John Högberg

I just have to ask you folks (cc @ThomasArts), have these benchmarks been used as target for performance improvements for a long time? If so I fear there's not much...

Currently just `min_bin_vheap_size`, but we might add options to change the per-process growth curve etc in the future.

> The leveled perf_SUITE test we're using was only recently created, we've been using this test just for the past few months. > > Most of the resultant optimisations have...

Huge thanks for putting that together, I’ll remove the vheap curve stuff from #8347 and focus more on adding flags to tweak it in the near term. 🙂

Given that the random number is used to pick a random pivot element in quicksort, performance is expected to vary as a result of this change, and the consistently worse...

Thanks for bringing this up! We don't think we'll have the time to have a deep look at it before OTP 25, so we'll try to do it for OTP...

Thanks for the PR! I'll add it to testing once the patch is released, and merge to `maint` and `master` shortly afterwards.

> @sverker, @jhogberg and @IngelaAndin: as it says in the commit message, the patch can only go to the 26 branch, and should not be propagated to maint as it...

Thanks for your report! We do not consider this a bug, when you ask `dialyzer` to go beyond its capabilities (e.g. by specifying `-Wmissing_return` or similar options), it may report...

Unfortunately there's not much we can do about this other than to have the JIT disabled by default on that configuration, and/or force people to [(partially) disable the hardened runtime](https://developer.apple.com/documentation/bundleresources/entitlements/com_apple_security_cs_allow-jit)....