Search Results

Search found 4133 results on 166 pages for 'boost graph'.

Page 58/166 | < Previous Page | 54 55 56 57 58 59 60 61 62 63 64 65  | Next Page >

  • How to get the next prefix in C++?

    - by Vicente Botet Escriba
    Given a sequence (for example a string "Xa"), I want to get the next prefix in order lexicographic (i.e "Xb"). As I don't want to reinvent the wheel, I'm wondering if there is any function in C++ STL or boost that can help to define this generic function easily? If not, do you think that this function can be useful? Notes The next of "aZ" should be "b". Even if the examples are strings, the function should work for any Sequence. The lexicographic order should be a template parameter of the function.

    Read the article

  • TCP 30 small packets per second polutes connection with server

    - by Denis Ermolin
    I'm testing connection with flash client and cloud server(boost::asio for software) over TCP connection. My connection with server already is really poor - 120 ms ping in average. I found when i start to send packets with 2 bytes size (without tcp header) with speed 30 packets/s ping grow to 170-200 average. I think that it's really bad and my bad connection and bad cloud provider is reason for this high ping without any load. What do you think? (I tested my software and can compute about 50k packets/s so software is not a problem).

    Read the article

  • How to know if the argument that is passed to the function is a class, union or enum in c++?

    - by Narek
    I want to define an operator<< for all enums, to cout the value and print that it is an enum like this: code: enum AnyEnum{A,B,C}; AnyEnum enm = A; cout << enm <<endl; output: This is an enum which has a value equal to 0 I know a way of doing this with Boost library by using is_enum struct. But I don’t understand how it works. So that's why, in general, I am interested how to identify if the veriable is a class type, union type or an enum (in compile time).

    Read the article

  • How to copy a structure with pointers to data inside (so to copy pointers and data they point to)?

    - by Kabumbus
    so I have a structure like struct GetResultStructure { int length; char* ptr; }; I need a way to make a full copy of it meaning I need a copy to have a structure with new ptr poinnting on to copy of data I had in original structure. Is It any how possible? I mean any structure I have which contains ptrs will have some fields with its lengths I need a function that would copy my structure coping all ptrs and data they point to by given array of lengthes... Any cool boost function for it? Or any way how to create such function?

    Read the article

  • Portable Socket programming in C/C++ possible?

    - by questions
    I am thinking of creating a multi-platform portable C++ server-client application. Is it even possible while using only standard libraries? If no, what other libraries are there? Are there any improvements in this direction in C++11x? Like for threads, now we have std::threads. To make it more clear.. I want something like boost::thread, which provides multiplatform portable multithreading, for networking. And why C++ doesn't have libraries(standard) for such basic things like networking?

    Read the article

  • Quickest way to compute the number of shared elements between two vectors

    - by shn
    Suppose I have two vectors of the same size vector< pair<float, NodeDataID> > v1, v2; I want to compute how many elements from both v1 and v2 have the same NodeDataID. For example if v1 = {<3.7, 22>, <2.22, 64>, <1.9, 29>, <0.8, 7>}, and v2 = {<1.66, 7>, <0.03, 9>, <5.65, 64>, <4.9, 11>}, then I want to return 2 because there are two elements from v1 and v2 that share the same NodeDataIDs: 7 and 64. What is the quickest way to do that in C++ ? Just for information, note that the type NodeDataIDs is defined as I use boost as: typedef adjacency_list<setS, setS, undirectedS, NodeData, EdgeData> myGraph; typedef myGraph::vertex_descriptor NodeDataID; But it is not important since we can compare two NodeDataID using the operator == (that is, possible to do v1[i].second == v2[j].second)

    Read the article

  • Core i5 Turboboost/C states Freezing computer

    - by Aaron Smith
    I'm not sure which, but I just had a heck of a time getting my computer to boot up and not freeze. It would run until it finished booting windows, then everything would freeze. This happened until I turned off turboboost and all the c states on the processor. What could be causing this? Is the processor going bad?

    Read the article

  • Oracle Database 12c Spatial: Vector Performance Acceleration

    - by Okcan Yasin Saygili-Oracle
    Most business information has a location component, such as customer addresses, sales territories and physical assets. Businesses can take advantage of their geographic information by incorporating location analysis and intelligence into their information systems. This allows organizations to make better decisions, respond to customers more effectively, and reduce operational costs – increasing ROI and creating competitive advantage. Oracle Database, the industry’s most advanced database,  includes native location capabilities, fully integrated in the kernel, for fast, scalable, reliable and secure spatial and massive graph applications. It is a foundation for deploying enterprise-wide spatial information systems and locationenabled business applications. Developers can extend existing Oracle-based tools and applications, since they can easily incorporate location information directly in their applications, workflows, and services. Spatial Features The geospatial data features of Oracle Spatial and Graph option support complex geographic information systems (GIS) applications, enterprise applications and location services applications. Oracle Spatial and Graph option extends the spatial query and analysis features included in every edition of Oracle Database with the Oracle Locator feature, and provides a robust foundation for applications that require advanced spatial analysis and processing in the Oracle Database. It supports all major spatial data types and models, addressing challenging business-critical requirements from various industries, including transportation, utilities, energy, public sector, defense and commercial location intelligence. Network Data Model Graph Features The Network Data Model graph explicitly stores and maintains a persistent data model withnetwork connectivity and provides network analysis capability such as shortest path, nearest neighbors, within cost and reachability. It loads partitioned networks into memory on demand, overcomingthe limitations of in-memory analysis. Partitioning massive networks into manageable sub-networkssimplifies the network analysis. RDF Semantic Graph Features RDF Semantic Graph has native support for World Wide Web Consortium standards. It has open, scalable, and secure features for storing RDF/OWL ontologies anddata; native inference with OWL 2, SKOS and user-defined rules; and querying RDF/OWL data withSPARQL 1.1, Java APIs, and SPARQLgraph patterns in SQL. Video: Oracle Spatial and Graph Overview Oracle spatial is embeded on oracle database product. So ,we can use oracle installer (OUI).The Oracle Universal Installer (OUI) is used to install Oracle Database software. OUI is a graphical user interface utility that enables you to view the Oracle software that is installed on your machine, install new Oracle Database software, and delete Oracle software that you no longer need to use. Online Help is available to guide you through the installation process. One of the installation options is to create a database. If you select database creation, OUI automatically starts Oracle Database Configuration Assistant (DBCA) to guide you through the process of creating and configuring a database. If you do not create a database during installation, you must invoke DBCA after you have installed the software to create a database. You can also use DBCA to create additional databases. For installing Oracle Database 12c you may check the Installing Oracle Database Software and Creating a Database tutorial under the Oracle Database 12c 2-Day DBA Series.You can always check if spatial is available in your database using  "select comp_id, version, status, comp_name from dba_registry where comp_id='SDO';"   One of the most notable improvements with Oracle Spatial and Graph 12c can be seen in performance increases in vector data operations. Enabling the Spatial Vector Acceleration feature (available with the Spatial option) dramatically improves the performance of commonly used vector data operations, such as sdo_distance, sdo_aggr_union, and sdo_inside. With 12c, these operations also run more efficiently in parallel than in prior versions through the use of metadata caching. For organizations that have been facing processing limitations, these enhancements enable developers to make a small set of configuration changes and quickly realize significant performance improvements. Results include improved index performance, enhanced geometry engine performance, optimized secondary filter optimizations for Spatial operators, and improved CPU and memory utilization for many advanced vector functions. Vector performance acceleration is especially beneficial when using Oracle Exadata Database Machine and other large-scale systems. Oracle Spatial and Graph vector performance acceleration builds on general improvements available to all SDO_GEOMETRY operations in these areas: Caching of index metadata, Concurrent update mechanisms, and Optimized spatial predicate selectivity and cost functions. These optimizations enable more efficient use of: CPU, Memory, and Partitioning Resulting in substantial query performance improvements.UsageTo accelerate the performance of spatial operators, it is recommended that you set the SPATIAL_VECTOR_ACCELERATION database system parameter to the value TRUE. (This parameter is authorized for use only by licensed Oracle Spatial users, and its default value is FALSE.) You can set this parameter for the whole system or for a single session. To set the value for the whole system, do either of the following:Enter the following statement from a suitably privileged account:   ALTER SYSTEM SET SPATIAL_VECTOR_ACCELERATION = TRUE;Add the following to the database initialization file (xxxinit.ora):   SPATIAL_VECTOR_ACCELERATION = TRUE;To set the value for the current session, enter the following statement from a suitably privileged account:   ALTER SESSION SET SPATIAL_VECTOR_ACCELERATION = TRUE; Checkout the complete list of new features on Oracle.com @ http://www.oracle.com/technetwork/database/options/spatialandgraph/overview/index.html Spatial and Graph Data Sheet (PDF) Spatial and Graph White Paper (PDF)

    Read the article

  • can I force server to always use turboboost?

    - by javapowered
    I'm using HP DL360p Gen8 with 2 * Xeon E5-2640. I do not load CPU 100%, i load it only ~10% and so I guess turboboost is not activated. However I'm using my server for trading so I absolutely don't care about CPU loading but I always want to process my data asap. So I want server to operate using maximum 3 GHz. I.e. 90% of CPU time I don't have anything to process. 10% of CPU time I have data to process. But I need to process it ASAP. I need every single microsecond. So I want server to operate always at maximum "turboboosted" mode. Is it possible?

    Read the article

  • What is a good practice for 2D scene graph partitioning for culling?

    - by DevilWithin
    I need to know an efficient way to cull the scene graph objects, to render exclusively the ones in the view, and as fast as possible. I am thinking of doing it the following way, having in each object a local boundingbox which holds the object bounds, and a global boundingbox which holds the bounds of the object and all children. When a camera is moved, the render list is updated by traversing the global boundingboxes. When only the object is being moved, it tries to enlarge or shrink the ancestors global boundingboxes, and in the end updating or not the renderlist. What do you think of this approach? Do you think it will provide a fast and efficient culling? Also, because the render list is a contiguous list, it could accelerate the rendering, right? Any further tips for a 2D scene graphs are highly appreciated!

    Read the article

  • Prim's MST algorithm implementation with Java

    - by user1290164
    I'm trying to write a program that'll find the MST of a given undirected weighted graph with Kruskal's and Prim's algorithms. I've successfully implemented Kruskal's algorithm in the program, but I'm having trouble with Prim's. To be more precise, I can't figure out how to actually build the Prim function so that it'll iterate through all the vertices in the graph. I'm getting some IndexOutOfBoundsException errors during program execution. I'm not sure how much information is needed for others to get the idea of what I have done so far, but hopefully there won't be too much useless information. This is what I have so far: I have a Graph, Edge and a Vertex class. Vertex class mostly just an information storage that contains the name (number) of the vertex. Edge class can create a new Edge that has gets parameters (Vertex start, Vertex end, int edgeWeight). The class has methods to return the usual info like start vertex, end vertex and the weight. Graph class reads data from a text file and adds new Edges to an ArrayList. The text file also tells us how many vertecis the graph has, and that gets stored too. In the Graph class, I have a Prim() -method that's supposed to calculate the MST: public ArrayList<Edge> Prim(Graph G) { ArrayList<Edge> edges = G.graph; // Copies the ArrayList with all edges in it. ArrayList<Edge> MST = new ArrayList<Edge>(); Random rnd = new Random(); Vertex startingVertex = edges.get(rnd.nextInt(G.returnVertexCount())).returnStartingVertex(); // This is just to randomize the starting vertex. // This is supposed to be the main loop to find the MST, but this is probably horribly wrong.. while (MST.size() < returnVertexCount()) { Edge e = findClosestNeighbour(startingVertex); MST.add(e); visited.add(e.returnStartingVertex()); visited.add(e.returnEndingVertex()); edges.remove(e); } return MST; } The method findClosesNeighbour() looks like this: public Edge findClosestNeighbour(Vertex v) { ArrayList<Edge> neighbours = new ArrayList<Edge>(); ArrayList<Edge> edges = graph; for (int i = 0; i < edges.size() -1; ++i) { if (edges.get(i).endPoint() == s.returnVertexID() && !visited(edges.get(i).returnEndingVertex())) { neighbours.add(edges.get(i)); } } return neighbours.get(0); // This is the minimum weight edge in the list. } ArrayList<Vertex> visited and ArrayList<Edges> graph get constructed when creating a new graph. Visited() -method is simply a boolean check to see if ArrayList visited contains the Vertex we're thinking about moving to. I tested the findClosestNeighbour() independantly and it seemed to be working but if someone finds something wrong with it then that feedback is welcome also. Mainly though as I mentioned my problem is with actually building the main loop in the Prim() -method, and if there's any additional info needed I'm happy to provide it. Thank you. Edit: To clarify what my train of thought with the Prim() method is. What I want to do is first randomize the starting point in the graph. After that, I will find the closest neighbor to that starting point. Then we'll add the edge connecting those two points to the MST, and also add the vertices to the visited list for checking later, so that we won't form any loops in the graph. Here's the error that gets thrown: Exception in thread "main" java.lang.IndexOutOfBoundsException: Index: 0, Size: 0 at java.util.ArrayList.rangeCheck(Unknown Source) at java.util.ArrayList.get(Unknown Source) at Graph.findClosestNeighbour(graph.java:203) at Graph.Prim(graph.java:179) at MST.main(MST.java:49) Line 203: return neighbour.get(0); in findClosestNeighbour() Line 179: Edge e = findClosestNeighbour(startingVertex); in Prim()

    Read the article

  • cannot receive UDP broadcast packets

    - by user292792
    Hello I have 2 boxes: - an embedded device (ARM Omap with linux) which I'll call "Omap". - a PC (can either be Windows or linux). Scenario 1 Both boxes are in the same network (example: my office). The Omap gets its address from a DHCP server (ex: 192.168.10.110). The PC has always the same address (ex. 192.168.10.104). I can successfully exchange UDP broadcast packets on any port. Success. Scenario 2 The 2 boxes are in a network withOUT a DHCP server. The PC has a static IP address (example: 10.10.10.20). The Omap boots, looks for a DHCP server, doesn't find it, and is in what I call "bad IP address" state. Now... Broadcasting UDP packets from the Omap works: the PC can see them. The opposite doesn't work: UDP packets broadcasted by the PC are not seen by the Omap. I am using Wireshark on another PC to verify that the packets are being sent. Failure. I tried to change the Omap ip address (with ifconfig)... no luck. What am I missing? To complete the picture, when the Omap is in Scenario 2, if I run udhcpc ... it can communicate with the DHCP server and get an IP address. I also see the packets with Wireshark. So this means that the DHCP client is able to broadcast UDP packets. (Yes, I tried to use DHCP ports 67/68 but it doesn't work). I am using Boost C++ Asio UDP sockets. Specifically, I took the multicast examples and changed them to do broadcasting. Any help is appreciated. Thanks, Benedetto

    Read the article

  • Reinterpret a CGImageRef using PyObjC in Python

    - by Michael Rondinelli
    Hi, I'm doing something that's a little complicated to sum up in the title, so please bear with me. I'm writing a Python module that provides an interface to my C++ library, which provides some specialized image manipulation functionality. It would be most convenient to be able to access image buffers as CGImageRefs from Python, so they could be manipulated further using Quartz (using PyObjC, which works well). So I have a C++ function that provides a CGImageRef representation from my own image buffers, like this: CGImageRef CreateCGImageRefForImageBuffer(shared_ptr<ImageBuffer> buffer); I'm using Boost::Python to create my Python bridge. What is the easiest way for me to export this function so that I can use the CGImageRef from Python? Problems: The CGImageRef type can't be exported directly because it is a pointer to an undefined struct. So I could make a wrapper function that wraps it in a PyCObject or something to get it to send the pointer to Python. But then how do I "cast" this object to a CGImageRef from Python? Is there a better way to go about this?

    Read the article

  • Instantiating class with custom allocator in shared memory

    - by recipriversexclusion
    I'm pulling my hair due to the following problem: I am following the example given in boost.interprocess documentation to instantiate a fixed-size ring buffer buffer class that I wrote in shared memory. The skeleton constructor for my class is: template<typename ItemType, class Allocator > SharedMemoryBuffer<ItemType, Allocator>::SharedMemoryBuffer( unsigned long capacity ){ m_capacity = capacity; // Create the buffer nodes. m_start_ptr = this->allocator->allocate(); // allocate first buffer node BufferNode* ptr = m_start_ptr; for( int i = 0 ; i < this->capacity()-1; i++ ) { BufferNode* p = this->allocator->allocate(); // allocate a buffer node } } My first question: Does this sort of allocation guarantee that the buffer nodes are allocated in contiguous memory locations, i.e. when I try to access the n'th node from address m_start_ptr + n*sizeof(BufferNode) in my Read() method would it work? If not, what's a better way to keep the nodes, creating a linked list? My test harness is the following: // Define an STL compatible allocator of ints that allocates from the managed_shared_memory. // This allocator will allow placing containers in the segment typedef allocator<int, managed_shared_memory::segment_manager> ShmemAllocator; //Alias a vector that uses the previous STL-like allocator so that allocates //its values from the segment typedef SharedMemoryBuffer<int, ShmemAllocator> MyBuf; int main(int argc, char *argv[]) { shared_memory_object::remove("MySharedMemory"); //Create a new segment with given name and size managed_shared_memory segment(create_only, "MySharedMemory", 65536); //Initialize shared memory STL-compatible allocator const ShmemAllocator alloc_inst (segment.get_segment_manager()); //Construct a buffer named "MyBuffer" in shared memory with argument alloc_inst MyBuf *pBuf = segment.construct<MyBuf>("MyBuffer")(100, alloc_inst); } This gives me all kinds of compilation errors related to templates for the last statement. What am I doing wrong?

    Read the article

  • A map and set which uses contiguous memory and has a reserve function

    - by edA-qa mort-ora-y
    I use several maps and sets. The lack of contiguous memory, and high number of (de)allocations, is a performance bottleneck. I need a mainly STL-compatbile map and set class which can use a contiguous block of memory for internal objects (or multiple blocks). It also needs to have a reserve function so that I can preallocate for expected sizes. Before I write my own I'd like to check what is available first. Is there something in Boost which does this? Does somebody know of an available implementation elsewhere? Intrusive collection types are not usable here as the same objects need to exist in several collections. As far as I know STL memory pools are per-type, not per instance. These global pools are not efficient with respect to memory locality in mutli-cpu/core processing. Object pools don't work as the types will be shared between instance but their pool should not. In many cases a hash map may be an option in some cases.

    Read the article

  • Parallel version of loop not faster than serial version

    - by Il-Bhima
    I'm writing a program in C++ to perform a simulation of particular system. For each timestep, the biggest part of the execution is taking up by a single loop. Fortunately this is embarassingly parallel, so I decided to use Boost Threads to parallelize it (I'm running on a 2 core machine). I would expect at speedup close to 2 times the serial version, since there is no locking. However I am finding that there is no speedup at all. I implemented the parallel version of the loop as follows: Wake up the two threads (they are blocked on a barrier). Each thread then performs the following: Atomically fetch and increment a global counter. Retrieve the particle with that index. Perform the computation on that particle, storing the result in a separate array Wait on a job finished barrier The main thread waits on the job finished barrier. I used this approach since it should provide good load balancing (since each computation may take differing amounts of time). I am really curious as to what could possibly cause this slowdown. I always read that atomic variables are fast, but now I'm starting to wonder whether they have their performance costs. If anybody has some ideas what to look for or any hints I would really appreciate it. I've been bashing my head on it for a week, and profiling has not revealed much.

    Read the article

  • What makes deployment successful for some users and unsuccessful for others?

    - by Julien
    I am trying to deploy a Visual C++ application (developed with Microsoft Visual Studio 2008) using a Setup and Deployment Project. After installation, users on some target computers get the following error message after launching the application executable: “This application has failed to start because the application configuration is incorrect. Reinstalling the application may fix the problem.” Another user after installation could run the application properly. I cannot find the root cause of this problem, despite spending several hours on the Visual Studio help files and online forums (most postings date back to 2006). Does anyone at Stack Overflow have a suggestion? Thanks in advance. Additional details appear below. The application uses FLTK 1.1.9 for a GUI library, as well as some Boost 1.39 libraries (regex, lexical_cast, date_time, math). I made sure I am trying to deploy the release version (not the debug version) of the application. The Runtime library in the Code Generation settings is Multi-threaded DLL (/MD). The dependency walker of myapp.exe lists the following DLLs: wsock32.dll, comctl32.dll, kernel32.dll, user32.dll, gdi32.dll, shell32.dll, ole32.dll, mvcp90.dll, msvcr90.dll. In the Setup and Deployment Project, I add the following DLLs to the File System on Target Machine: fltkdlld.dll, and a folder named Microsoft.VC90.CRT with msvcm90.dll, msvcp90.dll, mcvcr90.dll and Microsoft.VC90.CRT.manifest. The installation process on the target computers getting the error message requires having the .Net Framework 3.5 installed first. Any suggestion? Thanks in advance!

    Read the article

  • What is the nicest way to parse this in C++ ?

    - by ereOn
    Hi, In my program, I have a list of "server address" in the following format: host[:port] The brackets here, indicate that the port is optional. host can be a hostname, an IPv4 or IPv6 address. port, if present can be a numeric port number or a service string (like: "http" or "ssh"). If port is present and host is an IPv6 address, host must be in "bracket-enclosed" notation (Example: [::1]) Here are some valid examples: localhost localhost:11211 127.0.0.1:http [::1]:11211 ::1 [::1] And an invalid example: ::1:80 // Invalid: Is this the IPv6 address ::1:80 and a default port, or the IPv6 address ::1 and the port 80 ? ::1:http // This is not ambigous, but for simplicity sake, let's consider this is forbidden as well. My goal is to separate such entries in two parts (obviously host and port). I don't care if either the host or port are invalid as long as they don't contain a : (290.234.34.34.5 is ok for host, it will be rejected in the next process); I just want to separate the two parts, or if there is no port part, to know it somehow. I tried to do something with std::stringstream but everything I come up to seems hacky and not really elegant. How would you do this in C++ ? I don't mind answers in C but C++ is prefered. Any boost solution is welcome as well. Thank you.

    Read the article

  • Calling a method at a specific interval rate in C++

    - by Aetius
    This is really annoying me as I have done it before, about a year ago and I cannot for the life of me remember what library it was. Basically, the problem is that I want to be able to call a method a certain number of times or for a certain period of time at a specified interval. One example would be I would like to call a method "x" starting from now, 10 times, once every 0.5 seconds. Alternatively, call method "x" starting from now, 10 times, until 5 seconds have passed. Now I thought I used a boost library for this functionality but I can't seem to find it now and feeling a bit annoyed. Unfortunately I can't look at the code again as I'm not in possession of it any more. Alternatively, I could have dreamt this all up and it could have been proprietary code. Assuming there is nothing out there that does what I would like, what is currently the best way of producing this behaviour? It would need to be high-resolution, up to a millisecond. It doesn't matter if it blocks the thread that it is executed from or not. Thanks!

    Read the article

  • Loading specific files from arbitrary directories?

    - by Haydn V. Harach
    I want to load foo.txt. foo.txt might exist in the data/bar/ directory, or it might exist in the data/New Folder/ directory. There might be a different foo.txt in both of these directories, in which case I would want to either load one and ignore the other according to some order that I've sorted the directories by (perhaps manually, perhaps by date of creation), or else load them both and combine the results somehow. The latter (combining the results of both/all foo.txt files) is circumstantial and beyond the scope of this question, but something I want to be able to do in the future. I'm using SDL and boost::filesystem. I want to keep my list of dependencies as small as possible, and as cross-platform as possible. I'm guessing that my best bet would be to get a list of every directory (within the data/ folder), sort/filter this list, then when I go to load foo.txt, I search for it in each potential directory? This sounds like it would be very inefficient, if I have dozens of potential directories to search through every time. What's the best way to go about accomplishing this? Bonus: What if I want some of the directories to be archives? ie. considering both data/foo/ and data/bar.zip to both be valid, and pull foobar.txt from either one without caring.

    Read the article

  • I don't understand how work call_once

    - by SABROG
    Please help me understand how work call_once Here is thread-safe code. I don't understand why this need Thread Local Storage and global_epoch variables. Variable _fast_pthread_once_per_thread_epoch can be changed to constant/enum like {FAST_PTHREAD_ONCE_INIT, BEING_INITIALIZED, FINISH_INITIALIZED}. Why needed count calls in global_epoch? I think this code can be rewriting with logc: if flag FINISH_INITIALIZED do nothing, else go to block with mutexes and this all. #ifndef FAST_PTHREAD_ONCE_H #define FAST_PTHREAD_ONCE_H #include #include typedef sig_atomic_t fast_pthread_once_t; #define FAST_PTHREAD_ONCE_INIT SIG_ATOMIC_MAX extern __thread fast_pthread_once_t _fast_pthread_once_per_thread_epoch; #ifdef __cplusplus extern "C" { #endif extern void fast_pthread_once( pthread_once_t *once, void (*func)(void) ); inline static void fast_pthread_once_inline( fast_pthread_once_t *once, void (*func)(void) ) { fast_pthread_once_t x = *once; /* unprotected access */ if ( x _fast_pthread_once_per_thread_epoch ) { fast_pthread_once( once, func ); } } #ifdef __cplusplus } #endif #endif FAST_PTHREAD_ONCE_H Source fast_pthread_once.c The source is written in C. The lines of the primary function are numbered for reference in the subsequent correctness argument. #include "fast_pthread_once.h" #include static pthread_mutex_t mu = PTHREAD_MUTEX_INITIALIZER; /* protects global_epoch and all fast_pthread_once_t writes */ static pthread_cond_t cv = PTHREAD_COND_INITIALIZER; /* signalled whenever a fast_pthread_once_t is finalized */ #define BEING_INITIALIZED (FAST_PTHREAD_ONCE_INIT - 1) static fast_pthread_once_t global_epoch = 0; /* under mu */ __thread fast_pthread_once_t _fast_pthread_once_per_thread_epoch; static void check( int x ) { if ( x == 0 ) abort(); } void fast_pthread_once( fast_pthread_once_t *once, void (*func)(void) ) { /*01*/ fast_pthread_once_t x = *once; /* unprotected access */ /*02*/ if ( x _fast_pthread_once_per_thread_epoch ) { /*03*/ check( pthread_mutex_lock(µ) == 0 ); /*04*/ if ( *once == FAST_PTHREAD_ONCE_INIT ) { /*05*/ *once = BEING_INITIALIZED; /*06*/ check( pthread_mutex_unlock(µ) == 0 ); /*07*/ (*func)(); /*08*/ check( pthread_mutex_lock(µ) == 0 ); /*09*/ global_epoch++; /*10*/ *once = global_epoch; /*11*/ check( pthread_cond_broadcast(&cv;) == 0 ); /*12*/ } else { /*13*/ while ( *once == BEING_INITIALIZED ) { /*14*/ check( pthread_cond_wait(&cv;, µ) == 0 ); /*15*/ } /*16*/ } /*17*/ _fast_pthread_once_per_thread_epoch = global_epoch; /*18*/ check (pthread_mutex_unlock(µ) == 0); } } This code from BOOST: #ifndef BOOST_THREAD_PTHREAD_ONCE_HPP #define BOOST_THREAD_PTHREAD_ONCE_HPP // once.hpp // // (C) Copyright 2007-8 Anthony Williams // // Distributed under the Boost Software License, Version 1.0. (See // accompanying file LICENSE_1_0.txt or copy at // http://www.boost.org/LICENSE_1_0.txt) #include #include #include #include "pthread_mutex_scoped_lock.hpp" #include #include #include namespace boost { struct once_flag { boost::uintmax_t epoch; }; namespace detail { BOOST_THREAD_DECL boost::uintmax_t& get_once_per_thread_epoch(); BOOST_THREAD_DECL extern boost::uintmax_t once_global_epoch; BOOST_THREAD_DECL extern pthread_mutex_t once_epoch_mutex; BOOST_THREAD_DECL extern pthread_cond_t once_epoch_cv; } #define BOOST_ONCE_INITIAL_FLAG_VALUE 0 #define BOOST_ONCE_INIT {BOOST_ONCE_INITIAL_FLAG_VALUE} // Based on Mike Burrows fast_pthread_once algorithm as described in // http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2007/n2444.html template void call_once(once_flag& flag,Function f) { static boost::uintmax_t const uninitialized_flag=BOOST_ONCE_INITIAL_FLAG_VALUE; static boost::uintmax_t const being_initialized=uninitialized_flag+1; boost::uintmax_t const epoch=flag.epoch; boost::uintmax_t& this_thread_epoch=detail::get_once_per_thread_epoch(); if(epoch #endif I right understand, boost don't use atomic operation, so code from boost not thread-safe?

    Read the article

  • Questioning the motivation for dependency injection: Why is creating an object graph hard?

    - by oberlies
    Dependency injection frameworks like Google Guice give the following motivation for their usage (source): To construct an object, you first build its dependencies. But to build each dependency, you need its dependencies, and so on. So when you build an object, you really need to build an object graph. Building object graphs by hand is labour intensive (...) and makes testing difficult. But I don't buy this argument: Even without dependency injection, I can write classes which are both easy to instantiate and convenient to test. E.g. the example from the Guice motivation page could be rewritten in the following way: class BillingService { private final CreditCardProcessor processor; private final TransactionLog transactionLog; // constructor for tests, taking all collaborators as parameters BillingService(CreditCardProcessor processor, TransactionLog transactionLog) { this.processor = processor; this.transactionLog = transactionLog; } // constructor for production, calling the (productive) constructors of the collaborators public BillingService() { this(new PaypalCreditCardProcessor(), new DatabaseTransactionLog()); } public Receipt chargeOrder(PizzaOrder order, CreditCard creditCard) { ... } } So dependency injection may really be an advantage in advanced use cases, but I don't need it for easy construction and testability, do I?

    Read the article

  • Questioning one of the arguments for dependency injection: Why is creating an object graph hard?

    - by oberlies
    Dependency injection frameworks like Google Guice give the following motivation for their usage (source): To construct an object, you first build its dependencies. But to build each dependency, you need its dependencies, and so on. So when you build an object, you really need to build an object graph. Building object graphs by hand is labour intensive (...) and makes testing difficult. But I don't buy this argument: Even without dependency injection, I can write classes which are both easy to instantiate and convenient to test. E.g. the example from the Guice motivation page could be rewritten in the following way: class BillingService { private final CreditCardProcessor processor; private final TransactionLog transactionLog; // constructor for tests, taking all collaborators as parameters BillingService(CreditCardProcessor processor, TransactionLog transactionLog) { this.processor = processor; this.transactionLog = transactionLog; } // constructor for production, calling the (productive) constructors of the collaborators public BillingService() { this(new PaypalCreditCardProcessor(), new DatabaseTransactionLog()); } public Receipt chargeOrder(PizzaOrder order, CreditCard creditCard) { ... } } So there may be other arguments for dependency injection (which are out of scope for this question!), but easy creation of testable object graphs is not one of them, is it?

    Read the article

  • Structuring cascading properties - parent only or parent + entire child graph?

    - by SB2055
    I have a Folder entity that can be Moderated by users. Folders can contain other folders. So I may have a structure like this: Folder 1 Folder 2 Folder 3 Folder 4 I have to decide how to implement Moderation for this entity. I've come up with two options: Option 1 When the user is given moderation privileges to Folder 1, define a moderator relationship between Folder 1 and User 1. No other relationships are added to the db. To determine if the user can moderate Folder 3, I check and see if User 1 is the moderator of any parent folders. This seems to alleviate some of the complexity of handling updates / moved entities / additions under Folder 1 after the relationship has been defined, and reverting the relationship means I only have to deal with one entity. Option 2 When the user is given moderation privileges to Folder 1, define a new relationship between User 1 and Folder 1, and all child entities down to the grandest of grandchildren when the relationship is created, and if it's ever removed, iterate back down the graph to remove the relationship. If I add something under Folder 2 after this relationship has been made, I just copy all Moderators into the new Entity. But when I need to show only the top-level Folders that a user is Moderating, I need to query all folders that have a parent folder that the user does not moderate, as opposed to option 1, where I just query any items that the user is moderating. I think it comes down to determining if users will be querying for all parent items more than they'll be querying child items... if so, then option 1 seems better. But I'm not sure. Is either approach better than the other? Why? Or is there another approach that's better than both? I'm using Entity Framework in case it matters.

    Read the article

  • How to choose an integer linear programming solver ?

    - by Cassie
    Hi all, I am newbie for integer linear programming. I plan to use a integer linear programming solver to solve my combinatorial optimization problem. I am more familiar with C++/object oriented programming on an IDE. Now I am using NetBeans with Cygwin to write my applications most of time. May I ask if there is an easy use ILP solver for me? Or it depends on the problem I want to solve ? I am trying to do some resources mapping optimization. Please let me know if any further information is required. Thank you very much, Cassie.

    Read the article

< Previous Page | 54 55 56 57 58 59 60 61 62 63 64 65  | Next Page >