Changes between Version 13 and Version 14 of Profiling

Timestamp:

12/11/13 10:26:02 (9 years ago)

Author:

Antoine Martin

Comment:

--

Profiling

@@ +1 @@
+{{{#!div class="box"
 = Profiling =
 
@@ -4 +5 @@
 [[BR]]
 [/wiki/WindowRefresh Window Refresh] speed is a special case as it depends on many factors and not just pure code performance, though some parts of this code can obviously benefit from profiling.
+}}}
 
+
+{{{#!div class="box"
 == cProfile ==
 [http://docs.python.org/2/library/profile.html#module-cProfile cProfile] outputs text only, which may be more appropriate in some cases. Simply replace {{{xpra}}} with {{{tests/scripts/cProfile}}} on the command line:
@@ -28 +32 @@
 (...)
 }}}
+}}}
 
+
+{{{#!div class="box"
 == pycallgraph ==
 
@@ -53 +60 @@
 * some [../Debugging Debugging] environment variables may be useful
 * save useful graphs with all the information relating to the version used, modifications, environment, etc. So these can be reproduced later.
+}}}
 
-
+{{{#!div class="box"
 == Fixing Bottlenecks ==
 First, write a test to measure/isolate the problematic part of the code.
@@ -64 +72 @@
 * better algorithms
 
-[[BR]]
+
 ----
-[[BR]]
 
-= Examples =
-
-== maths.py re-written in Cython ==
-Merged in r2462
-
-[[BR]]
-
-* Before:
-{{{
-$ XPRA_CYTHON_MATH=0 ./tests/xpra/test_maths.py 
-calculate_time_weighted_average(100000 records)=(0.5129547191923681, 0.529864013016996)
-elapsed time: 72.6ms
-test_calculate_timesize_weighted_average(100000 records)=(70109.42497899215, 22337.16293859974)
-elapsed time: 204.9ms
+You can find a detailed example of how Cython was used to eliminate a performance bottleneck here: [/wiki/Profiling/Examples]
 }}}
-* After:
-{{{
-$ XPRA_CYTHON_MATH=1 ./tests/xpra/test_maths.py 
-calculate_time_weighted_average(100000 records)=(0.5226684212684631, 0.5245391130447388)
-elapsed time: 4.0ms
-test_calculate_timesize_weighted_average(100000 records)=(180154.15625, 35980.69140625)
-elapsed time: 19.7ms
-}}}
-Which is 10 to 20 times faster.
-
-== high network cost in UI thread ==
-Work in progress, see #231
-
-== high cost of logging ==
-This is a recurring problem, see: r2294, r2290, etc
-
-By running the server with:
-{{{
-xpra start :10 --start-child=glxgears
-}}}
-And the client with:
-{{{
-./tests/scripts/pycallgraph -t draw -i '*' -r 20 -- attach :10 --no-mmap
-}}}
-We find that logging is really costing us dearly:
-[[BR]]
-[[Image(pycallgraph-draw-loggingisexpensive.png)]]
-[[BR]]
-So we remove logging from this critical path in r2484, and we now find:
-[[BR]]
-[[Image(pycallgraph-draw-loggingremoved.png)]]
-[[BR]]
-That's a saving of about 35% - not bad for such a small change!