Typical LISA experiment
This notebook shows a typical LISA-use case:
Connecting to a target
Configuring an rt-app workload
Collecting a trace while executing a workload
Displaying the trace
Analysing the trace
It can serve as a template for different kind of experiments, - you could only change the workload to execute & the trace events to collect
[1]:
import logging
from lisa.utils import setup_logging
setup_logging()
2021-07-01 17:08:58,411 INFO : root : Using LISA logging configuration: /home/dourai01/Work/lisa/lisa/logging.conf
Target configuration
Target communication is abstracted away under a Target class. We’re going to create an instance of it and that’ll let us run whatever experiment we want on a given target.
Relevant documentation:
Target: https://tooling.sites.arm.com/lisa/latest/target.html#lisa.target.Target
TargetConf: https://tooling.sites.arm.com/lisa/latest/target.html#lisa.target.TargetConf
[2]:
from lisa.target import Target, TargetConf
[3]:
# target = Target(
# kind='linux',
# name='myhikey960',
# host='192.168.0.1',
# username='root',
# password='root',
# )
# Uses settings from target_conf.yml
target = Target.from_default_conf()
2021-07-01 17:09:01,454 INFO : lisa.target.Target : Target configuration:
+- devlib:
|- excluded-modules from default (list): []
+- platform:
|- class from default (str): devlib.platform.Platform
|- host from user (str): 10.7.6.218
|- kind from user (str): linux
|- lazy-platinfo from default (bool): False
|- name from user (str): juno-r0
|- password from user (str): <password>
|- strict-host-check from user (bool): False
|- tools from default (list): []
|- username from user (str): root
+- wait-boot:
|- enable from default (bool): True
|- timeout from default (int): 10
2021-07-01 17:09:01,564 INFO : lisa.target.Target : Creating result directory: /home/dourai01/Work/lisa/lisa/results/Target-juno-r0-20210701_170901.564655
2021-07-01 17:09:01,565 INFO : lisa.target.Target : User-defined platform information:
2021-07-01 17:09:05,344 INFO : lisa.target.Target : Connected to target juno-r0
2021-07-01 17:09:05,648 INFO : lisa.energy_model.EnergyModel.from_target : Attempting to load EM using LinuxEnergyModel
2021-07-01 17:09:07,192 INFO : lisa.target.Target : Loading target devlib module cpuidle
2021-07-01 17:09:09,126 INFO : lisa.target.Target : Loading target devlib module cpufreq
2021-07-01 17:09:10,581 INFO : lisa.target.Target : Loading target devlib module sched
2021-07-01 17:09:11,964 INFO : sched : Scheduler sched_domain procfs entries not found
2021-07-01 17:09:11,974 INFO : sched : Detected kernel compiled with SCHED_DEBUG=n
2021-07-01 17:09:11,977 INFO : sched : CPU capacity sysfs entries found
2021-07-01 17:09:14,965 INFO : lisa.target.Target : Effective platform information:
|- abi from target (str): arm64
+- cpu-capacities:
|- orig from target (dict): {0: 446, 1: 1024, 2: 1024, 3: 446, 4: 446, 5: 446}
|- writeable from target (bool): True
|- rtapp from target(platform-info/rtapp/calib),target(platform-info/cpu-capacities/orig),target(platform-info/cpu-capacities/writeable) (str): <depends on lazy keys: platform-info/rtapp/calib>
|- cpus-count from target (int): 6
|- freq-domains from target (list): [[0, 3, 4, 5], [1, 2]]
|- freqs from target (dict): {0: [450000, 575000, 700000, 775000, 850000], 1: [450000, 625000, 800000, 950000, 1100000], 2: [450000, 625000, 800000, 950000, 1100000], 3: [450000, 575000, 700000, 775000, 850000], 4: [450000, 575000, 700000, 775000, 850000], 5: [450000, 575000, 700000, 775000, 850000]}
+- kernel:
|- config from target (TypedKernelConfig): <kernel config>
|- symbols-address from target (DeferredValue): <symbols address>
|- version from target (KernelVersion): 5.13.0-rc3-g5c95b23cc7dd 1 SMP PREEMPT Thu Jul 1 16:07:56 BST 2021
|- name from target-conf (str): juno-r0
|- nrg-model from target (LinuxEnergyModel): <lisa.energy_model.LinuxEnergyModel object at 0x7f598c2faee0>
|- numa-nodes-count from target (int): 1
|- os from target (str): linux
+- rtapp:
|- calib from target (DeferredValue): <lazy value of RTA.get_cpu_calibrations>
|- capacity-classes from target(platform-info/cpu-capacities/orig) (list): [[0, 3, 4, 5], [1, 2]]
Setting up an rt-app workload
rt-app is very convenient for scheduler experiments, and the majority of the tests within LISA rely on it. Here we’re going to create a somewhat useless workload just to show off the API.
Relevant documentation:
rt-app LISA class: https://tooling.sites.arm.com/lisa/latest/workloads.html#lisa.wlgen.rta.RTA
RTAPhase class: https://tooling.sites.arm.com/lisa/latest/workloads.html#lisa.wlgen.rta.RTAPhase
PeriodicWload class: https://tooling.sites.arm.com/lisa/latest/workloads.html#lisa.wlgen.rta.PeriodicWload
[4]:
from lisa.wlgen.rta import RTA, RTAPhase, PeriodicWload
[5]:
rtapp_profile = {
f'tsk{cpu}_{i}': RTAPhase(
prop_wload=PeriodicWload(
duty_cycle_pct=20,
period=16e-3,
duration=1,
)
)
for cpu in range(target.number_of_cpus)
for i in range(3)
}
rt-app needs some calibration information (20% duty cycle isn’t the same amount of work on all platforms!). It can be manually specified like so:
[6]:
def provide_calibration(calibration):
target.plat_info["rtapp"].add_src("user", {"calib" : calibration})
[7]:
# Uncomment if you want to use this
# provide_calibration({0: 307, 1: 302, 2: 302, 3: 302, 4: 155, 5: 155, 6: 155, 7: 155})
However, it is automatically collected when first creating an rt-app workload if it is not specified, so you can forego the above step and let the calibration happen on-demand:
[8]:
wload = RTA.from_profile(target, rtapp_profile, name='experiment_workload')
2021-07-01 17:09:15,029 INFO : lisa.target.Target : Creating result directory: /home/dourai01/Work/lisa/lisa/results/Target-juno-r0-20210701_170901.564655/RTA-experiment_workload-20210701_170915.028850
2021-07-01 17:09:19,224 INFO : lisa.target.Target : Disabling idle states for all domains
2021-07-01 17:09:24,703 INFO : lisa.wlgen.rta.RTA : Created workload's run target directory: /root/devlib-target/lisa/wlgen/20210701_170924_e999fea4ccb94f3889f8de836d4a7918
2021-07-01 17:09:25,374 INFO : lisa.target.Target : Loading target devlib module cgroups
2021-07-01 17:09:25,574 INFO : CGroups : Available controllers:
2021-07-01 17:09:26,004 INFO : CGroups : cpuset : /root/devlib-target/cgroups/devlib_cgh1
2021-07-01 17:09:26,443 INFO : CGroups : cpu : /root/devlib-target/cgroups/devlib_cgh1
2021-07-01 17:09:26,878 INFO : CGroups : cpuacct : /root/devlib-target/cgroups/devlib_cgh1
2021-07-01 17:09:27,308 INFO : CGroups : blkio : /root/devlib-target/cgroups/devlib_cgh1
2021-07-01 17:09:27,747 INFO : CGroups : memory : /root/devlib-target/cgroups/devlib_cgh1
2021-07-01 17:09:28,210 INFO : CGroups : devices : /root/devlib-target/cgroups/devlib_cgh1
2021-07-01 17:09:28,877 INFO : CGroups : freezer : /root/devlib-target/cgroups/devlib_cgh1
2021-07-01 17:09:29,319 INFO : CGroups : perf_event : /root/devlib-target/cgroups/devlib_cgh1
2021-07-01 17:09:29,749 INFO : CGroups : hugetlb : /root/devlib-target/cgroups/devlib_cgh1
2021-07-01 17:09:30,200 INFO : CGroups : pids : /root/devlib-target/cgroups/devlib_cgh1
2021-07-01 17:09:30,331 INFO : lisa.target.Target : Freezing all tasks except: init,systemd[^-],dbus,sh,ssh,rsyslogd,jbd2
2021-07-01 17:09:33,663 INFO : lisa.wlgen.rta.RTA : Execution start: rt-app /root/devlib-target/lisa/wlgen/20210701_170924_e999fea4ccb94f3889f8de836d4a7918/rta_calib_cpu0.json 2>&1
2021-07-01 17:09:40,224 INFO : lisa.target.Target : Un-freezing userspace tasks
2021-07-01 17:09:40,782 INFO : lisa.wlgen.rta.RTA : Wiping target run directory: /root/devlib-target/lisa/wlgen/20210701_170924_e999fea4ccb94f3889f8de836d4a7918
2021-07-01 17:09:41,039 INFO : lisa.wlgen.rta.RTA : CPU0 calibration=335
2021-07-01 17:09:41,348 INFO : lisa.wlgen.rta.RTA : Created workload's run target directory: /root/devlib-target/lisa/wlgen/20210701_170941_ea3bba3f9ae6408ab7026db86fb3f3da
2021-07-01 17:09:41,676 INFO : lisa.target.Target : Freezing all tasks except: init,systemd[^-],dbus,sh,ssh,rsyslogd,jbd2
2021-07-01 17:09:44,843 INFO : lisa.wlgen.rta.RTA : Execution start: rt-app /root/devlib-target/lisa/wlgen/20210701_170941_ea3bba3f9ae6408ab7026db86fb3f3da/rta_calib_cpu1.json 2>&1
2021-07-01 17:09:51,248 INFO : lisa.target.Target : Un-freezing userspace tasks
2021-07-01 17:09:51,832 INFO : lisa.wlgen.rta.RTA : Wiping target run directory: /root/devlib-target/lisa/wlgen/20210701_170941_ea3bba3f9ae6408ab7026db86fb3f3da
2021-07-01 17:09:52,091 INFO : lisa.wlgen.rta.RTA : CPU1 calibration=206
2021-07-01 17:09:52,456 INFO : lisa.wlgen.rta.RTA : Created workload's run target directory: /root/devlib-target/lisa/wlgen/20210701_170952_37f23292cad449c19fba0c7d35b93e9c
2021-07-01 17:09:52,894 INFO : lisa.target.Target : Freezing all tasks except: init,systemd[^-],dbus,sh,ssh,rsyslogd,jbd2
2021-07-01 17:09:56,014 INFO : lisa.wlgen.rta.RTA : Execution start: rt-app /root/devlib-target/lisa/wlgen/20210701_170952_37f23292cad449c19fba0c7d35b93e9c/rta_calib_cpu2.json 2>&1
2021-07-01 17:10:02,572 INFO : lisa.target.Target : Un-freezing userspace tasks
2021-07-01 17:10:03,134 INFO : lisa.wlgen.rta.RTA : Wiping target run directory: /root/devlib-target/lisa/wlgen/20210701_170952_37f23292cad449c19fba0c7d35b93e9c
2021-07-01 17:10:03,389 INFO : lisa.wlgen.rta.RTA : CPU2 calibration=191
2021-07-01 17:10:03,722 INFO : lisa.wlgen.rta.RTA : Created workload's run target directory: /root/devlib-target/lisa/wlgen/20210701_171003_6ce86454f5f04772b0ed1641fa14f1f4
2021-07-01 17:10:04,030 INFO : lisa.target.Target : Freezing all tasks except: init,systemd[^-],dbus,sh,ssh,rsyslogd,jbd2
2021-07-01 17:10:07,140 INFO : lisa.wlgen.rta.RTA : Execution start: rt-app /root/devlib-target/lisa/wlgen/20210701_171003_6ce86454f5f04772b0ed1641fa14f1f4/rta_calib_cpu3.json 2>&1
2021-07-01 17:10:13,801 INFO : lisa.target.Target : Un-freezing userspace tasks
2021-07-01 17:10:14,380 INFO : lisa.wlgen.rta.RTA : Wiping target run directory: /root/devlib-target/lisa/wlgen/20210701_171003_6ce86454f5f04772b0ed1641fa14f1f4
2021-07-01 17:10:14,624 INFO : lisa.wlgen.rta.RTA : CPU3 calibration=335
2021-07-01 17:10:15,064 INFO : lisa.wlgen.rta.RTA : Created workload's run target directory: /root/devlib-target/lisa/wlgen/20210701_171014_41170e6e2b254b5084418c62243e247f
2021-07-01 17:10:15,367 INFO : lisa.target.Target : Freezing all tasks except: init,systemd[^-],dbus,sh,ssh,rsyslogd,jbd2
2021-07-01 17:10:18,450 INFO : lisa.wlgen.rta.RTA : Execution start: rt-app /root/devlib-target/lisa/wlgen/20210701_171014_41170e6e2b254b5084418c62243e247f/rta_calib_cpu4.json 2>&1
2021-07-01 17:10:24,967 INFO : lisa.target.Target : Un-freezing userspace tasks
2021-07-01 17:10:25,545 INFO : lisa.wlgen.rta.RTA : Wiping target run directory: /root/devlib-target/lisa/wlgen/20210701_171014_41170e6e2b254b5084418c62243e247f
2021-07-01 17:10:25,809 INFO : lisa.wlgen.rta.RTA : CPU4 calibration=335
2021-07-01 17:10:26,154 INFO : lisa.wlgen.rta.RTA : Created workload's run target directory: /root/devlib-target/lisa/wlgen/20210701_171025_3113de7c35de4ed191cef0d13122a690
2021-07-01 17:10:26,470 INFO : lisa.target.Target : Freezing all tasks except: init,systemd[^-],dbus,sh,ssh,rsyslogd,jbd2
2021-07-01 17:10:29,547 INFO : lisa.wlgen.rta.RTA : Execution start: rt-app /root/devlib-target/lisa/wlgen/20210701_171025_3113de7c35de4ed191cef0d13122a690/rta_calib_cpu5.json 2>&1
2021-07-01 17:10:36,132 INFO : lisa.target.Target : Un-freezing userspace tasks
2021-07-01 17:10:36,709 INFO : lisa.wlgen.rta.RTA : Wiping target run directory: /root/devlib-target/lisa/wlgen/20210701_171025_3113de7c35de4ed191cef0d13122a690
2021-07-01 17:10:36,975 INFO : lisa.wlgen.rta.RTA : CPU5 calibration=335
2021-07-01 17:10:37,010 INFO : lisa.wlgen.rta.RTA : Platform info rt-app calibration configuration:
platform-info:
conf:
cpu-capacities: {}
kernel: {}
rtapp:
calib:
0: 335
1: 206
2: 191
3: 335
4: 335
5: 335
2021-07-01 17:10:37,043 INFO : lisa.wlgen.rta.RTA : CPU capacities according to rt-app workload: {0: 606, 3: 606, 4: 606, 5: 606, 1: 1024, 2: 1024}
2021-07-01 17:10:37,044 WARNING : lisa.wlgen.rta.RTA : The calibration values are not inversely proportional to the CPU capacities, the duty cycles will be up to 35.87% off on some CPUs: {0: 135.8744394618834, 1: 100.0, 2: 100.0, 3: 135.8744394618834, 4: 135.8744394618834, 5: 135.8744394618834}
2021-07-01 17:10:37,045 WARNING : lisa.wlgen.rta.RTA : The calibration values are not inversely proportional to the CPU capacities. Either rt-app calibration failed, or the rt-app busy loops has a very different instruction mix compared to the workload used to establish the CPU capacities: {0: 135.8744394618834, 1: 100.0, 2: 100.0, 3: 135.8744394618834, 4: 135.8744394618834, 5: 135.8744394618834}
2021-07-01 17:10:37,045 INFO : lisa.target.Target : Re-enabling idle states for all domains
Running the workload
[9]:
import os
from lisa.trace import FtraceCollector
We need to specify the trace events we want to record. We could list what’s available like so:
[10]:
available_events = target.execute("cat /sys/kernel/debug/tracing/available_events").splitlines()
# That's gonna be a pretty big list, let's focus on the scheduler events
sched_events = [
event
for event in available_events
if (
event.startswith("sched:") or
event.startswith("task:")
)
]
print(sched_events)
['task:task_rename', 'task:task_newtask', 'sched:sched_wake_idle_without_ipi', 'sched:sched_swap_numa', 'sched:sched_stick_numa', 'sched:sched_move_numa', 'sched:sched_pi_setprio', 'sched:sched_stat_runtime', 'sched:sched_stat_blocked', 'sched:sched_stat_iowait', 'sched:sched_stat_sleep', 'sched:sched_stat_wait', 'sched:sched_process_exec', 'sched:sched_process_fork', 'sched:sched_process_wait', 'sched:sched_wait_task', 'sched:sched_process_exit', 'sched:sched_process_free', 'sched:sched_migrate_task', 'sched:sched_switch', 'sched:sched_wakeup_new', 'sched:sched_wakeup', 'sched:sched_waking', 'sched:sched_kthread_work_execute_end', 'sched:sched_kthread_work_execute_start', 'sched:sched_kthread_work_queue_work', 'sched:sched_kthread_stop_ret', 'sched:sched_kthread_stop', 'sched:sched_cpu_capacity', 'sched:uclamp_util_cfs', 'sched:uclamp_util_se', 'sched:sched_util_est_cfs', 'sched:sched_util_est_se', 'sched:sched_overutilized', 'sched:sched_pelt_se', 'sched:sched_pelt_irq', 'sched:sched_pelt_dl', 'sched:sched_pelt_rt', 'sched:sched_pelt_cfs']
Let’s just collect the base events required to plot task scheduling:
[11]:
events = [
'sched_switch',
'sched_wakeup',
'sched_wakeup_new',
'task_rename',
]
And now we can actually record traces while running our workload:
[12]:
trace_path = os.path.join(wload.res_dir, "trace.dat")
ftrace_coll = FtraceCollector(target, events=events, buffer_size=10240, output_path=trace_path)
with wload, ftrace_coll:
wload.run()
2021-07-01 17:10:49,741 INFO : lisa.wlgen.rta.RTA : Created workload's run target directory: /root/devlib-target/lisa/wlgen/20210701_170915_915f1b7cc21d4dadb9eac4de0ec2f72e
2021-07-01 17:10:50,039 INFO : lisa.wlgen.rta.RTA : Will update CPU capacities in sysfs: {0: 606, 3: 606, 4: 606, 5: 606, 1: 1024, 2: 1024}
2021-07-01 17:11:00,847 INFO : lisa.wlgen.rta.RTA : Execution start: rt-app /root/devlib-target/lisa/wlgen/20210701_170915_915f1b7cc21d4dadb9eac4de0ec2f72e/experiment_workload.json 2>&1
2021-07-01 17:11:11,142 INFO : lisa.wlgen.rta.RTA : Wiping target run directory: /root/devlib-target/lisa/wlgen/20210701_170915_915f1b7cc21d4dadb9eac4de0ec2f72e
Loading up the trace
We have a Trace class that lets us easily access trace events. It can also do some post-processing to provide different kinds of analysis.
[13]:
from lisa.trace import Trace
We also save some platform information (number of CPUs, available frequencies, kernel version…) that comes in handy for doing some analysis:
[14]:
print(target.plat_info)
|- abi from target (str): arm64
+- cpu-capacities:
|- orig from target (dict): {0: 446, 1: 1024, 2: 1024, 3: 446, 4: 446, 5: 446}
|- writeable from target (bool): True
|- rtapp from target(platform-info/rtapp/calib),target(platform-info/cpu-capacities/orig),target(platform-info/cpu-capacities/writeable) (dict): {0: 606, 3: 606, 4: 606, 5: 606, 1: 1024, 2: 1024}
|- cpus-count from target (int): 6
|- freq-domains from target (list): [[0, 3, 4, 5], [1, 2]]
|- freqs from target (dict): {0: [450000, 575000, 700000, 775000, 850000], 1: [450000, 625000, 800000, 950000, 1100000], 2: [450000, 625000, 800000, 950000, 1100000], 3: [450000, 575000, 700000, 775000, 850000], 4: [450000, 575000, 700000, 775000, 850000], 5: [450000, 575000, 700000, 775000, 850000]}
+- kernel:
|- config from target (TypedKernelConfig): <kernel config>
|- symbols-address from target (DeferredValue): <symbols address>
|- version from target (KernelVersion): 5.13.0-rc3-g5c95b23cc7dd 1 SMP PREEMPT Thu Jul 1 16:07:56 BST 2021
|- name from target-conf (str): juno-r0
|- nrg-model from target (LinuxEnergyModel): <lisa.energy_model.LinuxEnergyModel object at 0x7f598c1fccd0>
|- numa-nodes-count from target (int): 1
|- os from target (str): linux
+- rtapp:
|- calib from target (dict): {0: 335, 1: 206, 2: 191, 3: 335, 4: 335, 5: 335}
|- capacity-classes from target(platform-info/cpu-capacities/orig) (list): [[0, 3, 4, 5], [1, 2]]
You can pass the platform info directly from the Target:
[15]:
trace_path = os.path.join(wload.res_dir, 'trace.dat')
trace = Trace(trace_path, target.plat_info, events=events)
2021-07-01 17:11:12,517 INFO : lisa.platforms.platinfo.PlatformInfo : Attempting to read kallsyms from target
… But it’s a good idea to save it on the disk so that you can re-run whatever analysis code you want several months down the line after the platform was lost in a tragic fire.
It’s why we save this information somewhere instead of polling the target when we want to use them - we can run analysis code offline.
Here we show how to save to/restore this platform information from the disk.
[16]:
plat_info_path = os.path.join(wload.res_dir, "platinfo.yaml")
target.plat_info.to_yaml_map(plat_info_path)
[17]:
from lisa.platforms.platinfo import PlatformInfo
[18]:
plat_info = PlatformInfo.from_yaml_map(plat_info_path)
trace = Trace(trace_path, plat_info, events=events)
Looking at the trace
Kernelshark can be opened from the notebook:
[19]:
# trace.show()
Analysing the trace
Relevant documentation: https://tooling.sites.arm.com/lisa/latest/trace_analysis.html
Reading trace events
[20]:
df = trace.df_event("sched_switch")
df
[20]:
| __comm | __pid | __cpu | next_pid | prev_comm | prev_prio | next_comm | prev_pid | next_prio | prev_state | |
|---|---|---|---|---|---|---|---|---|---|---|
| Time | ||||||||||
| 2745.781282 | trace-cmd | 5635 | 0 | 3981 | trace-cmd | 120 | sshd | 5635 | 120 | 256 |
| 2745.781622 | <idle> | 0 | 5 | 12 | swapper/5 | 120 | rcu_preempt | 0 | 120 | 0 |
| 2745.781709 | sshd | 3981 | 0 | 5635 | sshd | 120 | trace-cmd | 3981 | 120 | 2 |
| 2745.781749 | rcu_preempt | 12 | 5 | 0 | rcu_preempt | 120 | swapper/5 | 12 | 120 | 128 |
| 2745.781848 | trace-cmd | 5635 | 0 | 3981 | trace-cmd | 120 | sshd | 5635 | 120 | 256 |
| ... | ... | ... | ... | ... | ... | ... | ... | ... | ... | ... |
| 2747.986134 | <idle> | 0 | 3 | 12 | swapper/3 | 120 | rcu_preempt | 0 | 120 | 0 |
| 2747.986260 | rcu_preempt | 12 | 3 | 0 | rcu_preempt | 120 | swapper/3 | 12 | 120 | 128 |
| 2747.987432 | <idle> | 0 | 0 | 3981 | swapper/0 | 120 | sshd | 0 | 120 | 0 |
| 2747.987845 | sshd | 3981 | 0 | 0 | sshd | 120 | swapper/0 | 3981 | 120 | 1 |
| 2747.993821 | <idle> | 0 | 3 | 12 | swapper/3 | 120 | rcu_preempt | 0 | 120 | 0 |
2644 rows × 10 columns
The standard DataFrame operations are available, so you can filter/slice it however you wish:
[21]:
df[df.next_comm == "tsk4_1"][1.5:2]
[21]:
| __comm | __pid | __cpu | next_pid | prev_comm | prev_prio | next_comm | prev_pid | next_prio | prev_state | |
|---|---|---|---|---|---|---|---|---|---|---|
| Time |
Using the trace analysis
Example dataframes
LISA ships a number of namespaced trace analysis methods. They can all be called on a trace with trace.ana.<analysis name>.<method name>(). They fall mostly into two categories: * Method starting with df_: returns a pandas DataFrame * Method starting with plot_: returns a holoviews element ready to be displayed
[22]:
trace.ana.tasks.df_tasks_runtime()
[22]:
| runtime | comm | |
|---|---|---|
| pid | ||
| 0 | 1.389661 | swapper/4 |
| 5659 | 0.362221 | tsk2_0-6 |
| 5662 | 0.361775 | tsk3_0-9 |
| 5657 | 0.356759 | tsk1_1-4 |
| 5658 | 0.351268 | tsk1_2-5 |
| 5670 | 0.332309 | tsk5_2-17 |
| 5654 | 0.331242 | tsk0_1-1 |
| 5669 | 0.330871 | tsk5_1-16 |
| 5664 | 0.319778 | tsk3_2-11 |
| 5655 | 0.316757 | tsk0_2-2 |
| 5668 | 0.308678 | tsk5_0-15 |
| 5661 | 0.279637 | tsk2_2-8 |
| 5663 | 0.269628 | tsk3_1-10 |
| 5660 | 0.267880 | tsk2_1-7 |
| 5656 | 0.263531 | tsk1_0-3 |
| 5667 | 0.260929 | tsk4_2-14 |
| 5666 | 0.252416 | tsk4_1-13 |
| 5665 | 0.239848 | tsk4_0-12 |
| 5653 | 0.227776 | tsk0_0-0 |
| 3981 | 0.108538 | sshd |
| 5652 | 0.078533 | rt-app |
| 5639 | 0.077796 | shutils |
| 5637 | 0.064597 | shutils |
| 5671 | 0.063522 | shutils |
| 5674 | 0.048984 | sh |
| 5636 | 0.045780 | sh |
| 5673 | 0.045675 | sh |
| 5640 | 0.026702 | true |
| 5648 | 0.025580 | true |
| 5646 | 0.025421 | true |
| 5650 | 0.024910 | true |
| 5638 | 0.023517 | busybox |
| 5644 | 0.022757 | true |
| 5672 | 0.020670 | busybox |
| 5642 | 0.020435 | true |
| 12 | 0.015584 | rcu_preempt |
| 5641 | 0.014737 | busybox |
| 5649 | 0.014633 | busybox |
| 5647 | 0.013716 | busybox |
| 5643 | 0.012695 | busybox |
| 5645 | 0.012236 | busybox |
| 5651 | 0.008759 | busybox |
| 1 | 0.001109 | init |
| 10 | 0.001085 | kworker/0:1 |
| 11 | 0.000950 | ksoftirqd/0 |
| 46 | 0.000875 | kworker/4:1 |
| 5635 | 0.000815 | trace-cmd |
| 26 | 0.000666 | migration/3 |
| 47 | 0.000633 | kworker/5:1 |
| 44 | 0.000618 | kworker/2:1 |
| 3980 | 0.000540 | kworker/u12:0 |
| 36 | 0.000439 | migration/5 |
| 45 | 0.000385 | kworker/3:1 |
| 22 | 0.000329 | ksoftirqd/2 |
| 126 | 0.000307 | kworker/1:2 |
| 51 | 0.000286 | kcompactd0 |
| 31 | 0.000282 | migration/4 |
[23]:
df = trace.ana.tasks.df_task_states("tsk4_1-13", stringify=True)
df
[23]:
| target_cpu | cpu | curr_state | next_state | delta | curr_state_str | next_state_str | |
|---|---|---|---|---|---|---|---|
| Time | |||||||
| 2746.543802 | -1 | 3.0 | 1 | 512 | 0.011418 | S | W |
| 2746.555220 | 2 | 2.0 | 512 | 8192 | 0.000046 | W | A |
| 2746.555266 | -1 | 2.0 | 8192 | 256 | 0.000061 | A | |
| 2746.555327 | -1 | 2.0 | 256 | 8192 | 0.000157 | A | |
| 2746.555484 | -1 | 2.0 | 8192 | 1 | 0.000071 | A | S |
| ... | ... | ... | ... | ... | ... | ... | ... |
| 2747.530640 | -1 | 2.0 | 8192 | 1 | 0.003432 | A | S |
| 2747.534072 | -1 | 2.0 | 1 | 512 | 0.012552 | S | W |
| 2747.546624 | 2 | 2.0 | 512 | 8192 | 0.000019 | W | A |
| 2747.546642 | -1 | 2.0 | 8192 | 16 | 0.000205 | A | X |
| 2747.546847 | -1 | 2.0 | 16 | -1 | NaN | X | U |
211 rows × 7 columns
The trace.ana object can be used to set default values to analysis methods. Simply calling it with keyword arguments will set default values, which can later be overridden when the method is called if necessary.
This avoids repetition of fixed parameters such as ŧask, tasks, cpu etc. Just be careful as some methods might take more parameters than you expect: some task-related methods also accept a cpu parameter to restrict to a given CPU, so it might be a good idea to have a proxy object for all CPU-related calls and another one for task-related calls.
[24]:
ana = trace.ana(task='tsk4_1-13')
df = ana.tasks.df_task_states(stringify=True)
# Default values can be overridden by calling it again
ana2 = ana(task='trace-cmd')
# And overridden again when calling the method
df = ana2.tasks.df_task_states(task='tsk4_1-13', stringify=True)
[25]:
from lisa.analysis.tasks import TaskState
[26]:
df[df.curr_state == TaskState.TASK_ACTIVE][1:1.2]
[26]:
| target_cpu | cpu | curr_state | next_state | delta | curr_state_str | next_state_str | |
|---|---|---|---|---|---|---|---|
| Time |
Example plots
[30]:
import holoviews as hv
# Before rendering any plot, a plotting backend for holoviews has to be chosen.
# Bokeh provides much better interactive (and HTML) plots than matplotlib.
# THIS MUST BE DONE AFTER ALL IMPORTS.
# Otherwise there might be issues that lead to
# not displaying plots until hv.extension() is called again.
hv.extension('bokeh')
[31]:
trace.ana.tasks.plot_task_total_residency("tsk4_1-13")
[31]:
[35]:
trace.ana.tasks.plot_tasks_wakeups_heatmap(bins=200)
[35]:
[36]:
trace.ana.tasks.plot_tasks_forks_heatmap(bins=200)
[36]:
[ ]: