期刊文献+

GCC后端中四路双精度短向量寄存器的实现 被引量:1

Implementation of Four-way Double Precision Short Vector Registers in GCC Backend
下载PDF
导出
摘要 设计和实现一个新的产品化的编译器通常需要几年时间。基于已有的编译器进行修改和扩展,是研发面向新体系结构的编译器的主要途径。GNU编译器集合(GCC)支持多种高级语言和多种目标处理器平台、文档及源代码开放等。基于GCC的Sparc后端,实现了支持四路双精度SIMD指令的四路双精度短向量寄存器的描述。在此过程中,定义了新的目标机,扩充了一类向量模式,定义了一类新的寄存器约束,实现了四路双精度寄存器的描述,定义了四路双精度SIMD指令的机器描述。对于面向此类SIMD指令的内嵌函数,GCC编译器能够正确使用该类向量寄存器来生成对应的SIMD指令。 It will cost several years to design and implement a new product-level compiler.Designing and implementing based on an already-issued product-level compiler are the main approach to develop a compiler for a new architecture.GNU compiler collection(GCC)supports multiple high level languages and multiple platforms,and its internal documents and source code are open.Based on the Sparc backend of GCC,we implemented the description of four-way double-precision short vector registers which support four-way double-precision SIMD instructions.In this process,we defined a new target machine,expanded a new vector mode,defined a new class of register constraints,provided the descriptions of four-way double-precision short vector registers,designed the machine descriptions of the four-way doubleprecision SIMD instructions.For the builtin functions for this kind of SIMD instructions,our GCC can produce correct SIMD instructions using such kind of vector registers.
出处 《计算机科学》 CSCD 北大核心 2012年第9期292-295,306,共5页 Computer Science
基金 国家自然科学基金项目(61170046 61170045)资助
关键词 GCC后端 四路双精度 向量寄存器 GCC backend Four-way double precision Vector registers
  • 相关文献

参考文献7

  • 1GCC, GNU Compiler Collection[OL]. http://gcc, gnu. org/.
  • 2The GNU General Public License[OL]. http://www, gnu. org/ licenses/licenses, html GPL.
  • 3OpenSPARCTM T2 Core Microarchitecture Specification[Z]. Re- vision A, Sun Microsystems, Inc. , December 2007.
  • 4Firasta N, Buxton M,Jinbo P, et al. Intel AVX: New Frontiers in Performance Improvements and Energy Efficiency[Z]. Intel white paper, 2008.
  • 5The VISTM Instruction Set V1. O[Z]. White paper, Sun Micrc: systems Inc. ,June 2002.
  • 6Makarov V N. The Integrated Register Allocator for GCC[C]// Proceedings of the CA2C Developers' Summit. Ottawa, Ontario, Canada,July 2007 ; 77-90.
  • 7Stallman R M. The CA2C Developer Community. GNU Compiler Collection Internals. For GCC version 4. 6 [Z]. Free Software Foundation, 2010.

同被引文献10

  • 1OpenSPARCTM T2 Core Microarchitecture Specification. Revi- sion A[S]. Sun Microsystems, Inc. , Dec. 2007.
  • 2Firasta N, Buxton M,Jinbo P, et al. Intel AVX: New Frontiers in Performance Improvements and Energy Efficiency[M]. 2008.
  • 3UltraSPARC Architecture 200T[M]. Sun Microsystems, Inc. , Aug. 2007.
  • 4GCC, GNU Compiler Colleetion[OL]. http://gce, gnu. org/.
  • 5The GNU General Public Lieense[OL]. http.-//www, gnu. org/ licenses/licenses, html GPL.
  • 6Ganguin M. C,C/: Back-end for the Ulysse Processor [D]. [COLE POLYT[CHNIQUE FDRALE DE LAUSANNE, Sept. 2007.
  • 7林秉毅,刘兴僳,陈立杰,等.追踪GCC核心原始码与移植相关之研究[R].专题报告.台湾逢甲大学资讯工程学系.
  • 8Str/itling A. Optimizing the Suite for a VLIW Architecture [D]. Chemnitz University of Technology, 2004.
  • 9林凯伟.移植GCC的系统方法之研究[D].台湾国立中正大学资讯工程研究所,2008.
  • 10李春江,杜云飞,易会战,杨灿群.GCC中内嵌函数实现剖析[J].计算机科学,2012,39(B06):357-359. 被引量:2

引证文献1

二级引证文献3

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部