spurious failures with OpenBSD 7.4 on xhyve

[manxorist]

I am seeing spurious failures due to timeout when executing consecutive commands in multiple steps in an OpenBSD 7.4 VM:

Run cross-platform-actions/action@v0.22.0
  with:
    architecture: x86_64
    hypervisor: xhyve
    memory: 4G
    operating_system: openbsd
    version: 7.4
    shell: bash
    sync_files: false
    shutdown_vm: false
    run: gmake -j$(sysctl -n hw.ncpu) STRICT=1 VERBOSE=1 AUTO_DEPS=1 check
  
Setting up VM
  VM is ready
Running command
  gmake -j$(sysctl -n hw.ncpu) STRICT=1 VERBOSE=1 AUTO_DEPS=1 check
/usr/bin/ssh -t runner@cross_platform_actions_host sh -c 'cd "/Users/runner/work/openmpt/openmpt" && exec "bash" -e'
Pseudo-terminal will not be allocated because stdin is not a terminal.
Error: The operation was canceled.

example run: https://github.com/OpenMPT/openmpt/actions/runs/7376167362/job/20068624263

workflow file: https://github.com/OpenMPT/openmpt/blob/91abfddfd96e8d8a3edee28b4f92c31e7eaa70df/.github/workflows/OpenBSD-7.4-Makefile.yml

Failure rate appears to be 1 failure in 5..10 runs (we only had about 15 runs since switching to multi-step runs).

I cannot compare to QEMU because of #73.

[jacob-carlborg]

I'm suspecting that it's the same problem that persists that has been reported before: #72 and #61. The root cause has not been fixed.

[manxorist]

After #47, I do not have any use case for this any more. I am running OpenBSD on QEMU on Ubuntu now, which appears to work fine so far.