[Buildroot] [GIT PULL] avr32 updates

Thiago A. Corrêa thiago.correa at gmail.com
Fri May 8 10:56:25 UTC 2009


On Fri, May 8, 2009 at 6:48 AM, Peter Korsgaard <jacmet at uclibc.org> wrote:
>>>>>> "Thiago" == Thiago A Corrêa <thiago.correa at gmail.com> writes:
>  Thiago> Ok, after an hour of wrestling with git and it's man pages, I
>  Thiago> think I got it right :) I was even able to split the kernel
>  Thiago> update commit.
>
> Heh, was git rebase -i that bad?

Just adding the signoff was easy... commit --amend -s did the trick,
the problem was me trying to split the other commit. I somehow managed
to completely trash my local tree and had to clone it again from the
remote. I really miss svn revert... If I remove some file from the
local tree, I can't bring it back with git fetch, git checkout or git
reset.

>  >>  Thiago> Something is wrong in the kernel headers in previous versions.
>  >>  Thiago> Something in byteorder.h prevended it from building earlier.
>  >>
>  >> But the Linux.advanced stuff doesn't have anything to do with kernel
>  >> headers, right?
>
>  Thiago> Unfortunally, it doesn't look like it. Even if I select
>  Thiago> 2.6.29 for the linux headers (toolchain), the build
>  Thiago> fails. Unless I actually update the kernel on the target, it
>  Thiago> fails to build for avr32. According to the avr32-linux
>  Thiago> mailling list, there was a byteorder.h restructuring a while
>  Thiago> back, so I guess some versions of the headers are not what
>  Thiago> iptables expects.
>
> I still don't get it - I've never used the Linux.advanced stuff, but
> it's just used to build a kernel for the target and doesn't have
> anything to do with the rootfs, right? How can it then influence the
> iptables build?

Perhaps it's how it touches all the LINUX26* variables. I have no idea
either. I just tried to update and it worked. I might just be loosing
my mind and tested it wrong, could have forgot to run a toolchain with
2.6.29 and kernel 2.6.28 test and thought that I had. I've been awake
for 36 hours already and still have to study for German test tomorrow
(as second foreign language).

I will leave this building with toolchain kernel headers 2.26.9 and
advanced linux 2.26.7 and see if iptables fails to compile again.

>  >> Ok. git am is easy - Simply save the mail and run git am -s <file>
>
>  Thiago> A bit easier said than done *smile*.
>  Thiago> I do all of my buildroot development over ssh.
>
> And? Do you mean that your desktop machine is running Windows?

Yes, my gentoo build environment runs inside vmware in a quad core.

>  Thiago> Anyway, what happens if I apply a patch and then you apply it
>  Thiago> a while later before pulling from my tree? I guess we will
>  Thiago> conflict right?
>
> No, if none of us modify the patch then the git sha1 id will be the
> same and the commit just ignored (if one of us modifies it you will
> most likely get a merge conflict that you'll have to resolve when you
> rebase your changes against me).

Nice. Then I don't need to skip all patches :)

>  Thiago> my git pull-request: (thanks for the tip)
>
> pulled and pushed.
>

Thanks.


More information about the buildroot mailing list