-
Notifications
You must be signed in to change notification settings - Fork 56
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
AlertLogPkg: Make it possible so that on ReportAlerts, a vacuous pass results in failure. #44
Comments
Ideally, it would be possible to set Vacuous Pass failure per AlertID/Scoreboard. I have Scoreboards that I expect to actually check things, and sometimes scoreboards exist in VIP modules that are not used in a specific test. (eg. I have a AXI4 VIP module that has checkers for AWADDR, WDATA and ARADDR, and a particular testbench may only be performing writes to the AXI4 i/f so ARADDR is unused). My current implementation requires a finish signal to mark the end of a test:
|
@SkydiverTricky If we add G_ALLOW_VACUOUS_PASS for scoreboards, why would we not make it a general thing for each AlertLogID? If we invert the check, it then amounts to a coverage goal on an AlertLogID basis. I have also been working on requirements tracking. It is interesting and challenging as Still brainstorming that requirements tracking. |
Having affirmation goals per ID would certainly be useful. As I said here in a test where multiple checkers are involved, having the test report a pass when one channel locked up and didnt meet an Affirmation goal could currently lead a CI system to pass - potentially leading to confusion in the delivered system. |
If you are talking about goals - it might be useful to be able to specify specific numbers and ranges. I can see all of the following being useful goals:
|
Given that you can do this, why add something inside. If we add something internal, you would loose the insight as to what the value means. Said another way, abstraction is cool - except when it obfuscates things. |
Note to self: If the alert is configured to have a requirement goal or set up as a requirement, then there are settings that will allow a requirement that is not met to be treated as a test failure. I expect to be doing another pass through requirement tracking later in 2022. I will look at this in more detail then. |
Currently, if you call ReportAlerts, if no affirmations were checked then a test always passes. In a testbench where the UUT has locked up and the testbench has timed out, this may give a false sense of security in a CI environment.
Currently I have the following check in my generic OSVVM control entity:
Would it be nicer to pass the required affirm count to the ReportAlerts method, or some machinism to turn a Vacuous pass into a failure within OSVVM itself?
The text was updated successfully, but these errors were encountered: