PropGCC for P2

1235»

Comments

  • Wouldn't it be easier to start with a version of PropGCC that already supports the P1? This would have all the necessary mods to handle the PASM quirks, and it also has some modifications for the P2.
  • gcc 4?

    Mike
  • msrobots wrote: »
    gcc 4?

    Mike
    Eric has already gotten GCC 6 working for the Propeller. It was just never adopted by Parallax.

  • @Dave Hein Aha! I looked back at the P1 modified binutils and found the magic line to add chars to the symbol list. Back to '#' as usual :-)

    I certainly started with the gcc4 port details. Perhaps it would have been better to try to move up one major version at a time, but I made the calculated choice to jump straight to present.
    extern const char propeller_symbol_chars[];
    #define tc_symbol_chars propeller_symbol_chars
    
  • As a personal preference, I would much rather see this port end up in upstream GCC/Binutils. The original P1 port used non-GPL code in a few places. I don't believe such code would be accepted. This is primarily why I started from the mainline branch.
  • ntosme2 wrote: »
    As a personal preference, I would much rather see this port end up in upstream GCC/Binutils. The original P1 port used non-GPL code in a few places. I don't believe such code would be accepted. This is primarily why I started from the mainline branch.
    You mean you can't upstream code to GCC/binutils if it has a less restrictive license than GPL? Ugh.

  • Hmm, it seems I'm incorrect. IANAL, but it appears MIT code may be included in GPL code and released under the GPL. The copyright notices just have to remain in the MIT-licensed files. What doesn't work is releasing GPL+MIT as MIT.

    The support libraries could also live separately and be pulled in as part of building the tools.
  • forget the license aspect for a while, but @ntosme2 THANK YOU for doing this.

    I do not write compilers, I am just using them, so I admire the work you put in there and want just to cheer you up a bit!

    Enjoy!

    Mike
  • ntosme2 wrote: »
    To get further with this I'd need some help from someone experienced with gcc's instruction constraint mapping.
    In this crowd that would probably be Eric Smith. I've done a bit of tweaking of the propgcc compiler but the port was done by Eric.

Sign In or Register to comment.