[M3devel] Stuck on adding new cm3 target (AMD64_DRAGONFLY)
jay.krell at cornell.edu
Sun Jan 19 13:32:44 CET 2014
upgrade.sh (note the .sh) rebuilds everything
I wrote upgrade.py based on it, but missed the "everything" part.
upgrade.py only upgrades "the compiler" (and m3core, libm3..)
A better practise although inefficient is:
upgrade.py && ./do-cm3-all.py realclean skipgcc && ./do-cm3-all.py buildship
I didn't suggest that to you here in order to get further in the pipeline faster.
However upgrade.py w/o other rebuild could leave m3core.so too new for many programs.
But for purposes of building/cross-building just the compiler, my instructions are good.
John, I see you added cm3 to the FreeBSD ports and you clearly know what you are doing.
Nevertheless, I am a lazy snob, don't quite see what you are doing wrong, and am suggesting "just different", not clearly better.
But, to my credit, this "just different" is what I use all the time.
I have used it to cross to several systems, w/ and w/o cm3cg (w/ and w/o C backend).
It isn't fast, and it doesn't always have the best incrementality, but it does usually work.
There is redundancy. pylib.py essentially duplicates parts of the config files.
You misunderstood what I meant by "C backend" and from the ports history you will somewhat appreciate it.
The C backend, instead of having a lot to do with gcc, is a backend that generates C, and then invokes a C compiler.
The problem you had with -gstabs for example, goes away. We'll use whatever debugging format regular C/C++ use.
There is a little hope of taking the one C output and compiling it for different targets -- at least FreeBSD 8,9,10.
Currently the C is very word size dependent, endian dependent, and somewhat otherwise target-dependent.
I'd like to fix these aspects but it isn't easy by far, will take possibly controversial changes in the frontend and interface to backend.
From: jay.krell at cornell.edu
To: adacore at marino.st
Date: Sun, 19 Jan 2014 12:17:24 +0000
CC: m3devel at elegosoft.com
Subject: Re: [M3devel] Stuck on adding new cm3 target (AMD64_DRAGONFLY)
Hey..I like that that does capture some of My Way.
Here is my independent off the cuff write up though:
I know the ports system is nice.
I know we really need to be in it for many operating systems.
I suspect your problem is that you didn't "upgrade" to apply you changes at the right time.
Do any of these work for you:
Anyway, please try this:
Ignore DragonFly at first.
Go to a system with a working cm3. Such as FreeBSD. Linux. Almost anything.
Such as from ports.
Make a backup, e.g. of /usr/local/cm3.
My instructions are unfortunately destructive of the existing install.
Checkout the full current CVS repository.
In that repository, run scripts/python/upgrade.py.
If that doesn't work, stop, and we'll help.
If that does work, apply your diffs. And upgrade.py again. You can say "upgrade.py skipgcc" here.
Then boot1.py c amd64_dragonflybsd
Make sure you say "c" and the target "amd64_dragonflybsd", anywhere on the command line.
Ok, "c" is actually up to you. Since nobody messes with ABI much, gcc/amd64 works for the same
for Linux, FreeBSD, OpenBSD, NetBSD, and Dragonfly.
If this works, it will give you, roughly amd64_dragonflybsd*.tar.gz
Copy that to your Dragonfly system, extract it, cd into it, make.
If that works, you have a native cm3.
Run it. It should say "can't find cm3.cfg".
Now on Dragonfly, mkdir -p /usr/local/cm3/bin
cp cm3 /usr/local/cm3/bin
full CVS checkout
cd scripts/python ; ./boot2.sh
You will need to have made small changes to pylib.py for this to work.
(And your config file should say to use the C backend, like Darwin and ARM_LINUX do.)
> Date: Sun, 19 Jan 2014 13:09:37 +0100
> From: adacore at marino.st
> To: jay.krell at cornell.edu
> CC: m3devel at elegosoft.com
> Subject: Re: [M3devel] Stuck on adding new cm3 target (AMD64_DRAGONFLY)
> On 1/19/2014 12:40, Jay K wrote:
> > Your change looks about right.
> > Did you first "upgrade" and existing install, with these changes, and
> > then cross?
> > That is what you must do.
> It's probably easier that I show you.
> I've attached "Makefile.local" (I added the .txt extension just now to
> help out some mail clients) which is my recipe.
> So I run three targets in a freshly patched workarea:
> 1) make cross-compiler
> 2) make assembler-code (called from cross-archive)
> 3) make cross-archive
> In the cross-compiler target, the FreeBSD bootstrap compiler builds
> m3middle, m3linker, m3front, mequake, and CM3 for the cross compiler.
> The cross-compiler then builds libm3core and libm3 with FreeBSD cm3cg
> Everything is cleaned, then the cross compiler builds everything again
> (including libm3core and libm3) along with m3objfile, sysutils, and
> Then the cm3.cfg is changed to specify HOST=freebsd, TARGET=dragonfly,
> cm3cg is cm3cg-AMD64_DRAGONFLY.
> I would not be surprised is there is a logic flaw or two in "make
> cross-compiler". The "how to port CM3" tutorial was slightly out of
> date and some of the instructions were a bit ambiguous so this was a
> best guess.
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the M3devel