[vtkusers] Delete vtkObjects
Robbie Jaeger
rjaeger at gmail.com
Thu Aug 14 15:53:02 EDT 2008
Hi,
don't all VTK objects have a counter which keeps track of how many times it
has been referenced by other objects? Could that be the reason that it
wouldn't be deleted (if # referencing objects > 0) ? Can you check that
variable to see if there is any chance its value is greater than zero?
I'm really new to VTK so if this is completely wrong, I apologize.
On Thu, Aug 14, 2008 at 9:55 AM, David Cole <david.cole at kitware.com> wrote:
> You must be creating some other vtkObjects from somewhere else in your code
> that are not being cleaned up.
> Can you set a breakpoint in the vtkObject::vtkObject constructor and see
> where you are creating them from? You should only hit that breakpoint from
> your Init_3D method if that is the only place where you are creating them
> from...
>
>
> On Thu, Aug 14, 2008 at 11:53 AM, Andi2008 <mail-andi at web.de> wrote:
>
>>
>> No, the DebugLeakReport appears when i close the application. Therefore i
>> just include the vtkDebugLeaks.h
>>
>> the method that instantiates and delete the surrounding class is called
>> from
>> from a combobox and the whole method is:
>>
>> void ImageCtrl::Init_3D(void)
>> {
>> ImageVolumeData *vtkModul = new ImageVolumeData();
>>
>> delete vtkModul;
>>
>> return;
>> }
>>
>> The class ImageCtrl is derived from CCmdTarget and handles the
>> Messagerouting in the program.
>>
>> I debugged thought the code and watched the ProcessExplorers
>> PrivateByteHistory of the running programm. I did the same in the
>> vtkexamples and there everything works fine. I know it sounds strange and
>> obviously it is a mistake on my side.
>>
>> Does anybody has an idea
>> Andi
>>
>>
>>
>>
>>
>>
>> David Cole wrote:
>> >
>> > Where is your code that instantiates/deletes the surrounding C++ class?
>> > Are you saying that instantiating one of those and then deleting it
>> > results
>> > in a VTK mem leak report?
>> >
>> >
>> > On Thu, Aug 14, 2008 at 10:48 AM, Andi2008 <mail-andi at web.de> wrote:
>> >
>> >>
>> >> thank you for the quick answer.But my case is different. I fixed the
>> >> leaks
>> >> he
>> >> reported already by loading all vtl dlls delayed. I am using MFC too.
>> >> Visual
>> >> Studio report leaks only if i create vtkObjects.vtkDebugLeaks and the
>> >> fact
>> >> that the private byte history in the ProcessExplorer doesnt decrease
>> >> after
>> >> i
>> >> the objects tells me that the leaks come from vtk.
>> >>
>> >> I have no idea anymore what iam doing wrong.
>> >> Andi
>> >>
>> >>
>> >>
>> >> malat wrote:
>> >> >
>> >> > Andreas,
>> >> >
>> >> > I you are using MFC, I suggest to read the following post:
>> >> >
>> >> > http://www.vtk.org/pipermail/vtkusers/2007-May/090968.html
>> >> >
>> >> > The non-compiling example you posted looks fine to me.
>> >> >
>> >> > 2cts
>> >> > -Mathieu
>> >> >
>> >> > On Thu, Aug 14, 2008 at 2:41 PM, Andreas Brüning <mail-andi at web.de>
>> >> wrote:
>> >> >> Hi,
>> >> >>
>> >> >> yes it is me again and i still have questions about deleting
>> >> vtkobjects
>> >> >> :)
>> >> >> Regarding to my post from Aug/12:
>> >> >>
>> >> >> http://www.vtk.org/pipermail/vtkusers/2008-August/096455.html
>> >> >>
>> >> >> i still have problems to delete vtkobjects in my program and it is
>> not
>> >> >> only the thread issue. To encapsulate the whole vtk stuff i created
>> it
>> >> in
>> >> >> a single class. I created some vtk objects in the constructor and
>> call
>> >> >> the Delete() method in the destructor for testing. Than i created
>> one
>> >> of
>> >> >> these class and delete it while i checked the private bytes history
>> >> on
>> >> >> Sysinternals ProcessExplorer(which is by the way a great tool). I
>> also
>> >> >> enabled VTK_DEGUG_LEAKS.
>> >> >> I thought when i delete the class the whole vtk stuff is gone but it
>> >> is
>> >> >> not. i can see no memory reduction in the process explorer and the
>> >> >> DebugLeaks MessageBox tells me that always the objects remains. When
>> i
>> >> >> create a vtkPolyDataMapper there are even more. If i create more of
>> my
>> >> >> vtkClasses there are no changes in the memory but the debugleaks
>> >> >> MessageBox report that all the vtkObjects remain in memory.
>> >> >>
>> >> >> I really have no idea where is the mistake. I did the same procedure
>> >> with
>> >> >> the vtkExamples and there everything is fine and the memory always
>> >> cleans
>> >> >> up. I compared the project properies and they are the same.
>> >> >>
>> >> >> Here is the code of my very short vtkClass:
>> >> >>
>> >> >>
>> >> >> #include "StdAfx.h"
>> >> >>
>> >> >> #include "ImageVolumeData.h"
>> >> >>
>> >> >> #undef THIS_FILE
>> >> >> static char BASED_CODE THIS_FILE[]=__FILE__;
>> >> >>
>> >> >> #define new DEBUG_NEW
>> >> >>
>> >> >>
>> >> >> ImageVolumeData::ImageVolumeData(void)
>> >> >> {
>> >> >> vtk_importer = vtkImageImport::New(); //global Variables
>> >> defined
>> >> >> in the header
>> >> >> vtkflipper = vtkImageFlip::New();
>> >> >>
>> >> >> //debugger = vtkDebugLeaks::New();
>> >> >> ren = vtkRenderer::New();
>> >> >>
>> >> >> }
>> >> >>
>> >> >> ImageVolumeData::~ImageVolumeData(void)
>> >> >> {
>> >> >> vtk_importer->SetImportVoidPointer(NULL,0);
>> >> >>
>> >> >> vtkflipper->Delete();
>> >> >> vtk_importer->Delete();
>> >> >>
>> >> >> ren->Delete();
>> >> >>
>> >> >> }
>> >> >>
>> >> >> I hope someone can give me an advice, i am running out of ideas
>> >> >>
>> >> >> Thanks in advance
>> >> >> Andi
>> >> >>
>> _____________________________________________________________________
>> >> >> Der WEB.DE SmartSurfer hilft bis zu 70% Ihrer Onlinekosten zu
>> sparen!
>> >> >> http://smartsurfer.web.de/?mc=100071&distributionid=000000000066
>> >> >>
>> >> >> _______________________________________________
>> >> >> This is the private VTK discussion list.
>> >> >> Please keep messages on-topic. Check the FAQ at:
>> >> >> http://www.vtk.org/Wiki/VTK_FAQ
>> >> >> Follow this link to subscribe/unsubscribe:
>> >> >> http://www.vtk.org/mailman/listinfo/vtkusers
>> >> >>
>> >> >
>> >> >
>> >> >
>> >> > --
>> >> > Mathieu
>> >> > _______________________________________________
>> >> > This is the private VTK discussion list.
>> >> > Please keep messages on-topic. Check the FAQ at:
>> >> > http://www.vtk.org/Wiki/VTK_FAQ
>> >> > Follow this link to subscribe/unsubscribe:
>> >> > http://www.vtk.org/mailman/listinfo/vtkusers
>> >> >
>> >> >
>> >>
>> >> --
>> >> View this message in context:
>> >> http://www.nabble.com/Delete-vtkObjects-tp18980765p18983087.html
>> >> Sent from the VTK - Users mailing list archive at Nabble.com.
>> >>
>> >> _______________________________________________
>> >> This is the private VTK discussion list.
>> >> Please keep messages on-topic. Check the FAQ at:
>> >> http://www.vtk.org/Wiki/VTK_FAQ
>> >> Follow this link to subscribe/unsubscribe:
>> >> http://www.vtk.org/mailman/listinfo/vtkusers
>> >>
>> >
>> > _______________________________________________
>> > This is the private VTK discussion list.
>> > Please keep messages on-topic. Check the FAQ at:
>> > http://www.vtk.org/Wiki/VTK_FAQ
>> > Follow this link to subscribe/unsubscribe:
>> > http://www.vtk.org/mailman/listinfo/vtkusers
>> >
>> >
>>
>> --
>> View this message in context:
>> http://www.nabble.com/Delete-vtkObjects-tp18980765p18984404.html
>> Sent from the VTK - Users mailing list archive at Nabble.com.
>>
>> _______________________________________________
>> This is the private VTK discussion list.
>> Please keep messages on-topic. Check the FAQ at:
>> http://www.vtk.org/Wiki/VTK_FAQ
>> Follow this link to subscribe/unsubscribe:
>> http://www.vtk.org/mailman/listinfo/vtkusers
>>
>
>
> _______________________________________________
> This is the private VTK discussion list.
> Please keep messages on-topic. Check the FAQ at:
> http://www.vtk.org/Wiki/VTK_FAQ
> Follow this link to subscribe/unsubscribe:
> http://www.vtk.org/mailman/listinfo/vtkusers
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.vtk.org/pipermail/vtkusers/attachments/20080814/9378425d/attachment.htm>
More information about the vtkusers
mailing list