Ensuring the correctness of code coverage profilers is of critical importance as code coverage has been widely adopted to facilitate substantial software engineering tasks. Recent studies have proposed to leverage differential testing and metamorphic testing to validate code coverage profilers, which have been shown to be effective in uncovering bugs for coverage profilers. However, existing techniques still have a bunch of drawbacks as they only use \textit{homogeneous source}, i.e., different coverage profilers or the profilers themselves, to validate coverage profilers. In this paper, we propose Decov, a \textit{heterogeneous testing} technique, to validate coverage profilers with the support of information provided by a \textit{heterogeneous source}, which are the debuggers. The coverage profilers record the \textit{execution count} for each line of a program during execution, while the debuggers monitor the \textit{hit count} for each line of a program when running the program in debuggers. Our key insight is that the execution count obtained from the coverage profilers for any statement should not conflict with the hit count monitored by the debuggers. As such, Decov constructs a number of \textit{heterogeneous relations} and utilizes them to uncover bugs in coverage profilers. The experiments on Gcov and LLVM-cov, two popular code coverage profilers, have led to 21 new bug reports and 18 of them are directly confirmed by developers. Of which, 5 bugs are fixed/resolved by developers in the latest trunk version. Decov is a simple yet effective coverage profiler validator that can be viewed as a complementary approach to existing ones.