Changes between Version 41 and Version 42 of OpenCL_Memo
- Timestamp:
- Dec 12, 2011 10:57:59 PM (13 years ago)
Legend:
- Unmodified
- Added
- Removed
- Modified
-
OpenCL_Memo
v41 v42 1 1 = OpenCL Notes = 2 2 == Disable auto vectorization == 3 In the section 6.7.2 (187 page) in the OpenCL Specification Version: 1.1 (Revision: 33), "__attribute__((vec_type_hint(<typen>))" is described. This hint controls the autovectorizer in the compiler for OpenCL C. I tested this feature by dumping the assembly code for a kernel ( using "float8" ) targeted for AVX instructions with Intel SDK (version 1.5).3 In the section 6.7.2 (187 page) in the OpenCL Specification Version: 1.1 (Revision: 33), "__attribute__((vec_type_hint(<typen>))" is described. This hint controls the autovectorizer in the compiler for OpenCL C. I tested this feature by dumping the assembly code for a kernel (heavily using "float8" ) targeted for AVX instructions with Intel SDK (version 1.5). 4 4 5 5 || attribute || lines || comment || … … 10 10 Without this attribute, the generated code includes two functions: (1) the code without unrolling and (2) the code over unrolled. 11 11 Due to this, the generated assembly file is large but we don't know which code is really used. 12 13 In both code, the core part is calculated with AVX instructions but the performance of the two codes is slightly different. .... now investigating. 14 12 15 13 16 == How to use "ioc" command equipped with Intel SDK. ==