Search Results

Search found 5015 results on 201 pages for 'compiler construction'.

Page 70/201 | < Previous Page | 66 67 68 69 70 71 72 73 74 75 76 77  | Next Page >

• In a Linux user space process what is the address of the vsyscall page?

  - by TomMD

  I would like to acquire the address of the vsyscall page for my own uses. I only have two ideas here: alter the compiler to store this information in some known location after it is given to __start, or read /proc/[pid]/maps. I really don't want to read /proc/ as that is slow and shouldn't be necessary. I also don't want to make compiler modifications. Does anyone have an alternative? Is there a symbol I should know about? Its at the point I'm tempted to stuff this functionality into an ioctl call in a module I've developed as part of this work!

  Read the article

• Porting C++-code from Windows to Unix: systemcalls colliding with name of functions

  - by marvin2k

  Hi I'm porting some crufty C++ Windows-code to Linux, which uses functions called "open" and "close" inside every class... Very bad style, or? Luckily that wasn't a problem in windows, since their systemcalls are named different. When I try to call the systemcalls open() or close() I'm getting some compiler error about "no matching function for call for class:open()". I can't rename all our functions named "class::open" and "class::close" in the whole code, and I have to use open() and close() since I'm working with serial ports. So my question is: How can I tell the compiler, which open I mean? How can I escape or hide the namespace of a class in C++?

  Read the article

• Passing a pointer to a function that doesn't match the requirements of the formal parameter

  - by Andreas Grech

  int valid (int x, int y) { return x + y; } int invalid (int x) { return x; } int func (int *f (int, int), int x, int y) { //f is a pointer to a function taking 2 ints and returning an int return f(x, y); } int main () { int val = func(valid, 1, 2), inval = func(invalid, 1, 2); // <- 'invalid' does not match the contract printf("Valid: %d\n", val); printf("Invalid: %d\n", inval); /* Output: * Valid: 3 * Invalid: 1 */ } At the line inval = func(invalid, 1, 2);, why am I not getting a compiler error? If func expects a pointer to a function taking 2 ints and I pass a pointer to a function that takes a single int, why isn't the compiler complaining? Also, since this is happening, what happens to the second parameter y in the invalid function?

  Read the article

• Visual Studio 2008 compiles anything in C++ file?

  - by Brad Pepers

  I noticed today that a source code file in a project was compiling even though it had junk at the top of it. It got me wondering what all would pass without error through the compiler. Here is an example of code that will not generate any error messages: what kind of weird behaviour is this??? #include "stdafx.h" // what is up? int foo(int bar) { bla bla bla????? return bar; } and more junk??? What in the world is the compiler doing to allow this code to compile without giving any error messages? I'm using Visual Studio 2008 and this is unmanaged C++ code. The foo function isn't actually generated in the object file so it can't be used but why no errors???

  Read the article

• VS2005 and VSS. Should .exes, .dlls, pdbs, .resource and generated .xml files go in VSS?

  - by Velika

  It seems like the following type of files need to be writable to be able to compile a solution. .exe .dll .pdb (if debugging) .resource (?) .xml (at least the system generated ones. I'm not sure if they are just all XML documentationf iles) If they were checked in and could only be overwritten by the compiler by the person who had them checked out, then checking them into sourcesafe and requiring a checkout to modify them, which is attemtped by the compiler each time you run, would interfer with other developer's ability to debug and run. What is the best practice? It seems like VSS automatically adds everything.

  Read the article

• Why is 'virtual' optional for overridden methods in derived classes?

  - by squelart

  When a method is declared as virtual in a class, its overrides in derived classes are automatically considered virtual as well, and the C++ language makes this keyword virtual optional in this case: class Base { virtual void f(); }; class Derived : public Base { void f(); // 'virtual' is optional but implied. }; My question is: What is the rationale for making virtual optional? I know that it is not absolutely necessary for the compiler to be told that, but I would think that developers would benefit if such a constraint was enforced by the compiler. E.g., sometimes when I read others' code I wonder if a method is virtual and I have to track down its superclasses to determine that. And some coding standards (Google) make it a 'must' to put the virtual keyword in all subclasses.

  Read the article

• How to include parenthesis for AS-IS output in format string

  - by DotDot

  I want to see, Hello {} in the output, but the following gives compiler errors Console.WriteLine("{0} \{\}", "Hello");

  Read the article

• Is there a way to serialize automatically enums as int?

  - by FireAphis

  Hello, Is there a way to serialize enums automatically as int? Every time I define a new enum and write std::stringstream stream; stream << myenum1; stream >> myenum2; the compiler complains that the operators << and are not defined. Do you know a way to tell the compiler to treat enums as plain int's? What makes the problem harder is that, actually, the serialization is inside a template. Something like this: template <typename T> void serialize(const T& value) { std::stringstream stream; stream << value; } So I cannot add any casts :( Maybe I can specialize it somehow? Thank you.

  Read the article

• Build multiple executables with small changes in objects

  - by Morgoth

  Consider the following Makefile COMP = compiler OBJECTS = file1 \ file2 \ file3 \ file4 \ file5_suffix \ file6 \ file7 \ file8 \ file9_suffix \ file10 all: $(OBJECTS) $(COMP) $(OBJECTS) -o bin/executable_suffix Is there an easy way to compile multiple executables for different values of suffix? For example, the equivalent of COMP = compiler OBJECTS1 = file1 \ file2 \ file3 \ file4 \ file5_s1 \ file6 \ file7 \ file8 \ file9_s1 \ file10 OBJECTS2 = file1 \ file2 \ file3 \ file4 \ file5_s2 \ file6 \ file7 \ file8 \ file9_s2 \ file10 all: $(OBJECTS1) $(OBJECTS2) $(COMP) $(OBJECTS1) -o bin/executable_s1 $(COMP) $(OBJECTS2) -o bin/executable_s2 but without redefining the whole list of objects? In the real life case I am dealing with, there might be 50+ objects and a dozen binaries to build, with only small changes between the object list each time, so it would be nice not to have to list all the objects each time.

  Read the article

• Is sizeof in C++ evaluated at compilation time or run time?

  - by ogzylz

  For example result of this code snippet depends on which machine: the compiler machine or the machine executable file works? sizeof(short int)

  Read the article

• F#: Can't hide a type abbreviation in a signature? Why not?

  - by Nels Beckman

  In F#, I'd like to have what I see as a fairly standard Abstract Datatype: // in ADT.fsi module ADT type my_Type // in ADT.fs module ADT type my_Type = int In other words, code inside the module knows that my_Type is an int, but code outside does not. However, F# seems to have a restriction where type abbreviations specifically cannot be hidden by a signature. This code gives a compiler error, and the restriction is described here. If my_Type were instead a discriminated union, then there is no compiler error. My question is, why the restriction? I seem to remember being able to do this in SML and Ocaml, and furthermore, isn't this a pretty standard thing to do when creating an abstract datatype? Thanks

  Read the article

• Can you use #defined values in if statements (In C programs)?

  - by Jordan S

  I am new at C programming. I thought when you type something like #define Const 5000 that the compiler just replaces every instance of Const with 5000 at compile time. Is that wrong? I try doing this in my code and I get a syntax error. Why can't i do this? #define STEPS_PER_REV 12345 ... in some function if(CurrentPosition >= STEPS_PER_REV) { // do some stuff here } The compiler complains about the if statement with a syntax error that gives me no details.

  Read the article

• C++ packing a typdef enum

  - by Sagar

  typedef enum BeNeLux { BELGIUM, NETHERLANDS, LUXEMBURG } _PACKAGE_ BeNeLux; When I try to compile this with C++ Compiler, I am getting errors, but it seems to work fine with a C compiler. So here's the question. Is it possible to pack an enum in C++, or can someone see why I would get the error? The error is: "semicolon missing after declaration of BeNeLux". I know, after checking and rechecking, that there definitely is a semicolon there, and in any places required in the rest of the code.

  Read the article

• error C2065: 'AfxBeginThread' : undeclared identifier

  - by bdhar

  I have a simple ATL Service in which I have included the following statement. AfxBeginThread(WorkerThread, this, THREAD_PRIORITY_NORMAL, 0, 0); But the compiler complains that error C2065: 'AfxBeginThread' : undeclared identifier But in MSDN i found that AfxBeginThread is found in AFXWIN.H If I include AFXWIN.H, the compiler complains that fatal error C1189: #error : WINDOWS.H already included. MFC apps must not #include <windows.h> What am I missing here? Thanks. PS: I am using Visual Studio 6.0 Enterprise Edition in Windows XP SP2.

  Read the article

• [c++] How to create a std::ofstream to a temp file?

  - by dehmann

  Okay, mkstemp is the preferred way to create a temp file in POSIX. But it opens the file and returns an int, which is a file descriptor. From that I can only create a FILE*, but not an std::ofstream, which I would prefer in C++. (Apparently, on AIX and some other systems, you can create an std::ofstream from a file descriptor, but my compiler complains when I try that.) I know I could get a temp file name with tmpnam and then open my own ofstream with it, but that's apparently unsafe due to race conditions, and results in a compiler warning (g++ v3.4. on Linux): warning: the use of `tmpnam' is dangerous, better use `mkstemp' So, is there any portable way to create an std::ofstream to a temp file?

  Read the article

• Setup Jetty 7 with JSP engine

  - by Justin

  I've been trying to get Jetty to run my web app via a custom launcher (embedded). I am trying to figure out how to tell Jetty which java compiler to use for JSPs. I want to do what java -jar start.jar -OPTIONS=jsp does, but without using start.jar. Here is what shows on the console: Javac exception, Unable to find a javac compiler; com.sun.tools.javac.Main is not on the classpath. Perhaps JAVA_HOME does not point to the JDK Do I need to put the javac libraries into my classpath?

  Read the article

• Forward declaring an enum in c++

  - by szevvy

  Hi guys, I'm trying to do something like the following: enum E; void Foo(E e); enum E {A, B, C}; which the compiler rejects. I've had a quick look on Google and the consensus seems to be "you can't do it", but I can't understand why. Can anyone explain? Many thanks. Clarification 2: I'm doing this as I have private methods in a class that take said enum, and I do not want the enum's values exposed - so, for example, I do not want anyone to know that E is defined as enum E { FUNCTIONALITY_NORMAL, FUNCTIONALITY_RESTRICTED, FUNCTIONALITY_FOR_PROJECT_X } as project X is not something I want my users to know about. So, I wanted to forward declare the enum so I could put the private methods in the header file, declare the enum internally in the cpp, and distribute the built library file and header to people. As for the compiler - it's GCC.

  Read the article

• Inheritance and Implicit Type Casting

  - by Josué Molina

  Suppose I have the following three classes: class Animal {}; class Human : public Animal {}; class Dog : public Animal { public: void setOwner(Animal* owner) { this->owner = owner; } private: Animal* owner; }; Why is the following allowed, and what exactly is happening? Dog d; Human h; d.setOwner(&h); // ? At first, I tried to cast it like this d.setOwner(&(Animal)h), but the compiler gave me a warning, and I hit a run-time error. Edit: the warning the compiler gave me was "taking address of temporary". Why is this so?

  Read the article

• Xcode 3.2 not recompiling changed files

  - by Jason

  I've been programming a new iPhone application and everything has been chugging along fine, until this afternoon when I noticed that any changes I was making were not being reflected in the actual app! Even code that throws errors isn't getting picked up by the compiler. To get around this, I have taken to the following procedure: Finish making my edits, and add code to the file which will always throw a compiler error. Right-click the file I have just edited, and hit "Compile" to just compile that one file; it will throw the error. Remove the error-producing code, and again right-click and "Compile" the one file. Build the whole project This is quite annoying, and I can't figure out why this would suddenly happen to my Xcode project. Any thoughts on what could be causing this, and how to fix it?

  Read the article

• C++ packing a typedef enum

  - by Sagar

  typedef enum BeNeLux { BELGIUM, NETHERLANDS, LUXEMBURG } _PACKAGE_ BeNeLux; When I try to compile this with C++ Compiler, I am getting errors, but it seems to work fine with a C compiler. So here's the question. Is it possible to pack an enum in C++, or can someone see why I would get the error? The error is: "semicolon missing after declaration of BeNeLux". I know, after checking and rechecking, that there definitely is a semicolon there, and in any places required in the rest of the code.

  Read the article

• Problem with default member functions of class in C++ (constructor, destructor, operator=, copy cons

  - by Narek

  We know that compiler generates some member functions for user-defined class if that member functions are not defined but used, isn't it. So I have this kind of code: class AA { }; void main() { AA a; AA b(a); a = b; } This code works fine. I mean no compiler error. But the following code.... class AA { int member1; int member2; }; But this code gives an run time error, because variable "a" is used without being iniltialized!!! So my question is this: when we instantiate an int, it has a value. So why the default constructer doesn't work and by using those two int numbers initializes variable "a"??

  Read the article

• Is 'bool' a basic datatype in C++ ?

  - by Naveen

  I got this doubt while writing some code. Is 'bool' a basic datatype defined in the C++ standard or is it some sort of extension provided by the compiler ? I got this doubt because Win32 has 'BOOL' which is nothing but a typedef of long. Also what happens if I do something like this: int i = true; Is it "always" guaranteed that variable i will have value 1 or is it again depends on the compiler I am using ? Further for some Win32 APIs which accept BOOL as the parameter what happens if I pass bool variable?

  Read the article

• What to do when using Contract.Assert(true) and the method must return something?

  - by devoured elysium

  I have a bit of code with the following logic: //pseudo-code foreach (element in elementList) { if (element is whatever) return element; } } In theory, there is always one element that is whatever, so this method should pose no problems. In any case, I've put an assertion on the end of the method just to be sure: //pseudo-code foreach (element in elementList) { if (element is whatever) return element; } } Contract.Assert(true, "Invalid state!"); The problem is that as this method has to return something, and the compiler doesn't understand that the assertion will break the program execution. Before using Contracts, in these kind of situations, I used to throw an Exception, which solved the problem. How would you handle this with Contract.Assert()? Returning null or default(element_type) after the Contract.Assert() call knowing it will never be called and shutting up the compiler? Or is there any other more elegant way of doing this? Thanks

  Read the article

• Why not System.Void?

  - by Stewart

  I have no practical reason for knowing this answer, but I'm curious anyway... In C#, trying to use System.Void will produce a compilation error: error CS0673: System.Void cannot be used from C# -- use typeof(void) to get the void type object As I understood it, void is simply an alias of System.Void. So, I don't understand why 'System.Void' can't be used directly as you might with 'string' for 'System.String' for example. I would love to read an explanation for this! Incidentally, System.Void can be successfully used with the Mono compiler, instead of Microsoft's .Net, and there it appears equivalent to using the void keyword. This must therefore be a compiler-enforced restriction rather than a CLR restriction, right?

  Read the article

• auto_inline - inadequate documentation

  - by Mick

  I want to disable inlining for a particular function. What the compiler does for everything else should be as specified in the project properties. I found a page on a forum which suggested the following: #pragma auto_inline(off) void func() { } #pragma auto_inline() The author suggested that calling auto_inline() with no arguments will set the compiler to revert to doing whatever the default action was before the call to auto_inline(off). Can anyone confirm that this works for visual studio 2008? I ask because the VS2008 documentation makes no mention at all of what happens if you call this function with no arguments.

  Read the article

< Previous Page | 66 67 68 69 70 71 72 73 74 75 76 77  | Next Page >