The project team is adding one new command line option, "-s", to the
latencytop tool. Below are the output from the current and the proposed
command, respectively. We would like to archive this enhancement in the
case directory.
Thanks,
Krish
----------------------------------------------------------------------------------------------------------------------------------------------------------
# latencytop -h
Usage: latencytop [option(s)]
Options:
-h, --help
Print this help.
-t, --interval TIME
Set refresh interval to TIME. Valid range [1...60] seconds,
default = 5
-o, --output-log-file FILE
Output kernel log to FILE. Default = /var/log/latencytop.log
-k, --kernel-log-level LEVEL
Set kernel log level to LEVEL.
0(default) = None, 1 = Unmapped, 2 = Mapped, 3 = All.
-f, --feature [no]feature1,[no]feature2,...
Enable/disable features in LatencyTOP.
[no]filter:
Filter large interruptible latencies, e.g. sleep.
[no]sched:
Monitors sched (PID=0).
[no]sobj:
Monitors synchronization objects.
[no]low:
Lower overhead by sampling small latencies.
-l, --log-period TIME
Write and restart log every TIME seconds, TIME >= 60
#latencytop -h
Usage: latencytop [option(s)]
Options:
-h, --help
Print this help.
-t, --interval TIME
Set refresh interval to TIME. Valid range [1...60] seconds,
default = 5
-o, --output-log-file FILE
Output kernel log to FILE. Default = /var/log/latencytop.log
-k, --kernel-log-level LEVEL
Set kernel log level to LEVEL.
0(default) = None, 1 = Unmapped, 2 = Mapped, 3 = All.
-f, --feature [no]feature1,[no]feature2,...
Enable/disable features in LatencyTOP.
[no]filter:
Filter large interruptible latencies, e.g. sleep.
[no]sched:
Monitors sched (PID=0).
[no]sobj:
Monitors synchronization objects.
[no]low:
Lower overhead by sampling small latencies.
-l, --log-period TIME
Write and restart log every TIME seconds, TIME >= 60
-s --select [ pid=<pid> | pgid=<pgid> ]
Monitor only the given process or processes in the given process
group.
