Serious bug: "finally" block is run whenever "continue" is reached

[Tal Liron]

Thank you, Atilla, your response does inspire confidence.

I will just add that it's not just me that would need this tool: none of 
us can afford to have a broken try/finally.

On 03/09/2015 05:58 PM, Attila Szegedi wrote:
> So my action plan for this is: 1. nominate the fix for inclusion into 
> a CPU release 2. whip up a small one-off tool that'd show you the 
> differences in finally inlining between 8u40 and 8u60 code. That 
> should help you find the trouble spots in your code. 3. understand 
> what caused the faulty inlining in your case that will hopefully let 
> me suggest a workaround for the time being. Attila.