Like Tree1Likes
  • 1 Post By ledow

Help with the size of the object files and extended source file

This is a discussion on Help with the size of the object files and extended source file within the C Programming forums, part of the General Programming Boards category; 1.c Code: #include<stdio.h> main() { } commands used $ cc -E 1.c -o 1.i <--to create extd source file $ ...

  1. #1
    Registered User
    Join Date
    Nov 2011
    Posts
    32

    Help with the size of the object files and extended source file

    1.c
    Code:
    #include<stdio.h>
    main()
    {
    }
    commands used

    $ cc -E 1.c -o 1.i <--to create extd source file
    $ cc -c 1.c <----to create object file
    2.c
    Code:
    #include<stdio.h>
    #include<string.h>
    main()
    {
    }
    commands used

    $ cc -E 2.c -o 2.i
    $ cc -c 2.c
    When I use 'll' command (long format)
    -rw-r--r-- 1 user user 29 2011-12-14 01:18 1.c
    -rw-r--r-- 1 user user 17036 2011-12-14 01:16 1.i
    -rw-r--r-- 1 user user 666 2011-12-14 01:17 1.o
    -rw-r--r-- 1 user user 48 2011-12-14 01:22 2.c
    -rw-r--r-- 1 user user 38432 2011-12-14 01:18 2.i
    -rw-r--r-- 1 user user 666 2011-12-14 01:18 2.o
    red=> size of file.

    As you can see the size of the 1.o and 2.o are same and the size of the 1.i is much greater than that of the 1.o (object file).

    Can someone enlighten me on this?
    why and how the included files (header files) are removed after 'PREPROCESSING STAGE'.

  2. #2
    Registered User hk_mp5kpdw's Avatar
    Join Date
    Jan 2002
    Location
    Northern Virginia/Washington DC Metropolitan Area
    Posts
    3,806
    Header files are text files that mostly contain things like function prototypes and define statements. They can be lengthy and therefore may have a large file size. This does not at all equate to compiled code size. The size of the header files that get included - and are represented by those .i files - ultimately don't have any bearing at all with the size of the object code files. The only thing that does matter is actual code, not function prototypes and defines which are meaningless once you have an actual compiled object file.
    "Owners of dogs will have noticed that, if you provide them with food and water and shelter and affection, they will think you are god. Whereas owners of cats are compelled to realize that, if you provide them with food and water and shelter and affection, they draw the conclusion that they are gods."
    -Christopher Hitchens

  3. #3
    Registered User ledow's Avatar
    Join Date
    Dec 2011
    Posts
    435
    Compilers are intelligent. If you have a 7,000,000 line header file but only use one definition or variable in it, it will ONLY include that one variable. Your two programs are identical in operation and don't use anything from string.h at all, so first the preprocessor will actually make one large file that contains stdio.h, string.h and your code all joined together (that's what a preprocessor does). The compiler will read through it all, realise that 99% of it is never used and only "compile" the bits that actually do something. It won't even compile functions that are never called, in most modern compilers, so you might have a 1,000,000 line function that never even makes it into the final executable.

    Hence your two output binaries are not only the same size but probably absolutely identical. In the same way, changing a variable name won't change the object code at all, because variable names aren't understood by the machine and the compiler has no need to put them in there (unless you've deliberately included them by enabled debug modes, profiling, etc.). However, the compiler is *not* always smart enough to detect two pieces of similar *code* that do the same job (e.g. i=2+2 and i = 2; i+= 2 so even the most minor of changes to the "processing path" (if you like, the instructions the computer has to execute to come to your answer) can change the object code (some compilers will spot simple changes like the above, of course, but not all).

    It's no different to sending a translator a pile of documents but telling them that only the main document and it's footnotes needs translation. If the main document doesn't reference anything in the 1000's of pages you've given them, they won't translate them all because you don't need them. And if the documents reference only one paragraph in a huge volume, they will probably only translate that paragraph for you.

    Be thankful it works this way, or every program would probably be twice the size it is! There are tools that can do this sort of "editing" to finished binaries too - see the Unix "strip" program - and they can often chop your executables in half if you've linked in too much stuff that never actually gets used.
    chibi.coder likes this.

Popular pages Recent additions subscribe to a feed

Similar Threads

  1. Calling c object file into c source file
    By prabu in forum C Programming
    Replies: 5
    Last Post: 11-30-2010, 11:15 AM
  2. Source-code file size? [Poll]
    By Devil Panther in forum General Discussions
    Replies: 13
    Last Post: 09-18-2009, 03:47 PM
  3. Appending to files and getting file size
    By mike_g in forum C Programming
    Replies: 4
    Last Post: 07-24-2008, 01:34 PM
  4. compiling source files and header file at once?
    By 1jackjack in forum C Programming
    Replies: 10
    Last Post: 05-04-2007, 12:06 PM
  5. Makefiles, Object files, Source files, OH MY!
    By Inquirer in forum Linux Programming
    Replies: 2
    Last Post: 04-29-2003, 11:36 PM

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21