Miro Kropacek wrote:
But this fact doesn't change anything on my hint to remove pml from compilation chain -- gcc is OK with its built-in stuff from mpfr (try to list libgcc.a with 'nm' -- a lot of math functions here), it needs libm / pml for nothing (tested -- I had no math lib installed during compilation).
Ok, I didn't understand... You said that no additional libm is required in order to build gcc itself. So the libm (PML or other) can be considered like a standard library, and built after GCC, outside of the compiler build chain.
True... but it isn't such nice after all ;-) I don't get that idea but it compiles all that m68020-60 etc stuff and then -- don't use it! When I ask gcc for multilib support without anything in t-mint, it says just 68000... why is m68020-60 etc compiled is really mystery for me...
Patrice Mandin has worked on that multilib stuff. Thanks to his contribution, I fixed the regression about bad directory output for 68000 libraries. Patrice also made a patch to enable 68020 compilation without using the FPU. I haven't included it yet: we have to discuss here of what set of multilibs need to be provided with our GCC distribution.
OK then, it's on my web in download section.
Nice ! Link added. http://vincent.riviere.free.fr/soft/m68k-atari-mint/
Hehe, it wasn't such hard... but without cross compiler, this would be really 'funny' :)
So you build a native GCC 4.3.2 using a cross GCC 4.3.2. If someone would build a native GCC 4.3.2 using a native GCC 2.95, the resulting compiler should be strictly equivalent (producing the very same code). If someone has a week to spend in order to check this assertion ;-)
By the way, I've made also something I call "gnu as democoder edition",
That's a good idea ! I read somewhere that gas was primarily made to assemble GCC output, not to be used by humans. However, it has been improved over the time. If I remember well, the assembly routines in the Linux kernel doesn't use gas: if I remember well, it uses nasm instead.
-- Vincent Rivière