Gerwin Klein

I'm not feeling too good about this one. The comment is good and should stay in, but I don't think the inlining improves readability (it does make it worse for...

> I see your point. Inlining makes this a bit more porting/experimenting friendly, as you have more context available there for debug messages. And you can abort the whole loop...

> Some TX1, TX2 and PC99 tests can't find seL4 images. I enabled some more configs after the build had already started and it looks like the hardware-run matrix picked...

Ok, so after a few re-runs, failures for these boards seem to be real: `IMX8MQ_EVK`, `IMX8MM_EVK`, `TX2`, `ODROID_XU4` I agree that the cache flush failure for Odroid doesn't seem to...

I'll kick off another run which should now exclude the failing platforms. https://github.com/seL4/seL4/actions/runs/3181453603/attempts/2 was the run with the failing tests.

Ok, with the current exclusions in seL4/ci-actions#224 the hardware tests are now all succeeding. If there are no further concerns, I'd close the PR here and merge the one in...

This seems to be working now and the ci-actions PR has been merged.

@ssrg-bamboo `test`

(rebased and fixed the sign-off typo)

When you say > Could not run fault benchmarks on RISCV for some reason. do you mean there weren't any or they were failing?