Home
last modified time | relevance | path

Searched refs:workload (Results 1 – 25 of 102) sorted by relevance

12345

/linux/Documentation/admin-guide/
H A Dworkload-tracing.rst4 Discovering Linux kernel subsystems used by a workload
14 * Understanding system resources necessary to build and run a workload
17 in use by a workload. The completeness of the system usage information
18 depends on the completeness of coverage of a workload.
24 * Once we discover and understand the workload needs, we can focus on them
32 the system resources in use by a workload. Once we discover and understand
33 the workload needs, we can focus on them to avoid regressions and use it
37 the workload and doesn't include all the system calls that can be invoked
39 these system calls that are invoked. As an example, if a workload opens a
42 is a workload that provides full coverage of a workload then the method
[all …]
/linux/tools/perf/tests/shell/
H A Dstat_bpf_counters.sh7 workload="perf test -w sqrtloop"
44 base_instructions=$(perf stat --no-big-num -e instructions -- $workload 2>&1 | \
48 bpf_instructions=$(perf stat --no-big-num --bpf-counters -e instructions -- $workload 2>&1 | \
60 stat_output=$(perf stat --no-big-num -e instructions/name=base_instructions/,instructions/name=bpf_instructions/b -- $workload 2>&1)
H A Dtrace_summary.sh17 workload="true"
20 echo "testing: perf trace ${args} -- ${workload}"
21 perf trace ${args} -- ${workload} >${OUTPUT} 2>&1
H A Dstat_metrics_values.sh14 workload="perf bench futex hash -r 2 -s"
22 $PYTHON $pythonvalidator -rule $rulefile -output_dir $tmpdir -wl "${workload}" \
H A Dtest_intel_pt.sh24 workload="${temp_dir}/workload"
64 cat << _end_of_file_ | /usr/bin/cc -o "${workload}" -xc - -pthread && have_workload=true
255 $workload &
257 $workload &
/linux/tools/perf/tests/
H A Dperf-record.c112 * Prepare the workload in argv[] to run, it'll fork it, and then wait in test__PERF_RECORD()
119 pr_debug("Couldn't run the workload!\n"); in test__PERF_RECORD()
132 err = sched__get_first_possible_cpu(evlist->workload.pid, cpu_mask); in test__PERF_RECORD()
145 if (sched_setaffinity(evlist->workload.pid, cpu_mask_size, cpu_mask) < 0) { in test__PERF_RECORD()
179 * count just on workload.pid, which will start... in test__PERF_RECORD()
234 if ((pid_t)sample.pid != evlist->workload.pid) { in test__PERF_RECORD()
236 name, evlist->workload.pid, sample.pid); in test__PERF_RECORD()
240 if ((pid_t)sample.tid != evlist->workload.pid) { in test__PERF_RECORD()
242 name, evlist->workload.pid, sample.tid); in test__PERF_RECORD()
251 (pid_t)event->comm.pid != evlist->workload in test__PERF_RECORD()
[all...]
H A Dbuiltin-test.c163 #define workloads__for_each(workload) \ argument
164 for (unsigned i = 0; i < ARRAY_SIZE(workloads) && ({ workload = workloads[i]; 1; }); i++)
788 const char *workload = NULL; in cmd_test() local
800 …OPT_STRING('w', "workload", &workload, "work", "workload to run for testing, use '--list-workloads… in cmd_test()
828 if (workload) in cmd_test()
829 return run_workload(workload, argc, argv); in cmd_test()
/linux/tools/perf/tests/shell/lib/
H A Dperf_metric_validation.py22 \tis {1} in workload(s): {2} \n\
31 \tis {1} in workload(s): {2}\n\
39 workload='true', metrics='', cputype='cpu'): argument
49 self.workloads = [x for x in workload.split(",") if x]
380 def _run_perf(self, metric, workload: str):
383 wl = workload.split()
392 def collect_perf(self, workload: str):
416 wl = workload
425 workload = self.workloads[self.wlidx]
427 data = self._run_perf(metric, workload)
[all …]
/linux/Documentation/admin-guide/mm/
H A Didle_page_tracking.rst9 accessed by a workload and which are idle. This information can be useful for
10 estimating the workload's working set size, which, in turn, can be taken into
11 account when configuring the workload parameters, setting memory cgroup limits,
12 or deciding where to place the workload within a compute cluster.
51 workload one should:
53 1. Mark all the workload's pages as idle by setting corresponding bits in
55 ``/proc/pid/pagemap`` if the workload is represented by a process, or by
56 filtering out alien pages using ``/proc/kpagecgroup`` in case the workload
59 2. Wait until the workload accesses its working set.
/linux/rust/pin-init/examples/
H A Dstatic_init.rs94 let workload = 1_000; in main()
101 for _ in 0..workload { in main()
110 for _ in 0..workload { in main()
123 assert_eq!(*mtx.lock(), workload * thread_count * 2); in main()
93 let workload = 1_000; main() localVariable
H A Dpthread_mutex.rs159 let workload = 1_000_000; in main()
166 for _ in 0..workload { in main()
171 for _ in 0..workload { in main()
183 assert_eq!(*mtx.lock(), workload * thread_count * 2); in main()
158 let workload = 1_000_000; main() localVariable
H A Dmutex.rs200 let workload = if cfg!(miri) { 100 } else { 1_000 }; in main()
207 for _ in 0..workload { in main()
212 for _ in 0..workload { in main()
224 assert_eq!(*mtx.lock(), workload * thread_count * 2); in main()
199 let workload = if cfg!(miri) { 100 } else { 1_000 }; main() localVariable
/linux/drivers/gpu/drm/i915/gvt/
H A Dcmd_parser.h50 int intel_gvt_scan_and_shadow_ringbuffer(struct intel_vgpu_workload *workload);
56 int intel_gvt_scan_engine_context(struct intel_vgpu_workload *workload);
H A Dscheduler.h137 void intel_vgpu_queue_workload(struct intel_vgpu_workload *workload);
164 void intel_vgpu_destroy_workload(struct intel_vgpu_workload *workload);
/linux/tools/perf/bench/
H A Dfind-bit-bench.c34 static noinline void workload(int val) in workload() function
82 workload(bit); in do_for_each_set_bit()
97 workload(bit); in do_for_each_set_bit()
/linux/Documentation/filesystems/nfs/
H A Dknfsd-stats.rst54 Depending on the NFS workload patterns and various network stack
58 However this is a more accurate and less workload-dependent measure
74 pool for the NFS workload (the workload is thread-limited), in which
76 performance of the NFS workload.
93 threads configured than can be used by the NFS workload. This is
99 slow; the idle timeout is 60 minutes. Unless the NFS workload
/linux/drivers/gpu/drm/amd/pm/powerplay/hwmgr/
H A Dpp_psm.c275 long workload[1]; in psm_adjust_power_state_dynamic() local
300 workload[0] = hwmgr->workload_setting[index]; in psm_adjust_power_state_dynamic()
302 if (hwmgr->power_profile_mode != workload[0] && hwmgr->hwmgr_func->set_power_profile_mode) in psm_adjust_power_state_dynamic()
303 hwmgr->hwmgr_func->set_power_profile_mode(hwmgr, workload, 0); in psm_adjust_power_state_dynamic()
/linux/Documentation/gpu/
H A Ddrm-compute.rst11 have their memory swapped out from them. Or they simply want their workload
19 waiting for the workload to complete. Effectively this terminates the workload
22 Since this is undesirable, there need to be mitigations to prevent a workload
H A Ddrm-vm-bind-async.rst43 * ``long-running workload``: A workload that may take more than the
106 fences gated by a long-running workload and used as in-syncobjs for the
116 workload itself, which is allowed by dma-fence rules, but rather for
140 buffers. The workload execution can then easily be pipelined behind
142 for a GPU semaphore embedded by UMD in the workload.
/linux/Documentation/scheduler/
H A Dsched-capacity.rst73 With a workload that periodically does a fixed amount of work, you will get an
104 Executing the same workload as described in 1.3.1, which each CPU running at its
112 workload on CPU1
152 One issue that needs to be taken into account is that a workload's duty cycle is
154 periodic workload at a given frequency F::
163 Now, consider running the *same* workload at frequency F/2::
185 identical workload on CPUs of different capacity values will yield different
193 Executing a given periodic workload on each CPU at their maximum frequency would
384 workload on CPU0
391 workload on CPU1
[all …]
/linux/tools/perf/Documentation/
H A Dperf-sched.txt18 of an arbitrary workload.
21 and other scheduling properties of the workload.
43 'perf sched script' to see a detailed trace of the workload that
46 'perf sched replay' to simulate the workload that was recorded
48 that mimic the workload based on the events in the trace. These
50 of the workload as it occurred when it was recorded - and can repeat
54 workload captured via perf sched record. Columns stand for
90 scheduler behavior for the workload.
499 repeat the workload n times (0: infinite). Default is 10.
/linux/Documentation/translations/zh_CN/scheduler/
H A Dsched-capacity.rst108 workload on CPU1
339 workload on CPU0
346 workload on CPU1
/linux/tools/perf/util/
H A Devlist.c82 evlist->workload.pid = -1; in evlist__init()
1482 evlist->workload.cork_fd = -1; in evlist__prepare_workload()
1494 evlist->workload.pid = fork(); in evlist__prepare_workload()
1495 if (evlist->workload.pid < 0) { in evlist__prepare_workload()
1500 if (!evlist->workload.pid) { in evlist__prepare_workload()
1530 * writing exactly one byte, in workload.cork_fd, usually via in evlist__prepare_workload()
1533 * For cancelling the workload without actually running it, in evlist__prepare_workload()
1534 * the parent will just close workload.cork_fd, without writing in evlist__prepare_workload()
1571 perf_thread_map__set_pid(evlist->core.threads, 0, evlist->workload.pid); in evlist__prepare_workload()
1585 evlist->workload in evlist__prepare_workload()
[all...]
/linux/Documentation/timers/
H A Dno_hz.rst52 However, if you are instead running a light workload with long idle
59 In addition, if you are running either a real-time workload or an HPC
60 workload with short iterations, the scheduling-clock interrupts can
61 degrade your applications performance. If this describes your workload,
203 but do not see any change in your workload's behavior. Is this because
204 your workload isn't affected that much by OS jitter, or is it because
215 possible, then you can conclude that your workload is not all that
291 constraining the workload. For example, the only way to eliminate
/linux/Documentation/admin-guide/mm/damon/
H A Dstart.rst101 with your real workload. The last line asks ``damo`` to record the access
174 >=60 seconds in your workload to be swapped out. ::
178 --target_pid <pid of your workload>

12345