Re: Improving build system

Il 15/05/2025 11:03, David Brown ha scritto:

On 14/05/2025 23:51, pozz wrote:

Il 14/05/2025 11:03, David Brown ha scritto:

On 13/05/2025 17:57, pozz wrote:

[...]

>
You are asking a lot of questions here.  They are good questions, but it would be a very long post if I tried to answer them all fully.  So instead, I will try to give a few hints and suggestions that you can take further.  I'll put numbers on them in case you want to reference them in replies.

>
Ok, thank you very much for your time.
>
>

1.
>
Windows path naming is insane.  Fortunately, you can almost always override it.  Whenever you install any serious program in Windows, especially if you ever want to refer to it from the command line, batch files, makefiles, etc., avoid names with spaces or "awkward" characters.   I recommend making top-level directories like "progs" or "compilers" and putting the tools in there as appropriate.  This also makes it vastly easier to copy tools to other machines.  And since you should never upgrade your toolchains - merely add new versions to your collection, in separate directories - it is easier if they are better organised.

>
I know, but not all software installers work well if you change their default installation path.  When it comes to stupid and big IDEs (such as Atmel/Microchip Studio), I prefer to avoid changing the default installation path (C:\Program Files (x86)\Atmel...) to avoid other obscure issues.

 Almost all of them work fine in different directories (and indeed different drives).  First assume they work - only fall back on the crappy defaults if there is no other option.

Yes, you are right.  However I don't think it would be simple to understand if a problem is caused by the installation path or other obscure reason.  I'm talking about an installation of an IDE (Atmel Studio) that takes a very long time to finish.  The installation process is a nightmare.

Of course, the best answer is to avoid any tool made by Microchip - they are the worst of the bunch.  (Atmel was always a bit behind in second place for the title of worst toolchain supplier, but Microchip has gradually integrated them.)  I have many fine things to say about Microchip and Atmel as hardware suppliers, but I make a point of avoiding their microcontrollers because of their tools.

:-)
I worked on PIC8 and AVR8 and IMHO AVR8 is much better then PIC8.
Regarding Cortex-M, SAM devices are fine for me.

My strong preference - regardless of the manufacturer - is to use the ARM-supplied gcc toolchains (for ARM microcontrollers, obviously) rather than the usually older tools supplied by manufacturers.

It is something I learned.

It is already a miracle if that software runs without problems with the default installation path.  I don't want to imagine what happens if I changed it.
>
Anyway until now I didn't find issues with spaces.  Even in msys2 shell I can use "/c/Program\ Files\ (x86)/...".

 It usually works - but that does not stop it being a PITA and an insane choice of pathnames.

Sure.

Still, you have to find what works best for you - I am giving recommendations and suggestions, not a unique solution or single "correct" answer.
 

>
The other IDE I use is MCUXpresso.  It is Eclipse based so I installed it in c:\ without any temptations.
>

 Yes, that has always worked for me.
 (Of course I normally have it on Linux, rather than Windows, and most manufacturer-supplied software uses a sensible default path - in /opt or in /usr/local.)
 

>

2.
>
You don't need to use bash or other *nix shells for makefile or other tools if you don't want to.  When I do builds on Windows, I run "make" from a normal command line (or from an editor / IDE).  It is helpful to have msys2's usr/bin on your path so that make can use *nix command-line utilities like cp, mv, sed, etc.  But if you want to make a minimal build system, you don't need a full msys2 installation - you only need the utilities you want to use, and they can be copied directly (unlike with Cygwin or WSL).
>
Of course you /can/ use fuller shells if you want.  But don't make your makefiles depend on that, as it will be harder to use them from IDEs, editors, or any other automation.

>
In the beginning (some years ago) I started installing GNU Make for Windows, putting it in c:\tools\make.  Then I created a simple Makefile and tried to process it on a standard Windows command line.  It was a mess!  I remember there were many issues regarding: slash/backslash on file paths, lack of Unix commands (rm, mv, ...) and so on.  Native Windows tools need backslash in the paths, but some unix tools need slash.  It was a mess to transform the paths between the two forms.
>

 Most tools on Windows are happy with forward slash for path separators as well. 

mkdir, just to name one?  And you need mkdir in a Makefile.

Certainly everything that is originally a *nix tool will be fine with that.
 Of course if you have a makefile that uses commands like "rm" and you don't have them on your path, and don't specify the path in the makefile, then it won't work.  This is why the norm in advanced makefiles is to use macros for these things :
 # Put this in the host-specific file, with blank for no path needed
bin_path :=
 # Use this instead of "rm".
RM := $(bin_path) rm

Initially I insisted using native Windows commands: DEL, MKDIR, COPY and so on.  Finally I gave up.

After this attempt, I gave up.  I thought it was much better to use the IDE and build system suggested by the MCU manufacturer.
>

 For most IDEs, the build system is "make".  But the IDE generates the makefiles - slowly for big projects, and usually overly simplistic with far too limited options.
 But IDE's are certainly much easier for getting started.  On new projects, or new devices, I will often use the IDE to get going and then move it over to an independent makefile.  (And I'll often continue to use the IDE after that as a solid editor and debugger - IDE's are generally happy with external makefiles.)

I'm going to create a new post regarding editors and debugger... stay tuned :-D

Now I'm trying a Unix shell in Windows (msys, WSL or even the bash installed with git) and it seems many issues I had are disappearing.
>
>

And of course you will want an msys2/mingw64 (/not/ old mingw32) for native gcc compilation.

>
The goal of the simulator is to detect problems on the software that runs directly on Windows, without flashing, debug probes and so on.  I increased my productivity a lot when I started this approach.
>
Obviously, the software running on Windows (the simulator) should be very similar to the sofware running on the embedded target.  Cortex-M MCUs are 32-bits so I thought it should be better to use a 32-bits compiler even for the simulator.
>

 mingw-w64 can happily generate 32-bit Windows executables.  IIRC you just use the "-m32" flag.  It is significantly better than old mingw in a number of ways - in particular it has vastly better standard C library support.

Why doesn't it work for me?  I open a Msys2/mingw64 shell and...
$ gcc -m32 -o main.exe main.c
C:/msys64/mingw64/bin/../lib/gcc/x86_64-w64-mingw32/11.2.0/../../../../x86_64-w64-mingw32/bin/ld.exe: skipping incompatible C:/msys64/mingw64/bin/../lib/gcc/x86_64-w64-mingw32/11.2.0/../../../../x86_64-w64-mingw32/lib/libmingw32.a when searching for -lmingw32
C:/msys64/mingw64/bin/../lib/gcc/x86_64-w64-mingw32/11.2.0/../../../../x86_64-w64-mingw32/bin/ld.exe: skipping incompatible C:/msys64/mingw64/bin/../lib/gcc/x86_64-w64-mingw32/11.2.0/../../../../x86_64-w64-mingw32/lib\libmingw32.a when searching for -lmingw32
...
... and much more

Moreover, I think many issues aries on a 64-bits compilation, for example static allocated buffers that would be too small on a 64-bits platforms.  Or some issues on serializers.
>
>

Don't bother with WSL unless you actually need a fuller Linux system - and if you /do/ need that, dump the wasteful crap that is Windows and use Linux for your development.  Build speeds will double on the same hardware.  (In my testing, done a good while back, I did some comparisons of a larger build on different setups on the same PC, using native Windows build as the baseline.  Native Linux builds were twice the speed.  Running VirtualBox on Windows host, with a Linux virtual machine, or running VirtualBox on Linux with a Windows virtual machine, both beat native Windows solidly.)

>
I completely agree with you.  At the moment msys2 seems ok.
>
>

3.
>
Makefiles can be split up.  Use "include" - and remember that you can do so using macros.  In my makefile setups, I have a file "host.mk" that is used to identify the build host, then pull in a file that is specific to the host:
>
# This is is for identifying host computer to get the paths right
>
ifeq ($(OS),Windows_NT)
   # We are on a Windows machine
   host_os := windows
   host := $(COMPUTERNAME)
else
   # Linux machine
   host_os := linux
   host := $(shell hostname)
endif
>
ifeq "$(call file-exists,makes/host_$(host).mk)" "1"
   include makes/host_$(host).mk
else
   $(error No host makefile host_$(host).mk found)
endif
>
Then I have files like "host_xxx.mk" for a computer named "xxx", containing things like :
>
toolchain_path := /opt/gcc-arm-none-eabi-10-2020-q4-major/bin/
>
or
>
toolchain_path := c:/micros/gcc-arm-none-eabi-10_2020-q4-major/bin/
>
>
All paths to compilers and other build-related programs are specified in these files.  The only things that are taken from the OS path are standard and common programs that do not affect the resulting binary files.

>
It is an interesting and uncommon (at least for me) approach.
>
What happens if multiple developers work on the same repository?  Are they forced to create a host_xxx.mk for all their development machines? Should the host_xxx.mk files be added to the repository?

 Yes, that is /exactly/ what you do.  It also applies to a single developer using multiple different machines.  For any long-term project, you want to be sure you can check out the repository, do a clean build, and get an identical binary from more than one machine.  Having individual "host_XXX.mk" files means that the build adapts automatically to the machine.  What you don't want is each developer making changes to a single makefile so that it works on their machine - and then either not checking in the changes, or checking them in and messing things up for someone else.
 

>
I guess the only goal of host_xxx.mk is to avoid changing PATH before make.  Why don't you like setting the PATH according to the project you're working on?
>

 No, that is not the only goal - there can be many differences between machines.  For example, I usually have ccache on my Linux systems but it is rare to have it on (native) Windows systems - thus that can be enabled or disabled in a host_xxx.mk file.  Some machines might also support building the documentation, or running a simulator, or signing binaries.
 Setting the path would be an extra complication of no benefit, but a significant source of risk or error.  How do you make sure your IDE is using the right PATH settings before it runs "make"?  How do you deal with multiple projects - do you keep swapping PATHs?  (I usually have a half-dozen projects "open" at a time, in different workspaces on my Linux machine.)  Do you now have a makefile and a separate path-setting batch file or shell script that you need to run before doing a project build?  How do you handle things when you install some new Windows program that messes with your path?
 It is /vastly/ simpler and safer to put the paths to the binaries in a couple of macros in your makefile(s).  It also gives clear and unequivocal documentation of the tools you need -  if your makefile has this line :
 toolchain_path := c:/micros/gcc-arm-none-eabi-10_2020-q4-major/bin/
 then there is never any doubt as to exactly which toolchain is used for the project.

I see your points.  The only drawback seems putting a bunch of host_xxx.mk files in the repository.  If the developer team and their development machines are well defined and static, everything goes well.
However what happens when a new developer pulls your repository and want to build?  At first, he must create his host_xxx.mk and starting polluting the original repository.  Instead, by using the PATH, it could build without touching any files in the repo.
Maybe this isn't our situation, but a public open-source repository can't use your approach.  It's impossible to include in the public repository tenths or hundreds host_xxx.mk.
Moreover, what happens if two developers like astronomy and set the hostname of their development machine JUPITER?  Maybe one uses Linux, the other Windows.
In your make, it seems you include the correct host_xxx.mk file automatically from the hostname.

Then I have a "commands.mk" file with things like :
>
ATDEP := @
>
toolchain_prefix := arm-none-eabi-
>
CCDEP := $(ATDEP)$(toolchain_path)$(toolchain_prefix)gcc
CC := $(AT)$(CCACHE) $(toolchain_path)$(toolchain_prefix)gcc
LD := $(AT)$(toolchain_path)$(toolchain_prefix)gcc
OBJCOPY := $(AT)$(CCACHE) $(toolchain_path)$(toolchain_prefix)objcopy
OBJDUMP := $(AT)$(CCACHE) $(toolchain_path)$(toolchain_prefix)dump
SIZE := $(AT)$(CCACHE) $(toolchain_path)$(toolchain_prefix)size
>
>
Put CONFIG dependent stuff in "config_full.mk" and similar files.  Put TARGET specific stuff in "target_simulator.mk".  And so on.  It makes it much easier to keep track of things, and you only need a few high-level "ifeq".
>
>
Keep your various makefiles in a separate directory.  Your project makefile is then clear and simple - much of it will be comments about usage (parameters like CONFIG).

>
Yes, splitting makefiles is a good suggestion.
>
>

4.
>
Generate dependency files, using the same compiler and the same include flags and -D flags as you have for the normal compilation, but with flags like -MM -MP -MT and -MF to make .d dependency files. Include them all in the makefile, using "-include" so that your makefile does not stop before they are generated.

>
I have to admit that ChatGPT helped me to create the Makefile.  The CFLAGS include -MMD and -MP and at the end I have
>
   -include $(DEP_FILES)
>
Of course, DEP_FILES are:
>
   DEP_FILES := $(OBJ_FILES:.o=.d)

 That's a good start.  There are quite a few articles and blog posts about automatic generation of makefile dependencies that can be worth reading.
 

Sincerely I don't know if it is good, but I tried to change an include file and related C files are compiled again as expected (so I think the dependency are correctly managed).
>
There's a thing that doesn't work.  If I change the Makefile itsel, for example changing CFLAGS adding a new compiler option, I need to manually invoke a clean.

 depfiles_src := $(cfiles:.c=.d) $(cppfiles:.cpp=.d)
depfiles := $(addprefix $(dep_dir),$(patsubst ../%,%,$(depfiles_src)))
 -include $(depfiles)
 alldepends := makefile $(wildcard makes/*.mk)
all : $(alldepends) $(depfiles)
depends : $(alldepends)
 # "depends" target just makes dep files
depends : $(depfiles)
         @echo Updated dependencies
  Vary according to your needs.  But basically, if something has $(alldepends) in its dependency list, it will be rebuild if one of your makefiles changes.
 

>

5.
>
Keep your build directories neat, separate from all source directories, and mirroring the tree structure of the source files.  So if you have a file "src/gui/main_window.c", and you are building with CONFIG=FULL TARGET=embedded, the object file generated should go in something akin to "builds/FULL/embedded/obj/src/gui/main_window.o".  I like to have separate parts for obj (.o files), dep (.d files), and bin (linked binaries, map files, etc.).  You could also mix .d and .o files in the same directory if you prefer.
>
This means you can happily do incremental builds for all your configurations and targets, and don't risk mixing object files from different setups.

>
Yes, perfectly agreed.
>
>

6.
>
Learn to use submakes.  When you use plain "make" (or, more realistically, "make -j") to build multiple configurations, have each configuration spawned off in a separate submake.  Then you don't need to track multiple copies of your "TARGET" macro in the same build - each submake has just one target, and one config.

>
I don't think I got the point.  Now I invoke the build of a single build configuration.  Are you talking about running make to build multiple configurations at the same time?

 Yes.
 Obviously it depends on the stage you are in development and the kind of project - much of the time, you will want to build just one configuration.  But sometimes you will also want to make multiple builds to check that a small change has not caused trouble elsewhere, or for different kinds of testing?  Why run multiple "make" commands when you can do a full project build from one "make" ?

Are you thinking something similar to:
all_configs:
$(MAKE) -j 4 CONFIG=FULL
$(MAKE) -j 4 CONFIG=STANDARD
$(MAKE) -j 4 CONFIG=LITE
With my actual Makefile, "make all_configs" returns an errore because CONFIG is not specified.