Changes between Version 15 and Version 16 of JITRewrite

Show
Ignore:
Timestamp:
11/03/09 03:15:06 (4 years ago)
Author:
plobsing
Comment:

put tasks at the top

Legend:

Unmodified
Added
Removed
Modified
  • JITRewrite

    v15 v16  
    11The current JIT system is not a suitable implementation of JIT for Parrot. It should be deprecated and replaced with a better system 
     2 
     3= Path Forward = 
     4 
     5Here are some general steps forward 
     6 
     7 - Decide on a standard intermediate language for our Ops (C what we have now, Lorito, or JIT Definitions) 
     8 - Write a parser for that intermediate language 
     9 - Write code generation backends for the intermediate language to output C Functions and JIT Definitions for at least one JIT engine * 
     10 - Write all the ops in the intermediate language * 
     11 
     12Note: if our intermediate language for Ops is C, we can ignore at least part of the starred steps, but will have to do a lot more work building a suitable parser. 
     13 
     14The specific plan is: 
     15 
     16  * Design Lorito 
     17  * Create a Lorito parser with multiple backends (C code in several formats, LLVM code) 
     18  * read LLVM documentation like mad men (whiteknight, help wanted!) 
     19  * Make a C code backend for PCT 
     20 
     21Accomplished steps: 
     22 
     23  * Redirect --runcore=jit to the fast core (also redirected switch JIT and CGP JIT cores and their Makefile targets) 
     24  * rip out the current JIT system 
    225 
    326= Definitions = 
     
    101124 
    102125http://luajit.org/dynasm.html 
    103  
    104 = Path Forward = 
    105  
    106 Here are some general steps forward 
    107  
    108  - Decide on a standard intermediate language for our Ops (C what we have now, Lorito, or JIT Definitions) 
    109  - Write a parser for that intermediate language 
    110  - Write code generation backends for the intermediate language to output C Functions and JIT Definitions for at least one JIT engine * 
    111  - Write all the ops in the intermediate language * 
    112  
    113 Note: if our intermediate language for Ops is C, we can ignore at least part of the starred steps, but will have to do a lot more work building a suitable parser. 
    114  
    115 The specific plan is: 
    116  
    117   * Design Lorito 
    118   * Create a Lorito parser with multiple backends (C code in several formats, LLVM code) 
    119   * read LLVM documentation like mad men (whiteknight, help wanted!) 
    120   * Make a C code backend for PCT 
    121  
    122 Accomplished steps: 
    123  
    124   * Redirect --runcore=jit to the fast core (also redirected switch JIT and CGP JIT cores and their Makefile targets) 
    125   * rip out the current JIT system