Lines Matching refs:range_k
340 const struct fxdiv_divisor_size_t range_k = threadpool->params.parallelize_3d.range_k; in thread_parallelize_3d() local
341 const struct fxdiv_result_size_t index_ij_k = fxdiv_divide_size_t(range_start, range_k); in thread_parallelize_3d()
350 if (++k == range_k.value) { in thread_parallelize_3d()
369 const struct fxdiv_result_size_t index_ij_k = fxdiv_divide_size_t(linear_index, range_k); in thread_parallelize_3d()
397 const size_t range_k = threadpool->params.parallelize_3d_tile_1d.range_k; in thread_parallelize_3d_tile_1d() local
399 task(argument, i, j, start_k, min(range_k - start_k, tile_k)); in thread_parallelize_3d_tile_1d()
401 if (start_k >= range_k) { in thread_parallelize_3d_tile_1d()
423 task(argument, index_i_j.quotient, index_i_j.remainder, start_k, min(range_k - start_k, tile_k)); in thread_parallelize_3d_tile_1d()
450 const size_t range_k = threadpool->params.parallelize_3d_tile_2d.range_k; in thread_parallelize_3d_tile_2d() local
453 …task(argument, i, start_j, start_k, min(range_j - start_j, tile_j), min(range_k - start_k, tile_k)… in thread_parallelize_3d_tile_2d()
455 if (start_k >= range_k) { in thread_parallelize_3d_tile_2d()
479 …ndex_i_j.quotient, start_j, start_k, min(range_j - start_j, tile_j), min(range_k - start_k, tile_k… in thread_parallelize_3d_tile_2d()
515 const size_t range_k = threadpool->params.parallelize_3d_tile_2d_with_uarch.range_k; in thread_parallelize_3d_tile_2d_with_uarch() local
518 …task(argument, uarch_index, i, start_j, start_k, min(range_j - start_j, tile_j), min(range_k - sta… in thread_parallelize_3d_tile_2d_with_uarch()
520 if (start_k >= range_k) { in thread_parallelize_3d_tile_2d_with_uarch()
544 …ndex_i_j.quotient, start_j, start_k, min(range_j - start_j, tile_j), min(range_k - start_k, tile_k… in thread_parallelize_3d_tile_2d_with_uarch()
572 const size_t range_k = threadpool->params.parallelize_4d.range_k; in thread_parallelize_4d() local
577 if (++k == range_k) { in thread_parallelize_4d()
629 const size_t range_k = threadpool->params.parallelize_4d_tile_1d.range_k; in thread_parallelize_4d_tile_1d() local
636 if (++k == range_k) { in thread_parallelize_4d_tile_1d()
691 const size_t range_k = threadpool->params.parallelize_4d_tile_2d.range_k; in thread_parallelize_4d_tile_2d() local
693 …task(argument, i, j, start_k, start_l, min(range_k - start_k, tile_k), min(range_l - start_l, tile… in thread_parallelize_4d_tile_2d()
698 if (start_k >= range_k) { in thread_parallelize_4d_tile_2d()
723 …task(argument, index_i_j.quotient, index_i_j.remainder, start_k, start_l, min(range_k - start_k, t… in thread_parallelize_4d_tile_2d()
763 const size_t range_k = threadpool->params.parallelize_4d_tile_2d_with_uarch.range_k; in thread_parallelize_4d_tile_2d_with_uarch() local
765 …task(argument, uarch_index, i, j, start_k, start_l, min(range_k - start_k, tile_k), min(range_l - … in thread_parallelize_4d_tile_2d_with_uarch()
770 if (start_k >= range_k) { in thread_parallelize_4d_tile_2d_with_uarch()
795 …ch_index, index_i_j.quotient, index_i_j.remainder, start_k, start_l, min(range_k - start_k, tile_k… in thread_parallelize_4d_tile_2d_with_uarch()
814 const struct fxdiv_divisor_size_t range_k = threadpool->params.parallelize_5d.range_k; in thread_parallelize_5d() local
815 const struct fxdiv_result_size_t index_ij_k = fxdiv_divide_size_t(index_ijk_lm.quotient, range_k); in thread_parallelize_5d()
833 if (++k == range_k.value) { in thread_parallelize_5d()
855 … const struct fxdiv_result_size_t index_ij_k = fxdiv_divide_size_t(index_ijk_lm.quotient, range_k); in thread_parallelize_5d()
891 const size_t range_k = threadpool->params.parallelize_5d_tile_1d.range_k; in thread_parallelize_5d_tile_1d() local
899 if (++k == range_k) { in thread_parallelize_5d_tile_1d()
945 const struct fxdiv_divisor_size_t range_k = threadpool->params.parallelize_5d_tile_2d.range_k; in thread_parallelize_5d_tile_2d() local
946 …t struct fxdiv_result_size_t index_ij_k = fxdiv_divide_size_t(tile_index_ijk_lm.quotient, range_k); in thread_parallelize_5d_tile_2d()
969 if (++k == range_k.value) { in thread_parallelize_5d_tile_2d()
991 …t struct fxdiv_result_size_t index_ij_k = fxdiv_divide_size_t(tile_index_ijk_lm.quotient, range_k); in thread_parallelize_5d_tile_2d()
1016 const struct fxdiv_divisor_size_t range_k = threadpool->params.parallelize_6d.range_k; in thread_parallelize_6d() local
1017 const struct fxdiv_result_size_t index_ij_k = fxdiv_divide_size_t(index_ijk_lmn.quotient, range_k); in thread_parallelize_6d()
1040 if (++k == range_k.value) { in thread_parallelize_6d()
1064 …const struct fxdiv_result_size_t index_ij_k = fxdiv_divide_size_t(index_ijk_lmn.quotient, range_k); in thread_parallelize_6d()
1087 const struct fxdiv_divisor_size_t range_k = threadpool->params.parallelize_6d_tile_1d.range_k; in thread_parallelize_6d_tile_1d() local
1088 … struct fxdiv_result_size_t index_ij_k = fxdiv_divide_size_t(tile_index_ijk_lmn.quotient, range_k); in thread_parallelize_6d_tile_1d()
1114 if (++k == range_k.value) { in thread_parallelize_6d_tile_1d()
1138 … struct fxdiv_result_size_t index_ij_k = fxdiv_divide_size_t(tile_index_ijk_lmn.quotient, range_k); in thread_parallelize_6d_tile_1d()
1182 const size_t range_k = threadpool->params.parallelize_6d_tile_2d.range_k; in thread_parallelize_6d_tile_2d() local
1193 if (++k == range_k) { in thread_parallelize_6d_tile_2d()
1564 size_t range_k, in pthreadpool_parallelize_3d() argument
1568 … || (threads_count = threadpool->threads_count.value) <= 1 || (range_i | range_j | range_k) <= 1) { in pthreadpool_parallelize_3d()
1577 for (size_t k = 0; k < range_k; k++) { in pthreadpool_parallelize_3d()
1586 const size_t range = range_i * range_j * range_k; in pthreadpool_parallelize_3d()
1589 .range_k = fxdiv_init_size_t(range_k), in pthreadpool_parallelize_3d()
1610 size_t range_k, in pthreadpool_parallelize_3d_tile_1d() argument
1615 …count = threadpool->threads_count.value) <= 1 || ((range_i | range_j) <= 1 && range_k <= tile_k)) { in pthreadpool_parallelize_3d_tile_1d()
1624 for (size_t k = 0; k < range_k; k += tile_k) { in pthreadpool_parallelize_3d_tile_1d()
1625 task(argument, i, j, k, min(range_k - k, tile_k)); in pthreadpool_parallelize_3d_tile_1d()
1633 const size_t tile_range_k = divide_round_up(range_k, tile_k); in pthreadpool_parallelize_3d_tile_1d()
1636 .range_k = range_k, in pthreadpool_parallelize_3d_tile_1d()
1660 size_t range_k, in pthreadpool_parallelize_3d_tile_2d() argument
1666 …hreadpool->threads_count.value) <= 1 || (range_i <= 1 && range_j <= tile_j && range_k <= tile_k)) { in pthreadpool_parallelize_3d_tile_2d()
1675 for (size_t k = 0; k < range_k; k += tile_k) { in pthreadpool_parallelize_3d_tile_2d()
1676 task(argument, i, j, k, min(range_j - j, tile_j), min(range_k - k, tile_k)); in pthreadpool_parallelize_3d_tile_2d()
1685 const size_t tile_range_k = divide_round_up(range_k, tile_k); in pthreadpool_parallelize_3d_tile_2d()
1690 .range_k = range_k, in pthreadpool_parallelize_3d_tile_2d()
1716 size_t range_k, in pthreadpool_parallelize_3d_tile_2d_with_uarch() argument
1722 …hreadpool->threads_count.value) <= 1 || (range_i <= 1 && range_j <= tile_j && range_k <= tile_k)) { in pthreadpool_parallelize_3d_tile_2d_with_uarch()
1740 for (size_t k = 0; k < range_k; k += tile_k) { in pthreadpool_parallelize_3d_tile_2d_with_uarch()
1741 task(argument, uarch_index, i, j, k, min(range_j - j, tile_j), min(range_k - k, tile_k)); in pthreadpool_parallelize_3d_tile_2d_with_uarch()
1750 const size_t tile_range_k = divide_round_up(range_k, tile_k); in pthreadpool_parallelize_3d_tile_2d_with_uarch()
1757 .range_k = range_k, in pthreadpool_parallelize_3d_tile_2d_with_uarch()
1781 size_t range_k, in pthreadpool_parallelize_4d() argument
1786 …ds_count = threadpool->threads_count.value) <= 1 || (range_i | range_j | range_k | range_l) <= 1) { in pthreadpool_parallelize_4d()
1795 for (size_t k = 0; k < range_k; k++) { in pthreadpool_parallelize_4d()
1806 const size_t range_kl = range_k * range_l; in pthreadpool_parallelize_4d()
1809 .range_k = range_k, in pthreadpool_parallelize_4d()
1833 size_t range_k, in pthreadpool_parallelize_4d_tile_1d() argument
1839 …s_count = threadpool->threads_count.value) <= 1 || ((range_i | range_j | range_k) <= 1 && range_l … in pthreadpool_parallelize_4d_tile_1d()
1848 for (size_t k = 0; k < range_k; k++) { in pthreadpool_parallelize_4d_tile_1d()
1860 const size_t tile_range_kl = range_k * tile_range_l; in pthreadpool_parallelize_4d_tile_1d()
1863 .range_k = range_k, in pthreadpool_parallelize_4d_tile_1d()
1889 size_t range_k, in pthreadpool_parallelize_4d_tile_2d() argument
1896 … = threadpool->threads_count.value) <= 1 || ((range_i | range_j) <= 1 && range_k <= tile_k && rang… in pthreadpool_parallelize_4d_tile_2d()
1905 for (size_t k = 0; k < range_k; k += tile_k) { in pthreadpool_parallelize_4d_tile_2d()
1908 min(range_k - k, tile_k), min(range_l - l, tile_l)); in pthreadpool_parallelize_4d_tile_2d()
1918 const size_t tile_range_kl = divide_round_up(range_k, tile_k) * tile_range_l; in pthreadpool_parallelize_4d_tile_2d()
1921 .range_k = range_k, in pthreadpool_parallelize_4d_tile_2d()
1950 size_t range_k, in pthreadpool_parallelize_4d_tile_2d_with_uarch() argument
1957 … = threadpool->threads_count.value) <= 1 || ((range_i | range_j) <= 1 && range_k <= tile_k && rang… in pthreadpool_parallelize_4d_tile_2d_with_uarch()
1975 for (size_t k = 0; k < range_k; k += tile_k) { in pthreadpool_parallelize_4d_tile_2d_with_uarch()
1978 min(range_k - k, tile_k), min(range_l - l, tile_l)); in pthreadpool_parallelize_4d_tile_2d_with_uarch()
1988 const size_t tile_range_kl = divide_round_up(range_k, tile_k) * tile_range_l; in pthreadpool_parallelize_4d_tile_2d_with_uarch()
1993 .range_k = range_k, in pthreadpool_parallelize_4d_tile_2d_with_uarch()
2020 size_t range_k, in pthreadpool_parallelize_5d() argument
2026 …ds_count = threadpool->threads_count.value) <= 1 || (range_i | range_j | range_k | range_l | range… in pthreadpool_parallelize_5d()
2035 for (size_t k = 0; k < range_k; k++) { in pthreadpool_parallelize_5d()
2049 const size_t range = range_i * range_j * range_k * range_lm; in pthreadpool_parallelize_5d()
2053 .range_k = fxdiv_init_size_t(range_k), in pthreadpool_parallelize_5d()
2076 size_t range_k, in pthreadpool_parallelize_5d_tile_1d() argument
2083 …s_count = threadpool->threads_count.value) <= 1 || ((range_i | range_j | range_k | range_l) <= 1 &… in pthreadpool_parallelize_5d_tile_1d()
2092 for (size_t k = 0; k < range_k; k++) { in pthreadpool_parallelize_5d_tile_1d()
2106 const size_t range_kl = range_k * range_l; in pthreadpool_parallelize_5d_tile_1d()
2109 .range_k = range_k, in pthreadpool_parallelize_5d_tile_1d()
2136 size_t range_k, in pthreadpool_parallelize_5d_tile_2d() argument
2144 …s_count = threadpool->threads_count.value) <= 1 || ((range_i | range_j | range_k) <= 1 && range_l … in pthreadpool_parallelize_5d_tile_2d()
2153 for (size_t k = 0; k < range_k; k++) { in pthreadpool_parallelize_5d_tile_2d()
2169 const size_t tile_range = range_i * range_j * range_k * tile_range_lm; in pthreadpool_parallelize_5d_tile_2d()
2176 .range_k = fxdiv_init_size_t(range_k), in pthreadpool_parallelize_5d_tile_2d()
2199 size_t range_k, in pthreadpool_parallelize_6d() argument
2206 …ds_count = threadpool->threads_count.value) <= 1 || (range_i | range_j | range_k | range_l | range… in pthreadpool_parallelize_6d()
2215 for (size_t k = 0; k < range_k; k++) { in pthreadpool_parallelize_6d()
2231 const size_t range = range_i * range_j * range_k * range_lmn; in pthreadpool_parallelize_6d()
2235 .range_k = fxdiv_init_size_t(range_k), in pthreadpool_parallelize_6d()
2259 size_t range_k, in pthreadpool_parallelize_6d_tile_1d() argument
2267 …s_count = threadpool->threads_count.value) <= 1 || ((range_i | range_j | range_k | range_l | range… in pthreadpool_parallelize_6d_tile_1d()
2276 for (size_t k = 0; k < range_k; k++) { in pthreadpool_parallelize_6d_tile_1d()
2293 const size_t tile_range = range_i * range_j * range_k * tile_range_lmn; in pthreadpool_parallelize_6d_tile_1d()
2299 .range_k = fxdiv_init_size_t(range_k), in pthreadpool_parallelize_6d_tile_1d()
2323 size_t range_k, in pthreadpool_parallelize_6d_tile_2d() argument
2332 …s_count = threadpool->threads_count.value) <= 1 || ((range_i | range_j | range_k | range_l) <= 1 &… in pthreadpool_parallelize_6d_tile_2d()
2341 for (size_t k = 0; k < range_k; k++) { in pthreadpool_parallelize_6d_tile_2d()
2357 const size_t range_kl = range_k * range_l; in pthreadpool_parallelize_6d_tile_2d()
2362 .range_k = range_k, in pthreadpool_parallelize_6d_tile_2d()