Hi All,

I have a makefile(Clearmake Makefile) and I have a target processor compiler for cross-compiling. One of the problems here is that I can not use GCC because I do not want to introduce another tool. So, in order to implement an incremental build, I had to use target processor option to generate the dependency file which will capture all the dependent file information to build the .o file and I was able to generate the same(i.e., a .d file). Now, I want to use this file(i.e., .d) to do my incremental build.

Tried referring GCC makefile options -MD and -MP and I was able to successfully understand the process. But, I was not able to implement the same using Clearlake with a dependent file generated by a target processor compiler.

clearmake can do it by itself and much better than gcc will ever be able. These are called derived objects (and also you may check about wink-in). These are not the depend files (which are generated by gcc) and which are not really required if you use the full power of clearmake.
If you use a pure clearmake environment (= a view) clearmake will automatically handle it and you do not need to take care about it at all (= you do not need those depend files or anything similar).

pan64, that is true. But, my files are view private files in a snapshot view.

ok. So the best solution would be to create derived objects (instead of view private files).
But if you don't want to do that you have still several possibilities:
1. go with gcc (gcc is not a make, but a c compiler). You need to use a compiler anyway (by the way which one do you use?)
2. use cpp instead of gcc, but probably it is just the same issue.
3. analyse c files as text files and process includes. You need to implement a script to collect the dependencies (although you may try to find it on the net).
4. use strace or similar low level tool to collect and produce that list. This solution may work with other targets too. Also requires a new script.

Hm. This sounds familiar:
https://www.linuxquestions.org/quest...ry-4175605788/

The problem is that I do not want to create derived objects because I have few files which are completely view private files. I am using the Windriver compiler option to generate the .d files. Now, how can I make use of these .d files to do an incremental build?

You modify your makefile accordingly. After working with this setup for over a year now, you should be familiar with how to make those modifications.

https://www.linuxquestions.org/quest...ry-4175605788/

You were referring to the above link which was of a different concept and purpose of using it. I think both the questions are different.

Yet the answer remains the same; modify your makefile to use/output/read/whatever you'd like in whatever location.

you can use the .d files as it was generated by gcc, exactly the same way (if I know it well).

Awesome !! ..I did not realize that modifying a makefile will really resolve my problem?

great answers