[Webkit-unassigned] [Bug 88419] Memory corruption on HashTable.h

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=88419





--- Comment #3 from Joe Mason <jmason at rim.com>  2012-06-06 09:26:26 PST ---
So I guess the heap is being corrupted by something earlier, and glibc isn't detecting the corruption until that innocuous malloc call.  Can you run the test through Valgrind to see if it can detect the corruption when it happens?

Tracing through the code, the sequence is basically:

m_mutex = adoptPtr(new Mutex)) (in HashTable constructor)
MutexLocker lock(*m_mutex); (in invalidateIterators called from add)
fastZeroedMalloc(64 * sizeof(int)); (in allocateTable called from add)

The only way I can see that failing is if "new Mutex" returned 0, so the MutexLocker operates on random memory.  In which case you've got bigger problems...  I suggest adding ASSERT(m_mutex) in the constructor.  And possibly before each MutexLocker call.

-- 
Configure bugmail: https://bugs.webkit.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.