Fix TODOs made during migration to libbinder_ndk

Test: atest GtsStatsdHostTestCases
Bug: 149254662
Change-Id: I50a1998e693f9a0c447cd655f8efa11afa28803b
diff --git a/cmds/statsd/src/external/StatsPullerManager.cpp b/cmds/statsd/src/external/StatsPullerManager.cpp
index 1c38542..0115ba2 100644
--- a/cmds/statsd/src/external/StatsPullerManager.cpp
+++ b/cmds/statsd/src/external/StatsPullerManager.cpp
@@ -85,12 +85,9 @@
         return;
     }
 
-    // TODO(b/149254662): Why are we creating a copy here? This is different
-    // from the other places where we create a copy because we don't reassign
-    // mStatsCompanionService so a destructor can't implicitly be called...
-    shared_ptr<IStatsCompanionService> statsCompanionServiceCopy = mStatsCompanionService;
-    if (statsCompanionServiceCopy != nullptr) {
-        statsCompanionServiceCopy->setPullingAlarm(mNextPullTimeNs / 1000000);
+    // TODO(b/151045771): do not hold a lock while making a binder call
+    if (mStatsCompanionService != nullptr) {
+        mStatsCompanionService->setPullingAlarm(mNextPullTimeNs / 1000000);
     } else {
         VLOG("StatsCompanionService not available. Alarm not set.");
     }
@@ -99,8 +96,6 @@
 
 void StatsPullerManager::SetStatsCompanionService(
         shared_ptr<IStatsCompanionService> statsCompanionService) {
-    // TODO(b/149254662): Why are we using AutoMutex instead of lock_guard?
-    // Additionally, do we need the temporary shared_ptr to prevent deadlocks?
     AutoMutex _l(mLock);
     shared_ptr<IStatsCompanionService> tmpForLock = mStatsCompanionService;
     mStatsCompanionService = statsCompanionService;