harmony-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Mikhail Fursov" <mike.fur...@gmail.com>
Subject Re: [drlvm][Jitrino\Jet] Jitrino memory leaks due to insufficient JFrame internal data disposing
Date Wed, 14 Nov 2007 09:27:35 GMT
please check if the problem solved

https://issues.apache.org/jira/browse/HARMONY-5111

On Nov 13, 2007 12:53 PM, Mikhail Fursov <mike.fursov@gmail.com> wrote:
> I'm going to check it today.
>
>
> On Nov 12, 2007 10:09 PM, Ilya Leviev <ilya.leviev@gmail.com> wrote:
> > Hello JIT gurus,
> >
> > I have profiled a simple EIOffice application on Harmony using
> > BoundsChecker and the tool detected more memory leaks. These memory
> > leaks fall to your area today.
> >
> > I've noticed that when JET is used for method compilation, JFrame
> > class at Jitrino\Jet does not dispose its local variables and operand
> > stack data fields (m_stack and m_vars). During simple scenario
> > (start-shutdown) it gives us 11078 leaking Val objects.
> >
> > Here is stack trace:
> >
> > JIT_compile_method_with_params->
> > Jet::compile_with_params->
> > Jet::Compiler::compile>
> >   m_bbStates[0]->jframe.init(m_infoBlock.get_stack_max(),
> >   m_infoBlock.get_num_locals());
> > Jet::JFrame::init
> >   m_stack = new Val[max_stack+1];
> >
> > I wonder if you think that these leaks might be worth to be fixed. As
> > for my personal point of view the less leaks we have, the better,
> > especially when the fix is as simple as adding delete operator. Thank
> > you for your time.
> >
> > With best regards, Ilya.
> >
>
>
>
> --
> Mikhail Fursov
>



-- 
Mikhail Fursov

Mime
View raw message