From: Timur Tabi <hidden> Date: 2007-10-08 21:33:24
Looks like the problem is back:
BOOTCC arch/powerpc/boot/treeboot-walnut.o
Assembler messages:
Error: Internal assembler error for instruction icbt
Internal error, aborting at
/tmp/crosstool/crosstool-0.42/build/powerpc-unknown-linux-gnu/gcc-4.0.2_e300-enabled-glibc-2.3.6/binutils-2.16.1-complete/gas/config/tc-ppc.c
line 1314 in ppc_setup_opcodes
Please report this bug.
make[1]: *** [arch/powerpc/boot/treeboot-walnut.o] Error 2
make: *** [uImage] Error 2
Question: I'm building a kernel for the 8610. Why is treeboot-walnut.c being
compiled at all?
Grant Likely wrote:
Building Paul's current powerpc branch with a ppc6xx compiler causes
the following build failure:
SYSMAP System.map
BOOTCC arch/powerpc/boot/treeboot-walnut.o
{standard input}: Assembler messages:
{standard input}:184: Error: Unrecognized opcode: `mfdcr'
{standard input}:185: Error: Unrecognized opcode: `mfdcr'
{standard input}:186: Error: Unrecognized opcode: `mfdcr'
{standard input}:217: Error: Unrecognized opcode: `mtdcr'
make[1]: *** [arch/powerpc/boot/treeboot-walnut.o] Error 1
make: *** [uImage] Error 2
This is using ELDK 4.1:
GNU assembler version 2.16.1 (powerpc-linux) using BFD version 2.16.1
gcc version 4.0.0 (DENX ELDK 4.1 4.0.0)
Cheers,
g.
From: David Gibson <hidden> Date: 2007-10-09 03:07:29
On Mon, Oct 08, 2007 at 04:33:24PM -0500, Timur Tabi wrote:
Looks like the problem is back:
BOOTCC arch/powerpc/boot/treeboot-walnut.o
Assembler messages:
Error: Internal assembler error for instruction icbt
Internal error, aborting at
/tmp/crosstool/crosstool-0.42/build/powerpc-unknown-linux-gnu/gcc-4.0.2_e300-enabled-glibc-2.3.6/binutils-2.16.1-complete/gas/config/tc-ppc.c
line 1314 in ppc_setup_opcodes
Please report this bug.
make[1]: *** [arch/powerpc/boot/treeboot-walnut.o] Error 2
make: *** [uImage] Error 2
I thought that got fixed already, did the CFLAGS override fall out
somehow?
Question: I'm building a kernel for the 8610. Why is treeboot-walnut.c being
compiled at all?
Policy. Compiling everything means build bugs - like this one - can
be found by everybody, not just those building for the specific
obscure platform.
--
David Gibson | I'll have my music baroque, and my code
david AT gibson.dropbear.id.au | minimalist, thank you. NOT _the_ _other_
| _way_ _around_!
http://www.ozlabs.org/~dgibson
From: Grant Likely <hidden> Date: 2007-10-09 04:00:18
On 10/8/07, Timur Tabi [off-list ref] wrote:
Looks like the problem is back:
BOOTCC arch/powerpc/boot/treeboot-walnut.o
Assembler messages:
Error: Internal assembler error for instruction icbt
Internal error, aborting at
/tmp/crosstool/crosstool-0.42/build/powerpc-unknown-linux-gnu/gcc-4.0.2_e300-enabled-glibc-2.3.6/binutils-2.16.1-complete/gas/config/tc-ppc.c
line 1314 in ppc_setup_opcodes
Please report this bug.
make[1]: *** [arch/powerpc/boot/treeboot-walnut.o] Error 2
make: *** [uImage] Error 2
No; this is something different. The assember itself is failing
internally. What version of binutils are you running?
g.
--
Grant Likely, B.Sc., P.Eng.
Secret Lab Technologies Ltd.
grant.likely@secretlab.ca
(403) 399-0195
From: Kumar Gala <hidden> Date: 2007-10-09 13:30:31
On Oct 8, 2007, at 11:00 PM, Grant Likely wrote:
On 10/8/07, Timur Tabi [off-list ref] wrote:
quoted
Looks like the problem is back:
BOOTCC arch/powerpc/boot/treeboot-walnut.o
Assembler messages:
Error: Internal assembler error for instruction icbt
Internal error, aborting at
/tmp/crosstool/crosstool-0.42/build/powerpc-unknown-linux-gnu/
gcc-4.0.2_e300-enabled-glibc-2.3.6/binutils-2.16.1-complete/gas/
config/tc-ppc.c
line 1314 in ppc_setup_opcodes
Please report this bug.
make[1]: *** [arch/powerpc/boot/treeboot-walnut.o] Error 2
make: *** [uImage] Error 2
No; this is something different. The assember itself is failing
internally. What version of binutils are you running?
2.61.1 if we believe his path :)
the assembler is failing because 'icbt' has several possible
different instruction encodings and I'm guessing he's getting a
conflict set.
What's the actual compile line? Want to see the flags to assembler
of -Wa to the compiler.
- k
What's the actual compile line? Want to see the flags to assembler of
-Wa to the compiler.
How do I modify the makefiles to spit out that command line?
make V=1
With kind regards,
Geert Uytterhoeven
Software Architect
Sony Network and Software Technology Center Europe
The Corporate Village · Da Vincilaan 7-D1 · B-1935 Zaventem · Belgium
Phone: +32 (0)2 700 8453
Fax: +32 (0)2 700 8622
E-mail: Geert.Uytterhoeven@sonycom.com
Internet: http://www.sony-europe.com/
Sony Network and Software Technology Center Europe
A division of Sony Service Centre (Europe) N.V.
Registered office: Technologielaan 7 · B-1840 Londerzeel · Belgium
VAT BE 0413.825.160 · RPR Brussels
Fortis Bank Zaventem · Swift GEBABEBB08A · IBAN BE39001382358619
From: Timur Tabi <hidden> Date: 2007-10-09 16:06:05
David Gibson wrote:
Policy. Compiling everything means build bugs - like this one - can
be found by everybody, not just those building for the specific
obscure platform.
Is this a new policy? Modules in the kernel are not built unless you want
them. Even in arch/powerpc/platforms, only the specific platform file I'm
targeting is built. So I don't really understand why you claim it's normal
for platform-specific files to be built, regardless of the actual platform.
And frankly, I don't like this "policy". Build bugs for 4xx should not
interfere in my 83xx development. I can't build any kernels now because of
this bug!
--
Timur Tabi
Linux Kernel Developer @ Freescale
From: Scott Wood <hidden> Date: 2007-10-09 16:06:45
On Tue, Oct 09, 2007 at 01:07:29PM +1000, David Gibson wrote:
On Mon, Oct 08, 2007 at 04:33:24PM -0500, Timur Tabi wrote:
quoted
Question: I'm building a kernel for the 8610. Why is treeboot-walnut.c
being compiled at all?
Policy. Compiling everything means build bugs - like this one - can
be found by everybody, not just those building for the specific
obscure platform.
Of course, it also introduces bugs that wouldn't have been an issue if we
didn't try to build everything with the same toolchain. :-P
I'm also somewhat worried what it'll do to build time as platforms
accumulate.
-Scott
On Tue, 09 Oct 2007 11:06:05 -0500
Timur Tabi [off-list ref] wrote:
David Gibson wrote:
quoted
Policy. Compiling everything means build bugs - like this one - can
be found by everybody, not just those building for the specific
obscure platform.
Is this a new policy? Modules in the kernel are not built unless you want
them. Even in arch/powerpc/platforms, only the specific platform file I'm
targeting is built. So I don't really understand why you claim it's normal
for platform-specific files to be built, regardless of the actual platform.
And frankly, I don't like this "policy". Build bugs for 4xx should not
interfere in my 83xx development. I can't build any kernels now because of
this bug!
I fixed this bug. It should be in Paul's latest tree. You seem to be
having a different issue with your binutils though.
josh
From: Kumar Gala <hidden> Date: 2007-10-09 19:14:08
On Oct 9, 2007, at 1:52 PM, Josh Boyer wrote:
On Tue, 09 Oct 2007 11:06:05 -0500
Timur Tabi [off-list ref] wrote:
quoted
David Gibson wrote:
quoted
Policy. Compiling everything means build bugs - like this one - can
be found by everybody, not just those building for the specific
obscure platform.
Is this a new policy? Modules in the kernel are not built unless
you want
them. Even in arch/powerpc/platforms, only the specific platform
file I'm
targeting is built. So I don't really understand why you claim
it's normal
for platform-specific files to be built, regardless of the actual
platform.
And frankly, I don't like this "policy". Build bugs for 4xx
should not
interfere in my 83xx development. I can't build any kernels now
because of
this bug!
I fixed this bug. It should be in Paul's latest tree. You seem to be
having a different issue with your binutils though.
Just to clarify this is an issue with the particular patched binutils
Timur is using.
I'll leave it to you guys to fight over if we should be building
everything or not.
- k
From: David Gibson <hidden> Date: 2007-10-10 02:35:15
On Tue, Oct 09, 2007 at 11:06:05AM -0500, Timur Tabi wrote:
David Gibson wrote:
quoted
Policy. Compiling everything means build bugs - like this one - can
be found by everybody, not just those building for the specific
obscure platform.
Is this a new policy? Modules in the kernel are not built unless you want
them. Even in arch/powerpc/platforms, only the specific platform file I'm
targeting is built. So I don't really understand why you claim it's normal
for platform-specific files to be built, regardless of the actual
platform.
It has been policy as long as we've had the platform+library structure
in arch/powerpc/boot.
And frankly, I don't like this "policy". Build bugs for 4xx should not
interfere in my 83xx development. I can't build any kernels now because of
this bug!
Argue with Paulus.
--
David Gibson | I'll have my music baroque, and my code
david AT gibson.dropbear.id.au | minimalist, thank you. NOT _the_ _other_
| _way_ _around_!
http://www.ozlabs.org/~dgibson
From: David Gibson <hidden> Date: 2007-10-10 02:35:54
On Tue, Oct 09, 2007 at 11:06:45AM -0500, Scott Wood wrote:
On Tue, Oct 09, 2007 at 01:07:29PM +1000, David Gibson wrote:
quoted
On Mon, Oct 08, 2007 at 04:33:24PM -0500, Timur Tabi wrote:
quoted
Question: I'm building a kernel for the 8610. Why is treeboot-walnut.c
being compiled at all?
Policy. Compiling everything means build bugs - like this one - can
be found by everybody, not just those building for the specific
obscure platform.
Of course, it also introduces bugs that wouldn't have been an issue if we
didn't try to build everything with the same toolchain. :-P
I'm also somewhat worried what it'll do to build time as platforms
accumulate.
It also provides incentive to keep the bootwrapper small...
--
David Gibson | I'll have my music baroque, and my code
david AT gibson.dropbear.id.au | minimalist, thank you. NOT _the_ _other_
| _way_ _around_!
http://www.ozlabs.org/~dgibson
From: Paul Mackerras <hidden> Date: 2007-10-11 00:12:35
Timur Tabi writes:
Is this a new policy? Modules in the kernel are not built unless you want
them. Even in arch/powerpc/platforms, only the specific platform file I'm
targeting is built. So I don't really understand why you claim it's normal
for platform-specific files to be built, regardless of the actual platform.
The wrapper script is intended to be a standalone tool which is
independent of the kernel configuration, and can be used separately
from the kernel build process.
And frankly, I don't like this "policy". Build bugs for 4xx should not
interfere in my 83xx development. I can't build any kernels now because of
this bug!
I'm sure you're not actually that helpless. :) Just carry a patch to
arch/powerpc/boot/Makefile in your local tree.
Paul.
From: Scott Wood <hidden> Date: 2007-10-11 20:02:37
On Thu, Oct 11, 2007 at 10:12:35AM +1000, Paul Mackerras wrote:
Timur Tabi writes:
quoted
Is this a new policy? Modules in the kernel are not built unless you want
them. Even in arch/powerpc/platforms, only the specific platform file I'm
targeting is built. So I don't really understand why you claim it's normal
for platform-specific files to be built, regardless of the actual platform.
The wrapper script is intended to be a standalone tool which is
independent of the kernel configuration, and can be used separately
from the kernel build process.
Is it ever actually used that way? I wonder if this "intent" is worth the
pain...
-Scott