GraphBLAS Lowering Pass¶
These tutorials and examples will cover how to use the GraphBLAS dialect’s --graphblas-lower pass by using the JIT Engine to lower several ops from the GraphBLAS dialect.
The main purpose of the --graphblas-lower pass is to lower from the GraphBLAS dialect into a lower level dialect, e.g. the SCF dialect or the Sparse Tensor dialect. Rather than simply showing the code transformations, we’ll use the JIT Engine to take some example MLIR code using the GraphBLAS dialect and create executable Python code from it. Since we won’t be able to go over all of the ops in the GraphBLAS dialect and since all of the ops are documented with examples in the GraphBLAS Dialect Op Reference, our examples in this section will mostly cover ops that are commonly used.
The content of the tutorials are somewhat sequentially dependent as some later tutorials assume completion of previous tutorials. They also assume familiarity with the content of the GraphBLAS Dialect Op Reference.