Changes between Version 1 and Version 2 of NestedCPR


Ignore:
Timestamp:
Dec 10, 2013 6:18:27 PM (16 months ago)
Author:
nomeata
Comment:

--

Legend:

Unmodified
Added
Removed
Modified
  • NestedCPR

    v1 v2  
    1919 * Shouldn’t nested CPR help a lot with Complex-heavy code? Is there something in nofib? 
    2020 * Try passing CPR information from the scrunitee to the pattern variables. For that: Reverse flow of analysis for complex scrunitees (for simple, we want the demand coming from the body, for complex, this is not so important.) 
    21  * Check nofib of better-ho-cardinality and possibly merge into master, including all non-nested-cpr improvements from the nested-cpr branch. 
    2221 * Why is `cacheprof` not deterministic? 
    2322 * Use ticky-profiling to learn more about the effects of nested CPR. 
    2423 * Look at !DmdAnal-related [SLPJ-Tickets] and see which ones are affected by nested-cpr. 
    2524 
     25==== better-ho-cardinality ==== 
     26 
     27It would be nice to merge the code structure improvements and notes into master, to keep my branch short. But it is based on `better-ho-cardinality`, and that is not suitable for merging because of unexpected regressions even in `nofib` and ´rtak`. So I am investigating. 
     28 
     29In these tests, it is related to reading and showing data. Small example: 
     30{{{ 
     31#!haskell 
     32main = (read "10" :: Int) `seq` return () 
     33}}} 
     34Baseline: 49832, `better-ho-cardinality`: 49968. Unfortunately, the changes to, for example, `GHC.Read` are not small, and probably mostly benign... 
     35 
    2636=== Side tracks === 
    2737