Don't set_checked() on is_override files.

If a file is overridden and then overridden again, this caused us to
rebuild only the first thing that depends on it, but not any subsequent
things, which is a pretty serious bug.

It turned out that t/350-deps-forget is already supposed to test this,
but I had cleverly encoded the wrong behaviour into the expected
results in the table-driven test.  I blame lack of sleep.  Anyway, I
fixed the test, which made it fail, and then fixed the code, which made
it pass.

redo/builder.py

@@ -165,11 +165,10 @@ class _BuildJob(object):
                 warn('%s - old: %r\n' % (_nice(t), sf.stamp))
                 warn('%s - new: %r\n' % (_nice(t), newstamp))
                 sf.set_override()
-            sf.set_checked()
             sf.save()
-            return self._finalize(0)
+            # fall through and treat it the same as a static file
         if (os.path.exists(t) and not os.path.isdir(t + '/.')
-                and not sf.is_generated):
+                and (sf.is_override or not sf.is_generated)):
             # an existing source file that was not generated by us.
             # This step is mentioned by djb in his notes.
             # For example, a rule called default.c.do could be used to try