In recent years, the MLIR platform has had explosive growth due to the need of building extensible deep learning compilers and hardware accelerator compilers. Such examples include Triton, CIRCT, and ONNX-MLIR. MLIR compilers introduce significant complexities in localizing bugs or inefficiencies because of their layered optimization and transformation process with compilation passes. While existing delta debugging techniques can be used to identify a minimum subset of IR code that reproduces a given bug symptom, their naive application to MLIR is time-consuming, because real-world MLIR compilers usually involve a large number of compilation passes and compiler developers must also identify a minimized set of relevant compilation passes simultaneously, in order to reduce the footprint of MLIR compiler code to be inspected for a bug fix. We propose DuoReduce, a dual-dimensional reduction approach for MLIR bug localization. DuoReduce leverages three key ideas in tandem to design an efficient MLIR debugger. First, DuoReduce reduces the bug-irrelevant compilation passes by identifying ordering dependencies among different compilation passes. Second, DuoReduce uses MLIR-semantics aware transformations to expedite IR code reduction. Finally, DuoReduce leverages cross-dependence between the IR code dimension and the compilation pass dimension by accounting for which IR code segments are related to which compilation passes to reduce the unused passes.

Experiments with three large-scale MLIR compiler projects find that DuoReduce outperforms syntax-aware reducers such as Perses and Vulcan in terms of IR code reduction by 31.6% and 21.5% respectively. If one uses these reducers by enumerating all possible compilation passes (on average 18 passes), it could take up to 145 hours. By identifying ordering dependencies among compilation passes, DuoReduce reduces this time to 9.5 minutes. By identifying which compilation passes are unused for compiling reduced IR code, DuoReduce reduces the number of passes by 14.6%. This translates to not needing to examine 281 lines of MLIR compiler code on average to fix the bugs. DuoReduce has the potential to significantly reduce debugging effort in multi-layer extensible compilers, which serves as an important basis for the current landscape of machine learning and hardware accelerators.