[pypy-svn] r70164 - pypy/extradoc/talk/jitdec09

pedronis at codespeak.net pedronis at codespeak.net
Thu Dec 17 09:47:31 CET 2009


Author: pedronis
Date: Thu Dec 17 09:47:30 2009
New Revision: 70164

Modified:
   pypy/extradoc/talk/jitdec09/talk.html
   pypy/extradoc/talk/jitdec09/talk.txt
Log:
tweaks

Modified: pypy/extradoc/talk/jitdec09/talk.html
==============================================================================
--- pypy/extradoc/talk/jitdec09/talk.html	(original)
+++ pypy/extradoc/talk/jitdec09/talk.html	Thu Dec 17 09:47:30 2009
@@ -382,8 +382,8 @@
 <h1>Status</h1>
 <ul class="simple">
 <li>Intel x86-32</li>
-<li>Small, medium examples run</li>
-<li>Algorithmic code is sped up, see benchmarks</li>
+<li>Small and some medium examples run</li>
+<li>Good results for algorithmic code, see benchmarks</li>
 <li>Reasonable heuristic when traces get too long</li>
 </ul>
 </div>
@@ -397,15 +397,15 @@
 <li>Further work on memory footprint</li>
 <li>Correct and improved behavior in the presence of exceptions: work in progress</li>
 <li>Better results for recursion, asm to asm calls</li>
-<li>Bugs, correctness (sys.setrace for example)</li>
+<li>Bugs, correctness (sys.settrace for example)</li>
 </ul>
 </div>
 <div class="slide" id="work-left-ii">
 <h1>Work Left (ii)</h1>
 <ul class="simple">
-<li>Machine code management, discard no longer valid  traces</li>
-<li>Make the interpreter more JIT friendly</li>
-<li>Interpreter level algorithmic improvements</li>
+<li>Machine code management, discard no longer valid traces</li>
+<li>Make the interpreter more JIT-friendly</li>
+<li>Interpreter-level algorithmic improvements</li>
 <li>Intel x86-64</li>
 </ul>
 </div>

Modified: pypy/extradoc/talk/jitdec09/talk.txt
==============================================================================
--- pypy/extradoc/talk/jitdec09/talk.txt	(original)
+++ pypy/extradoc/talk/jitdec09/talk.txt	Thu Dec 17 09:47:30 2009
@@ -46,11 +46,9 @@
 Status
 =======
 
-xxx weak
-
 - Intel x86-32
-- Small, medium examples run
-- Algorithmic code is sped up, see benchmarks
+- Small and some medium examples run
+- Good results for algorithmic code, see benchmarks
 - Reasonable heuristic when traces get too long
 
 Benchmarks
@@ -70,7 +68,7 @@
 Work Left (ii)
 ===============
 
-- Machine code management, discard no longer valid  traces
-- Make the interpreter more JIT friendly
-- Interpreter level algorithmic improvements
+- Machine code management, discard no longer valid traces
+- Make the interpreter more JIT-friendly
+- Interpreter-level algorithmic improvements
 - Intel x86-64



More information about the Pypy-commit mailing list