https://bugzilla.novell.com/show_bug.cgi?id=362598

User [EMAIL PROTECTED] added comment
https://bugzilla.novell.com/show_bug.cgi?id=362598#c8





--- Comment #8 from Rodrigo Kumpera <[EMAIL PROTECTED]>  2008-02-19 11:52:57 
MST ---
I think it would be better to have a better heuristics for method inlining than
just bytecode size, this way this kind of issue would not exist.

One thing that makes me wonder is why we don't cache inlining failures, which
makes inlining more expensive than it should be. If we cache failures, it would
be ok to inline larger methods as the failure rate would not be an issue.

Once the verifier is integrated, the cost of calling mono_method_to_ir will
increase significantly and caching inline failures might be one way to mitigate
this cost. If we are going to that, we might just store a better metric that
just bytecode size for the inlining heristics - opcode count, for example.


-- 
Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the QA contact for the bug.
_______________________________________________
mono-bugs maillist  -  [email protected]
http://lists.ximian.com/mailman/listinfo/mono-bugs

Reply via email to