Gerwin Klein

In https://github.com/seL4/seL4/pull/1109#issuecomment-1749063200 a new failure mode for MULTICORE0004 popped up, triggering an assertion instead: ``` Running test MULTICORE0004 (Test core stalling is behaving properly (flaky)) seL4 failed assertion 'isRunnable(NODE_STATE(ksCurThread))' at...

Entirely possibly that it is the same problem (that would be great), I just hadn't seen this failure mode yet.

Adding as an observation here that I have seen this most frequently in the combination SMP+MCS+clang in the last few months.

Ok, so some more bizarre observations. - The [seL4 commit c28e52a1ec](https://github.com/seL4/seL4/commit/c28e52a1ece2020723d1857c893e75897559563e) made the test MULTICORE0004 hang almost reliably on the `tqma` board in the config `SMP_MCS_release_clang` (and only that config,...

Happy to try that: https://github.com/seL4/ci-actions/pull/326

@wom-bat do you know if `skylake` has any USB devices attached that would be causing this?

Peter says it might be the keyboard/mouse that is plugged in.

To summarise some of the discussion from the seL4 dev hangout on this: the proposed testing method for this is to run the normal C sel4test, but to use Rust...

> Could this be something from lone time ago when SMP support was added. So checking `CONFIG_MAX_NUM_NODES' implied checking SMP support? It's possible that this is very old and made...

I should note that I have checked all other instances of `config_set` now and they all refer to boolean values only, so these should really be the last ones unless...