Re: C level mutices and condition variables

[Neil Jerram <neil@ossau.uklinux.net>]

>>>>> "Marius" == Marius Vollmer <mvo@zagadka.ping.de> writes:

    Marius> Hi,
    Marius> currently, we support the scm_t_mutex and scm_t_cond types, along with
    Marius> some functions.  You can declare a variable of type scm_t_mutex and
    Marius> initialize it with scm_mute_init, for example.

    Marius> This makes it hard to have SCM data in a scm_t_mutex since you have to
    Marius> protect it without the help from the user of the mutex.  Also, this is
    Marius> not 'astract' enough when we make thread package support a run-time
    Marius> option (which we should).

    Marius> Dirk did some work in this direction in the dirk-thread-factorization
    Marius> branch.  I propose something simpler:

    Marius> Let's just remove C level support for mutex and condition variable
    Marius> objects and tell them to use the Scheme level mutices, which are also
    Marius> available to C.  (With deprecation.)

    Marius> Ok?

I don't understand, I'm afraid.

- Why would you want "to have SCM data in a scm_t_mutex"? - isn't the
  data that you are protecting access to normally a separate variable
  (or set of variables) from the mutex itself?

- What do you mean by not abstract enough?

- In "tell them to use the Scheme level mutices", who is "them", and
  would it be possible for them to use Scheme mutexes from C code, or
  are you saying that people who need mutexes must code in Scheme?

Regards,
        Neil



_______________________________________________
Guile-devel mailing list
Guile-devel@gnu.org
http://mail.gnu.org/mailman/listinfo/guile-devel