Re: gio deadlock

[Alexander Larsson <alexl redhat com>]

On Fri, 2008-02-22 at 14:43 -0500, David Zeuthen wrote:
> Saw this earlier today when hacking on some other code ... perhaps we
> need to replace all the locks in gio with recursive ones?

I don't think just plugging in recursive locks is a good idea. That kind
of recursion is just begging for other types of issues. We should emit
signals like file change events and volume monitor events from an idle
or another toplevel mainloop source.