (k_iter), // 3 in bli_sgemm_armv8a_asm_8x12() 1125 uint64_t k_iter = k0 / 4; in bli_dgemm_armv8a_asm_6x8() local 2053 [k_iter] "m" (k_iter), // 3 in bli_dgemm_armv8a_asm_6x8()
|
/dports/math/tblis/tblis-1.2.0/src/configs/core2/ |
H A D | bli_gemm_asm_d4x4.c | 52 uint64_t k_iter = k / 4; in bli_sgemm_asm_8x4() local 819 "m" (k_iter), in bli_sgemm_asm_8x4() 854 uint64_t k_iter = k / 4; in bli_dgemm_asm_4x4() local 1461 "m" (k_iter), // 0 in bli_dgemm_asm_4x4()
|
/dports/math/libflame/libflame-5.2.0/src/lapack/dec/tevd/v/flamec/old/ |
H A D | FLA_Tevd_v_opt_var4r.c | 98 int k_iter, in FLA_Tevd_v_ops_var4r() argument 114 int k_iter, in FLA_Tevd_v_opd_var4r() argument 155 k_iter, in FLA_Tevd_v_opd_var4r() 160 k_iter += k; in FLA_Tevd_v_opd_var4r() 197 k_iter, in FLA_Tevd_v_opd_var4r() 207 k_iter, in FLA_Tevd_v_opd_var4r()
|
/dports/math/blis/blis-0.8.1/kernels/armsve/3/ |
H A D | bli_gemm_armsve256_asm_d8x8.c | 66 uint64_t k_iter = k0 / 4; in bli_dgemm_armsve256_asm_8x8() local 783 [k_iter] "m" (k_iter), // 3 in bli_dgemm_armsve256_asm_8x8()
|
/dports/math/blis/blis-0.8.1/sandbox/power10/ |
H A D | gemm_pack.c | 315 int k_iter = k/4; \ 328 for (int p=0; p<k_iter; p++) \ 380 for (int p=0; p<k_iter; p++) \ 461 int k_iter = k/4; \ 473 for ( int p=0; p<k_iter; p++ ) \ 556 for ( int p=0; p<k_iter; p++ ) \ 652 int p_idx, k_left, k_iter; \ 656 k_iter = k/8; \ 665 for (int p=0; p<k_iter; p++) { \ 763 int p_idx, k_left, k_iter; \ [all …]
|
/dports/math/tblis/tblis-1.2.0/src/configs/haswell/ |
H A D | bli_gemm_asm_d12x4.c | 83 dim_t k_iter = k / 4; in bli_sgemm_asm_24x4() local 641 "m" (k_iter), // 0 in bli_sgemm_asm_24x4() 702 dim_t k_iter = k / 4; in bli_dgemm_asm_12x4() local 1260 "m" (k_iter), // 0 in bli_dgemm_asm_12x4()
|
H A D | bli_gemm_asm_d4x12.c | 83 dim_t k_iter = k / 4; in bli_sgemm_asm_4x24() local 684 "m" (k_iter), // 0 in bli_sgemm_asm_4x24() 747 dim_t k_iter = k / 4; in bli_dgemm_asm_4x12() local 1337 "m" (k_iter), // 0 in bli_dgemm_asm_4x12()
|
H A D | bli_gemm_asm_d8x6.c | 83 dim_t k_iter = k / 4; in bli_sgemm_asm_16x6() local 634 "m" (k_iter), // 0 in bli_sgemm_asm_16x6() 695 dim_t k_iter = k / 4; in bli_dgemm_asm_8x6() local 1246 "m" (k_iter), // 0 in bli_dgemm_asm_8x6()
|
H A D | bli_gemm_asm_d6x8.c | 83 uint64_t k_iter = k / 4; in bli_sgemm_asm_6x16() local 602 "m" (k_iter), // 0 in bli_sgemm_asm_6x16() 665 uint64_t k_iter = k / 4; in bli_dgemm_asm_6x8() local 1182 "m" (k_iter), // 0 in bli_dgemm_asm_6x8()
|
/dports/math/blis/blis-0.8.1/kernels/old/x86/3/ |
H A D | bli_gemm_opt_d2x4.c | 58 dim_t k_iter; in bli_dgemm_opt_d2x4() local 61 k_iter = k / 8; in bli_dgemm_opt_d2x4() 350 "m" (k_iter), in bli_dgemm_opt_d2x4()
|
H A D | bli_gemm_opt_d4x2.c | 62 dim_t k_iter; in bli_dgemm_opt_d4x2() local 65 k_iter = k / 8; in bli_dgemm_opt_d4x2() 308 "m" (k_iter), in bli_dgemm_opt_d4x2()
|
H A D | bli_gemmtrsm_l_opt_d4x2.c | 66 dim_t k_iter; in bli_dgemmtrsm_l_opt_d4x2() local 69 k_iter = k / 8; in bli_dgemmtrsm_l_opt_d4x2() 400 "m" (k_iter), in bli_dgemmtrsm_l_opt_d4x2()
|
H A D | bli_gemmtrsm_u_opt_d4x2.c | 66 dim_t k_iter; in bli_dgemmtrsm_u_opt_d4x2() local 69 k_iter = k / 8; in bli_dgemmtrsm_u_opt_d4x2() 404 "m" (k_iter), in bli_dgemmtrsm_u_opt_d4x2()
|
/dports/math/blis/blis-0.8.1/kernels/zen/3/sup/broken/ |
H A D | bli_gemmsup_rv_zen_asm_z3x4n.c | 110 uint64_t k_iter = 0; in bli_zgemmsup_rv_zen_asm_3x4n() local 166 for (k_iter = 0; k_iter <k0; k_iter++) in bli_zgemmsup_rv_zen_asm_3x4n() 171 ymm0 = _mm256_loadu_pd((double const *)(tB + tb_inc_row * k_iter)); in bli_zgemmsup_rv_zen_asm_3x4n() 480 uint64_t k_iter = 0; in bli_zgemmsup_rv_zen_asm_2x4n() local 531 for (k_iter = 0; k_iter <k0; k_iter++) in bli_zgemmsup_rv_zen_asm_2x4n() 536 ymm0 = _mm256_loadu_pd((double const *)(tB + tb_inc_row * k_iter)); in bli_zgemmsup_rv_zen_asm_2x4n() 785 uint64_t k_iter = 0; in bli_zgemmsup_rv_zen_asm_1x4n() local 828 for (k_iter = 0; k_iter <k0; k_iter++) in bli_zgemmsup_rv_zen_asm_1x4n() 1006 uint64_t k_iter = 0; in bli_zgemmsup_rv_zen_asm_3x2() local 1043 for (k_iter = 0; k_iter <k0; k_iter++) in bli_zgemmsup_rv_zen_asm_3x2() [all …]
|