Do I have a gcc optimization bug or a C code problem?

Question

Test the following code:

#include 
#include 
main()
{
    const char *yytext=\"0\";
    const float f=(float)atof(yytext);
    siz         


        

Answer 1

This is bad C code. Your cast breaks C aliasing rules, and the optimiser is free do things that break this code. You will probably find that GCC has cheduled the size_t read before the floating-point write (to hide fp pipeline latency).

You can set the -fno-strict-aliasing switch, or use a union or a reinterpret_cast to reinterpret the value in a standards-compliant way.