Re: Automation of parsing of fstests xunit xml to kicdb kernel-ci
Gustavo Padovan <[email protected]> Fri, 18 Apr 2025 11:54:26 -0300
| Newsgroups | dev.linux.lists.kernelci,dev.linux.lists.kdevops,org.kernel.vger.fstests,org.kernel.vger.linux-fsdevel |
|---|---|
| Message-ID | <[email protected]> |
---- On Thu, 17 Apr 2025 13:54:15 -0300 Luis Chamberlain <[email protected]> wrote --- > On Thu, Apr 17, 2025 at 09:18:57AM -0700, Luis Chamberlain wrote: > > We're at the point that we're going to start enablish automatic push > > of tests for a few filesystems with kdevops. We now have automatic > > collection of results, parsing of them, etc. And so the last step > > really, is to just send results out to kicdb [0]. > > > > Since we have the xml file, I figured I'd ask if anyone has already > > done the processing of this file to kicdb, because it would be easier > > to share the same code rather than re-invent. We then just need to > > describe the source, kdevops, version, etc. > > > > If no one has done this yet, we can give it a shot and we can post > > here the code once ready. > > > > [0] https://docs.kernelci.org/kcidb/submitter_guide/ From KernelCI side, we are not aware of anyone who created that xml->kcidb processing. It is a good time to discuss that though as we are making some fundamental changes in KCIDB to improve performance and allow it to be more agile in how we deal with the various needs we are seeing from the community. In a nutshell, we are creating a new interface to publish the data to KCIDB and also allow easier local dev env for people to try it out. Thinking about your need here, maybe we could go as far as having an endpoint in the new KCIDB api that consumes the xml directly and do the translation internally. We should also look if the schema is enough for you or new additions may be needed. For context, for those not aware with the new KernelCI infra. KCIDB data is the backend of our https://dashboard.kernelci.org/ and https://kci.dev/ cmdline. Mind you that both project are quite new and will grow more and more functionality over the years. We'd be happy to collect your usecases and see how we can evolve our infra, dashboard, etc. Best, - Gus