Comments on: Writing a profiler for Silverlight applications – Part 1 http://lostechies.com/gabrielschenker/2009/11/26/writing-a-profiler-for-silverlight-applications-part-1/ Blog about architectural patterns, best practices, coding principles and techniques Wed, 22 May 2013 12:15:00 +0000 hourly 1 http://wordpress.org/?v=3.4.2 By: Jb Evain http://lostechies.com/gabrielschenker/2009/11/26/writing-a-profiler-for-silverlight-applications-part-1/#comment-198 Jb Evain Tue, 01 Dec 2009 12:46:34 +0000 /blogs/gabrielschenker/archive/2009/11/26/writing-a-profiler-for-silverlight-applications-part-1.aspx#comment-198 The multiple exit points issue is not really hard to fix, just replace all the `ret`s opcode to an unconditional branch to a block emitted at the end of the method. You'll just have to take care of some cases where you can't just jump outside of exceptions blocks. Have fun with Cecil, and pay us a visit on the mono-cecil google group if you have any question or issue. The multiple exit points issue is not really hard to fix, just replace all the `ret`s opcode to an unconditional branch to a block emitted at the end of the method. You’ll just have to take care of some cases where you can’t just jump outside of exceptions blocks.

Have fun with Cecil, and pay us a visit on the mono-cecil google group if you have any question or issue.

]]> By: Richard Flamsholt http://lostechies.com/gabrielschenker/2009/11/26/writing-a-profiler-for-silverlight-applications-part-1/#comment-197 Richard Flamsholt Mon, 30 Nov 2009 13:40:21 +0000 /blogs/gabrielschenker/archive/2009/11/26/writing-a-profiler-for-silverlight-applications-part-1.aspx#comment-197 I'm afraid not: CIL-methods often has multiple returns. It's an interesting project - it entails some hairy details but it's surely possible. Some good advice would be to test out your instrumentation on a large body of .NET code, which will reveal stuff like e.g. the multiple-returns. NASA's World Wind app is a good candidate. I’m afraid not: CIL-methods often has multiple returns.

It’s an interesting project – it entails some hairy details but it’s surely possible. Some good advice would be to test out your instrumentation on a large body of .NET code, which will reveal stuff like e.g. the multiple-returns. NASA’s World Wind app is a good candidate.

]]> By: Gabriel N. Schenker http://lostechies.com/gabrielschenker/2009/11/26/writing-a-profiler-for-silverlight-applications-part-1/#comment-196 Gabriel N. Schenker Sat, 28 Nov 2009 18:16:59 +0000 /blogs/gabrielschenker/archive/2009/11/26/writing-a-profiler-for-silverlight-applications-part-1.aspx#comment-196 @Juanma: as far as I have seen a (C#) method having multiple exit points is converted to a method with a single exit point in MSIL @Juanma: as far as I have seen a (C#) method having multiple exit points is converted to a method with a single exit point in MSIL

]]> By: Juanma http://lostechies.com/gabrielschenker/2009/11/26/writing-a-profiler-for-silverlight-applications-part-1/#comment-195 Juanma Fri, 27 Nov 2009 08:17:30 +0000 /blogs/gabrielschenker/archive/2009/11/26/writing-a-profiler-for-silverlight-applications-part-1.aspx#comment-195 Very interesting. How are you planning to handle methods with multiple exit points? Very interesting.

How are you planning to handle methods with multiple exit points?

]]>