End of test (EOT), is one of the most important and tricky parts of a test bench. EOT is important because it qualifies if a test has really passed or failed. Thus it potentially impacts all tests. However, EOT is tricky because the implementation is spread across all the components of a test bench. EOT has become more complicated due to self-checking nature of coverage driven constrained random verification environments.
Whenever EOT is discussed, emphasis is laid on the verification methodology’s support for “end of test” implementation. This is not sufficient. One needs to also understand what objective is the EOT check attempting to attain. This blog focuses on describing these objectives of the EOT.
In order to declare “end of test”, the following two objectives have to be met:
- End of Stimulus Generation: Ensure all specified stimulus has been generated
- End of Stimulus Execution: Ensure all generated stimulus has been executed or processed
In this blog we describe these two objectives in detail.