Jan Midtgaard

Overall, I can see benefit to the CI log anchors and would like to salvage them! I'm wondering whether a good solution would be to extend QCheck to output the...

Ah, this old brain is finally starting to understand :sweat_smile: That's why you need a "parent process" to catch crashes and print the anchor (rather than just crash and burn)....

BTW, thanks for the documentation link - much appreciated! :pray: Potentially QCheck could `::group` ... `::endgroup` a list of tests in a test suite in such a mode.

Thanks a bunch for sharing these! :pray: :smiley: I'm happy to see that we can trigger such failures with sufficiently high counts! ATM we are still moving things around and...

Note: We already have `Array` tests in https://github.com/jmid/multicoretests/blob/main/src/array/lin_tests_dsl.ml Because they test more `Array` functions they tend to find and report observed "sequential inconsistency" of these, rather than performing a focused...

I gave this a shot for `Lin` in #115 and then paused. Here is a recap of my understanding (if I recall correctly): It is relatively easy to run over...

That's indeed interesting that the `Semaphore` is faster than the "Atomic waiting loop" :+1: :thinking: I had a quick look: - When an exception is raised `mk_prop` does not increase...

Ah, that is indeed quite a difference! :open_mouth: i'm surprised by the number 36868 though! Because of the way `Util.repeat` is implemented it should stop early on the first failed...

There's also - [ ] `caml_sys_mkdir` in runtime/sys.c

Thanks a bunch for sharing this! :pray: Here's a CI summary as I was seeing several red lights: - 32bit 5.2 and 32bit trunk failed to compile with `0xdeadbeef` out...