bug-binutils
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Is this a bug?


From: Gonzalo Jimenez Sanchez
Subject: Is this a bug?
Date: Mon, 29 Aug 2005 13:03:01 +0200

Hi.
I have built cross compiler tool chain with:
binutils-2.16.1 cross compiled windows - powerpc-eabi.

This is my script for building:
export TARGET=powerpc-eabi
export PREFIX=/tgnuppc/$TARGET
export PATH=$PATH:$PREFIX/bin

mkdir t-build-binutils
cd t-build-binutils
../binutils-2.16.1/configure --target=$TARGET --prefix=$PREFIX
--enable-languages=c,c++
make all
make install




I find this error when assembling my code.
c:/tgnuppc/powerpc-eabi/powerpc-eabi/bin/as -I.. -W -m603 -mregnames -o
../_b8260/obj/regs.o   ../nivel1/regs/regs.s
init8260.s: Assembler messages:
regs.s:445: Error: operand out of range (0x0000aa39 is not between
0xffff8000 and 0x00007fff)

line 445 in that file is: li      r3,0xAA39


c:/tgnuppc/powerpc-eabi/powerpc-eabi/bin/as -I../.. -I.. -W -m603 -mregnames
-o ../_b8260/obj/init8260.o   ../init8260.s
init8260.s: Assembler messages:
init8260.s:1067: Error: operand out of range (0x0000ffff is not between
0xffff8000 and 0x00007fff)

line 1067 in that file is: addi    r3,0,0xFFFF



If I trie to assemble an initialization code from Motorola I have the same
error.

 <<8260_int.s>>  <<Init8260.s>> 

In 8260_int.s you can find in line 385:    addi     r2,0,0xFFFD      # Load
the mask - 0xFFFFFFFD




I tried with this change in binutils-2.16.1\opcodes\ppc-opc.c
Original:
{ "li",      OP(14),    DRA_MASK,       PPCCOM,         { RT, SI } },

{ "addi",    OP(14),    OP_MASK,        PPCCOM,         { RT, RA0, SI } },


My change:
{ "li",      OP(14),    DRA_MASK,       PPCCOM,         { RT, SISIGNOPT } },

{ "addi",    OP(14),    OP_MASK,        PPCCOM,         { RT, RA0, SISIGNOPT
} },


now assembling goes fine.

Is that a bug in binutils or I'm using SIMM field in addi and li instruction
erroneously?

If that is a bug, Is my change correct? Are there more instruction with de
same bug?


best regards,
Gonzalo Jimenez


Attachment: 8260_int.s
Description: Binary data

Attachment: Init8260.s
Description: Binary data


reply via email to

[Prev in Thread] Current Thread [Next in Thread]