[pypy-svn] r52208 - pypy/extradoc/talk/sfi2008
arigo at codespeak.net
arigo at codespeak.net
Thu Mar 6 09:42:25 CET 2008
Author: arigo
Date: Thu Mar 6 09:42:24 2008
New Revision: 52208
Modified:
pypy/extradoc/talk/sfi2008/talk.txt
Log:
Typos, minor clarifications
Modified: pypy/extradoc/talk/sfi2008/talk.txt
==============================================================================
--- pypy/extradoc/talk/sfi2008/talk.txt (original)
+++ pypy/extradoc/talk/sfi2008/talk.txt Thu Mar 6 09:42:24 2008
@@ -55,12 +55,12 @@
things such as object model orthogonal to language semantics
encoding
-Developement technology
+Development technology
=======================
-* We use test driven developement extensively
+* We use test driven development extensively
-* We've got a test suite which runs on a single processor for ~8 hours
+* We've got a test suite which would need ~8 hours to run on a single processor
* Our special testing tool, py.test, little hassle and a lot of features
@@ -126,7 +126,7 @@
Abstract interpretation
=======================
-* We start by importing a python module
+* We start by importing the RPython modules on top of CPython
* Next we analyze the bytecode and produce a forest of flow graphs
@@ -214,7 +214,8 @@
* Interpreters are much easier to write than compilers
-* One cannot achieve C-level performance with static analysis
+* For many languages, one cannot achieve C-level performance
+ with static analysis
* Dynamic compilation can produce faster results
than static compilation (e.g. a good Java VM versus gcj)
@@ -230,7 +231,7 @@
* We can do better!
-PyPy approach for JIT
+PyPy approach to JIT
=====================
.. raw:: html
@@ -246,7 +247,7 @@
* automatic generation of a compiler from an interpreter
-* relatively few practical applications so far
+* an old idea with relatively few practical applications so far
JIT - general idea
===================
@@ -256,7 +257,7 @@
* may not yield good performance (our experiments show
about 2x for removing intepretation overhead)
-* things such as the types are still not known
+* things such as the types of Python objects are still not known
Solution: promotion
===================
@@ -279,7 +280,8 @@
* Lazy allocation of objects (allocation only happens when the object escapes)
-* Use CPU stack and registers (we're not very good at it)
+* Use CPU stack and registers as a cache for Python frame
+ objects that live in the heap
Irrelevant to interpreter
==============================
More information about the Pypy-commit
mailing list