[Mono-bugs] [Bug 80508][Nor] Changed - code quality regression for array init

bugzilla-daemon at bugzilla.ximian.com bugzilla-daemon at bugzilla.ximian.com
Fri Jan 12 19:13:33 EST 2007

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 marek.safar at seznam.cz.


--- shadow/80508	2007-01-12 18:04:38.000000000 -0500
+++ shadow/80508.tmp.7984	2007-01-12 19:13:33.000000000 -0500
@@ -32,6 +32,24 @@
 ------- Additional Comments From lupus at ximian.com  2007-01-12 18:04 -------
 It is a bad change. Instead of filing a whishlist bug for the jit to
 optimize this case you cause bloat both in the assembly and in the
 runtime generated code, making the code unoptimizable. Your limit can
 cause 4-5 KB of generated code just to initialize an array. Please
 revert to more sane values, like 10 or 20 instead of 200.
+------- Additional Comments From marek.safar at seznam.cz  2007-01-12 19:13 -------
+Thank you for your comments.
+IIRC, this change was part of a bugfix about constant and non-constant
+array initializers optimization.
+I will try to do the performance testing again to see if the runtime
+made any improvements but when I did this change InitializeArray was
+significantly (5x?) slower. I agree, I should fill separate bug report
+(but it usually takes very long time to close runtime performance bugs
+:-( )
+I understand that on other architectures you can get different results
+ so it probably makes better sense to optimize InitializeArray to be
+architecture specific and set threshold lower to cover only cost of
+managed->unmanaged transition.

More information about the mono-bugs mailing list