Changes between Version 4 and Version 5 of squirrelfish


Ignore:
Timestamp:
Mar 24, 2008 6:10:44 PM (14 years ago)
Author:
cwzwarich@uwaterloo.ca
Comment:

--

Legend:

Unmodified
Added
Removed
Modified
  • squirrelfish

    v4 v5  
    1616
    1717Function calls to native functions (requires revamping the List class to be an alias to the register file -- Geoff has some ideas for making this fast)
     18
     19'''Cameron is working on:
     20'''
     21
     22Investigating performance regressions caused by the introduction of new opcodes. In particular, a simple for loop with no body regresses about 25%. It seemed at first that this has to do with the fact that these instructions call arbitrary external functions, but some odd performance differences still occur without these calls.
     23
     24Better code generation. We have been pondering whether to have a separate peephole optimization pass or to incorporate peephole optimization into code generation. Either way, we should look at some code generation algorithms based on tile matching. We also want to choose an approach that will be compatible with planned extensions, e.g. superinstructions.
    1825
    1926'''You could take something from Geoff, or make something up yourself, or do one of these:'''
     
    6572Future optimizations:
    6673
    67 Tile matching algorithm for emitting faster patterns of code
    68 
    69 The introduction of op_new_func, op_call, and op_ret regressed the simple for loop tests by about 25%. Seems to have to do with the fact that these opcodes call arbitrary external functions.
    70 
    7174Use RefPtr to indicate use of register -- moves to un-refed registers should be stripped or consolidated to other instructions.
    7275    - i++ => ++i