bitbake: event/cooker/runqueue: Add ability to interrupt longer running code

Bitbake is now able to understand when a UI wants it to stop the current
processing. There are some long running functions which currently have no
mechanism to interrupt them however.

This patch adds a call, bb.event.check_for_interrupts(d) which can be
placed in long running code and allows an internal state flag within
bitbake to be checked. If set, that flag will trigger an exit.

This means that Ctrl+C can be made to work in a variety of places where
it currently would not.

Long running event handlers in OE-Core can also then benefit from this
new approach with the addition of the function call as well.

(Bitbake rev: b7ed7e9a815c4e10447fd499508be3dbb47f06e8)

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

bitbake/lib/bb/runqueue.py

@@ -655,6 +655,7 @@ class RunQueueData:
 
         self.init_progress_reporter.start()
         self.init_progress_reporter.next_stage()
+        bb.event.check_for_interrupts(self.cooker.data)
 
         # Step A - Work out a list of tasks to run
         #
@@ -803,6 +804,7 @@ class RunQueueData:
         #self.dump_data()
 
         self.init_progress_reporter.next_stage()
+        bb.event.check_for_interrupts(self.cooker.data)
 
         # Resolve recursive 'recrdeptask' dependencies (Part B)
         #
@@ -899,6 +901,7 @@ class RunQueueData:
             self.runtaskentries[tid].depends.difference_update(recursivetasksselfref)
 
         self.init_progress_reporter.next_stage()
+        bb.event.check_for_interrupts(self.cooker.data)
 
         #self.dump_data()
 
@@ -980,6 +983,7 @@ class RunQueueData:
                 mark_active(tid, 1)
 
         self.init_progress_reporter.next_stage()
+        bb.event.check_for_interrupts(self.cooker.data)
 
         # Step C - Prune all inactive tasks
         #
@@ -1019,6 +1023,7 @@ class RunQueueData:
                 bb.msg.fatal("RunQueue", "Could not find any tasks with the tasknames %s to run within the recipes of the taskgraphs of the targets %s" % (str(self.cooker.configuration.runall), str(self.targets)))
 
         self.init_progress_reporter.next_stage()
+        bb.event.check_for_interrupts(self.cooker.data)
 
         # Handle runonly
         if self.cooker.configuration.runonly:
@@ -1059,6 +1064,7 @@ class RunQueueData:
         logger.verbose("Assign Weightings")
 
         self.init_progress_reporter.next_stage()
+        bb.event.check_for_interrupts(self.cooker.data)
 
         # Generate a list of reverse dependencies to ease future calculations
         for tid in self.runtaskentries:
@@ -1066,6 +1072,7 @@ class RunQueueData:
                 self.runtaskentries[dep].revdeps.add(tid)
 
         self.init_progress_reporter.next_stage()
+        bb.event.check_for_interrupts(self.cooker.data)
 
         # Identify tasks at the end of dependency chains
         # Error on circular dependency loops (length two)
@@ -1082,12 +1089,14 @@ class RunQueueData:
         logger.verbose("Compute totals (have %s endpoint(s))", len(endpoints))
 
         self.init_progress_reporter.next_stage()
+        bb.event.check_for_interrupts(self.cooker.data)
 
         # Calculate task weights
         # Check of higher length circular dependencies
         self.runq_weight = self.calculate_task_weights(endpoints)
 
         self.init_progress_reporter.next_stage()
+        bb.event.check_for_interrupts(self.cooker.data)
 
         # Sanity Check - Check for multiple tasks building the same provider
         for mc in self.dataCaches:
@@ -1188,6 +1197,7 @@ class RunQueueData:
 
         self.init_progress_reporter.next_stage()
         self.init_progress_reporter.next_stage()
+        bb.event.check_for_interrupts(self.cooker.data)
 
         # Iterate over the task list looking for tasks with a 'setscene' function
         self.runq_setscene_tids = set()
@@ -1200,6 +1210,7 @@ class RunQueueData:
                 self.runq_setscene_tids.add(tid)
 
         self.init_progress_reporter.next_stage()
+        bb.event.check_for_interrupts(self.cooker.data)
 
         # Invalidate task if force mode active
         if self.cooker.configuration.force:
@@ -1216,6 +1227,7 @@ class RunQueueData:
                     invalidate_task(fn + ":" + st, True)
 
         self.init_progress_reporter.next_stage()
+        bb.event.check_for_interrupts(self.cooker.data)
 
         # Create and print to the logs a virtual/xxxx -> PN (fn) table
         for mc in taskData:
@@ -1228,6 +1240,7 @@ class RunQueueData:
                 bb.parse.siggen.tasks_resolved(virtmap, virtpnmap, self.dataCaches[mc])
 
         self.init_progress_reporter.next_stage()
+        bb.event.check_for_interrupts(self.cooker.data)
 
         bb.parse.siggen.set_setscene_tasks(self.runq_setscene_tids)
 
@@ -1240,6 +1253,7 @@ class RunQueueData:
                     dealtwith.add(tid)
                     todeal.remove(tid)
                     self.prepare_task_hash(tid)
+                bb.event.check_for_interrupts(self.cooker.data)
 
         bb.parse.siggen.writeout_file_checksum_cache()
 
@@ -1483,6 +1497,7 @@ class RunQueue:
         """
 
         retval = True
+        bb.event.check_for_interrupts(self.cooker.data)
 
         if self.state is runQueuePrepare:
             # NOTE: if you add, remove or significantly refactor the stages of this