bitbake: cooker/cookerdata: Improve configuration object handling

Originally it seemed like a good idea to keep the parameters around. Having
seen this in real life use, its incorrect, we should pull all the data we need
into the cooker's configuguration and then use this to build the datastore.

Being able to just build the datastore from the parameters seemed like a good
idea but having a dummy cooker configuration object is now looking like
the better option.

This also fixes failures in hob since the parseFiles command can call
into cooker directly now and reset the configuration prefiles and postfiles
at will, rather than the indirect calls before which were breaking the datastore
(e.g. BBPATH wasn't set).

The cleanup this allows in tinfoil illustrates how this change makes more sense.

(Bitbake rev: f50df5b891bf318f12fc61c74adfcc626cc6f836)

Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

1 files changed, 1 insertions, 1 deletions

diff --git a/bitbake/lib/bb/cooker.py b/bitbake/lib/bb/cooker.py
index 4c0b569439..cd9cccdfce 100644
--- a/bitbake/lib/bb/cooker.py
+++ b/bitbake/lib/bb/cooker.py
@@ -165,7 +165,7 @@ class BBCooker:
         if not self.configuration.server_register_idlecallback:
             worker = True
 
-        self.databuilder = bb.cookerdata.CookerDataBuilder(self.configuration.params, worker)
+        self.databuilder = bb.cookerdata.CookerDataBuilder(self.configuration, worker)
         self.configuration.data = self.databuilder.data
 
     def enableDataTracking(self):