Aurora Perego

The performance does not change: | nThreads | base[@c9dbac0](https://github.com/AuroraPerego/alpaka/tree/c9dbac0e8f347a00c6bb93861d0c38bf191fcc8c) | base + PR | |-------|-------|-------| | 1 | 148.44 +/- 3.48) | 148.32 +/- 0.35) | | 2 | 172.73...

FYI in debug mode with the CUDA back-end I see: ```bash 93% tests passed, 2 tests failed out of 30 Total Test time (real) = 316.42 sec The following tests...

Good! The tests that fail are due to the OpenCL CPU runtime :(

Something is broken in the runtime. I do not think is a bug in alpaka because all the tests pass with an older runtime (see #2140).

As I wrote in the README, the only problem is that there is a bug in `all_of_group`/`any_of_group` methods (that does not affect the results of the tests). For that we...

Yes, i've built and run the tests for CPU and GPU with the usual result (all pass on GPU, 4 out of resources on CPU)

(I'll check also the impact on the performance)

The situation is worse than expected, apparently there are many more synchronization introduced with these changes. I compared the baseline with this PR and with another version that has only...

Yes, sorry. The tests have been done on pixeltrack-standalone[@dcf2898](https://github.com/AuroraPerego/pixeltrack-standalone/tree/dcf2898da7d66d89fdff2607d016e03c1671f1c3). These numbers are the throughput, i.e. the number of concurrent events processed per second, computed processing 10000 events and `nThreads` is...

Not sure the handling it's correct, it was done mostly to remove the `ext_oneapi_submit_barrier()` and `ext_oneapi_empty()` calls to check the impact. At this point I'm not sure if it's worth...