How to release memory in managed C++?

Hi guys,
We meet a memory leak issue when to invoke managed C++.
In our project, we use a managed C++ method to invoke a native C++ method (through C++ interop), and by using of
a unit test case to find the different memory usage between managed and natived.
We believe the managed C++ consume some more memory and it will need to create a managed/proxy object for each unmanaged object. I have tried
to call GC.Collect() manually, but it doesn't work memory out.
Following is our target managed method, do you have any ideas to release memory?
IProxyList<INamedElement^>^ NamedElementProxy::GetAllMembers() { return gcnew SetOfNamedElementProxy<INamedElement^>(metaNamedElement->GetAllMembers()); }
Thanks in advanced

Thanks Viorel for your quick response!
I tried to use "delete[]", but it doesn't work either.
void NamedElementProxy::ReleaseProxy(IProxyList<INamedElement^>^ myProxy)
delete[] myProxy;
I also did another try to verify if the pointer address is different between native c++ object and managed c++ object, I find the answer is negative,
the two objects have the same pointer address, take the following code for example, allMembers0 and allMembers1 point to the same address.
IProxyList<INamedElement^>^ NamedElementProxy::GetAllMembers(ManagedProxy::Enums::ElementKind kinds, AttributeMask^ required)
ReferenceTo<Meta::SetOfNamedElement> allMembers0 = metaNamedElement->GetAllMembers((Meta::ElementKind)kinds, *required->ToNoGC());
IProxyList<INamedElement^>^ allMembers1 = gcnew SetOfNamedElementProxy<INamedElement^>(allMembers0);
return allMembers1;
I am confused why the release from managed c++ doesn't help to release native c++?

Similar Messages

How do I release memory when done with a large Image?

I've got a sample program here. Enter a filename of a .jpg file, click the button and it will load and display a thumbnail of it. However memory is not released so repeatedly clicking the button will let you watch the memory use grow and grow. What should be done in the code to release the large original image once the thumbnail is obtained? Here's the class:
import java.awt.*;
import java.awt.event.*;
import java.awt.image.*;
import java.io.*;
import java.util.Iterator;
import javax.imageio.*;
import javax.imageio.stream.*;
import javax.swing.*;
public class ImageMemoryLeak extends JFrame implements ActionListener {
     private JLabel thumbnail = null;
     private JTextField tf = null;
     private JButton button = null;
     public static void main(String[] args) {
          ImageMemoryLeak leaker = new ImageMemoryLeak();
          try { UIManager.setLookAndFeel(UIManager.getSystemLookAndFeelClassName()); }
          catch (Exception ex) { }
          leaker.showDialog();
     private void showDialog() {
          setDefaultCloseOperation(JFrame.EXIT_ON_CLOSE);
          setBounds(100,100,200,200);
          setBackground(new Color(255,255,255));
          Container cont = getContentPane();
          cont.setBackground(Color.lightGray);
          cont.setLayout(new FlowLayout());
          thumbnail = new JLabel("thumbnail here");
          cont.add(thumbnail);
          tf = new JTextField("type filename here");
          cont.add(tf);
          button = new JButton("Load Image");
          button.addActionListener(this);
          cont.add(button);
          pack();
          setVisible(true);
     public void actionPerformed(ActionEvent e) {
          if (e.getSource() == button) {
               String fname = tf.getText();
               File f = new File(fname);
               if (f.exists()) {
                    try {
                         // This is where a file is loaded and thumbnail created
                         Iterator<ImageReader> iter = ImageIO.getImageReadersByFormatName("jpeg");
                         ImageReader imgrdr = iter.next();
                         ImageInputStream iis = ImageIO.createImageInputStream(f);
                         imgrdr.setInput(iis, true);
                         ImageReadParam param = imgrdr.getDefaultReadParam();
                         BufferedImage fullSizeImage = imgrdr.read(0, param);
                         imgrdr.dispose();     // is this enough?
                         iis.close();
                         int thWidth = 150;
                         int thHeight = 150;
                         int w = fullSizeImage.getWidth(null);
                         int h = fullSizeImage.getHeight(null);
                         double ratio = (double)w/(double)h;
                         if (w>h) thHeight = (int)((double)thHeight /ratio);
                         else if (w<h) thWidth = (int)((double)thWidth * ratio);
                         BufferedImage thumbImage = new BufferedImage(thWidth, thHeight, BufferedImage.TYPE_INT_RGB);
                         Graphics2D graphics2D = thumbImage.createGraphics();
                         graphics2D.setRenderingHint(RenderingHints.KEY_INTERPOLATION,
                         RenderingHints.VALUE_INTERPOLATION_BILINEAR);
                         graphics2D.drawImage(fullSizeImage, 0, 0, thWidth, thHeight, null);
                         // done with fullSizeImage now - how to release it though?
                         fullSizeImage.flush();     // doesn't seem to do the trick
                         ImageIcon oldicon = (ImageIcon)thumbnail.getIcon();
                         if (oldicon != null) oldicon.getImage().flush();
                         ImageIcon icon = new ImageIcon(thumbImage);
                         thumbnail.setIcon(icon);
                         thumbnail.setText("");
                         pack();
                    } catch (IOException e1) {
                         e1.printStackTrace();
               else {
                    thumbnail.setText("file not found");
}

lecisco wrote:
AndrewThompson64 wrote:
lecisco wrote:
I don't mind having GC getting to it but it's not doing so now. .. Isn't it? GC is only called when the JRE feels it is necessary. Definitive evidence of a memory leak is generally revealed by an OutOfMemoryError. Does the code throw OOMEs?Excellent point. I tried loading over and over to force an OOME. I found that after the memory footprint grew to about 750mb, it reset down to about 130mb again, so it seems that GC does eventually kick in. Perhaps what I have is fine.
That question brings me to the code sample. Typing an image file name in a text field is soooo 1980s. It would take a long time and much hard work on the part of the person testing it, in order to get to an OOME.In my actual application I'm using drag-and-drop to specify the image. The text field was just to simplify the code sample. How you get the file location is not relevant for the question - it's how the resources are released. If you're saying my code is fine regarding and shouldn't be holding onto any resources, that's great. I'm not saying that. So far I've not looked closely at the code, and am no expert on resource caching in any case.
..I am looking to confirm I'm following best practices regarding image resources.Good show. There is too much rubbish code out there.
..With the code sample, you just have to specify one .jpg file and repeatedly click the button to see the memory grow, so there's no burden to type a new file each time.Oh right, my bad. I had presumed it required a different image each time. Still (grumbles) a file chooser to select the image file would not have gone astray - for us lazy types. ;)

How to release the budget in SAP Funds Management

Hi
How to release the budget in SAP Funds Management. Please let me know the proceedure and the transaction code
Tks

Hi,
The answer depends on if you are using Former Budgeting or BCS.
1) If you are using Former Budgeting:
a. You should use transaction FR51 to enter Release or FR55 to distribute it.
2) If you are using BCS:
a. Within FMBBC transaction you have the option to release the budget, use document type as REL (release).
b. You can activate the automatic release for each budget profile in Customizing of Funds Management Government. To do this, in IMG select Funds Management Government ® Budgeting and Availability Control (Former Budgeting) ® Budget Profiles (Former Budgeting) ® Set Up Budget Profiles.
With the automatic release, the corresponding background releases for expenditures FM account assignments with budget transfers are also posted. Take note that the maximum amount that this automated release can transfer from the sender is the amount that has already been released.
If you do not activate the automated release, the maximum amount that can be transferred from the sender is that which has not yet been released.
c. You can use the mass transaction for release, which is FMMPRELE.
I hope that this answer your inquiry.
Best Regards,
Vanessa Barth.

How to see memory setting of a managed server

Hello,
Can anybody explain me how to see Memory settings of a managed server instance in weblogic server using unix box
Thanks
Ravi

Hi,
The memory setting on the WC_Spaces managed servers can be seen as mentioned below.
ps -eaf | grep Spaces
Above command shows something like this at the top of the output:
/a01/app/oraweb/middle/jrrt-3.1.2-1.6.0/bin/java -jrockit -Xms512m -Xmx1024m -Dweblogic.Name=WC_Spaces1 ...
Regards
Sreedhar

How to reduce CompositionElements/RAM consumption, changing MediaElement won't release memory.

Hello there,
First of all thanks for such a great framework, its pretty nice architected and really helpfull.
I've been using media composition elements In custom made player experience,
What I came across and can't get rid off is memory usage by the flash player whilst media elements are in series and parallel.
Whenever I try to remove media elements from composition element,
or add/change/remove mediaElement it won't affect player to release memory.
'nulling' player media reference, removing element from media container, also won't help.
Aprox. adding up to twenty media elements causes 400-500 Mb of RAM usage, the more lements the more RAM..
Could guru's tell me how can I reduce/avoid such an issue, any workaround? (Is that issue related to flash player VideoObject/NetStream scope thats out of OSMF?)
Any advance will be apreciated.Thank you.
Env. Flex SDK 4.1 + OSMF 1.5 release
Pseudo code as follows:
SomeMediaFactory
.. function constructMediaElement():MediaElement
    var aggrElement:ParallelElement = new ParallelElement();
    var serial1:SerialElement = new SerialElement();
    var serial2:SerialElement = new SerialElement();
    var videoOfSerial1:LightweightVideoElement = new LightweightVideoElement(new URLResource("some.media1.mp4");
    var videoOfSerial2:LightweightVideoElement = new LightweightVideoElement(new URLResource("some.media2.mp4");
    //add to serial
    serial1.addChild(video1);
    serial1.addChild(video2);
    serial1.addChild(video5);
    serial2.addChild(video6);
    serial2.addChild(video7);
    serial2.addChild(video10);
    //add to parallel
    aggrElement.addChild(series1);
    aggrElement.addChild(series2);
    return aggrElement;
//release container and media
...function releaseMedia()
    if (mediaPlayer.media)
        mediaContainer.removeMediaElement(mediaPlayer.media)
    mediaPlayer.media = null;
...function constructMedia()
    releaseMedia();
    var newMedia:MediaElmnt = SomeMediaFactory.constructMediaElement();
    mediaPlayer.media = newMedia;
    mediaContainer.addMediaElement(newMedia)
..function init()
    constructMedia();
.. function onSomeEventThatWeNeedChangeMediaSource()
     releaseMedia()..
     mediaPlayer.media = Somefactory.newMedia();

Hi msrustem,
It looks like there's indeed a memory build up when using multiple media elements in a serial composition.
I've currently logged a bug in Jira (link: http://bugs.adobe.com/jira/browse/FM-1227), and it will go throught a review process (hence I'm not sure when the fix will be out, but as soon as it is, I'll let you know).
Off the top of my head it looks like the netStreams used to create the serial composition are not reused, but rather new ones are added and cached with every new media element that starts playing, and that could be the cause. However developers need to take a closer look.
Feel free to check out the bug description and provide additional information if necessary.

How is PeopleSoft Memory Managed?

I'm running a custom app engine with one PeopleCode section which is very brief and simple calls an app package. This app package calls another app package that is using a CI. The process is failing after processing approximately 10,000 out of 20,000 rows (the actual number of rows varies from run to run, same data), with a simple out of memory error. I've had our SysAdmin increase the memory for Java without any change. So, my question is how does PeopleSoft manage it's memory. Is it best to cancel the CI after each row or keep one instance and do a new get/create (I believe that's doable)? Should I should I re-instantiate the app package class objects after each row, etc. Does anyone have any clue as to how the memory is managed?

Jim - As to fine tuning, I probably won't get around to it. I suspect that you are right about somewhere there's a tipping point. But the CI, is CRM's Consumer Component, so there's just a lot to it, which is why it takes a while to process 20,000+ rows. To answer your restart question, it's probably worth describing the higher levels of how it all fits together.
The way we use CRM, requires that we are frequently required to load large amounts of data from spreadsheets into various components - Users, Workers, Consumers, configuration tables, etc. Often times there's some complex logic associated with processing the raw data. So, to accommodate that need we started out by building an almost abstract packages/class to handle all the common needs of opening, reading and manipulating an input file, trap errors, write to a custom log, etc. We then extend this class to meet the specific needs of a given load. At the same time since we were frequently using the same code for a given CI, we started creating custom classes for each CI. So that all this would be reusable whether we were doing a load or needing to use it somewhere else. Somewhere along the ling PeopleSoft has probably caught up to what we've been doing, but we're too invested in our own methodology to take a look. But, since we were always loading from files, our users wanted to see the progress each process was making. We incorporated logging each row processed out into our logs. (I know this is not very efficient, but when users were watching very long running processes, they wanted to be able to know it was making progress, so we forced the app engine/classes to write to the log by opening and closing the file.) I guess this a very wordy way to say we don't use restart. We just remove the processed rows from the file and start again. I'm sure we can find more efficient ways to do all this, given the added features of tools over the years, but we can very quickly turn out new CI loading processes, and we don't have the time to rework what we're currently using.
Thanks for the input.
I should also tell you that your original blogs on 64 Base encoding enabled us to figure out how to get all our attachments into our document app.
Thanks,
Russ
P.S. I forgot to answer your question. Yes basically you psuedocode is what we're doing. But, instead of ci = null, we use ci.cancel(). I assume that has the same affect?
Edited by: user4414415 on Apr 11, 2012 12:51 PM

Plug-in request to release memory in Photoshop

Photoshop has never released memory. I read about 10 discussions about it, with entertaining side-steps to virus scanners and video drivers, and it all comes down to the same thing. Photoshop does not release memory. As Adobe knows this since a couple versions back and never patched or updated it in a new release, it would therefore be great if some company could pick up the task to write a plug-in for this. Almost every serious Photoshop user will buy that plugin as the only other option is to buy more memory.
Just a little illustration for newcomers what the problem is. Here's what Photoshop uses according to the task manager:
start Photoshop -> 110mb
open a 2.65gb psb -> 7975mb
close the 2.65gb psb -> still 7975mb
open a different 2.07gb psb -> 10106mb and any edit goes tedious slow because it is partly scratching because it did not release the memory when closing the 2.65gb file. The only solution was that I had to restart Photoshop after I closed the first psb. This is by the way with 16gb ram on win 7 64, where photoshop is assigned 10gb (67%).
The same thing happens if you Automate a Batch because with every opened file the memory never is given back until Photoshop ends up on the scratch disc slowing down to about one third the original speed if it isn't slower.
Anyway, I really hope there are some talented coders out there, using Photoshop, who would want to developer a plug-in. I have 16gb ram in my computer and the only program that uses that is Photoshop. I would have never upgraded my ram if Photoshop released memory.

Why do you think it needs to work differently than it does? What specific problem are you having?
If you want the best possible Photoshop performance, having it allocate then reuse its own memory (without deallocation) is more efficient than returning it to the operating system, then requesting it again later. It doesn't sound right at first, but the Adobe engineer Chris Cox has said that they've explicitly tried it the other way and it just doesn't work as well. Your basic assumption (that Photoshop is doing something wrong) is simply flawed.
Now, if you need your system to be doing a number of RAM-intensive things at the same time, including Photoshop, you have to outfit it with sufficient RAM that it can work with Photoshop using all that you have allowed (in the Performance preferences) PLUS the other RAM usage you anticipate. In reality, few folks use multiple applications simultaneously that are all RAM-intensive, and those who do simply buy huge amounts of it. I personally outfitted my system with 48GB and it's never starved for RAM (and I do multitask a lot).
There's little hope for a plug-in that will cause Photoshop to release all of its RAM. The only way I know of for a plug-in to affect Photoshop's memory usage, and that is to provide a large (up to 2GB because it's a signed 32 bit field) estimate of how much RAM a plug-in will need in the gFilterRecord->bufferSpace field. That field is 32 bits no matter whether the 32 bit or 64 bit Photoshop is running, so the most a plug-in could possibly ask Photoshop to release (indirectly) is 2GB.
-Noel

CS4 32 Bit Not Releasing Memory

Photoshop CS4 does not seem to release memory until I close the software. As an example I will open a D3x image file and maybe work on 3 layers of the image and at this point Photoshop is using up to 3GB of system memory.
I then take this file flatten it and reduce it down to around 1333px and in Task Manager Photoshop is still using up 3GB of system memory. Only after about 5-10 minutes does it then drop down to 1.5 GB but keep in mind I only have a file that's about 2.4 MB in size open.
Could this be due to the fact that I am running the 32-bit version of CS4 under a 64 bit OS? Any other ideas why Photoshop could be doing this?

Chris
I was wondering if you could help me. I have Photoshop CS4 and it works great after a fresh boot but hours later it annoyingly slower. I have to reboot to enjoy it again. Do you have any suggestions? I have Defragged, run spybot and cleared memory and nothing seems to make any difference.I was wondering if it had something to do with Memory and how it used or leadked. Thanks, Steve

How to cleam memory in InDesign.

Dear all,
We are developing a extension for Adobe InDesign and after many months of development we are still stuck with various things we cannot find a proper solution to them. The extension exports a lot of images, texts, creates new pages, basically we are exporting all the content in a document as long as 300-400 pages in different assets, in JPG and PNG formats. So is really memory consuming
When running the extension we have developed, the main thread is bloqued and the InDesign is not responding and if the export we are doing is really large it just crashes InDesign. We would like to reléase memory after each page is finished, but we cannot do so. Once the complete export is finished, then memory is released, but not while the export is happening
As we export the memory is increasing more and more without releasing anything, we have tried different options, but seems like we never get to release memory as everything is blocked
How can we release memory while the extension is running ? Or is there a way to stop the extension each iteration so it releases memory?
We use the extension manager and flash builder.
Thanks!
Cesar.

Hi Cesar,
I'm also making an export script, (not extension), mine seems to be quite different to yours but we'll might see when both are published.
Off the point.
The first thing you have to be aware of is using flash builder and extension manager, If you are making a flash based extension, from what I understand it's not going to work for the next version of InDesign which is switching to HTML5 based panels and dropping support for flash based panels (you'll still need the flash based panels to support the current version (9) and below. This is going to be a major pain for all extensions for the next lot of years that all of them will need 2 completely different interface bases to be compatible with CC(1), CS6 (which is going to be around for a long time for obvious reasons) and below on the one side and CC(2) + on the other. I can't actually blame Adobe for this one as there should be some serious benefits in the HTML5 panels.
Back to the point.
See Ariel's Phenomenal optimization technique! which is based on the same Idea of breaking down the script into small parts.
See also Re: Big performance issue while removing tabs by indents including the link brought by Xavier in 7 over there. Loic reduced his script run time from 6 hours + to 90 seconds by using efficient techniques.
I made a script for making swatch books (sort of thing than Pantone publish - involved a quite a bit of maths and color conversions) it was workable for 3,000 - 4,000 swatches but when it got to 15,000 or so swatches the machines wouldn't budge one bit. Separating the output into separate document (which any I needed to do because the proofing and print machines can only handle up to 27 spot colors) resulted in the script working very efficiently.
As Uwe mentioned redraw can make a big difference and even more so hidden documents app.open(myFile, false)
Using myObjectCollection.everyItem().getElements().slice(0). Very significant for big collections.
Better still don't make big collections in the first place. Don't if possible try process all the export items in the document myItems = doc.allGraphics or myItems = doc.spreads.everyItem().allGraphics rather make a collection myItems = doc.spread[n].allGraphics this will take up much less memory and work much quicker.
Try playing with $.memCache see if increasing or reducing the memory allocation makes a difference. Please let me know if it does!
There's a very lot of methods that would depend on the actual script the differences can be huge, from scripts that just won't move to taking less than 2 minutes.
Don't forget to report back with the results
Trevor

How to configure Oracle Enterprise Manager for ASM RAC Database ?

Dears,,
We have two databases (Primary & Standby), each database has two instances
Database version: Oracle Database 10g Enterprise Edition Release 10.2.0.4.0 - 64bit
How to configure Oracle Enterprise Manager for this environment ?
I need documentation for this please.
Many thanks & Regards,,

Assuming an agent is running on the servers you want to monitor,
navigate to the Agent home page (Via setup --> Agent)
When in the agent home page select 'Add Database'and press [Go]
Assuming yopur database now gets recognized, select the Configure icon and enter the password for dbsnmp.
When done, press [Ok] to return to the agent home page.
Regards
Rob
http://oemgc.wordpress.com

Release memory for nested top-level classes

Hello experts.
I have a top-level class which consists of several static methods. Inside such a method I create several new instances of a nested top-level class inside my original class. Later I want to release one of these instances i.e. destroy instance and free memory. How can I do that?
Thanks.

sure, but I do not see any references to this nested top-level class. With "Releasing" I mean something like directly assigning a NULL value to the nested top-level clas in order to enable GC.
But as I said there are no references and while monitoring the memory I see that the GC does not collect some memory. (I know that GC is not reliable, but after ten attempts of this scenario the memory allocation is equal the whole time).
I guess the GC does not work with classes, but the classloader. But how can I control releasing memory in this scenario?
Any further hints?

How to Release jstring (C++ dll calls Java)

I'm trying to call a method from a Dll (written in C++) to Java. Pseudo might look like this
String cstr = "abcdefghijklmnopqrstuvwxyz";
jstring jstr = NewStringUTF(cstr);
CallJavaMethod(classid, methodid, jstr);
// so problem is how to release this jstr because it is causing memory leaksI've tried DeleteLocalRef and ReleaseStringUTFChars(jstr, cstr) but they don't work and Memory keeps on increasing.
Can anybody suggest a solution plz

...but they don't work and Memory keeps on increasing.
Can anybody suggest a solution.Looking at the rest of your code too?
You might also want to examine the assumption that you are making when you say "Memory keeps on increasing". Java can keep allocating memory from the OS and never give it back. That has nothing to do with a memory leak (either in C or java.) So unless you have already excercised your application without JNI or you are using a profiling tool then you might want examine if 'increasing memory' is the same as 'memory leak'.

PS Touch Android does not release memory when closed.

I have to force closure with Task Manager on my Samsung Note 2 running JB 4.1. Why and when will it be fixed?

Hi Ignacio,
Sorry for the lack of information - I had typed more complete description, but had trouble posting it.
The app is installed on a Samsung Galaxy Note 2 phone that is running Jellybean 4.1.1.
The only way I can see to exit the application is by using the "back" soft key under the screen on the right. Some apps have an "exit" choice under the "menu" soft key, but that key is not active with PSTouch.
I see the same issue with Google Chrome. The menu key is active with this app, but it does not have an "exit" choice.
With both these apps, I have to go to the "Active Apps" tab of the Task Manager to release memory.
There is one difference with PS Touch. Google Chrome always starts the same way whether or not I release it's memory. With PS Touch, if I release memory after closing with the back key, the next time I start it I get an initial royal blue screen with the PS Touch icon in the middle followed by a second screen with a menu in the middle titled "New Project From". If I exit at that point, it takes two presses of the back key to first close the menu and then exit the program and return to a home screen. If I do not release memory, the next time I start PS Touch, it goes directly to the second screen, but without the "New Project From" menu.
I hope this gives you the information you need.
Pete

How to calculate memory usage base on graphic utilization

Dear All ,
We have t2000 server with solaris 10 and 15 zones inside , and install SMC server include module ,Harddware configuration is 16 Gb Memory , 4 x 72 gb Hardisk and Swap 4Gb .,from menu Manage container manager , we select host of the server then click utilization , but i see memory usage 19759 Mb , How to calculate memory from this graph ? cause maximum ream RAM only 16 Gb in our server.
Regards
Hadi

PL/SQL collections are stored in the PGA. So you can monitor the PGA utilization of the session(s) to see how much PGA they use.
SELECT sid, name, value
FROM v$statname name
JOIN v$sesstat using (statistic#)
WHERE name.name in ('session pga memory', 'session pga memory max' )That will show you, for each session, the current PGA consumed by the session and the high water mark of PGA consumption by that session. You can join to V$SESSION and add additional predicates to narrow things down to the particular sessions you are interested in.
Justin

How to integrate free download manager and mozilla firefox?

how to integrate free download manager and mozilla firefox?

The Free Download Manager version 3.5 will support Firefox 4. The developers have made a release candidate available, for details including a download link see http://www.freedownloadmanager.org/board/viewtopic.php?f=1&t=15061

How to release memory in managed C++?

Similar Messages

Maybe you are looking for