Results 1 to 3 of 3
I am trying to install a couple of programs on Angstrom linux on a gumstix overo with an armv7 chip. The problem I am running into right now is that ...
Enjoy an ad free experience by logging in. Not a member yet? Register.
- 10-09-2010 #1
- Join Date
- Oct 2010
Questions about header files
The problem I am running into right now is that the last 2 programs I need to get installed for now are looking for 2 header files. One says its in /net the other /sys. So I assume it was looking in this area for header files as it finds some in both the /net and /sys but there is a couple headers that are there that it does not see as well as a header thats not on there. So what I thought I should do is just copy the .h from the kernel source and copy it to the location /usr/include/net and /usr/include/sys But when I re ran the program it still did not see the new header file I put in there nor the couple that were there to begin with.
Does this mean I need to include these specific headers in my $PATH or do the need to be linked or what would cause the ./configure of a program to only find some files in /usr/include/net and usr/include/sys but not others in there. I hope this is something easy that I am just not thinking of or not grasping from my recent endless reading I have been doing.
Thanks for any tips or help with this. If you need more info from me about anything please ask and I will happily provide it, tried to give all I thought was relevant.
- 10-10-2010 #2
- Join Date
- Apr 2009
- I can be found either 40 miles west of Chicago, or in a galaxy far, far away.
There are several issues here, but the biggest is that you apparently don't understand how C programs are written, compiled, and how you tell them where to find header files. So, the first question I have for you is how are you trying to install these components? What are the steps you take to install them?Sometimes, real fast is almost as good as real time.
Just remember, Semper Gumbi - always be flexible!
- 10-10-2010 #3
- Join Date
- Oct 2010
Basically there are two programs I can not get to install on my system through openembedded right now so I need to compile them on my overo natively. I understand the basics but do not get why if there is a header that was built and installed while the OS was being built and packaged that can not be found with its prebuilt paths when others in the same directory are found. I also dont want to have to modify my kernel recipe right now to include the couple missing header files. So thought would be easier to just put the couple where they need to be and link them if thats a possible way.
The steps I am doing is putting the tarball on the overo
The programs error out in the ./configure complaining of the few missing headers. I searched the entire file system to verify they did not exsist at all in case they did but just didnt have the right link or path. But they are not any where on there. So next I went to the Kernel source I am using to see if they exsisted in the source, which they did so I know eventually I need to fix my kernel recipe to include those missing ones, but would like to to try and fix the image I am working on now while it takes the next few days to rebuild the new kernel once I get the recipe fixed and building again.
Everything else seems to work fine and have been running on this image for a few weeks now. I try and do as much with a image until I have no choice but to build all the fixes I have done on the previous image to the new image. Hope this better explains why I asked this.
edit: Forgot to say also that the 2 programs I am trying to install, install fine with no complaints on Ubuntu, Fedora and BSD, and I verified the only difference is between these headers. Thanks