I have a new (may be very well known) wisdom — “The bug count* found in a software program or app is directly proportional to the count of test cases or number of test personnel” :).
* I mean the count of bugs you will find, not the total number of bugs.
In other words, the more you test your software, the more bugs you will find in your software program or app. And except most trivial programs, there will always be some bugs left, always! Bugs due to misinterpreted API documentation, race conditions, deadlocks (like these from DllMain), bounds check related — min/max limits, etc., etc., etc…
I strongly recommend to test your software extensively before releasing it to the users. A good start is Wikipedia article about Software testing.
