Sorry I forgot to pull from sf before creating the patch
I'm going to refactor based on your hook-support commit(
https://github.com/ThomasAdam/tmux/commit/ea2f662950f0bc12eb673c2cbd946f2579c62c7f)
ok?
About the SIMPLEQ, I used it because I only needed a simple stack/queue but
I will change it to TAILQ which sould be a drop-in replacement.
On Tue, Feb 26, 2013 at 5:04 AM, Thomas Adam <tho...@xteddy.org> wrote:
> On Mon, Feb 25, 2013 at 09:27:24PM -0300, Thiago Padilha wrote:
> > Command that provides features for all synchronization needs in tmux
> scripts.
>
> I can't apply this with git-apply.
>
> > 1- When unlocking waiting panes, I guess it should be done in the
> > event loop, however I was getting random crashes when doing that, so
>
> Yes it should. I'd have to see what you tried to say why it was crashing.
>
> > now everything is done synchronously inside the command.
> > 2- It must take action when a pane that is holding the monitor exits,
> > but I couldn't figure out how to do it
>
> In the server loop.
>
> > +enum monitor_options {
>
> This should go in tmux.h
>
> Note also in the general case you need to read OpenBSD style(9):
>
>
> http://www.openbsd.org/cgi-bin/man.cgi?query=style&apropos=0&sektion=0&manpath=OpenBSD+Current&arch=i386&format=html
>
> No C99 comments, no static functions, etc.
>
> > + MONITOR_ENQUEUE = 1,
> > + // if this is set, as soon as the pane acquires the lock, it will
> > + // temporarily give up the lock rights and be put on the wait
> stack until
> > + // a signal is received, at which point it will be put back in
> front
> > + // of the lock queue again
> > + MONITOR_WAIT = 2,
> > + MONITOR_SIGNAL = 4,
> > + // if the pane is not holding the lock when it starts waiting for
> a signal,
> > + // then this flag is set, which means that as soon as the the
> signal
> > + // arrives, it will leave the wait stack and release the lock
> > + MONITOR_RELEASE = 8
> > +};
> > +
> > +// queue of panes, used for tracking panes waiting to lock the monitor,
> or to
> > +// receive a signal from another pane
> > +struct cmdq_node {
>
> Again; in tmux.h, but why isn't this factored in to struct window_pane?
>
> > +
> > +// name, and the monitor node only exists while there's some pane using
> it
> > +struct monitor_node {
>
> This in tmux.h
>
> > +
> > + // monitor name
> > + char *name;
> > +
> > + // client/pane id
> > + int pane_id;
> > +
> > + // number of times the pane has locked the monitor.
> > + int count;
> > +
> > + SIMPLEQ_HEAD(, cmdq_node) lock_queue; // panes waiting to lock the
> monitor
>
> We use either TAILQ or SLIST, RB_TREE, etc. Not SIMPLEQ_HEAD.
>
> Until I can mess around with this, I've superficial comments only.
>
> -- Thomas Adam
>
------------------------------------------------------------------------------
Everyone hates slow websites. So do we.
Make your web apps faster with AppDynamics
Download AppDynamics Lite for free today:
http://p.sf.net/sfu/appdyn_d2d_feb
_______________________________________________
tmux-users mailing list
tmux-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/tmux-users
