Need support regarding the execution order of GNU Make rules

[Basil Mathew]

Dear Support,

My name is Basil Mathew. I work as a support engineer, for a Windows
Eclipse based build platform that uses GNU Make scripts, in my
organization. I have been asked to take over the GNU Make script support
and maintenance for the future. I was going through our GNU Make scripts
for build with the help of the GNU Make Manual and the book “Managing
Projects With Gnu Make” for the last week and half. It has been difficult,
but I am slowly making progress.

I have a question(not related to implementation, but more related to why it
happens the way it happens) for which I have not been able to find a
suitable explanation neither in the manual nor over the internet. Would it
be possible for some one to have a look into this question and help me out?
I would really appreciate the help.

*Version Info:*

Running "gmake -version" at the command prompt gives me this output.

C:\>gmake -version
GNU Make 3.80
Copyright (C) 2002  Free Software Foundation, Inc.
This is free software; see the source for copying conditions.
There is NO warranty; not even for MERCHANTABILITY or FITNESS FOR A
PARTICULAR PURPOSE.

The cpp.exe version that we are using as of now is "GNU CPP version
egcs-2.91.57 19980901 (egcs-1.1 release) (80386, BSD syntax)"

*Build Environment Basics:*

In our build environment, the compilation of a C-file into the object file
is a 2 step process as described below. We use the GNU’s “cpp.exe” to
generate the dependency file and then compile the C source file using the
specific compiler that is configured for the project. We have 2 separate
make rules to get this done.

We generate the list of object files to be created as mentioned below

sources_c = $(filter %.c,$(GEN_FILE_C))
objects_c = $(sources_c:.c=.o)

where the variable “GEN_FILE_C” will contain the list of input C source
files. A project usually contains more than 30 C Source files.

The rule for compilation is defined as

build_c: $(objects_c)
%.o: %.c %.d
    !!!!Execute the Compiler Call to create the object file

The rule for dependency generation is defined as

%.d: %.c
    !!!!Execute the cpp.exe call to generate the dependency file

We invoke the make rule “build_c”. Also we invoke the makefile with the
“-j” option that takes “ENV{NUMBER_OF_PROCESSORS}” as the argument.

*Question:*

I find that the build works fine, no problem with the build.
Only problem that I have is that I cannot explain why the dependencies
generation phase for all the 30 or more C source files are executed first
completely before the compilation commands for any C-source files are
executed.
I was assuming that the dependency generation for one C-source file will be
completed followed by the compilation commands for the particular C-source
file followed by the dependency generation for next C-source file and so on.

Can you please explain the reason behind this if you are aware of this?

Thanks & Regards
Basil Mathew.