[
Date Prev][
Date Next][
Thread Prev][
Thread Next][
Date Index][
Thread Index]
[
List Home]
|
Re: [aspectj-users] Re: aspectj and stack traces
|
When the weaving is done (compile time, load time) shouldn't make a
difference as the woven class files are the same.
-XnoInline is required because the JSR-45 information inserted into
the woven code is not correct when inlining occurs. If you use
-XnoInline you will be getting an extra class generated representing a
closure - these are small but obviously consume some 'memory' in terms
of permgen (as they are a loaded class) and a bit of heap. I've not
heard of anyone who was treading such a fine line on memory such that
switching from normal weaving to -XnoInline triggered an out of memory
or permgen issue.
cheers,
Andy
On 5 July 2010 20:06, Ashank <k_arvind_shankar@xxxxxxxxx> wrote:
>
> I don't have much experience using java debuggers but upon searching further,
> I realize that the applications woven by aspectj can still be debugged. How
> does this apply with respect to load time weaving? Any pointers, greatly
> appreciated.
>
> For compatibility with all debuggers one can apparently use the -XnoInline
> option for ajc. In general, does the -XnoInline option cause the woven
> application to use more memory than if that option were not used (I read
> that the efficiency is less for -XnoInline, but don't know details)?
>
> Thanks
> -Arvind
> --
> View this message in context: http://aspectj.2085585.n4.nabble.com/aspectj-and-stack-traces-tp2278335p2279002.html
> Sent from the AspectJ - users mailing list archive at Nabble.com.
> _______________________________________________
> aspectj-users mailing list
> aspectj-users@xxxxxxxxxxx
> https://dev.eclipse.org/mailman/listinfo/aspectj-users
>
