Re: gvariant threading snafu ?

Hi Michael

> 	One of them is from the slightly unfortunate weak-reference /
> hash-table scheme in gvarianttypeinfo.c (patch attached).

Thanks for catching this.  This is truly an impressive bug.  I'm
surprised you were able to track it down -- I hope it didn't take too
much time. :)

I fixed the issue you caught along with another closely-related one.
That patch is now on glib master.

>                                                           Of course, a
> truly 31337 solution starts to look like g_object_unref [ which is a
> fierce piece of code ], and makes you wonder: why not just use a
> GObject :-)

GObject weak references are actually not thread-safe, due to a very
similar problem.  See

I understand from your following message that the other issue is not a
glib bug?


[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]