Question

Let’s say we have MethodA and MethodB where whole methods’ bodies are within single big lock block. Both methods use the same object for locking:

lock (objectX) 
{ 
    // methodbody
}

There are multiple threads that call these two methods.

Two questions:

1. Is it possible to measure total wait time? Including: time to obtain lock and time required by “other” method’s lock block to finish execution.
2. Would it be faster to introduce locks in both methods on appropriate places instead of encapsulating whole method body with single lock? I realize that here i need to measure performance of both solutions, but I would like to hear your experience and recommendations

Answer 1

2) Would it be faster to introduce locks in both methods on appropriate places

Yes. Always keep locked segments as short as possible. The locking itself is cheap, dont keep others waiting while you do non-critical things.

1) Is it possible to measure total wait time? Including: time to obtain lock and time required by “other” method’s lock block to finish execution.

Not this precisely I think. For long stretches you might look into CPU time vs Elapsed time but in general: use a good profiler.