On 2024-10-16 12:30, PICCA Frederic-Emmanuel wrote: > Yes it is a test case for me and I will upload some real OpenCL test with > pyfai, silx and pyvkfft > thanks a lot for the rocm ci infra :))
Great. FYI, we have a helper script coming that should reduce writing tests to just a very simple Test-Command in d/test/control. It just needs to pass its final tests, and then clear NEW once uploaded. It might be simpler/less work for you to hold off on this until then. >> Also, since there is no direct dependency on ROCm packages, our CI will >> never trigger for clpeak. I will need to adapt our scheduler to account >> for opencl-icd. > > the dependencies are only in the test. This is not enough ? No, that is sufficient, I just misremembered something here. Our scheduler currently only triggers on d/control Depends, not d/tests/control Depends, but that is just a limitation of the scheduler's implementation. I remember from a related issue that I performed deeper analysis and ran into some show-stopper [4], though the finer details currently escape me. I'll have to look into it again. Best, Christian [3]: https://salsa.debian.org/rocm-team/rocm-dev-tools/-/issues/17 [4]: https://salsa.debian.org/rocm-team/rocm-dev-tools/-/issues/15#note_480293
