In the source code of C library functions, are native system calls to the OS directly made (eg. in the fopen() function, the read system call) or is it that the C program calls the corresponding win32 API functions (in Windows) and POSIX functions (in Linux) ?
I can't answer for Win32, but for POSIX systems, there's generally no difference between a system call and the corresponding C function provided (as in the open system call vs POSIX open()). The POSIX functions are just wrappers around a system call. As such, a C library's fopen() will just call its own open() wrapper instead of directly making a system call, but in the end there's no real difference.
In windows the library functions do have to make system calls, just like in Linux...
Sometimes they have to do a lot of work to make that call for you....
Compare fopen() to the windows OpenFile()... you'll get the idea.
On Windows systems, the C library functions are usually accessed from a DLL. Internally, that code converts calls to the native format (eg: setting up an LPSECURITY_ATTRIBUTES structure, invoking CreateFile, etc), returning the results in a format that complies with C conventions. Static linking is also used at times, although much more rarely so simply because it isn't very flexible/extensible (making security updates and such problematic, as well). However, in that case the application is linked directly with the LIB file; ie: the above code would be present within the application itself. Also, a C library call would most likely translate to a "CALL to a hard-coded address", which is somewhat more efficient than the "JMPXX to index into jump table" (symbols which are fixed-up by the *loader* (ie: not at link time), that is) technique used in "dynamic" linking.Originally Posted by daudiam
Last edited by Sebastiani; 01-23-2011 at 05:56 PM. Reason: gramer, clarific, rmv
Code:#include <cmath> #include <complex> bool euler_flip(bool value) { return std::pow ( std::complex<float>(std::exp(1.0)), std::complex<float>(0, 1) * std::complex<float>(std::atan(1.0) *(1 << (value + 2))) ).real() < 0; }
