Makefiles in multiple directories - makefile

I want to build an app and I have multiple modules stored in multiple directories. I've decided to follow this idea, i.e. to have a makefile in each directory and then to merge it. But - as a beginner programmer - I still do not see how to do that. First of all, how would such "partial" makefiles look like. They cannot have main function as there can be only one per binary, though when I try to compile it gcc complains for the undefined reference to main. Secondly, I have no idea how would putting all those modules together look like.
I would appreciate any help, but please try to keep your answers simple. Makefiles are still a bit of black magic to me.

Before you can do anything with a makefile, you must know how to do it without a makefile.
Since you are using gcc, I will assume that your source code is C++.
You haven't told us what your directory structure looks like, so I'll suppose that you have three source files in two directories: primary/main.cc, other/foo.cc and other/bar.cc. (We can deal with header files like foo.h later.) And you want to build myApp.
STEP 1: Doing It By Hand
To do this in one command, you might use:
gcc -Wall primary/main.cc other/foo.cc other/bar.cc -o myApp
This will compile the three source files and link the binary objects together into the executable myApp.
STEP 2: Doing It In Pieces (Do not attempt this until you can get the previous step to work perfectly.)
Instead of building with one command, you could take an intermediate step, compiling the source files into binary object files:
gcc -Wall -c primary/main.cc -o primary/main.o
gcc -Wall -c other/foo.cc -o other/foo.o
gcc -Wall -c other/bar.cc -o other/bar.o
This will produce alpha/main.o, beta/foo.o and beta/bar.o. The compiler won't complain about foo and bar lacking a main() function, because an object file doesn't need one. Then link the objects together into an executable:
gcc -Wall primary/main.o other/foo.o other/bar.o -o myApp
STEP 3: Doing It Locally (Do not attempt this until you can get the previous step to work perfectly.)
Just like the previous step, but we act in primary/ and other/:
cd primary
gcc -Wall -c main.cc -o main.o
cd ../other
gcc -Wall -c foo.cc -o foo.o
gcc -Wall -c bar.cc -o bar.o
cd ..
gcc -Wall primary/main.o other/foo.o other/bar.o -o myApp
STEP 4: Using a Makefile (Do not attempt this until you can get the previous step to work perfectly.)
We could have a makefile perform STEP 1, but that isn't really necessary. Write a makefile in primary (i.e. primary/makefile) like this:
main.o:
gcc -Wall -c main.cc -o main.o
(That whitespace in fromt of gcc... is a TAB.)
Now try this:
cd primary
make
cd ../other
gcc -Wall -c foo.cc -o foo.o
gcc -Wall -c bar.cc -o bar.o
cd ..
gcc -Wall primary/main.o other/foo.o other/bar.o -o myApp
STEP 5: Using Several Makefiles (Do not attempt this until you can get the previous step to work perfectly.)
Write a other/makefile:
both: foo.o bar.o
foo.o:
gcc -Wall -c foo.cc -o foo.o
bar.o:
gcc -Wall -c bar.cc -o bar.o
and a makefile in the top directory, where you're building myApp:
myApp:
gcc -Wall primary/main.o other/foo.o other/bar.o -o myApp
Now try this:
cd primary
make
cd ../other
make
cd ..
make
STEP 6: Using One Makefile That Calls Others (Do not attempt this until you can get the previous step to work perfectly.)
Edit the top makefile:
myApp:
cd primary; make
cd other; make
gcc -Wall primary/main.o other/foo.o other/bar.o -o myApp
Now try:
make
If all of this works, what you have is a crude but effective makefile system. There are many refinements possible, when you're ready to take the training wheels off.
EDIT:
If there are many source files in a subdirectory (e.g. other/) and you don't want to maintain a list in the top makefile by hand, there are several ways to handle it. This is one:
OTHER_SOURCES := $(wildcard other/*.cc)
OTHER_OBJECTS := $(OTHER_SOURCES:.cc=.o)
myApp:
cd primary; make
cd other; make
gcc -Wall primary/main.o $(OTHER_OBJECTS) -o myApp
But you should get these makefiles working and understand them, before you try any more streamlining.

Related

Im trying to compile program on Ubuntu and dont understand some things

Im a Windows dev who has no expirience on building C/C++ programs on Linux, but now I need to. Right way would be to go and learn Make and g++ compiler, but before I commit to that I want to figure out some basic stuff.
So I have .c program which is compiled with this makefile:
CUDA_VER=11.5
ifeq ($(CUDA_VER),)
$(error "CUDA_VER is not set")
endif
APP:= deepstream-test3-app
TARGET_DEVICE = $(shell gcc -dumpmachine | cut -f1 -d -)
NVDS_VERSION:=6.0
LIB_INSTALL_DIR?=/opt/nvidia/deepstream/deepstream-$(NVDS_VERSION)/lib/
APP_INSTALL_DIR?=/opt/nvidia/deepstream/deepstream-$(NVDS_VERSION)/bin/
ifeq ($(TARGET_DEVICE),aarch64)
CFLAGS:= -DPLATFORM_TEGRA
endif
SRCS:= $(wildcard *.c)
$(info info is $(SRCS))
INCS:= $(wildcard *.h)
PKGS:= gstreamer-1.0
OBJS:= $(SRCS:.c=.o)
CFLAGS+= -I../../../includes \
-I /usr/local/cuda-$(CUDA_VER)/include
CFLAGS+= $(shell pkg-config --cflags $(PKGS))
LIBS:= $(shell pkg-config --libs $(PKGS))
LIBS+= -L/usr/local/cuda-$(CUDA_VER)/lib64/ -lcudart -lnvdsgst_helper -lm \
-L$(LIB_INSTALL_DIR) -lnvdsgst_meta -lnvds_meta \
-lcuda -Wl,-rpath,$(LIB_INSTALL_DIR)
$(info info is $(CFLAGS))
all: $(APP)
%.o: %.c $(INCS) Makefile
gcc -c -o $# $(CFLAGS) $<
$(APP): $(OBJS) Makefile
gcc -o $(APP) $(OBJS) $(LIBS)
install: $(APP)
cp -rv $(APP) $(APP_INSTALL_DIR)
clean:
rm -rf $(OBJS) $(APP)
First thing I tried is to change this Makefile to compile it as C++ program. I changed .c file into .cpp, in makefile I change gcc to g++ everywhere and .c to .cpp everywhere. It gave me error that it couldnt find "main" entry point.
I gave up on that pretty fast and decided just to use lines output of original makefile, ending up with this:
g++ -c -o deepstream_test3_app.o -I../../../includes -I /usr/local/cuda-11.5/include -pthread -I/usr/include/gstreamer-1.0 -I/usr/include/glib-2.0 -I/usr/lib/x86_64-linux-gnu/glib-2.0/include ./deepstream_test3_app.cpp
g++ -o deepstream-test3-app deepstream_test3_app.o -lgstreamer-1.0 -lgobject-2.0 -lglib-2.0 -L/usr/local/cuda-11.5/lib64/ -lcudart -lnvdsgst_helper -lm -L/opt/nvidia/deepstream/deepstream-6.0/lib/ -lnvdsgst_meta -lnvds_meta -lcuda -Wl,-rpath,/opt/nvidia/deepstream/deepstream-6.0/lib/
First question, can I combine this 2 launches of g++ into one?
Second, when I make changes to "./deepstream_test3_app.cpp" they are not noticed by compiler. I added
#include <iostream>
...
std::cout << "hello!" << std::endl;
and they are ignored. Its like g++ gets as input some other copy/older version of the file and I dont understand how to go about it.
Hope for any help, sorry if it's all sounds stupid.
Ignoring for the moment the issues surrounding compiling C code with a C++ compiler,
g++ -c -o deepstream_test3_app.o -I../../../includes -I /usr/local/cuda-11.5/include -pthread -I/usr/include/gstreamer-1.0 -I/usr/include/glib-2.0 -I/usr/lib/x86_64-linux-gnu/glib-2.0/include ./deepstream_test3_app.cpp
g++ -o deepstream-test3-app deepstream_test3_app.o -lgstreamer-1.0 -lgobject-2.0 -lglib-2.0 -L/usr/local/cuda-11.5/lib64/ -lcudart -lnvdsgst_helper -lm -L/opt/nvidia/deepstream/deepstream-6.0/lib/ -lnvdsgst_meta -lnvds_meta -lcuda -Wl,-rpath,/opt/nvidia/deepstream/deepstream-6.0/lib/
First question, can I combine this 2 launches of g++ into one?
Yes. It is a common practice in makefiles to separate the compilation and linking steps, but that is not mandatory. When there are multiple sources, the separation makes it possible to limit recompilations to only the source files that have changed, but it doesn't make much difference, makefile or not, when there is only one source file.
The one-command version would be mostly a concatenation of the two commands you gave. One would omit the -c option, which instructs g++ to compile but not link, and one would omit the -o deepstream_test3_app.o, which specifies the name of the object file that we are no longer going to create. One would also omit the appearance of deepstream_test3_app.o drawn from the link (second) command, as we are going straight from source file to program. The rest of the options can be reordered to some extent, but all the -l options need to remain in the same order relative to each other and to any object files among the inputs. Here is how I would write it:
g++ -c -o deepstream_test3_app -I../../../includes -I /usr/local/cuda-11.5/include -pthread -I/usr/include/gstreamer-1.0 -I/usr/include/glib-2.0 -I/usr/lib/x86_64-linux-gnu/glib-2.0/include -Wl,-rpath,/opt/nvidia/deepstream/deepstream-6.0/lib/ ./deepstream_test3_app.cpp -lgstreamer-1.0 -lgobject-2.0 -lglib-2.0 -L/usr/local/cuda-11.5/lib64/ -lcudart -lnvdsgst_helper -lm -L/opt/nvidia/deepstream/deepstream-6.0/lib/ -lnvdsgst_meta -lnvds_meta -lcuda
Second, when I make changes to "./deepstream_test3_app.cpp" they are not noticed by compiler.
The compiler compiles the source file(s) you tell it to.
Its like g++ gets as input some other copy/older version of the file
It is possible that you are indeed telling it to compile a different version than the one you modified. It is also possible that compilation fails, so you don't get a new executable. And it is possible that when you try to run the result, you are not running the program you think you are running. We don't have enough information to know.
With regard to the last, however, do be aware that on Linux, unlike on Windows, the working directory is not automatically in the executable search path. If you want to run the compiled result from the above command, you would want to specify the path to it, which you could most easily do by prepending ./ to its simple name: ./deepstream-test3-app.

'make: nothing to be done for p1 and' no rule to make target 'clean''

I'm trying to create a makefile titled 'p1' for a project.
When I try the command make p1 it returns with make: nothing to be done for p1
Also, when I try the command make p1 clean it returns no rule to make p1 'clean.' Stop
Here is my makefile:
a.out: main.o P1LinkedList.o const_iterator.o iterator.o Node.o
g++ -std=c++11 main.o const_iterator.o iterator.o Node.o
main.o:
g++ -std=c++11 -c main.cpp
P1LinkedList.o:
g++ -std=c++11 -c P1LinkedList.cpp
iterator.o:
g++ -std=c++11 -c iterator.cpp
const_iterator.o:
g++ -std=c++11 -c const_iterator.cpp
Node.o:
g++ -std=c++11 -c Node.cpp
depend:
g++ -MM main.cpp > p1.dep
clean:
rm -f a.out *.o
What do I need to fix to have the makefile compile .o files from my .cpp files and how do I fix the issue with the clean command?
Edit:
Here are the commands I've used to compile manually:
Helens-Air:p1a helenade$ g++ -std=c++11 *.cpp
Helens-Air:p1a helenade$ ./a.out
^^ and this just continues with the program execution from there
We may have to take this in stages.
First, you seem to misunderstand the difference between a makefile name and a target name. This appears to have been a miscommunication between you and your teacher, but it's easy to clear up.
Suppose you have a makefile named "Makefile", containing the following:
foo:
#echo running the foo rule
bar:
#echo running the bar rule
If you make foo, you will get:
running the foo rule
The argument (foo) tells Make which target to attempt to build. And how did Make know which makefile to use? (After all, you could have a dozen makefiles in the working directory.) You can specify which makefile to use, but if you don't then by default Make will look for a makefile named Makefile (or makefile or GNUmakefile, don't worry about this for now). To specify a makefile with another name, like "Buildfile", you can use the -f flag:
make -f Buildfile
So "p1" ought to have been the name of a target, not a makefile. Within the makefile, rename your a.out rule to p1. Then rename the whole makefile to Makefile. Then
make p1
should work (or at least run).
Edit:
I'll go out on a limb. In the a.out rule (which should now be called the p1 rule), I notice that you have left P1LinkedList.o out of the list of object files to be linked. So try changing it:
p1: main.o P1LinkedList.o const_iterator.o iterator.o Node.o
g++ -std=c++11 main.o P1LinkedList.o const_iterator.o iterator.o Node.o
If that works, you can simplify it with an automatic variable:
p1: main.o P1LinkedList.o const_iterator.o iterator.o Node.o
g++ -std=c++11 $^
And there will be other small improvements you can make.
If it doesn't work, try ls *.cpp and see if you've overlooked some other source file.

Simple makefile for MPI-based C program

I am writing my first MPI-based C program. I have four files namely, MPI_Program.c and Helper.c, and their corresponding .h files. MPI_Program.c includes mpi.h and Helper.h.
I wrote the following simple makefile to compile MPI_Program but I get the error "No rule to make target mpi.h"
Here's the makefile:
# Rules to produce the target
all: MPI_Program
MPI_Program: MPI_Program.o Helper.o
mpixlc MPI_Program.o Helper.o -f machineFile -std=c99 -g -o MPI_Program
# Rules to produce the object files
Helper.o: Helper.c Helper.h
gcc -c Helper.c -std=c99 -g
MPI_Program.o: MPI_Program.c MPI_Program.h Helper.h mpi.h
mpixlc -c MPI_Program.c -std=c99 -g -f machineFile
I am certainly missing something...Can someone please advise.

Makefile: Why can't I see compilation output?

I have a master Makefile that calls submakefiles inside directories.
Those submakefiles are generated by autotools starting from Makefile.am and configure.ac files.
The first target of the entire project is a shared library, while the others are executables that link against it (so there is a dependency in the main Makefile).
The problem is that I can only see the compilation output of the shared library, while the other projects are just printing warnings and errors (just stderr text, I believe).
I can't understand what is causing this behaviour since all the projects are called from the very same rule in the master Makefile and all the submakefiles are autogenerated from very seemed configuration files.
The snippet of master Makefile that compiles all the targets looks like this:
SUBDIRS = libMylib app1 app2 app3
$(SUBDIRS):
$(ECHO) "-> Building $#"
$(MAKE) -C $#
$(ECHO) "-> Build of $# finished."
The output of "libMylib" is like:
-> Building libMylib
libtool: compile: ppc-linux-gcc -DHAVE_CONFIG_H -I. -I/opt/ELDK/ppc_8xx/usr/include/ -I/opt/ELDK/ppc_8xx/include/ -Wall -std=gnu99 -O2 -MT libMylibF1.lo -MD -MP -MF .deps/libMylibF1.Tpo -c libMylibF1.c -fPIC -DPIC -o .libs/libMylibF1.o
...
libtool: link: ppc-linux-gcc -shared -fPIC -DPIC .libs/libMylibF1.o .libs/libMylibF2.o .libs/libMylibF3.o -Wl,-rpath -Wl,/opt/ELDK/ppc_8xx/lib -Wl,-rpath -Wl,/opt/ELDK/ppc_8xx/lib -lz -lpthread -O2 -Wl,-soname -Wl,libMylib.so.0 -o .libs/libMylib.so.0.0.0
libtool: link: (cd ".libs" && rm -f "libMylib.so.0" && ln -s "libMylib.so.0.0.0" "libawmg.so.0")
libtool: link: (cd ".libs" && rm -f "libMylib.so" && ln -s "libMylib.so.0.0.0" "libMylib.so")
libtool: link: ( cd ".libs" && rm -f "libMylib.la" && ln -s "../libMylib.la" "libMylib.la" )
-> Build of libAwmg finished.
While the output of any "appN" is:
-> Building app1
app1F1.c: In function `app1F1Func1':
app1F1.c:161: warning: unused variable `varA'
app1F2.c:85: warning: `app1F2FuncX' defined but not used
-> Build of app1 finished.
Could anyone please help me?
EDIT:
I found out that I was able to see the compile stuff in "libMylib" because it is "libtoolized".
If I "libtoolize" another project, I can see the linker part of the output (libtool: link:...).
So, considering that make calls a shell (sh) for each line in the rule, that $(MAKE) equals to just "make" and, by default, make is verbose, why do the submake is not printing its output to stdout?
Which flags may be passed to it so that this happens?
It depends on the Makefiles in the sub directories. If, for example, you have a # in front of the compile commands in the app Makefiles, then they will not output the commands that were being run (also look for $(Q), which is used quite commonly when you want to enable/disable verboseness in makefiles). Alternatively, your make commands in the apps directories may be pushing stdout to some files (which would still allow you to see the warnings).
By the same token your library makefile may be incorrectly redirecting stderr elsewhere, which would explain why you don't see any error outputs on that (or it just doens't have any errors...)
Please post the build rules in the other directories (and all associated variable definitions) if you want more help.
Well, my master makefile includes another file that puts a .SILENT rule in.
I haven't noticed that until today.
Removing this rule makes everithing work as expected.
Oh my...

Compile C++ with Cygwin

How do I compile my C++ programs in Cygwin. I have gcc installed. What command should I use? Also, how do I run my console application when it is in a .cpp extension. I am trying to learn C++ with some little programs, but in Visual C++, I don't want to have to create a seperate project for each little .cpp file.
You need to use a command like:
g++ -o prog prog.cpp
That's a simple form that will turn a one-file C++ project into an executable. If you have multiple C++ files, you can do:
g++ -o prog prog.cpp part2.cpp part3.cpp
but eventually, you'll want to introduce makefiles for convenience so that you only have to compile the bits that have changed. Then you'll end up with a Makefile like:
prog: prog.o part2.o part3.o
g++ -o prog prog.o part2.o part3.o
prog.o: prog.cpp
g++ -c -o prog.o prog.cpp
part2.o: part2.cpp
g++ -c -o part2.o part2.cpp
part3.o: part3.cpp
g++ -c -o part3.o part3.cpp
And then, you'll start figuring how to write your makefiles to make them more flexible (such as not needing a separate rule for each C++ file), but that can be left for another question.
Regarding having a separate project for each C++ file, that's not necessary at all. If you've got them all in one directory and there's a simple mapping of C++ files to executable files, you can use the following makefile:
SRCS=$(wildcard *.cpp)
EXES=$(SRCS:.cpp=.exe)
all: $(EXES)
%.exe: %.cpp
g++ -o $# $^
Then run the make command and it will (intelligently) create all your executables. $# is the target and $^ is the list of pre-requisites.
And, if you have more complicated rules, just tack them down at the bottom. Specific rules will be chosen in preference to the pattern rules:
SRCS=$(wildcard *.cpp)
EXES=$(SRCS:.cpp=.exe)
all: $(EXES)
%.exe: %.cpp
g++ -o $# $^
xx.exe: xx.cpp xx2.cpp xx3.cpp
g++ -o $# $^
echo Made with special rule.
You will need g++. Then try g++ file.cpp -o file.exe as a start. Later you can avoid much typing by learning about Makefiles.
if you want to use cygwin you should use the normal gcc syntax
g++ -o foobar foobar.cpp
but that doesn't really play well with Visual C++. I advise you to take a look into Eclipse CDT if you prefer using GCC over the visual C++ compiler.
What I do to compile a cpp program:
g++ -Wall Test.cpp -o Test
-Wall enables warning and error messages to be shown
-o Test creates an Test.exe after compilation
If you want to compile files separately:
g++ -Wall -c File1.cpp
g++ -Wall -c File2.cpp
Now create an executable with the combined object files as:
g++ -Wall File1.o File2.o -o File.exe
This way you can compile your header files and you can include in your application programs.

Resources