Jan Kotas

cc @dotnet/samsung Do you see a problem with getting this deleted?

> Benchmark results on Intel I am not able to reproduce this regression on Dev Box Amd. The PR is ~7% faster compared to baseline on that machine.

#106578 has workaround that should fix the build break. We may still need to figure out the permanent fix as a follow up.

This was fixed in upstream zlibng by https://github.com/zlib-ng/zlib-ng/pull/1771 . We will pick up the proper upstream fix with the next zlibng update.

This is hit almost once a day.

> Is CI so bogged down that it takes 5 seconds to spin up a process and run this functionality? It is not unusual for process start to take this...

> It checked for success exit code before. So, buffers in process wrapper? We keep seeing reports of the redirected output being truncated mysteriously. For example, https://github.com/dotnet/runtime/issues/108395 filled a few...

This is hit multiple times per week currently.

@steveisok We have picked up intermittent crossgen crash on arm64 in `_beginInlining` method. It is causing a lot of noise (see duplicate issues). Could you please take a look?

I think it was likely introduced by the #104696 toolset update, so it is not easy to roll-back.