How to write XTA command scripts

• version

  1.0.1

• scope

  Example.

  This code is provided as example code for a user to base their code on.

• description

  How to write XTA command scripts

• boards

  Unless otherwise specified, this example runs on the SliceKIT Core Board, but can easily be run on any XMOS device by using a different XN file.

The XTA can be used interactively to investigate and measure the timing properties of you program. Once satisfied, it is often a requirement to capture these timing requirements in a script thus allowing the timing of your program to be automatically re-validated on future modifications. This can be achieved by placing XTA commands into a script file.

For example, compile the following code:

#include <stdlib.h>
#include <xs1.h>

port p1 = XS1_PORT_1A;
port p2 = XS1_PORT_1B;

int main() {
  int x;

  #pragma xta endpoint "input"
  p1 :> x;

  // Checks for errors..
  if (x == 1) {
    #pragma xta label "error_case"
    exit(1);
  }

  // do some computation here..

  #pragma xta endpoint "output"
  p2 <: 0;
  return 0;
}

Assume that there is a timing requirement between the input and the output of 100.0 ns. Assume also assume that you are not interested in the timing of the error_case.

Add the following lines to a file script.xta:

load a.xe
analyze endpoints input output
set exclusion - error_case
set required - 100.0 ns
print summary
exit

Note: In the above script, the ‘-‘ refers to the ID of the most recently created route. Using this construct can result in more maintainable scripts.

On the command line, this script can then be executed:

xta source script.xta exit

Which will display the following:

Route(0) endpoints: input to output
    Pass with 1 unknown, Num Paths: 1, Slack: 20.0 ns, Required: 100.0 ns, Worst: 80.0 ns, Min Core Frequency: 320 MHz

Pass, Min Core Frequency: 320 MHz