Changes between Version 13 and Version 14 of NestedCPR


Ignore:
Timestamp:
Jan 7, 2014 5:24:15 PM (20 months ago)
Author:
nomeata
Comment:

--

Legend:

Unmodified
Added
Removed
Modified
  • NestedCPR

    v13 v14  
    1616
    1717 * Does Nick Frisby’s late λ-lifting alleviate problems when CPR’ing join-points?
     18   * Need to see if his branch can be merged onto master.
    1819 * Paper-Writeup of CPR
    1920 * Shouldn’t nested CPR help a lot with Complex-heavy code? Is there something in nofib?
     
    2223 * Look at !DmdAnal-related [SLPJ-Tickets] and see which ones are affected by nested-cpr.
    2324 * Do not destroy join points (see below).
    24  * Can we make sure more stuff gets the `Converging` flag, e.g. after a `case` of an unboxed value? Should case binders get the `Converging` flag? What pattern match variables in strict data constructors? Unboxed values?
     25 * Can we make sure more stuff gets the `Converging` flag, e.g. after a `case` of an unboxed value? Should case binders get the `Converging` flag? What about pattern match variables in strict data constructors? Unboxed values?
     26 * Why does nested CPR make some stuff so bad?
     27  * Possibly because of character reboxing. Try avoiding CPR’ing `C#` alltogether!
    2528
    2629=== join points ===
     
    4952 * Should `runSTRep` be inlined (see ticket:1600#comment:34)?
    5053 * Can we use `Terminates` CPR information to eagerly evaluate thunks? Yes, and there is a small gain there: #8655
     54  * But why no allocation change? Understand this better!
     55  * Can we statically and/or dynamically count the number of thunks, and the number of CBV’ed thunks?
    5156 * Why is `cacheprof` not deterministic? (→ #8611)
    5257 * What became of Simon’s better-ho-cardinality branch? See [./better-ho-cardinality].
     58 * Try vtunes to get better numbers.