[oilpan]: Don't enter safepoint in SafePointAwareMutexLocker if sweeping.

Source/platform/heap/ThreadState.h

Index: Source/platform/heap/ThreadState.h
diff --git a/Source/platform/heap/ThreadState.h b/Source/platform/heap/ThreadState.h
index ee55938a23a43ebaaa05e2493daf6b03f8d1c6a5..5269820c6417a1cff55af4d7d8237da95eed2fb1 100644
--- a/Source/platform/heap/ThreadState.h
+++ b/Source/platform/heap/ThreadState.h
@@ -617,7 +617,11 @@ public:
         ThreadState* state = ThreadState::current();
         do {
             bool leaveSafePoint = false;
-            if (!state->isAtSafePoint()) {
+            // We cannot enter a safepoint if we are currently sweeping. In that
+            // case we just try to acquire the lock without being at a safepoint.
+            // If another thread tries to do a GC at that time it might time out
+            // due to this thread not being at a safepoint and waiting on the lock.

[haraken, 2014/07/01 10:35:10]

Would you elaborate on why this doesn't cause dead lock?

(1) Thread 1 is sweeping.
(2) Thread 2 calls SafePointAwareMutexLocker. It acquires the mutex.
(3) Thread 2 triggers a GC.
(4) Thread 1 calls SafePointAwareMutexLocker. It tries to acquire the mutex.
(5) Dead lock?

Actually I don't fully understand why we cannot enter a safe point while
sweeping. It seems we can just enter a safe point, and if a GC is triggered we
can just clear the mark bits and restart the GC. I'd be happy if you could just
help me understand :)

+            if (!state->isSweepInProgress() && !state->isAtSafePoint()) {
                 state->enterSafePoint(ThreadState::HeapPointersOnStack, this);
                 leaveSafePoint = true;
             }