128 bit math routines

TheMSsForum.com: The Microsoft Software Forum

The MSS Forum ‹ VC
- Archive
  - Biz
  - MCSE
  - CRM
  - Drivers
  - Framework
  - ADO
  - ASP
  - Compact
  - Forms
  - Dotnet
  - C#
  - VB
  - FontpageGen
  - Excel
  - WorkSheet
  - Exchange
  - Setup
  - Fox
  - Fontpage
  - ASP
  - IIS
  - Entourage
  - Money
  - Messanger
  - PocketPC
  - Powerpoint
  - Project
  - Publisher
  - Excel
  - VB
  - Security
  - Portal
  - Services
  - SQLServerDev
  - SVCS
  - SQLServer
  - VB
  - VC
  - MFC
  - ExcelGen

- Previous
  - 1
    - Declaration and Definition I declare a function that is virutal without implementation code in my base class, the derived class does not implement it so link error occurs. My question is how do you make the base class code swung into action? class A { virtual int Run(void); }; class B : public A { want to use class A's code, what can be done? Thanks Jack Tag: 128 bit math routines Tag: 245734
  - 2
    - Stealth activation of exe upon right-button -down event? I appologise for bringing this here, however I don't know of more savvy group of folks to ask this question to. I was in the process of deleteing an exe, I decided to not install it. It took about 3-5 seconds for my right-click menu to activate. I sensed that there was alot going on behind the scenes on this event. Am I being paranoid or is it possible that this is by design to secretly get into a computer? I'm using Win2000 with symantic corprate edition av. any help appreciated. Tag: 128 bit math routines Tag: 245732
  - 3
    - Using STL with classes generated dynamically Hi -- I am currently working on a Forth-inspired virtual machine that I use as a run-time for executing domain-specific languages that are generated on the fly from visual models. This machine is created out of thin air at runtime by writing Intel assembler codes into memory for the 32 primitive instructions, the execution engine and for a rudimentary parser that can interpret and compile Forth instruction streams. Once this initialization is done, then the Forth machine accepts streams of forth words and bootstraps itself into a more full-fledged execution environment that includes a type system, classes, an intel code assembler, floating point math, etc. This is all kicked off from within some host program that must be written in a language that supports writing bytes to memory and then jumping to one of these locations to execute the code thus assembled. (I've tried to keep the host requirements limited to these two so that this scheme can also eventually work in an embedded environment with only ROM-based byte sequences to kick it off.) Please bear with me and I will get to my queson in one second; I just have to talk about one more thing to frame it correctly. The runtime environment that emerges from this scheme involves a few different levels: there's the basic intel assembler level that I have access to through the virtual machine; there's the virtual machine itself, which executes any named sequence of the underlying primitives that has been compiled into the system's code dictionary or that is presented to it for run-time interpretation (or any sequence of other such named sequences) ; and then there's the runtime of the host program. While I'm trying to keep the virtual machine simple and capable of executing in just about any environment that satisfies the two requirements I mentioned earlier and that has had an appropriate execution engine generated for it, I also want it to be able, when it's useful, to interact with both the underlying processor and with the host executable environment. I accomplish this through two of the primitives -- one that executes underlying assembly code and the other that makes far calls to the host environment. This presents a great degree of choice about which level to program specific functionality at and allows me to use facilities of a host language to handle stuff that had not yet made its way into the forth-based portable environment or that can be handled better by the host in a context where I am willing to sacrifice portability. More specifically (and more relevant to this newsgroup) when the host programming environment is a C++ executable, I would like to enable as much interaction as possible with the C++ run-time. For example I have a project coming up for a client for a simulation system for which I would like to be able to use the STL and the C++/CLI extensions for the .NET framework library. So here's my question. I want to use the type-generating and dynamic execution capabilities of my embedded Forth environment to create execution streams on the fly from visual models that the user creates and manipulates at run-time. On the other hand, in the course of executing these I would like to leverage STL containers and algorithms to take advantage of rich programming facilities like Boost and Loki (in particular the Boost Graph Library). Now I realize that by the time my application runs, code templates are gone and the only thing that remains of them are type-specific classes that have been expanded at compile time. So my original thought to somehow create new expansions with my dynamically generated types is impossible, since at that point there is nothing to expand from. However, I'm envisioning something like the following, and would just like some basic validation on whether it seems feasible to minds more experienced and wise than my own: If I gnerate my dynamic types in such a way that they are set up in memory exactly like C++ would set up its own objects, then I should be able to pass pointers to them as parameters in far calls to C++ functions and the C++ functions shouldn't know the difference. Right? (I already know this is true with simple structures, because that's how I'm invoking OS calls.) And continuing along this path, if I compile into the C++ host a template expansion of say, an STL vector or a map that has as its value type paramater a pointer to some class, then I should be able to generate my own forth types in such a way that they are set up in memory as though they inherited from this base class type. And I should be able to pass instances pointers of these classes in calls to the corresponding push_back() or find() functions of my STL container classes. As long as the instance pointed to has a correctly constructed vtable, I should be okay. Right? And the same concept should, I'm imagining, hold true for functors that I might want to pass as predicates to STL algorithms. And in fact, if I set up my functor objects so that they appear correctly to derive from the actual type I specified in the C++ template expansion, then the actual predicate functors that I pass could be typed to be aware of the actual type of object that I passed to the container as a value type and could do type-specific manipulations on those objects. I've probably lost y'all by now and of course you might ask "Why don't you just try it and see?". Unfortuinately, I'm currently doing major surgery on my underlying execution engine and am not in a position to test this out, since I didn't have the foresight to preserve a working version that would allow me to test. (Idiot!) Yet I need to decide if this approach is even feasible as I will be specing the system out for the client over the next few weeks and would love to have the comfort of knowing that I have available the full expressive power of the STL in addition to my dyamic execution engine. I realize that if my value type for the STL containers is a pointer to something, then I will have to use proper iterators that are able to call comparison functions and the like dynamically (Boost has this already, I believe). I also realize that this totally violates the spirit of compile-time template expansion type safety. But given my circumstances, I don't think I should have to give up the STL just because my needs are a little exotic. Whoever's still with me, thanks very much for slogging through this. I'd appreciate very much some informed view of the feasibility of what I've described here. And if you feel really charitable, any insights into how to generate well-formed objects with correctly formulated vtables would also be lovely. Thanks and best wishes to all, Michael Clagett Tag: 128 bit math routines Tag: 245730
  - 4
    - Programming for Outlook Express I know there are programs that interact with Outlook Express (mostly anti-spam ones) but am curious as to how I can make a program that will interact with OE. Looking so hard on Google, and the best I came up with was for C#, but I am just C...no C++, no C#, just C. Is there any hope for me? Abd if there is, could you provide the link too? Thank Tag: 128 bit math routines Tag: 245729
  - 5
    - Internal Compiler error in MSVC++ Using MSVC++ 7.1.3088 I have been getting an Internal Error (contact MS tech support) from my VC++ compiler lately. I finally figured out what causes it. Suppose I have a class X, then the compiler has a problem with the following lines of code: typedef X Y; class Y; The forward declare of "class Y" should be treated as a redefinition of Y, but it confuses the compiler and causes the Internal Error message. Is this a known bug? Can it be fixed in the next version? From, Stephen R Mercer -= LabVIEW R&D =- National Instruments Tag: 128 bit math routines Tag: 245726
  - 6
    - How to call placement delete operator ? HI, Is there any way I can call placement delete operator ? Maybe somehow thanks... a.a.cpp Tag: 128 bit math routines Tag: 245713
  - 7
    - [error lnk 2001] hello i got the following linking error : error LNK2001: unresolved external symbol "public: __thiscall kplAnonKernFromCndKern::kplAnonKernFromCndKern(class std::list<class kplVariable *,class std::allocator<class kplVariable *> > const &,class s td::list<class kplVariable *,class std::allocator<class kplVariable *> > const &,class kplExternalProbFunction *,class plDataValues const &)" (??0kplAnonKernFromCndKern@@QAE@ABV?$list@PAVkplVariable@@V?$allocator@PAVkplVariable@@@std@@@std@@0PAVkplE xternalProbFunction@@ABVplDataValues@@@Z) Debug/test2.exe : fatal error LNK1120: 1 unresolved externals and i'm wondering if this is due to the lack of a specific library abd how to determine which library is missing? Thanks François Tag: 128 bit math routines Tag: 245708
  - 8
    - Writing a library in C I am looking for some example how to write a dll in pure C, not C++. As a compiler and IDE I've got Visual C++ 5 installed. Any tips? Tag: 128 bit math routines Tag: 245706
  - 9
    - waveOutOpen I made a post earlier about trying to write my own sound control (within my app). through a few of the replies I did find "waveOutOpen" and it's other functions. but i can't see the correct way to use "waveOutOpen". my pc only has one audio device, and i just need my app to be able to open the standard volume on any flavor of windows, and change it. i only knnow C, not C++ so please keep that in mind. can anyone else give me some tips or point me in the right direction? thanks Tag: 128 bit math routines Tag: 245697
  - 10
    - Someway to let new & new[] do not allocate storage !!! Hi, Is there some way to not to allow new operator to allocate storage..somehow maybe if it can be done.... Thanks, a.a.cpp Tag: 128 bit math routines Tag: 245695
  - 11
    - number of days between dates Is there a good simple method of calculating the difference between dates? I have strings of yyyymmdd, and want to do something like 20050304 - 20050223, and get the number of days between those two dates. I thought about creating a whole date class, but is there an easier method? Like maybe creating just one function. string sdate1 = "20050203"; string sdate2 = "20050201"; string date = date1 .substr(6.2); int ndate1 = atoi(sdate1.c_str()); date = date2 .substr(6.2); int ndate2 = atoi(sdate2.c_str()); //calculate how many days are missing days = ndate1 - ndate2 ; but this just does the dd part, maybe someone can post some code, ideas to help. thanks. Tag: 128 bit math routines Tag: 245681
  - 12
    - IViewObject::Draw in Internet Explorer hi Im loading HTML content in a hidden Internet Explorer control and redirect the painting messages in a Memory Device context for other uses. So far here's what i did : I selected a DIBSECTION in a HDC and with a timer , i use the IViewObject::Draw of the document object to draw in my HDC. But the problem is for exemple, when other objects have compound document like IFRAME or FLASH objects. Is there a way to get a hook the main document 's drawing proceedure ? In other words, can i subclass a message somewhere and dispense of the (Timer + IViewObject::Draw) combo ? BEcause obviously the (Timer + IViewObject::Draw) forces the docuemtn to draw TWICE !!! thanks Tag: 128 bit math routines Tag: 245664
  - 13
    - How to Disable Debug? I recently installed C++ 6.0 (old, I know)... and I said "Yes" when it asked about adding debug information to Windows or SDK (or something like that). The unfortunate side-affect has been that when the server program I run crashes, a dialog box now pops up demanding user interaction to allow the process to die or "cancel and debug." This would be fine if Visual Studio was running, but it's not--I only want this sort of crash interception to occur when I'm in Visual Studio. As it stands now, the watchdog process to restart my server when it crashes is effectively nerfed because it needs to run unattended, but the debug dialog box demands a user be present. :( I un-installed all components and this crash intercept debug feature still persists. Unfortunately I didn't save a backup of my Windows folder before the install (which would have made undoing this much easier). I found some information on Microsoft's site about retail and debug versions of the SDK, but it was for an older version of DirectX. Given that background, how do I disable this crash debug feature? Thanks, ~Bonesly Tag: 128 bit math routines Tag: 245647
  - 14
    - DDEConnect question I'm using the DDEConnect function and when I call it with no DDE server in action the code run inside itself and don't return. I think that the normal think is that if the function fall it have to return a code in a bit. I call the funcion like this: hconvCppServer = DdeConnect(m_InstId, hszCppServer, hszCppTopic, 0); All the function parameters are valid. In anyway if I use the function with a valid DDE server it work fine. Can you explain me how is possible this? Tag: 128 bit math routines Tag: 245644
  - 15
    - May STL destructors throw? The C++ standard says (17.4.4.8/3): "No destructor operation defined in the C++ Standard Library will throw an exception." It also does not put an excception specification for basic_string::~basic_string() (21.3/6, pp. 385). Is this NOT a contradiction? May the basic_string dtor throw or not? Tag: 128 bit math routines Tag: 245639
  - 16
    - PrintDlg() method : hDevMode and hDevName help Hello all, My code seems to be working, however I have a question about the returned PRINTDLG::hDevMode and PRINTDLG ::hDevName fields. //****Code*** PRINTDLG g_PrintDlg; //hDevMode and hDevNames = NULL DEVMODE *pdv; BOOL bRetVal, bColor; TCHAR m_szPrinterName[32]; if(bRetVal = PrintDlg(&g_PrintDlg)) { ... if(g_PrintDlg.hDevMode != NULL) { pdv = (DEVMODE *) GlobalLock(g_PrintDlg.hDevMode); lstrcpy(szPrinterName, (LPSTR) pdv->dmDeviceName); bColor = (DMCOLOR_COLOR == pdv->dmColor); GlobalUnlock(g_PrintDlg.hDevMode); } ... // When I free the global memory (below), a second call to // PrintDlg() will cause an insufficient memory error. When I // do not free the memory, everything works fine with repeated // calls to PrintDlg(). if(g_PrintDlg.hDevMode != NULL) GlobalFree(g_PrintDlg.hDevMode); if(g_PrintDlg.hDevNames != NULL) GlobalFree(g_PrintDlg.hDevMode); ... } Question : If I do not free the global memory, will this code create a leak or is there a mechanism which frees the memory that I couldn't find in MSDN? Thank You - M Tag: 128 bit math routines Tag: 245637
  - 17
    - Which has higher precedence & or == I have this example in a window procedure: fBlocking is BOOL and wParam is an argument of the window procedure. switch (iMsg){ // case WM_CHAR: if (fBlocking & wParam =3D=3D '\x1B'){ // } retrurn 0; // } Like this we get a warning: warning C4554: '&' : check operator precedence for possible error; use = parentheses to clarify precedence But I don't know what the precedence was expected here to put the parens = in correctly. Any ideas? --=20 George Hester _________________________________ Tag: 128 bit math routines Tag: 245625
  - 18
    - Simpel WIN32 Message box Hi, i got a problem. I've got VS2005 Beta 1 installed, and the newest platform SDK. I'm a newbie in the WIN32 Programming. but if i try to compile this piece of code: #define WIN32_LEAN_AND_MEAN #include <windows.h> // the main windows headers #include <windowsx.h> // a lot of cool macros int WINAPI WinMain(HINSTANCE hinstance, HINSTANCE hprevinstance, LPSTR lpcmdline, int ncmdshow) { MessageBox(NULL, "THERE CAN BE ONLY ONE!!!", "MY FIRST WINDOWS PROGRAM", MB_OK | MB_ICONEXCLAMATION); return(0); } The compiller gives this error back: Linking... main.obj : error LNK2019: unresolved external symbol __imp__MessageBoxA@16 referenced in function _WinMain@16 Debug/Game_02.exe : fatal error LNK1120: 1 unresolved externals Build log was saved at "file://c:\Documents and Settings\Administrator\Mijn documenten\Visual Studio\Projects\Game_02\Game_02\Debug\BuildLog.htm" Game_02 - 2 error(s), 0 warning(s) ========== Build: 0 succeeded, 1 failed, 0 up-to-date, 0 skipped ========== Anyone knows how to solve this problem? I really like the compilers GUI, and i want to learn on this compiler :) Tag: 128 bit math routines Tag: 245610
  - 19
    - SetWindowWord() I have this in a Window procedure in a C program: SetWindowWord(hwnd, 0, 1^GetWindowWord(hwnd, 0)); hwnd is a handle to a CreateWindow done earlier. Now this gives a = Warning C4761: integral size mismatch in argument; conversion supplied. = So this is my fix: SetWindowWord(hwnd, 0, (WORD)(1^GetWindowWord(hwnd, 0))); I'm wondering if this would be the correct way to get rid of C4761 and = why don't I still get the warning since 1 is a long and GetWindowWord is = an unsigned short? Like 1L would be a long 1 is there a (WORD)1? --=20 George Hester _________________________________ Tag: 128 bit math routines Tag: 245606
  - 20
    - where is thye bext place to buy MSDN subscription? Hello, I want to buy a MSDN subscription. Where is the best place to buy it. I found that OSROnline selling it a little chaper than microsoft itself. Is there any other place that sell it cheaper? Best regards Tag: 128 bit math routines Tag: 245597
  - 21
    - Boolean returns I have a routine that I want to return a boolean value from, i.e: DLL_API bool STDCALL MyRoutine(HWND hWnd) { //some code... return true; } This routine is in a dll which I'm calling from a VB exe. Sometimes the routine returns true when it clearly shouldn't have. The routine is quite simple with a few if's and some &'s and |'s. Anyhow, the code is not important because I can easily do a dry-run in my head, but the routine is not consistent in it's return values. I also use the PrintWindow api which returns a boolean which can't be relied on in c++. I know my routine works ok, because if I return an int instead, so return 1 is true and return 0 is false, everything is fine. Is there a problem with transferring boolean results from a c++ dll to a vb exe? Or is there something I've missed here? Tag: 128 bit math routines Tag: 245595
  - 22
    - copying data Hi, I wrote the following code: BYTE *pBuffer = new BYTE[MYBUFFERSIZE]; //[...] filling buffer with data... BYTE *currpos = NULL; UINT offset = 20;//random value, smaller than MYBUFFERSIZE, of course currpos = &pBuffer[offset]; //creates a pointer to pBuffer increased by 20 bytes This statement compiles fine, but when I run it, I find the first 4 bytes currpos points to corrupted (these 32bits contain the pointers address). Obviously, I'm casting a wrong way, but I cannot figure out what exactly I'm doing wrong....can anyone help? Thanks Peter Tag: 128 bit math routines Tag: 245592
  - 23
    - Runtime libraries In VC++ environment are C, C++ and win32 API in separate libraries/files? What are the filenames for each of them? Tag: 128 bit math routines Tag: 245591
  - 24
    - newbie: Addition of vars Hi, just one question: What's the difference between: int a;//replace int by any value or class that supports this int b; a = a + b; versus a += b; Is there any performance difference? Thanks Peter Tag: 128 bit math routines Tag: 245587
  - 25
    - A silly question Hi All, If there are two classes, they both have a pure virtual function which the signatures are exactly same. Now, can I implement & invoke them respectively if another class extends them at the same time? For example: class A { virtual void Func(void) = 0; }; class B { virtual void Func(void) = 0; }; class C : public B, public A { void Func(void); }; Thanks in advance, Best Regards, Elliott Tag: 128 bit math routines Tag: 245580
- Next
  - 1
    - Mouse hooks I have a neat little dll that allows resizing and dragging by clicking anywhere on the window (pressing ALT at the same time). However, I noted something interesting. If I click the title bar, my dll does nothing. Does clicking the titlebar prevent mouse hooks? If so, what would be a way around it, if there is one? Tag: 128 bit math routines Tag: 245579
  - 2
    - what do I need to develop applications for win64? Hello, What do I need to develop applications for win64 on AMD64 or Intel Xeon processors? Can I do this with MSVC 2003? what about MSVC 2005? Best regards Tag: 128 bit math routines Tag: 245576
  - 3
    - Passing a vector to a function Hi; Is it possible for to pass a vector into a function and access the elements of the vector inside that function? I tried but I am getting an error, so please help! // the following is what I am trying to do but seems like there is sothing I am missing. void readFile(char * fileName) { ifstream inFile(fileName); vector<char> doc; while (!inFile.eof()) { char ch[1]; inFile.read(ch,1); do.push_back(ch[0]); } inFile.close(); compareAndProcess(doc); } void compareAndProcess(char doc[]) { vector<char>::iterator iter; for (iter = doc.begin(); iter < doc.end(); iter++) // process each character } When I compile Iam getting the follwoing error: error C2228: left of '.begin' must have class/struct/union type error C2228: left of '.end' must have class/struct/union type So, could someone please explain if it is even possible to do; if it is possible what changes do I need to make so that it will compile. thanks in advance. JP Tag: 128 bit math routines Tag: 245570
  - 4
    - "Command Arguments" redirection failure in IDE I'm using VS 2003 and have 2 separate simple console projects. One has one source file (main.cpp), which I use for developing simple C++ programs and the other has one source file (main.c), which I use for developing simple C programs. I've written a simple C++ program using the cin.get and cout.put and an equivalent C program using getchar and putchar. In both programs I've setup the "Command Arguments" field of the debug properties for command redirection of stdin using "<\infile" (no actual double-quotes, of course) . The strange thing (to me, at least) is that it works correctly for the C++ version, but completely ignores the redirection in the C version and simply sits there waiting for me to enter something from the keyboard. If I run the programs from a command window they both work as expected when redirected. I know I've gotten this to work in the past so I can't help thinking there must be some subtle thing I have setup incorrectly in the C project. Tag: 128 bit math routines Tag: 245569
  - 5
    - Questions about fopen and printers This is more of a theoretical question. is it possible to use fopen() and fprintf() to send basic text to a printer through the printer's USB port? I guess my real question is, can I use a printer's USB input port as a simple serial port the same way fopen("COM1:", "w") does? thanks in advance, ed Tag: 128 bit math routines Tag: 245564
  - 6
    - About loading an image Hi Group, I'm going to load an image file into memory and modify it then write it out. Please consider this couldn't be using Microsoft classes. The code should be running under both windows and linux so I have to go with standard C++. What classes should I consider to do thing in terms of array and memory managment? Any suggestions on this? Thanks in advance, Tag: 128 bit math routines Tag: 245559
  - 7
    - Efficient File Processing I've written a little utility program that reads binary data (from a file) in big endian format and converts it to little endian. Unfortunately, the size of the file(s) that I typically deal with are quite large (~400MB). I've chosen to read the data using the read() method of the fstream class (I actually read the entire contents of the file into a Byte array...not byte-by-byte or "pieces" at a time). It takes my application a little longer to run than I'd like...is there a better way to access this data than the implementation I've chosen? Here's an example of how I read the data: ifstream faic; filebuf* fbuf; faic.open(argv[1], ios::binary); fbuf = faic.rdbuf(); //get file size filesize=fbuf->pubseekoff (0,ios::end,ios::in); fbuf->pubseekpos (0,ios::in); //allocate byte array b = new byte[filesize]; //read data into b faic.read(b, filesize); faic.close(); Tag: 128 bit math routines Tag: 245534
  - 8
    - Working with Visual C++ Toolkit 2003 Hello! I need some advice concerning the Visual Toolkit 2003. I have been working with Visual C++ 6.0 Professional for several years now, and I am still using it for a running project. Unfortunately, VC++ 6.0 is rather out of date and I am having some trouble with its lack of conformance to the C++ standard. Visual C++ .NET turned out to be rather expensive, especially for the reason that the Professional Edition is not available as a stand-alone any more, only together with the rest of Visual Studio. I guess that this is their way of selling that Visual Basic and Visual J crap that nobody wants to have - well, thanks Microsoft... This more or less forced me to take a look at the Visual C++ Toolkit 2003. I have downloaded and installed it, and it basically works fine. However, I cannot imagine that it is at all possible to build a large-scale project from the command line. The ReadMe file include some link to internet resources on using the Toolkit, but I have checked them out and they are nothing but useless. Other so-called "tutorials" only provide an installation guide and tell you how to compile a "hello world" application. Is anyone out there seriously using the Toolkit for their projects? If so, how do you use it? Are you working with Notepad or with a custom IDE? Are you using makefiles, or are you typing a dozen of commands each time you compile? How do you set your path variables? When I type "cl" into a DOS prompt, all I get is an error because the command cannot be found... Sorry, but I don't even know where to start - or maybe I shouldn't do so anyway, but rather throw the Toolkit away and get the full version? Best regards, Matthias Hofmann Tag: 128 bit math routines Tag: 245533
  - 9
    - HELP:CreateInstance Failing Hi guys, I haven't gotten any luck with this yet so ny suggestion or help is really appreciated. I have a COM exe server from which I'm trying to create an instance of SQLServer object so I use this: if(SUCCEEDED(OleInitialize(NULL))) { try { CoInitialize(NULL); _SQLServerPtr spSQLServer; HRESULT hr2 = spSQLServer.CreateInstance(__uuidof(SQLServer)); if (FAILED(hr2)) _com_issue_error(hr2); The problem is CreateInstance always fails with this error : Class not Registered error number : -2147221164 When I try the same method in an MFC API it works fine... Is it possible that being in a COM and instantiating another COM is a problem?? Thanks guys. Tag: 128 bit math routines Tag: 245529
  - 10
    - ocx callback function convert VC++ to VB6 Hi all, I'm using a 3rd part component (ocx) in vb6 and I have an example (see below) in VC++ how to use a callback function. I need to convert that example in VB6, any help would be appreciate. Thanks, Juky ....... // define callback function. int GetBuffer(int s, int w, int h, LPBYTE p) { ... // any code return 0; } ...... myOCX.SetCallBack((long)GetBuffer); Tag: 128 bit math routines Tag: 245528
  - 11
    - access to fonctions in the main application from dlls Hello, I am trying to have a set of functions (and variables) that I can access both from several dlls and from the core application. Right now, i am building a static library with these functions, but this same lib is loaded several times (in the core app and in each dll). To avoid that, does it make sense to build a dll rather than a static lib ? i.e. can I be sure that the dll will be loaded once only ? Is there a better solution such as exporting functions (and variables) directly from the core application rather than through this "in-between" lib ? Can you tell me how to proceed ? Thanks yomgui Tag: 128 bit math routines Tag: 245526
  - 12
    - get iocontrol code constants of NDIS IM using C# Hello, I'm try to access the extended Passthru from http://www.wd-3.com/031504/PTEx3_index.htm using C#. Any body know how to get the correct iocontrol codes of : IOCTL_PTUSERIO_ENUMERATE = ?; IOCTL_NDISPROT_OPEN_DEVICE = ?; There is an example using with ndisprot protocol driver: http://www.codeproject.com/csharp/SendRawPacket.asp?df=100&forumid=25245&select=1082543 // iocontrol code constants private const uint IOCTL_NDISPROT_QUERY_BINDING = 0x12C80C; private const uint IOCTL_NDISPROT_OPEN_DEVICE = 0x12C800; howee i dont khnow this codes from where ? Tag: 128 bit math routines Tag: 245521
  - 13
    - odd first character from read(0,...) Given this trivial little snippet, #include <stdio.h> int main() { char ch1,ch2; _sleep(1000); printf("? "); read(0,&ch1,1); read(0,&ch2,1); printf("%c%c\n",ch1,ch2); return 0; } I'm seeing behavior that I'm at loss to explain: c:\src> cl /o curious.exe foo.c ... c:\src> curious <type in 'a' prior to appearance of prompt> ? ax fx c:\src> curious <delay typing anything until prompt appears> ? ax ax c:\src> Why is the first character in the type-ahead case reported as 'f'? Same behavior seen when compiling with /MD (cl.exe's version string is 12.00.8804) Anyone know if this is a known issue and/or there is a workaround? (unfortunately, I'm stuck with using read() in the app that this snippet has been extricated from.) cheers zeb Tag: 128 bit math routines Tag: 245514
  - 14
    - Trapping First-Chance Exception Hi, I asked this question before but I can't remember it. In .NET 2002, how do you trap first-chance exceptions so that the debugger stops right at that point? Thanks Jack Tag: 128 bit math routines Tag: 245510
  - 15
    - error - type cast Hello Group, I will appreciate it if you guys could give me insight into this. I have : class CImageArrray { public: CImageArrray(unsigned long inElementSize = 0, long inBlockSize = 10); virtual ~CImageArrray(void); long numElements(void) const; bool isEmpty(void) const; unsigned long elementSize(void) const; long blockSize(void) const; etc ... etc ... private: etc ... protected: // Instance variables: CMemoryBlock m_elements; // Buffer of array items etc ... }; class CMemoryBlock { public: CMemoryBlock(unsigned long inSize = 0, bool inClearMem = false); virtual ~CMemoryBlock(void); etc ... private: // Instance variables void* m_memBlock; }; //Return a pointer into the array at the specified index. void* CImageArrray::elementPtr(long inIndex) const { return ((unsigned char*) (void*) m_elements) + (inIndex * m_elementSize); } I get compiler error as : error C2440: 'type cast' : cannot convert from 'const class CMemoryBlock' to 'void *' any suggestions? Thanks, Tag: 128 bit math routines Tag: 245507
  - 16
    - How to use initialization list? Suppose I have class as follows: class Test { private: TCHAR m_tchBuf[BUFSIZ + 1]; public: Test(void) //here I want to use the initialization list to initialize the member cariable m_tchBuf { m_tchBuf[0] = 0; //but I am initialzing it this way } }; How to initialize the above member variable in the initialization list of the class? Thanks, Tag: 128 bit math routines Tag: 245503
  - 17
    - WndRetProc messages I want to set up a global window hook to receive messages from windows after they have processed a message. The WndRetProc seems suitable for this. Will it give me info such as when a window receives or loses focus or if it resizes? Tag: 128 bit math routines Tag: 245489
  - 18
    - Check Profile Is there a way to check if the profile is loaded? What happens if LoadUserProfile is called when the profile is already loaded? Thanks. Tag: 128 bit math routines Tag: 245488
  - 19
    - Member ordering of struct/class I notice that some structs for instance NMTREEVIEW have their first item as an NMHDR, which enables the WndProc to treat it as an NMHDR and access those members without knowing that it's actually also an NMTREEVIEW. Question is will the same be possible with a class, e.g. class NMCustom { public: NMHDR //...(other member vars. and methods) }; and then in WM_NOTIFY handler LPNMHDR pnmtv = (LPNMHDR)lParam; if(pnmtv->code == TVN_MYCUSTOMCODE) { NMCustom* myobj = (NMCustom*)pnmtv; } Also would it still work if I used derivation of an NMHDR rather than placing it as the first member. Are there any reason why this shouldn't be done? Granted it is an upcast, but this is determined reliably at runtime by the application-specific code.... Thanks Tag: 128 bit math routines Tag: 245486
  - 20
    - destructor is not called Hi, How is this possible that destructor is not called when I quit my application? This is toolbar for IE. For one of my classes destructot is not called causing memory leaks. Class was created on the stack and there is no inheritance. thanks for advice Tag: 128 bit math routines Tag: 245482
  - 21
    - PostMessage synchronicity Hello If I call PostMessage twice to post a message to an HWND's WndProc, once before the other in the same function on a different thread (2) to the WndProc's thread (1), will it be guaranteed that they will be processed by the HWND's WndProc in that order, or is it possible one may 'jump the queue'? It doesn't matter on the timing, or the order respective to a different thread (3), but are the messages posted both from thread (2) both processed in the correct order? I need this to be true in order for my application to succeed with PostMessage. Thanks Tag: 128 bit math routines Tag: 245481
  - 22
    - Visual C++ /clr Porting Booth at TechEd 2005 Hi, my name is Kang Su Gatlin and I'm a program manager on the Visual C++ team. I'm writing to let you all know about an exciting thing that the Visual C++ team is doing at TechEd in Orlando this year (2005). We're hosting what we are calling a "C++ /clr Porting Booth", although it's actually in a room. The idea of this "C++ /clr Porting Booth" is that developers bring some subset of their C++ code (or something representative of their code), which they can currently build and run, and we'll help them port it to VS2005 and to get it compiled with the /clr switch. (It's important to bring code that you can compile and run, so that we can verify at the end that it works.) There are two benefits of your doing this Porting Booth activity: (1) You'll now have some code that can directly make use of the .NET functionality, but more importantly (2) You'll have first-hand knowledge and experience on how to do this with all of the code-base when you head back home. To get more info for yourself (or friends and colleagues) on participating in the "C++ /clr Porting Booth" at TechEd email us here: porting@microsoft.com. Thanks, Kang Su Gatlin Visual C++ Program Manager Report bugs and make suggestions here: http://msdn.microsoft.com/feedback Tag: 128 bit math routines Tag: 245472
  - 23
    - Get the Program Files directory How can I get the Program Files directory? Tks. Rui Tag: 128 bit math routines Tag: 245454
  - 24
    - Create a shortcup How can I create a shortcut? Tks. Rui Tag: 128 bit math routines Tag: 245453
  - 25
    - ISAPI Filter - HTTP_USER_AGENT Hi I would like to read the value of HTTP_USER_AGENT using GetHeader function, but it returns nothing :( CODE : DWORD CSagecrmfilterFilter::OnPreprocHeaders(CHttpFilterContext* pCtxt, PHTTP_FILTER_PREPROC_HEADERS pHeaderInfo) { char szUserAgent[1024]; DWORD dwSize = sizeof(szUserAgent); BOOL bHeader = pHeaderInfo->GetHeader(pCtxt->m_pFC, "HTTP_USER_AGENT", szUserAgent, &dwSize); // TODO: React to this notification accordingly and // return the appropriate status code return SF_STATUS_REQ_NEXT_NOTIFICATION; } Please help me. Thanks in Advance. Ram. Tag: 128 bit math routines Tag: 245446

Does any one know if the New Visual Studio 2005 will support QUAD precision
math routines?
Where sizeof(long double)==16 bytes

Top

Re: 128 bit math routines by Tim

Tim
Mon Apr 18 23:51:48 CDT 2005

TOB <TOB> wrote:
>
>Does any one know if the New Visual Studio 2005 will support QUAD precision
>math routines?
>Where sizeof(long double)==16 bytes

Quad precision floating point? Not a chance. No chip that Visual Studio
supports has such a thing in hardware (IEEE 754 does not even define such a
format), and no one is about to undertake a floating point simulation at
this point.
--
- Tim Roberts, timr@probo.com
Providenza & Boekelheide, Inc

Top