1 |
On Sat, 2004-07-10 at 02:58, Travis Tilley wrote: |
2 |
> just a heads up - i've moved all the 2.6 header packages to |
3 |
> sys-kernel/linux26-headers so that they can be keyworded without having them |
4 |
> accidentally install on the system of a 2.4 kernel user. hopefully soon this |
5 |
> will mean we can unmask the newer glibc ebuilds on x86, which have a dep on |
6 |
> the 2.6 headers when using nptl (instead of using /usr/src/linux, which is |
7 |
> problematic at best and quite often broken at worst). |
8 |
> |
9 |
|
10 |
I and Solar is working on getting bootstrap-2.6.sh (or maby an |
11 |
merged version) ready to work seemlessly with linux26-headers. |
12 |
Hopefully in the next two weeks if testing are going as planned. |
13 |
|
14 |
> another note re: NPTL is that nptl no longer has a version number of it's own |
15 |
> as of 2004-06-29. i'm guessing this means that ulrich drepper considers nptl |
16 |
> to be fairly stable, but i dont think i'll be so quick to put words in his |
17 |
> mouth. when i'm sure that everything's good and calm with the linux26-headers |
18 |
> change i'll poke at making a new snapshot to see if it passes "make check" |
19 |
> yet. :) |
20 |
|
21 |
I do not know about non-x86 arches, but I make sure 'make check' pass |
22 |
for NPTL as well as run it for a day or two before commit. |
23 |
|
24 |
I think the biggest problem with getting it into stable previously |
25 |
(and for the next few weeks), is the headers issue - we could not |
26 |
just unmask 2.6 headers. Hopefully if 2.6.6-r1 or latest 2.6.7 |
27 |
linux26-headers test ok, and we get headway with bootstrap.sh, we |
28 |
can really think about 'offical' NPTL support ... |
29 |
|
30 |
|
31 |
Cheers, |
32 |
|
33 |
-- |
34 |
|
35 |
Martin Schlemmer |
36 |
Gentoo Linux Developer, Desktop/System Team Developer |
37 |
Cape Town, South Africa |