Fix data race in nestkernel/per_thread_bool_indicator.cpp

[jprotze]

No description provided.

[heplesser]

@jprotze Thank you for the PR. I just have a few suggestions mainly related to NEST coding conventions.

[heplesser]

@ackurth Could you benchmark this branch on a single node with microcircuit, up to 128 threads? Just to check that the change does not introduce a performance regression.

[ackurth]

@ackurth Could you benchmark this branch on a single node with microcircuit, up to 128 threads? Just to check that the change does not introduce a performance regression.

Will do.

[heplesser]

@jprotze This looks good to me now. I still wait for results from systematic benchmarks on larger models (first checks look good) and will wait for final approval until then.

[ackurth]

@ackurth Could you benchmark this branch on a single node with microcircuit, up to 128 threads? Just to check that the change does not introduce a performance regression.

I compared the suggested changes to the current NEST Master, there is no performance regression in our usual benchmarks of the microcircuit model.

[heplesser]

@ackurth Could you benchmark this branch on a single node with microcircuit, up to 128 threads? Just to check that the change does not introduce a performance regression.

I compared the suggested changes to the current NEST Master, there is no performance regression in our usual benchmarks of the microcircuit model.

Excellent, then we can proceed. Thanks!

[heplesser]

With positive benchmark results, I approve.

[suku248]

Thanks @jprotze for this PR and @ackurth for benchmarking!