Enabling RISC-V Vector Code Generation in MLIR through Custom xDSL Lowerings

TL;DR

This paper introduces a compilation method combining MLIR and xDSL to generate optimized RISC-V Vector code, enabling portable high-performance kernels for scientific and machine learning workloads.

Contribution

It presents a novel approach using custom xDSL lowering passes to translate high-level operations into RISC-V Vector intrinsics, improving code portability and performance.

Findings

Generated kernels outperform OpenBLAS in GFLOPS.

Achieved up to 12.2 GFLOPS on RISC-V platforms.

Performance improvements of 10-35% on evaluated workloads.

Abstract

The growing adoption of RISC-V in high-performance and scientific computing has increased the need for performance-portable code targeting the RISC-V Vector (RVV) extension. However, current compiler infrastructures provide limited end-to-end support for generating optimized RVV code from high-level representations to low-level implementations. In particular, existing MLIR distributions lack practical lowering paths that map high-level abstractions to RVV intrinsics, limiting their applicability for production-ready RISC-V kernels. This paper presents a compilation approach that combines MLIR with xDSL to bridge the missing lowering stages required for RVV code generation. Using custom intermediate representations and transformation passes implemented in xDSL, we systematically translate high-level operations into specialized, hardware-aware C code invoking RVV intrinsics. The resulting kernels are emitted as portable C functions that can be directly integrated into existing applications, enabling incremental adoption without modifying surrounding software stacks. We demonstrate the approach on the General Matrix Multiplication (GEMM) kernel and evaluate the generated micro-kernels on two real RISC-V platforms, the K230 and the BananaPi F3, comparing against OpenBLAS for both square-matrix benchmarks and transformer-based workloads derived from the BERT-Large model. When integrated into a matrix multiplication kernel, the proposed approach consistently outperforms OpenBLAS, reaching up to 12.2 GFLOPS compared to the baseline's 5.1 GFLOPS and providing performance improvements between 10--35\% across the evaluated workloads. These results demonstrate that combining MLIR with xDSL provides a practical pathway to portable, optimized code generation for RISC-V platforms.