Intel® Stratix® 10 Hard Processor System Remote System Update User Guide

ID 683021
Date 2/23/2024
Public
Document Table of Contents

2.12. Retrying when Configuration Fails

The max retry option, introduced in Intel® Quartus® Prime Pro Edition software version 19.3, allows you the ability to configure the number of times both the application images and the factory image are tried again, if any configuration failure occurs, such as:
  • Flash corruptions
  • Authentication and encryption errors
  • RSU watchdog timeouts

The max retry option is selected in Intel® Quartus® Prime for the factory image project and is stored in the decision firmware data structure in flash. The default value for max retry is 1, which means each image is tried only once, the standard behavior before Intel® Quartus® Prime software version 19.3. The maximum value for max retry is 3, which means each image is tried up to three times.

Note: Although the parameter is called max retry, its value actually denotes the total number of times each image is tried.

You can query the value of the max retry parameter from both U-Boot and LIBRSU. You must first query it in U-Boot for the value to also be available in Linux for LIBRSU.

The SDM maintains the retry counter, which counts the number of times the current image has failed to configure. Once that counter reaches max retry, the next image is loaded. You can query the value of the retry counter and also request it to be reset to zero from both U-Boot and LIBRSU.

Both the decision firmware and the currently running application or factory image firmware must be generated with Intel® Quartus® Prime Programmer software version 19.3 or above so that all of the max retry functionality is available.
Note: If only the decision firmware version is 19.3 or above, then the images are tried the specified number of times, but querying the retry counter or resetting it to zero is not available.
The retry behavior applies to all configuration failures, including:
  • Failure to initially configure images after a power up or nCONFIG event.
  • Failure of an image after it was configured, due to an HPS watchdog timeout, when the watchdog was configured to trigger an RSU failure.
  • Failure to load a specific image, applying to all possible failures: requested image failure, or failure of next images that are tried in case the requested image failed.

For examples of using the max retry option from both U-Boot and Linux, refer to the Remote System Update Example section.