Where are C/C++ main function's parameters?
C++CParametersLocationMainC++ Problem Overview
In C/C++, the main function receives parameters which are of type char*
.
int main(int argc, char* argv[]){
return 0;
}
argv
is an array of char*
, and points to strings. Where are these string located? Are they on the heap, stack, or somewhere else?
C++ Solutions
Solution 1 - C++
They are compiler magic, and implementation-dependent.
Solution 2 - C++
Here's what the C standard (n1256) says:
5.1.2.2.1 Program startup
...
2 If they are declared, the parameters to the main function shall obey the following constraints:
- The value of argc shall be nonnegative.
- argv[argc] shall be a null pointer.
- If the value of argc is greater than zero, the array members argv[0] through argv[argc-1] inclusive shall contain pointers to strings, which are given implementation-defined values by the host environment prior to program startup. The intent is to supply to the program information determined prior to program startup from elsewhere in the hosted environment. If the host environment is not capable of supplying strings with letters in both uppercase and lowercase, the implementation shall ensure that the strings are received in lowercase.
- If the value of argc is greater than zero, the string pointed to by argv[0] represents the program name; argv[0][0] shall be the null character if the program name is not available from the host environment. If the value of argc is greater than one, the strings pointed to by argv1 through argv[argc-1] represent the program parameters.
- The parameters argc and argv and the strings pointed to by the argv array shall be modifiable by the program, and retain their last-stored values between program startup and program termination.
The last bullet is the most interesting wrt where the string values are stored. It doesn't specify heap or stack, but it does require that the strings be writable and have static extent, which places some limits on where the string contents may be located. As others have said, the exact details will depend on the implementation.
Solution 3 - C++
It's actually a combination of compiler dependence and operating system dependence. main()
is a function just like any other C function, so the location of the two parameters argc
and argv
will follow standard for the compiler on the platform. e.g. for most C compilers targeting x86 they will be on the stack just above the return address and the saved base pointer (the stack grows downwards, remember). On x86_64 parameters are passed in registers, so argc
will be in %edi
and argv
will be in %rsi
. Code in the main function generated by the compiler then copies them to the stack, and that is where later references point. This is so the registers can be used for function calls from main
.
The block of char*
s that argv points to and the actual sequences of characters could be anywhere. They will start in some operating system defined location and may be copied by the pre-amble code that the linker generates to the stack or somewhere else. You'll have to look at the code for exec()
and the assembler pre-amble generated by the linker to find out.
Solution 4 - C++
The answer to this question is compiler-dependent. This means it is not treated in the C standard, so anyone can implement that as he or she would like to. This is normal since also operating systems don't have a common accepted, standard way to start processes and finish them.
Let's imagine a simple, why-not scenario.
The process receives by some mechanism the arguments written in the command line. argc is then just an int which is pushed to the stack by the bootstrap function the compiler put as the entry point for the process of the program (part of the runtime). The actual values are get from the operating system, and can be, say, written in a memory block of the Heap. Then the argv vector is built and the address to its first position also pushed into the stack.
Then the function main(), which must be provided by the programmer, is called, and its return value is saved for later (nearly inmediate) use. The structures in the Heap are freed, and the exit code obtained for main is exported to the operating system. The process finishes.
Solution 5 - C++
These parameters are no different than any other function's parameters. If the architecture's calling sequence requires parameters to go through stack they are on stack. If, like on, x86-64 some parameters go in registers these also go in registers.
Solution 6 - C++
As pmg
mentions, when main
is called recursively, it's up to the caller where the arguments point to. Basically the answer is the same on the original invocation of main
, except that the "caller" is the C implementation/OS.
On UNIX-y systems, the strings that argv
points to, the argv
pointers themselves, and the process's initial environment variables are almost always stored at the very top of the stack.
Solution 7 - C++
As many other answers here point out, the precise mechanism a compiler implementation uses to pass arguments to main is unspecified by the standard (as is the mechanism a compiler uses to pass any arguments to a function). Strictly speaking, the compiler need not even pass anything useful in those parameters, since the values are implementation-defined. But neither of these are particularly helpful answers.
The typical C (or C++) program is compiled for what's known as a 'hosted' execution environment (using function main()
as the starting point of your program is one of the requirements for a hosted environment). The key thing to know is that the compiler arranges things so that when the executable is launched by the operating system, the compiler's runtime gets control initially - not the main()
function. The runtime's initialization code performs whatever initialization is necessary, including allocating memory for the arguments to main()
, then it transfers control to main()
.
The memory for the arguments to main()
could come from the heap, could be allocated on the stack (possibly using techniques that aren't available to standard C code), or could use statically allocated memory, though that's a less likely option just because it's less flexible. The standard does require that the memory used for the strings pointed to by argv
are modifiable and that modifications made to those string persist throughout the program's lifetime.
Just be aware that before execution reaches main()
, quite a bit of code has already been run that's setting up the environment for your program to run in.
Solution 8 - C++
The argument list is part of the process environment, similar to (but distinct from) environment variables.
Solution 9 - C++
Usually it is unknown where they are.
#include <stdlib.h>
#include <string.h>
int main(int argc, char *argv[]) {
char **foo;
char *bar[] = {"foo", "bar"};
(void)argv; /* avoid unused argv warning */
foo = malloc(sizeof *foo);
foo[0] = malloc(42);
strcpy(foo[0], "forty two");
/* where is foo located? stack? heap? somewhere else? */
if (argc != 42) main(42, foo); else return 0;
/* where is bar located? stack? heap? somewhere else? */
if (argc != 43) main(43, bar); else return 0;
/* except for the fact that bar elements
** point to unmodifiable strings
** this call to main is perfectably reasonable */
return 0;
/* please ignore memory leaks, thank you */
}
Solution 10 - C++
While you are able to access to the actual parameters, I think their actual location does not matter at all.