| 1 |
On Mon, 2005-06-06 at 10:03 -0500, Heath H Holcomb wrote: |
| 2 |
> Version 0.06 |
| 3 |
> Please add, delete, modify. Thanks! |
| 4 |
> This procedure produces a working embedded rootfs, that boots form a hard |
| 5 |
> drive on a Geode based SBC (single board computer); at least on my |
| 6 |
> development system. |
| 7 |
|
| 8 |
|
| 9 |
> #----- Step 2 - Build the development_rootfs --------------------------------- |
| 10 |
|
| 11 |
|
| 12 |
> rm /etc/make.profile |
| 13 |
> ln -s /usr/portage/profiles/uclibc/x86 /etc/make.profile |
| 14 |
|
| 15 |
# This should be. |
| 16 |
cd /etc |
| 17 |
rm make.profile |
| 18 |
ln -s ../usr/portage/profiles/uclibc/x86 make.profile |
| 19 |
|
| 20 |
# Note the ../ vs / ; This is how portage expects things to be for the |
| 21 |
cascading profile setup. |
| 22 |
If you do it the other wrong way you encounter strange bugs here and |
| 23 |
there. |
| 24 |
|
| 25 |
> # Workaround 2 |
| 26 |
> # During emerge -e system, python-fchksum failes complaing about |
| 27 |
> # gcc-config error: Could not run/locate "i386-gentoo-linux-uclibc-gcc" |
| 28 |
> # The following commands work around this problem. |
| 29 |
> emerge python |
| 30 |
|
| 31 |
|
| 32 |
> emerge -e system |
| 33 |
You already built python. So you know need to rebuild python-fscksum vs all of '-e system' |
| 34 |
That bootstrap+system+system would end up having the user rebuild |
| 35 |
gcc/uclibc etc about 3 times total. |
| 36 |
|
| 37 |
|
| 38 |
-- |
| 39 |
Ned Ludd <solar@g.o> |
| 40 |
|
| 41 |
-- |
| 42 |
gentoo-embedded@g.o mailing list |
Archives