One way to use the shellshock.h header is to:
- Create a 'Shellshock' object
- Resolve functions that you intend to use with the 'load_' member functions
- Utilize a singular function and make sure that all variables are created
on the stack
Once something like this is achieved, you can compile the binary to an object
file and dump the .text
section out to a whatever you like. That dump should
be position independent.
#include "shellshock/shellshock.h"
typedef int (WINAPI * MessageBoxA_t)(HWND, LPCSTR, LPCSTR, UINT);
extern "C" int payload_cpp(void) {
auto ss = ss::shellshock();
// Load target function into a temporary variable.
char szMessageBoxA[] = "MessageBoxA";
auto fMessageBoxA = ss.find_user32_func<MessageBoxA_t>(szMessageBoxA);
// Perform function call
char szTitle[] = "Shellshock";
char szMessage[] = "Success.";
fMessageBoxA && fMessageBoxA(nullptr, szMessage, szTitle, 0);
return 0;
}
If you want the payload to be immediately exported to a file,
you can do the following:
// Payload that will be exported to shellcode
extern "C" auto payload() noexcept -> void { (...) }
// Immediately after the function ends
void stub() {
return;
}
int main() {
auto pd = ss::payload_data::build_from_payload(payload, stub);
pd.extract_to_file("shellcode.bin");
}
Compiling this code to an executable should export the code to the specified
file location.
Compiling this code to an object should mean that the payload
function of
the .text
section is out new position independent executable.
NOTE: Compilation may fail if position-independent-code is not enabled AND/OR if function sections are enabled.