[Buildroot] 2009.11-rc1 build failure

[Buildroot] 2009.11-rc1 build failure

[Joachim Pihl]

Just trying out building RC1 straight from the "box", basically all  
defaults and adding a few packages. Set to use headers matching a  
previously configured kernel. Running on Ubuntu 9.10



make[1]: Entering directory  
`/home/jpihl/system_controller/buildroot-2009.11_rc1/output/toolchain/linux-2.6.27.10'
   CHK     include/linux/version.h
   HOSTCC  scripts/basic/fixdep
   UPD     include/linux/version.h
   HOSTCC  scripts/basic/docproc
   HOSTCC  scripts/unifdef
scripts/unifdef.c:209: error: conflicting types for ?getline?
/usr/include/stdio.h:651: note: previous declaration of ?getline? was here
make[2]: *** [scripts/unifdef] Error 1
make[1]: *** [__headers] Error 2
make[1]: Leaving directory  
`/home/jpihl/system_controller/buildroot-2009.11_rc1/output/toolchain/linux-2.6.27.10'
make: ***  
[/home/jpihl/system_controller/buildroot-2009.11_rc1/output/toolchain/linux/.configured]  
Error 2
jpihl at jap-lappis:~/system_controller/buildroot-2009.11_rc1$ make

Checking build system dependencies:
BUILDROOT_DL_DIR clean:				Ok
CC clean:					Ok
CXX clean:					Ok
CPP clean:					Ok
CFLAGS clean:					Ok
INCLUDES clean:					Ok
CXXFLAGS clean:					Ok
which installed:				Ok
sed works:					Ok (/bin/sed)
GNU make version '3.81':			Ok
C compiler '/usr/bin/gcc'
C compiler version '4.4.1':			Ok
C++ compiler '/usr/bin/g++'
C++ compiler version '4.4.1':			Ok
awk installed:					Ok
bash installed:					Ok
bison installed:				Ok
flex installed:					Ok
gettext installed:				Ok
makeinfo installed:				Ok
patch installed:				Ok
Build system dependencies:			Ok

rm -rf  
/home/jpihl/system_controller/buildroot-2009.11_rc1/output/build/buildroot-config
mkdir -p /home/jpihl/system_controller/buildroot-2009.11_rc1/output/build
cp -dpRf package/config/buildroot-config  
/home/jpihl/system_controller/buildroot-2009.11_rc1/output/build/buildroot-config
(cd  
/home/jpihl/system_controller/buildroot-2009.11_rc1/output/toolchain/linux-2.6.27.10;  
\
	 /usr/bin/make -j2 ARCH=arm \
		HOSTCC="/usr/bin/gcc" HOSTCFLAGS="" \
		HOSTCXX="/usr/bin/g++" \
		INSTALL_HDR_PATH=/home/jpihl/system_controller/buildroot-2009.11_rc1/output/toolchain/linux  
headers_install; \
	)
make[1]: Entering directory  
`/home/jpihl/system_controller/buildroot-2009.11_rc1/output/toolchain/linux-2.6.27.10'
   CHK     include/linux/version.h
   HOSTCC  scripts/unifdef
scripts/unifdef.c:209: error: conflicting types for ?getline?
/usr/include/stdio.h:651: note: previous declaration of ?getline? was here
make[2]: *** [scripts/unifdef] Error 1
make[1]: *** [__headers] Error 2
make[1]: Leaving directory  
`/home/jpihl/system_controller/buildroot-2009.11_rc1/output/toolchain/linux-2.6.27.10'
make: ***  
[/home/jpihl/system_controller/buildroot-2009.11_rc1/output/toolchain/linux/.configured]  
Error 2



-- 
Using Opera's revolutionary e-mail client: http://www.opera.com/mail/
-------------- next part --------------
A non-text attachment was scrubbed...
Name: .config
Type: application/octet-stream
Size: 20027 bytes
Desc: not available
URL: <http://lists.busybox.net/pipermail/buildroot/attachments/20091126/67ca1779/attachment-0001.obj>

[Buildroot] 2009.11-rc1 build failure

[Peter Korsgaard]

>>>>> "Joachim" == Joachim Pihl <jpihl@nevion.com> writes:

 Joachim> Just trying out building RC1 straight from the "box", basically all
 Joachim> defaults and adding a few packages. Set to use headers matching a
 Joachim> previously configured kernel. Running on Ubuntu 9.10

Huh, that's wrong - The config you attached has:

BR2_KERNEL_HEADERS_2_6_31=y
# BR2_KERNEL_HEADERS_VERSION is not set
# BR2_KERNEL_HEADERS_SNAP is not set
BR2_DEFAULT_KERNEL_HEADERS="2.6.31.6"

And still you are trying to use headers from 2.6.27.10.

The getline problem is a known issue with old kernels and new
distributions (fixed in 2.6.29+ and later versions of 2.6.27.x).

Are you sure you're using that exact config and just running make?

 Joachim> make[1]: Entering directory
 Joachim> /home/jpihl/system_controller/buildroot-2009.11_rc1/output/toolchain/linux-2.6.27.10'

-- 
Bye, Peter Korsgaard

[Buildroot] 2009.11-rc1 build failure

[Joachim Pihl]

On Thu, 26 Nov 2009 15:31:17 +0100, Peter Korsgaard <jacmet@uclibc.org>  
wrote:

> Huh, that's wrong - The config you attached has:
>
> BR2_KERNEL_HEADERS_2_6_31=y
> # BR2_KERNEL_HEADERS_VERSION is not set
> # BR2_KERNEL_HEADERS_SNAP is not set
> BR2_DEFAULT_KERNEL_HEADERS="2.6.31.6"
>
> And still you are trying to use headers from 2.6.27.10.

This is what I extracted from the email that came back from this list, I  
think you made a mistake somewhere:

# BR2_KERNEL_HEADERS_2_6_31 is not set
BR2_KERNEL_HEADERS_VERSION=y
# BR2_KERNEL_HEADERS_SNAP is not set
BR2_DEFAULT_KERNEL_VERSION="2.6.27.10"
BR2_DEFAULT_KERNEL_HEADERS="2.6.27.10"

> Are you sure you're using that exact config and just running make?

AFAIK .config should be the right config file,
sequence was as follows:

download and unpack
make xconfig
manually specify kernel version (matches the already patched kernel I had  
lying about)
exit config
make




> The getline problem is a known issue with old kernels and new
> distributions (fixed in 2.6.29+ and later versions of 2.6.27.x).

OK.

Some time back, I remember hearing that 2.6.27 was the branch that was to  
be maintained for XScale PXA. Is this correct, or are all XScale oddities  
handled in later kernel versions?





-- 
Joachim Pihl
Senior R&D Engineer
Nevion

Tel: +47 33 48 94 66
Fax: +47 33 48 99 98
Mobile: +47 91 33 98 91
jpihl at nevion.com
www.nevion.com

-----------------------------------
The Global Video Transport Leader
-----------------------------------

[Buildroot] 2009.11-rc1 build failure

[Peter Korsgaard]

>>>>> "Joachim" == Joachim Pihl <jpihl@nevion.com> writes:

Hi,

 Joachim> This is what I extracted from the email that came back from
 Joachim> this list, I think you made a mistake somewhere:

 Joachim> # BR2_KERNEL_HEADERS_2_6_31 is not set
 Joachim> BR2_KERNEL_HEADERS_VERSION=y
 Joachim> # BR2_KERNEL_HEADERS_SNAP is not set
 Joachim> BR2_DEFAULT_KERNEL_VERSION="2.6.27.10"
 Joachim> BR2_DEFAULT_KERNEL_HEADERS="2.6.27.10"

Argh, sorry - I have too many /tmp/.config* files :/

Your fix is then to use the 2.6.27.x version instead of hardcoding
2.6.27.10 - That will make you use 2.6.27.39 which has the getline fix.

Notice that your kernel headers version doesn't need to be the same as
the kernel version you're using.

 >> The getline problem is a known issue with old kernels and new
 >> distributions (fixed in 2.6.29+ and later versions of 2.6.27.x).

 Joachim> OK.

 Joachim> Some time back, I remember hearing that 2.6.27 was the branch
 Joachim> that was to be maintained for XScale PXA. Is this correct, or
 Joachim> are all XScale oddities handled in later kernel versions?

I don't know about PXA in particular (but even if so, please use the
latest in the 2.6.27.x series). In general, the latest stable kernel
release (currently 2.6.31.6) is the most sensible choice.

-- 
Bye, Peter Korsgaard

[Buildroot] 2009.11-rc1 build failure

[Michael S. Zick]

On Fri November 27 2009, Peter Korsgaard wrote:
> 
> Notice that your kernel headers version doesn't need to be the same as
> the kernel version you're using.
> 
- - - -
>
> I don't know about PXA in particular (but even if so, please use the
> latest in the 2.6.27.x series). In general, the latest stable kernel
> release (currently 2.6.31.6) is the most sensible choice.
> 

Although a newcomer to using Buildroot -
the above is not Buildroot specific advice -
It is a good rule of thumb when building for Linux.

For example (my current use case):

I am building user land code against the 2.6.31.6 *headers*
even though that code will be running on a 2.6.12.6 kernel
released Aug. 29, 2005. (At least until I replace it.)

That is an amazing level of backwards compatibility.
Of course, it allows a person to build code today that uses
system calls that might not exist at run-time. ;)
But in cases like that, your userland code will get an error,
not break something.

On the other hand, I can see your case for wanting to build
against the oldest, stable (I.E: still supported) kernel.
Which is 2.6.27.x, most recently released version Nov. 10, 2009.
Doing that might avoid some of those run-time surprises.  ;)

So it probably depends on your own project goals -
Greatest degree of compatibility now (2.6.27.x), or
longest lasting compatibility going forwarded (2.6.31.x).

I am shooting for "rebuild userland every 5 years" myself. ;)

Mike

[Buildroot] 2009.11-rc1 build failure

[Joachim Pihl]

On Fri, 27 Nov 2009 11:02:14 +0100, Peter Korsgaard <jacmet@uclibc.org>
wrote:
> Argh, sorry - I have too many /tmp/.config* files :/
>
> Your fix is then to use the 2.6.27.x version instead of hardcoding
> 2.6.27.10 - That will make you use 2.6.27.39 which has the getline fix.
>
> Notice that your kernel headers version doesn't need to be the same as
> the kernel version you're using.

Noted. Tried punching in 2.6.27.x, that failed miserably, but 2.6.27.39
worked - all the way up to the next hiccup:

...
checking dynamic linker characteristics... GNU/Linux ld.so
checking how to hardcode library paths into programs... immediate
appending configuration tag "F77" to libtool
checking for arm-linux-gcc... (cached)
/home/jpihl/system_controller/buildroot-2009.11_rc1/output/staging/usr/bin/arm-linux-uclibcgnueabi-gcc
-Os -pipe -O3
-I/home/jpihl/system_controller/buildroot-2009.11_rc1/output/staging/usr/include
-I/home/jpihl/system_controller/buildroot-2009.11_rc1/output/staging/include
--sysroot=/home/jpihl/system_controller/buildroot-2009.11_rc1/output/staging/
-isysroot
/home/jpihl/system_controller/buildroot-2009.11_rc1/output/staging
-mtune=xscale -march=armv5te -mabi=aapcs-linux -msoft-float
checking whether we are using the GNU C compiler... (cached) yes
checking whether
/home/jpihl/system_controller/buildroot-2009.11_rc1/output/staging/usr/bin/arm-linux-uclibcgnueabi-gcc
-Os -pipe -O3
-I/home/jpihl/system_controller/buildroot-2009.11_rc1/output/staging/usr/include
-I/home/jpihl/system_controller/buildroot-2009.11_rc1/output/staging/include
--sysroot=/home/jpihl/system_controller/buildroot-2009.11_rc1/output/staging/
-isysroot
/home/jpihl/system_controller/buildroot-2009.11_rc1/output/staging
-mtune=xscale -march=armv5te -mabi=aapcs-linux -msoft-float accepts -g...
(cached) yes
checking for
/home/jpihl/system_controller/buildroot-2009.11_rc1/output/staging/usr/bin/arm-linux-uclibcgnueabi-gcc
-Os -pipe -O3
-I/home/jpihl/system_controller/buildroot-2009.11_rc1/output/staging/usr/include
-I/home/jpihl/system_controller/buildroot-2009.11_rc1/output/staging/include
--sysroot=/home/jpihl/system_controller/buildroot-2009.11_rc1/output/staging/
-isysroot
/home/jpihl/system_controller/buildroot-2009.11_rc1/output/staging
-mtune=xscale -march=armv5te -mabi=aapcs-linux -msoft-float option to
accept ISO C89... (cached) none needed
checking dependency style of
/home/jpihl/system_controller/buildroot-2009.11_rc1/output/staging/usr/bin/arm-linux-uclibcgnueabi-gcc
-Os -pipe -O3
-I/home/jpihl/system_controller/buildroot-2009.11_rc1/output/staging/usr/include
-I/home/jpihl/system_controller/buildroot-2009.11_rc1/output/staging/include
--sysroot=/home/jpihl/system_controller/buildroot-2009.11_rc1/output/staging/
-isysroot
/home/jpihl/system_controller/buildroot-2009.11_rc1/output/staging
-mtune=xscale -march=armv5te -mabi=aapcs-linux -msoft-float... (cached)
gcc3
checking how to run the C preprocessor...
/home/jpihl/system_controller/buildroot-2009.11_rc1/output/staging/usr/bin/arm-linux-uclibcgnueabi-cpp
-Os -pipe -O3
-I/home/jpihl/system_controller/buildroot-2009.11_rc1/output/staging/usr/include
-I/home/jpihl/system_controller/buildroot-2009.11_rc1/output/staging/include
--sysroot=/home/jpihl/system_controller/buildroot-2009.11_rc1/output/staging/
-isysroot
/home/jpihl/system_controller/buildroot-2009.11_rc1/output/staging
-mtune=xscale -march=armv5te -mabi=aapcs-linux -msoft-float
checking whether we are using the GNU C++ compiler... (cached) yes
checking whether
/home/jpihl/system_controller/buildroot-2009.11_rc1/output/staging/usr/bin/arm-linux-uclibcgnueabi-g++
-Os -pipe -O3
-I/home/jpihl/system_controller/buildroot-2009.11_rc1/output/staging/usr/include
-I/home/jpihl/system_controller/buildroot-2009.11_rc1/output/staging/include
--sysroot=/home/jpihl/system_controller/buildroot-2009.11_rc1/output/staging/
-isysroot
/home/jpihl/system_controller/buildroot-2009.11_rc1/output/staging
-mtune=xscale -march=armv5te -mabi=aapcs-linux -msoft-float accepts -g...
(cached) yes
checking dependency style of
/home/jpihl/system_controller/buildroot-2009.11_rc1/output/staging/usr/bin/arm-linux-uclibcgnueabi-g++
-Os -pipe -O3
-I/home/jpihl/system_controller/buildroot-2009.11_rc1/output/staging/usr/include
-I/home/jpihl/system_controller/buildroot-2009.11_rc1/output/staging/include
--sysroot=/home/jpihl/system_controller/buildroot-2009.11_rc1/output/staging/
-isysroot
/home/jpihl/system_controller/buildroot-2009.11_rc1/output/staging
-mtune=xscale -march=armv5te -mabi=aapcs-linux -msoft-float... (cached)
gcc3
checking for ld...
/home/jpihl/system_controller/buildroot-2009.11_rc1/output/staging/usr/bin/arm-linux-uclibcgnueabi-ld
checking for kernel OProfile support... yes
checking whether malloc attribute is understood... no
checking whether __builtin_expect is understood... yes
checking for sched_setaffinity... yes
checking for perfmonctl... no
checking for poptGetContext in -lpopt... yes
checking libiberty.h usability... yes
checking libiberty.h presence... yes
checking for libiberty.h... yes
checking for cplus_demangle in -liberty... yes
checking for xcalloc... yes
checking for xmemdup... yes
checking for dlopen in -ldl... yes
checking for main in -lintl... no
checking for bfd_openr in -lbfd... no
configure: error: bfd library not found
make: ***
[/home/jpihl/system_controller/buildroot-2009.11_rc1/output/build/oprofile-0.9.4/.stamp_configured]
Error 1


OProfile does not want to build, and a brief search in the config menus
didn't reveal any options containing "bfd". As far as I could tell from
the OProfile pages, it should work for the PXA processors?

-- 
Using Opera's revolutionary e-mail client: http://www.opera.com/mail/

[Buildroot] 2009.11-rc1 build failure

[Peter Korsgaard]

>>>>> "Joachim" == Joachim Pihl <jpihl@nevion.com> writes:

[Please keep communication on the list instead of mailing me directly]

 >> Notice that your kernel headers version doesn't need to be the same as
 >> the kernel version you're using.

 Joachim> Noted. Tried punching in 2.6.27.x, that failed miserably, but
 Joachim> 2.6.27.39  worked - all the way up to the next hiccup:

Failed in what way? Please provide more details so we can fix the
problem if there is one.

 Joachim> OProfile does not want to build, and a brief search in the
 Joachim> config menus didn't reveal any options containing "bfd". As
 Joachim> far as I could tell from the OProfile pages, it should work
 Joachim> for the PXA processors?

The oprofile package hasn't seen any updates in a very long time (and I
don't use it myself), so chances are that it is broken.

I'll take a look now.

-- 
Bye, Peter Korsgaard

[Buildroot] 2009.11-rc1 build failure

[Michael S. Zick]

On Fri November 27 2009, Peter Korsgaard wrote:
> >>>>> "Joachim" == Joachim Pihl <jpihl@nevion.com> writes:
> 
> [Please keep communication on the list instead of mailing me directly]
> 
>  >> Notice that your kernel headers version doesn't need to be the same as
>  >> the kernel version you're using.
> 
>  Joachim> Noted. Tried punching in 2.6.27.x, that failed miserably, but
>  Joachim> 2.6.27.39  worked - all the way up to the next hiccup:
> 
> Failed in what way? Please provide more details so we can fix the
> problem if there is one.
> 
>  Joachim> OProfile does not want to build, and a brief search in the
>  Joachim> config menus didn't reveal any options containing "bfd". As
>  Joachim> far as I could tell from the OProfile pages, it should work
>  Joachim> for the PXA processors?
> 

"BFD" is the library on which binutils is built.
If you don't have that, you may go home now.  ;)

Mike
> The oprofile package hasn't seen any updates in a very long time (and I
> don't use it myself), so chances are that it is broken.
> 
> I'll take a look now.
> 

[Buildroot] 2009.11-rc1 build failure

[Joachim Pihl]

On Fri, 27 Nov 2009 19:13:49 +0100, Peter Korsgaard <jacmet@uclibc.org>  
wrote:

>>>>>> "Joachim" == Joachim Pihl <jpihl@nevion.com> writes:
>
> [Please keep communication on the list instead of mailing me directly]
>

Sorry for that, the mailing list is set up with reply-to the original  
author. I habitually do NOT send a reply-to-all, and I didn't catch the  
error until a few minutes after I sent it.

>  >> Notice that your kernel headers version doesn't need to be the same  
> as
>  >> the kernel version you're using.
>
>  Joachim> Noted. Tried punching in 2.6.27.x, that failed miserably, but
>  Joachim> 2.6.27.39  worked - all the way up to the next hiccup:
>
> Failed in what way? Please provide more details so we can fix the
> problem if there is one.

...
flex installed:					Ok
gettext installed:				Ok
makeinfo installed:				Ok
patch installed:				Ok
Build system dependencies:			Ok

rm -rf  
/home/jpihl/system_controller/buildroot-2009.11_rc1/output/build/buildroot-config
mkdir -p /home/jpihl/system_controller/buildroot-2009.11_rc1/output/build
cp -dpRf package/config/buildroot-config  
/home/jpihl/system_controller/buildroot-2009.11_rc1/output/build/buildroot-config
--2009-11-27 19:38:34--   
http://www.kernel.org/pub//linux/kernel/v2.6//linux-2.6.27.x.tar.bz2
Resolving www.kernel.org... 130.239.17.4
Connecting to www.kernel.org|130.239.17.4|:80... connected.
HTTP request sent, awaiting response... 404 Not Found
2009-11-27 19:38:34 ERROR 404: Not Found.

--2009-11-27 19:38:34--   
http://buildroot.net/downloads/sources//linux-2.6.27.x.tar.bz2
Resolving buildroot.net... 140.211.167.224
Connecting to buildroot.net|140.211.167.224|:80... connected.
HTTP request sent, awaiting response... 404 Not Found
2009-11-27 19:38:35 ERROR 404: Not Found.

make: *** [dl/linux-2.6.27.x.tar.bz2] Error 1


I may have misunderstood something here...


>  Joachim> OProfile does not want to build, and a brief search in the
>  Joachim> config menus didn't reveal any options containing "bfd". As
>  Joachim> far as I could tell from the OProfile pages, it should work
>  Joachim> for the PXA processors?
>
> The oprofile package hasn't seen any updates in a very long time (and I
> don't use it myself), so chances are that it is broken.
>
> I'll take a look now.
>


-- 
Using Opera's revolutionary e-mail client: http://www.opera.com/mail/

[Buildroot] 2009.11-rc1 build failure

[Peter Korsgaard]

>>>>> "Joachim" == Joachim Pihl <jpihl@nevion.com> writes:

Hi,

 >> [Please keep communication on the list instead of mailing me directly]

 Joachim> Sorry for that, the mailing list is set up with reply-to the original
 Joachim> author. I habitually do NOT send a reply-to-all, and I didn't catch
 Joachim> the  error until a few minutes after I sent it.

OK. The reply-to setting is pretty much standard and done on purpose.

 >> Failed in what way? Please provide more details so we can fix the
 >> problem if there is one.

 Joachim> ...
 Joachim> flex installed:					Ok
 Joachim> gettext installed:				Ok
 Joachim> makeinfo installed:				Ok
 Joachim> patch installed:				Ok
 Joachim> Build system dependencies:			Ok

 Joachim> rm -rf
 Joachim> /home/jpihl/system_controller/buildroot-2009.11_rc1/output/build/buildroot-config
 Joachim> mkdir -p /home/jpihl/system_controller/buildroot-2009.11_rc1/output/build
 Joachim> cp -dpRf package/config/buildroot-config
 Joachim> /home/jpihl/system_controller/buildroot-2009.11_rc1/output/build/buildroot-config
 Joachim> --2009-11-27 19:38:34--
 Joachim> http://www.kernel.org/pub//linux/kernel/v2.6//linux-2.6.27.x.tar.bz2

Ahh yes, you have to select 2.6.27.x in the choice box above instead of
typing the string (Where you have now 'manually selected version')

-- 
Bye, Peter Korsgaard

[Buildroot] 2009.11-rc1 build failure

[Joachim Pihl]

On Fri, 27 Nov 2009 19:51:36 +0100, Peter Korsgaard <jacmet@uclibc.org>  
wrote:

>>>>>> "Joachim" == Joachim Pihl <jpihl@nevion.com> writes:
>
> Hi,
>
>  >> [Please keep communication on the list instead of mailing me  
> directly]
>
>  Joachim> Sorry for that, the mailing list is set up with reply-to the  
> original
>  Joachim> author. I habitually do NOT send a reply-to-all, and I didn't  
> catch
>  Joachim> the  error until a few minutes after I sent it.
>
> OK. The reply-to setting is pretty much standard and done on purpose.
>
>  >> Failed in what way? Please provide more details so we can fix the
>  >> problem if there is one.
>
>  Joachim> ...
>  Joachim> flex installed:					Ok
>  Joachim> gettext installed:				Ok
>  Joachim> makeinfo installed:				Ok
>  Joachim> patch installed:				Ok
>  Joachim> Build system dependencies:			Ok
>
>  Joachim> rm -rf
>  Joachim>  
> /home/jpihl/system_controller/buildroot-2009.11_rc1/output/build/buildroot-config
>  Joachim> mkdir -p  
> /home/jpihl/system_controller/buildroot-2009.11_rc1/output/build
>  Joachim> cp -dpRf package/config/buildroot-config
>  Joachim>  
> /home/jpihl/system_controller/buildroot-2009.11_rc1/output/build/buildroot-config
>  Joachim> --2009-11-27 19:38:34--
>  Joachim>  
> http://www.kernel.org/pub//linux/kernel/v2.6//linux-2.6.27.x.tar.bz2
>
> Ahh yes, you have to select 2.6.27.x in the choice box above instead of
> typing the string (Where you have now 'manually selected version')
>

No such setting in xconfig... Only 2.6.30, 2.6.31, snapshot and manual.


-- 
Using Opera's revolutionary e-mail client: http://www.opera.com/mail/

[Buildroot] 2009.11-rc1 build failure

[Peter Korsgaard]

>>>>> "Joachim" == Joachim Pihl <jpihl@nevion.com> writes:

Hi,

 >> Ahh yes, you have to select 2.6.27.x in the choice box above instead of
 >> typing the string (Where you have now 'manually selected version')
 >> 

 Joachim> No such setting in xconfig... Only 2.6.30, 2.6.31, snapshot
 Joachim> and manual

You'll have to enable BR2_DEPRECATED to see such an old kernel
(Build options->Show packages that are deprecated or obsolete)

-- 
Bye, Peter Korsgaard