/* Copyright (C) 2011 Fredrik Johansson Copyright (C) 2013, 2014 Martin Lee Copyright (C) 2020 William Hart This file is part of FLINT. FLINT is free software: you can redistribute it and/or modify it under the terms of the GNU Lesser General Public License (LGPL) as published by the Free Software Foundation; either version 2.1 of the License, or (at your option) any later version. See . */ #undef ulong #define ulong ulongxx/* interferes with system includes */ #include #include #undef ulong #include #include #define ulong mp_limb_t #include "flint.h" #include "nmod_poly.h" #include "ulong_extras.h" #include "thread_support.h" int main(void) { #if FLINT_USES_PTHREAD && (FLINT_USES_TLS || FLINT_REENTRANT) int i; #endif FLINT_TEST_INIT(state); flint_printf("compose_mod_brent_kung_precomp_preinv_threaded...."); fflush(stdout); #if FLINT_USES_PTHREAD && (FLINT_USES_TLS || FLINT_REENTRANT) /* check precomputation */ for (i = 0; i < 100 * flint_test_multiplier(); i++) { nmod_poly_t a, b, c, cinv; nmod_poly_struct * tmp; nmod_mat_t B; nmod_mat_struct * C; mp_limb_t m = n_randtest_prime(state, 0); slong j, num_threads; nmod_poly_matrix_precompute_arg_t * args1; thread_pool_handle * threads; flint_set_num_threads(1 + n_randint(state, 3)); num_threads = flint_request_threads(&threads, FLINT_DEFAULT_THREAD_LIMIT); tmp = (nmod_poly_struct *) flint_malloc(sizeof(nmod_poly_struct)*(num_threads + 1)); nmod_poly_init(a, m); nmod_poly_init(b, m); nmod_poly_init(c, m); nmod_poly_init(cinv, m); for (j = 0; j < num_threads + 1; j++) nmod_poly_init(tmp + j, m); nmod_poly_randtest_not_zero(a, state, 1 + n_randint(state, 20)); nmod_poly_randtest_not_zero(b, state, 1 + n_randint(state, 20)); do { nmod_poly_randtest_not_zero(c, state, 1 + n_randint(state, 20)); } while (c->length < 2); nmod_poly_reverse(cinv, c, c->length); nmod_poly_inv_series(cinv, cinv, c->length); nmod_mat_init(B, n_sqrt (c->length - 1) + 1, c->length - 1, m); nmod_poly_precompute_matrix(B, b, c, cinv); args1 = (nmod_poly_matrix_precompute_arg_t *) flint_malloc(sizeof(nmod_poly_matrix_precompute_arg_t) *(num_threads + 1)); C = (nmod_mat_struct *) flint_malloc(sizeof(nmod_mat_struct)*(num_threads + 1)); for (j = 0; j < num_threads + 1; j++) { nmod_mat_init(C + j, n_sqrt(c->length - 1) + 1, c->length - 1, m); nmod_poly_set(tmp + j, b); nmod_poly_rem(tmp + j, tmp + j, c); if (tmp[j].length < c->length - 1) { nmod_poly_fit_length(tmp + j, c->length - 1); _nmod_vec_zero(tmp[j].coeffs + tmp[j].length, c->length - 1 - b->length); } args1[j].A = C + j; args1[j].poly1 = tmp + j; args1[j].poly2 = c; args1[j].poly2inv = cinv; } for (j = 1; j < num_threads + 1; j++) { thread_pool_wake(global_thread_pool, threads[j - 1], 0, _nmod_poly_precompute_matrix_worker, &args1[j]); } _nmod_poly_precompute_matrix_worker(&args1[0]); for (j = 1; j < num_threads + 1; j++) thread_pool_wait(global_thread_pool, threads[j - 1]); for (j = 0; j < num_threads + 1; j++) { if (!nmod_mat_equal(B, C + j)) { flint_printf("FAIL (precomputation):\n"); flint_printf("B:\n"); nmod_mat_print_pretty(B); flint_printf("\n"); flint_printf("C[j]:\n"); nmod_mat_print_pretty(C + j); flint_printf("\n"); flint_printf("a:\n"); nmod_poly_print(a); flint_printf("\n"); flint_printf("b:\n"); nmod_poly_print(b); flint_printf("\n"); flint_printf("c:\n"); nmod_poly_print(c); flint_printf("\n"); abort(); } } flint_give_back_threads(threads, num_threads); nmod_poly_clear(a); nmod_poly_clear(b); nmod_mat_clear (B); nmod_poly_clear(c); nmod_poly_clear(cinv); for (j = 0; j < num_threads + 1; j++) { nmod_poly_clear(tmp + j); nmod_mat_clear(C + j); } flint_free(C); flint_free(tmp); flint_free(args1); } /* check composition */ for (i = 0; i < 100 * flint_test_multiplier(); i++) { nmod_poly_t a, b, c, cinv, d; nmod_poly_struct * res; nmod_mat_t B; mp_limb_t m = n_randtest_prime(state, 0); slong j, num_threads; nmod_poly_compose_mod_precomp_preinv_arg_t * args1; thread_pool_handle * threads; flint_set_num_threads(1 + n_randint(state, 3)); num_threads = flint_request_threads(&threads, FLINT_DEFAULT_THREAD_LIMIT); res = (nmod_poly_struct *) flint_malloc(sizeof(nmod_poly_struct)*(num_threads + 1)); nmod_poly_init(a, m); nmod_poly_init(b, m); nmod_poly_init(c, m); nmod_poly_init(cinv, m); nmod_poly_init(d, m); for (j = 0; j < num_threads + 1; j++) nmod_poly_init(res + j, m); nmod_poly_randtest_not_zero(a, state, 1 + n_randint(state, 20)); nmod_poly_randtest_not_zero(b, state, 1 + n_randint(state, 20)); do { nmod_poly_randtest_not_zero(c, state, n_randint(state, 20) + 1); } while (c->length < 2); nmod_poly_reverse(cinv, c, c->length); nmod_poly_inv_series(cinv, cinv, c->length); nmod_mat_init(B, n_sqrt(c->length - 1) + 1, c->length - 1, m); nmod_poly_precompute_matrix(B, b, c, cinv); nmod_poly_rem(a, a, c); nmod_poly_compose_mod(d, a, b, c); args1 = (nmod_poly_compose_mod_precomp_preinv_arg_t *) flint_malloc((num_threads + 1)* sizeof(nmod_poly_compose_mod_precomp_preinv_arg_t)); for (j = 0; j < num_threads + 1; j++) { nmod_poly_fit_length(res + j, c->length - 1); _nmod_poly_set_length(res + j, c->length - 1); flint_mpn_zero(res[j].coeffs, c->length - 1); args1[j].A = B; args1[j].res = res + j; args1[j].poly1 = a; args1[j].poly3 = c; args1[j].poly3inv = cinv; } for (j = 1; j < num_threads + 1; j++) { thread_pool_wake(global_thread_pool, threads[j - 1], 0, _nmod_poly_compose_mod_brent_kung_precomp_preinv_worker, &args1[j]); } _nmod_poly_compose_mod_brent_kung_precomp_preinv_worker(&args1[0]); _nmod_poly_normalise(res + 0); for (j = 1; j < num_threads + 1; j++) { thread_pool_wait(global_thread_pool, threads[j - 1]); _nmod_poly_normalise(res + j); } for (j = 0; j < num_threads + 1; j++) { if (!nmod_poly_equal(d, res + j)) { flint_printf("FAIL (composition):\n"); flint_printf("res[j]:\n"); nmod_poly_print(res + j); flint_printf("\n"); flint_printf("d:\n"); nmod_poly_print(d); flint_printf("\n"); flint_printf("a:\n"); nmod_poly_print(a); flint_printf("\n"); flint_printf("b:\n"); nmod_poly_print(b); flint_printf("\n"); flint_printf("c:\n"); nmod_poly_print(c); flint_printf("\n"); abort(); } } flint_give_back_threads(threads, num_threads); nmod_poly_clear(a); nmod_poly_clear(b); nmod_mat_clear (B); nmod_poly_clear(c); nmod_poly_clear(cinv); nmod_poly_clear(d); for (j = 0; j < num_threads + 1; j++) nmod_poly_clear(res + j); flint_free(res); flint_free(args1); } FLINT_TEST_CLEANUP(state); flint_printf("PASS\n"); return 0; #else FLINT_TEST_CLEANUP(state); flint_printf("SKIPPED\n"); return 0; #endif }