[Buildroot] make on MAC OS X

[Buildroot] make on MAC OS X

[Fan Zhang]

Hi,

I am new to buildroot and want to build it on MAC OS X. The first try  
is failed. Anybody successfully build it on MAC OS X?

Regards,
Fan
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://busybox.net/lists/buildroot/attachments/20071129/45c7b9ab/attachment.htm 

[Buildroot] uClibc-nptl?

[Ryan Ordway]

Has anyone gotten uClibc-nptl working with buildroot? Particularly on  
the ARM arch.

Thanks!

Ryan

[Buildroot] uClibc-nptl?

[Steven J. Hill]

> Has anyone gotten uClibc-nptl working with buildroot? Particularly on  
> the ARM arch.
> 
The uClibc-NPTL patches for ARM have been posted to the list. They are
not currently in trunk, but will get there in the next couple of weeks.
For now, you could get those patches, and create a tarball of the uClibc
code and call it 'uClibc-snapshot.tar.bz2'. Next, create the directory
'dl' in your buildroot tree and copy that tarball in there. Finally,
select uClibc snapshot from the toolchain configuration menu and you
should be set.

-Steve

[Buildroot] uClibc-nptl?

[Ryan Ordway]

On Nov 29, 2007, at 5:30 AM, Steven J. Hill wrote:

>> Has anyone gotten uClibc-nptl working with buildroot? Particularly on
>> the ARM arch.
>>
> The uClibc-NPTL patches for ARM have been posted to the list. They are
> not currently in trunk, but will get there in the next couple of  
> weeks.
> For now, you could get those patches, and create a tarball of the  
> uClibc
> code and call it 'uClibc-snapshot.tar.bz2'. Next, create the directory
> 'dl' in your buildroot tree and copy that tarball in there. Finally,
> select uClibc snapshot from the toolchain configuration menu and you
> should be set.


Great! Very good news!

Unfortunately I must have missed seeing the patches. All I can find  
are some old patches from CodeSourcery for 0.9.28 that look to be  
against r16550. I've browsed through the mailing list archives and the  
closest thing I can find is your message from 11/12. I'd appreciate a  
clue bat if you have the time.

Thanks,

Ryan

--
Ryan Ordway                           E-mail: rordway at oregonstate.edu
Unix Systems Administrator               rordway at library.oregonstate.edu
OSU Libraries, Corvallis, OR 97331    Office: Valley Library #4657

[Buildroot] Buildroot and Linux/ARM on AT91SAM9260

[Ryan Ordway]

I've been using Buildroot with a board based on the AT91SAM9260. I was  
able to get a successful kernel build (2.6.21.6) and buildroot build  
(buildroot-atmel-20070521) using gcc 4.0.4, binutils 2.17 and uClibc  
0.9.28.3. I've been trying to get an updated build from various  
different versions of buildroot from buildrooot-atmel-20070607,  
buildroot-trunk, etc. and while things build fine with some tweaking  
none of the toolchain combinations seem to generate working binaries.  
I've tried using EABI vs. OABI, softfloat for gcc, NWFPE vs. FastFPE,  
etc. The only combination that I can get to work reliably are gcc  
4.0.4, binutils 2.17 and uClibc 0.9.28.3 with my 2.6.21.6 kernel with  
AT91 patches for 2.6.21 applied.

Is there a magic compiler, binutils and uClibc trinity for Linux/ARM  
in particular?

Thanks!

Ryan

--
Ryan Ordway                           E-mail: rordway at oregonstate.edu
Unix Systems Administrator               rordway at library.oregonstate.edu
OSU Libraries, Corvallis, OR 97331    Office: Valley Library #4657

[Buildroot] Buildroot and Linux/ARM on AT91SAM9260

[Ivan Kuten]

Ryan Ordway wrote:
> I've been using Buildroot with a board based on the AT91SAM9260. I was  
> able to get a successful kernel build (2.6.21.6) and buildroot build  
> (buildroot-atmel-20070521) using gcc 4.0.4, binutils 2.17 and uClibc  
> 0.9.28.3. I've been trying to get an updated build from various  
> different versions of buildroot from buildrooot-atmel-20070607,  
> buildroot-trunk, etc. and while things build fine with some tweaking  
> none of the toolchain combinations seem to generate working binaries.  
> I've tried using EABI vs. OABI, softfloat for gcc, NWFPE vs. FastFPE,  
> etc. The only combination that I can get to work reliably are gcc  
> 4.0.4, binutils 2.17 and uClibc 0.9.28.3 with my 2.6.21.6 kernel with  
> AT91 patches for 2.6.21 applied.
> 
> Is there a magic compiler, binutils and uClibc trinity for Linux/ARM  
> in particular?
> 
> Thanks!
> 
> Ryan
> 

Indeed, I tried svn buildroot with kernel 2.6.22 for Cirrus Logic ARM board
BR2_ARM_TYPE="ARM920T"
BR2_ARM_EABI=y
BR2_SOFT_FLOAT=y
BR2_TARGET_OPTIMIZATION="-Os -pipe"
BR2_BINUTILS_VERSION_2_18=y
BR2_GCC_VERSION_4_2_1=y

rootfs was created but when booting:

/sbin/init: can't Kernel panic - not syncing: Attempted to kill init!
load library 'libgcc_s.so.1'

I checked that there is no libgcc_s.so.1 in rootfs image but I have BR2_GCC_SHARED_LIBGCC=y
May be that is a cause why you not getting working rootfs.


Regards,
Ivan
--------------------------------
Embedded Linux engineer,
Promwad Company: http://www.promwad.com/
Homepage : http://www.ivankuten.com/
--------------------------------

[Buildroot] Buildroot and Linux/ARM on AT91SAM9260

[Ryan Ordway]

On Dec 17, 2007, at 8:14 AM, Ivan Kuten wrote:

> Ryan Ordway wrote:
>> I've been using Buildroot with a board based on the AT91SAM9260. I  
>> was
>> able to get a successful kernel build (2.6.21.6) and buildroot build
>> (buildroot-atmel-20070521) using gcc 4.0.4, binutils 2.17 and uClibc
>> 0.9.28.3. I've been trying to get an updated build from various
>> different versions of buildroot from buildrooot-atmel-20070607,
>> buildroot-trunk, etc. and while things build fine with some tweaking
>> none of the toolchain combinations seem to generate working binaries.
>> I've tried using EABI vs. OABI, softfloat for gcc, NWFPE vs. FastFPE,
>> etc. The only combination that I can get to work reliably are gcc
>> 4.0.4, binutils 2.17 and uClibc 0.9.28.3 with my 2.6.21.6 kernel with
>> AT91 patches for 2.6.21 applied.
>>
>> Is there a magic compiler, binutils and uClibc trinity for Linux/ARM
>> in particular?
>>
>
>
>
> /sbin/init: can't Kernel panic - not syncing: Attempted to kill init!
> load library 'libgcc_s.so.1'
>
> I checked that there is no libgcc_s.so.1 in rootfs image but I have  
> BR2_GCC_SHARED_LIBGCC=y
> May be that is a cause why you not getting working rootfs.


I'm not using a shared libgcc. It's just freezing after the freeing of  
the initrd memory about the time it should be booting up init. I guess  
my next step is to start throwing in some printk's to see exactly  
where it's dying.

Of course, I can't get a 2.6.23 kernel to even boot at all, but that  
may be the same toolchain issues. My 2.6.21.6 kernel seems to be  
working ok -- that was built with my gcc 4.0.4 based toolchain, the  
only version I've been able to get successful boots from.

Ryan

--
Ryan Ordway                           E-mail: rordway at oregonstate.edu
Unix Systems Administrator               rordway at library.oregonstate.edu
OSU Libraries, Corvallis, OR 97331    Office: Valley Library #4657

[Buildroot] Buildroot and Linux/ARM on AT91SAM9260

[Ivan Kuten]

>>
>> /sbin/init: can't Kernel panic - not syncing: Attempted to kill init!
>> load library 'libgcc_s.so.1'
>>
>> I checked that there is no libgcc_s.so.1 in rootfs image but I have
>> BR2_GCC_SHARED_LIBGCC=y
>> May be that is a cause why you not getting working rootfs.
>
>
> I'm not using a shared libgcc. It's just freezing after the freeing of
> the initrd memory about the time it should be booting up init. I guess
> my next step is to start throwing in some printk's to see exactly
> where it's dying.
>
> Of course, I can't get a 2.6.23 kernel to even boot at all, but that
> may be the same toolchain issues. My 2.6.21.6 kernel seems to be
> working ok -- that was built with my gcc 4.0.4 based toolchain, the
> only version I've been able to get successful boots from.
>
> Ryan
>
> --
> Ryan Ordway                           E-mail: rordway at oregonstate.edu
> Unix Systems Administrator               rordway at library.oregonstate.edu
> OSU Libraries, Corvallis, OR 97331    Office: Valley Library #4657
>
>

Ryan,

Yes, now I have the same symptoms
After buildroot rebuild libgcc_s.so.1 placed correctly (btw need to investigate why it was not placed there for the
first time). but when booting ramdisk is mounted and then freeze.
I suggest to try run your rootfs via qemu emulator, adjust only console device.

Folks who using ARM buildroot, please tell us your working configs!

Regards,
Ivan

[Buildroot] Buildroot and Linux/ARM on AT91SAM9260

[Ivan Kuten]

>
>
> Of course, I can't get a 2.6.23 kernel to even boot at all, but that
> may be the same toolchain issues. My 2.6.21.6 kernel seems to be
> working ok

As I understood 2.6.23 is not bootable at all? Is it differs by size from 2.6.21.6 built by
correct toolchain?

I have not bootable 2.6.23 kernel image of 4MB built by svn toolchain and
the same kernel built by toolchain of 1O Oct 2007 is ok and is only 1.2MB

May be 2.18 binutils used in svn buildroot has some issues.

Definitely svn ARM buildroot has some issues. Need to investigate more.

Regards,
Ivan

[Buildroot] Buildroot and Linux/ARM on AT91SAM9260

[Ryan Ordway]

On Dec 17, 2007, at 2:49 PM, Ivan Kuten wrote:

>>
>>
>> Of course, I can't get a 2.6.23 kernel to even boot at all, but that
>> may be the same toolchain issues. My 2.6.21.6 kernel seems to be
>> working ok
>
> As I understood 2.6.23 is not bootable at all? Is it differs by size  
> from 2.6.21.6 built by
> correct toolchain?
>
> I have not bootable 2.6.23 kernel image of 4MB built by svn  
> toolchain and
> the same kernel built by toolchain of 1O Oct 2007 is ok and is only  
> 1.2MB
>
> May be 2.18 binutils used in svn buildroot has some issues.
>
> Definitely svn ARM buildroot has some issues. Need to investigate  
> more.


Correct. My 2.6.21.6 kernel is about 1.1MB. My 2.6.23.9 kernel is >  
4MB. I get the "Uncompressing Linux....." etc. then a few gibberish  
characters and then it freezes. I'm assuming something is wacky about  
that particular toolchain combination. I'll try using binutils 2.17  
and see if that helps.

Ryan


--
Ryan Ordway                           E-mail: rordway at oregonstate.edu
Unix Systems Administrator               rordway at library.oregonstate.edu
OSU Libraries, Corvallis, OR 97331    Office: Valley Library #4657

[Buildroot] Buildroot and Linux/ARM on AT91SAM9260

[Ryan Ordway]

On Dec 17, 2007, at 2:53 PM, Ryan Ordway wrote:

>
> On Dec 17, 2007, at 2:49 PM, Ivan Kuten wrote:
>
>>>
>>>
>>> Of course, I can't get a 2.6.23 kernel to even boot at all, but that
>>> may be the same toolchain issues. My 2.6.21.6 kernel seems to be
>>> working ok
>>
>> As I understood 2.6.23 is not bootable at all? Is it differs by size
>> from 2.6.21.6 built by
>> correct toolchain?
>>
>> I have not bootable 2.6.23 kernel image of 4MB built by svn
>> toolchain and
>> the same kernel built by toolchain of 1O Oct 2007 is ok and is only
>> 1.2MB
>>
>> May be 2.18 binutils used in svn buildroot has some issues.
>>
>> Definitely svn ARM buildroot has some issues. Need to investigate
>> more.
>
>
> Correct. My 2.6.21.6 kernel is about 1.1MB. My 2.6.23.9 kernel is >
> 4MB. I get the "Uncompressing Linux....." etc. then a few gibberish
> characters and then it freezes. I'm assuming something is wacky about
> that particular toolchain combination. I'll try using binutils 2.17
> and see if that helps.


Oh, and the build that I have working thus far is:

svn snapshot from 05/21/2007 + Atmel patches from Ulf
gcc 4.0.4
binutils 2.17
uClibc 0.9.28.3
kernel 2.6.21.6 + atmel patches for 2.6.21
EABI

This has been working great for me since May/June, but I want to start  
working towards uClibc 0.9.29 and/or the most recent snapshots once  
the ARM NPTL stuff hits trunk. I've been having some weird threading  
problems using LinuxThreads.

Thanks,

Ryan

--
Ryan Ordway                           E-mail: rordway at oregonstate.edu
Unix Systems Administrator               rordway at library.oregonstate.edu
OSU Libraries, Corvallis, OR 97331    Office: Valley Library #4657

[Buildroot] Buildroot and Linux/ARM on AT91SAM9260

[Ivan Kuten]

>> Correct. My 2.6.21.6 kernel is about 1.1MB. My 2.6.23.9 kernel is >
>> 4MB. I get the "Uncompressing Linux....." etc. then a few gibberish
>> characters and then it freezes. I'm assuming something is wacky about
>> that particular toolchain combination. I'll try using binutils 2.17
>> and see if that helps.
> Oh, and the build that I have working thus far is:
>
> svn snapshot from 05/21/2007 + Atmel patches from Ulf
> gcc 4.0.4
> binutils 2.17
> uClibc 0.9.28.3
> kernel 2.6.21.6 + atmel patches for 2.6.21
> EABI
>
> This has been working great for me since May/June, but I want to start
> working towards uClibc 0.9.29 and/or the most recent snapshots once
> the ARM NPTL stuff hits trunk. I've been having some weird threading
> problems using LinuxThreads.
>
> Thanks,
>
> Ryan
>

As I see we have 2 problems:
1) binutils 2.18 generate incorrect kernel image at least for 2.6.23 kernel: probably on this step:
arm-linux-objcopy -O binary -R .note -R comment -S vmlinux arch/arm/boot/Image

2) invalid rootfs binaries generated by svn toolchain

Regards,
Ivan

[Buildroot] Buildroot and Linux/ARM on AT91SAM9260

[Ryan Ordway]

On Dec 17, 2007, at 3:36 PM, Ivan Kuten wrote:

>>> Correct. My 2.6.21.6 kernel is about 1.1MB. My 2.6.23.9 kernel is >
>>> 4MB. I get the "Uncompressing Linux....." etc. then a few gibberish
>>> characters and then it freezes. I'm assuming something is wacky  
>>> about
>>> that particular toolchain combination. I'll try using binutils 2.17
>>> and see if that helps.
>> Oh, and the build that I have working thus far is:
>>
>> svn snapshot from 05/21/2007 + Atmel patches from Ulf
>> gcc 4.0.4
>> binutils 2.17
>> uClibc 0.9.28.3
>> kernel 2.6.21.6 + atmel patches for 2.6.21
>> EABI
>>
>> This has been working great for me since May/June, but I want to  
>> start
>> working towards uClibc 0.9.29 and/or the most recent snapshots once
>> the ARM NPTL stuff hits trunk. I've been having some weird threading
>> problems using LinuxThreads.
>>
>> Thanks,
>>
>> Ryan
>>
>
> As I see we have 2 problems:
> 1) binutils 2.18 generate incorrect kernel image at least for 2.6.23  
> kernel: probably on this step:
> arm-linux-objcopy -O binary -R .note -R comment -S vmlinux arch/arm/ 
> boot/Image
>
> 2) invalid rootfs binaries generated by svn toolchain


 From what I've seen, at least for ARM, it's very toolchain dependent.  
I've tried different versions of gcc, binutils and uClibc, with and  
without EABI, using soft float and not. The only combination that has  
worked reliably for me has been the combination I listed above. You  
might try an older snapshot of buildroot from around May and try gcc  
4.0.4. That's the only compiler version I've gotten to work reliably.

I did finally get a build of the latest trunk of buildroot, but it had  
a lot of problems. sshd and other non-busybox applications segfault.  
This was using gcc 4.2.1, binutils 2.18 and a snapshot of uClibc from  
early december, using softfloat and EABI. I've had to focus on some  
other issues so I haven't gotten back to it yet. This was still with  
my 2.6.21.6 kernel I'd built previously, still cannot get a working  
2.6.23 kernel with that toolchain.

Ryan

--
Ryan Ordway                           E-mail: rordway at oregonstate.edu
Unix Systems Administrator               rordway at library.oregonstate.edu
OSU Libraries, Corvallis, OR 97331    Office: Valley Library #4657

[Buildroot] Buildroot and Linux/ARM on AT91SAM9260

[Ivan Kuten]

> 
> 
>  From what I've seen, at least for ARM, it's very toolchain dependent.  
> I've tried different versions of gcc, binutils and uClibc, with and  
> without EABI, using soft float and not. The only combination that has  
> worked reliably for me has been the combination I listed above. You  
> might try an older snapshot of buildroot from around May and try gcc  
> 4.0.4. That's the only compiler version I've gotten to work reliably.
> 
> I did finally get a build of the latest trunk of buildroot, but it had  
> a lot of problems. sshd and other non-busybox applications segfault.  
> This was using gcc 4.2.1, binutils 2.18 and a snapshot of uClibc from  
> early december, using softfloat and EABI. I've had to focus on some  
> other issues so I haven't gotten back to it yet. This was still with  
> my 2.6.21.6 kernel I'd built previously, still cannot get a working  
> 2.6.23 kernel with that toolchain.
> 

This link seems mention the same problem (for 2.6.23):
Subject: binutils 2.17.50.0.17 --build-id feature breaks arm kernel build
http://lists.arm.linux.org.uk/lurker/message/20070828.080711.574bb3f0.en.html

Regards,
Ivan
--------------------------------
Embedded Linux engineer,
Promwad Company: http://www.promwad.com/
Homepage : http://www.ivankuten.com/
--------------------------------

[Buildroot] Buildroot and Linux/ARM on AT91SAM9260

[Ivan Kuten]

>> I did finally get a build of the latest trunk of buildroot, but it had  
>> a lot of problems. sshd and other non-busybox applications segfault.  
>> This was using gcc 4.2.1, binutils 2.18 and a snapshot of uClibc from  
>> early december, using softfloat and EABI. I've had to focus on some  
>> other issues so I haven't gotten back to it yet. This was still with  
>> my 2.6.21.6 kernel I'd built previously, still cannot get a working  
>> 2.6.23 kernel with that toolchain.
>>
> 

This commit:
http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=1e621a8e3752367d4aae78a8ab00a18fb2793f34

fixes issue with build-id section introduced in 2.6.23

Hope you will finally have working 2.6.23 kernel.

Regards,
Ivan
-- 

--------------------------------
Embedded Linux engineer,
Promwad Company: http://www.promwad.com/
Homepage : http://www.ivankuten.com/
--------------------------------

[Buildroot] Buildroot and Linux/ARM on AT91SAM9260

[Wayne Keenan]

Hi,

I am runing into a similarissue, there would be no  outpu after :
Uncompressing
Linux.........................................................................................................................................T


o I applied the patch mentioned
(http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=1e621a8e3752367d4aae78a8ab00a18fb2793f34)<http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=1e621a8e3752367d4aae78a8ab00a18fb2793f34>

Now it gets futher but outputs gibberish, after:

... done, booting the kernel.   <gibberish>

is there anything else that needs to be applied?

My Environment:

host: ubuntu 7.04   2.6.20-16-generic
target: AT91 SAM 9263
buildroot: r18347 | aldot | 2007-04-06 12:55:37 +0100 (Fri, 06 Apr 2007)
gcc: 3.4.6
uClibc:0.9.28.3
binutils: 2.1.7
kernel: 2.6.20.4
busybox: 1.5.0

Thanks,
Wayn

On Dec 27, 2007 12:42 PM, Ivan Kuten <ivan.kuten@promwad.com> wrote:

> >> I did finally get a build of the latest trunk of buildroot, but it had
> >> a lot of problems. sshd and other non-busybox applications segfault.
> >> This was using gcc 4.2.1, binutils 2.18 and a snapshot of uClibc from
> >> early december, using softfloat and EABI. I've had to focus on some
> >> other issues so I haven't gotten back to it yet. This was still with
> >> my 2.6.21.6 kernel I'd built previously, still cannot get a working
> >> 2.6.23 kernel with that toolchain.
> >>
> >
>
> This commit:
>
> http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=1e621a8e3752367d4aae78a8ab00a18fb2793f34
>
> fixes issue with build-id section introduced in 2.6.23
>
> Hope you will finally have working 2.6.23 kernel.
>
> Regards,
> Ivan
> --
>
> --------------------------------
> Embedded Linux engineer,
> Promwad Company: http://www.promwad.com/
> Homepage : http://www.ivankuten.com/
> --------------------------------
> _______________________________________________
> buildroot mailing list
> buildroot at uclibc.org
> http://busybox.net/mailman/listinfo/buildroot
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://busybox.net/lists/buildroot/attachments/20080112/489dde2a/attachment.htm 

[Buildroot] Buildroot and Linux/ARM on AT91SAM9260

[Wayne Keenan]

Sorry, that should have read

kernel:  2.6.23

Wayne

On Jan 12, 2008 4:42 PM, Wayne Keenan <wayne.keenan@gmail.com> wrote:

> Hi,
>
> I am runing into a similarissue, there would be no  outpu after :
> Uncompressing
> Linux.........................................................................................................................................T
>
>
>
> o I applied the patch mentioned (http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=1e621a8e3752367d4aae78a8ab00a18fb2793f34)<http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=1e621a8e3752367d4aae78a8ab00a18fb2793f34>
>
> Now it gets futher but outputs gibberish, after:
>
> ... done, booting the kernel.   <gibberish>
>
> is there anything else that needs to be applied?
>
> My Environment:
>
> host: ubuntu 7.04   2.6.20-16-generic
> target: AT91 SAM 9263
> buildroot: r18347 | aldot | 2007-04-06 12:55:37 +0100 (Fri, 06 Apr 2007)
> gcc: 3.4.6
> uClibc:0.9.28.3
> binutils: 2.1.7
> kernel: 2.6.20.4
> busybox: 1.5.0
>
> Thanks,
> Wayn
>
>
> On Dec 27, 2007 12:42 PM, Ivan Kuten <ivan.kuten@promwad.com> wrote:
>
> > >> I did finally get a build of the latest trunk of buildroot, but it
> > had
> > >> a lot of problems. sshd and other non-busybox applications segfault.
> > >> This was using gcc 4.2.1, binutils 2.18 and a snapshot of uClibc from
> > >> early december, using softfloat and EABI. I've had to focus on some
> > >> other issues so I haven't gotten back to it yet. This was still with
> > >> my 2.6.21.6 kernel I'd built previously, still cannot get a working
> > >> 2.6.23 kernel with that toolchain.
> > >>
> > >
> >
> > This commit:
> >
> > http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=1e621a8e3752367d4aae78a8ab00a18fb2793f34
> >
> > fixes issue with build-id section introduced in 2.6.23
> >
> > Hope you will finally have working 2.6.23 kernel.
> >
> > Regards,
> > Ivan
> > --
> >
> > --------------------------------
> > Embedded Linux engineer,
> > Promwad Company: http://www.promwad.com/
> > Homepage : http://www.ivankuten.com/
> > --------------------------------
> > _______________________________________________
> > buildroot mailing list
> > buildroot at uclibc.org
> > http://busybox.net/mailman/listinfo/buildroot
> >
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://busybox.net/lists/buildroot/attachments/20080112/8be73c4c/attachment-0001.htm 

[Buildroot] Buildroot and Linux/ARM on AT91SAM9260

[Wayne Keenan]

I have tried a couple of combinations and a pre-built binary, my
environments have been:

1.

buildroot: r20692 | thomasez | 2007-12-27 12:00:12 +0000 (Thu, 27 Dec 2007)
gcc: 4.2.1
uClibc:2.28.3
binutils: 2.1.8
busybox: 1.7.2


2.
I then tried with binutils 2.1.7 - the result was a larger quantity of
gibberish  than before was output.

3.
I then tried a pre-built binary from
ftp://www.linux4sam.org/pub/linux/linux-2.6.22-at91sam9263ek.bin

and the pre-built binary output even more that 2.1.8.

so, I have managed to go from no gibber to increasingly amounts of gibber,
Im hoping someone can help me get sanity back again please?

This is a Calao systems USB9263 device, and I'm wondering if I need to
perform some kind of extra initialisation.

Thanks
Wayne





On Jan 12, 2008 4:43 PM, Wayne Keenan <wayne.keenan@gmail.com> wrote:

> Sorry, that should have read
>
> kernel:  2.6.23
>
> Wayne
>
>
> On Jan 12, 2008 4:42 PM, Wayne Keenan < wayne.keenan@gmail.com> wrote:
>
> > Hi,
> >
> > I am runing into a similarissue, there would be no  outpu after :
> > Uncompressing
> > Linux.........................................................................................................................................T
> >
> >
> >
> > o I applied the patch mentioned (http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=1e621a8e3752367d4aae78a8ab00a18fb2793f34)<http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=1e621a8e3752367d4aae78a8ab00a18fb2793f34>
> >
> > Now it gets futher but outputs gibberish, after:
> >
> > ... done, booting the kernel.   <gibberish>
> >
> > is there anything else that needs to be applied?
> >
> > My Environment:
> >
> > host: ubuntu 7.04   2.6.20-16-generic
> > target: AT91 SAM 9263
> > buildroot: r18347 | aldot | 2007-04-06 12:55:37 +0100 (Fri, 06 Apr 2007)
> >
> > gcc: 3.4.6
> > uClibc:0.9.28.3
> > binutils: 2.1.7
> > kernel: 2.6.20.4
> > busybox: 1.5.0
> >
> > Thanks,
> > Wayn
> >
> >
> > On Dec 27, 2007 12:42 PM, Ivan Kuten <ivan.kuten@promwad.com> wrote:
> >
> > > >> I did finally get a build of the latest trunk of buildroot, but it
> > > had
> > > >> a lot of problems. sshd and other non-busybox applications
> > > segfault.
> > > >> This was using gcc 4.2.1, binutils 2.18 and a snapshot of uClibc
> > > from
> > > >> early december, using softfloat and EABI. I've had to focus on some
> > > >> other issues so I haven't gotten back to it yet. This was still
> > > with
> > > >> my 2.6.21.6 kernel I'd built previously, still cannot get a working
> > > >> 2.6.23 kernel with that toolchain.
> > > >>
> > > >
> > >
> > > This commit:
> > >
> > > http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=1e621a8e3752367d4aae78a8ab00a18fb2793f34
> > >
> > > fixes issue with build-id section introduced in 2.6.23
> > >
> > > Hope you will finally have working 2.6.23 kernel.
> > >
> > > Regards,
> > > Ivan
> > > --
> > >
> > > --------------------------------
> > > Embedded Linux engineer,
> > > Promwad Company: http://www.promwad.com/
> > > Homepage : http://www.ivankuten.com/
> > > --------------------------------
> > > _______________________________________________
> > > buildroot mailing list
> > > buildroot at uclibc.org
> > > http://busybox.net/mailman/listinfo/buildroot
> > >
> >
> >
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://busybox.net/lists/buildroot/attachments/20080112/dc3772d6/attachment.htm