Question

Whenever any new object is created, the object is created on heap. The memory allocated for each object has two additional fields 1) The type object pointer 2) sync block index.

What exactly is the usage of these two fields. Can anybody shed light on this?

Was it helpful?

Solution

The type object pointer is used to represent the type of the object. This is required for:

  • Method lookup (the vtable)
  • Checking casts
  • Finding the Type object if you call GetType.

The syncblock field is primarily used for locking. It's only filled in when it needs to be, and when a lock is always uncontested the CLR makes do with a "thin" lock which doesn't require any external data. Otherwise, it's an entry in a process-wide table - I don't know the details of what's in the table, but I would imagine it's things like a list of threads waiting on the object's monitor. Of course the most important bit of information is whether or not the lock is currently held, by which thread, and what its count is (due to the reentrant nature of .NET locks).

The syncblock is also filled in if you call GetHashCode() and it's not overridden - it uses the process-wide table to allocate a stable number, basically. (The address of the object isn't good enough as it can change over time.)

OTHER TIPS

Type object is what returned by obj.GetType call

sync block used for synchronization

See:

The sync block index is used under the hood by the Monitor class and thus the lock statement also.

Some bits of sync block index are also used by GC to mark an object as garbage in case it is no longer referenced.

Licensed under: CC-BY-SA with attribution
Not affiliated with StackOverflow
scroll top