Re: 8sync error in procedure select

[Christopher Lemmer Webber]

Hi!  Hm... it might make sense to try switching to the "fibers" branch
of 8sync and see if that does better?

I guess I haven't worked on 8sync in some time, but really, the
intention was to switch it over to being on top of Fibers... Andy knew
more about building an event loop than I did ;P


Alexey Abramov writes:

> Hi,
>
> I am trying to test an actor I wrote with 8sync. I created a probe actor
> which sends messages to the main one. A problem I am having is that when
> I want to send a *cleanup* to the main (server) actor from the probe, I
> am getting the following backtrace:
>
> --8<---------------cut here---------------start------------->8---
> λ guile --debug -s tests/test.scm
> Backtrace:
> In ice-9/boot-9.scm:
>   1736:10 11 (with-exception-handler _ _ #:unwind? _ # _)
> In unknown file:
>           10 (apply-smob/0 #<thunk 7fac09a7b9a0>)
> In ice-9/boot-9.scm:
>     718:2  9 (call-with-prompt _ _ #<procedure default-prompt-handle…>)
> In ice-9/eval.scm:
>     619:8  8 (_ #(#(#<directory (guile-user) 7fac096a0f00>)))
> In ice-9/boot-9.scm:
>    2806:4  7 (save-module-excursion _)
>   4351:12  6 (_)
> In 8sync/actors.scm:
>     812:6  5 (run-hive #<<hive> 7fac08bdd0c0> _ #:cleanup _ # _)
> In ice-9/control.scm:
>     91:24  4 (call-with-escape-continuation _)
> In 8sync/agenda.scm:
>     569:6  3 (run-agenda #<<agenda> queue: (() . #f) prompt-tag: ("…> …)
>     470:7  2 (update-agenda-from-select! #<<agenda> queue: (() . #f)…>)
> In ice-9/boot-9.scm:
>   1731:15  1 (with-exception-handler #<procedure 7fac08bbd240 at ic…> …)
> In unknown file:
>            0 (select (#<closed: file 7fac07ff6460>) () () #f #f)
>
> ERROR: In procedure select:
> In procedure select: Wrong type argument in position 1: #<closed: file 
> 7fac07ff6460>
> --8<---------------cut here---------------end--------------->8---
>
> I have attached a simple snippet to reproduce the problem, and also a
> naive patch I did. I am not sure if it is me doing the shutdown wrong, or
> it's a bug.