47
edits
LYH
Joined 29 April 2013
→Y: Your task
| Line 104: | Line 104: | ||
:* '''If yes, what task did you choose? Please describe what part of it you especially want to focus on if you can already provide this information.''' | :* '''If yes, what task did you choose? Please describe what part of it you especially want to focus on if you can already provide this information.''' | ||
http://wiki.octave.org/Summer_of_Code_Project_Ideas#Improve_JIT_compiling | |||
::I choose “Improveing JIT compiling” project, focusing on “Enhance JITC functionality” and “Migrate from LLVM JIT to LLVM MCJIT”. Current JITC lacks some feature. In addition disablility of built-in function linking, JITC cannot work with matrices data type but only scalar type. I want to concentrate on expanding JITC to handle this non-trivial but fundamental case. | ::I choose “Improveing JIT compiling” project, focusing on “Enhance JITC functionality” and “Migrate from LLVM JIT to LLVM MCJIT”. Current JITC lacks some feature. In addition disablility of built-in function linking, JITC cannot work with matrices data type but only scalar type. I want to concentrate on expanding JITC to handle this non-trivial but fundamental case. | ||
::Second, I would begin to migrate from LLVM JIT to LLVM MCJIT. JITC use JIT now, however LLVM community has no longer maintain JIT and move on MCJIT. I think we could move on MCJIT (and still support JIT). This is really important if we want to get the benefit from LLVM subsequently. | ::Second, I would begin to migrate from LLVM JIT to LLVM MCJIT. JITC use JIT now, however LLVM community has no longer maintain JIT and move on MCJIT. I think we could move on MCJIT (and still support JIT). This is really important if we want to get the benefit from LLVM subsequently. | ||