インテル® VTune™ Amplifier 2018 ヘルプ
Use the collect-with action to configure and run a custom analysis using any of the following data collectors:
The hardware event-based sampling collector of the VTune Amplifier profiles your application using the counter overflow feature of the Performance Monitoring Unit (PMU).
Syntax:
$ amplxe-cl -collect-with runsa [-knob <knobName=knobValue>] [--] <target>
Knobs:collect-io-waits, io-mode, enable-stack-collection, stack-size, stack-type, enable-call-counts, enable-trip-counts, event-config, chipset-event-config, ftrace-config, enable-user-tasks, enable-system-cswitch, collect tsx-cycles, enable-gpu-usage, gpu-counters-mode, gpu-sampling-interval, enable-gpu-runtimes, analyze-loops, mrte-type, event-mode, enable-context-switches, sampling-interval, analyze-mem-objects (Linux* targets only), mem-object-size-min-thres (Linux targets only), dram-bandwidth-limits, analyze-openmp.
For the most current information on available knobs (configuration options) for the hardware event-based sampling, enter:
$ amplxe-cl -help collect-with runsa
To display a list of events available on the target PMU, enter:
$ amplxe-cl -collect-with runsa -knob event-config=? <target>
The command returns names and short descriptions of available events. For more information on the events, use Intel Processor Events Reference
Example:
This example runs a custom hardware event-based sampling collection for the sample application with the specified events:
> amplxe-cl -collect-with runsa -knob event-config=CPU_CLK_UNHALTED.CORE,CPU_CLK_UNHALTED.REF,INST_RETIRED.ANY -- home/test/sampleExample:
This example configures and runs a custom event-based sampling data collection with the stack size limited to 8192 bytes and defines a custom Sample After value for the CPU_CLK_UNHALTED.REF_TSC event using the sa option:
> amplxe-cl -collect-with runsa -knob enable-stack-collection=true -knob stack-size=8192 -knob enable-call-counts=true -knob event-config=CPU_CLK_UNHALTED.REF_TSC:sa=1800000,CPU_CLK_UNHALTEDThe user-mode sampling and tracing collector profiles an application execution and takes snapshots of how that application utilizes the processors in the system. The collector interrupts a process, collects the value of all active instruction addresses and captures a calling sequence for each of these samples.
Syntax:
$ amplxe-cl -collect-with runss [-knob <knobName=knobValue>] [--] <target>
Knobs: cpu-sampling-mode, waits-mode, signals-mode, io-mode, enable-user-tasks, enable-user-sync, stackwalk-mode, stack-stitching, ftrace-config,no-altstack, enable-gpu-usage, gpu-counters-mode, gpu-sampling-interval, enable-gpu-runtimes, analyze-loops, mrte-type, sampling-interval, analyze-openmp.
For the most current information on available knobs (configuration options) for the user-mode sampling and tracing, enter:
$ amplxe-cl -help collect-with runss
Example:
This example runs user-mode sampling and tracing collection for the sample application with enabled loop analysis.
> amplxe-cl -collect-with runss -knob analyze-loops=true -- home/test/sampleWhen the data collection is complete, do one of the following to view the result:
Use the -report action to view the data from command line.
Use the -report-output action to write report to a .txt or .csv file
Open the data collection result (*.amplxe) in the VTune Amplifier graphical interface.