[Mono-bugs] [Bug 81663][Wis] Changed - Performance: Delegate optimization, DLR and IronPython

[bugzilla-daemon at bugzilla.ximian.com]

Please do not reply to this email- if you want to comment on the bug, go to the
URL shown below and enter your comments there.

Changed by kumpera at gmail.com.

http://bugzilla.ximian.com/show_bug.cgi?id=81663

--- shadow/81663	2007-05-24 14:24:13.000000000 -0400
+++ shadow/81663.tmp.12474	2007-05-24 22:42:01.000000000 -0400
@@ -178,6 +178,17 @@
 ------- Additional Comments From vargaz at gmail.com  2007-05-24 14:24 -------
 I agree with your approach, just doesn't understand why my patch was
 causing a slowdown, when the original call to delegate_invoke was also
 virtual and thus indirect, and the patch removed lots of loads and
 compares.
 
+
+------- Additional Comments From kumpera at gmail.com  2007-05-24 22:42 -------
+Wouldn't a guard on the call site make the whole thing a lot easier?
+This way it would optimize for the non-multicast way using a branch
+with a not-taken hint, which is a lot cheaper than an indirect jump.
+
+So, instead of having all these problems with generating/selecting the
+right delegate method and fixing the calling convention, a simple
+patch in the invoke call site might fix it. I just don't know if the
+code increase overhead would be significant.
+