User Guide

Contents

Pane: Call Stack

The
Call Stack
pane is available for the results of the user-mode sampling and tracing collection, such as Hotspots and Threading, and for the results of the hardware event-based sampling with the stack collection enabled.
Use the
Call Stack
pane to identify the call sequences (stacks) that called the program unit currently selected in the grid. Call stacks from different threads are aggregated together, showing all the call stacks for a function, without providing information on what threads were calling. See the table below to understand how to use the data provided in the
Call Stack
pane for the Threading analysis results.
Call Stack Pane
Stack metric drop-down menu
. Select a performance metric to explore the distribution of this metric over stacks of the selected object. For example, for the Threading Efficiency viewpoint the Wait Time metric is pre-selected.
Navigation bar
. Click the next/previous arrows to view stacks for the selected program unit(s). The stack types are classified by metrics and depend on the selected viewpoint. For example, for the
Threading Efficiency
viewpoint the
Wait Time
stack type displays call stacks where the object selected in the grid contributed to the application Wait time.
When multiple stacks lead to the selected program unit, the
Call Stack
pane shows the stack that contributed most to the metric value, the
hottest path
, as the first stack. To see other stacks, click the navigation arrows.
  • If several stacks go to the same functions in different code lines, the bottom-up tree shown in the
    Bottom-up
    grid aggregates these stacks in one line. But the
    Call Stack
    pane shows each as a separate stack.
  • If a selected stack type is not applicable to a selected program unit, the
    VTune
    Profiler
    automatically uses the first applicable stack type from the stack type list instead.
Contribution bar
. Analyze the indicator of the contribution of the currently visible stack to the overall metric data for the selected program unit(s). If you select a single stack in the result window, the Contribution bar shows 100%. If more than one program unit is selected, all the related stacks are added to the calculation.
In the example above, the function selected in the
Bottom-up
grid had 3 Wait Time stacks leading to it with the total Wait time 23.718 seconds. The first stack is responsible for 97.9% (or 23.230s) of the overall 23.718 seconds. Note that the Bottom-up grid aggregates all 3 stacks into one since all of them go to the same function in different code lines.
Call stack for a program unit selected in the grid or in the Timeline pane
. Analyze the call sequence for the selected function according to the stack metric selected in the navigation bar. Each row in the stack represents a function (with an RVA and a line number of the call site, if available) that called the function in the row above it. When the
Call Stack Mode
on the filter toolbar is set to
Only user functions
, the system functions are shown at the bottom of the stack. When set to
User/system functions
, the system functions are shown in the correct location, according to the call sequence.
Click a hyperlink or double-click a function in the stack to open the source exactly where this function was called.
If you see
[Unknown frame(s)]
identifiers in the stack, it means that the
VTune
Profiler
could not locate symbol files for system or your application modules. See the Resolving Unknown Frames topic for more details.
Context menu.
Manage the call stack representation in the
Call Stack
pane (applicable to all stacks). Right-click and select an option. For example, you may de-select the
Show in One-Line Mode
option to view functions in two lines:
When you compare two analysis results, the
Call Stack
pane does not show any call stacks.

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