Jan Midtgaard

Rebased and updated - to reflect the macOS ARM64 vs Intel workflow split - to add scheduled runs on these new workflows

Rebased and updated - to reflect the new GitHub actions setup and - adding a 5.4.0 workflow too This is now limited to just Linux bytecode debug. We have seen...

Note to self: Like the native debug workflows, this reuses the cache entries for the non-debug workflows, as both debug and non-debug can be run from the same compiler build....

FTR, these are the times for the bytecode debug workflows: - Bytecode 5.2 debug / Successful in 180m - Bytecode 5.3 debug / Successful in 182m - Bytecode trunk debug...

Just observed this again - now on FreeBSD 5.1 trying out the latest QCheck memory usage improvement https://github.com/ocaml-multicore/multicoretests/tree/try-qcheck-mem-improvement https://ocaml-multicoretests.ci.dev:8100/job/2024-06-28/125000-ci-ocluster-build-caa2d4 ``` random seed: 362318939 generated error fail pass / total time...

I've created a standalone reproducer for this one: ```ocaml let test () = let path = Filename.temp_file "stm-" "" in let channel = Atomic.make (Out_channel.open_text path) in (* First, a...

The branch https://github.com/ocaml-multicore/multicoretests/tree/flushtest-repro-focus triggers a focused test of it on the CI. Across all runs, this reproducer only triggered in multicoretests-ci where `flush` may trigger a `Sys_error` in - linux-arm64-5.1...

In trying to understand this better, I ran `dune exec src/io/stm_tests.exe -- -v` locally under a 5.2.0+fp switch, where it fails spectacularly - even the sequential one: ``` $ dune...

To summarize - the Cygwin `5.2.0` failure was actually in `5.2.0~alpha1`. It hasn't shown up since #12678 was merged (into 5.2.0~beta1) - the FreeBSD 5.1 failure was in `5.1.1` -...

Thanks for sharing! :pray: Your explanation makes sense to me. I can't reproduce the error ATM though... I'll need to look more into it.