LinuxQuestions.org

LinuxQuestions.org (/questions/)
-   Linux - Newbie (https://www.linuxquestions.org/questions/linux-newbie-8/)
-   -   Header file problem - SVGAlib (https://www.linuxquestions.org/questions/linux-newbie-8/header-file-problem-svgalib-769365/)

Tim-Deale 11-15-2009 09:42 PM

Header file problem - SVGAlib
 
I'm a newbie to Linux programming (I'm a hardware engineer by practice) - the last C programming I did was 10+ years ago... It was K&R C with a Borland IDE.

I'm using Fedora 10, programming using the Anjuta IDE (although I ran into the same problem with gcc at the shell). I have installed the SVGAlib library with Fedora's yum utility. It seems to have loaded without any problems. However, when I compile code to include the <vga.h> and/or the <svga.h> header file(s) I get the error:

main.c:21:17: error: vga.h: No such file or directory

I also get the warning:

main.c:29:warning: implicit declaration of function 'vga_init'

which I expect since the compiler didn't recognize the header file in the first place. I'm suspecting the compiler is expecting the header file to be located in the same location as the rest of the header files (/user/include) but it is not. When yum installed it, it was located at:

/usr/src/kernels/2.6.27.38-170.2.113.fc10.i686/include/video

My question is "how can I explicitly state in my code to have the compiler look at the /user/src/kernels... location rather than the normal location for the header file?

For reference, below is my code. It's simple, just to test the compiling and see that it works... Of course it doesn't...

Can anyone help?

<><><><><><><><><>

#include <stdio.h>
#include <vga.h>

int main()
{
printf("Hello world.\n"); //started with this
//and added the following

vga_init(); // initialize VGA graphics
vga_setmode(1); // 320 x 200 pixels, 16 colors

vga_setmode(TEXT); // change back to text display

return 0;
}

Nylex 11-16-2009 12:21 AM

gcc's -I option lets specify further directories to be searched for headers, e.g.

gcc stuff.c -o stuff -I/stuff

will look for headers in /stuff, in addition to the standard directories. I'm not sure where you set flags for gcc in Anjuta (or even set which compiler it's using, but it's most likely to be gcc), I'm afraid. You'll also need to link the SVGA library, for which you use -lvga, according to the tutorial on the SVGAlib website.

knudfl 11-16-2009 08:40 AM

.
# yum -y install svgalib-devel

.. will provide the headers ' vga.h ' etc.
.. and usr/lib/libvga.so .

Tim-Deale 11-30-2009 07:09 PM

I'm back from a Thanksgiving trip.

I did install svgalib via yum - see initial posting. It still doesn't compile. Any other suggestions?

exvor 11-30-2009 07:45 PM

You should never use header files from the kernel for building external programs. This is a NO NO unless your building a driver for the running kernel. You should always use the sanitized ones that were used for building glibc. In your example you are trying to use vga.h and it is not being found. You probably don't have the library for it installed in your system. That or the development stuff is not there who knows. Install svgalib either from source or from the development package and this should clear things up.

knudfl 12-01-2009 02:38 AM

Quote:

It still doesn't compile.
Missing svgalib-devel


Did you install the -devel package ? ?

.....

Tim-Deale 12-02-2009 10:27 PM

OK, I found the problem, and the fix. Remember, I'm new to programming in the linux environment...

The way I found the fix was that I ran into the same problem with the math library using the sqrt() function. Web searches on that problem lead me to the solution. From the shell, using the gcc compiler, I needed to compile with the -lm option (linking in the libm.a library). Going back to the svgalib problem, I compiled with the -lvga option and it now all compiles just fine, no errors.

Now I want to use the Anjuta IDE. How do I tell the IDE to link using the -lm and -lvga options? Can anyone help?

Thanks in advance.


All times are GMT -5. The time now is 08:40 PM.