flex-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Alex Harui <aha...@adobe.com>
Subject Re: git commit: [flex-sdk] [refs/heads/develop] - FLEX-33904 memory fix
Date Sun, 29 Dec 2013 06:27:40 GMT
OK, took a look.  There is no activation object in the emptytrycatch
method.  The compiler must know there is no need for it.

On 12/28/13 9:49 PM, "Alex Harui" <aharui@adobe.com> wrote:

>I believe that an object (IIRC it was called an activation object and not
>a closure) is allocated for any try block.  I believe it tracks
>information to handle the catching of the error.  But it could be that if
>there is an empty catch block, it doesn't need that information and just
>needs to stop propagation of the error.
>Someday I'll try to look at the byte code and see what it is doing.
>On 12/28/13 9:38 PM, "Justin Mclean" <justin@classsoftware.com> wrote:
>>Even curiouser  that happens even if the catch block isn't run. Ouch.
>>Wonder if we could optimise that somehow?

View raw message