I was really only trying to get github CI to check the build on
windows(hence the PR title) - I don't know if the win pthread
implementation supports the PTHREAD_MUTEX_RECURSIVE attribute type.
Anyway, if it compiles it should work at least as well as the current
code, except for one big gotcha - if somewhere deep in the bowels of
hamlib I/O error handling tries to do rig_close()/rig_open() for
recovery, all bets are off. rig_close() will try to terminate all
threads, some of which may be waiting on a mutex. That's not a new
problem, but this code may make the problem more evident.
On 2/28/25 07:14, Michael Black via Hamlib-developer wrote:
> Branch: refs/heads/master
> Home: https://github.com/Hamlib/Hamlib
> Commit: 439bd96b5d01ddbaf811a1e1df98c08a1caff406
> https://github.com/Hamlib/Hamlib/commit/439bd96b5d01ddbaf811a1e1df98c08a1caff406
> Author: George Baltz N3GB <Geo...@gm...>
> Date: 2025-02-27 (Thu, 27 Feb 2025)
>
> Changed paths:
> M src/rig.c
>
> Log Message:
> -----------
> Fix the easy parts of missing thread locks
>
> - Fix depth initialization so LOCK() at least calls rig_lock()
> - Add message showing why rig_lock() does nothing
> - Add missing LOCK(0)s
>
>
> Commit: 0147da5a3ca8dc0f874d2a43ff5a2e702bbc4f41
> https://github.com/Hamlib/Hamlib/commit/0147da5a3ca8dc0f874d2a43ff5a2e702bbc4f41
> Author: George Baltz N3GB <Geo...@gm...>
> Date: 2025-02-28 (Fri, 28 Feb 2025)
>
> Changed paths:
> M include/hamlib/rig.h
> M src/rig.c
>
> Log Message:
> -----------
> Replace rs->multicast->mutex with a dedicated mutex for the API.
>
> Use PTHREAD_MUTEX_RECURSIVE and change LOCK() to always call rig_lock().
> Still needs a reliable way to determine availability.
>
>
> Commit: f2f947c51d45f5596afd0b8b3c8df7c3d88d93dd
> https://github.com/Hamlib/Hamlib/commit/f2f947c51d45f5596afd0b8b3c8df7c3d88d93dd
> Author: Michael Black <mdb...@ya...>
> Date: 2025-02-28 (Fri, 28 Feb 2025)
>
> Changed paths:
> M include/hamlib/rig.h
> M src/rig.c
>
> Log Message:
> -----------
> Merge pull request #1675 from GeoBaltz/fix26
>
> DO NOT MERGE!
>
>
> Compare: https://github.com/Hamlib/Hamlib/compare/3d46647808dc...f2f947c51d45
>
> To unsubscribe from these emails, change your notification settings at https://github.com/Hamlib/Hamlib/settings/notifications
>
>
> _______________________________________________
> Hamlib-developer mailing list
> Ham...@li...
> https://lists.sourceforge.net/lists/listinfo/hamlib-developer
|
