Re: [Boost-users] [signals] Automatic connection management and MT

From: Frank Mori Hess (frank.hess_at_[hidden])
Date: 2007-02-01 15:34:08

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On Thursday 01 February 2007 11:24 am, Timmo Stange wrote:
> Frank Mori Hess wrote:

> > Thread-safe automatic disconnect from non-static member functions
> > seems like it would require some kind of extension to the language
> > itself, like a pre-destructor.
>
> There is also the possibility of using a wrapper template instead of
> a base class to put the tracking functionality to the top of the class
> hierarchy, but I don't think I particularly like that approach,
> especially as measures to "finalize" the hierarchy would have to be
> taken.

One solution is to overload the signal::connect function with another 
version that accepts an additional weak_ptr<void> argument, which points 
to the object whose member function is being used as a slot.  The signal 
could try to convert the weak_ptr into a shared_ptr<void> whenever it runs 
the slot.  Then the signal can either clean up the connection if the 
object is dead, or be sure the object won't die while the slot is running.  
The disadvantage, of course, is it forces people to use shared_ptrs to 
manage the lifetimes of objects they want to use thread-safe automatic 
connection management with.  But that seems preferable to offering no 
means of thread-safe automatic connection management at all.

> Thanks for the code, it looks very promising. I was hoping it could be
> done with a little less runtime overhead (per slot mutex + polymorphism
> + shared_ptr), but I have to investigate the available options some
> more.

I imagine it could be.  Optimization wasn't high on my priority list while 
writing it, I was aiming more for correctness, followed by ease of 
implementation.

> I have one small suggestion for now: You could perhaps move the
> connection mutex to the base class to avoid too much per-Signature
> code repetition (use a private virtual do_disconnect() for forwarding,
> for example).

I'm not following what you're suggesting.  Would you spell it out in more 
detail, or even better say it with a patch?

- -- 
Frank
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.1 (GNU/Linux)

iD8DBQFFwk7B5vihyNWuA4URAqa+AJ9b35Dk01ZrWKpqPxp0uzGp99DcBwCg04UB
kIA0X8mCMjrMlfmkYeQ7Ask=
=IB4D
-----END PGP SIGNATURE-----