gcc - Questions about C Function Prototypes and Compilation -

with following code:

int main(){     printf("%f\n",multiply(2));     return 0; }  float multiply(float n){     return n * 2; } 

when try compile 1 warning: "'%f' expects 'double', argument has type 'int'" , 2 errors: "conflicting types 'multiply'", "previous implicit declaration of 'multiply' here."

question 1: guessing it's because, given compiler has no knowledge of function 'multiply' when comes across first time, invent prototype, , invented prototypes assume 'int' both returned , taken parameter. invented prototype "int multiply(int)", , hence errors. correct?

now, previous code won't compile. however, if break code in 2 files this:

#file1.c  int main(){     printf("%f\n",multiply(2));     return 0;  }  #file2.c float multiply(float n){     return n * 2; } 

and execute "gcc file1.c file2.c -o file" still give 1 warning (that printf expecting double getting int) errors won't show anymore , compile.

question 2: how come when break code 2 files compiles?

question 3: once run program above (the version split 2 files) result 0.0000 printed on screen. how come? guessing compiler again invented prototype doesn't match function, why 0 printed? , if change printf("%f") printf("%d") prints 1. again, explanation of what's going on behind scenes?

thanks lot in advance.

so invented prototype "int multiply(int)", , hence errors. correct?

absolutely. done backward compatibility pre-ansi c lacked function prototypes, , declared without type implicitly int. compiler compiles main, creates implicit definition of int multiply(int), when finds real definition, discovers lie, , tells it.

how come when break code 2 files compiles?

the compiler never discovers lie prototype, because compiles 1 file @ time: assumes multiply takes int, , returns int in main, , not find contradictions in multiply.c. running program produces undefined behavior, though.

once run program above (the version split 2 files) result 0.0000 printed on screen.

that's result of undefined behavior described above. program compile , link, because compiler thinks multiply takes int, never convert 2 2.0f, , multiply never find out. similarly, incorrect value computed doubling int reinterpreted float inside multiply function treated int again.