Hi!
The first commit adds a test that fails on current master due to fibers
not getting scheduled again when their port gets closed. This is fixed
by the second commit. I am not sure if we could still run into problems
with the hashv-set! when the scheduled task gets stolen from another
scheduler by steal-work!.
Hi!
The first commit adds a test that fails on current master due to fibers
not getting scheduled again when their port gets closed. This is fixed
by the second commit. I am not sure if we could still run into problems
with the `hashv-set!` when the scheduled task gets stolen from another
scheduler by `steal-work!`.