gcc/libgomp/testsuite/libgomp.c/doacross-3.c

229 lines
5.7 KiB
C
Raw Normal View History

omp-low.c (lower_omp_ordered): Add argument to GOMP_SMD_ORDERED_* internal calls - 0 if... gcc/ 2015-11-14 Jakub Jelinek <jakub@redhat.com> * omp-low.c (lower_omp_ordered): Add argument to GOMP_SMD_ORDERED_* internal calls - 0 if ordered simd and 1 for ordered threads simd. * tree-vectorizer.c (adjust_simduid_builtins): If GOMP_SIMD_ORDERED_* argument is 1, replace it with GOMP_ordered_* call instead of removing it. gcc/c/ 2015-11-14 Jakub Jelinek <jakub@redhat.com> * c-typeck.c (c_finish_omp_clauses): Don't mark GOMP_MAP_FIRSTPRIVATE_POINTER decls addressable. gcc/cp/ 2015-11-14 Jakub Jelinek <jakub@redhat.com> * semantics.c (finish_omp_clauses): Don't mark GOMP_MAP_FIRSTPRIVATE_POINTER decls addressable. libgomp/ 2015-11-14 Jakub Jelinek <jakub@redhat.com> Aldy Hernandez <aldyh@redhat.com> Ilya Verbin <ilya.verbin@intel.com> * ordered.c (gomp_doacross_init, GOMP_doacross_post, GOMP_doacross_wait, gomp_doacross_ull_init, GOMP_doacross_ull_post, GOMP_doacross_ull_wait): For GFS_GUIDED don't divide number of iterators or IV by chunk size. * parallel.c (gomp_resolve_num_threads): Don't assume that if thr->ts.team is non-NULL, then pool must be non-NULL. * libgomp-plugin.h (GOMP_PLUGIN_target_task_completion): Declare. * libgomp.map (GOMP_PLUGIN_1.1): New symbol version, export GOMP_PLUGIN_target_task_completion. * Makefile.am (libgomp_la_SOURCES): Add priority_queue.c. * Makefile.in: Regenerate. * libgomp.h: Shuffle prototypes and forward definitions around so priority queues can be defined. (enum gomp_task_kind): Add GOMP_TASK_ASYNC_RUNNING. (enum gomp_target_task_state): New enum. (struct gomp_target_task): Add state, tgt, task and team fields. (gomp_create_target_task): Change return type to bool, add state argument. (gomp_target_task_fn): Change return type to bool. (struct gomp_device_descr): Add async_run_func. (struct gomp_task): Remove children, next_child, prev_child, next_queue, prev_queue, next_taskgroup, prev_taskgroup. Add pnode field. (struct gomp_taskgroup): Remove children. Add taskgroup_queue. (struct gomp_team): Change task_queue type to a priority queue. (splay_compare): Define inline. (priority_queue_offset): New. (priority_node_to_task): New. (task_to_priority_node): New. * oacc-mem.c: Do not include splay-tree.h. * priority_queue.c: New file. * priority_queue.h: New file. * splay-tree.c: Do not include splay-tree.h. (splay_tree_foreach_internal): New. (splay_tree_foreach): New. * splay-tree.h: Become re-entrant if splay_tree_prefix is defined. (splay_tree_callback): Define typedef. * target.c (splay_compare): Move to libgomp.h. (GOMP_target): Don't adjust *thr in any way around running offloaded task. (GOMP_target_ext): Likewise. Handle target nowait. (GOMP_target_update_ext, GOMP_target_enter_exit_data): Check return value from gomp_create_target_task, if false, fallthrough as if no dependencies exist. (gomp_target_task_fn): Change return type to bool, return true if the task should have another part scheduled later. Handle target nowait. (gomp_load_plugin_for_device): Initialize async_run. * task.c (gomp_init_task): Initialize children_queue. (gomp_clear_parent_in_list): New. (gomp_clear_parent_in_tree): New. (gomp_clear_parent): Handle priorities. (GOMP_task): Likewise. (priority_queue_move_task_first, gomp_target_task_completion, GOMP_PLUGIN_target_task_completion): New functions. (gomp_create_target_task): Use priority queues. Change return type to bool, add state argument, return false if for async {{enter,exit} data,update} constructs no dependencies need to be waited for, handle target nowait. Set task->fn to NULL instead of gomp_target_task_fn. (verify_children_queue): Remove. (priority_list_upgrade_task): New. (priority_queue_upgrade_task): New. (verify_task_queue): Remove. (priority_list_downgrade_task): New. (priority_queue_downgrade_task): New. (gomp_task_run_pre): Use priority queues. Abstract code out to priority_queue_downgrade_task. (gomp_task_run_post_handle_dependers): Use priority queues. (gomp_task_run_post_remove_parent): Likewise. (gomp_task_run_post_remove_taskgroup): Likewise. (gomp_barrier_handle_tasks): Likewise. Handle target nowait target tasks specially. (GOMP_taskwait): Likewise. (gomp_task_maybe_wait_for_dependencies): Likewise. Abstract code to priority-queue_upgrade_task. (GOMP_taskgroup_start): Use priority queues. (GOMP_taskgroup_end): Likewise. Handle target nowait target tasks specially. If taskgroup is NULL, and thr->ts.level is 0, act as a barrier. * taskloop.c (GOMP_taskloop): Handle priorities. * team.c (gomp_new_team): Call priority_queue_init. (free_team): Call priority_queue_free. (gomp_free_thread): Call gomp_team_end if thr->ts.team is artificial team created for target nowait in implicit parallel region. (gomp_team_start): For nested check, test thr->ts.level instead of thr->ts.team != NULL. * testsuite/libgomp.c/doacross-3.c: New test. * testsuite/libgomp.c/ordered-5.c: New test. * testsuite/libgomp.c/priority.c: New test. * testsuite/libgomp.c/target-31.c: New test. * testsuite/libgomp.c/target-32.c: New test. * testsuite/libgomp.c/target-33.c: New test. * testsuite/libgomp.c/target-34.c: New test. liboffloadmic/ 2015-11-14 Ilya Verbin <ilya.verbin@intel.com> * runtime/offload_host.cpp (task_completion_callback): New variable. (offload_proxy_task_completed_ooo): Call task_completion_callback. (__offload_register_task_callback): New function. * runtime/offload_host.h (__offload_register_task_callback): New declaration. * plugin/libgomp-plugin-intelmic.cpp (offload): Add async_data argument, handle async offloading. (register_main_image): Call register_main_image. (GOMP_OFFLOAD_init_device, get_target_table, GOMP_OFFLOAD_alloc, GOMP_OFFLOAD_free, GOMP_OFFLOAD_host2dev, GOMP_OFFLOAD_dev2host, GOMP_OFFLOAD_dev2dev) Adjust offload callers. (GOMP_OFFLOAD_async_run): New function. (GOMP_OFFLOAD_run): Implement using GOMP_OFFLOAD_async_run. From-SVN: r230381
2015-11-14 19:42:13 +01:00
extern void abort (void);
#define N 256
int a[N], b[N / 16][8][4], c[N / 32][8][8], g[N / 16][8][6];
volatile int d, e;
volatile unsigned long long f;
int
main ()
{
unsigned long long i;
int j, k, l, m;
#pragma omp parallel private (l)
{
#pragma omp for schedule(guided, 3) ordered (1) nowait
for (i = 1; i < N + f; i++)
{
#pragma omp atomic write
a[i] = 1;
#pragma omp ordered depend(sink: i - 1)
if (i > 1)
{
#pragma omp atomic read
l = a[i - 1];
if (l < 2)
abort ();
}
#pragma omp atomic write
a[i] = 2;
if (i < N - 1)
{
#pragma omp atomic read
l = a[i + 1];
if (l == 3)
abort ();
}
#pragma omp ordered depend(source)
#pragma omp atomic write
a[i] = 3;
}
#pragma omp for schedule(guided) ordered (3) nowait
for (i = 3; i < N / 16 - 1 + f; i++)
for (j = 0; j < 8; j += 2)
for (k = 1; k <= 3; k++)
{
#pragma omp atomic write
b[i][j][k] = 1;
#pragma omp ordered depend(sink: i, j - 2, k - 1) \
depend(sink: i - 2, j - 2, k + 1)
#pragma omp ordered depend(sink: i - 3, j + 2, k - 2)
if (j >= 2 && k > 1)
{
#pragma omp atomic read
l = b[i][j - 2][k - 1];
if (l < 2)
abort ();
}
#pragma omp atomic write
b[i][j][k] = 2;
if (i >= 5 && j >= 2 && k < 3)
{
#pragma omp atomic read
l = b[i - 2][j - 2][k + 1];
if (l < 2)
abort ();
}
if (i >= 6 && j < N / 16 - 3 && k == 3)
{
#pragma omp atomic read
l = b[i - 3][j + 2][k - 2];
if (l < 2)
abort ();
}
#pragma omp ordered depend(source)
#pragma omp atomic write
b[i][j][k] = 3;
}
#define A(n) int n;
#define B(n) A(n##0) A(n##1) A(n##2) A(n##3)
#define C(n) B(n##0) B(n##1) B(n##2) B(n##3)
#define D(n) C(n##0) C(n##1) C(n##2) C(n##3)
D(m)
#undef A
#pragma omp for collapse (2) ordered(61) schedule(guided, 15)
for (i = 2; i < N / 32 + f; i++)
for (j = 7; j > 1; j--)
for (k = 6; k >= 0; k -= 2)
#define A(n) for (n = 4; n < 5; n++)
D(m)
#undef A
{
#pragma omp atomic write
c[i][j][k] = 1;
#define A(n) ,n
#define E(n) C(n##0) C(n##1) C(n##2) B(n##30) B(n##31) A(n##320) A(n##321)
#pragma omp ordered depend (sink: i, j, k + 2 E(m)) \
depend (sink:i - 2, j + 1, k - 4 E(m)) \
depend(sink: i - 1, j - 2, k - 2 E(m))
if (k <= 4)
{
omp-low.c (lower_omp_target): Fix up argument to is_reference. gcc/ * omp-low.c (lower_omp_target): Fix up argument to is_reference. (expand_omp_ordered_sink): Handle TREE_PURPOSE of deps being TRUNC_DIV_EXPR. * gimplify.c (gimplify_scan_omp_clauses): Likewise. Set ctx->target_map_scalars_firstprivate on OMP_TARGET even for Fortran. Remove omp_no_lastprivate callers. Propagate lastprivate on combined teams distribute parallel for simd even to distribute and teams construct. For OMP_CLAUSE_DEPEND add missing break at the end of OMP_CLAUSE_DEPEND_SINK case. (omp_notice_variable): Use lang_hooks.decls.omp_scalar_p. (omp_no_lastprivate): Removed. (gimplify_adjust_omp_clauses): Remove omp_no_lastprivate callers. (gimplify_omp_for): Likewise. (computable_teams_clause): Fail for automatic vars from current function not yet seen in bind expr. * langhooks.c (lhd_omp_scalar_p): New function. * langhooks.h (struct lang_hooks_for_decls): Add omp_scalar_p. * varpool.c (varpool_node::get_create): Set node->offloading even for DECL_EXTERNAL decls. * langhooks-def.h (lhd_omp_scalar_p): New prototype. (LANG_HOOKS_OMP_SCALAR_P): Define. (LANG_HOOKS_DECLS): Use it. gcc/fortran/ * openmp.c (gfc_free_omp_clauses): Free critical_name, grainsize, hint, num_tasks, priority and if_exprs. (gfc_match_omp_to_link, gfc_match_omp_depend_sink): New functions. (enum omp_mask1, enum omp_mask2): New enums. Change all OMP_CLAUSE_* defines into enum values, and change their values from ((uint64_t) 1 << bit) to just bit. (omp_mask, omp_inv_mask): New classes. Add ctors and operators. (gfc_match_omp_clauses): Change mask argument from uint64_t to const omp_mask. Assert OMP_MASK1_LAST and OMP_MASK2_LAST are at most 64. Move delete clause handling to where it alphabetically belongs. Parse defaultmap, grainsize, hint, is_device_ptr, nogroup, nowait, num_tasks, priority, simd, threads and use_device_ptr clauses. Parse if clause modifier. Parse map clause always modifier, and release and delete kinds. Parse ordered clause with argument. Parse schedule clause modifiers. Differentiate device clause parsing based on openacc flag. Guard link clause parsing with openacc flag. Add support for parsing linear clause modifiers. Parse depend(source) and depend(sink: ...). Use gfc_match_omp_to_link for to and link clauses in declare target construct. (match_acc): Change mask type from uint64_t to const omp_mask. (OMP_SINGLE_CLAUSES, OMP_ORDERED_CLAUSES, OMP_DECLARE_TARGET_CLAUSES, OMP_TASKLOOP_CLAUSES, OMP_TARGET_ENTER_DATA_CLAUSES, OMP_TARGET_EXIT_DATA_CLAUSES): Define. (OACC_PARALLEL_CLAUSES, OACC_KERNELS_CLAUSES, OACC_DATA_CLAUSES, OACC_LOOP_CLAUSES, OACC_HOST_DATA_CLAUSES, OACC_DECLARE_CLAUSES, OACC_ENTER_DATA_CLAUSES, OACC_EXIT_DATA_CLAUSES, OACC_WAIT_CLAUSES, OACC_ROUTINE_CLAUSES, OMP_PARALLEL_CLAUSES, OMP_DECLARE_SIMD_CLAUSES, OMP_SECTIONS_CLAUSES, OMP_TEAMS_CLAUSES, OMP_DISTRIBUTE_CLAUSES): Replace first or only OMP_CLAUSE_* value in bitset with omp_mask (OMP_CLAUSE_*). (OMP_DO_CLAUSES): Likewise. Add OMP_CLAUSE_LINEAR. (OMP_SIMD_CLAUSES): Replace first or only OMP_CLAUSE_* value in bitset with omp_mask (OMP_CLAUSE_*). Add OMP_CLAUSE_SIMDLEN. (OACC_UPDATE_CLAUSES): Replace first or only OMP_CLAUSE_* value in bitset with omp_mask (OMP_CLAUSE_*). Replace OMP_CLAUSE_OACC_DEVICE with OMP_CLAUSE_DEVICE. (OMP_TASK_CLAUSES): Replace first or only OMP_CLAUSE_* value in bitset with omp_mask (OMP_CLAUSE_*). Add OMP_CLAUSE_PRIORITY. (OMP_TARGET_CLAUSES): Replace first or only OMP_CLAUSE_* value in bitset with omp_mask (OMP_CLAUSE_*). Add OMP_CLAUSE_DEPEND, OMP_CLAUSE_NOWAIT, OMP_CLAUSE_PRIVATE, OMP_CLAUSE_FIRSTPRIVATE, OMP_CLAUSE_DEFAULTMAP and OMP_CLAUSE_IS_DEVICE_PTR. (OMP_TARGET_DATA_CLAUSES): Replace first or only OMP_CLAUSE_* value in bitset with omp_mask (OMP_CLAUSE_*). Add OMP_CLAUSE_USE_DEVICE_PTR. (OMP_TARGET_UPDATE_CLAUSES): Replace first or only OMP_CLAUSE_* value in bitset with omp_mask (OMP_CLAUSE_*). Add OMP_CLAUSE_DEPEND and OMP_CLAUSE_NOWAIT. (match_omp): Change mask argument from unsigned int to const omp_mask. (gfc_match_omp_critical): Parse optional clauses and use omp_clauses union member instead of omp_name. (gfc_match_omp_end_critical): New function. (gfc_match_omp_distribute_parallel_do): Remove ordered and linear clauses from the mask. (gfc_match_omp_distribute_parallel_do_simd): Use & ~(omp_mask (OMP_CLAUSE_*)) instead of & ~OMP_CLAUSE_*. (gfc_match_omp_target_teams_distribute_parallel_do_simd): Likewise. (gfc_match_omp_teams_distribute_parallel_do_simd): Likewise. (gfc_match_omp_do_simd): Likewise. Don't remove ordered clause from the mask. (gfc_match_omp_parallel_do_simd): Likewise. (gfc_match_omp_target_teams_distribute_parallel_do): Likewise. (gfc_match_omp_teams_distribute_parallel_do): Likewise. (gfc_match_omp_declare_simd): If not using the form with (proc-name), require space before first clause. Make (proc-name) optional. If not present, set proc_name to NULL. (gfc_match_omp_declare_target): Rewritten for OpenMP 4.5. (gfc_match_omp_single): Use OMP_SINGLE_CLAUSES. (gfc_match_omp_task, gfc_match_omp_taskwait, gfc_match_omp_taskyield): Move around to where they belong alphabetically. (gfc_match_omp_target_enter_data, gfc_match_omp_target_exit_data, gfc_match_omp_target_parallel, gfc_match_omp_target_parallel_do, gfc_match_omp_target_parallel_do_simd, gfc_match_omp_target_simd, gfc_match_omp_taskloop, gfc_match_omp_taskloop_simd): New functions. (gfc_match_omp_ordered): Parse clauses. (gfc_match_omp_ordered_depend): New function. (gfc_match_omp_cancel, gfc_match_omp_end_single): Use omp_mask (OMP_CLAUSE_*) instead of OMP_CLAUSE_*. (resolve_oacc_scalar_int_expr): Renamed to ... (resolve_scalar_int_expr): ... this. Fix up formatting. (resolve_oacc_positive_int_expr): Renamed to ... (resolve_positive_int_expr): ... this. Fix up formatting. (resolve_nonnegative_int_expr): New function. (resolve_omp_clauses): Adjust callers, use the above functions even for OpenMP clauses, add handling of new OpenMP 4.5 clauses. Require orderedc >= collapse if specified. Handle depend(sink:) and depend(source) restrictions. Disallow linear clause when orderedc is non-zero. Diagnose linear clause modifiers when not in declare simd. Only check for integer type if ref modifier is not used. Remove diagnostics for required VALUE attribute. Diagnose VALUE attribute with ref or uval modifiers. Allow non-constant linear-step, if it is a dummy argument alone and is mentioned in uniform clause. Diagnose map kinds not allowed for various constructs. Diagnose target {enter ,exit ,}data without any map clauses. Add dummy OMP_LIST_IS_DEVICE_PTR and OMP_LIST_USE_DEVICE_PTR cases. (gfc_resolve_omp_do_blocks): Set omp_current_do_collapse to orderedc if non-zero. (gfc_resolve_omp_parallel_blocks): Handle new OpenMP 4.5 constructs, replace underscores with spaces in a few construct names. (resolve_omp_do): Set collapse to orderedc if non-zero. Handle new OpenMP 4.5 constructs. (resolve_oacc_loop_blocks): Call resolve_positive_int_expr instead of resolve_oacc_positive_int_expr. (gfc_resolve_omp_directive): Handle new OpenMP 4.5 constructs. (gfc_resolve_omp_declare_simd): Allow ods->proc_name to be NULL. * trans-openmp.c (gfc_omp_scalar_p): New function. (doacross_steps): New variable. (gfc_trans_omp_clauses): Handle new OpenMP 4.5 clauses and new clause modifiers. (gfc_trans_omp_critical): Adjust EXEC_OMP_CRITICAL handling. (gfc_trans_omp_do): Handle doacross loops. Clear sched_simd flag. Handle EXEC_OMP_TASKLOOP. (gfc_trans_omp_ordered): Translate omp clauses, allow NULL code->block. (GFC_OMP_SPLIT_TASKLOOP, GFC_OMP_MASK_TASKLOOP): New enum constants. (gfc_split_omp_clauses): Copy orderedc together with ordered. Change firstprivate and lastprivate handling for OpenMP 4.5. Handle EXEC_OMP_TARGET_SIMD, EXEC_OMP_TARGET_PARALLEL{,_DO,_DO_SIMD} and EXEC_OMP_TASKLOOP{,_SIMD}. Add handling for new OpenMP 4.5 clauses and clause modifiers and handle if clause without/with modifiers. (gfc_trans_omp_teams): Add omp_clauses argument, add it to other teams clauses. Don't wrap into OMP_TEAMS if -fopenmp-simd. (gfc_trans_omp_target): For -fopenmp, translate num_teams and thread_limit clauses on combined target teams early and pass to gfc_trans_omp_teams. Set OMP_TARGET_COMBINED if needed. Handle EXEC_OMP_TARGET_PARALLEL{,_DO,_DO_SIMD} and EXEC_OMP_TARGET_SIMD. (gfc_trans_omp_taskloop, gfc_trans_omp_target_enter_data, gfc_trans_omp_target_exit_data): New functions. (gfc_trans_omp_directive): Handle EXEC_OMP_TARGET_{ENTER,EXIT}_DATA EXEC_OMP_TASKLOOP{,_SIMD}, EXEC_OMP_TARGET_PARALLEL{,_DO,_DO_SIMD} and EXEC_OMP_TARGET_SIMD. Adjust gfc_trans_omp_teams caller. * symbol.c (check_conflict): Handle omp_declare_target_link. (gfc_add_omp_declare_target_link): New function. (gfc_copy_attr): Copy omp_declare_target_link. * dump-parse-tree.c (show_omp_namelist): Handle OMP_DEPEND_SINK_FIRST depend_op. Print linear clause modifiers. (show_omp_clauses): Adjust for OpenMP 4.5 clause changes. (show_omp_node): Print clauses for EXEC_OMP_ORDERED. Allow NULL c->block for EXEC_OMP_ORDERED. Formatting fixes. Adjust handling of EXEC_OMP_CRITICAL, handle new OpenMP 4.5 constructs and some forgotten OpenMP 4.0 constructs. (show_code_node): Handle new OpenMP 4.5 constructs and some forgotten OpenMP 4.0 constructs. * gfortran.h (symbol_attribute): Add omp_declare_target_link bitfield. (struct gfc_omp_namelist): Add u.common and u.linear_op fields. (struct gfc_common_head): Change omp_declare_target into bitfield. Add omp_declare_target_link bitfield. (gfc_add_omp_declare_target_link): New prototype. (enum gfc_statement): Add ST_OMP_TARGET_PARALLEL, ST_OMP_END_TARGET_PARALLEL, ST_OMP_TARGET_PARALLEL_DO, ST_OMP_END_TARGET_PARALLEL_DO, ST_OMP_TARGET_PARALLEL_DO_SIMD, ST_OMP_END_TARGET_PARALLEL_DO_SIMD, ST_OMP_TARGET_ENTER_DATA, ST_OMP_TARGET_EXIT_DATA, ST_OMP_TARGET_SIMD, ST_OMP_END_TARGET_SIMD, ST_OMP_TASKLOOP, ST_OMP_END_TASKLOOP, ST_OMP_TASKLOOP_SIMD, ST_OMP_END_TASKLOOP_SIMD and ST_OMP_ORDERED_DEPEND. (enum gfc_omp_depend_op): Add OMP_DEPEND_SINK_FIRST and OMP_DEPEND_SINK. (enum gfc_omp_linear_op): New. (struct gfc_omp_clauses): Add critical_name, depend_source, orderedc, defaultmap, nogroup, sched_simd, sched_monotonic, sched_nonmonotonic, simd, threads, grainsize, hint, num_tasks, priority and if_exprs fields. (enum gfc_exec_op): Add EXEC_OMP_END_CRITICAL, EXEC_OMP_TARGET_ENTER_DATA, EXEC_OMP_TARGET_EXIT_DATA, EXEC_OMP_TARGET_PARALLEL, EXEC_OMP_TARGET_PARALLEL_DO, EXEC_OMP_TARGET_PARALLEL_DO_SIMD, EXEC_OMP_TARGET_SIMD, EXEC_OMP_TASKLOOP, EXEC_OMP_TASKLOOP_SIMD. (enum gfc_omp_map_op): Add OMP_MAP_RELEASE, OMP_MAP_ALWAYS_TO, OMP_MAP_ALWAYS_FROM and OMP_MAP_ALWAYS_TOFROM. (OMP_LIST_IS_DEVICE_PTR, OMP_LIST_USE_DEVICE_PTR): New. (enum gfc_omp_if_kind): New. * module.c (enum ab_attribute): Add AB_OMP_DECLARE_TARGET_LINK. (attr_bits): Add AB_OMP_DECLARE_TARGET_LINK entry. (mio_symbol_attribute): Save and restore omp_declare_target_link bit. * trans.h (gfc_omp_scalar_p): New prototype. * frontend-passes.c (gfc_code_walker): Handle new OpenMP 4.5 expressions. * trans.c (trans_code): Handle new OpenMP 4.5 constructs. * resolve.c (gfc_resolve_blocks): Likewise. (gfc_resolve_code): Likewise. * f95-lang.c (LANG_HOOKS_OMP_SCALAR_P): Redefine to gfc_omp_scalar_p. (gfc_attribute_table): Add "omp declare target link". * st.c (gfc_free_statement): Handle EXEC_OMP_END_CRITICAL like EXEC_OMP_CRITICAL before, free clauses for EXEC_OMP_CRITICAL and new OpenMP 4.5 constructs. Free omp clauses even for EXEC_OMP_ORDERED. * match.c (match_exit_cycle): Rename collapse variable to count, set it to orderedc if non-zero, instead of collapse. * trans-decl.c (add_attributes_to_decl): Add "omp declare target link" instead of "omp declare target" for omp_declare_target_link. * trans-common.c (build_common_decl): Likewise. * match.h (gfc_match_omp_target_enter_data, gfc_match_omp_target_exit_data, gfc_match_omp_target_parallel, gfc_match_omp_target_parallel_do, gfc_match_omp_target_parallel_do_simd, gfc_match_omp_target_simd, gfc_match_omp_taskloop, gfc_match_omp_taskloop_simd, gfc_match_omp_end_critical, gfc_match_omp_ordered_depend): New prototypes. * parse.c (decode_omp_directive): Use gfc_match_omp_end_critical instead of gfc_match_omp_critical for !$omp end critical. Handle new OpenMP 4.5 constructs. If ordered directive has depend clause as the first of the clauses, use gfc_match_omp_ordered_depend and ST_OMP_ORDERED_DEPEND instead of gfc_match_omp_ordered and ST_OMP_ORDERED. (case_executable): Add ST_OMP_TARGET_ENTER_DATA, ST_OMP_TARGET_EXIT_DATA and ST_OMP_ORDERED_DEPEND cases. (case_exec_markers): Add ST_OMP_TARGET_PARALLEL, ST_OMP_TARGET_PARALLEL_DO, ST_OMP_TARGET_PARALLEL_DO_SIMD, ST_OMP_TARGET_SIMD, ST_OMP_TASKLOOP and ST_OMP_TASKLOOP_SIMD cases. (gfc_ascii_statement): Handle new OpenMP 4.5 constructs. (parse_omp_do): Handle ST_OMP_TARGET_PARALLEL_DO, ST_OMP_TARGET_PARALLEL_DO_SIMD, ST_OMP_TASKLOOP and ST_OMP_TASKLOOP_SIMD. (parse_omp_structured_block): Handle EXEC_OMP_END_CRITICAL instead of EXEC_OMP_CRITICAL, adjust for EXEC_OMP_CRITICAL having omp clauses now. (parse_executable): Handle ST_OMP_TARGET_PARALLEL, ST_OMP_TARGET_PARALLEL_DO, ST_OMP_TARGET_PARALLEL_DO_SIMD, ST_OMP_TASKLOOP and ST_OMP_TASKLOOP_SIMD. gcc/testsuite/ * gfortran.dg/gomp/pr77516.f90: Add dg-warning. * gfortran.dg/gomp/target1.f90: Remove ordered clause where it is no longer allowed and corresponding ordered construct. * gfortran.dg/gomp/linear-1.f90: New test. * gfortran.dg/gomp/declare-simd-2.f90: New test. * gfortran.dg/gomp/declare-target-1.f90: New test. * gfortran.dg/gomp/declare-target-2.f90: New test. libgomp/ * testsuite/libgomp.fortran/examples-4/declare_target-1.f90 (fib_wrapper): Add map(from: x) clause. * testsuite/libgomp.fortran/examples-4/declare_target-2.f90 (e_53_2): Likewise. * testsuite/libgomp.fortran/examples-4/declare_target-4.f90 (accum): Add map(tmp) clause. * testsuite/libgomp.fortran/examples-4/declare_target-5.f90 (accum): Add map(tofrom: tmp) clause. * testsuite/libgomp.fortran/examples-4/target_data-3.f90 (gramSchmidt): Likewise. * testsuite/libgomp.fortran/examples-4/teams-2.f90 (dotprod): Add map(tofrom: sum) clause. * testsuite/libgomp.fortran/nestedfn5.f90 (foo): Add twice map (alloc: a, l) clause. Add defaultmap(tofrom: scalar) clause. * testsuite/libgomp.fortran/pr66199-2.f90: Adjust for linear clause only allowed on the loop iterator. * testsuite/libgomp.fortran/target4.f90 (foo): Add map(t) clause. * testsuite/libgomp.fortran/taskloop2.f90: New test. * testsuite/libgomp.fortran/taskloop4.f90: New test. * testsuite/libgomp.fortran/doacross1.f90: New test. * testsuite/libgomp.fortran/doacross3.f90: New test. * testsuite/libgomp.fortran/taskloop1.f90: New test. * testsuite/libgomp.fortran/taskloop3.f90: New test. * testsuite/libgomp.fortran/doacross2.f90: New test. * testsuite/libgomp.c/doacross-1.c (main): Add missing #pragma omp atomic read. * testsuite/libgomp.c/doacross-2.c (main): Likewise. * testsuite/libgomp.c/doacross-3.c (main): Likewise. From-SVN: r242037
2016-11-10 12:38:05 +01:00
#pragma omp atomic read
omp-low.c (lower_omp_ordered): Add argument to GOMP_SMD_ORDERED_* internal calls - 0 if... gcc/ 2015-11-14 Jakub Jelinek <jakub@redhat.com> * omp-low.c (lower_omp_ordered): Add argument to GOMP_SMD_ORDERED_* internal calls - 0 if ordered simd and 1 for ordered threads simd. * tree-vectorizer.c (adjust_simduid_builtins): If GOMP_SIMD_ORDERED_* argument is 1, replace it with GOMP_ordered_* call instead of removing it. gcc/c/ 2015-11-14 Jakub Jelinek <jakub@redhat.com> * c-typeck.c (c_finish_omp_clauses): Don't mark GOMP_MAP_FIRSTPRIVATE_POINTER decls addressable. gcc/cp/ 2015-11-14 Jakub Jelinek <jakub@redhat.com> * semantics.c (finish_omp_clauses): Don't mark GOMP_MAP_FIRSTPRIVATE_POINTER decls addressable. libgomp/ 2015-11-14 Jakub Jelinek <jakub@redhat.com> Aldy Hernandez <aldyh@redhat.com> Ilya Verbin <ilya.verbin@intel.com> * ordered.c (gomp_doacross_init, GOMP_doacross_post, GOMP_doacross_wait, gomp_doacross_ull_init, GOMP_doacross_ull_post, GOMP_doacross_ull_wait): For GFS_GUIDED don't divide number of iterators or IV by chunk size. * parallel.c (gomp_resolve_num_threads): Don't assume that if thr->ts.team is non-NULL, then pool must be non-NULL. * libgomp-plugin.h (GOMP_PLUGIN_target_task_completion): Declare. * libgomp.map (GOMP_PLUGIN_1.1): New symbol version, export GOMP_PLUGIN_target_task_completion. * Makefile.am (libgomp_la_SOURCES): Add priority_queue.c. * Makefile.in: Regenerate. * libgomp.h: Shuffle prototypes and forward definitions around so priority queues can be defined. (enum gomp_task_kind): Add GOMP_TASK_ASYNC_RUNNING. (enum gomp_target_task_state): New enum. (struct gomp_target_task): Add state, tgt, task and team fields. (gomp_create_target_task): Change return type to bool, add state argument. (gomp_target_task_fn): Change return type to bool. (struct gomp_device_descr): Add async_run_func. (struct gomp_task): Remove children, next_child, prev_child, next_queue, prev_queue, next_taskgroup, prev_taskgroup. Add pnode field. (struct gomp_taskgroup): Remove children. Add taskgroup_queue. (struct gomp_team): Change task_queue type to a priority queue. (splay_compare): Define inline. (priority_queue_offset): New. (priority_node_to_task): New. (task_to_priority_node): New. * oacc-mem.c: Do not include splay-tree.h. * priority_queue.c: New file. * priority_queue.h: New file. * splay-tree.c: Do not include splay-tree.h. (splay_tree_foreach_internal): New. (splay_tree_foreach): New. * splay-tree.h: Become re-entrant if splay_tree_prefix is defined. (splay_tree_callback): Define typedef. * target.c (splay_compare): Move to libgomp.h. (GOMP_target): Don't adjust *thr in any way around running offloaded task. (GOMP_target_ext): Likewise. Handle target nowait. (GOMP_target_update_ext, GOMP_target_enter_exit_data): Check return value from gomp_create_target_task, if false, fallthrough as if no dependencies exist. (gomp_target_task_fn): Change return type to bool, return true if the task should have another part scheduled later. Handle target nowait. (gomp_load_plugin_for_device): Initialize async_run. * task.c (gomp_init_task): Initialize children_queue. (gomp_clear_parent_in_list): New. (gomp_clear_parent_in_tree): New. (gomp_clear_parent): Handle priorities. (GOMP_task): Likewise. (priority_queue_move_task_first, gomp_target_task_completion, GOMP_PLUGIN_target_task_completion): New functions. (gomp_create_target_task): Use priority queues. Change return type to bool, add state argument, return false if for async {{enter,exit} data,update} constructs no dependencies need to be waited for, handle target nowait. Set task->fn to NULL instead of gomp_target_task_fn. (verify_children_queue): Remove. (priority_list_upgrade_task): New. (priority_queue_upgrade_task): New. (verify_task_queue): Remove. (priority_list_downgrade_task): New. (priority_queue_downgrade_task): New. (gomp_task_run_pre): Use priority queues. Abstract code out to priority_queue_downgrade_task. (gomp_task_run_post_handle_dependers): Use priority queues. (gomp_task_run_post_remove_parent): Likewise. (gomp_task_run_post_remove_taskgroup): Likewise. (gomp_barrier_handle_tasks): Likewise. Handle target nowait target tasks specially. (GOMP_taskwait): Likewise. (gomp_task_maybe_wait_for_dependencies): Likewise. Abstract code to priority-queue_upgrade_task. (GOMP_taskgroup_start): Use priority queues. (GOMP_taskgroup_end): Likewise. Handle target nowait target tasks specially. If taskgroup is NULL, and thr->ts.level is 0, act as a barrier. * taskloop.c (GOMP_taskloop): Handle priorities. * team.c (gomp_new_team): Call priority_queue_init. (free_team): Call priority_queue_free. (gomp_free_thread): Call gomp_team_end if thr->ts.team is artificial team created for target nowait in implicit parallel region. (gomp_team_start): For nested check, test thr->ts.level instead of thr->ts.team != NULL. * testsuite/libgomp.c/doacross-3.c: New test. * testsuite/libgomp.c/ordered-5.c: New test. * testsuite/libgomp.c/priority.c: New test. * testsuite/libgomp.c/target-31.c: New test. * testsuite/libgomp.c/target-32.c: New test. * testsuite/libgomp.c/target-33.c: New test. * testsuite/libgomp.c/target-34.c: New test. liboffloadmic/ 2015-11-14 Ilya Verbin <ilya.verbin@intel.com> * runtime/offload_host.cpp (task_completion_callback): New variable. (offload_proxy_task_completed_ooo): Call task_completion_callback. (__offload_register_task_callback): New function. * runtime/offload_host.h (__offload_register_task_callback): New declaration. * plugin/libgomp-plugin-intelmic.cpp (offload): Add async_data argument, handle async offloading. (register_main_image): Call register_main_image. (GOMP_OFFLOAD_init_device, get_target_table, GOMP_OFFLOAD_alloc, GOMP_OFFLOAD_free, GOMP_OFFLOAD_host2dev, GOMP_OFFLOAD_dev2host, GOMP_OFFLOAD_dev2dev) Adjust offload callers. (GOMP_OFFLOAD_async_run): New function. (GOMP_OFFLOAD_run): Implement using GOMP_OFFLOAD_async_run. From-SVN: r230381
2015-11-14 19:42:13 +01:00
l = c[i][j][k + 2];
if (l < 2)
abort ();
}
#pragma omp atomic write
c[i][j][k] = 2;
if (i >= 4 && j < 7 && k >= 4)
{
omp-low.c (lower_omp_target): Fix up argument to is_reference. gcc/ * omp-low.c (lower_omp_target): Fix up argument to is_reference. (expand_omp_ordered_sink): Handle TREE_PURPOSE of deps being TRUNC_DIV_EXPR. * gimplify.c (gimplify_scan_omp_clauses): Likewise. Set ctx->target_map_scalars_firstprivate on OMP_TARGET even for Fortran. Remove omp_no_lastprivate callers. Propagate lastprivate on combined teams distribute parallel for simd even to distribute and teams construct. For OMP_CLAUSE_DEPEND add missing break at the end of OMP_CLAUSE_DEPEND_SINK case. (omp_notice_variable): Use lang_hooks.decls.omp_scalar_p. (omp_no_lastprivate): Removed. (gimplify_adjust_omp_clauses): Remove omp_no_lastprivate callers. (gimplify_omp_for): Likewise. (computable_teams_clause): Fail for automatic vars from current function not yet seen in bind expr. * langhooks.c (lhd_omp_scalar_p): New function. * langhooks.h (struct lang_hooks_for_decls): Add omp_scalar_p. * varpool.c (varpool_node::get_create): Set node->offloading even for DECL_EXTERNAL decls. * langhooks-def.h (lhd_omp_scalar_p): New prototype. (LANG_HOOKS_OMP_SCALAR_P): Define. (LANG_HOOKS_DECLS): Use it. gcc/fortran/ * openmp.c (gfc_free_omp_clauses): Free critical_name, grainsize, hint, num_tasks, priority and if_exprs. (gfc_match_omp_to_link, gfc_match_omp_depend_sink): New functions. (enum omp_mask1, enum omp_mask2): New enums. Change all OMP_CLAUSE_* defines into enum values, and change their values from ((uint64_t) 1 << bit) to just bit. (omp_mask, omp_inv_mask): New classes. Add ctors and operators. (gfc_match_omp_clauses): Change mask argument from uint64_t to const omp_mask. Assert OMP_MASK1_LAST and OMP_MASK2_LAST are at most 64. Move delete clause handling to where it alphabetically belongs. Parse defaultmap, grainsize, hint, is_device_ptr, nogroup, nowait, num_tasks, priority, simd, threads and use_device_ptr clauses. Parse if clause modifier. Parse map clause always modifier, and release and delete kinds. Parse ordered clause with argument. Parse schedule clause modifiers. Differentiate device clause parsing based on openacc flag. Guard link clause parsing with openacc flag. Add support for parsing linear clause modifiers. Parse depend(source) and depend(sink: ...). Use gfc_match_omp_to_link for to and link clauses in declare target construct. (match_acc): Change mask type from uint64_t to const omp_mask. (OMP_SINGLE_CLAUSES, OMP_ORDERED_CLAUSES, OMP_DECLARE_TARGET_CLAUSES, OMP_TASKLOOP_CLAUSES, OMP_TARGET_ENTER_DATA_CLAUSES, OMP_TARGET_EXIT_DATA_CLAUSES): Define. (OACC_PARALLEL_CLAUSES, OACC_KERNELS_CLAUSES, OACC_DATA_CLAUSES, OACC_LOOP_CLAUSES, OACC_HOST_DATA_CLAUSES, OACC_DECLARE_CLAUSES, OACC_ENTER_DATA_CLAUSES, OACC_EXIT_DATA_CLAUSES, OACC_WAIT_CLAUSES, OACC_ROUTINE_CLAUSES, OMP_PARALLEL_CLAUSES, OMP_DECLARE_SIMD_CLAUSES, OMP_SECTIONS_CLAUSES, OMP_TEAMS_CLAUSES, OMP_DISTRIBUTE_CLAUSES): Replace first or only OMP_CLAUSE_* value in bitset with omp_mask (OMP_CLAUSE_*). (OMP_DO_CLAUSES): Likewise. Add OMP_CLAUSE_LINEAR. (OMP_SIMD_CLAUSES): Replace first or only OMP_CLAUSE_* value in bitset with omp_mask (OMP_CLAUSE_*). Add OMP_CLAUSE_SIMDLEN. (OACC_UPDATE_CLAUSES): Replace first or only OMP_CLAUSE_* value in bitset with omp_mask (OMP_CLAUSE_*). Replace OMP_CLAUSE_OACC_DEVICE with OMP_CLAUSE_DEVICE. (OMP_TASK_CLAUSES): Replace first or only OMP_CLAUSE_* value in bitset with omp_mask (OMP_CLAUSE_*). Add OMP_CLAUSE_PRIORITY. (OMP_TARGET_CLAUSES): Replace first or only OMP_CLAUSE_* value in bitset with omp_mask (OMP_CLAUSE_*). Add OMP_CLAUSE_DEPEND, OMP_CLAUSE_NOWAIT, OMP_CLAUSE_PRIVATE, OMP_CLAUSE_FIRSTPRIVATE, OMP_CLAUSE_DEFAULTMAP and OMP_CLAUSE_IS_DEVICE_PTR. (OMP_TARGET_DATA_CLAUSES): Replace first or only OMP_CLAUSE_* value in bitset with omp_mask (OMP_CLAUSE_*). Add OMP_CLAUSE_USE_DEVICE_PTR. (OMP_TARGET_UPDATE_CLAUSES): Replace first or only OMP_CLAUSE_* value in bitset with omp_mask (OMP_CLAUSE_*). Add OMP_CLAUSE_DEPEND and OMP_CLAUSE_NOWAIT. (match_omp): Change mask argument from unsigned int to const omp_mask. (gfc_match_omp_critical): Parse optional clauses and use omp_clauses union member instead of omp_name. (gfc_match_omp_end_critical): New function. (gfc_match_omp_distribute_parallel_do): Remove ordered and linear clauses from the mask. (gfc_match_omp_distribute_parallel_do_simd): Use & ~(omp_mask (OMP_CLAUSE_*)) instead of & ~OMP_CLAUSE_*. (gfc_match_omp_target_teams_distribute_parallel_do_simd): Likewise. (gfc_match_omp_teams_distribute_parallel_do_simd): Likewise. (gfc_match_omp_do_simd): Likewise. Don't remove ordered clause from the mask. (gfc_match_omp_parallel_do_simd): Likewise. (gfc_match_omp_target_teams_distribute_parallel_do): Likewise. (gfc_match_omp_teams_distribute_parallel_do): Likewise. (gfc_match_omp_declare_simd): If not using the form with (proc-name), require space before first clause. Make (proc-name) optional. If not present, set proc_name to NULL. (gfc_match_omp_declare_target): Rewritten for OpenMP 4.5. (gfc_match_omp_single): Use OMP_SINGLE_CLAUSES. (gfc_match_omp_task, gfc_match_omp_taskwait, gfc_match_omp_taskyield): Move around to where they belong alphabetically. (gfc_match_omp_target_enter_data, gfc_match_omp_target_exit_data, gfc_match_omp_target_parallel, gfc_match_omp_target_parallel_do, gfc_match_omp_target_parallel_do_simd, gfc_match_omp_target_simd, gfc_match_omp_taskloop, gfc_match_omp_taskloop_simd): New functions. (gfc_match_omp_ordered): Parse clauses. (gfc_match_omp_ordered_depend): New function. (gfc_match_omp_cancel, gfc_match_omp_end_single): Use omp_mask (OMP_CLAUSE_*) instead of OMP_CLAUSE_*. (resolve_oacc_scalar_int_expr): Renamed to ... (resolve_scalar_int_expr): ... this. Fix up formatting. (resolve_oacc_positive_int_expr): Renamed to ... (resolve_positive_int_expr): ... this. Fix up formatting. (resolve_nonnegative_int_expr): New function. (resolve_omp_clauses): Adjust callers, use the above functions even for OpenMP clauses, add handling of new OpenMP 4.5 clauses. Require orderedc >= collapse if specified. Handle depend(sink:) and depend(source) restrictions. Disallow linear clause when orderedc is non-zero. Diagnose linear clause modifiers when not in declare simd. Only check for integer type if ref modifier is not used. Remove diagnostics for required VALUE attribute. Diagnose VALUE attribute with ref or uval modifiers. Allow non-constant linear-step, if it is a dummy argument alone and is mentioned in uniform clause. Diagnose map kinds not allowed for various constructs. Diagnose target {enter ,exit ,}data without any map clauses. Add dummy OMP_LIST_IS_DEVICE_PTR and OMP_LIST_USE_DEVICE_PTR cases. (gfc_resolve_omp_do_blocks): Set omp_current_do_collapse to orderedc if non-zero. (gfc_resolve_omp_parallel_blocks): Handle new OpenMP 4.5 constructs, replace underscores with spaces in a few construct names. (resolve_omp_do): Set collapse to orderedc if non-zero. Handle new OpenMP 4.5 constructs. (resolve_oacc_loop_blocks): Call resolve_positive_int_expr instead of resolve_oacc_positive_int_expr. (gfc_resolve_omp_directive): Handle new OpenMP 4.5 constructs. (gfc_resolve_omp_declare_simd): Allow ods->proc_name to be NULL. * trans-openmp.c (gfc_omp_scalar_p): New function. (doacross_steps): New variable. (gfc_trans_omp_clauses): Handle new OpenMP 4.5 clauses and new clause modifiers. (gfc_trans_omp_critical): Adjust EXEC_OMP_CRITICAL handling. (gfc_trans_omp_do): Handle doacross loops. Clear sched_simd flag. Handle EXEC_OMP_TASKLOOP. (gfc_trans_omp_ordered): Translate omp clauses, allow NULL code->block. (GFC_OMP_SPLIT_TASKLOOP, GFC_OMP_MASK_TASKLOOP): New enum constants. (gfc_split_omp_clauses): Copy orderedc together with ordered. Change firstprivate and lastprivate handling for OpenMP 4.5. Handle EXEC_OMP_TARGET_SIMD, EXEC_OMP_TARGET_PARALLEL{,_DO,_DO_SIMD} and EXEC_OMP_TASKLOOP{,_SIMD}. Add handling for new OpenMP 4.5 clauses and clause modifiers and handle if clause without/with modifiers. (gfc_trans_omp_teams): Add omp_clauses argument, add it to other teams clauses. Don't wrap into OMP_TEAMS if -fopenmp-simd. (gfc_trans_omp_target): For -fopenmp, translate num_teams and thread_limit clauses on combined target teams early and pass to gfc_trans_omp_teams. Set OMP_TARGET_COMBINED if needed. Handle EXEC_OMP_TARGET_PARALLEL{,_DO,_DO_SIMD} and EXEC_OMP_TARGET_SIMD. (gfc_trans_omp_taskloop, gfc_trans_omp_target_enter_data, gfc_trans_omp_target_exit_data): New functions. (gfc_trans_omp_directive): Handle EXEC_OMP_TARGET_{ENTER,EXIT}_DATA EXEC_OMP_TASKLOOP{,_SIMD}, EXEC_OMP_TARGET_PARALLEL{,_DO,_DO_SIMD} and EXEC_OMP_TARGET_SIMD. Adjust gfc_trans_omp_teams caller. * symbol.c (check_conflict): Handle omp_declare_target_link. (gfc_add_omp_declare_target_link): New function. (gfc_copy_attr): Copy omp_declare_target_link. * dump-parse-tree.c (show_omp_namelist): Handle OMP_DEPEND_SINK_FIRST depend_op. Print linear clause modifiers. (show_omp_clauses): Adjust for OpenMP 4.5 clause changes. (show_omp_node): Print clauses for EXEC_OMP_ORDERED. Allow NULL c->block for EXEC_OMP_ORDERED. Formatting fixes. Adjust handling of EXEC_OMP_CRITICAL, handle new OpenMP 4.5 constructs and some forgotten OpenMP 4.0 constructs. (show_code_node): Handle new OpenMP 4.5 constructs and some forgotten OpenMP 4.0 constructs. * gfortran.h (symbol_attribute): Add omp_declare_target_link bitfield. (struct gfc_omp_namelist): Add u.common and u.linear_op fields. (struct gfc_common_head): Change omp_declare_target into bitfield. Add omp_declare_target_link bitfield. (gfc_add_omp_declare_target_link): New prototype. (enum gfc_statement): Add ST_OMP_TARGET_PARALLEL, ST_OMP_END_TARGET_PARALLEL, ST_OMP_TARGET_PARALLEL_DO, ST_OMP_END_TARGET_PARALLEL_DO, ST_OMP_TARGET_PARALLEL_DO_SIMD, ST_OMP_END_TARGET_PARALLEL_DO_SIMD, ST_OMP_TARGET_ENTER_DATA, ST_OMP_TARGET_EXIT_DATA, ST_OMP_TARGET_SIMD, ST_OMP_END_TARGET_SIMD, ST_OMP_TASKLOOP, ST_OMP_END_TASKLOOP, ST_OMP_TASKLOOP_SIMD, ST_OMP_END_TASKLOOP_SIMD and ST_OMP_ORDERED_DEPEND. (enum gfc_omp_depend_op): Add OMP_DEPEND_SINK_FIRST and OMP_DEPEND_SINK. (enum gfc_omp_linear_op): New. (struct gfc_omp_clauses): Add critical_name, depend_source, orderedc, defaultmap, nogroup, sched_simd, sched_monotonic, sched_nonmonotonic, simd, threads, grainsize, hint, num_tasks, priority and if_exprs fields. (enum gfc_exec_op): Add EXEC_OMP_END_CRITICAL, EXEC_OMP_TARGET_ENTER_DATA, EXEC_OMP_TARGET_EXIT_DATA, EXEC_OMP_TARGET_PARALLEL, EXEC_OMP_TARGET_PARALLEL_DO, EXEC_OMP_TARGET_PARALLEL_DO_SIMD, EXEC_OMP_TARGET_SIMD, EXEC_OMP_TASKLOOP, EXEC_OMP_TASKLOOP_SIMD. (enum gfc_omp_map_op): Add OMP_MAP_RELEASE, OMP_MAP_ALWAYS_TO, OMP_MAP_ALWAYS_FROM and OMP_MAP_ALWAYS_TOFROM. (OMP_LIST_IS_DEVICE_PTR, OMP_LIST_USE_DEVICE_PTR): New. (enum gfc_omp_if_kind): New. * module.c (enum ab_attribute): Add AB_OMP_DECLARE_TARGET_LINK. (attr_bits): Add AB_OMP_DECLARE_TARGET_LINK entry. (mio_symbol_attribute): Save and restore omp_declare_target_link bit. * trans.h (gfc_omp_scalar_p): New prototype. * frontend-passes.c (gfc_code_walker): Handle new OpenMP 4.5 expressions. * trans.c (trans_code): Handle new OpenMP 4.5 constructs. * resolve.c (gfc_resolve_blocks): Likewise. (gfc_resolve_code): Likewise. * f95-lang.c (LANG_HOOKS_OMP_SCALAR_P): Redefine to gfc_omp_scalar_p. (gfc_attribute_table): Add "omp declare target link". * st.c (gfc_free_statement): Handle EXEC_OMP_END_CRITICAL like EXEC_OMP_CRITICAL before, free clauses for EXEC_OMP_CRITICAL and new OpenMP 4.5 constructs. Free omp clauses even for EXEC_OMP_ORDERED. * match.c (match_exit_cycle): Rename collapse variable to count, set it to orderedc if non-zero, instead of collapse. * trans-decl.c (add_attributes_to_decl): Add "omp declare target link" instead of "omp declare target" for omp_declare_target_link. * trans-common.c (build_common_decl): Likewise. * match.h (gfc_match_omp_target_enter_data, gfc_match_omp_target_exit_data, gfc_match_omp_target_parallel, gfc_match_omp_target_parallel_do, gfc_match_omp_target_parallel_do_simd, gfc_match_omp_target_simd, gfc_match_omp_taskloop, gfc_match_omp_taskloop_simd, gfc_match_omp_end_critical, gfc_match_omp_ordered_depend): New prototypes. * parse.c (decode_omp_directive): Use gfc_match_omp_end_critical instead of gfc_match_omp_critical for !$omp end critical. Handle new OpenMP 4.5 constructs. If ordered directive has depend clause as the first of the clauses, use gfc_match_omp_ordered_depend and ST_OMP_ORDERED_DEPEND instead of gfc_match_omp_ordered and ST_OMP_ORDERED. (case_executable): Add ST_OMP_TARGET_ENTER_DATA, ST_OMP_TARGET_EXIT_DATA and ST_OMP_ORDERED_DEPEND cases. (case_exec_markers): Add ST_OMP_TARGET_PARALLEL, ST_OMP_TARGET_PARALLEL_DO, ST_OMP_TARGET_PARALLEL_DO_SIMD, ST_OMP_TARGET_SIMD, ST_OMP_TASKLOOP and ST_OMP_TASKLOOP_SIMD cases. (gfc_ascii_statement): Handle new OpenMP 4.5 constructs. (parse_omp_do): Handle ST_OMP_TARGET_PARALLEL_DO, ST_OMP_TARGET_PARALLEL_DO_SIMD, ST_OMP_TASKLOOP and ST_OMP_TASKLOOP_SIMD. (parse_omp_structured_block): Handle EXEC_OMP_END_CRITICAL instead of EXEC_OMP_CRITICAL, adjust for EXEC_OMP_CRITICAL having omp clauses now. (parse_executable): Handle ST_OMP_TARGET_PARALLEL, ST_OMP_TARGET_PARALLEL_DO, ST_OMP_TARGET_PARALLEL_DO_SIMD, ST_OMP_TASKLOOP and ST_OMP_TASKLOOP_SIMD. gcc/testsuite/ * gfortran.dg/gomp/pr77516.f90: Add dg-warning. * gfortran.dg/gomp/target1.f90: Remove ordered clause where it is no longer allowed and corresponding ordered construct. * gfortran.dg/gomp/linear-1.f90: New test. * gfortran.dg/gomp/declare-simd-2.f90: New test. * gfortran.dg/gomp/declare-target-1.f90: New test. * gfortran.dg/gomp/declare-target-2.f90: New test. libgomp/ * testsuite/libgomp.fortran/examples-4/declare_target-1.f90 (fib_wrapper): Add map(from: x) clause. * testsuite/libgomp.fortran/examples-4/declare_target-2.f90 (e_53_2): Likewise. * testsuite/libgomp.fortran/examples-4/declare_target-4.f90 (accum): Add map(tmp) clause. * testsuite/libgomp.fortran/examples-4/declare_target-5.f90 (accum): Add map(tofrom: tmp) clause. * testsuite/libgomp.fortran/examples-4/target_data-3.f90 (gramSchmidt): Likewise. * testsuite/libgomp.fortran/examples-4/teams-2.f90 (dotprod): Add map(tofrom: sum) clause. * testsuite/libgomp.fortran/nestedfn5.f90 (foo): Add twice map (alloc: a, l) clause. Add defaultmap(tofrom: scalar) clause. * testsuite/libgomp.fortran/pr66199-2.f90: Adjust for linear clause only allowed on the loop iterator. * testsuite/libgomp.fortran/target4.f90 (foo): Add map(t) clause. * testsuite/libgomp.fortran/taskloop2.f90: New test. * testsuite/libgomp.fortran/taskloop4.f90: New test. * testsuite/libgomp.fortran/doacross1.f90: New test. * testsuite/libgomp.fortran/doacross3.f90: New test. * testsuite/libgomp.fortran/taskloop1.f90: New test. * testsuite/libgomp.fortran/taskloop3.f90: New test. * testsuite/libgomp.fortran/doacross2.f90: New test. * testsuite/libgomp.c/doacross-1.c (main): Add missing #pragma omp atomic read. * testsuite/libgomp.c/doacross-2.c (main): Likewise. * testsuite/libgomp.c/doacross-3.c (main): Likewise. From-SVN: r242037
2016-11-10 12:38:05 +01:00
#pragma omp atomic read
omp-low.c (lower_omp_ordered): Add argument to GOMP_SMD_ORDERED_* internal calls - 0 if... gcc/ 2015-11-14 Jakub Jelinek <jakub@redhat.com> * omp-low.c (lower_omp_ordered): Add argument to GOMP_SMD_ORDERED_* internal calls - 0 if ordered simd and 1 for ordered threads simd. * tree-vectorizer.c (adjust_simduid_builtins): If GOMP_SIMD_ORDERED_* argument is 1, replace it with GOMP_ordered_* call instead of removing it. gcc/c/ 2015-11-14 Jakub Jelinek <jakub@redhat.com> * c-typeck.c (c_finish_omp_clauses): Don't mark GOMP_MAP_FIRSTPRIVATE_POINTER decls addressable. gcc/cp/ 2015-11-14 Jakub Jelinek <jakub@redhat.com> * semantics.c (finish_omp_clauses): Don't mark GOMP_MAP_FIRSTPRIVATE_POINTER decls addressable. libgomp/ 2015-11-14 Jakub Jelinek <jakub@redhat.com> Aldy Hernandez <aldyh@redhat.com> Ilya Verbin <ilya.verbin@intel.com> * ordered.c (gomp_doacross_init, GOMP_doacross_post, GOMP_doacross_wait, gomp_doacross_ull_init, GOMP_doacross_ull_post, GOMP_doacross_ull_wait): For GFS_GUIDED don't divide number of iterators or IV by chunk size. * parallel.c (gomp_resolve_num_threads): Don't assume that if thr->ts.team is non-NULL, then pool must be non-NULL. * libgomp-plugin.h (GOMP_PLUGIN_target_task_completion): Declare. * libgomp.map (GOMP_PLUGIN_1.1): New symbol version, export GOMP_PLUGIN_target_task_completion. * Makefile.am (libgomp_la_SOURCES): Add priority_queue.c. * Makefile.in: Regenerate. * libgomp.h: Shuffle prototypes and forward definitions around so priority queues can be defined. (enum gomp_task_kind): Add GOMP_TASK_ASYNC_RUNNING. (enum gomp_target_task_state): New enum. (struct gomp_target_task): Add state, tgt, task and team fields. (gomp_create_target_task): Change return type to bool, add state argument. (gomp_target_task_fn): Change return type to bool. (struct gomp_device_descr): Add async_run_func. (struct gomp_task): Remove children, next_child, prev_child, next_queue, prev_queue, next_taskgroup, prev_taskgroup. Add pnode field. (struct gomp_taskgroup): Remove children. Add taskgroup_queue. (struct gomp_team): Change task_queue type to a priority queue. (splay_compare): Define inline. (priority_queue_offset): New. (priority_node_to_task): New. (task_to_priority_node): New. * oacc-mem.c: Do not include splay-tree.h. * priority_queue.c: New file. * priority_queue.h: New file. * splay-tree.c: Do not include splay-tree.h. (splay_tree_foreach_internal): New. (splay_tree_foreach): New. * splay-tree.h: Become re-entrant if splay_tree_prefix is defined. (splay_tree_callback): Define typedef. * target.c (splay_compare): Move to libgomp.h. (GOMP_target): Don't adjust *thr in any way around running offloaded task. (GOMP_target_ext): Likewise. Handle target nowait. (GOMP_target_update_ext, GOMP_target_enter_exit_data): Check return value from gomp_create_target_task, if false, fallthrough as if no dependencies exist. (gomp_target_task_fn): Change return type to bool, return true if the task should have another part scheduled later. Handle target nowait. (gomp_load_plugin_for_device): Initialize async_run. * task.c (gomp_init_task): Initialize children_queue. (gomp_clear_parent_in_list): New. (gomp_clear_parent_in_tree): New. (gomp_clear_parent): Handle priorities. (GOMP_task): Likewise. (priority_queue_move_task_first, gomp_target_task_completion, GOMP_PLUGIN_target_task_completion): New functions. (gomp_create_target_task): Use priority queues. Change return type to bool, add state argument, return false if for async {{enter,exit} data,update} constructs no dependencies need to be waited for, handle target nowait. Set task->fn to NULL instead of gomp_target_task_fn. (verify_children_queue): Remove. (priority_list_upgrade_task): New. (priority_queue_upgrade_task): New. (verify_task_queue): Remove. (priority_list_downgrade_task): New. (priority_queue_downgrade_task): New. (gomp_task_run_pre): Use priority queues. Abstract code out to priority_queue_downgrade_task. (gomp_task_run_post_handle_dependers): Use priority queues. (gomp_task_run_post_remove_parent): Likewise. (gomp_task_run_post_remove_taskgroup): Likewise. (gomp_barrier_handle_tasks): Likewise. Handle target nowait target tasks specially. (GOMP_taskwait): Likewise. (gomp_task_maybe_wait_for_dependencies): Likewise. Abstract code to priority-queue_upgrade_task. (GOMP_taskgroup_start): Use priority queues. (GOMP_taskgroup_end): Likewise. Handle target nowait target tasks specially. If taskgroup is NULL, and thr->ts.level is 0, act as a barrier. * taskloop.c (GOMP_taskloop): Handle priorities. * team.c (gomp_new_team): Call priority_queue_init. (free_team): Call priority_queue_free. (gomp_free_thread): Call gomp_team_end if thr->ts.team is artificial team created for target nowait in implicit parallel region. (gomp_team_start): For nested check, test thr->ts.level instead of thr->ts.team != NULL. * testsuite/libgomp.c/doacross-3.c: New test. * testsuite/libgomp.c/ordered-5.c: New test. * testsuite/libgomp.c/priority.c: New test. * testsuite/libgomp.c/target-31.c: New test. * testsuite/libgomp.c/target-32.c: New test. * testsuite/libgomp.c/target-33.c: New test. * testsuite/libgomp.c/target-34.c: New test. liboffloadmic/ 2015-11-14 Ilya Verbin <ilya.verbin@intel.com> * runtime/offload_host.cpp (task_completion_callback): New variable. (offload_proxy_task_completed_ooo): Call task_completion_callback. (__offload_register_task_callback): New function. * runtime/offload_host.h (__offload_register_task_callback): New declaration. * plugin/libgomp-plugin-intelmic.cpp (offload): Add async_data argument, handle async offloading. (register_main_image): Call register_main_image. (GOMP_OFFLOAD_init_device, get_target_table, GOMP_OFFLOAD_alloc, GOMP_OFFLOAD_free, GOMP_OFFLOAD_host2dev, GOMP_OFFLOAD_dev2host, GOMP_OFFLOAD_dev2dev) Adjust offload callers. (GOMP_OFFLOAD_async_run): New function. (GOMP_OFFLOAD_run): Implement using GOMP_OFFLOAD_async_run. From-SVN: r230381
2015-11-14 19:42:13 +01:00
l = c[i - 2][j + 1][k - 4];
if (l < 2)
abort ();
}
if (i >= 3 && j >= 4 && k >= 2)
{
omp-low.c (lower_omp_target): Fix up argument to is_reference. gcc/ * omp-low.c (lower_omp_target): Fix up argument to is_reference. (expand_omp_ordered_sink): Handle TREE_PURPOSE of deps being TRUNC_DIV_EXPR. * gimplify.c (gimplify_scan_omp_clauses): Likewise. Set ctx->target_map_scalars_firstprivate on OMP_TARGET even for Fortran. Remove omp_no_lastprivate callers. Propagate lastprivate on combined teams distribute parallel for simd even to distribute and teams construct. For OMP_CLAUSE_DEPEND add missing break at the end of OMP_CLAUSE_DEPEND_SINK case. (omp_notice_variable): Use lang_hooks.decls.omp_scalar_p. (omp_no_lastprivate): Removed. (gimplify_adjust_omp_clauses): Remove omp_no_lastprivate callers. (gimplify_omp_for): Likewise. (computable_teams_clause): Fail for automatic vars from current function not yet seen in bind expr. * langhooks.c (lhd_omp_scalar_p): New function. * langhooks.h (struct lang_hooks_for_decls): Add omp_scalar_p. * varpool.c (varpool_node::get_create): Set node->offloading even for DECL_EXTERNAL decls. * langhooks-def.h (lhd_omp_scalar_p): New prototype. (LANG_HOOKS_OMP_SCALAR_P): Define. (LANG_HOOKS_DECLS): Use it. gcc/fortran/ * openmp.c (gfc_free_omp_clauses): Free critical_name, grainsize, hint, num_tasks, priority and if_exprs. (gfc_match_omp_to_link, gfc_match_omp_depend_sink): New functions. (enum omp_mask1, enum omp_mask2): New enums. Change all OMP_CLAUSE_* defines into enum values, and change their values from ((uint64_t) 1 << bit) to just bit. (omp_mask, omp_inv_mask): New classes. Add ctors and operators. (gfc_match_omp_clauses): Change mask argument from uint64_t to const omp_mask. Assert OMP_MASK1_LAST and OMP_MASK2_LAST are at most 64. Move delete clause handling to where it alphabetically belongs. Parse defaultmap, grainsize, hint, is_device_ptr, nogroup, nowait, num_tasks, priority, simd, threads and use_device_ptr clauses. Parse if clause modifier. Parse map clause always modifier, and release and delete kinds. Parse ordered clause with argument. Parse schedule clause modifiers. Differentiate device clause parsing based on openacc flag. Guard link clause parsing with openacc flag. Add support for parsing linear clause modifiers. Parse depend(source) and depend(sink: ...). Use gfc_match_omp_to_link for to and link clauses in declare target construct. (match_acc): Change mask type from uint64_t to const omp_mask. (OMP_SINGLE_CLAUSES, OMP_ORDERED_CLAUSES, OMP_DECLARE_TARGET_CLAUSES, OMP_TASKLOOP_CLAUSES, OMP_TARGET_ENTER_DATA_CLAUSES, OMP_TARGET_EXIT_DATA_CLAUSES): Define. (OACC_PARALLEL_CLAUSES, OACC_KERNELS_CLAUSES, OACC_DATA_CLAUSES, OACC_LOOP_CLAUSES, OACC_HOST_DATA_CLAUSES, OACC_DECLARE_CLAUSES, OACC_ENTER_DATA_CLAUSES, OACC_EXIT_DATA_CLAUSES, OACC_WAIT_CLAUSES, OACC_ROUTINE_CLAUSES, OMP_PARALLEL_CLAUSES, OMP_DECLARE_SIMD_CLAUSES, OMP_SECTIONS_CLAUSES, OMP_TEAMS_CLAUSES, OMP_DISTRIBUTE_CLAUSES): Replace first or only OMP_CLAUSE_* value in bitset with omp_mask (OMP_CLAUSE_*). (OMP_DO_CLAUSES): Likewise. Add OMP_CLAUSE_LINEAR. (OMP_SIMD_CLAUSES): Replace first or only OMP_CLAUSE_* value in bitset with omp_mask (OMP_CLAUSE_*). Add OMP_CLAUSE_SIMDLEN. (OACC_UPDATE_CLAUSES): Replace first or only OMP_CLAUSE_* value in bitset with omp_mask (OMP_CLAUSE_*). Replace OMP_CLAUSE_OACC_DEVICE with OMP_CLAUSE_DEVICE. (OMP_TASK_CLAUSES): Replace first or only OMP_CLAUSE_* value in bitset with omp_mask (OMP_CLAUSE_*). Add OMP_CLAUSE_PRIORITY. (OMP_TARGET_CLAUSES): Replace first or only OMP_CLAUSE_* value in bitset with omp_mask (OMP_CLAUSE_*). Add OMP_CLAUSE_DEPEND, OMP_CLAUSE_NOWAIT, OMP_CLAUSE_PRIVATE, OMP_CLAUSE_FIRSTPRIVATE, OMP_CLAUSE_DEFAULTMAP and OMP_CLAUSE_IS_DEVICE_PTR. (OMP_TARGET_DATA_CLAUSES): Replace first or only OMP_CLAUSE_* value in bitset with omp_mask (OMP_CLAUSE_*). Add OMP_CLAUSE_USE_DEVICE_PTR. (OMP_TARGET_UPDATE_CLAUSES): Replace first or only OMP_CLAUSE_* value in bitset with omp_mask (OMP_CLAUSE_*). Add OMP_CLAUSE_DEPEND and OMP_CLAUSE_NOWAIT. (match_omp): Change mask argument from unsigned int to const omp_mask. (gfc_match_omp_critical): Parse optional clauses and use omp_clauses union member instead of omp_name. (gfc_match_omp_end_critical): New function. (gfc_match_omp_distribute_parallel_do): Remove ordered and linear clauses from the mask. (gfc_match_omp_distribute_parallel_do_simd): Use & ~(omp_mask (OMP_CLAUSE_*)) instead of & ~OMP_CLAUSE_*. (gfc_match_omp_target_teams_distribute_parallel_do_simd): Likewise. (gfc_match_omp_teams_distribute_parallel_do_simd): Likewise. (gfc_match_omp_do_simd): Likewise. Don't remove ordered clause from the mask. (gfc_match_omp_parallel_do_simd): Likewise. (gfc_match_omp_target_teams_distribute_parallel_do): Likewise. (gfc_match_omp_teams_distribute_parallel_do): Likewise. (gfc_match_omp_declare_simd): If not using the form with (proc-name), require space before first clause. Make (proc-name) optional. If not present, set proc_name to NULL. (gfc_match_omp_declare_target): Rewritten for OpenMP 4.5. (gfc_match_omp_single): Use OMP_SINGLE_CLAUSES. (gfc_match_omp_task, gfc_match_omp_taskwait, gfc_match_omp_taskyield): Move around to where they belong alphabetically. (gfc_match_omp_target_enter_data, gfc_match_omp_target_exit_data, gfc_match_omp_target_parallel, gfc_match_omp_target_parallel_do, gfc_match_omp_target_parallel_do_simd, gfc_match_omp_target_simd, gfc_match_omp_taskloop, gfc_match_omp_taskloop_simd): New functions. (gfc_match_omp_ordered): Parse clauses. (gfc_match_omp_ordered_depend): New function. (gfc_match_omp_cancel, gfc_match_omp_end_single): Use omp_mask (OMP_CLAUSE_*) instead of OMP_CLAUSE_*. (resolve_oacc_scalar_int_expr): Renamed to ... (resolve_scalar_int_expr): ... this. Fix up formatting. (resolve_oacc_positive_int_expr): Renamed to ... (resolve_positive_int_expr): ... this. Fix up formatting. (resolve_nonnegative_int_expr): New function. (resolve_omp_clauses): Adjust callers, use the above functions even for OpenMP clauses, add handling of new OpenMP 4.5 clauses. Require orderedc >= collapse if specified. Handle depend(sink:) and depend(source) restrictions. Disallow linear clause when orderedc is non-zero. Diagnose linear clause modifiers when not in declare simd. Only check for integer type if ref modifier is not used. Remove diagnostics for required VALUE attribute. Diagnose VALUE attribute with ref or uval modifiers. Allow non-constant linear-step, if it is a dummy argument alone and is mentioned in uniform clause. Diagnose map kinds not allowed for various constructs. Diagnose target {enter ,exit ,}data without any map clauses. Add dummy OMP_LIST_IS_DEVICE_PTR and OMP_LIST_USE_DEVICE_PTR cases. (gfc_resolve_omp_do_blocks): Set omp_current_do_collapse to orderedc if non-zero. (gfc_resolve_omp_parallel_blocks): Handle new OpenMP 4.5 constructs, replace underscores with spaces in a few construct names. (resolve_omp_do): Set collapse to orderedc if non-zero. Handle new OpenMP 4.5 constructs. (resolve_oacc_loop_blocks): Call resolve_positive_int_expr instead of resolve_oacc_positive_int_expr. (gfc_resolve_omp_directive): Handle new OpenMP 4.5 constructs. (gfc_resolve_omp_declare_simd): Allow ods->proc_name to be NULL. * trans-openmp.c (gfc_omp_scalar_p): New function. (doacross_steps): New variable. (gfc_trans_omp_clauses): Handle new OpenMP 4.5 clauses and new clause modifiers. (gfc_trans_omp_critical): Adjust EXEC_OMP_CRITICAL handling. (gfc_trans_omp_do): Handle doacross loops. Clear sched_simd flag. Handle EXEC_OMP_TASKLOOP. (gfc_trans_omp_ordered): Translate omp clauses, allow NULL code->block. (GFC_OMP_SPLIT_TASKLOOP, GFC_OMP_MASK_TASKLOOP): New enum constants. (gfc_split_omp_clauses): Copy orderedc together with ordered. Change firstprivate and lastprivate handling for OpenMP 4.5. Handle EXEC_OMP_TARGET_SIMD, EXEC_OMP_TARGET_PARALLEL{,_DO,_DO_SIMD} and EXEC_OMP_TASKLOOP{,_SIMD}. Add handling for new OpenMP 4.5 clauses and clause modifiers and handle if clause without/with modifiers. (gfc_trans_omp_teams): Add omp_clauses argument, add it to other teams clauses. Don't wrap into OMP_TEAMS if -fopenmp-simd. (gfc_trans_omp_target): For -fopenmp, translate num_teams and thread_limit clauses on combined target teams early and pass to gfc_trans_omp_teams. Set OMP_TARGET_COMBINED if needed. Handle EXEC_OMP_TARGET_PARALLEL{,_DO,_DO_SIMD} and EXEC_OMP_TARGET_SIMD. (gfc_trans_omp_taskloop, gfc_trans_omp_target_enter_data, gfc_trans_omp_target_exit_data): New functions. (gfc_trans_omp_directive): Handle EXEC_OMP_TARGET_{ENTER,EXIT}_DATA EXEC_OMP_TASKLOOP{,_SIMD}, EXEC_OMP_TARGET_PARALLEL{,_DO,_DO_SIMD} and EXEC_OMP_TARGET_SIMD. Adjust gfc_trans_omp_teams caller. * symbol.c (check_conflict): Handle omp_declare_target_link. (gfc_add_omp_declare_target_link): New function. (gfc_copy_attr): Copy omp_declare_target_link. * dump-parse-tree.c (show_omp_namelist): Handle OMP_DEPEND_SINK_FIRST depend_op. Print linear clause modifiers. (show_omp_clauses): Adjust for OpenMP 4.5 clause changes. (show_omp_node): Print clauses for EXEC_OMP_ORDERED. Allow NULL c->block for EXEC_OMP_ORDERED. Formatting fixes. Adjust handling of EXEC_OMP_CRITICAL, handle new OpenMP 4.5 constructs and some forgotten OpenMP 4.0 constructs. (show_code_node): Handle new OpenMP 4.5 constructs and some forgotten OpenMP 4.0 constructs. * gfortran.h (symbol_attribute): Add omp_declare_target_link bitfield. (struct gfc_omp_namelist): Add u.common and u.linear_op fields. (struct gfc_common_head): Change omp_declare_target into bitfield. Add omp_declare_target_link bitfield. (gfc_add_omp_declare_target_link): New prototype. (enum gfc_statement): Add ST_OMP_TARGET_PARALLEL, ST_OMP_END_TARGET_PARALLEL, ST_OMP_TARGET_PARALLEL_DO, ST_OMP_END_TARGET_PARALLEL_DO, ST_OMP_TARGET_PARALLEL_DO_SIMD, ST_OMP_END_TARGET_PARALLEL_DO_SIMD, ST_OMP_TARGET_ENTER_DATA, ST_OMP_TARGET_EXIT_DATA, ST_OMP_TARGET_SIMD, ST_OMP_END_TARGET_SIMD, ST_OMP_TASKLOOP, ST_OMP_END_TASKLOOP, ST_OMP_TASKLOOP_SIMD, ST_OMP_END_TASKLOOP_SIMD and ST_OMP_ORDERED_DEPEND. (enum gfc_omp_depend_op): Add OMP_DEPEND_SINK_FIRST and OMP_DEPEND_SINK. (enum gfc_omp_linear_op): New. (struct gfc_omp_clauses): Add critical_name, depend_source, orderedc, defaultmap, nogroup, sched_simd, sched_monotonic, sched_nonmonotonic, simd, threads, grainsize, hint, num_tasks, priority and if_exprs fields. (enum gfc_exec_op): Add EXEC_OMP_END_CRITICAL, EXEC_OMP_TARGET_ENTER_DATA, EXEC_OMP_TARGET_EXIT_DATA, EXEC_OMP_TARGET_PARALLEL, EXEC_OMP_TARGET_PARALLEL_DO, EXEC_OMP_TARGET_PARALLEL_DO_SIMD, EXEC_OMP_TARGET_SIMD, EXEC_OMP_TASKLOOP, EXEC_OMP_TASKLOOP_SIMD. (enum gfc_omp_map_op): Add OMP_MAP_RELEASE, OMP_MAP_ALWAYS_TO, OMP_MAP_ALWAYS_FROM and OMP_MAP_ALWAYS_TOFROM. (OMP_LIST_IS_DEVICE_PTR, OMP_LIST_USE_DEVICE_PTR): New. (enum gfc_omp_if_kind): New. * module.c (enum ab_attribute): Add AB_OMP_DECLARE_TARGET_LINK. (attr_bits): Add AB_OMP_DECLARE_TARGET_LINK entry. (mio_symbol_attribute): Save and restore omp_declare_target_link bit. * trans.h (gfc_omp_scalar_p): New prototype. * frontend-passes.c (gfc_code_walker): Handle new OpenMP 4.5 expressions. * trans.c (trans_code): Handle new OpenMP 4.5 constructs. * resolve.c (gfc_resolve_blocks): Likewise. (gfc_resolve_code): Likewise. * f95-lang.c (LANG_HOOKS_OMP_SCALAR_P): Redefine to gfc_omp_scalar_p. (gfc_attribute_table): Add "omp declare target link". * st.c (gfc_free_statement): Handle EXEC_OMP_END_CRITICAL like EXEC_OMP_CRITICAL before, free clauses for EXEC_OMP_CRITICAL and new OpenMP 4.5 constructs. Free omp clauses even for EXEC_OMP_ORDERED. * match.c (match_exit_cycle): Rename collapse variable to count, set it to orderedc if non-zero, instead of collapse. * trans-decl.c (add_attributes_to_decl): Add "omp declare target link" instead of "omp declare target" for omp_declare_target_link. * trans-common.c (build_common_decl): Likewise. * match.h (gfc_match_omp_target_enter_data, gfc_match_omp_target_exit_data, gfc_match_omp_target_parallel, gfc_match_omp_target_parallel_do, gfc_match_omp_target_parallel_do_simd, gfc_match_omp_target_simd, gfc_match_omp_taskloop, gfc_match_omp_taskloop_simd, gfc_match_omp_end_critical, gfc_match_omp_ordered_depend): New prototypes. * parse.c (decode_omp_directive): Use gfc_match_omp_end_critical instead of gfc_match_omp_critical for !$omp end critical. Handle new OpenMP 4.5 constructs. If ordered directive has depend clause as the first of the clauses, use gfc_match_omp_ordered_depend and ST_OMP_ORDERED_DEPEND instead of gfc_match_omp_ordered and ST_OMP_ORDERED. (case_executable): Add ST_OMP_TARGET_ENTER_DATA, ST_OMP_TARGET_EXIT_DATA and ST_OMP_ORDERED_DEPEND cases. (case_exec_markers): Add ST_OMP_TARGET_PARALLEL, ST_OMP_TARGET_PARALLEL_DO, ST_OMP_TARGET_PARALLEL_DO_SIMD, ST_OMP_TARGET_SIMD, ST_OMP_TASKLOOP and ST_OMP_TASKLOOP_SIMD cases. (gfc_ascii_statement): Handle new OpenMP 4.5 constructs. (parse_omp_do): Handle ST_OMP_TARGET_PARALLEL_DO, ST_OMP_TARGET_PARALLEL_DO_SIMD, ST_OMP_TASKLOOP and ST_OMP_TASKLOOP_SIMD. (parse_omp_structured_block): Handle EXEC_OMP_END_CRITICAL instead of EXEC_OMP_CRITICAL, adjust for EXEC_OMP_CRITICAL having omp clauses now. (parse_executable): Handle ST_OMP_TARGET_PARALLEL, ST_OMP_TARGET_PARALLEL_DO, ST_OMP_TARGET_PARALLEL_DO_SIMD, ST_OMP_TASKLOOP and ST_OMP_TASKLOOP_SIMD. gcc/testsuite/ * gfortran.dg/gomp/pr77516.f90: Add dg-warning. * gfortran.dg/gomp/target1.f90: Remove ordered clause where it is no longer allowed and corresponding ordered construct. * gfortran.dg/gomp/linear-1.f90: New test. * gfortran.dg/gomp/declare-simd-2.f90: New test. * gfortran.dg/gomp/declare-target-1.f90: New test. * gfortran.dg/gomp/declare-target-2.f90: New test. libgomp/ * testsuite/libgomp.fortran/examples-4/declare_target-1.f90 (fib_wrapper): Add map(from: x) clause. * testsuite/libgomp.fortran/examples-4/declare_target-2.f90 (e_53_2): Likewise. * testsuite/libgomp.fortran/examples-4/declare_target-4.f90 (accum): Add map(tmp) clause. * testsuite/libgomp.fortran/examples-4/declare_target-5.f90 (accum): Add map(tofrom: tmp) clause. * testsuite/libgomp.fortran/examples-4/target_data-3.f90 (gramSchmidt): Likewise. * testsuite/libgomp.fortran/examples-4/teams-2.f90 (dotprod): Add map(tofrom: sum) clause. * testsuite/libgomp.fortran/nestedfn5.f90 (foo): Add twice map (alloc: a, l) clause. Add defaultmap(tofrom: scalar) clause. * testsuite/libgomp.fortran/pr66199-2.f90: Adjust for linear clause only allowed on the loop iterator. * testsuite/libgomp.fortran/target4.f90 (foo): Add map(t) clause. * testsuite/libgomp.fortran/taskloop2.f90: New test. * testsuite/libgomp.fortran/taskloop4.f90: New test. * testsuite/libgomp.fortran/doacross1.f90: New test. * testsuite/libgomp.fortran/doacross3.f90: New test. * testsuite/libgomp.fortran/taskloop1.f90: New test. * testsuite/libgomp.fortran/taskloop3.f90: New test. * testsuite/libgomp.fortran/doacross2.f90: New test. * testsuite/libgomp.c/doacross-1.c (main): Add missing #pragma omp atomic read. * testsuite/libgomp.c/doacross-2.c (main): Likewise. * testsuite/libgomp.c/doacross-3.c (main): Likewise. From-SVN: r242037
2016-11-10 12:38:05 +01:00
#pragma omp atomic read
omp-low.c (lower_omp_ordered): Add argument to GOMP_SMD_ORDERED_* internal calls - 0 if... gcc/ 2015-11-14 Jakub Jelinek <jakub@redhat.com> * omp-low.c (lower_omp_ordered): Add argument to GOMP_SMD_ORDERED_* internal calls - 0 if ordered simd and 1 for ordered threads simd. * tree-vectorizer.c (adjust_simduid_builtins): If GOMP_SIMD_ORDERED_* argument is 1, replace it with GOMP_ordered_* call instead of removing it. gcc/c/ 2015-11-14 Jakub Jelinek <jakub@redhat.com> * c-typeck.c (c_finish_omp_clauses): Don't mark GOMP_MAP_FIRSTPRIVATE_POINTER decls addressable. gcc/cp/ 2015-11-14 Jakub Jelinek <jakub@redhat.com> * semantics.c (finish_omp_clauses): Don't mark GOMP_MAP_FIRSTPRIVATE_POINTER decls addressable. libgomp/ 2015-11-14 Jakub Jelinek <jakub@redhat.com> Aldy Hernandez <aldyh@redhat.com> Ilya Verbin <ilya.verbin@intel.com> * ordered.c (gomp_doacross_init, GOMP_doacross_post, GOMP_doacross_wait, gomp_doacross_ull_init, GOMP_doacross_ull_post, GOMP_doacross_ull_wait): For GFS_GUIDED don't divide number of iterators or IV by chunk size. * parallel.c (gomp_resolve_num_threads): Don't assume that if thr->ts.team is non-NULL, then pool must be non-NULL. * libgomp-plugin.h (GOMP_PLUGIN_target_task_completion): Declare. * libgomp.map (GOMP_PLUGIN_1.1): New symbol version, export GOMP_PLUGIN_target_task_completion. * Makefile.am (libgomp_la_SOURCES): Add priority_queue.c. * Makefile.in: Regenerate. * libgomp.h: Shuffle prototypes and forward definitions around so priority queues can be defined. (enum gomp_task_kind): Add GOMP_TASK_ASYNC_RUNNING. (enum gomp_target_task_state): New enum. (struct gomp_target_task): Add state, tgt, task and team fields. (gomp_create_target_task): Change return type to bool, add state argument. (gomp_target_task_fn): Change return type to bool. (struct gomp_device_descr): Add async_run_func. (struct gomp_task): Remove children, next_child, prev_child, next_queue, prev_queue, next_taskgroup, prev_taskgroup. Add pnode field. (struct gomp_taskgroup): Remove children. Add taskgroup_queue. (struct gomp_team): Change task_queue type to a priority queue. (splay_compare): Define inline. (priority_queue_offset): New. (priority_node_to_task): New. (task_to_priority_node): New. * oacc-mem.c: Do not include splay-tree.h. * priority_queue.c: New file. * priority_queue.h: New file. * splay-tree.c: Do not include splay-tree.h. (splay_tree_foreach_internal): New. (splay_tree_foreach): New. * splay-tree.h: Become re-entrant if splay_tree_prefix is defined. (splay_tree_callback): Define typedef. * target.c (splay_compare): Move to libgomp.h. (GOMP_target): Don't adjust *thr in any way around running offloaded task. (GOMP_target_ext): Likewise. Handle target nowait. (GOMP_target_update_ext, GOMP_target_enter_exit_data): Check return value from gomp_create_target_task, if false, fallthrough as if no dependencies exist. (gomp_target_task_fn): Change return type to bool, return true if the task should have another part scheduled later. Handle target nowait. (gomp_load_plugin_for_device): Initialize async_run. * task.c (gomp_init_task): Initialize children_queue. (gomp_clear_parent_in_list): New. (gomp_clear_parent_in_tree): New. (gomp_clear_parent): Handle priorities. (GOMP_task): Likewise. (priority_queue_move_task_first, gomp_target_task_completion, GOMP_PLUGIN_target_task_completion): New functions. (gomp_create_target_task): Use priority queues. Change return type to bool, add state argument, return false if for async {{enter,exit} data,update} constructs no dependencies need to be waited for, handle target nowait. Set task->fn to NULL instead of gomp_target_task_fn. (verify_children_queue): Remove. (priority_list_upgrade_task): New. (priority_queue_upgrade_task): New. (verify_task_queue): Remove. (priority_list_downgrade_task): New. (priority_queue_downgrade_task): New. (gomp_task_run_pre): Use priority queues. Abstract code out to priority_queue_downgrade_task. (gomp_task_run_post_handle_dependers): Use priority queues. (gomp_task_run_post_remove_parent): Likewise. (gomp_task_run_post_remove_taskgroup): Likewise. (gomp_barrier_handle_tasks): Likewise. Handle target nowait target tasks specially. (GOMP_taskwait): Likewise. (gomp_task_maybe_wait_for_dependencies): Likewise. Abstract code to priority-queue_upgrade_task. (GOMP_taskgroup_start): Use priority queues. (GOMP_taskgroup_end): Likewise. Handle target nowait target tasks specially. If taskgroup is NULL, and thr->ts.level is 0, act as a barrier. * taskloop.c (GOMP_taskloop): Handle priorities. * team.c (gomp_new_team): Call priority_queue_init. (free_team): Call priority_queue_free. (gomp_free_thread): Call gomp_team_end if thr->ts.team is artificial team created for target nowait in implicit parallel region. (gomp_team_start): For nested check, test thr->ts.level instead of thr->ts.team != NULL. * testsuite/libgomp.c/doacross-3.c: New test. * testsuite/libgomp.c/ordered-5.c: New test. * testsuite/libgomp.c/priority.c: New test. * testsuite/libgomp.c/target-31.c: New test. * testsuite/libgomp.c/target-32.c: New test. * testsuite/libgomp.c/target-33.c: New test. * testsuite/libgomp.c/target-34.c: New test. liboffloadmic/ 2015-11-14 Ilya Verbin <ilya.verbin@intel.com> * runtime/offload_host.cpp (task_completion_callback): New variable. (offload_proxy_task_completed_ooo): Call task_completion_callback. (__offload_register_task_callback): New function. * runtime/offload_host.h (__offload_register_task_callback): New declaration. * plugin/libgomp-plugin-intelmic.cpp (offload): Add async_data argument, handle async offloading. (register_main_image): Call register_main_image. (GOMP_OFFLOAD_init_device, get_target_table, GOMP_OFFLOAD_alloc, GOMP_OFFLOAD_free, GOMP_OFFLOAD_host2dev, GOMP_OFFLOAD_dev2host, GOMP_OFFLOAD_dev2dev) Adjust offload callers. (GOMP_OFFLOAD_async_run): New function. (GOMP_OFFLOAD_run): Implement using GOMP_OFFLOAD_async_run. From-SVN: r230381
2015-11-14 19:42:13 +01:00
l = c[i - 1][j - 2][k - 2];
if (l < 2)
abort ();
}
#pragma omp ordered depend (source)
#pragma omp atomic write
c[i][j][k] = 3;
}
#pragma omp for schedule(guided, 5) ordered (3) nowait
for (j = 0; j < N / 16 - 1; j++)
for (k = 0; k < 8; k += 2)
for (i = 3; i <= 5 + f; i++)
{
#pragma omp atomic write
g[j][k][i] = 1;
#pragma omp ordered depend(sink: j, k - 2, i - 1) \
depend(sink: j - 2, k - 2, i + 1)
#pragma omp ordered depend(sink: j - 3, k + 2, i - 2)
if (k >= 2 && i > 3)
{
#pragma omp atomic read
l = g[j][k - 2][i - 1];
if (l < 2)
abort ();
}
#pragma omp atomic write
g[j][k][i] = 2;
if (j >= 2 && k >= 2 && i < 5)
{
#pragma omp atomic read
l = g[j - 2][k - 2][i + 1];
if (l < 2)
abort ();
}
if (j >= 3 && k < N / 16 - 3 && i == 5)
{
#pragma omp atomic read
l = g[j - 3][k + 2][i - 2];
if (l < 2)
abort ();
}
#pragma omp ordered depend(source)
#pragma omp atomic write
g[j][k][i] = 3;
}
#pragma omp for collapse(2) ordered(4) lastprivate (i, j, k)
for (i = 2; i < f + 3; i++)
for (j = d + 1; j >= 0; j--)
for (k = 0; k < d; k++)
for (l = 0; l < d + 2; l++)
{
#pragma omp ordered depend (source)
#pragma omp ordered depend (sink:i - 2, j + 2, k - 2, l)
if (!e)
abort ();
}
#pragma omp single
{
if (i != 3 || j != -1 || k != 0)
abort ();
i = 8; j = 9; k = 10;
}
#pragma omp for collapse(2) ordered(4) lastprivate (i, j, k, m)
for (i = 2; i < f + 3; i++)
for (j = d + 1; j >= 0; j--)
for (k = 0; k < d + 2; k++)
for (m = 0; m < d; m++)
{
#pragma omp ordered depend (source)
#pragma omp ordered depend (sink:i - 2, j + 2, k - 2, m)
abort ();
}
#pragma omp single
if (i != 3 || j != -1 || k != 2 || m != 0)
abort ();
#pragma omp for collapse(2) ordered(4) nowait
for (i = 2; i < f + 3; i++)
for (j = d; j > 0; j--)
for (k = 0; k < d + 2; k++)
for (l = 0; l < d + 4; l++)
{
#pragma omp ordered depend (source)
#pragma omp ordered depend (sink:i - 2, j + 2, k - 2, l)
if (!e)
abort ();
}
#pragma omp for nowait
for (i = 0; i < N; i++)
if (a[i] != 3 * (i >= 1))
abort ();
#pragma omp for collapse(2) private(k) nowait
for (i = 0; i < N / 16; i++)
for (j = 0; j < 8; j++)
for (k = 0; k < 4; k++)
if (b[i][j][k] != 3 * (i >= 3 && i < N / 16 - 1 && (j & 1) == 0 && k >= 1))
abort ();
#pragma omp for collapse(3) nowait
for (i = 0; i < N / 32; i++)
for (j = 0; j < 8; j++)
for (k = 0; k < 8; k++)
if (c[i][j][k] != 3 * (i >= 2 && j >= 2 && (k & 1) == 0))
abort ();
#pragma omp for collapse(2) private(k) nowait
for (i = 0; i < N / 16; i++)
for (j = 0; j < 8; j++)
for (k = 0; k < 6; k++)
if (g[i][j][k] != 3 * (i < N / 16 - 1 && (j & 1) == 0 && k >= 3))
abort ();
}
return 0;
}