This is the mail archive of the
gcc-help@gcc.gnu.org
mailing list for the GCC project.
cross-compiler is not "naming" anymore object files
- From: Lapo <lapo_pasqui at libero dot it>
- To: gcc-help at gcc dot gnu dot org
- Date: Mon, 18 Aug 2003 14:07:08 +0200
- Subject: cross-compiler is not "naming" anymore object files
- Organization: P.P.R.
- Reply-to: lapo_pasqui at libero dot it
Hi all,
I entered a new (for me) problem.
I'm trying to pack to some RPM a full version of the cross compiler I've
already built.
I followed the classic three steps approach: a minimal gcc cross compiler with
the "inibhit libc" patch to use for building the glibc, the glibc and,
finally, the full gcc.
For some strange reason the process stops at the last steps, when it is trying
to use the already built glibc.
To my big surprise I've noticed the linker is refusing to use the libc.so.6
object files because it cannot test which machine it is for.
Thus, if a look at the libc.so.6 file,
powerpc-linux-readelf -h libc-2.3.2.so
ELF Header:
Magic: 7f 45 4c 46 01 02 01 00 00 00 00 00 00 00 00 00
Class: ELF32
Data: 2's complement, big endian
Version: 1 (current)
OS/ABI: UNIX - System V
ABI Version: 0
Type: DYN (Shared object file)
Machine: None
Version: 0x1
Entry point address: 0x1ce40
Start of program headers: 52 (bytes into file)
Start of section headers: 1266804 (bytes into file)
Flags: 0x0
Size of this header: 52 (bytes)
Size of program headers: 32 (bytes)
Number of program headers: 7
Size of section headers: 40 (bytes)
Number of section headers: 65
Section header string table index: 64
as you can see the Machine entry is set to None.
Is this the expected behaviour?
What may have I changed to blow it up? The configure options are the same
(believe me or not) as for the successfully RPM-outside installation.
Shall I look to the binutils or to the minimal cross complier?
Thanks for your help.
Lapo