public inbox for crossgcc@sourceware.org
 help / color / mirror / Atom feed
From: Khem Raj <raj.khem@gmail.com>
To: "Franzi Edo." <edo.franzi@ukos.ch>
Cc: crossgcc@sourceware.org
Subject: Re: gcc-4.3.3 and coldfire
Date: Mon, 22 Jun 2009 21:41:00 -0000	[thread overview]
Message-ID: <19c1b8a90906221441l12b6d60fu1f39c28019287c04@mail.gmail.com> (raw)
In-Reply-To: <FE4757BE-FDD9-4DF1-9BE2-57E3BB62B5EC@ukos.ch>

On Sun, Jun 21, 2009 at 3:39 AM, Franzi Edo.<edo.franzi@ukos.ch> wrote:
> Hi All,
> I am trying to move from the gcc-4.2.4 to the gcc-4.3.3 version. I would
> like to generate compiler versions for cpu32 and for coldfire (m5234).
> My toolchain running with the 4.2.4 works fine for both cpus. The version
> 4.3.3 works fine only for cpu32, and I noticed that something is wrong with
> the library location. I had to change the file t-m68kelf in order to have a
> multilib and to have a fPIC libraries. Here is the t-m68kelf that works fine
> for the 4.2.4 version:
>
> LIB1ASMSRC = m68k/lb1sf68.asm
> LIB1ASMFUNCS = _mulsi3 _udivsi3 _divsi3 _umodsi3 _modsi3 \
>   _double _float _floatex \
>   _eqdf2 _nedf2 _gtdf2 _gedf2 _ltdf2 _ledf2 \
>   _eqsf2 _nesf2 _gtsf2 _gesf2 _ltsf2 _lesf2
>
> LIB2FUNCS_EXTRA = fpgnulib.c xfgnulib.c
>
> fpgnulib.c: $(srcdir)/config/m68k/fpgnulib.c
>        cp $(srcdir)/config/m68k/fpgnulib.c fpgnulib.c
> xfgnulib.c: $(srcdir)/config/m68k/fpgnulib.c
>        echo '#define EXTFLOAT' > xfgnulib.c
>        cat $(srcdir)/config/m68k/fpgnulib.c >> xfgnulib.c
>
> MULTILIB_OPTIONS = m528x/mcpu32 m68881/msoft-float
> MULTILIB_OPTIONS += fPIC
> MULTILIB_DIRNAMES =
> MULTILIB_MATCHES = mcpu32=m68332
> MULTILIB_EXCEPTIONS = m528x/m68881 m528x/msoft-float \
>                    mcpu32/m68881 mcpu32/msoft-float
> LIBGCC = stmp-multilib
> INSTALL_LIBGCC = install-multilib
>
> Here is the same file I modified for the version 4.3.3
>
> LIB1ASMSRC = m68k/lb1sf68.asm
> LIB1ASMFUNCS = _mulsi3 _udivsi3 _divsi3 _umodsi3 _modsi3 \
>   _double _float _floatex \
>   _eqdf2 _nedf2 _gtdf2 _gedf2 _ltdf2 _ledf2 \
>   _eqsf2 _nesf2 _gtsf2 _gesf2 _ltsf2 _lesf2
>
> LIB2FUNCS_EXTRA = fpgnulib.c xfgnulib.c
>
> fpgnulib.c: $(srcdir)/config/m68k/fpgnulib.c
>        cp $(srcdir)/config/m68k/fpgnulib.c fpgnulib.c
> xfgnulib.c: $(srcdir)/config/m68k/fpgnulib.c
>        echo '#define EXTFLOAT' > xfgnulib.c
>        cat $(srcdir)/config/m68k/fpgnulib.c >> xfgnulib.c
>

The above is also included through t-floatlib fragment so you
do not need to add it into  t-m68kelf

> # Custom multilibs for uKOS
>
> M68K_MLIB_CPU += && (match(MLIB, "^68") \
>                     || MLIB == "cpu32" \
>                     || MLIB == "5206" \
>                     || MLIB == "5235" \
>                     || MLIB == "528x")
>
> M68K_MLIB_DIRNAMES = msoft-float fPIC
> M68K_MLIB_OPTIONS = msoft-float fPIC

Should this be M68K_MLIB_OPTIONS = msoft-float/fPIC

> MULTILIB_DIRNAMES = msoft-float fPIC
> MULTILIB_OPTIONS = msoft-float fPIC

these are not needed I think.

> LIBGCC = stmp-multilib
> INSTALL_LIBGCC = install-multilib
>
> Now, when I compile for cpu32 in the .map file I can see that gcc points to
> the right libraries, but this is not the case for the coldfire version:
> Here is what I get:
>
> OK for cpu32:
> LOAD
> /Users/Shared/cross/gcc-4.3.3/m68k-elf/lib/gcc/m68k-elf/4.3.3/../../../../m68k-elf/lib/mcpu32/fPIC/libc.a
> LOAD
> /Users/Shared/cross/gcc-4.3.3/m68k-elf/lib/gcc/m68k-elf/4.3.3/mcpu32/fPIC/libgcc.a
> LOAD
> /Users/Shared/cross/gcc-4.3.3/m68k-elf/lib/gcc/m68k-elf/4.3.3/../../../../m68k-elf/lib/mcpu32/fPIC/libc.a
> LOAD
> /Users/Shared/cross/gcc-4.3.3/m68k-elf/lib/gcc/m68k-elf/4.3.3/mcpu32/fPIC/libgcc.a
>
> Not OK for coldfire:
> LOAD
> /Users/Shared/cross/gcc-4.3.3/m68k-elf/lib/gcc/m68k-elf/4.3.3/../../../../m68k-elf/lib/libc.a
> LOAD
> /Users/Shared/cross/gcc-4.3.3/m68k-elf/lib/gcc/m68k-elf/4.3.3/m5206/fPIC/libgcc.a
> LOAD
> /Users/Shared/cross/gcc-4.3.3/m68k-elf/lib/gcc/m68k-elf/4.3.3/../../../../m68k-elf/lib/libc.a
> LOAD
> /Users/Shared/cross/gcc-4.3.3/m68k-elf/lib/gcc/m68k-elf/4.3.3/m5206/fPIC/libgcc.a
>  Here, in the first and the third line the m5206 is missed. Another problem
> is that gcc wants to use the m5206 instead of the m528x as forced in the
> makefile.
>
> Any ideas for helping me with the coldfire cores?
> Regards
>  Edo.
>
>
>
> --
> For unsubscribe information see http://sourceware.org/lists.html#faq
>
>

--
For unsubscribe information see http://sourceware.org/lists.html#faq

      reply	other threads:[~2009-06-22 21:41 UTC|newest]

Thread overview: 2+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2009-06-21 10:39 Franzi Edo.
2009-06-22 21:41 ` Khem Raj [this message]

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=19c1b8a90906221441l12b6d60fu1f39c28019287c04@mail.gmail.com \
    --to=raj.khem@gmail.com \
    --cc=crossgcc@sourceware.org \
    --cc=edo.franzi@ukos.ch \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).