[gentoo-dev] Linux 2.6.21 plans

[gentoo-dev] Linux 2.6.21 plans

[Daniel Drake]

Hi,

2.6.21 was released today. Testing muchly appreciated as usual -- please 
file bugs and clearly mark them as 2.6.21 regressions if that is the case.

There will probably be several packages unable to compile/load due to 
internal kernel API changes, as usual. Please make these block bug 
#176188 and please do treat these bugs with relatively high priority.

I'm hoping that we'll be able to return to our usual release cycle of 
pushing to get 2.6.21 marked stable in 3 weeks time, plus a week for 
ironing out the final few issues.

This means that we may be pushing for 2.6.21 stable on x86 and amd64 on 
May 17th. If important issues come up (which they may well do), this 
will obviously be delayed, but do keep this date in mind.

Thanks,
Daniel
-- 
gentoo-dev@gentoo.org mailing list

[gentoo-dev] Re: Linux 2.6.21 plans

[Duncan]

Daniel Drake <dsd@gentoo.org> posted 46313C29.7050501@gentoo.org,
excerpted below, on  Thu, 26 Apr 2007 19:56:25 -0400:

> This means that we may be pushing for 2.6.21 stable on x86 and amd64 on
> May 17th. If important issues come up (which they may well do), this
> will obviously be delayed, but do keep this date in mind.

As I'm sure you are aware, there are more known regressions remaining 
this time around.  As with some others, I'm not all that confident this 
kernel was ready for release just yet.  Oh, well, I suppose there'll be a 
2.6.21.1 and etc...

I'm running (vanilla) rc7-git10 ATM, and have two possible regressions 
remaining here, which <details> kept me from properly testing and bugging 
upstream in a timely manner.  Both are regressions as they worked before, 
but I'm not yet totally sure they are kernel regressions.  Briefly, 
amd64, (1) system clock to hardware clock isn't syncing now, I'm not sure 
about hw2system as I use ntp, and (2) X was restoring after suspend to 
disk, but now breaks, such that I immediately return to the console the 
first time I try to switch to it after a suspend, and I must restart X.  
I'm still working on isolating both of them to the kernel, and didn't 
notice them until after rc7.

Just heads-up on bugs you may see... if others who haven't been testing 
run into them as well.

-- 
Duncan - List replies preferred.   No HTML msgs.
"Every nonfree program has a lord, a master --
and if you use the program, he is your master."  Richard Stallman

-- 
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] Re: Linux 2.6.21 plans

[Greg KH]

On Fri, Apr 27, 2007 at 03:33:59AM +0000, Duncan wrote:
> Daniel Drake <dsd@gentoo.org> posted 46313C29.7050501@gentoo.org,
> excerpted below, on  Thu, 26 Apr 2007 19:56:25 -0400:
> 
> > This means that we may be pushing for 2.6.21 stable on x86 and amd64 on
> > May 17th. If important issues come up (which they may well do), this
> > will obviously be delayed, but do keep this date in mind.
> 
> As I'm sure you are aware, there are more known regressions remaining 
> this time around.  As with some others, I'm not all that confident this 
> kernel was ready for release just yet.  Oh, well, I suppose there'll be a 
> 2.6.21.1 and etc...

As no one was actually fixing any of the remaining bugs that were
reported, what were the kernel developers supposed to do, just sit
around and wait another week for no reason?

Now we have more people testing :)

And yes, this release might be a bit more unstable due to the large core
changes, but in my testing, I have had no problems.

thanks,

greg k-h
-- 
gentoo-dev@gentoo.org mailing list

[gentoo-dev] Re: Linux 2.6.21 plans

[Duncan]

Greg KH <gregkh@gentoo.org> posted 20070427041626.GA4841@kroah.com,
excerpted below, on  Thu, 26 Apr 2007 21:16:26 -0700:

> As no one was actually fixing any of the remaining bugs that were
> reported, what were the kernel developers supposed to do, just sit
> around and wait another week for no reason?

I wasn't intending to second-guess the decision (who me, a little 
nobody?), just pointing out that the ride to stability might not be so 
smooth this time, both from my experience and in the opinion of the guy 
who has been doing the regression tracking upstream...

> Now we have more people testing :)
> 
> And yes, this release might be a bit more unstable due to the large core
> changes, but in my testing, I have had no problems.

Agreed with both sentiments, even if I'm having a couple minor problems 
that could be related (unverified as yet).

-- 
Duncan - List replies preferred.   No HTML msgs.
"Every nonfree program has a lord, a master --
and if you use the program, he is your master."  Richard Stallman

-- 
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] Re: Linux 2.6.21 plans

[Daniel Drake]

Duncan wrote:
> I'm running (vanilla) rc7-git10 ATM, and have two possible regressions 
> remaining here

If reproducible on gentoo-soures-2.6.21, please file bug reports for 
them or they will get lost.

Daniel

-- 
gentoo-dev@gentoo.org mailing list

[gentoo-dev] Re: Linux 2.6.21 plans

[Duncan]

Daniel Drake <dsd@gentoo.org> posted 46321019.9000207@gentoo.org,
excerpted below, on  Fri, 27 Apr 2007 11:00:41 -0400:

> Duncan wrote:
>> I'm running (vanilla) rc7-git10 ATM, and have two possible regressions
>> remaining here
> 
> If reproducible on gentoo-soures-2.6.21, please file bug reports for
> them or they will get lost.

Ugh.  I hate it when I so publicly mis-type! =8^(  It was 2.6.21-rc7-
git8, not git10.  Anyway, the X restore thing was (apparently) fixed 
since git4, but the clock thing's still an issue.  I still have to 
eliminate a couple more things on my end, however, before bugging it.

-- 
Duncan - List replies preferred.   No HTML msgs.
"Every nonfree program has a lord, a master --
and if you use the program, he is your master."  Richard Stallman

-- 
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] Linux 2.6.21 plans

[Petteri Räty]

[-- Attachment #1: Type: text/plain, Size: 441 bytes --]

Daniel Drake kirjoitti:
>
> This means that we may be pushing for 2.6.21 stable on x86 and amd64 on
> May 17th. If important issues come up (which they may well do), this
> will obviously be delayed, but do keep this date in mind.
> 

Why would the kernel have to go stable before the usual month dictated
by policy? Yes there are usually security bugs but you did not mention
that as a reason in your post.

Regards,
Petteri


[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 252 bytes --]

Re: [gentoo-dev] Linux 2.6.21 plans

[Daniel Drake]

Petteri Räty wrote:
> Why would the kernel have to go stable before the usual month dictated
> by policy? Yes there are usually security bugs but you did not mention
> that as a reason in your post.

At last check this was a recommendation, not a policy, plus nobody 
objected timeframe-wise before.

Also, as noted in my mail I anticipate this taking more than a week from 
the point where we ask arch teams to consider stabling.

Daniel
-- 
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] Linux 2.6.21 plans

[Petteri Räty]

[-- Attachment #1: Type: text/plain, Size: 852 bytes --]

Daniel Drake kirjoitti:
> Petteri Räty wrote:
>> Why would the kernel have to go stable before the usual month dictated
>> by policy? Yes there are usually security bugs but you did not mention
>> that as a reason in your post.
> 
> At last check this was a recommendation, not a policy, plus nobody
> objected timeframe-wise before.

http://www.gentoo.org/proj/en/glep/glep-0040.html

"The package has spent a reasonable amount of time in ~arch first.
Thirty days is the usual figure, although this is clearly only a
guideline. For critical packages, a much longer duration is expected.
For small packages which have only minor changes between versions, a
shorter period is sometimes appropriate."

I would consider the kernel a critical package. Sure I could have worded
my original mail a little better.

Regards,
Petteri



[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 252 bytes --]

Re: [gentoo-dev] Linux 2.6.21 plans

[Alec Warner]

Petteri Räty wrote:
> Daniel Drake kirjoitti:
>> Petteri Räty wrote:
>>> Why would the kernel have to go stable before the usual month dictated
>>> by policy? Yes there are usually security bugs but you did not mention
>>> that as a reason in your post.
>> At last check this was a recommendation, not a policy, plus nobody
>> objected timeframe-wise before.
> 
> http://www.gentoo.org/proj/en/glep/glep-0040.html
> 
> "The package has spent a reasonable amount of time in ~arch first.
> Thirty days is the usual figure, although this is clearly only a
> guideline. For critical packages, a much longer duration is expected.
> For small packages which have only minor changes between versions, a
> shorter period is sometimes appropriate."
> 
> I would consider the kernel a critical package. Sure I could have worded
> my original mail a little better.
> 
'is expected'.

Portage is also a critical package and I doubt it's ever spent 30 days
in ~arch.  As always, maintainer knows best (and you can obviously blame
dsd if all hell breaks loose :))

-Alec
-- 
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] Linux 2.6.21 plans

[Florian D.]

Daniel Drake wrote:
> Hi,
> 
> 2.6.21 was released today. Testing muchly appreciated as usual -- please
> file bugs and clearly mark them as 2.6.21 regressions if that is the case.
> 

hello,
2.6.21 will break the current *stable* VMware worstation. VMware 6 will work again. please see the
following thread on LKML:
http://marc.info/?t=117799835200003&r=1&w=2

I didn´t file this on bugzilla, because its an upsteam thing anyway. I still think its a good idea
to upgrade to 2.6.21 -- just put a note on GWN or something.

cheers, f
-- 
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] Linux 2.6.21 plans

[Daniel Ostrow]

[-- Attachment #1: Type: text/plain, Size: 794 bytes --]

On Wed, 2007-05-09 at 01:11 +0200, Florian D. wrote:
> Daniel Drake wrote:
> > Hi,
> > 
> > 2.6.21 was released today. Testing muchly appreciated as usual -- please
> > file bugs and clearly mark them as 2.6.21 regressions if that is the case.
> > 
> 
> hello,
> 2.6.21 will break the current *stable* VMware worstation. VMware 6 will work again. please see the
> following thread on LKML:
> http://marc.info/?t=117799835200003&r=1&w=2
> 
> I didn´t file this on bugzilla, because its an upsteam thing anyway. I still think its a good idea
> to upgrade to 2.6.21 -- just put a note on GWN or something.
> 
> cheers, f

Ummm....I use VMWare Workstation every single day for wonk on my laptop,
I'm running 5.5.3.34685 and 2.6.21.1 without any issues what so ever...

--Dan

[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 189 bytes --]

Re: [gentoo-dev] Linux 2.6.21 plans

[Mike Auty]

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

Hiya,
	Reading over the discussion on lkml, it appears that it only affects
x86_64 systems...
	Mike  5:)
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.3 (GNU/Linux)

iD8DBQFGQQlpu7rWomwgFXoRAhzPAJ94Dcg/S0a6dtHodXRyPRgRT4CS0gCdHSW2
kszd0QRaPlWLg8zhoTZlc/I=
=2/I+
-----END PGP SIGNATURE-----
-- 
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] Linux 2.6.21 plans

[Daniel Ostrow]

[-- Attachment #1: Type: text/plain, Size: 396 bytes --]

On Wed, 2007-05-09 at 00:36 +0100, Mike Auty wrote:
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
> 
> Hiya,
> 	Reading over the discussion on lkml, it appears that it only affects
> x86_64 systems...
> 	Mike  5:)

Mine is an x86_64 system...it also only seems to affect early adopters
of VMWare Workstation 6, which hasn't been released yet no less
considered *stable*.

--Dan

[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 189 bytes --]

Re: [gentoo-dev] Linux 2.6.21 plans

[Florian D.]

Daniel Ostrow wrote:

> Mine is an x86_64 system...it also only seems to affect early adopters
> of VMWare Workstation 6, which hasn't been released yet no less
> considered *stable*.
> 
> --Dan

ok, then it only affects *some* amd64 users, but app-emulation/vmware-workstation-5.5.3.34685 seems
to be affected -- I can reproduce the crash here (by e.g. starting a windows update *ahem*).
-- 
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] Linux 2.6.21 plans

[Neil Bothwick]

[-- Attachment #1: Type: text/plain, Size: 489 bytes --]

On Tue, 08 May 2007 16:43:13 -0700, Daniel Ostrow wrote:

> Mine is an x86_64 system...it also only seems to affect early adopters
> of VMWare Workstation 6, which hasn't been released yet no less
> considered *stable*.

Once I tried WS6, the same fault showed up after going back to 5.5.
However, there is a fix on the VMware forums and apparently the next WS6
RC won't have this problem.


-- 
Neil Bothwick

A pessimist complains about the noise when opportunity knocks.

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 189 bytes --]