xilinx.analyzer is a MATLAB® class that provides an interface between the System Generator model and Vivado® timing paths.
The System Generator timing analysis is supported for all compilation targets. The Perform analysis drop down menu under the Clocking tab of the System Generator token provides two options for the trade-off between total runtime vs. accuracy of the Vivado timing data. If you select either the Post Synthesis or the Post Implementation option of Perform analysis and click the Generate button, then Vivado timing paths information is collected during the netlist generation. The xilinx.analyzer class is used to access Vivado timing paths information. The xilinx.analyzer class object processes Vivado timing paths to find 50 unique paths with the worst slack value. The unique timing paths are sorted in increasing value of slack and saved in the analyzer object.
The cross-probing between Vivado timing paths and the System Generator model is made possible using the following API functions in the xilinx.analyzer class.
Function Name | Description | Function Argument |
---|---|---|
xilinx.analyzer |
This is a constructor of the class. A call to the xilinx.analyzer constructor returns object of the class. |
First argument is System Generator model name. Second argument is path to already generated netlist directory. |
isValid | Indicates if timing analysis data is valid or not. Use this API to make sure that the xilinx.analyzer class construction was successful. | No argument |
getErrorMessage | Returns an error message string if the call to the class constructor or other API function had an error. | No argument |
getStatus | Returns ‘FAILED’ if any of the timing paths in the model have a violation, i.e., negative slack. | No argument |
getVivadoStage | Returns either Post Synthesis or Post Implementation. This is the Vivado design stage after which timing analysis was performed. | No argument |
paths | Returns an array of MATLAB structures. Each structure contains data for a timing path. | A string that is equal to either ‘setup’ or ‘hold’ |
violations | Returns an array of MATLAB structures. Each member of the array is a path structure with a timing violation. | A string that is equal to either ‘setup’ or ‘hold’ |
Prints timing path information such as Slack, Path Delay, Levels of Logic, Name of Source and Destination blocks, and Source and Destination clocks. | An array of MATLAB structures for timing path data. The array can have one or more structures. | |
highlight | In the System Generator model, highlights blocks for the timing path passed in the argument. Blocks that are already highlighted in the model will remain highlighted. | MATLAB structure for one timing path |
highlightOnePath | In the System Generator model, highlights blocks for the timing path passed in the argument. Before highlighting blocks for this path, the blocks that are already highlighted in the model will be unhighlighted. | MATLAB® structure for one timing path |
unhighlight | In the Simulink® model, unhighlights all blocks currently highlighted. | No argument |
disp | Displays a summary of timing analysis results on the MATLAB console, including the worst slack value among all timing paths. | No argument |
delete | This is a destructor of the xilinx.analyzer class | No argument |
Field Name | Description |
---|---|
Slack | The double value containing timing slack for the path |
Delay | Total Data Path delay for the path |
Levels_of_Logic | Number of elements in Vivado design for the timing path. The number of System Generator blocks in the timing path may be different from Levels_of_Logic. |
Source | First System Generator block in the timing path |
Destination | Last System Generator block in the timing path |
Source_Clock | Name of the clock domain for the source block |
Destination_Clock | Name of the clock domain for the destination block |
Path_Constraints | Timing constraint used for the path. For a multi-clock design, the path constraint can be a multi-clock timing constraint. |
Block_Masks | Cell array where each element contains mask information for a System Generator block. |
Simulink_Names | Cell array where each element contains hierarchical name of a block in System Generator model |
Vivado_Names | Cell array where each element contains name of System Generator block in Vivado database |
Type | A timing violation type. The value is either ‘setup’ or ‘hold’. |
xilinx.analyzer - Construct xilinx.analyzer class object
Syntax
analyzer_object = xilinx.analyzer(<name_of_the_model>', '<path_to_netlist_directory>')
- Description
-
A call to xilinx.analyzer constructor returns object of the class.
The first argument is the name of the System Generator model. The model must be open before the class constructor is called.
The second argument is an absolute or relative path to the netlist directory. You must have read permission to the netlist directory.
To access API functions of the xilinx.analyzer class use the object of the class as described below. To get more details for a specific API function type the following at the MATLAB command prompt:
help xilinx.analyzer.<API_function>
- Example
-
//Construct class. Must give the model name and absolute or relative path to the //target directory >> timing_object = xilinx.analyzer('fixed_point_IIR', './netlist_for_timing_analysis') timing_object = Number of setup paths = 9 Worst case setup slack = -1.6430
isValid – Check validity of Vivado timing paths
- Syntax
-
result = analyzer_object.isValid();
- Description
-
If timing analysis data is valid then the result equals '1', otherwise it is '0'. Use this API to make sure that the xilinx.analyzer class construction was successful and the timing data was valid.
- Example
-
//Determine if timing analysis data is valid >> valid_status = timing_object.isValid() valid_status = 1
getErrorMessage - Get an error message
- Syntax
-
result = analyzer_object.getErrorMessage();
- Description
-
Returns an error message string if the call to the class constructor or other API function had an error.
- Example
-
//Determine if there was an error in the xilinx.analyzer constructor //or in any of the API functions >> err_msg = timing_object.getErrorMessage() err_msg = ''
getStatus - Timing analysis status
- Syntax
-
string = analyzer_object.getStatus();
- Description
-
The returned string is either 'PASSED' or 'FAILED'. If any of the timing paths have a violation, i.e. negative slack, then the timing analysis status is considered failed.
- Example
-
//Determine if there were timing path violations in Simulink model >> analysis_status = timing_object.getStatus() analysis_status = FAILED
getVivadoStage - Get Vivado design stage for timing analysis
- Syntax
-
string = analyzer_object.getVivadoStage();
- Description
-
The returned string is the Vivado design stage after which timing analysis was performed and data collected in Vivado. The value is either 'Post Synthesis' or 'Post Implementation'.
- Example
-
//Determine Vivado stage when timing data was collected >> design_stage = timing_object.getVivadoStage() design_stage = Post Synthesis
paths - Access all timing paths
- Syntax
-
<array_of_timing_paths_structure> = analyzer_object.paths('<violation_type>');
- Description
-
The returned value is an array of MATLAB structures. Each structure contains data for a timing path, sorted in decreasing order of timing violation, i.e. in increasing order of slack value.
The argument 'violation_type' is either 'setup' or 'hold' string.
- Example
-
//Return an array of the timing path structures >> all_timing_paths = timing_object.paths('setup') all_timing_paths = 1x9 struct array with fields: Slack Delay Levels_of_Logic Source Destination Source_Clock Destination_Clock Path_Constraints Block_Masks Simulink_Names Vivado_Names Type
Note:There are a total of nine timing paths in this timing analysis.
You can find the data fields in each timing path as shown in Example 1 in Additional Information.
violations - Access paths with timing violations
- Syntax
-
<array_of_timing_paths_structure> = analyzer_object.violations('<violation_type>');
- Description
-
The returned value is an array of MATLAB structures. Each member of the array is data for a path with a timing violation. The array elements are sorted in decreasing order of timing violation. If there are no timing violations in the design then the API function returns an empty array.
The argument 'violation_type' is either 'setup' or 'hold'.
- Example
-
//Return an array of timing paths with setup violations >> violating_paths = timing_object.violations('setup') violating_paths = 1x2 struct array with fields: Slack Delay Levels_of_Logic Source Destination Source_Clock Destination_Clock Path_Constraints Block_Masks Simulink_Names Vivado_Names Type
There are a total of two paths with violations in this timing analysis.
You can find the data fields in each timing path as shown in Example 3 in Additional Information.
print - Print timing path information
- Syntax
-
analyzer_object.print(<timing_path_structures>);
- Description
-
Prints timing data such as Slack, Path Delay, Levels of Logic, Name of Source and Destination blocks, Source and Destination clocks, Path Constraints, etc. for the input timing path structure.
The argument is an array of MATLAB structures with one or more elements.
- Examples
-
//Print timing path information for path #1 >> timing_object.print(all_timing_paths(1)) Path Num Slack (ns) Delay (ns) Levels of Logic Source/Destination Blocks Source Clock Destination Clock Path Constraints 1 -1.6430 11.5690 6 fixed_point_IIR/Delay1 clk clk create_clock -name clk -period 2 [get_ports clk] fixed_point_IIR/IIR Filter Subsystem/Delay4 ans = 1 //Print timing path information for path #3 >> timing_object.print(all_timing_paths(3)) Path Num Slack (ns) Delay (ns) Levels of Logic Source/Destination Blocks Source Clock Destination Clock Path Constraints 1 1.1320 0.5270 0 fixed_point_IIR/Delay1 clk clk create_clock -name clk -period 2 [get_ports clk] fixed_point_IIR/Delay1 ans = 1 //Print timing path information for path #2 from violating_paths array >> timing_obj.print(violating_paths(2)) Path Num Slack (ns) Delay (ns) Levels of Logic Source/Destination Blocks Source Clock Destination Clock Path Constraints 1 -1.3260 11.2520 6 fixed_point_IIR/Delay1 clk clk create_clock -name clk -period 2 [get_ports clk] fixed_point_IIR/Delay2 ans = 1
highlight - Highlight design blocks for a timing path
- Syntax
-
analyzer_object.highlight(<timing_path_structure>);
- Description
-
This API highlights System Generator blocks for the timing path passed in the argument. It doesn't change the highlighting of a block from other paths, so more than one timing path can be highlighted if you use this function repeatedly.
The argument is the MATLAB structure for one timing path.
- Example
-
//Highlight Simulink model blocks in the selected path //Don't change highlighting of currently highlighted blocks in the model >> [result, err_msg] = timing_object.highlight(all_timing_paths(1));
Highlighted System Generator model blocks appear as shown below.
Figure 2. System Generator Model Blocks
highlightOnePath - Highlight design blocks for one timing path
- Syntax
-
analyzer_object.highlightOnePath(<timing_path_structure>);
- Description
-
This API highlights System Generator blocks for the timing path passed in the argument. If a block from other paths is already highlighted then it will be unhighlighted first, so only one path is highlighted at a time.
The argument is the MATLAB structure for one timing path.
- Example
-
//Highlight a single path in System Generator model, and unhighlight currently //highlighted paths >> [result, err_msg] = timing_object.highlightOnePath(violating_paths(2));
unhighlight - Unhighlight design blocks
- Syntax
-
analyzer_object.unhighlight();
- Description
-
This API unhighlights blocks that are already highlighted. The blocks in System Generator model are displayed in their original colors.
- Example
-
//Unhighlight any Simulink block that is currenly highlighted >> [result, err_msg] = timing_object.unhighlight();
disp - Display summary of timing analysis
- Syntax
-
analyzer_object.disp();
- Description
-
This API displays the summary of timing paths on the MATLAB console, including the worst slack value.
- Example
-
//Display a summary of timing analysis >> timing_object.disp() Number of setup paths = 9 Worst case setup slack = -1.6430
delete - Delete xilinx.analyzer class object
- Syntax
-
analyzer_object.delete();
- Description
-
This is a destructor for the xilinx.analyzer class.
- Example
-
//Delete xilinx.analyzer object, i.e., timing_object >> timing_object.delete();
Additional Information
Accessing data fields of timing path structures:
- Example 1: Data for timing path #1
-
//Return the data fields for the timing path with the worst slack >> all_timing_paths(1) ans = Slack: -1.6430 Delay: 11.5690 Levels_of_Logic: 6 Source: 'fixed_point_IIR/Delay1' Destination: 'fixed_point_IIR/IIR Filter Subsystem/Delay4' Source_Clock: 'clk' Destination_Clock: 'clk' Path_Constraints: 'create_clock -name clk -period 2 [get_ports ...' Block_Masks: {1x5 cell} Simulink_Names: {1x5 cell} Vivado_Names: {1x5 cell} Type: 'setup'
- Example 2: Data for timing path #3
-
//Return the data fields for a timing path >> all_timing_paths(3) ans = Slack: 1.1320 Delay: 0.5270 Levels_of_Logic: 0 Source: 'fixed_point_IIR/Delay1' Destination: 'fixed_point_IIR/Delay1' Source_Clock: 'clk' Destination_Clock: 'clk' Path_Constraints: 'create_clock -name clk -period 2 [get_ports ...' Block_Masks: {'fprintf('','COMMENT: begin icon graphics')...'} Simulink_Names: {'fixed_point_IIR/Delay1'} Vivado_Names: {'fixed_point_iir.fixed_point_iir_struct.delay1'}
Type: 'setup'
- Example 3: Data for path #1 in violating_paths array
-
//Return the data fields in a timing path with timing violations >> violating_paths(1) ans = Slack: -1.6430 Delay: 11.5690 Levels_of_Logic: 6 Source: 'fixed_point_IIR/Delay1' Destination: 'fixed_point_IIR/IIR Filter Subsystem/Delay4' Source_Clock: 'clk' Destination_Clock: 'clk' Path_Constraints: 'create_clock -name clk -period 2 [get_ports ...' Block_Masks: {1x5 cell} Simulink_Names: {1x5 cell} Vivado_Names: {1x5 cell} Type: 'setup'