@hardboprobot mentioned in an OOB discussion:
Currently, the data we keep about a test result includes the kernel version used for it, but for other types of tests where the userspace is a key ingredient (ChromeOS - Yocto - Android), we might want to also encode the ChromeOS userspace version used as part of the test results. Does it makes sense?
Retrieving and storing such information would probably make sense (for ChromeOS it could be the content of /etc/os-release -- which is already cat'ed in LAVA jobs -- for example).
This needs to be carefully designed as it involves multiple components (e.g. making this information a test result in LAVA) and should be generic enough to be used across various userspaces (Android likely has different way of storing OS-related information).
@hardboprobot mentioned in an OOB discussion:
Retrieving and storing such information would probably make sense (for ChromeOS it could be the content of
/etc/os-release-- which is alreadycat'ed in LAVA jobs -- for example).This needs to be carefully designed as it involves multiple components (e.g. making this information a test result in LAVA) and should be generic enough to be used across various userspaces (Android likely has different way of storing OS-related information).