Search Results

Search found 16731 results on 670 pages for 'memory limit'.

Page 599/670 | < Previous Page | 595 596 597 598 599 600 601 602 603 604 605 606  | Next Page >

• Using SMO to call Database.ExecuteNonQuery() concurrently?

  - by JimDaniel

  I have been banging my head against the wall trying to figure out how I can run update scripts concurrently against multiple databases in a single SQL Server instance using SMO. Our environments have an ever-increasing number of databases which need updating, and iterating through one at a time is becoming a problem (too slow). From what I understand SMO does not support concurrent operations, and my tests have bore that out. There seems to be shared memory at the Server object level, for things like DataReader context, keeps throwing exceptions such as "reader is already open." I apologize for not having the exact exceptions I am getting. I will try to get them and update this post. I am no expert on SMO and just feeling my way through to be honest. Not really sure I am approaching it the right way, but it's something that has to be done, or our productivity will slow to a crawl. So how would you guys do something like this? Am I using the wrong technology with SMO? All I am wanting to do is execute sql scripts against databases in a single sql server instance in parallel. Thanks for any help you can give, Daniel

  Read the article

• How would I instruct extconf.rb to use additional g++ optimization flags, and which are advisable?

  - by mohawkjohn

  I'm using Rice to write a C++ extension for a Ruby gem. The extension is in the form of a shared object (.so) file. This requires 'mkmf-rice' instead of 'mkmf', but the two (AFAIK) are pretty similar. By default, the compiler uses the flags -g -O2. Personally, I find this kind of silly, since it's hard to debug with any optimization enabled. I've resorted to editing the Makefile to take out the flags I don't like (e.g., removing -fPIC -shared when I need to debug using main() instead of Ruby's hooks). But I figure there's got to be a better way. I know I can just do $CPPFLAGS += " -DRICE" to add additional flags. But how do I remove things without editing the Makefile directly? A secondary question: what optimizations are safe for shared objects loaded by Ruby? Can I do things like -funroll-loops? What do you all recommend? It's a scientific computing project, so the faster the better. Memory is not much of an issue. Many thanks!

  Read the article

• Building "isolated" and "automatically updated" caches (java.util.List) in Java.

  - by Aidos

  Hi Guys, I am trying to write a framework which contains a lot of short-lived caches created from a long-living cache. These short-lived caches need to be able to return their entier contents, which is a clone from the original long-living cache. Effectively what I am trying to build is a level of transaction isolation for the short-lived caches. The user should be able to modify the contents of the short-lived cache, but changes to the long-living cache should not be propogated through (there is also a case where the changes should be pushed through, depending on the Cache type). I will do my best to try and explain: master-cache contains: [A,B,C,D,E,F] temporary-cache created with state [A,B,C,D,E,F] 1) temporary-cache adds item G: [A,B,C,D,E,F] 2) temporary-cache removes item B: [A,C,D,E,F] master-cache contains: [A,B,C,D,E,F] 3) master-cache adds items [X,Y,Z]: [A,B,C,D,E,F,X,Y,Z] temporary-cache contains: [A,C,D,E,F] Things get even harder when the values in the items can change and shouldn't always be updated (so I can't even share the underlying object instances, I need to use clones). I have implemented the simple approach of just creating a new instance of the List using the standard Collection constructor on ArrayList, however when you get out to about 200,000 items the system just runs out of memory. I know the value of 200,000 is excessive to iterate, but I am trying to stress my code a bit. I had thought that it might be able to somehow "proxy" the list, so the temporary-cache uses the master-cache, and stores all of it's changes (effectively a Memento for the change), however that quickly becomes a nightmare when you want to iterate the temporary-cache, or retrieve an item at a specific index. Also given that I want some modifications to the contents of the list to come through (depending on the type of the temporary-cache, whether it is "auto-update" or not) and I get completly out of my depth. Any pointers to techniques or data-structures or just general concepts to try and research will be greatly appreciated. Cheers, Aidos

  Read the article

• Explanation of `self` usage during dealloc?

  - by Greg

  I'm trying to lock down my understanding of proper memory management within Objective-C. I've gotten into the habit of explicitly declaring self.myProperty rather than just myProperty because I was encountering occasional scenarios where a property would not be set to the reference that I intended. Now, I'm reading Apple documentation on releasing IBOutlets, and they say that all outlets should be set to nil during dealloc. So, I put this in place as follows and experienced crashes as a result: - (void)dealloc { [self.dataModel close]; [self.dataModel release], self.dataModel = nil; [super dealloc]; } So, I tried taking out the "self" references, like so: - (void)dealloc { [dataModel close]; [dataModel release], dataModel = nil; [super dealloc]; } This second system seems to work as expected. However, it has me a bit confused. Why would self cause a crash in that case, when I thought self was a fairly benign reference more used as a formality than anything else? Also, if self is not appropriate in this case, then I have to ask: when should you include self references, and when should you not?

  Read the article

• How to get the size of a binary tree ?

  - by Andrei Ciobanu

  I have a very simple binary tree structure, something like: struct nmbintree_s { unsigned int size; int (*cmp)(const void *e1, const void *e2); void (*destructor)(void *data); nmbintree_node *root; }; struct nmbintree_node_s { void *data; struct nmbintree_node_s *right; struct nmbintree_node_s *left; }; Sometimes i need to extract a 'tree' from another and i need to get the size to the 'extracted tree' in order to update the size of the initial 'tree' . I was thinking on two approaches: 1) Using a recursive function, something like: unsigned int nmbintree_size(struct nmbintree_node* node) { if (node==NULL) { return(0); } return( nmbintree_size(node->left) + nmbintree_size(node->right) + 1 ); } 2) A preorder / inorder / postorder traversal done in an iterative way (using stack / queue) + counting the nodes. What approach do you think is more 'memory failure proof' / performant ? Any other suggestions / tips ? NOTE: I am probably going to use this implementation in the future for small projects of mine. So I don't want to unexpectedly fail :).

  Read the article

• What is the fastest way to filter a list of strings when making an Intellisense/Autocomplete list?

  - by user559548

  Hello everyone, I'm writing an Intellisense/Autocomplete like the one you find in Visual Studio. It's all fine up until when the list contains probably 2000+ items. I'm using a simple LINQ statement for doing the filtering: var filterCollection = from s in listCollection where s.FilterValue.IndexOf(currentWord, StringComparison.OrdinalIgnoreCase) >= 0 orderby s.FilterValue select s; I then assign this collection to a WPF Listbox's ItemSource, and that's the end of it, works fine. Noting that, the Listbox is also virtualised as well, so there will only be at most 7-8 visual elements in memory and in the visual tree. However the caveat right now is that, when the user types extremely fast in the richtextbox, and on every key up I execute the filtering + binding, there's this semi-race condition, or out of sync filtering, like the first key stroke's filtering could still be doing it's filtering or binding work, while the fourth key stroke is also doing the same. I know I could put in a delay before applying the filter, but I'm trying to achieve a seamless filtering much like the one in Visual Studio. I'm not sure where my problem exactly lies, so I'm also attributing it to IndexOf's string operation, or perhaps my list of string's could be optimised in some kind of index, that could speed up searching. Any suggestions of code samples are much welcomed. Thanks.

  Read the article

• Negative number representation across multiple architechture

  - by Donotalo

  I'm working with OKI 431 micro controller. It can communicate with PC with appropriate software installed. An EEPROM is connected in the I2C bus of the micro which works as permanent memory. The PC software can read from and write to this EEPROM. Consider two numbers, B and C, each is two byte integer. B is known to both the PC software and the micro and is a constant. C will be a number so close to B such that B-C will fit in a signed 8 bit integer. After some testing, appropriate value for C will be determined by PC and will be stored into the EEPROM of the micro for later use. Now the micro can store C in two ways: The micro can store whole two byte representing C The micro can store B-C as one byte signed integer, and can later derive C from B and B-C I think that two's complement representation of negative number is now universally accepted by hardware manufacturers. Still I personally don't like negative numbers to be stored in a storage medium which will be accessed by two different architectures because negative number can be represented in different ways. For you information, 431 also uses two's complement. Should I get rid of the headache that negative number can be represented in different ways and accept the one byte solution as my other team member suggested? Or should I stick to the decision of the two byte solution because I don't need to deal with negative numbers? Which one would you prefer and why?

  Read the article

• python crc32 woes

  - by lazyr

  I'm writing a python program to extract data from the middle of a 6 GB bz2 file. A bzip2 file is made up of independently decryptable blocks of data, so I only need to find a block (they are delimited by magic bits), then create a temporary one-block bzip2 file from it in memory, and finally pass that to the bz2.decompress function. Easy, no? The bzip2 format has a crc32 checksum for the file at the end. No problem, binascii.crc32 to the rescue. But wait. The data to be checksummed does not necessarily end on a byte boundary, and the crc32 function operates on a whole number of bytes. My plan: use the binascii.crc32 function on all but the last byte, and then a function of my own to update the computed crc with the last 1-7 bits. But hours of coding and testing has left me bewildered, and my puzzlement can be boiled down to this question: how come crc32("\x00") is not 0x00000000? Shouldn't it be, according to the wikipedia article? You start with 0b00000000 and pad with 32 0's, then do polynomial division with 0x04C11DB7 until there are no ones left in the first 8 bits, which is immediately. Your last 32 bits is the checksum, and how can that not be all zeroes? I've searched google for answers and looked at the code of several crc32 implementations without finding any clue to why this is so.

  Read the article

• How do the operators < and > work with pointers?

  - by Øystein

  Just for fun, I had a std::list of const char*, each element pointing to a null-terminated text string, and ran a std::list::sort() on it. As it happens, it sort of (no pun intended) did not sort the strings. Considering that it was working on pointers, that makes sense. According to the documentation of std::list::sort(), it (by default) uses the operator < between the elements to compare. Forgetting about the list for a moment, my actual question is: How do these (, <, =, <=) operators work on pointers in C++ and C? Do they simply compare the actual memory addresses? char* p1 = (char*) 0xDAB0BC47; char* p2 = (char*) 0xBABEC475; e.g. on a 32-bit, little-endian system, p1 p2 because 0xDAB0BC47 0xBABEC475? Testing seems to confirm this, but I thought it'd be good to put it on StackOverflow for future reference. C and C++ both do some weird things to pointers, so you never really know...

  Read the article

• Char C question about encoding signed/unsigned.

  - by drigoSkalWalker

  Hi guys. I read that C not define if a char is signed or unsigned, and in GCC page this says that it can be signed on x86 and unsigned in PowerPPC and ARM. Okey, I'm writing a program with GLIB that define char as gchar (not more than it, only a way for standardization). My question is, what about UTF-8? It use more than an block of memory? Say that I have a variable unsigned char *string = "My string with UTF8 enconding ~ çã"; See, if I declare my variable as unsigned I will have only 127 values (so my program will to store more blocks of mem) or the UTF-8 change to negative too? Sorry if I can't explain it correctly, but I think that i is a bit complex. NOTE: Thanks for all answer I don't understand how it is interpreted normally. I think that like ascii, if I have a signed and unsigned char on my program, the strings have diferently values, and it leads to confuse, imagine it in utf8 so.

  Read the article

• Code Golf: Counting XML elements in file on Android

  - by CSharperWithJava

  Take a simple XML file formatted like this: <Lists> <List> <Note/> ... <Note/> </List> <List> <Note/> ... <Note/> </List> </Lists> Each node has some attributes that actually hold the data of the file. I need a very quick way to count the number of each type of element, (List and Note). Lists is simply the root and doesn't matter. I can do this with a simple string search or something similar, but I need to make this as fast as possible. Design Parameters: Must be in java (Android application). Must AVOID allocating memory as much as possible. Must return the total number of Note elements and the number of List elements in the file, regardless of location in file. Number of Lists will typically be small (1-4), and number of notes can potentially be very large (upwards of 1000, typically 100) per file. I look forward to your suggestions.

  Read the article

• Calling a WPF Application and modify exposed properties?

  - by Justin

  I have a WPF Keyboard Application, it is developed in such a way that an application could call it and modify its properties to adapt the Keyboard to do what it needs to. Right now I have a file *.Keys.Set which tells the application (on open) to style itself according to that new style. I know this file could be passed as a command line argument into the application. That would not be a problem. My concern is, is there a way via a managed environment to change the properties of the executable as long as they are exposed properly, an example: 'Creates a new instance of the Keyboard Application Dim e_key as new WpfApplication("C:\egt\components\keyboard.exe") 'Sets the style path e_key.SetStylePath("c:\users\joe\apps\me\default.keys.set") e_key.Refresh() 'Applies the style e_key.HideMenu() 'Hides the menu e_key.ShowDeck("PIN") 'Shows the custom "deck" of keyboard keys the developer 'Created in the style application. ''work with events and response 'Clear the instance from memory e_key.close e_key.dispose e_key = nothing This would allow my application to become easily accessible to other Touch Screen Application Developers, allowing them to use my keyboard and keep the functionality they need. It seems like it might be possible because (name of executable).application shows all the exposed functions, properties, and values. I just have never done this before. Any help would be appreciated, thank you in advance.

  Read the article

• Multi-reader IPC solution?

  - by gct

  I'm working on a framework in C++ (just for fun for now), that lets the user write plugins that use a standard API to stream data between each other. There's going to be three basic transport mechanisms for the data: files, sockets, and some kind of IPC piping system. The system is set up so that for the non-file transport, each stream can have multiple readers. IE once a server socket it setup, multiple computers can connect and stream the data. I'm a little stuck at the multi-reader IPC system though. All my plugins run in threads so they live in the same address space, so some kind of shared memory system would work fine, I was thinking I'd write my own circular buffer with a write pointer and read pointers chassing it around the buffer, but I have my doubts that I can achieve the same performance as something like linux pipes. I'm curious what people would suggest for a multi-reader solution to something like this? Is the overhead for pipes or domain sockets low enough that I could just open a connection to each reader and issue separate writes to each reader? This is intended to be significant volumes of data (tens of mega-samples/sec), so performance is a must.

  Read the article

• Ember Data Sycn - LocalStorage+REST+RealTime+Online/Offline

  - by Miguel Madero

  We have a combination of requirements in terms o data access. Pre-load some reference data. We need reference data to survive browser restarts instead of just living in memory to avoid loading it all the time. I'm currently using the LocalStorageAdapter for that. Once we have it, we would like to sync changes (polling or using Socket.IO in the background and updating the LocalStorage could do the trick) There're other models that are more transactional, where we would need to directly go to the Server and get/save them. It would be nice to use something like the RESTAdapter for that. Lastly, there're some operations that should work off-line and changes should be synced later. To make it more concrete: * We pre-load vendor and "favorite products" into Local Storage. We work offline with those. * We need to sync server changes to vendor and product information. * If they search the full catalog, that requires them to be online. * When offline, we need to allow users to add something to their cart or even submit and order. We would like to queue this action and submit it when they have an Internet Connection. So a few questions are derived from this: * Is there a way to user RESTAdapter in combination with LocalStorage? * Is there some Socket.IO support? (Happy to do this part manually) * Is there Queueing support? Ideally at the Ember-Data level. I know we will have to do a lot of this manually and pull together the different lego pieces, but I wanted to ask for some perspective from experience Ember devs.

  Read the article

• OpenCL Matrix Multiplication - Getting wrong answer

  - by Yash

  here's a simple OpenCL Matrix Multiplication kernel which is driving me crazy: __kernel void matrixMul( __global int* C, __global int* A, __global int* B, int wA, int wB){ int row = get_global_id(1); //2D Threas ID x int col = get_global_id(0); //2D Threas ID y //Perform dot-product accumulated into value int value; for ( int k = 0; k < wA; k++ ){ value += A[row*wA + k] * B[k*wB+col]; } C[row*wA+col] = value; //Write to the device memory } Where (inputs) A = [72 45 75 61] B = [26 53 46 76] Output I am getting: C = [3942 7236 3312 5472] But the output should be: C = [3943 7236 4756 8611] The problem I am facing here is that for any dimension array the elements of the first row of the resulting matrix is correct. The elements of all the other rows of the resulting matrix is wrong. By the way I am using pyopencl. I don't know what I mistake I am doing here. I have spent the entire day with no luck. Please help me with this

  Read the article

• Alternatives to static methods on interfaces for enforcing consistency

  - by jayshao

  In Java, I'd like to be able to define marker interfaces, that forced implementations to provide static methods. For example, for simple text-serialization/deserialization I'd like to be able to define an interface that looked something like this: public interface TextTransformable<T>{ public static T fromText(String text); public String toText(); Since interfaces in Java can't contain static methods though (as noted in a number of other posts/threads: here, here, and here this code doesn't work. What I'm looking for however is some reasonable paradigm to express the same intent, namely symmetric methods, one of which is static, and enforced by the compiler. Right now the best we can come up with is some kind of static factory object or generic factory, neither of which is really satisfactory. Note: in our case our primary use-case is we have many, many "value-object" types - enums, or other objects that have a limited number of values, typically carry no state beyond their value, and which we parse/de-parse thousands of time a second, so actually do care about reusing instances (like Float, Integer, etc.) and its impact on memory consumption/g.c. Any thoughts?

  Read the article

• function returns after an XMLHttpRequest

  - by ashays

  Alright, I know questions like this have probably been asked dozens of times, but I can't seem to find a working solution for my project. Recently, while using jQuery for a lot of AJAX calls, I've found myself in a form of callback hell. Whether or not jQuery is too powerful for this project is beyond the scope of this question. So basically here's some code that shows what's going on: function check_form(table) { var file = "/models/"+table+".json"; var errs = {}; var xhr = $.getJSON(file, function(json) { for (key in json) { var k = key; var r = json[k]; $.extend(errs, check_item("#"+k,r)); } }); return errs; } And... as you can probably guess, I get an empty object returned. My original idea was to use some sort of onReadyStateChange idea that would return whenever the readyState had finally hit 4. This causes my application to hang indefinitely, though. I need these errors to decide whether or not the form is allowed to submit or not (as well as to tell the user where the errors are in the application. Any ideas? Edit. It's not the prettiest solution, but I've managed to get it to work. Basically, check_form has the json passed to it from another function, instead of loading it. I was already loading it there, too, so it's probably best that I don't continue to load the same file over and over again anyways. I was just worried about overloading memory. These files aren't absolutely huge, though, so I guess it's probably okay.

  Read the article

• What C++ templates issue is going on with this error?

  - by WilliamKF

  Running gcc v3.4.6 on the Botan v1.8.8 I get the following compile time error building my application after successfully building Botan and running its self test: ../../src/Botan-1.8.8/build/include/botan/secmem.h: In member function `Botan::MemoryVector<T>& Botan::MemoryVector<T>::operator=(const Botan::MemoryRegion<T>&)': ../../src/Botan-1.8.8/build/include/botan/secmem.h:310: error: missing template arguments before '(' token What is this compiler error telling me? Here is a snippet of secmem.h that includes line 130: [...] /** * This class represents variable length buffers that do not * make use of memory locking. */ template<typename T> class MemoryVector : public MemoryRegion<T> { public: /** * Copy the contents of another buffer into this buffer. * @param in the buffer to copy the contents from * @return a reference to *this */ MemoryVector<T>& operator=(const MemoryRegion<T>& in) { if(this != &in) set(in); return (*this); } // This is line 130! [...]

  Read the article

• XML Decryption Bug (referencing issue)

  - by OrangePekoe

  Hi, Needing some explanation of what exactly the decryption is doing, in addition to some help on solving the problem. Currently, when a portion of XML is encrypted, and then decrypted, the DOM appears to work correctly. We can see the element is encrypted and then see it return back once it is decrypted. Our problem lies when a user tries to change data in that same element after decryption has occurred. When a user changes some settings, data in the XML should change. However, if the user attempts to change an XML element that has been decrypted the changes are not reflected in the DOM. We have a reference pointer to the XML element that is used to bind the element to an object. If you encrypt the node and then decrypt it, the reference pointer now points to a valid orphaned XML element that is no longer part of the DOM. After decryption, there will be 2 copies of the XML element. One in the DOM as expected (though will not reflect new changes), and one orphaned element in memory that is still referenced by our pointer. The orphaned element is valid (reflects new changes). We can see that this orphaned element is owned by the DOM, but when we try to return its parent, it returns null. The question is: Where did this orphaned xml element come from? And how can we get it to correctly append (replace old data) to the DOM? The code resembles: public static void Decrypt(XmlDocument Doc, SymmetricAlgorithm Alg) { if (Doc == null) throw new ArgumentNullException("Doc"); if (Alg == null) throw new ArgumentNullException("Alg"); XmlElement encryptedElement = Doc.GetElementsByTagName("EncryptedData")[0] as XmlElement; if (encryptedElement == null) { throw new XmlException("The EncryptedData element was not found."); } EncryptedData edElement = new EncryptedData(); edElement.LoadXml(encryptedElement); EncryptedXml exml = new EncryptedXml(); byte[] rgbOutput = exml.DecryptData(edElement, Alg); exml.ReplaceData(encryptedElement, rgbOutput); }

  Read the article

• object / class methods serialized as well?

  - by Mat90

  I know that data members are saved to disk but I was wondering whether object's/class' methods are saved in binary format as well? Because I found some contradictionary info, for example: Ivor Horton: "Class objects contain function members as well as data members, and all the members, both data and functions, have access specifiers; therefore, to record objects in an external file, the information written to the file must contain complete specifications of all the class structures involved." and: Are methods also serialized along with the data members in .NET? Thus: are method's assembly instructions (opcodes and operands) stored to disk as well? Just like a precompiled LIB or DLL? During the DOS ages I used assembly so now and then. As far as I remember from Delphi and the following site (answer by dan04): Are methods also serialized along with the data members in .NET? sizeof(<OBJECT or CLASS>) will give the size of all data members together (no methods/procedures). Also a nice C example is given there with data and members declared in one class/struct but at runtime these methods are separate procedures acting on a struct of data. However, I think that later class/object implementations like Pascal's VMT may be different in memory.

  Read the article

• Should I pass a SqlDataReader by reference or not when passing it out to multiple threads.

  - by deroby

  Hi all, being new to c# I've run into this 'conundrum' when passing around a SqlDataReader between different threads. Without going into too much detail, the idea is to have a main thread fetching data from the database (a large recordset) and then have a helper-task run through this record by record and doing some stuff based upon the contents of this. There is no feedback to the recordset, it simply wades through until no records are left. This works fine, but given the nature of the job at hand it should be possible to have this job spread over different threads (CPUs) to maximize throughput (the order of execution is of no significance). The question then becomes, when I pass this recordset in a SqlDataReader, do I have to use ref or not ? It kind of boils down to the question : if I pass the object around without specifying ref, won't it create new copies in memory and have records processed n times ? Or, don't I risk having the record-position being moved forward while not all fields have been fully read yet ? The latter seems more like a 'data racing' issue and probably is covered by the lock()ing mechanism (or not?). My initial take on the problem was that it doesn't really hurt passing the variable using ref, yet as a colleague put it : "you only need ref when you're doing something wrong" =) Additionally using ref restricts me from applying a Using() construction too which isn't very nice either. I thus create a "basic" project that tackles the same approach but without the ref notation. Tests so far show that it works flawlessly on a Core2Duo (2cpu) using any number of threads, yet I'm still a bit wary... What do you experts think about this ? Use ref or not ? You can find the test-project here as it seems I can't upload it to this question directly ?!? ps: it's just a test-project and I'm new to c#, so please be gentle on me when breaking down the code =P

  Read the article

• Usage of autorelease pools for fetch method

  - by Matthias

  Hi, I'm a little bit confused regarding the autorelease pools when programming for the iPhone. I've read a lot and the oppionions seem to me from "Do-NOT-use" to "No problem to use". My specific problem is, I would like to have a class which encapsulates the SQLite3 Access, so I have for example the following method: -(User*)fetchUserWithId:(NSInteger)userId Now, within this method a SQL query is done and a new user object is created with the data from the database and then returned. Within this DB Access class I don't need this object anymore, so I can do a release, but since the calling method needs it, I would do an autorelease, wouldn't I? So, is it okay to use autorelease here oder would it gain too much memory, if this method is called quite frequently? Some websites say, that the autorelease pool is released first at the end of the application, some say, at every event (e.g. user touches something). If I should not use autorelease, how can I make sure, that the object is released correctly? Can I do a release in the fetch method and hope, that the object is still there until the calling method can do a retain? Thanks for your help! Regards Matthias

  Read the article

• CUDA - multiple kernels to compute a single value

  - by Roger

  Hey, I'm trying to write a kernel to essentially do the following in C float sum = 0.0; for(int i = 0; i < N; i++){ sum += valueArray[i]*valueArray[i]; } sum += sum / N; At the moment I have this inside my kernel, but it is not giving correct values. int i0 = blockIdx.x * blockDim.x + threadIdx.x; for(int i=i0; i<N; i += blockDim.x*gridDim.x){ *d_sum += d_valueArray[i]*d_valueArray[i]; } *d_sum= __fdividef(*d_sum, N); The code used to call the kernel is kernelName<<<64,128>>>(N, d_valueArray, d_sum); cudaMemcpy(&sum, d_sum, sizeof(float) , cudaMemcpyDeviceToHost); I think that each kernel is calculating a partial sum, but the final divide statement is not taking into account the accumulated value from each of the threads. Every kernel is producing it's own final value for d_sum? Does anyone know how could I go about doing this in an efficient way? Maybe using shared memory between threads? I'm very new to GPU programming. Cheers

  Read the article

• How to append new elements to Xml from stream

  - by Wololo

  I have a method which returns some xml in a memory stream private MemoryStream GetXml() { XmlWriterSettings settings = new XmlWriterSettings(); settings.Indent = true; using (MemoryStream memoryStream = new MemoryStream()) { using (XmlWriter writer = XmlWriter.Create(memoryStream, settings)) { writer.WriteStartDocument(); writer.WriteStartElement("root"); writer.WriteStartElement("element"); writer.WriteString("content"); writer.WriteEndElement(); writer.WriteEndElement(); writer.WriteEndDocument(); writer.Flush(); } return memoryStream; } } In this example the format of the xml will be: <?xml version="1.0" encoding="utf-8"?> <root> <element>content</element> </root> How can i insert a new element under the root e.g <?xml version="1.0" encoding="utf-8"?> <root> <element>content</element> ----->New element here <------ </root>

  Read the article

• C++ design related question

  - by Kotti

  Hi! Here is the question's plot: suppose I have some abstract classes for objects, let's call it Object. It's definition would include 2D position and dimensions. Let it also have some virtual void Render(Backend& backend) const = 0 method used for rendering. Now I specialize my inheritance tree and add Rectangle and Ellipse class. Guess they won't have their own properties, but they will have their own virtual void Render method. Let's say I implemented these methods, so that Render for Rectangle actually draws some rectangle, and the same for ellipse. Now, I add some object called Plane, which is defined as class Plane : public Rectangle and has a private member of std::vector<Object*> plane_objects; Right after that I add a method to add some object to my plane. And here comes the question. If I design this method as void AddObject(Object& object) I would face trouble like I won't be able to call virtual functions, because I would have to do something like plane_objects.push_back(new Object(object)); and this should be push_back(new Rectangle(object)) for rectangles and new Circle(...) for circles. If I implement this method as void AddObject(Object* object), it looks good, but then somewhere else this means making call like plane.AddObject(new Rectangle(params)); and this is generally a mess because then it's not clear which part of my program should free the allocated memory. ["when destroying the plane? why? are we sure that calls to AddObject were only done as AddObject(new something).] I guess the problems caused by using the second approach could be solved using smart pointers, but I am sure there have to be something better. Any ideas?

  Read the article

< Previous Page | 595 596 597 598 599 600 601 602 603 604 605 606  | Next Page >