Fix: replace guard with cek-try in test runner, clear stale reuse_stack

The guard form (call/cc + handler-bind expansion) doesn't survive async
IO suspension — the CEK continuation from guard's call/cc captures frames
that become invalid after the VM resumes from hs-wait. Replacing guard
with cek-try (which compiles to VM-native OP_PUSH_HANDLER/OP_POP_HANDLER)
avoids the CEK boundary crossing.

The test runner now executes: suspends on hs-wait, resumes, runs test
actions, and test assertions fire correctly. The "Not callable: nil"
error is eliminated. Remaining: test assertion errors from iframe content
not loading fast enough (timing issue, not a framework bug).

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
This commit is contained in:
2026-04-17 18:10:10 +00:00
parent d8fec1305b
commit b12ec746a2
4 changed files with 14 additions and 15 deletions

View File

@@ -854,11 +854,12 @@ let resume_vm vm result =
(try run vm
with VmSuspended _ as e ->
(* Re-suspension during resume: the VM hit another perform.
Clear reuse_stack — these entries are stale from the PREVIOUS
suspension and don't apply to the current VM frame state.
The new VmSuspended carries the current VM state correctly. *)
vm.reuse_stack <- [];
The new VmSuspended carries the current VM state. *)
raise e);
(* Clear reuse_stack — any entries here are stale from the original
suspension and don't apply to the current state. The VM just
completed its execution successfully. *)
vm.reuse_stack <- [];
(* Restore call_closure_reuse continuations saved during suspension.
reuse_stack is in catch order (outermost first from prepend)
reverse to get innermost first, matching callback→caller unwinding. *)