Max Horn

Adjust to jl_boundp change in Julia nightly

1

Companion to https://github.com/gap-system/gap/pull/5767

CI: stop testing for invalidations

1

This test is unreliable, often misleading, and we mostly ignore it anyway. See also https://github.com/julia-actions/julia-invalidations/issues/19

Provide a stale_age to mkpidlock

3

Normally the pidlock should be held for just a few milliseconds; waiting for its age to exceed 10 seconds until we consider it stale should be very safe. In fact...

CI: test against Julia 1.11-rc1

1

... and stop testing against 1.9

Error handling incomplete / clashes with GAP's Test/TestDirectory, leaves GAP in bad state, subsequent crash

3

Tried to run the GAP test suite in Julia. This failed because it includes testing errors and we hijack them (so the code in GAP's `Test` which also "hijacks" error...

Doing this: ``` using GAP GAP.prompt() # now on the GAP prompt Test("SOMEGAPROOT/tst/testinstall/kernel/streams.tst"); CollectGarbage(true); # force a GC from GAP ``` results in a crash for me, always: ``` [35914]...

... in order to pass the GAP test suite: ``` ########> Diff in /Users/mhorn/Projekte/GAP/gap.spielwiese/tst/testinstall/kernel/streams.tst:293 # Input is: ExecuteProcess("","",0,fail,fail); # Expected output: Error, ExecuteProcess: must be a small integer (not the...

Add tests involving Julia tasks

2

In particular, add such tests which stress the garbage collector, as it is not quite clear how well our current task stack scanning code works (or if it indeed even...

When starting multiple Julia's and using GAP.jl in each of them, there can be issues caused by the shared faux gaproot in a scratch space, despite pidlocking. To quote myself...

This test is unreliable, often misleading, and we mostly ignore it anyway. See also https://github.com/julia-actions/julia-invalidations/issues/19