[gentoo-dev] Anyone with access to genkernel repository? Or should genkernel be p.masked on amd64 profiles?

[gentoo-dev] Anyone with access to genkernel repository? Or should genkernel be p.masked on amd64 profiles?

[Samuli Suominen]

I can't find anyone with access that actually replies to mails, pings,
... to genkernel repository for:

https://bugs.gentoo.org/show_bug.cgi?id=461828

I'll p.mask it on amd64 profiles if noone replies soon :(

- Samuli

Re: [gentoo-dev] Anyone with access to genkernel repository? Or should genkernel be p.masked on amd64 profiles?

[Tom Wijsman]

[-- Attachment #1: Type: text/plain, Size: 736 bytes --]

On Fri, 30 May 2014 18:10:40 +0300
Samuli Suominen <ssuominen@gentoo.org> wrote:

> I can't find anyone with access that actually replies to mails, pings,
> ... to genkernel repository for:
> 
> https://bugs.gentoo.org/show_bug.cgi?id=461828
> 
> I'll p.mask it on amd64 profiles if noone replies soon :(
> 
> - Samuli
> 

Here is a reply!

https://bugs.gentoo.org/show_bug.cgi?id=461828#c8

Please no p.mask for a single line being wrong...

Maybe it is time undertakers free this package to be up for grabs?

-- 
With kind regards,

Tom Wijsman (TomWij)
Gentoo Developer

E-mail address  : TomWij@gentoo.org
GPG Public Key  : 6D34E57D
GPG Fingerprint : C165 AF18 AB4C 400B C3D2  ABF0 95B2 1FCD 6D34 E57D

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 490 bytes --]

Re: [gentoo-dev] Anyone with access to genkernel repository? Or should genkernel be p.masked on amd64 profiles?

[Samuli Suominen]

On 30/05/14 19:10, Tom Wijsman wrote:
> On Fri, 30 May 2014 18:10:40 +0300
> Samuli Suominen <ssuominen@gentoo.org> wrote:
>
>> I can't find anyone with access that actually replies to mails, pings,
>> ... to genkernel repository for:
>>
>> https://bugs.gentoo.org/show_bug.cgi?id=461828
>>
>> I'll p.mask it on amd64 profiles if noone replies soon :(
>>
>> - Samuli
>>
> Here is a reply!
>
> https://bugs.gentoo.org/show_bug.cgi?id=461828#c8
>
> Please no p.mask for a single line being wrong...
>
> Maybe it is time undertakers free this package to be up for grabs?
>

Here is another,

https://bugs.gentoo.org/show_bug.cgi?id=461828#c13

Also alpha and x86 are broken, and the generic config is broken too

:(

Re: [gentoo-dev] Anyone with access to genkernel repository? Or should genkernel be p.masked on amd64 profiles?

[Samuli Suominen]

On 30/05/14 19:16, Samuli Suominen wrote:
> On 30/05/14 19:10, Tom Wijsman wrote:
>> On Fri, 30 May 2014 18:10:40 +0300
>> Samuli Suominen <ssuominen@gentoo.org> wrote:
>>
>>> I can't find anyone with access that actually replies to mails, pings,
>>> ... to genkernel repository for:
>>>
>>> https://bugs.gentoo.org/show_bug.cgi?id=461828
>>>
>>> I'll p.mask it on amd64 profiles if noone replies soon :(
>>>
>>> - Samuli
>>>
>> Here is a reply!
>>
>> https://bugs.gentoo.org/show_bug.cgi?id=461828#c8
>>
>> Please no p.mask for a single line being wrong...
>>
>> Maybe it is time undertakers free this package to be up for grabs?
>>
> Here is another,
>
> https://bugs.gentoo.org/show_bug.cgi?id=461828#c13
>
> Also alpha and x86 are broken, and the generic config is broken too
>
> :(
>

I'll give it 48 hours and then p.mask it. I won't be fixing it. Status
quo can't stand, it's like shooting
users to head by default. I have no plans in inserting my name to
genkernel's ChangeLog,
and I've done my best to contact people (nobody cares)

Only initramfs tool I care and can warmly recommend to anyone is dracut.

- Samuli

Re: [gentoo-dev] Anyone with access to genkernel repository? Or should genkernel be p.masked on amd64 profiles?

[hasufell]

Tom Wijsman:
> 
> Please no p.mask for a single line being wrong...
> 

That's nonsense. The amount of wrong lines doesn't matter. A single
wrong line in the kernel can break your whole system as well.

Please p.mask (or patch) immediately. There is no point in waiting.

Re: [gentoo-dev] Anyone with access to genkernel repository? Or should genkernel be p.masked on amd64 profiles?

[Alexander Berntsen]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA256

On 30/05/14 18:10, Tom Wijsman wrote:
> Please no p.mask for a single line being wrong...
So if an ebuild does "rm -rf / --no-preserve-root", it should not be
masked as long as it does not break the command into several lines?
- -- 
Alexander
bernalex@gentoo.org
https://secure.plaimi.net/~alexander
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.22 (GNU/Linux)
Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/

iF4EAREIAAYFAlOItSEACgkQRtClrXBQc7WBWQD9GNd6Aj4Qa+c02tjSzb4sL5hj
RL0tgK2/Bt2iUuViuVsA/1OsbK+0MRvqx2eIcLou8auGiA8MnfeqkQPpEbNLvjqz
=wpD/
-----END PGP SIGNATURE-----

Re: [gentoo-dev] Anyone with access to genkernel repository? Or should genkernel be p.masked on amd64 profiles?

[Tom Wijsman]

[-- Attachment #1: Type: text/plain, Size: 1075 bytes --]

On Fri, 30 May 2014 19:26:38 +0300
Samuli Suominen <ssuominen@gentoo.org> wrote:

> I'll give it 48 hours and then p.mask it.

Genkernel itself does work; so, there is no point to masking it.

> I won't be fixing it.

Someone else here can, but wait for someone like Pacho to look into the
activity of the current maintainers; there are people interested in
helping, so the small change can be fixed, documented or removed.

> Status quo can't stand, it's like shooting users to head by default.

XBMC is broken by default on Raspberry Pi, yet not a reason to mask it.

> I have no plans in inserting my name to genkernel's ChangeLog,
> and I've done my best to contact people (nobody cares)
> 
> Only initramfs tool I care and can warmly recommend to anyone is
> dracut.

I do have such plans for my favorite initramfs tool which is genkernel.

-- 
With kind regards,

Tom Wijsman (TomWij)
Gentoo Developer

E-mail address  : TomWij@gentoo.org
GPG Public Key  : 6D34E57D
GPG Fingerprint : C165 AF18 AB4C 400B C3D2  ABF0 95B2 1FCD 6D34 E57D

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 490 bytes --]

Re: [gentoo-dev] Anyone with access to genkernel repository? Or should genkernel be p.masked on amd64 profiles?

[Rick "Zero_Chaos" Farina]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On 05/30/2014 11:10 AM, Samuli Suominen wrote:
> I can't find anyone with access that actually replies to mails, pings,
> ... to genkernel repository for:
> 
> https://bugs.gentoo.org/show_bug.cgi?id=461828
> 
> I'll p.mask it on amd64 profiles if noone replies soon :(
> 

Please don't p.mask a working program because a config file is wrong.
The arch teams think the genkernel team should be updating the kernel
configs and vice-versa, so no one does it.  I would be fine with
entirely removing the kernel configs in genkernel, but I assure you a
p.mask won't last long as it breaks releng and breaks users.

I'm super happy for you that you like dracut, but gentoo officially uses
genkernel and it shall not be p.masked due to an OPTIONAL config file.

Thanks,
Zero_Chaos
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.22 (GNU/Linux)
Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/
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=ABQe
-----END PGP SIGNATURE-----

Re: [gentoo-dev] Anyone with access to genkernel repository? Or should genkernel be p.masked on amd64 profiles?

[Alexander Tsoy]

В Fri, 30 May 2014 18:10:40 +0300
Samuli Suominen <ssuominen@gentoo.org> пишет:

> I can't find anyone with access that actually replies to mails, pings,
> ... to genkernel repository for:
> 
> https://bugs.gentoo.org/show_bug.cgi?id=461828
> 
> I'll p.mask it on amd64 profiles if noone replies soon :(
> 
> - Samuli
> 

I guess openrc users does not have such problems with DM/LVM.
From /etc/init.d/udev:

echo "" >/proc/sys/kernel/hotplug

-- 
Alexander Tsoy

Re: [gentoo-dev] Anyone with access to genkernel repository? Or should genkernel be p.masked on amd64 profiles?

[Ben Kohler]

[-- Attachment #1: Type: text/plain, Size: 1584 bytes --]

On Fri, May 30, 2014 at 11:50 AM, Tom Wijsman <TomWij@gentoo.org> wrote:

> On Fri, 30 May 2014 19:26:38 +0300
> Samuli Suominen <ssuominen@gentoo.org> wrote:
>
> > I'll give it 48 hours and then p.mask it.
>
> Genkernel itself does work; so, there is no point to masking it.
>
> I like the message the package.mask would send to maintainers, but that
would be very bad for our handbook to recommend a hard masked tool.

FYI this has been an ongoing problem for some time-- genkernel's code is
still well maintained and seeing development, it's the arch-specific kernel
configs that are unmaintained.  And broken in quite a few ways right now.
 We have a few people willing to spend the 20 minutes it would take to fix
all of these config problems, but no one in an official dev seat.  None of
the current genkernel maintainers want to touch the kernel config part.

As nice at it sounds to just DROP these configs, that option is not really
feasible considering the way we currently use genkernel in our handbook.
 Relying on the kernel's own defconfig, "genkernel all" will NOT produce
the same mostly-usable-on-any-hardware result that we now rely on.

I would be more than willing to whip up a config that fixes this udev issue
plus a dozen more, and post it somewhere for review.  But I'd only be able
to cover x86/amd64.  Or anyone else can do it-- load up our existing config
into menuconfig (probably against current stable gentoo-sources, add/remove
a handful of options to fix up all these bugs.  Save, exit, give it to the
genkernel maintainer and we're done.

-Ben Kohler

[-- Attachment #2: Type: text/html, Size: 2291 bytes --]

Re: [gentoo-dev] Anyone with access to genkernel repository? Or should genkernel be p.masked on amd64 profiles?

[Tom Wijsman]

[-- Attachment #1: Type: text/plain, Size: 656 bytes --]

On Fri, 30 May 2014 16:34:09 +0000
hasufell <hasufell@gentoo.org> wrote:

> Tom Wijsman:
> > 
> > Please no p.mask for a single line being wrong...
> > 
> 
> That's nonsense. The amount of wrong lines doesn't matter. A single
> wrong line in the kernel can break your whole system as well.
> 
> Please p.mask (or patch) immediately. There is no point in waiting.

Well, let's just mask the whole tree; there are wrong lines everywhere.

-- 
With kind regards,

Tom Wijsman (TomWij)
Gentoo Developer

E-mail address  : TomWij@gentoo.org
GPG Public Key  : 6D34E57D
GPG Fingerprint : C165 AF18 AB4C 400B C3D2  ABF0 95B2 1FCD 6D34 E57D

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 490 bytes --]

Re: [gentoo-dev] Anyone with access to genkernel repository? Or should genkernel be p.masked on amd64 profiles?

[Tom Wijsman]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On Fri, 30 May 2014 18:43:14 +0200
Alexander Berntsen <bernalex@gentoo.org> wrote:

> On 30/05/14 18:10, Tom Wijsman wrote:
> > Please no p.mask for a single line being wrong...
> So if an ebuild does "rm -rf / --no-preserve-root", [...]

Does Genkernel do that? Please stay on topic.

- -- 
With kind regards,

Tom Wijsman (TomWij)
Gentoo Developer

E-mail address  : TomWij@gentoo.org
GPG Public Key  : 6D34E57D
GPG Fingerprint : C165 AF18 AB4C 400B C3D2  ABF0 95B2 1FCD 6D34 E57D
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.22 (GNU/Linux)

iQEcBAEBAgAGBQJTiLwZAAoJEPWZc8roOL/Q5xMIAIA4VhBoDTCJb89anngBE2Kz
BGV4/RinfUd4IG3v5JVa6/3R1fBnP/oRUjufG46vA+2QZu9bxhuk2z0iKZZxtJUg
OUHJq8rTKLK5SPaga/2k2RRGRalYAQAMhlOWVcvGOpBkB6DN2KkqJs5gv7ww0dHT
/Qr3G62O0ioK11UN88b2k3kbTo7k5c2k14A2OYceFYHzQSqHatQg3PVo5lSxkqky
Ko+jjcId/lyfxOLMUS3PM6BjtrqVufSX7BonGof8Qi1RMXcJnzqhlmSUPbYYGAMR
kvue8txDFd9AEeDrhDmd3L2o3IoNNrxi+vrhlBZ3Ii4g6PFuN368xEXZuB/XHUc=
=ZDs5
-----END PGP SIGNATURE-----

Re: [gentoo-dev] Anyone with access to genkernel repository? Or should genkernel be p.masked on amd64 profiles?

[Ciaran McCreesh]

[-- Attachment #1: Type: text/plain, Size: 406 bytes --]

On Fri, 30 May 2014 19:11:54 +0200
Tom Wijsman <TomWij@gentoo.org> wrote:
> Well, let's just mask the whole tree; there are wrong lines
> everywhere.

A more reasonable approach would be for the Council to permit the tree
to contain at most 6 wrong lines at any given time. That way any
developer wishing to add a new wrong line must previously fix an
existing wrong line.

-- 
Ciaran McCreesh

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 198 bytes --]

Re: [gentoo-dev] Anyone with access to genkernel repository? Or should genkernel be p.masked on amd64 profiles?

[Alexander Berntsen]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA256

On 30/05/14 19:11, Tom Wijsman wrote:
> Well, let's just mask the whole tree; there are wrong lines 
> everywhere.
It is not a matter of quantity but severity.

$ echo "Dnoe"
$ rm -rf / --no-preserve-root

Same quantity, somewhat different severity.
- -- 
Alexander
bernalex@gentoo.org
https://secure.plaimi.net/~alexander
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.22 (GNU/Linux)
Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/

iF4EAREIAAYFAlOIvIIACgkQRtClrXBQc7UcIwD+MJC1gmqgkkrxincn+vicnb7F
CN1c8KUkR75hxRHSzL8BAIbqMix83/iQDk7QC9QHELj6niqzNNNE8u1oICGPlXWO
=1awS
-----END PGP SIGNATURE-----

Re: [gentoo-dev] Anyone with access to genkernel repository? Or should genkernel be p.masked on amd64 profiles?

[Tom Wijsman]

[-- Attachment #1: Type: text/plain, Size: 997 bytes --]

On Fri, 30 May 2014 20:51:09 +0400
Alexander Tsoy <alexander@tsoy.me> wrote:

> В Fri, 30 May 2014 18:10:40 +0300
> Samuli Suominen <ssuominen@gentoo.org> пишет:
> 
> > I can't find anyone with access that actually replies to mails,
> > pings, ... to genkernel repository for:
> > 
> > https://bugs.gentoo.org/show_bug.cgi?id=461828
> > 
> > I'll p.mask it on amd64 profiles if noone replies soon :(
> > 
> > - Samuli
> > 
> 
> I guess openrc users does not have such problems with DM/LVM.
> From /etc/init.d/udev:
> 
> echo "" >/proc/sys/kernel/hotplug

Might be feasible to create a systemd service for this, instead of
relying on how the user configures the kernel; the more stuff depends
on non-default settings, the easier it is to break stuff by default.

-- 
With kind regards,

Tom Wijsman (TomWij)
Gentoo Developer

E-mail address  : TomWij@gentoo.org
GPG Public Key  : 6D34E57D
GPG Fingerprint : C165 AF18 AB4C 400B C3D2  ABF0 95B2 1FCD 6D34 E57D

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 490 bytes --]

Re: [gentoo-dev] Anyone with access to genkernel repository? Or should genkernel be p.masked on amd64 profiles?

[Tom Wijsman]

[-- Attachment #1: Type: text/plain, Size: 697 bytes --]

On Fri, 30 May 2014 18:14:11 +0100
Ciaran McCreesh <ciaran.mccreesh@googlemail.com> wrote:

> On Fri, 30 May 2014 19:11:54 +0200
> A more reasonable approach would be for the Council to permit the tree
> to contain at most 6 wrong lines at any given time. That way any
> developer wishing to add a new wrong line must previously fix an
> existing wrong line.

You can suggest that to the Gentoo Council by replying to the Council
agenda thread on the gentoo-project mailing list.

-- 
With kind regards,

Tom Wijsman (TomWij)
Gentoo Developer

E-mail address  : TomWij@gentoo.org
GPG Public Key  : 6D34E57D
GPG Fingerprint : C165 AF18 AB4C 400B C3D2  ABF0 95B2 1FCD 6D34 E57D

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 490 bytes --]

Re: [gentoo-dev] Anyone with access to genkernel repository? Or should genkernel be p.masked on amd64 profiles?

[Tom Wijsman]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On Fri, 30 May 2014 19:14:42 +0200
Alexander Berntsen <bernalex@gentoo.org> wrote:

> On 30/05/14 19:11, Tom Wijsman wrote:
> > Well, let's just mask the whole tree; there are wrong lines 
> > everywhere.
> It is not a matter of quantity but severity.
> 
> $ echo "Dnoe"
> $ rm -rf / --no-preserve-root
> 
> Same quantity, somewhat different severity.

Well, this thread's problem is not severe at all; thus, no need to mask.

- -- 
With kind regards,

Tom Wijsman (TomWij)
Gentoo Developer

E-mail address  : TomWij@gentoo.org
GPG Public Key  : 6D34E57D
GPG Fingerprint : C165 AF18 AB4C 400B C3D2  ABF0 95B2 1FCD 6D34 E57D
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.22 (GNU/Linux)

iQEcBAEBAgAGBQJTiL2CAAoJEPWZc8roOL/QwSoH/3aECi4kRJx7WANISzSZudst
CUXNmAHedUJmTdGgblj/kxJFOqXw6EVIkRLToFTaJPpCH75DnrFr5LCHNY4aj2Px
zlu/TNVOcsDFGJeiCg7WDB72q6FqvIXvgn41jd9TjSD4SxmER8jUQtMeljfitm9O
/ghfjgFpW2eXKZX///veU111ZktghrtUWwYdnJPavV8TVdmZP5NnxlHes/kwkGqD
uuiRLCcN4AjMuj5bnITCqGuw0wENdwk63AokYErX2s4jQh18qZRBLLgyxfgeg2SC
BAp1EUc6VIylOtypvIs5gGCfLf0Il12VQaa+0+EAmrRrUzQvX5iVUu8khlU957s=
=wrta
-----END PGP SIGNATURE-----

Re: [gentoo-dev] Anyone with access to genkernel repository? Or should genkernel be p.masked on amd64 profiles?

[Ian Stakenvicius]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA256

On 30/05/14 01:14 PM, Ciaran McCreesh wrote:
> On Fri, 30 May 2014 19:11:54 +0200 Tom Wijsman <TomWij@gentoo.org>
> wrote:
>> Well, let's just mask the whole tree; there are wrong lines 
>> everywhere.
> 
> A more reasonable approach would be for the Council to permit the
> tree to contain at most 6 wrong lines at any given time. That way
> any developer wishing to add a new wrong line must previously fix
> an existing wrong line.
> 

hehehe :)  let's get that on the agenda asap!  :D


-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.22 (GNU/Linux)

iF4EAREIAAYFAlOIvfMACgkQ2ugaI38ACPCiWQD/S9r0y3CvavplyIfgeICZahy4
5yLpESb+sP/Rmr2jczkBAKEl3tba/ET2Xemrnq8SOoX0O+w40UhenUUTZtXUtdfb
=IG37
-----END PGP SIGNATURE-----

Re: [gentoo-dev] Anyone with access to genkernel repository? Or should genkernel be p.masked on amd64 profiles?

[Alexander Berntsen]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA256

On 30/05/14 19:18, Tom Wijsman wrote:
> Well, this thread's problem is not severe at all; thus, no need to 
> mask.
I disagree. I'm with ssuominen. Do with that information what you will...
- -- 
Alexander
bernalex@gentoo.org
https://secure.plaimi.net/~alexander
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.22 (GNU/Linux)
Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/

iF4EAREIAAYFAlOIvhAACgkQRtClrXBQc7Uo+wEAuIGuuYNG6EKwDEwWMsM5qxJQ
ihZuZ0qC9VfBwdva5VABAJuq6TkYroOOHxI/ncsABDg2tyY1Q/8VSTwkqzEx9nXz
=2jVG
-----END PGP SIGNATURE-----

Re: [gentoo-dev] Anyone with access to genkernel repository? Or should genkernel be p.masked on amd64 profiles?

[Rich Freeman]

On Fri, May 30, 2014 at 12:50 PM, Tom Wijsman <TomWij@gentoo.org> wrote:
> On Fri, 30 May 2014 19:26:38 +0300
> Samuli Suominen <ssuominen@gentoo.org> wrote:
>
>> I have no plans in inserting my name to genkernel's ChangeLog,
>> and I've done my best to contact people (nobody cares)
>>
>> Only initramfs tool I care and can warmly recommend to anyone is
>> dracut.
>
> I do have such plans for my favorite initramfs tool which is genkernel.
>

While all this bickering is nice to read, what is really needed is:

1.  Somebody needs to step up and actually fix genkernel.
2.  Somebody needs to update the handbook to document how to use
dracut and a list of standard working kernel configs that go along
with it.

Getting into "the official tool is foo" and "foo doesn't work for bar"
really isn't terribly useful.  Those who like foo should make it work,
and those who like the alternatives should help make them
better-supported.

Personally I like dracut, but I'm not terribly happy that the last
dracut upgrade on my PXE+nfs-root box now takes 90 seconds to run DHCP
with the latest upgrade.  But it does look like it will support
mounting that nfs share over an IPv6 tunnel broker.  Whatever - it is
the price of living on the edge, and I haven't gotten around to filing
a bug so it obviously isn't that bad...  :)

Rich

Re: [gentoo-dev] Anyone with access to genkernel repository? Or should genkernel be p.masked on amd64 profiles?

[Rich Freeman]

On Fri, May 30, 2014 at 1:02 PM, Ben Kohler <bkohler@gmail.com> wrote:
> As nice at it sounds to just DROP these configs, that option is not really
> feasible considering the way we currently use genkernel in our handbook.
> Relying on the kernel's own defconfig, "genkernel all" will NOT produce the
> same mostly-usable-on-any-hardware result that we now rely on.

Considering that the configs are more generically useful than
genkernel, having them separately maintained sort-of makes sense.
Then genkernel is a kernel build/install/initramfs tool, not a config
management tool.

I'd stick them someplace where any dev can get to them, and separate
them from the genkernel functional code base.

As far as who takes care of them goes - I suggest that this stuff
comes out of the devmanual unless somebody steps up to take care of
them.  Those who take care of them become those who want to keep them
around.  You can't toss out a tool and ask for it to be a
recommendation but point to others that you think need to maintain its
configuration.

Rich

Re: [gentoo-dev] Anyone with access to genkernel repository? Or should genkernel be p.masked on amd64 profiles?

[Fabio Erculiani]

configs should have never gotten into genkernel in the first place.
it's each kernel pkg (or even version) that owns a valid config of
itself.
I am part of genkernel@ but I have no will nor time to fix it. And
when I have, I'd rather work on genkernel-next, that comes with a much
more readable initramfs code (that I managed to rewrite myself).

Wiping the whole config files has been on my agenda for very long time already.

On Fri, May 30, 2014 at 6:32 PM, Rich Freeman <rich0@gentoo.org> wrote:
> On Fri, May 30, 2014 at 1:02 PM, Ben Kohler <bkohler@gmail.com> wrote:
>> As nice at it sounds to just DROP these configs, that option is not really
>> feasible considering the way we currently use genkernel in our handbook.
>> Relying on the kernel's own defconfig, "genkernel all" will NOT produce the
>> same mostly-usable-on-any-hardware result that we now rely on.
>
> Considering that the configs are more generically useful than
> genkernel, having them separately maintained sort-of makes sense.
> Then genkernel is a kernel build/install/initramfs tool, not a config
> management tool.
>
> I'd stick them someplace where any dev can get to them, and separate
> them from the genkernel functional code base.
>
> As far as who takes care of them goes - I suggest that this stuff
> comes out of the devmanual unless somebody steps up to take care of
> them.  Those who take care of them become those who want to keep them
> around.  You can't toss out a tool and ask for it to be a
> recommendation but point to others that you think need to maintain its
> configuration.
>
> Rich
>



-- 
Fabio Erculiani

Re: [gentoo-dev] Anyone with access to genkernel repository? Or should genkernel be p.masked on amd64 profiles?

[Rick "Zero_Chaos" Farina]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On 05/30/2014 01:39 PM, Fabio Erculiani wrote:
> configs should have never gotten into genkernel in the first place.
> it's each kernel pkg (or even version) that owns a valid config of
> itself.
> I am part of genkernel@ but I have no will nor time to fix it. And
> when I have, I'd rather work on genkernel-next, that comes with a much
> more readable initramfs code (that I managed to rewrite myself).
> 
> Wiping the whole config files has been on my agenda for very long time already.

And replaced by what?  defconfig? allmodconfig?

- -Zero
> 
> On Fri, May 30, 2014 at 6:32 PM, Rich Freeman <rich0@gentoo.org> wrote:
>> On Fri, May 30, 2014 at 1:02 PM, Ben Kohler <bkohler@gmail.com> wrote:
>>> As nice at it sounds to just DROP these configs, that option is not really
>>> feasible considering the way we currently use genkernel in our handbook.
>>> Relying on the kernel's own defconfig, "genkernel all" will NOT produce the
>>> same mostly-usable-on-any-hardware result that we now rely on.
>>
>> Considering that the configs are more generically useful than
>> genkernel, having them separately maintained sort-of makes sense.
>> Then genkernel is a kernel build/install/initramfs tool, not a config
>> management tool.
>>
>> I'd stick them someplace where any dev can get to them, and separate
>> them from the genkernel functional code base.
>>
>> As far as who takes care of them goes - I suggest that this stuff
>> comes out of the devmanual unless somebody steps up to take care of
>> them.  Those who take care of them become those who want to keep them
>> around.  You can't toss out a tool and ask for it to be a
>> recommendation but point to others that you think need to maintain its
>> configuration.
>>
>> Rich
>>
> 
> 
> 

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.22 (GNU/Linux)
Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/
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=hNQF
-----END PGP SIGNATURE-----

Re: [gentoo-dev] Anyone with access to genkernel repository? Or should genkernel be p.masked on amd64 profiles?

[Tom Wijsman]

[-- Attachment #1: Type: text/plain, Size: 2260 bytes --]

On Fri, 30 May 2014 13:32:54 -0400
Rich Freeman <rich0@gentoo.org> wrote:

> On Fri, May 30, 2014 at 1:02 PM, Ben Kohler <bkohler@gmail.com> wrote:
> > As nice at it sounds to just DROP these configs, that option is not
> > really feasible considering the way we currently use genkernel in
> > our handbook. Relying on the kernel's own defconfig, "genkernel
> > all" will NOT produce the same mostly-usable-on-any-hardware result
> > that we now rely on.
> 
> Considering that the configs are more generically useful than
> genkernel, having them separately maintained sort-of makes sense.
> Then genkernel is a kernel build/install/initramfs tool, not a config
> management tool.

It has always been a kernel build/install/initramfs tool; some people
misinterpret it for being a "default big working config" tool, without
seeing what else it can do. See /etc/genkernel.conf and `man genkernel`.

> I'd stick them someplace where any dev can get to them, and separate
> them from the genkernel functional code base.

Good idea, we really could use some kind of kernel seeds in the Portage
tree; if someone is willing to maintain them, knowing that Pappy has
maintained them for years and spoke about it it seems like hard work.

Pappy has discontinued and sold his kernel seeds; a Funtoo developer
bought it, but we're yet to see the first release of this continuation.
 
> As far as who takes care of them goes - I suggest that this stuff
> comes out of the devmanual unless somebody steps up to take care of
> them.  Those who take care of them become those who want to keep them
> around.  You can't toss out a tool and ask for it to be a
> recommendation but point to others that you think need to maintain its
> configuration.

s/devmanual/Gentoo Installation Handbook/ to avoid confusion. Other
than that; I agree with this as Genkernel works perfectly itself and
its just a small problem in the config that could be solved in various
ways, like fixing/forking/splitting/removing the config that come along.

-- 
With kind regards,

Tom Wijsman (TomWij)
Gentoo Developer

E-mail address  : TomWij@gentoo.org
GPG Public Key  : 6D34E57D
GPG Fingerprint : C165 AF18 AB4C 400B C3D2  ABF0 95B2 1FCD 6D34 E57D

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 490 bytes --]

Re: [gentoo-dev] Anyone with access to genkernel repository? Or should genkernel be p.masked on amd64 profiles?

[Tom Wijsman]

[-- Attachment #1: Type: text/plain, Size: 641 bytes --]

On Fri, 30 May 2014 13:22:54 -0400
Rich Freeman <rich0@gentoo.org> wrote:

> While all this bickering is nice to read, what is really needed is:
> 
> 1.  Somebody needs to step up and actually fix genkernel.

And that's why we await the maintainers' or undertakers' response. With
undertakers, I mean the type of "up for grabs" activity that Pacho does.

Given this isn't severe, there's no hurry; thus no immediate need...

-- 
With kind regards,

Tom Wijsman (TomWij)
Gentoo Developer

E-mail address  : TomWij@gentoo.org
GPG Public Key  : 6D34E57D
GPG Fingerprint : C165 AF18 AB4C 400B C3D2  ABF0 95B2 1FCD 6D34 E57D

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 490 bytes --]

Re: [gentoo-dev] Anyone with access to genkernel repository? Or should genkernel be p.masked on amd64 profiles?

[Ben Kohler]

[-- Attachment #1: Type: text/plain, Size: 650 bytes --]

On Fri, May 30, 2014 at 12:59 PM, Tom Wijsman <TomWij@gentoo.org> wrote:
>
> Good idea, we really could use some kind of kernel seeds in the Portage
> tree; if someone is willing to maintain them, knowing that Pappy has
> maintained them for years and spoke about it it seems like hard work.
>
> Remember that these "kernel seeds" are actually pretty far from what we
need for genkernel defaults (at least the way we currently recommend using
it in our install handbook).  The kernel seeds provide a good sane kernel
config *with little or no specific hardware support*.

But I'm definitely in favor of separately packaging the kernel configs.

-Ben

[-- Attachment #2: Type: text/html, Size: 1017 bytes --]

Re: [gentoo-dev] Anyone with access to genkernel repository? Or should genkernel be p.masked on amd64 profiles?

[Tom Wijsman]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On Fri, 30 May 2014 19:21:20 +0200
Alexander Berntsen <bernalex@gentoo.org> wrote:

> On 30/05/14 19:18, Tom Wijsman wrote:
> > Well, this thread's problem is not severe at all; thus, no need to 
> > mask.
> I disagree. I'm with ssuominen. Do with that information what you
> will...

One can disagree about that, even claim a need for masking; that still
doesn't mean one can ignore the maintainers and mask their package.

It's not *that* severe for there to be *such* mask; anyway, the commit
watches have been set up by some of those that would be affected, which
includes parties like releng and infra that both rely on this package.

If this were really severe, a ton of parties (users, dev, releng,
infra, ...) would've complained about this; but in my participation
everywhere, this comes over like a single once in a lifetime sea wave.

- -- 
With kind regards,

Tom Wijsman (TomWij)
Gentoo Developer

E-mail address  : TomWij@gentoo.org
GPG Public Key  : 6D34E57D
GPG Fingerprint : C165 AF18 AB4C 400B C3D2  ABF0 95B2 1FCD 6D34 E57D
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.22 (GNU/Linux)

iQEcBAEBAgAGBQJTiMrIAAoJEPWZc8roOL/QI9QH/jCt1y4/elM0VUvvJhu8KFXB
uQxvAHwArsEAf1PAjheRHKuEw+JlB8rSRRxz6T//M7I3MClpQ7q6teVuWy8HrChe
xI6hUF2QROZWzXYpf46c6fvqE6z6sJtI9AM6vzUCJdjpGhmQyDYBQFk9oYR+OCkb
6ZJM/SMURWcKxmkl0SgoMmDFSQVxTX/Lx3mpgAlfu1axyarJEGxODCF+kGbppKoB
xlulodqyvQJmoZyVIqmXf5Tv7acmSRkGfUmq3vaPHA/Y+7OH/Z1sp+8k0uzn7Sgw
ipZdsm93SvBCvfntFWXjk+uNI0z7i6R0vFuRJMmn02bLSEb2koBx3EbNPfBm/jU=
=aCvB
-----END PGP SIGNATURE-----

Re: [gentoo-dev] Anyone with access to genkernel repository? Or should genkernel be p.masked on amd64 profiles?

[Tom Wijsman]

[-- Attachment #1: Type: text/plain, Size: 1134 bytes --]

On Fri, 30 May 2014 13:08:50 -0500
Ben Kohler <bkohler@gmail.com> wrote:

> On Fri, May 30, 2014 at 12:59 PM, Tom Wijsman <TomWij@gentoo.org>
> wrote:
> >
> > Good idea, we really could use some kind of kernel seeds in the
> > Portage tree; if someone is willing to maintain them, knowing that
> > Pappy has maintained them for years and spoke about it it seems
> > like hard work.
> >
> > Remember that these "kernel seeds" are actually pretty far from
> > what we
> need for genkernel defaults (at least the way we currently recommend
> using it in our install handbook).  The kernel seeds provide a good
> sane kernel config *with little or no specific hardware support*.
> 
> But I'm definitely in favor of separately packaging the kernel
> configs.
> 
> -Ben

Yeah, same thoughts; I meant kernel seeds in a more general way, but
indeed "seed" wouldn't be the right word, perhaps "kernel plants"? :)

-- 
With kind regards,

Tom Wijsman (TomWij)
Gentoo Developer

E-mail address  : TomWij@gentoo.org
GPG Public Key  : 6D34E57D
GPG Fingerprint : C165 AF18 AB4C 400B C3D2  ABF0 95B2 1FCD 6D34 E57D

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 490 bytes --]

Re: [gentoo-dev] Anyone with access to genkernel repository? Or should genkernel be p.masked on amd64 profiles?

[Sven Vermeulen]

On Fri, May 30, 2014 at 04:34:09PM +0000, hasufell wrote:
> Tom Wijsman:
> >
> > Please no p.mask for a single line being wrong...
> >
>
> That's nonsense. The amount of wrong lines doesn't matter. A single
> wrong line in the kernel can break your whole system as well.
>
> Please p.mask (or patch) immediately. There is no point in waiting.

I'd appreciate any patch on the Gentoo Handbook to use another tool for
initramfs. I have some experience with dracut, but the last dracut generated
initramfs failed on my system, whereas genkernel's still goes well.

The last Dracut generated initramfs also failed on SELinux systems, but that
isn't something that cannot be worked around...

Wkr,
	Sven Vermeulen

Re: [gentoo-dev] Anyone with access to genkernel repository? Or should genkernel be p.masked on amd64 profiles?

[hasufell]

Sven Vermeulen:
> On Fri, May 30, 2014 at 04:34:09PM +0000, hasufell wrote:
>> Tom Wijsman:
>>>
>>> Please no p.mask for a single line being wrong...
>>>
>>
>> That's nonsense. The amount of wrong lines doesn't matter. A single
>> wrong line in the kernel can break your whole system as well.
>>
>> Please p.mask (or patch) immediately. There is no point in waiting.
> 
> I'd appreciate any patch on the Gentoo Handbook to use another tool for
> initramfs. I have some experience with dracut, but the last dracut generated
> initramfs failed on my system, whereas genkernel's still goes well.
> 
> The last Dracut generated initramfs also failed on SELinux systems, but that
> isn't something that cannot be worked around...
> 
> Wkr,
> 	Sven Vermeulen
> 

I have never used any of them. I am a bit confused why this is the
standard gentoo way, but I don't really care what is. A lot of gentoo
standards don't make any sense, so...

If you need help with manual initramfs configuration, feel free to ping
me. The only thing that was a bit more complicated was zfs support.

Re: [gentoo-dev] Anyone with access to genkernel repository? Or should genkernel be p.masked on amd64 profiles?

[Andreas K. Huettel]

Am Freitag, 30. Mai 2014, 19:14:11 schrieb Ciaran McCreesh:
> On Fri, 30 May 2014 19:11:54 +0200
> 
> Tom Wijsman <TomWij@gentoo.org> wrote:
> > Well, let's just mask the whole tree; there are wrong lines
> > everywhere.
> 
> A more reasonable approach would be for the Council to permit the tree
> to contain at most 6 wrong lines at any given time. That way any
> developer wishing to add a new wrong line must previously fix an
> existing wrong line.

You just made my day. :D

-- 
Andreas K. Huettel
Gentoo Linux developer (council, kde)
dilfridge@gentoo.org
http://www.akhuettel.de/

Re: [gentoo-dev] Anyone with access to genkernel repository? Or should genkernel be p.masked on amd64 profiles?

[Alexander Tsoy]

On Fri May 30 22:30:32 2014 Sven Vermeulen <swift@gentoo.org> wrote:
> On Fri, May 30, 2014 at 04:34:09PM +0000, hasufell wrote:
> > Tom Wijsman:
> > > 
> > > Please no p.mask for a single line being wrong...
> > > 
> > 
> > That's nonsense. The amount of wrong lines doesn't matter. A single
> > wrong line in the kernel can break your whole system as well.
> > 
> > Please p.mask (or patch) immediately. There is no point in waiting.
> 
> I'd appreciate any patch on the Gentoo Handbook to use another tool for
> initramfs. I have some experience with dracut, but the last dracut
> generated initramfs failed on my system, whereas genkernel's still goes
> well.
> 

Hmm.. You are the second person in this thread who experienced problems with latest dracut. I suspect this is due to hostonly-cmdline option, which is disabled by default. If this is really the case, we should either enable it in gentoo.conf or revert corresponding changes like in fedora [1].

And where are bug reports? :)

Sorry for offtop.

[1] http://pkgs.fedoraproject.org/cgit/dracut.git/commit/?id=2fa76e2882883ea64be6c3210c257cff4e4b641a

> The last Dracut generated initramfs also failed on SELinux systems, but
> that isn't something that cannot be worked around...
> 
> Wkr,
>     Sven Vermeulen
> 

-- 
Alexander Tsoy

[gentoo-dev] Re: Anyone with access to genkernel repository? Or should genkernel be p.masked on amd64 profiles?

[Steven J. Long]

On Fri, May 30, 2014, Rick "Zero_Chaos" Farina wrote:
> On 05/30/2014 11:10 AM, Samuli Suominen wrote:
> > https://bugs.gentoo.org/show_bug.cgi?id=461828
> > 
> > I'll p.mask it on amd64 profiles if noone replies soon :(
> > 
> 
> Please don't p.mask a working program because a config file is wrong.
> The arch teams think the genkernel team should be updating the kernel
> configs and vice-versa, so no one does it.  I would be fine with
> entirely removing the kernel configs in genkernel, but I assure you a
> p.mask won't last long as it breaks releng and breaks users.
> 
> I'm super happy for you that you like dracut, but gentoo officially uses
> genkernel and it shall not be p.masked due to an OPTIONAL config file.

Makes sense; the suggestion to use releng configs also makes sense. No
configs also works, given that the relevant sources package is usually
patched within Gentoo, but a releng config would lead to more testing,
and easier collaboration, one might hope.

Though I'm at a loss as to how a package.mask is any less work than
simply setting the string to ""; worrying about a changelong entry in
an environment where everyone deliberately has commit to everything,
seems inane.

In any event, please move to simplify to one or none, rather than
current 2, as you discussed in the bug.

Regards,
steveL
-- 
#friendly-coders -- We're friendly, but we're not /that/ friendly ;-)

Re: [gentoo-dev] Anyone with access to genkernel repository? Or should genkernel be p.masked on amd64 profiles?

[Robin H. Johnson]

On Fri, May 30, 2014 at 06:10:40PM +0300, Samuli Suominen wrote:
> I can't find anyone with access that actually replies to mails, pings,
> ... to genkernel repository for:
> 
> https://bugs.gentoo.org/show_bug.cgi?id=461828
> 
> I'll p.mask it on amd64 profiles if noone replies soon :(
My excuse is AFK baby, literally sleeping on me right now, and not even
2 months old.

I saw floppym's original comment opening the bug, but none of the
followups due to baby eating my life.

I'll apply this patch later today if I have a chance, but I do agree
with the general sentiment of this thread that the kernel configs NEED
to get out of genkernel; so that arches can touch them at will, and
other initramfs/kernel build tools can start to use them.

In the absence of any other prompt complaints, I'll create a
kernel-configs repo, and move the configs there.

The one thing that WILL have to be maintained in genkernel still, and
in-sync with kernel changes, are the modules_load files,  esp when new
common stuff is added to the kernel (disk controllers & filesystems most
commonly).

-- 
Robin Hugh Johnson
Gentoo Linux: Developer, Infrastructure Lead
E-Mail     : robbat2@gentoo.org
GnuPG FP   : 11ACBA4F 4778E3F6 E4EDF38E B27B944E 34884E85

Re: [gentoo-dev] Anyone with access to genkernel repository? Or should genkernel be p.masked on amd64 profiles?

[Samuli Suominen]

On 31/05/14 22:41, Robin H. Johnson wrote:
> On Fri, May 30, 2014 at 06:10:40PM +0300, Samuli Suominen wrote:
>> I can't find anyone with access that actually replies to mails, pings,
>> ... to genkernel repository for:
>>
>> https://bugs.gentoo.org/show_bug.cgi?id=461828
>>
>> I'll p.mask it on amd64 profiles if noone replies soon :(
> My excuse is AFK baby, literally sleeping on me right now, and not even
> 2 months old.
>
> I saw floppym's original comment opening the bug, but none of the
> followups due to baby eating my life.
>
> I'll apply this patch later today if I have a chance, but I do agree
> with the general sentiment of this thread that the kernel configs NEED
> to get out of genkernel; so that arches can touch them at will, and
> other initramfs/kernel build tools can start to use them.
>
> In the absence of any other prompt complaints, I'll create a
> kernel-configs repo, and move the configs there.
>
> The one thing that WILL have to be maintained in genkernel still, and
> in-sync with kernel changes, are the modules_load files,  esp when new
> common stuff is added to the kernel (disk controllers & filesystems most
> commonly).
>

The patch in the bug is not enough. Notice
http://bugs.gentoo.org/show_bug.cgi?id=461828#c13
Those should be reseted to "" too.

Thanks,
Samuli

Re: [gentoo-dev] Anyone with access to genkernel repository? Or should genkernel be p.masked on amd64 profiles?

[Fabio Erculiani]

[-- Attachment #1: Type: text/plain, Size: 1609 bytes --]

On May 31, 2014 8:42 PM, "Robin H. Johnson" <robbat2@gentoo.org> wrote:
>
> On Fri, May 30, 2014 at 06:10:40PM +0300, Samuli Suominen wrote:
> > I can't find anyone with access that actually replies to mails, pings,
> > ... to genkernel repository for:
> >
> > https://bugs.gentoo.org/show_bug.cgi?id=461828
> >
> > I'll p.mask it on amd64 profiles if noone replies soon :(
> My excuse is AFK baby, literally sleeping on me right now, and not even
> 2 months old.
>
> I saw floppym's original comment opening the bug, but none of the
> followups due to baby eating my life.
>
> I'll apply this patch later today if I have a chance, but I do agree
> with the general sentiment of this thread that the kernel configs NEED
> to get out of genkernel; so that arches can touch them at will, and
> other initramfs/kernel build tools can start to use them.
>
> In the absence of any other prompt complaints, I'll create a
> kernel-configs repo, and move the configs there.

It would be better if those would be put in individual source pkgs in a way
that they can be picked up by genkernel. Kernel config belongs to kernel
pkgs, pretty much like init scripts or config files belong to their own
project pkgs.

>
> The one thing that WILL have to be maintained in genkernel still, and
> in-sync with kernel changes, are the modules_load files,  esp when new
> common stuff is added to the kernel (disk controllers & filesystems most
> commonly).
>
> --
> Robin Hugh Johnson
> Gentoo Linux: Developer, Infrastructure Lead
> E-Mail     : robbat2@gentoo.org
> GnuPG FP   : 11ACBA4F 4778E3F6 E4EDF38E B27B944E 34884E85
>

[-- Attachment #2: Type: text/html, Size: 2137 bytes --]

Re: [gentoo-dev] Anyone with access to genkernel repository? Or should genkernel be p.masked on amd64 profiles?

[Robin H. Johnson]

On Sat, May 31, 2014 at 10:46:35PM +0300, Samuli Suominen wrote:
> The patch in the bug is not enough. Notice
> http://bugs.gentoo.org/show_bug.cgi?id=461828#c13
> Those should be reseted to "" too.
Read what I applied, I did set it to "".

-- 
Robin Hugh Johnson
Gentoo Linux: Developer, Infrastructure Lead
E-Mail     : robbat2@gentoo.org
GnuPG FP   : 11ACBA4F 4778E3F6 E4EDF38E B27B944E 34884E85

Re: [gentoo-dev] Anyone with access to genkernel repository? Or should genkernel be p.masked on amd64 profiles?

[Robin H. Johnson]

On Sat, May 31, 2014 at 08:51:39PM +0100, Fabio Erculiani wrote:
> On May 31, 2014 8:42 PM, "Robin H. Johnson" <robbat2@gentoo.org> wrote:
> >
> > On Fri, May 30, 2014 at 06:10:40PM +0300, Samuli Suominen wrote:
> > > I can't find anyone with access that actually replies to mails, pings,
> > > ... to genkernel repository for:
> > >
> > > https://bugs.gentoo.org/show_bug.cgi?id=461828
> > >
> > > I'll p.mask it on amd64 profiles if noone replies soon :(
> > My excuse is AFK baby, literally sleeping on me right now, and not even
> > 2 months old.
> >
> > I saw floppym's original comment opening the bug, but none of the
> > followups due to baby eating my life.
> >
> > I'll apply this patch later today if I have a chance, but I do agree
> > with the general sentiment of this thread that the kernel configs NEED
> > to get out of genkernel; so that arches can touch them at will, and
> > other initramfs/kernel build tools can start to use them.
> >
> > In the absence of any other prompt complaints, I'll create a
> > kernel-configs repo, and move the configs there.
> 
> It would be better if those would be put in individual source pkgs in a way
> that they can be picked up by genkernel. Kernel config belongs to kernel
> pkgs, pretty much like init scripts or config files belong to their own
> project pkgs.
No, I don't agree that kernel configs "belong" to kernel packages.  In
general, barring the crazy option explosion, these are meant to be stock
working configs that should in combination with ANY kernel package,
produce a working kernel.

As for the rest of the 'kernel seeds' stuff; it never made it properly
into the tree, and because of that, didn't get much traction.

Infra has recently put together their own ebuild-that-runs-genkernel
setup, so we could roll out kernels more consistently, and if I can
solve one bug [1], I'm strongly considering putting that functionality
into an eclass, and giving ALL sys-kernel/*sources the ability to spit
out a built kernel with genkernel, for the price of an emerge.

[1] Taking binaries from the host system can lead to problems if you
want the initramfs to work reliably everywhere. glibc compiled with
SSE4.2 bit infra, because libm got linked into stuff being built with
-march=generic.

-- 
Robin Hugh Johnson
Gentoo Linux: Developer, Infrastructure Lead
E-Mail     : robbat2@gentoo.org
GnuPG FP   : 11ACBA4F 4778E3F6 E4EDF38E B27B944E 34884E85

Re: [gentoo-dev] Anyone with access to genkernel repository? Or should genkernel be p.masked on amd64 profiles?

[Samuli Suominen]

On 31/05/14 23:00, Robin H. Johnson wrote:
> On Sat, May 31, 2014 at 10:46:35PM +0300, Samuli Suominen wrote:
>> The patch in the bug is not enough. Notice
>> http://bugs.gentoo.org/show_bug.cgi?id=461828#c13
>> Those should be reseted to "" too.
> Read what I applied, I did set it to "".
>

Thanks, looks good.   Can we go with fast stabilizing this version?

Re: [gentoo-dev] Anyone with access to genkernel repository? Or should genkernel be p.masked on amd64 profiles?

[Fabio Erculiani]

On Sat, May 31, 2014 at 9:06 PM, Robin H. Johnson <robbat2@gentoo.org> wrote:
> No, I don't agree that kernel configs "belong" to kernel packages.  In
> general, barring the crazy option explosion, these are meant to be stock
> working configs that should in combination with ANY kernel package,
> produce a working kernel.
>

Then you are just moving the problem around.
I believe that kernel configs should be provided by their own kernel
packages (and there are some, not just gentoo-sources) because it is
much easier to keep them in sync on every new release and deal with
each version separately if/as needed (including testing!). How are you
dealing with config var name changes between different kernel versions
or just different pkgs then?
You cannot possibly support all kernel versions for all kernel pkgs
available in tree with just one single config file in a sane, clean
and maintainable way, hoping that a change in this file will not
affect previous or future kernel releases. How are you going to test
your config changes against old kernel pkgs? Each test is quite
expensive to run.

Good luck with that :-)

[...]

>
> --
> Robin Hugh Johnson
> Gentoo Linux: Developer, Infrastructure Lead
> E-Mail     : robbat2@gentoo.org
> GnuPG FP   : 11ACBA4F 4778E3F6 E4EDF38E B27B944E 34884E85
>



-- 
Fabio Erculiani

Re: [gentoo-dev] Anyone with access to genkernel repository? Or should genkernel be p.masked on amd64 profiles?

[Tom Wijsman]

[-- Attachment #1: Type: text/plain, Size: 2330 bytes --]

On Sat, 31 May 2014 22:42:17 +0100
Fabio Erculiani <lxnay@gentoo.org> wrote:

> On Sat, May 31, 2014 at 9:06 PM, Robin H. Johnson
> <robbat2@gentoo.org> wrote:
> > No, I don't agree that kernel configs "belong" to kernel packages.
> > In general, barring the crazy option explosion, these are meant to
> > be stock working configs that should in combination with ANY kernel
> > package, produce a working kernel.
> >
> 
> Then you are just moving the problem around.
> I believe that kernel configs should be provided by their own kernel
> packages (and there are some, not just gentoo-sources) because it is
> much easier to keep them in sync on every new release and deal with
> each version separately if/as needed (including testing!). How are you
> dealing with config var name changes between different kernel versions
> or just different pkgs then?

Different packages is not a problem; since the difference in terms of
config between separate packages is small enough, a dozen of options.

Different version may be a problem, a rather small one; nothing
prevents one from keeping config options around for both versions.

> You cannot possibly support all kernel versions for all kernel pkgs
> available in tree with just one single config file in a sane, clean
> and maintainable way, hoping that a change in this file will not
> affect previous or future kernel releases. How are you going to test
> your config changes against old kernel pkgs? Each test is quite
> expensive to run.
> 
> Good luck with that :-)

Does it really need to be sane, clean and maintainable for it to work?

Yes, maybe; but how sane, clean and maintainable? We can do better...

A fork (eg. hardened-sources config, geek-sources config) where needed
might still be a way out; however, one should consider to look into a
better architecture than plain forks. For example, a config that
sources a generic config and adds hardened changes on top of that;
kind of like the way GRUB 2's /etc/grub.d/ config generation works.

We should introduce this only when needed to avoid to over-design it.

-- 
With kind regards,

Tom Wijsman (TomWij)
Gentoo Developer

E-mail address  : TomWij@gentoo.org
GPG Public Key  : 6D34E57D
GPG Fingerprint : C165 AF18 AB4C 400B C3D2  ABF0 95B2 1FCD 6D34 E57D

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 490 bytes --]

Re: [gentoo-dev] Anyone with access to genkernel repository? Or should genkernel be p.masked on amd64 profiles?

[Robin H. Johnson]

On Sat, May 31, 2014 at 11:16:35PM +0300, Samuli Suominen wrote:
> 
> On 31/05/14 23:00, Robin H. Johnson wrote:
> > On Sat, May 31, 2014 at 10:46:35PM +0300, Samuli Suominen wrote:
> >> The patch in the bug is not enough. Notice
> >> http://bugs.gentoo.org/show_bug.cgi?id=461828#c13
> >> Those should be reseted to "" too.
> > Read what I applied, I did set it to "".
> >
> 
> Thanks, looks good.   Can we go with fast stabilizing this version?
Bug 511992; amd64 & x86 done already.

-- 
Robin Hugh Johnson
Gentoo Linux: Developer, Infrastructure Lead
E-Mail     : robbat2@gentoo.org
GnuPG FP   : 11ACBA4F 4778E3F6 E4EDF38E B27B944E 34884E85

Re: [gentoo-dev] Anyone with access to genkernel repository? Or should genkernel be p.masked on amd64 profiles?

[Robin H. Johnson]

On Sat, May 31, 2014 at 10:42:17PM +0100, Fabio Erculiani wrote:
> On Sat, May 31, 2014 at 9:06 PM, Robin H. Johnson <robbat2@gentoo.org> wrote:
> > No, I don't agree that kernel configs "belong" to kernel packages.  In
> > general, barring the crazy option explosion, these are meant to be stock
> > working configs that should in combination with ANY kernel package,
> > produce a working kernel.
> >
> 
> Then you are just moving the problem around.
> I believe that kernel configs should be provided by their own kernel
> packages (and there are some, not just gentoo-sources) because it is
> much easier to keep them in sync on every new release and deal with
> each version separately if/as needed (including testing!). How are you
> dealing with config var name changes between different kernel versions
> or just different pkgs then?
>
> You cannot possibly support all kernel versions for all kernel pkgs
> available in tree with just one single config file in a sane, clean
> and maintainable way, hoping that a change in this file will not
> affect previous or future kernel releases. How are you going to test
> your config changes against old kernel pkgs? Each test is quite
> expensive to run.
I never said I was going to support all different kernel sources.

genkernel only officially supports gentoo-sources & hardened-sources.
(and those are supersets of the vanilla tree).

The stock genkernel config actually works for most users, on most kernel
sources, on most systems; and parts of it date back to 2.6.14. Sure it
doesn't turn on all of the features requested, but it does actually
work.

That said, I do intend to included an official kernel config for each
major kernel release 3.x; for both hardened & gentoo-sources. Infra uses
the hardened one, and releng uses the gentoo-sources one.

The only change really is that the packaging is going to be separate
from genkernel, and it'll get a bit more care than before.

-- 
Robin Hugh Johnson
Gentoo Linux: Developer, Infrastructure Lead
E-Mail     : robbat2@gentoo.org
GnuPG FP   : 11ACBA4F 4778E3F6 E4EDF38E B27B944E 34884E85

Re: [gentoo-dev] Anyone with access to genkernel repository? Or should genkernel be p.masked on amd64 profiles?

[Jeroen Roovers]

On Fri, 30 May 2014 19:17:31 +0200
Tom Wijsman <TomWij@gentoo.org> wrote:

> On Fri, 30 May 2014 18:14:11 +0100
> Ciaran McCreesh <ciaran.mccreesh@googlemail.com> wrote:
> > A more reasonable approach would be for the Council to permit the
> > tree to contain at most 6 wrong lines at any given time. That way
> > any developer wishing to add a new wrong line must previously fix an
> > existing wrong line.
> 
> You can suggest that to the Gentoo Council by replying to the Council
> agenda thread on the gentoo-project mailing list.

whooosh


     jer

Re: [gentoo-dev] Anyone with access to genkernel repository? Or should genkernel be p.masked on amd64 profiles?

[Tom Wijsman]

[-- Attachment #1: Type: text/plain, Size: 907 bytes --]

On Tue, 3 Jun 2014 20:58:46 +0200
Jeroen Roovers <jer@gentoo.org> wrote:

> On Fri, 30 May 2014 19:17:31 +0200
> Tom Wijsman <TomWij@gentoo.org> wrote:
> 
> > On Fri, 30 May 2014 18:14:11 +0100
> > Ciaran McCreesh <ciaran.mccreesh@googlemail.com> wrote:
> > > A more reasonable approach would be for the Council to permit the
> > > tree to contain at most 6 wrong lines at any given time. That way
> > > any developer wishing to add a new wrong line must previously fix
> > > an existing wrong line.
> > 
> > You can suggest that to the Gentoo Council by replying to the
> > Council agenda thread on the gentoo-project mailing list.
> 
> whooosh
> 
> 
>      jer
> 

boom splat

-- 
With kind regards,

Tom Wijsman (TomWij)
Gentoo Developer

E-mail address  : TomWij@gentoo.org
GPG Public Key  : 6D34E57D
GPG Fingerprint : C165 AF18 AB4C 400B C3D2  ABF0 95B2 1FCD 6D34 E57D

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 490 bytes --]

Re: [gentoo-dev] Anyone with access to genkernel repository? Or should genkernel be p.masked on amd64 profiles?

[hasufell]

Tom Wijsman:
> On Tue, 3 Jun 2014 20:58:46 +0200
> Jeroen Roovers <jer@gentoo.org> wrote:
> 
>> On Fri, 30 May 2014 19:17:31 +0200
>> Tom Wijsman <TomWij@gentoo.org> wrote:
>>
>>> On Fri, 30 May 2014 18:14:11 +0100
>>> Ciaran McCreesh <ciaran.mccreesh@googlemail.com> wrote:
>>>> A more reasonable approach would be for the Council to permit the
>>>> tree to contain at most 6 wrong lines at any given time. That way
>>>> any developer wishing to add a new wrong line must previously fix
>>>> an existing wrong line.
>>>
>>> You can suggest that to the Gentoo Council by replying to the
>>> Council agenda thread on the gentoo-project mailing list.
>>
>> whooosh
>>
>>
>>      jer
>>
> 
> boom splat
> 

Stop spamming and stop telling obvious things. We can write a bot that
gives that kind of information.

Re: [gentoo-dev] Anyone with access to genkernel repository? Or should genkernel be p.masked on amd64 profiles?

[Tom Wijsman]

[-- Attachment #1: Type: text/plain, Size: 1472 bytes --]

On Tue, 03 Jun 2014 22:35:13 +0000
hasufell <hasufell@gentoo.org> wrote:

> Tom Wijsman:
> > On Tue, 3 Jun 2014 20:58:46 +0200
> > Jeroen Roovers <jer@gentoo.org> wrote:
> > 
> >> On Fri, 30 May 2014 19:17:31 +0200
> >> Tom Wijsman <TomWij@gentoo.org> wrote:
> >>
> >>> On Fri, 30 May 2014 18:14:11 +0100
> >>> Ciaran McCreesh <ciaran.mccreesh@googlemail.com> wrote:
> >>>> A more reasonable approach would be for the Council to permit the
> >>>> tree to contain at most 6 wrong lines at any given time. That way
> >>>> any developer wishing to add a new wrong line must previously fix
> >>>> an existing wrong line.
> >>>
> >>> You can suggest that to the Gentoo Council by replying to the
> >>> Council agenda thread on the gentoo-project mailing list.
> >>
> >> whooosh
> >>
> >>
> >>      jer
> >>
> > 
> > boom splat
> > 
> 
> Stop spamming and stop telling obvious things. We can write a bot that
> gives that kind of information.

We can also write a bot that points out that it is "nonsense", "doesn't
matter" and "breakage" as in your original response to this sub thread.

Can you instead tell us what you think makes sense, matters and works?

In this case, for me; whooosh boom splat makes sense, matters and works.

-- 
With kind regards,

Tom Wijsman (TomWij)
Gentoo Developer

E-mail address  : TomWij@gentoo.org
GPG Public Key  : 6D34E57D
GPG Fingerprint : C165 AF18 AB4C 400B C3D2  ABF0 95B2 1FCD 6D34 E57D

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 490 bytes --]