[OE-core] bitbake/runqueue: Fix 'full' stamp checking to be more efficient and cache results

[Richard Purdie]

On Wed, 2012-05-09 at 17:03 -0700, Chris Larson wrote:
> On Wed, May 9, 2012 at 4:57 PM, Richard Purdie
> <richard.purdie at linuxfoundation.org> wrote:
> > diff --git a/bitbake/lib/bb/runqueue.py b/bitbake/lib/bb/runqueue.py
> > index b870caf..48433be 100644
> > --- a/bitbake/lib/bb/runqueue.py
> > +++ b/bitbake/lib/bb/runqueue.py
> > @@ -875,7 +875,7 @@ class RunQueue:
> >             bb.msg.fatal("RunQueue", "check_stamps fatal internal error")
> >         return current
> >
> > -    def check_stamp_task(self, task, taskname = None, recurse = False):
> > +    def check_stamp_task(self, task, taskname = None, recurse = False, cache = {}):
> 
> When people do this, it's typically a bug, but I presume you're doing
> it intentionally here? Use of mutable default values is often
> problematic due to their being shared across all calls to that
> function, but that's okay for a cache. Maybe you intended this, given
> it's a cache, but I wanted to ensure it was a conscious choice. Also,
> this adds yet another global cache with no form of invalidation /
> clear at all, it'll continue to grow through the lifetime of the
> process.

I'll change it to cache = None and then default it to {} in the code. I
agree infinitely growing caches in memory are not a good idea as we need
to avoid them as this could really screw up a UI doing re-execution.

Cheers,

Richard