MVME2306 : kernel 2.2.6 SUCCESS !

Simone Piccardi Simone.Piccardi at
Thu Apr 22 01:04:47 EST 1999

David De Ridder wrote:
> Hi,
> a quick report :
> * I patched my 2.2.4 source tree to 2.2.6 (through 2.2.5)
> * I applied Gabriel's latest patches (for 2.2.6). First generic,
>   then vme. There were no messages during patching (``patch -p0 -s'').
> * ``make distclean'' on the patched tree
> * ``make oldconfig'' using Gabriel's mvme2600.config.nfsroot for 2.2.6
> * ``make bzImage'' (went flawless)
> * ``make modules'' - no problems this time
> * ``make modules_install''
> * ``depmod -a'' gives no messages (this is good)
I just finished to try this procedure (with some more step added) to
cross-compile the kernel on my PC, but it does not work.
I did all the step up to the distclean, then I modify the Makefile
(according to Matt Porter instructions), I copied the
mvme2600.config.nfsroot as .config and did make oldconfig.
Everything up to here went fine, without any error or warning.
Then I did make dep and make clean (always using the Matt Porter
instruction) and the make zImage, at the end the error was this:

powerpc-linux-gcc -D__KERNEL__ -I/home/ppc/linux/include -Wall
-Wstrict-prototypes -O2 -fomit-frame-pointer -D__powerpc__ -fsigned-char
-msoft-float -pipe -fno-builtin -ffixed-r2 -Wno-uninitialized -mmultiple
-mstring -mcpu=603e -D__powerpc__ -msoft-float -pipe -fno-builtin -Wall
-mmultiple -mstring -I../boot -I/home/ppc/linux -D__KERNEL__ -O2
-fomit-frame-pointer -mrelocatable -ffixed-r13 -D__BOOT__ -DDEBUG  -c -o
zlib.o zlib.c
zlib.c:83: string.h: No such file or directory
zlib.c: In function `inflate_blocks':
zlib.c:878: warning: implicit declaration of function `memcpy'
make[1]: *** [zlib.o] Error 1
make[1]: Leaving directory `/home/ppc/linux/arch/ppc/prepboot'
make: *** [zImage] Error 2

I'm a little bit worried about the -mcpu=603e switch but I don't think
this can be relate to the error. 
I looked inside the kernel tree and there are many of these file inside:


The code where the (1st) error come is:
         /* functions */

#include <string.h>
#define zmemcpy memcpy
#define zmemzero(dest, len)     memset(dest, 0, len)

and it seems absolutely normal to me.

The cross compiler was build installing egcs 1.1.2 and the glibc-2.1
over a Redhat 5.2, and then using the Matt Porter instruction (with the
exception of a newer version of binutils, because that one was not
accepted in the egcs upgrade).

If you have any idea about how to solve this I'll be happy.
Thanks to everybody
Simone Piccardi
Microsoft is NOT the answer. Microsoft is the Question.
The answer is: "NO!"

[[ This message was sent via the linuxppc-dev mailing list.  Replies are ]]
[[ not  forced  back  to the list, so be sure to Cc linuxppc-dev if your ]]
[[ reply is of general interest. Please check ]]
[[ and for useful information before posting.   ]]

More information about the Linuxppc-dev mailing list