-
Notifications
You must be signed in to change notification settings - Fork 0
/
TODO
44 lines (27 loc) · 1.35 KB
/
TODO
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
TODO:
A class extends an abstract class and some of the callbacks are not visible in
the concrete class, but they should be - the missing callbacks could be doing
something nasty.
Add edges from finsish() to the end of the lifecycle.
Create Manager for SpecialConditions.
Put result directories in ApkPaths.java
Look at intent creation while a WakeLock is held.
Sanitize CFG (check out exceptional edges, etc.)
Circular dependencies (check again ?)
Check what king of operations are called with the WakeLock held (Intents,
Services, Handlers, etc.)
There are cases where a class belongs to multiple components (eg. Service
implements Runnable)
Check cases that are not flagged at all.
Account for all kinds of WakeLocks (Screen, Full, Partial) - and create a more
fine grained domain, following the scheme in Pathak pg.4.
Unresolved Callbacks can have wakelock info too: propagate thread info for
them as well?
Get rid of JNI methods from CG construction - or just ignore them (as with
native and abstract).
Like WakeLocks, we can have "exotic" components, e.g. camera on/off.
--------------------------------------------------------------------------------------------------
Just did:
1. every wakelock object is treated as different (identified by its Binder
object) - So do continue to treat them separately based on
abstract representation.