This PR demonstrates how matcher table feedback can be used to aid fuzzing. If the new InsertIntrinsicStrategy is selected for mutation, the shadow map is first analyzed, and a list of uncovered intrinsic function IDs are dumped, one of which will be converted to a Function* and used for InsertFunctionStrategy::mutate. This PR depends on SecurityLab-UCD/llvm-project:insert-func-customize.
This PR demonstrates how matcher table feedback can be used to aid fuzzing. If the new
InsertIntrinsicStrategyis selected for mutation, the shadow map is first analyzed, and a list of uncovered intrinsic function IDs are dumped, one of which will be converted to aFunction*and used forInsertFunctionStrategy::mutate. This PR depends on SecurityLab-UCD/llvm-project:insert-func-customize.