On Mon, Oct 31, 2022 at 10:40:25AM -0700, Brian Norris wrote:
On Sun, Oct 30, 2022 at 10:51 PM kernelci.org bot bot@kernelci.org wrote:
chrome-platform/for-kernelci baseline: 98 runs, 5 regressions (v6.1-rc1-5-g27b86a65cd16)
Does anybody look at these? It's a bit weird to see "5 regressions", and then to look back on the last few weeks (probably months) and see the same errors...
Nobody seems to care about a lot of the Chromebook stuff AFAICT.
Details: https://kernelci.org/test/job/chrome-platform/branch/for-kernelci/kernel/v6....
Where can I find the test cases? (i.e., what's determining a "failure" for one of these?)
The first place to go is the details link above, though it's a bit less clear for the non-baseline tests you should at least be able to see what's going on on the console which will often help with finding the testsuite, or the error messages printed are descriptive as to what they were looking for specifically. The kernelci-core repo contains all the stuff to map from a test name inot running something (often with external assistance) but there's lots of templating in the way.
As a small guess, I see that we're missing some common configs, like:
CONFIG_PHY_ROCKCHIP_DP
This sometimes means that the display subsystem (and therefore audio subsystem that relies on DP for one of its components) is not going to fully set itself up properly. Usually things can still sort of work, but I don't really know what the tests are looking for.
The baseline tests are just making sure that the system comes up to a shell. The driver loaded tests are checking that particular devices have a driver bound to them.