User Guide

Contents

Import External Data

Correlate interval or discrete data provided by an external collector with the regular data provided by the
Intel® VTune™
Profiler
.
For example, you can see how the data captured from SoCs or peripheral devices (camera, touch screen, sensors, and so on) correlate with
VTune
Profiler
metrics collected for your analysis target.
VTune
Profiler
can load and process the following data types:
  • Interval data with start time and end time
  • Samples with a set of counters
Data may be optionally bound to process and thread ID.
To add external performance statistics to a
VTune
Profiler
result:

Launch a Custom Data Collector

Collect custom performance data using one of the following modes:
  • Application mode:
    You can leverage the statistics collected by your target application to enhance the
    VTune
    Profiler
    analysis. For example, a part of your application has many instances executed many times in one run and some of these instances exhibit a performance problem. You can retrieve time frames where problems occur from your application log file and supply this data to the
    VTune
    Profiler
    .
  • Custom collector mode:
    If you cannot/do not want to collect statistics directly by your application during the
    VTune
    Profiler
    analysis, you may either create a custom collector or use an existing external collector (for example, ftrace, ETW, logcatthat ) and launch it from the
    VTune
    Profiler
    . To enable this mode, configure a
    VTune
    Profiler
    analysis type to use the Custom collector option and specify a command starting your external collector.

Convert Custom Data to the CSV Format and Import It to
VTune
Profiler

To import the externally collected data to the
VTune
Profiler
:
  1. Convert the collected custom data to a
    csv
    file with a predefined structure .
    To do this for the custom collector mode, you need to configure the collector to output the data in the required CSV format using the
    VTUNE
    _HOSTNAME
    environment variable that identifies the name of the current host required for the
    csv
    file format. For the application mode, you may identify the hostname from the
    Computer name
    field provided in the
    Summary
    window for your result, or from the
    summary
    command line report.
  2. Import the
    csv
    file to the
    VTune
    Profiler
    result using any of the following options:
    in GUI:
    1. Open the
      VTune
      Profiler
      result that was launched in parallel with the external data collection.
    2. Open the
      Analysis Target
      tab, or
      Analysis Type
      tab.
    3. Click the
      Import from CSV
      button on the command toolbar on the left.
      The
      Choose a File to Import
      dialog box opens.
    4. Navigate to the required
      csv
      file and click
      Open
      . You may import several
      csv
      files at a time.
      Importing a
      csv
      file to the
      VTune
      Profiler
      result does not affect symbol resolution in the result. For example, you can safely import a
      csv
      file to a result located on a system where module and debug information is not available.
    in CLI: use the option as follows:
    vtune
    -r <
    existing result dir
    > -import <
    path to csv file
    >
    VTune
    Profiler
    processes the data gathered by its own collectors and the external application and provides an integrated picture of your code performance in its standard data views, such as the
    Timeline
    pane,
    Bottom-up
    pane and others.
If you develop a custom collector yourself, you may use the
VTUNE
_DATA_DIR
environment variable to make your collector identify the
VTune
Profiler
result directory and automatically save the custom collection result (in the CSV format) to this directory. In this case, external statistics will be imported to the
VTune
Profiler
result automatically.

Product and Performance Information

1

Intel's compilers may or may not optimize to the same degree for non-Intel microprocessors for optimizations that are not unique to Intel microprocessors. These optimizations include SSE2, SSE3, and SSSE3 instruction sets and other optimizations. Intel does not guarantee the availability, functionality, or effectiveness of any optimization on microprocessors not manufactured by Intel. Microprocessor-dependent optimizations in this product are intended for use with Intel microprocessors. Certain optimizations not specific to Intel microarchitecture are reserved for Intel microprocessors. Please refer to the applicable product User and Reference Guides for more information regarding the specific instruction sets covered by this notice.

Notice revision #20110804