The exploration of the relationships between decision trees and systems of decision rules is a significant research domain within computer science. While established methods exist for converting decision trees into systems of decision rules, the reverse transformation poses considerable challenges. In this paper, we focus on transforming a system of decision rules into a decision tree that solves the All Rules problem: for a given tuple of attribute values, it is required to retrieve all realizable rules from the system (rules with true left-hand side). We analyze and compare the minimum depths of deterministic and nondeterministic decision trees that solve this problem. Additionally, we propose a greedy algorithm that describes how a deterministic decision tree solves the All Rules problem for a given tuple of attribute values, and we evaluate the depth of this decision tree.