The content and format of a view can be modified using view management. Use view management to change the kinds of performance data shown in a view, to aggregate or separate data by core (CPU), to aggregate or separate data by task, to aggregate or separate data by thread, and to control display of a percentage column.
Two types of View Management dialog are available:
View management can be initiated in any of three ways:
Global View Management ![]()
Local View Management ![]()
The selected platform name in Platform Name drop-down menu list determines the types of AMD processor family, which in turn determines the available views shown in View name drop-down menu list. Availability of views are determined by the available performance counters on a particular AMD processor family. The selected view determines the choices displayed under Available data and under Columns shown.
The performance data to be shown in the selected view can be changed by moving items between the Available data list and the Columns shown list.
To add an item to the Columns shown list and make it visible in the view:
- Select the item in the Available data list.
- Click the right arrow -> button.
To remove item from the Columns shown list and remove it from the view:
- Select the item in the Columns shown list.
- Click the left arrow <- button.
NOTE : Changes made to view configurations are persistent.
The Platform Name field shows the name of the selected AMD processor family. Processor families are selected from a drop-down list, which is enabled only in the Global View dialog. In Local View dialog, the processor type is determined by the type of processor the profile data of the session was collected on.
The View Name field shows the name of the selected view configuration. View configurations are selected from a drop-down list.
The description changes according to the selection in View name.
CodeAnalyst can only display items listed in the Available data list.
Available data: The items in the Available data list are preselected according to the selected view configuration. Each item is available for inclusion in a view as a column.
Columns shown: Selects which columns are shown and can be changed by using the directional arrows to move an item from the list.
The aggregation check boxes control the separation or aggregation (summation) of sample data. Under Aggregation, three types of separation/aggregation are supported:
When the corresponding check box is selected, sample data is broken out separately by CPU, process, or thread, respectively.
Event data is shown for each processor in separate columns in the System Data table and as separate bars in the System Data graph. For example, on a dual-core processor, timer-based samples display in two columns with the headings:
For Separate Tasks, a separate row is displayed in the System Data table for each unique use of a module by a task. The module name is extended by the task name. A new bar is added in the System Data graph for each unique use of a module by a task.
Enabling Show Percentage displays the percentage column in the view.