/Linux-v5.10/tools/lib/perf/ |
D | cpumap.c | 15 struct perf_cpu_map *cpus = malloc(sizeof(*cpus) + sizeof(int)); in perf_cpu_map__dummy_new() local 17 if (cpus != NULL) { in perf_cpu_map__dummy_new() 18 cpus->nr = 1; in perf_cpu_map__dummy_new() 19 cpus->map[0] = -1; in perf_cpu_map__dummy_new() 20 refcount_set(&cpus->refcnt, 1); in perf_cpu_map__dummy_new() 23 return cpus; in perf_cpu_map__dummy_new() 50 struct perf_cpu_map *cpus; in cpu_map__default_new() local 57 cpus = malloc(sizeof(*cpus) + nr_cpus * sizeof(int)); in cpu_map__default_new() 58 if (cpus != NULL) { in cpu_map__default_new() 62 cpus->map[i] = i; in cpu_map__default_new() [all …]
|
D | evlist.c | 42 * We already have cpus for evsel (via PMU sysfs) so in __perf_evlist__propagate_maps() 46 perf_cpu_map__put(evsel->cpus); in __perf_evlist__propagate_maps() 47 evsel->cpus = perf_cpu_map__get(evlist->cpus); in __perf_evlist__propagate_maps() 48 } else if (!evsel->system_wide && perf_cpu_map__empty(evlist->cpus)) { in __perf_evlist__propagate_maps() 49 perf_cpu_map__put(evsel->cpus); in __perf_evlist__propagate_maps() 50 evsel->cpus = perf_cpu_map__get(evlist->cpus); in __perf_evlist__propagate_maps() 51 } else if (evsel->cpus != evsel->own_cpus) { in __perf_evlist__propagate_maps() 52 perf_cpu_map__put(evsel->cpus); in __perf_evlist__propagate_maps() 53 evsel->cpus = perf_cpu_map__get(evsel->own_cpus); in __perf_evlist__propagate_maps() 58 evlist->all_cpus = perf_cpu_map__merge(evlist->all_cpus, evsel->cpus); in __perf_evlist__propagate_maps() [all …]
|
/Linux-v5.10/drivers/cpuidle/ |
D | coupled.c | 3 * coupled.c - helper functions to enter the same idle state on multiple cpus 24 * cpus cannot be independently powered down, either due to 31 * shared between the cpus (L2 cache, interrupt controller, and 33 * be tightly controlled on both cpus. 36 * WFI state until all cpus are ready to enter a coupled state, at 38 * cpus at approximately the same time. 40 * Once all cpus are ready to enter idle, they are woken by an smp 42 * cpus will find work to do, and choose not to enter idle. A 43 * final pass is needed to guarantee that all cpus will call the 46 * ready counter matches the number of online coupled cpus. If any [all …]
|
/Linux-v5.10/Documentation/admin-guide/cgroup-v1/ |
D | cpusets.rst | 31 2.2 Adding/removing cpus 43 Cpusets provide a mechanism for assigning a set of CPUs and Memory 57 include CPUs in its CPU affinity mask, and using the mbind(2) and 60 CPUs or Memory Nodes not in that cpuset. The scheduler will not 67 cpusets and which CPUs and Memory Nodes are assigned to each cpuset, 75 The management of large computer systems, with many processors (CPUs), 113 Cpusets provide a Linux kernel mechanism to constrain which CPUs and 117 CPUs a task may be scheduled (sched_setaffinity) and on which Memory 122 - Cpusets are sets of allowed CPUs and Memory Nodes, known to the 126 - Calls to sched_setaffinity are filtered to just those CPUs [all …]
|
/Linux-v5.10/Documentation/timers/ |
D | no_hz.rst | 19 2. Omit scheduling-clock ticks on idle CPUs (CONFIG_NO_HZ_IDLE=y or 23 3. Omit scheduling-clock ticks on CPUs that are either idle or that 65 Omit Scheduling-Clock Ticks For Idle CPUs 74 scheduling-clock interrupts to idle CPUs, which is critically important 82 idle CPUs. That said, dyntick-idle mode is not free: 104 Omit Scheduling-Clock Ticks For CPUs With Only One Runnable Task 109 Note that omitting scheduling-clock ticks for CPUs with only one runnable 110 task implies also omitting them for idle CPUs. 113 sending scheduling-clock interrupts to CPUs with a single runnable task, 114 and such CPUs are said to be "adaptive-ticks CPUs". This is important [all …]
|
/Linux-v5.10/tools/perf/tests/ |
D | openat-syscall-all-cpus.c | 26 struct perf_cpu_map *cpus; in test__openat_syscall_event_on_all_cpus() local 39 cpus = perf_cpu_map__new(NULL); in test__openat_syscall_event_on_all_cpus() 40 if (cpus == NULL) { in test__openat_syscall_event_on_all_cpus() 54 if (evsel__open(evsel, cpus, threads) < 0) { in test__openat_syscall_event_on_all_cpus() 61 for (cpu = 0; cpu < cpus->nr; ++cpu) { in test__openat_syscall_event_on_all_cpus() 66 * without CPU_ALLOC. 1024 cpus in 2010 still seems in test__openat_syscall_event_on_all_cpus() 69 if (cpus->map[cpu] >= CPU_SETSIZE) { in test__openat_syscall_event_on_all_cpus() 70 pr_debug("Ignoring CPU %d\n", cpus->map[cpu]); in test__openat_syscall_event_on_all_cpus() 74 CPU_SET(cpus->map[cpu], &cpu_set); in test__openat_syscall_event_on_all_cpus() 77 cpus->map[cpu], in test__openat_syscall_event_on_all_cpus() [all …]
|
D | cpumap.c | 54 struct cpu_map_entries *cpus; in process_event_cpus() local 62 cpus = (struct cpu_map_entries *)data->data; in process_event_cpus() 64 TEST_ASSERT_VAL("wrong nr", cpus->nr == 2); in process_event_cpus() 65 TEST_ASSERT_VAL("wrong cpu", cpus->cpu[0] == 1); in process_event_cpus() 66 TEST_ASSERT_VAL("wrong cpu", cpus->cpu[1] == 256); in process_event_cpus() 80 struct perf_cpu_map *cpus; in test__cpu_map_synthesize() local 83 cpus = perf_cpu_map__new("0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19"); in test__cpu_map_synthesize() 86 !perf_event__synthesize_cpu_map(NULL, cpus, process_event_mask, NULL)); in test__cpu_map_synthesize() 88 perf_cpu_map__put(cpus); in test__cpu_map_synthesize() 91 cpus = perf_cpu_map__new("1,256"); in test__cpu_map_synthesize() [all …]
|
/Linux-v5.10/include/linux/ |
D | stop_machine.h | 13 * function to be executed on a single or multiple cpus preempting all 14 * other processes and monopolizing those cpus until it finishes. 18 * cpus are online. 94 * stop_machine: freeze the machine on all CPUs and run this function 97 * @cpus: the cpus to run the @fn() on (NULL = any online cpu) 109 int stop_machine(cpu_stop_fn_t fn, void *data, const struct cpumask *cpus); 112 * stop_machine_cpuslocked: freeze the machine on all CPUs and run this function 115 * @cpus: the cpus to run the @fn() on (NULL = any online cpu) 120 int stop_machine_cpuslocked(cpu_stop_fn_t fn, void *data, const struct cpumask *cpus); 123 const struct cpumask *cpus); [all …]
|
D | energy_model.h | 32 * @cpus: Cpumask covering the CPUs of the domain. It's here 37 * In case of CPU device, a "performance domain" represents a group of CPUs 38 * whose performance is scaled together. All CPUs of a performance domain 40 * a 1-to-1 mapping with CPUFreq policies. In case of other devices the @cpus 46 unsigned long cpus[]; member 49 #define em_span_cpus(em) (to_cpumask((em)->cpus)) 68 * In case of CPUs, the power is the one of a single CPU in the domain, 86 * em_cpu_energy() - Estimates the energy consumed by the CPUs of a 89 * @max_util : highest utilization among CPUs of the domain 90 * @sum_util : sum of the utilization of all CPUs in the domain [all …]
|
/Linux-v5.10/tools/perf/arch/arm64/util/ |
D | header.c | 17 static int _get_cpuid(char *buf, size_t sz, struct perf_cpu_map *cpus) in _get_cpuid() argument 26 cpus = perf_cpu_map__get(cpus); in _get_cpuid() 28 for (cpu = 0; cpu < perf_cpu_map__nr(cpus); cpu++) { in _get_cpuid() 33 sysfs, cpus->map[cpu]); in _get_cpuid() 57 perf_cpu_map__put(cpus); in _get_cpuid() 67 struct perf_cpu_map *cpus = perf_cpu_map__new(NULL); in get_cpuid() local 70 if (!cpus) in get_cpuid() 73 ret = _get_cpuid(buf, sz, cpus); in get_cpuid() 75 perf_cpu_map__put(cpus); in get_cpuid() 85 if (!pmu || !pmu->cpus) in get_cpuid_str() [all …]
|
/Linux-v5.10/drivers/clk/sunxi/ |
D | clk-sun9i-cpus.c | 7 * Allwinner A80 CPUS clock driver 22 * sun9i_a80_cpus_clk_setup() - Setup function for a80 cpus composite clk 55 struct sun9i_a80_cpus_clk *cpus = to_sun9i_a80_cpus_clk(hw); in sun9i_a80_cpus_clk_recalc_rate() local 60 reg = readl(cpus->reg); in sun9i_a80_cpus_clk_recalc_rate() 155 struct sun9i_a80_cpus_clk *cpus = to_sun9i_a80_cpus_clk(hw); in sun9i_a80_cpus_clk_set_rate() local 162 reg = readl(cpus->reg); in sun9i_a80_cpus_clk_set_rate() 170 writel(reg, cpus->reg); in sun9i_a80_cpus_clk_set_rate() 188 struct sun9i_a80_cpus_clk *cpus; in sun9i_a80_cpus_setup() local 193 cpus = kzalloc(sizeof(*cpus), GFP_KERNEL); in sun9i_a80_cpus_setup() 194 if (!cpus) in sun9i_a80_cpus_setup() [all …]
|
/Linux-v5.10/tools/lib/perf/tests/ |
D | test-evlist.c | 30 struct perf_cpu_map *cpus; in test_stat_cpu() local 43 cpus = perf_cpu_map__new(NULL); in test_stat_cpu() 44 __T("failed to create cpus", cpus); in test_stat_cpu() 59 perf_evlist__set_maps(evlist, cpus, NULL); in test_stat_cpu() 65 cpus = perf_evsel__cpus(evsel); in test_stat_cpu() 67 perf_cpu_map__for_each_cpu(cpu, tmp, cpus) { in test_stat_cpu() 78 perf_cpu_map__put(cpus); in test_stat_cpu() 200 struct perf_cpu_map *cpus; in test_mmap_thread() local 245 cpus = perf_cpu_map__dummy_new(); in test_mmap_thread() 246 __T("failed to create cpus", cpus); in test_mmap_thread() [all …]
|
D | test-cpumap.c | 15 struct perf_cpu_map *cpus; in main() local 21 cpus = perf_cpu_map__dummy_new(); in main() 22 if (!cpus) in main() 25 perf_cpu_map__get(cpus); in main() 26 perf_cpu_map__put(cpus); in main() 27 perf_cpu_map__put(cpus); in main()
|
/Linux-v5.10/Documentation/admin-guide/ |
D | cputopology.rst | 41 internal kernel map of CPUs within the same core. 46 human-readable list of CPUs within the same core. 51 internal kernel map of the CPUs sharing the same physical_package_id. 56 human-readable list of CPUs sharing the same physical_package_id. 61 internal kernel map of CPUs within the same die. 65 human-readable list of CPUs within the same die. 137 offline: CPUs that are not online because they have been 139 of CPUs allowed by the kernel configuration (kernel_max 140 above). [~cpu_online_mask + cpus >= NR_CPUS] 142 online: CPUs that are online and being scheduled [cpu_online_mask] [all …]
|
D | kernel-per-CPU-kthreads.rst | 13 - Documentation/core-api/irq/irq-affinity.rst: Binding interrupts to sets of CPUs. 15 - Documentation/admin-guide/cgroup-v1: Using cgroups to bind tasks to sets of CPUs. 18 of CPUs. 21 call to bind tasks to sets of CPUs. 50 2. Do all eHCA-Infiniband-related work on other CPUs, including 53 provisioned only on selected CPUs. 101 with multiple CPUs, force them all offline before bringing the 102 first one back online. Once you have onlined the CPUs in question, 103 do not offline any other CPUs, because doing so could force the 104 timer back onto one of the CPUs in question. [all …]
|
/Linux-v5.10/Documentation/scheduler/ |
D | sched-energy.rst | 9 the impact of its decisions on the energy consumed by CPUs. EAS relies on an 10 Energy Model (EM) of the CPUs to select an energy efficient CPU for each task, 59 In short, EAS changes the way CFS tasks are assigned to CPUs. When it is time 64 knowledge about the platform's topology, which include the 'capacity' of CPUs, 72 differentiate CPUs with different computing throughput. The 'capacity' of a CPU 76 tasks and CPUs computed by the Per-Entity Load Tracking (PELT) mechanism. Thanks 79 energy trade-offs. The capacity of CPUs is provided via arch-specific code 99 Let us consider a platform with 12 CPUs, split in 3 performance domains 102 CPUs: 0 1 2 3 4 5 6 7 8 9 10 11 108 containing 6 CPUs. The two root domains are denoted rd1 and rd2 in the [all …]
|
D | sched-domains.rst | 10 Each scheduling domain spans a number of CPUs (stored in the ->span field). 13 i. The top domain for each CPU will generally span all CPUs in the system 15 CPUs will never be given tasks to run unless the CPUs allowed mask is 17 CPUs". 23 to which the domain belongs. Groups may be shared among CPUs as they contain 27 shared between CPUs. 31 load of each of its member CPUs, and only when the load of a group becomes 49 If it succeeds, it looks for the busiest runqueue of all the CPUs' runqueues in 62 In SMP, the parent of the base domain will span all physical CPUs in the 80 CPUs using cpu_attach_domain.
|
/Linux-v5.10/drivers/cpufreq/ |
D | cpufreq-dt.c | 30 cpumask_var_t cpus; member 50 if (cpumask_test_cpu(cpu, priv->cpus)) in cpufreq_dt_find_data() 119 cpumask_copy(policy->cpus, priv->cpus); in cpufreq_init() 129 * Initialize OPP tables for all policy->cpus. They will be shared by in cpufreq_init() 130 * all CPUs which have marked their CPUs shared with OPP bindings. in cpufreq_init() 133 * before updating policy->cpus. Otherwise, we will end up creating in cpufreq_init() 134 * duplicate OPPs for policy->cpus. in cpufreq_init() 138 if (!dev_pm_opp_of_cpumask_add_table(policy->cpus)) in cpufreq_init() 180 dev_pm_opp_of_register_em(cpu_dev, policy->cpus); in cpufreq_init() 188 dev_pm_opp_of_cpumask_remove_table(policy->cpus); in cpufreq_init() [all …]
|
/Linux-v5.10/Documentation/devicetree/bindings/csky/ |
D | cpus.txt | 5 The device tree allows to describe the layout of CPUs in a system through 6 the "cpus" node, which in turn contains a number of subnodes (ie "cpu") 9 Only SMP system need to care about the cpus node and single processor 10 needn't define cpus node at all. 13 cpus and cpu node bindings definition 16 - cpus node 20 The node name must be "cpus". 22 A cpus node must define the following properties: 59 cpus {
|
/Linux-v5.10/Documentation/power/ |
D | suspend-and-cpuhotplug.rst | 27 |tasks | | cpus | | | | cpus | |tasks| 59 online CPUs 75 Note down these cpus in | P 100 | Call _cpu_up() [for all those cpus in the frozen_cpus mask, in a loop] 158 the non-boot CPUs are offlined or onlined, the _cpu_*() functions are called 177 update on the CPUs, as discussed below: 184 a. When all the CPUs are identical: 187 to apply the same microcode revision to each of the CPUs. 192 all CPUs, in order to handle case 'b' described below. 195 b. When some of the CPUs are different than the rest: [all …]
|
/Linux-v5.10/tools/perf/util/ |
D | cpumap.c | 21 static struct perf_cpu_map *cpu_map__from_entries(struct cpu_map_entries *cpus) in cpu_map__from_entries() argument 25 map = perf_cpu_map__empty_new(cpus->nr); in cpu_map__from_entries() 29 for (i = 0; i < cpus->nr; i++) { in cpu_map__from_entries() 35 if (cpus->cpu[i] == (u16) -1) in cpu_map__from_entries() 38 map->map[i] = (int) cpus->cpu[i]; in cpu_map__from_entries() 83 struct perf_cpu_map *cpus = malloc(sizeof(*cpus) + sizeof(int) * nr); in perf_cpu_map__empty_new() local 85 if (cpus != NULL) { in perf_cpu_map__empty_new() 88 cpus->nr = nr; in perf_cpu_map__empty_new() 90 cpus->map[i] = -1; in perf_cpu_map__empty_new() 92 refcount_set(&cpus->refcnt, 1); in perf_cpu_map__empty_new() [all …]
|
D | perf_api_probe.c | 63 struct perf_cpu_map *cpus; in perf_probe_api() local 66 cpus = perf_cpu_map__new(NULL); in perf_probe_api() 67 if (!cpus) in perf_probe_api() 69 cpu = cpus->map[0]; in perf_probe_api() 70 perf_cpu_map__put(cpus); in perf_probe_api() 128 struct perf_cpu_map *cpus; in perf_can_record_cpu_wide() local 131 cpus = perf_cpu_map__new(NULL); in perf_can_record_cpu_wide() 132 if (!cpus) in perf_can_record_cpu_wide() 134 cpu = cpus->map[0]; in perf_can_record_cpu_wide() 135 perf_cpu_map__put(cpus); in perf_can_record_cpu_wide()
|
/Linux-v5.10/tools/testing/selftests/cpu-hotplug/ |
D | cpu-on-off-test.sh | 44 echo -e "\t Cpus in online state: $online_cpus" 52 echo -e "\t Cpus in offline state: $offline_cpus" 56 # list all hot-pluggable CPUs 172 echo -e "\t run with -a option to offline all cpus" 206 echo "Full scope test: all hotplug cpus" 207 echo -e "\t online all offline cpus" 208 echo -e "\t offline all online cpus" 209 echo -e "\t online all offline cpus" 213 # Online all hot-pluggable CPUs 220 # Offline all hot-pluggable CPUs [all …]
|
/Linux-v5.10/tools/testing/selftests/rcutorture/bin/ |
D | jitter.sh | 4 # Alternate sleeping and spinning on randomly selected CPUs. The purpose 56 if cpus=`grep 1 /sys/devices/system/cpu/*/online 2>&1 | 61 cpus= 63 # Do not leave out non-hot-pluggable CPUs 64 cpus="$cpus $nohotplugcpus" 66 cpumask=`awk -v cpus="$cpus" -v me=$me -v n=$n 'BEGIN { 68 ncpus = split(cpus, ca);
|
/Linux-v5.10/Documentation/arm/ |
D | cluster-pm-race-avoidance.rst | 18 In a system containing multiple CPUs, it is desirable to have the 19 ability to turn off individual CPUs when the system is idle, reducing 22 In a system containing multiple clusters of CPUs, it is also desirable 27 of independently running CPUs, while the OS continues to run. This 92 CPUs in the cluster simultaneously modifying the state. The cluster- 104 referred to as a "CPU". CPUs are assumed to be single-threaded: 107 This means that CPUs fit the basic model closely. 216 A cluster is a group of connected CPUs with some common resources. 217 Because a cluster contains multiple CPUs, it can be doing multiple 272 which exact CPUs within the cluster play these roles. This must [all …]
|