1 |
On Thursday 04 January 2007 23:37, Dale wrote: |
2 |
> Peter Alfredsen wrote: |
3 |
> > On Thursday 04 January 2007 19:39, Dale wrote: |
4 |
> >> Anybody know how to fix this? |
5 |
> > |
6 |
> > Yeah, -r4 fails for me too. -r5 compiles beautifully, though. |
7 |
> > Just add this line to /etc/portage/package.keywords: |
8 |
> > =sys-apps/util-linux-2.12r-r5 ~x86 |
9 |
> > |
10 |
> > And re-emerge it. |
11 |
> > |
12 |
> > HTH |
13 |
> |
14 |
> Oh, may try that. I'm using a script that I found on the forums to do |
15 |
> this instead of emerge -ev world. It is something I found when updating |
16 |
> gcc a good while back. According to some emerge doesn't do them in the |
17 |
> correct order but this script does. It starts out with the kernel so it |
18 |
> is different. |
19 |
|
20 |
Yes, I've read about that. Looks nice. There should be file where you can |
21 |
alter stuff like this, IIRC. |
22 |
|
23 |
> Now to go figure out how to get this script to emerge that version |
24 |
> without starting over. ;) |
25 |
> |
26 |
> Now to figure out what the deal was on my CFLAGS. |
27 |
|
28 |
I think your CFLAGS are probably fine. I have sane (march, pipe, |
29 |
fomitframepointer) CFLAGS and the problem is still there. If they've worked |
30 |
for you before with no problems, they will probably work for you again. |
31 |
|
32 |
I think your problem probably is that you're trying to build a system based |
33 |
off a 2.6.19 kernel/headers while remaining in stable for the rest of the |
34 |
tree. I guess that this is the case because there is only one significant |
35 |
difference between -r4 and -r5 of util-linux: |
36 |
> 31 Oct 2006; Mike Frysinger <vapier@g.o> |
37 |
> +files/util-linux-2.12r-no-_syscall.patch, util-linux-2.12r-r5.ebuild: |
38 |
> Fix building with linux-headers-2.6.18+ as _syscall#() no longer exists |
39 |
> #150852 by Mario Fetka. |
40 |
This will not work without a lot of unmasking. Only ~x86 is fully patched for |
41 |
2.6.19. |
42 |
|
43 |
-- |
44 |
/PA |
45 |
-- |
46 |
gentoo-user@g.o mailing list |