Differences between revisions 3 and 30 (spanning 27 versions)
Revision 3 as of 2006-05-21 11:19:36
Size: 853
Editor: 213
Comment:
Revision 30 as of 2008-11-15 14:00:44
Size: 2612
Editor: localhost
Comment: converted to 1.6 markup
Deletions are marked like this. Additions are marked like this.
Line 1: Line 1:
* [ 1479611 ] speed up function calls Pending:
Line 3: Line 3:
* [ 1442927 ] PyLong_FromString optimization  * [[http://www.python.org/sf/1479611|[ 1479611 ]] ] speed up function calls TP
 * [[http://www.python.org/sf/876193|[ 876193 ]] ] reorganize, extend function call optimizations TP
 * [[http://www.python.org/sf/1107887|[ 1107887 ]] ] Speed up function calls/can add more introspection info TP
 * [[http://www.python.org/sf/1087418|[ 1087418 ]] ] long int bitwise ops speedup TP
 * [[http://www.python.org/sf/1366311|[ 1366311 ]] ] SRE engine do not release the GIL
 * [[http://www.python.org/sf/1346214|[ 1346214 ]] ] Better dead code elimination for the AST compiler GB
   * New patch posted, ready for review
 * [[http://www.python.org/sf/1243654|[ 1243654 ]] ] Faster output if message already has a boundary SH
Line 5: Line 12:
* [ 1366311 ] SRE engine do not release the GIL
Line 7: Line 13:
* [ 1359618 ] Speed charmap encoder Addressed:
Line 9: Line 15:
* [ 1353872 ] a faster Modulefinder (Related to py2exe runtime performance)  * [[http://www.python.org/sf/1145039|[ 1145039 ]] ] Remove some invariant conditions and assert in ceval TP
 * [[http://www.python.org/sf/1442927|[ 1442927 ]] ] `PyLong_FromString` optimization
 * [[http://www.python.org/sf/1335972|[ 1335972]] ] combo speedup and bugfix for string->int conversion
 * [[http://www.python.org/sf/876206|[ 876206 ]] ] scary frame speed hacks [SUCCESS]
 * [[http://www.python.org/sf/1281707|[ 1281707 ]] ] Speed up gzip.readline (~40%) [SUCCESS]
 * [[http://www.python.org/sf/1337051|[ 1337051 ]] ] remove 4 ints from PyFrameObject
 * [[http://www.python.org/sf/1243730|[ 1243730 ]] ] Big speedup in email message parsing SH '''no speedup observed'''
 * [[http://www.python.org/sf/921466|[ 921466 ]] ] Reduce number of open calls on startup GB
 * [[http://www.python.org/sf/1359618|[ 1359618 ]] ] Speed charmap encoder [patch approved, MAL to apply]
Line 11: Line 25:
* [ 1346238 ] A constant folding optimization pass for the AST

* [ 1346214 ] Better dead code elimination for the AST compiler (this appears to be a bugfix and not a speed patch -jackd)

* [ 1337051 ] remove 4 ints from PyFrameObject

* [ 1281707 ] Speed up gzip.readline (~40%)

* [ 1243730 ] Big speedup in email message parsing

* [ 1243654 ] Faster output if message already has a boundary

* [ 1145039 ] Remove some invariant conditions and assert in ceval

* [ 1107887 ] Speed up function calls/can add more introspection info

* [ 936813 ] fast modular exponentiation
Deferred:
 * [[http://www.python.org/sf/813436|[ 813436 ]] ] Scalable zipfile extension
   * Addressed by SoC student
 * [[http://www.python.org/sf/738094|[ 738094 ]] ] for i in range(N) optimization
   * Related to [[http://www.python.org/sf/1472639|[ 1472639 ]] ] make range be xrange
   * Topic was discussed endlessly for some times.
 * [[http://www.python.org/sf/936813|[ 936813 ]] ] fast modular exponentiation
   * Going to need a lot of attention, not suitable for the sprint.
 * [[http://www.python.org/sf/923643|[ 923643 ]] ] long <-> byte-string conversion
   * This is a feature request.
 * [[http://www.python.org/sf/1353872|[ 1353872 ]] ] a faster Modulefinder (Related to py2exe runtime performance)
   * No one interested in it.
 * [[http://www.python.org/sf/1492828|[ 1492828 ]] ] Improvements to ceval.c --> RaymondHettinger
 * [[http://www.python.org/sf/1346238|[ 1346238 ]] ] A constant folding optimization pass for the AST

Pending:

  • [ 1479611 ] speed up function calls TP

  • [ 876193 ] reorganize, extend function call optimizations TP

  • [ 1107887 ] Speed up function calls/can add more introspection info TP

  • [ 1087418 ] long int bitwise ops speedup TP

  • [ 1366311 ] SRE engine do not release the GIL

  • [ 1346214 ] Better dead code elimination for the AST compiler GB

    • New patch posted, ready for review
  • [ 1243654 ] Faster output if message already has a boundary SH

Addressed:

  • [ 1145039 ] Remove some invariant conditions and assert in ceval TP

  • [ 1442927 ] PyLong_FromString optimization

  • [ 1335972 ] combo speedup and bugfix for string->int conversion

  • [ 876206 ] scary frame speed hacks [SUCCESS]

  • [ 1281707 ] Speed up gzip.readline (~40%) [SUCCESS]

  • [ 1337051 ] remove 4 ints from PyFrameObject

  • [ 1243730 ] Big speedup in email message parsing SH no speedup observed

  • [ 921466 ] Reduce number of open calls on startup GB

  • [ 1359618 ] Speed charmap encoder [patch approved, MAL to apply]

Deferred:

  • [ 813436 ] Scalable zipfile extension

    • Addressed by SoC student
  • [ 738094 ] for i in range(N) optimization

    • Related to [ 1472639 ] make range be xrange

    • Topic was discussed endlessly for some times.
  • [ 936813 ] fast modular exponentiation

    • Going to need a lot of attention, not suitable for the sprint.
  • [ 923643 ] long <-> byte-string conversion

    • This is a feature request.
  • [ 1353872 ] a faster Modulefinder (Related to py2exe runtime performance)

    • No one interested in it.
  • [ 1492828 ] Improvements to ceval.c --> RaymondHettinger

  • [ 1346238 ] A constant folding optimization pass for the AST

ListOfPerformanceRelatedPatches (last edited 2008-11-15 14:00:44 by localhost)

Unable to edit the page? See the FrontPage for instructions.