关键词: Autovectorization is an optimal method for the compiler to improve the parallelism of programs.With the extensive use of a computing platform supporting the SIMD architecture processor, autovectorization becomes a hot research of compiler technology. The GCC compiler is a kind of open source, crossplatform compiler. Based on the internal autovectorization algorithm and the architecture and instruction set characteristics of the Matrix chip, this paper extends the Matrix vector instruction set at GCC&rsquo, s backend and realizes the basic autovectorization support. The measured results indicate that the extended compiler can support the Matrix vector instruction set, the basic autovectorization and the development of the Matrixbased parallel programs by way of builtin functions.

Abstract:

Autovectorization is an optimal method for the compiler to improve the parallelism of programs.With the extensive use of a computing platform supporting the SIMD architecture processor, autovectorization becomes a hot research of compiler technology. The GCC compiler is a kind of open source, crossplatform compiler. Based on the internal autovectorization algorithm and the architecture and instruction set characteristics of the Matrix chip, this paper extends the Matrix vector instruction set at GCC’s backend and realizes the basic autovectorization support. The measured results indicate that the extended compiler can support the Matrix vector instruction set, the basic autovectorization and the development of the Matrixbased parallel programs by way of builtin functions.

Key words: GCC;Matrix;autovectorization