aborting (or finishing) pending operations in GPGME

Werner Koch wk at gnupg.org
Tue Jan 15 16:58:02 CET 2002

On Tue, 15 Jan 2002 16:25:16 +0100, Marcus Brinkmann said:

> The stubs in posix-sema.c are empty.  This certainly needs to be fixed at

Given that most software is not MT (for good reasons) I won't like to
implement this unconditionally; it might be better to add a configure
flag and build 2 vesions of the library - gpgme-config can then be
used to figure out the needed one.

Or should we provide an additional object file to shortcut the locking

> some time, but it will take some effort to go through gpgme and to add
> proper locking.  That doesn't mean that you can't use it in a threaded

There are not many places which need to be changed.  gpgme_wait()
needs to be changed of course, the fail_on_pending_request macro and
the stuff which fires up the backend.

> environment, you just have to be very careful.  What you describe above is
> probably safe, but I didn't check (cancel just sets an int from zero to some
> non-zero value).

As long as an object is only used by one thread it should be save.
gpgme_cancel() should work from all threads.

Werner Koch        Omnis enim res, quae dando non deficit, dum habetur
g10 Code GmbH      et non datur, nondum habetur, quomodo habenda est.
Privacy Solutions                                        -- Augustinus

More information about the Gnupg-devel mailing list