The Performance Call Graph shows the dynamic call structure of the program you are profiling, and indicates where the program spends most of its time.
To open the Performance Call Graph, right-click a profiling resource in the Profiling Monitor and select
from the pop-up menu.Here are some tips for using the graph:
Data item | Description |
---|---|
Method | The name of the method over which you have paused the mouse cursor. |
Source File | If source code is available, the name of the source file for the selected method. |
Calls | The total number of calls made to the selected method. |
Base Time | The time spent in the selected method, excluding the times for any descendant methods. |
Cumulative Time | The Base Time for the selected method plus the times spent in all descendant methods on behalf of the selected method. |
Min Time | The smallest time recorded for a single invocation of the selected method. |
Max Time | The largest time recorded for a single invocation of the selected method. |
Avg Time | The average time per call spent in the selected method (Base Time/Calls). |
Caller | For an arc between nodes, the name of the calling node. |
Descendant | For an arc between nodes, the name of the descendant node. |
Calls | For an arc between nodes, the number of calls made to the Descendant by the Caller. |
Propagated Time | For an arc between nodes, the time spent in the Descendant on behalf of the Caller. |
Control | Command | Description |
---|---|---|
Redraws the call graph in the default layout. | ||
Displays or hides the call graph overview. | ||
![]() |
Opens the Method Details view for the currently selected method. | |
![]() |
Lets you specify a method name to locate within the graph. | |
![]() |
Lets you change the colors used in the graph. | |
In the node selection history, navigates to the previously selected node. | ||
In the node selection history, navigates to the next selected node. | ||
Pull-down menu | Line Scale Factor | Specifies how the lines in the graph are displayed. You can specify | or . Linear line scaling uses the thickness of lines linking nodes to represent the distribution of measured resources from a node to its immediate neighbor, as a percentage of the total for the profiling run. The default is .
Pull-down menu | Lets you change how method names are displayed in the Performance Call Graph view and the Method Details view. | |
Pull-down menu | Lets you select the time units to use: seconds, milliseconds, or microseconds. | |
Pull-down menu | Lets you select the precision to use: 0.0 to 0.00000, or none. | |
Pop-up menu | Refresh Views | Repopulates all views with the most recent data for the selected profiling resource. |
Pop-up menu | Filter | Includes or excludes methods from the analysis data set. For more information, see Performance Call Graph data commands. |
Pop-up menu | Subtree | Hides, deletes, or focuses on a subtree. For more information, see Performance Call Graph data commands. |
Pop-up menu | Expand/Collapse | Displays or hides the descendants of the selected method. For more information, see Performance Call Graph data commands. |
Pop-up menu | Open Source | When source code for the selected method is available, opens the source file in the default editor. |
Highlight drop-down list | Hightlight Methods | Highlights methods based upon criteria you select in the list. For more information, see Performance Call Graph data commands. |
Zoom slider | Zoom the graph | Changes the scale of the graph. |
Parent topic: Performance bottleneck detection user interface reference
Related tasks
Getting information from the Performance Call Graph view
Customizing the Performance Call Graph view
Related reference
Performance Call Graph data commands