Re: Fwd: Re: gcc 4.1.1 for mcore

[Alexander Grobman <agrobman at yahoo dot com>]

Hello Nick,

Thanks for your support.

this is the error message I'm getting:
"
...
make[4]: Leaving directory
`/proj/tec/alpeca_lite/users/alexgr/gcc/objdir1/gcc'
/proj/tec/alpeca_lite/users/alexgr/gcc/objdir1/./gcc/xgcc
-B/proj/tec/alpeca_lite/users/alexgr/gcc/objdir1/./gcc/
-B/usr/local/mcore-elf/bin/
-B/usr/local/mcore-elf/lib/ -isystem
/usr/local/mcore-elf/include -isystem
/usr/local/mcore-elf/sys-include -O2  -O2 -g -O2 
-DIN_GCC -DCROSS_COMPILE  -DDONT_HAVE_STDIO
-DDONT_HAVE_SETJMP  -Dinhibit_libc  -W -Wall
-Wwrite-strings -Wstrict-prototypes
-Wmissing-prototypes -Wold-style-definition  -isystem
./include  -O3 -DNO_FLOATLIB_FIXUNSDFSI  -g 
-DIN_LIBGCC2 -D__GCC_FLOAT_NOT_NEEDED -Dinhibit_libc
-I. -I. -I../../gcc-4.1.1/gcc -I../../gcc-4.1.1/gcc/.
-I../../gcc-4.1.1/gcc/../include
-I../../gcc-4.1.1/gcc/../libcpp/include  -DL_floatdisf
-c ../../gcc-4.1.1/gcc/libgcc2.c -o
libgcc/./_floatdisf.o
/tmp/ccvk5vjH.s: Assembler messages:
/tmp/ccvk5vjH.s:38: Error: operand must be absolute in
range 1..32, not 53
make[3]: *** [libgcc/./_floatdisf.o] Error 1
make[3]: Leaving directory
`/proj/tec/alpeca_lite/users/alexgr/gcc/objdir1/gcc'
make[2]: *** [stmp-multilib] Error 2
make[2]: Leaving directory
`/proj/tec/alpeca_lite/users/alexgr/gcc/objdir1/gcc'
make[1]: *** [all-gcc] Error 2
make[1]: Leaving directory
`/proj/tec/alpeca_lite/users/alexgr/gcc/objdir1'
make: *** [all] Error 2

attached are the library file compiled with -S option
with the 


   bmaski  r4,53

wrong instruction form.

and my configuration file.

I run on a Linux machine with AMD CPU (x86_64).

It runs 3.1.3-6.10 RedHat KDE release 2.4.21-32.EL


Thanks again
Alex.
--- Nick Clifton <nickc@redhat.com> wrote:

> Hi Alex,
> 
> > I have some troubles with the MCORE GCC target,
> (see
> > below). Do have any idea what it can be and how to
> > solve the problem?
> 
> Yes - First unless you are constrained to only use
> the 4.1.1 sources 
> please try using the current mainline gcc
> development sources.  It is 
> possible, but unlikely, that the bug has already
> been fixed.
> 
> Assuming that the problem persists, please could you
> put together a 
> small test case to demonstrate it, including details
> of how you 
> configured the toolchain and what sort of host
> machine you are using. 
> If you could send this to me, but, *important*, CC
> the email to the gcc 
> mailing list as well, then I can investigate and try
> to fix the problem.
> 
> >>> The compiler generates illigal form of bmaski
> >>> instruction with operand out of range.
> 
> Interesting.  I have looked at the code locations
> where this instruction 
> is generated and it appears to be correct.  I will
> be intrigued to see 
> how this bug is produced.
> 
> Cheers
>    Nick
> 
> 
>

Attachment: libgcc2.s
Description: 846145001-libgcc2.s

Attachment: config.status
Description: 881791177-config.status