Patching Locking Bugs Statically with Crayons

The Linux Kernel is a world-class operating system controlling most of our computing infrastructure: mobile devices, Internet routers and services, the supercomputers. also an example low-level software with no comprehensive regression test suite (for good reasons). kernel’s tremendous societal importance imposes strict stability correctness requirements. These properties make challenging relevant target for static automated program repair (APR). Over past decade, significant progress has been made in dynamic APR. However, APR techniques do not translate naturally to systems without tests. We present technique addressing sequential locking API misuse bugs Kernel. attack key challenge APR, namely, lack detailed specification, by combining analysis machine learning complement information presented analyzer. In experiments on historical real-world kernel, we were able automatically re-produce or propose equivalent patches 85% human-made patches, rank them among top three candidates 64% cases five 74%.