[arch-general] [signoff] kernel26 2.6.34.2-1 - Arch-general - lists.archlinux.org

newer
[arch-general] Packages Translate.

[arch-general] [signoff] kernel26 2.6.34.2-1

older
[arch-general] Initial AIF...

Tobias Powalowski

3 Aug 2010 3 Aug '10

3:39 p.m.

Latest kernel is in testing, please signoff for both arches. I already have .35 prepared, a fast signoff would be great. greetings tpowa -- Tobias Powalowski Archlinux Developer & Package Maintainer (tpowa) http://www.archlinux.org tpowa@archlinux.org

Attachments:

signature.asc (application/pgp-signature — 198 bytes)

Show replies by date

Lukáš Jirkovský

3 Aug 3 Aug

4:23 p.m.

On 3 August 2010 09:39, Tobias Powalowski wrote:

Latest kernel is in testing, please signoff for both arches.

I already have .35 prepared, a fast signoff would be great.

greetings tpowa -- Tobias Powalowski Archlinux Developer & Package Maintainer (tpowa) http://www.archlinux.org tpowa@archlinux.org

I get following error on x86_64: (1/1) checking for file conflicts [####################] 100% error: failed to commit transaction (conflicting files) kernel26: /lib/modules/2.6.34-ARCH/modules.devname exists in filesystem kernel26: /lib/modules/2.6.34-ARCH/modules.softdep exists in filesystem but it may be due to the experiments with kernel26-rc I did earlier today.

Tobias Powalowski

4:24 p.m.

On 3 August 2010 09:39, Tobias Powalowski wrote:

...
Latest kernel is in testing,

please signoff for both arches.

...
I already have .35 prepared,

a fast signoff would be great.

...
greetings tpowa --

Tobias

...
Powalowski Archlinux Developer & Package Maintainer

(tpowa)

...
http://www.archlinux.org tpowa@archlinux.org

I get following error on x86_64: (1/1) checking for file conflicts [####################] 100% error: failed to commit transaction (conflicting files) kernel26: /lib/modules/2.6.34-ARCH/modules.devname exists in filesystem kernel26: /lib/modules/2.6.34-ARCH/modules.softdep exists in filesystem

but it may be due to the experiments with kernel26-rc I did earlier today.

Am Dienstag 03 August 2010 schrieb Lukáš Jirkovský: pacman -Qo yourfiles to see which package owns those files. -- Tobias Powalowski Archlinux Developer & Package Maintainer (tpowa) http://www.archlinux.org tpowa@archlinux.org

Lukáš Jirkovský

4:31 p.m.

On 3 August 2010 10:24, Tobias Powalowski wrote:

...
On 3 August 2010 09:39, Tobias Powalowski wrote:

...
Latest kernel is in testing,

please signoff for both arches.

...
I already have .35 prepared,

a fast signoff would be great.

...
greetings tpowa --

Tobias

...
Powalowski Archlinux Developer & Package Maintainer

(tpowa)

...
http://www.archlinux.org tpowa@archlinux.org

I get following error on x86_64: (1/1) checking for file conflicts [####################] 100% error: failed to commit transaction (conflicting files) kernel26: /lib/modules/2.6.34-ARCH/modules.devname exists in filesystem kernel26: /lib/modules/2.6.34-ARCH/modules.softdep exists in filesystem

but it may be due to the experiments with kernel26-rc I did earlier today.

Am Dienstag 03 August 2010 schrieb Lukáš Jirkovský: pacman -Qo yourfiles to see which package owns those files.

-- Tobias Powalowski Archlinux Developer & Package Maintainer (tpowa) http://www.archlinux.org tpowa@archlinux.org

It wasn't owned by anything. But since it works fine on a different computer with i686 I guess it was my fault. Sign off i686. Sign off x86_64. (except for the minor glitch when updating which was most likely my fault). Lukas

Evangelos Foutras

4:51 p.m.

2010/8/3 Lukáš Jirkovský :

On 3 August 2010 09:39, Tobias Powalowski wrote:

...
Latest kernel is in testing, please signoff for both arches.

I already have .35 prepared, a fast signoff would be great.

greetings tpowa -- Tobias Powalowski Archlinux Developer & Package Maintainer (tpowa) http://www.archlinux.org tpowa@archlinux.org

I get following error on x86_64: (1/1) checking for file conflicts [####################] 100% error: failed to commit transaction (conflicting files) kernel26: /lib/modules/2.6.34-ARCH/modules.devname exists in filesystem kernel26: /lib/modules/2.6.34-ARCH/modules.softdep exists in filesystem

but it may be due to the experiments with kernel26-rc I did earlier today.

I get this too (on x86_64; didn't check on i686), when trying to upgrade from 2.6.34.1-1. Also, no package owns those two files, but the rest of the /lib/modules/2.6.34-ARCH/modules.* files are owned by kernel26 2.6.34.1-1.

Thomas Bächler

9:28 p.m.

Am 03.08.2010 10:51, schrieb Evangelos Foutras:

...
I get following error on x86_64: (1/1) checking for file conflicts [####################] 100% error: failed to commit transaction (conflicting files) kernel26: /lib/modules/2.6.34-ARCH/modules.devname exists in filesystem kernel26: /lib/modules/2.6.34-ARCH/modules.softdep exists in filesystem

but it may be due to the experiments with kernel26-rc I did earlier today.

I get this too (on x86_64; didn't check on i686), when trying to upgrade from 2.6.34.1-1. Also, no package owns those two files, but the rest of the /lib/modules/2.6.34-ARCH/modules.* files are owned by kernel26 2.6.34.1-1.

module-init-tools got updated recently, it tends to create new files on 'depmod' that we don't track yet. This is unfortunate, it means some users need to -Sf a kernel update every now and then.

Simon Stoakley

11:14 p.m.

On 03/08/10 09:51, Evangelos Foutras wrote:

2010/8/3 Lukáš Jirkovský:

...
On 3 August 2010 09:39, Tobias Powalowski wrote:

...
Latest kernel is in testing, please signoff for both arches.

I already have .35 prepared, a fast signoff would be great.

greetings tpowa -- Tobias Powalowski Archlinux Developer& Package Maintainer (tpowa) http://www.archlinux.org tpowa@archlinux.org

I get following error on x86_64: (1/1) checking for file conflicts [####################] 100% error: failed to commit transaction (conflicting files) kernel26: /lib/modules/2.6.34-ARCH/modules.devname exists in filesystem kernel26: /lib/modules/2.6.34-ARCH/modules.softdep exists in filesystem

but it may be due to the experiments with kernel26-rc I did earlier today.

I get this too (on x86_64; didn't check on i686), when trying to upgrade from 2.6.34.1-1. Also, no package owns those two files, but the rest of the /lib/modules/2.6.34-ARCH/modules.* files are owned by kernel26 2.6.34.1-1.

I also get this, if it's safe to pacman -Sf this, wouldn't it be a good idea to put up a news item,(as i think someone suggested earlier) to stop the forum getting flodded with posts and also to let people know (it is the kernel after all). Just a thought.

Mario Figueiredo

4 Aug 4 Aug

1:43 a.m.

On 03-08-2010 16:14, Simon Stoakley wrote:

On 03/08/10 09:51, Evangelos Foutras wrote:

...
2010/8/3 Lukáš Jirkovský:

...
On 3 August 2010 09:39, Tobias Powalowski wrote:

...
Latest kernel is in testing, please signoff for both arches.

I already have .35 prepared, a fast signoff would be great.

greetings tpowa -- Tobias Powalowski Archlinux Developer& Package Maintainer (tpowa) http://www.archlinux.org tpowa@archlinux.org

I get following error on x86_64: (1/1) checking for file conflicts [####################] 100% error: failed to commit transaction (conflicting files) kernel26: /lib/modules/2.6.34-ARCH/modules.devname exists in filesystem kernel26: /lib/modules/2.6.34-ARCH/modules.softdep exists in filesystem

but it may be due to the experiments with kernel26-rc I did earlier today.

I get this too (on x86_64; didn't check on i686), when trying to upgrade from 2.6.34.1-1. Also, no package owns those two files, but the rest of the /lib/modules/2.6.34-ARCH/modules.* files are owned by kernel26 2.6.34.1-1.

I also get this, if it's safe to pacman -Sf this, wouldn't it be a good idea to put up a news item,(as i think someone suggested earlier) to stop the forum getting flodded with posts and also to let people know (it is the kernel after all). Just a thought.

I'd indeed appreciate if a news item was generated with a little more detailed explanation of how to handle with the error. Because so far I don't understand from both this and arch-dev-public's thread on the same matter.

Stefano Avallone

3 Aug 3 Aug

10:19 p.m.

New subject: [arch-general] [signoff] kernel26 2.6.34.2-1 & intel pageflipping

On Tuesday 03 August 2010 10:23:19 Lukáš Jirkovský wrote:

On 3 August 2010 09:39, Tobias Powalowski wrote:

...
Latest kernel is in testing, please signoff for both arches.

I already have .35 prepared, a fast signoff would be great.

greetings tpowa -- Tobias Powalowski Archlinux Developer & Package Maintainer (tpowa) http://www.archlinux.org tpowa@archlinux.org

There are a number of fixes about pageflipping with drm/i915 in the changelog of 2.6.34.2. So, I installed this new kernel and rebuilt xf86-video-intel without the patch to disable pageflipping. I have a G45 graphics and I suffered from frequent hangs with previous kernel versions and a 2D driver with pageflipping enabled. Now, using 2.6.34.2 and pageflipping enabled, I had my desktop running for 6 hours and no hangs occurred so far. I'll keep it running and report back if an hang occurs. cheers, Stefano

Thomas Bächler

4 Aug 4 Aug

1:29 a.m.

New subject: [arch-general] [signoff] kernel26 2.6.34.2-1 & intel pageflipping

Am 03.08.2010 16:19, schrieb Stefano Avallone:

There are a number of fixes about pageflipping with drm/i915 in the changelog of 2.6.34.2. So, I installed this new kernel and rebuilt xf86-video-intel without the patch to disable pageflipping. I have a G45 graphics and I suffered from frequent hangs with previous kernel versions and a 2D driver with pageflipping enabled. Now, using 2.6.34.2 and pageflipping enabled, I had my desktop running for 6 hours and no hangs occurred so far. I'll keep it running and report back if an hang occurs.

Thank you for this, please keep reporting it. We actually don't want to have page flipping disabled indefinitely.

Stefano Avallone

8 Aug 8 Aug

11:35 p.m.

New subject: [arch-general] [signoff] kernel26 2.6.34.2-1 & intel pageflipping

On Tuesday 03 August 2010 19:29:32 Thomas Bächler wrote:

Am 03.08.2010 16:19, schrieb Stefano Avallone:

...
There are a number of fixes about pageflipping with drm/i915 in the changelog of 2.6.34.2. So, I installed this new kernel and rebuilt xf86-video-intel without the patch to disable pageflipping. I have a G45 graphics and I suffered from frequent hangs with previous kernel versions and a 2D driver with pageflipping enabled. Now, using 2.6.34.2 and pageflipping enabled, I had my desktop running for 6 hours and no hangs occurred so far. I'll keep it running and report back if an hang occurs.

Thank you for this, please keep reporting it. We actually don't want to have page flipping disabled indefinitely.

I continued using my desktop PC with intel G45 graphics, kernel 2.6.34.2 and intel driver with pageflipping enabled and no hangs occurred. Everything seems to work fine. I also compiled the intel 2D driver with pageflipping enabled on a notebook with intel GM965 graphics. I noticed an annoying issue: every few minutes (part of) the screen went black for an istant, maybe during a refresh. Then, I installed kernel 2.6.35 and this issue seems to be gone. I hope more people could test it, but I believe it is safe to enable pageflipping in the intel xorg driver with kernel 2.6.35. cheers, Stefano

Dan McGee

4 Aug 4 Aug

10:24 a.m.

On Tue, Aug 3, 2010 at 2:39 AM, Tobias Powalowski wrote:

Latest kernel is in testing, please signoff for both arches.

I already have .35 prepared, a fast signoff would be great.

This completely broke wireless for me on my laptop. b43 driver, i686 architecture. The wlan0 device would show up, but then you couldn't do much of anything with it. 03:00.0 Network controller: Broadcom Corporation BCM4312 802.11b/g (rev 01) Subsystem: Hewlett-Packard Company Device 365e -Dan

Dan McGee

10:34 a.m.

On Tue, Aug 3, 2010 at 9:24 PM, Dan McGee wrote:

On Tue, Aug 3, 2010 at 2:39 AM, Tobias Powalowski wrote:

...
Latest kernel is in testing, please signoff for both arches.

I already have .35 prepared, a fast signoff would be great.

This completely broke wireless for me on my laptop. b43 driver, i686 architecture. The wlan0 device would show up, but then you couldn't do much of anything with it.

03:00.0 Network controller: Broadcom Corporation BCM4312 802.11b/g (rev 01) Subsystem: Hewlett-Packard Company Device 365e

Interesting, because the driver hasn't changed at all since the base 2.6.34 was released: dmcgee@galway ~/projects/linux-2.6 (master) $ (find -name b43*) | xargs git diff v2.6.34..v2.6.34.2 | cat <no output here> The diff on drivers/net/wireless didn't look odd either; specific (other) drivers changed but nothing that sticks out as affecting b43. -Dan

Ionuț Bîru

2:03 p.m.

On 08/04/2010 05:34 AM, Dan McGee wrote:

On Tue, Aug 3, 2010 at 9:24 PM, Dan McGee wrote:

...
On Tue, Aug 3, 2010 at 2:39 AM, Tobias Powalowski wrote:

...
Latest kernel is in testing, please signoff for both arches.

I already have .35 prepared, a fast signoff would be great.

This completely broke wireless for me on my laptop. b43 driver, i686 architecture. The wlan0 device would show up, but then you couldn't do much of anything with it.

03:00.0 Network controller: Broadcom Corporation BCM4312 802.11b/g (rev 01) Subsystem: Hewlett-Packard Company Device 365e

Interesting, because the driver hasn't changed at all since the base 2.6.34 was released:

dmcgee@galway ~/projects/linux-2.6 (master) $ (find -name b43*) | xargs git diff v2.6.34..v2.6.34.2 | cat <no output here>

The diff on drivers/net/wireless didn't look odd either; specific (other) drivers changed but nothing that sticks out as affecting b43.

-Dan

it has ssb changes and b43 relies on it -- Ionuț

Dan McGee

7:23 p.m.

On Wed, Aug 4, 2010 at 1:03 AM, Ionuț Bîru wrote:

On 08/04/2010 05:34 AM, Dan McGee wrote:

...
On Tue, Aug 3, 2010 at 9:24 PM, Dan McGee wrote:

...
On Tue, Aug 3, 2010 at 2:39 AM, Tobias Powalowski wrote:

...
Latest kernel is in testing, please signoff for both arches.

I already have .35 prepared, a fast signoff would be great.

This completely broke wireless for me on my laptop. b43 driver, i686 architecture. The wlan0 device would show up, but then you couldn't do much of anything with it.

03:00.0 Network controller: Broadcom Corporation BCM4312 802.11b/g (rev 01) Subsystem: Hewlett-Packard Company Device 365e

Interesting, because the driver hasn't changed at all since the base 2.6.34 was released:

dmcgee@galway ~/projects/linux-2.6 (master) $ (find -name b43*) | xargs git diff v2.6.34..v2.6.34.2 | cat <no output here>

The diff on drivers/net/wireless didn't look odd either; specific (other) drivers changed but nothing that sticks out as affecting b43.

-Dan

it has ssb changes and b43 relies on it

Thanks, Ionuț. So much for a stable release...I should probably let them know upstream, no? Is this something to send to linux-stable? -Dan

Ionuț Bîru

7:26 p.m.

On 08/04/2010 02:23 PM, Dan McGee wrote:

On Wed, Aug 4, 2010 at 1:03 AM, Ionuț Bîru wrote:

...
On 08/04/2010 05:34 AM, Dan McGee wrote:

...
On Tue, Aug 3, 2010 at 9:24 PM, Dan McGee wrote:

...
On Tue, Aug 3, 2010 at 2:39 AM, Tobias Powalowski wrote:

...
Latest kernel is in testing, please signoff for both arches.

I already have .35 prepared, a fast signoff would be great.

This completely broke wireless for me on my laptop. b43 driver, i686 architecture. The wlan0 device would show up, but then you couldn't do much of anything with it.

03:00.0 Network controller: Broadcom Corporation BCM4312 802.11b/g (rev 01) Subsystem: Hewlett-Packard Company Device 365e

Interesting, because the driver hasn't changed at all since the base 2.6.34 was released:

dmcgee@galway ~/projects/linux-2.6 (master) $ (find -name b43*) | xargs git diff v2.6.34..v2.6.34.2 | cat <no output here>

The diff on drivers/net/wireless didn't look odd either; specific (other) drivers changed but nothing that sticks out as affecting b43.

-Dan

it has ssb changes and b43 relies on it

Thanks, Ionuț. So much for a stable release...I should probably let them know upstream, no? Is this something to send to linux-stable?

-Dan

it was reported and we have a patch available http://bugs.archlinux.org/task/20353 -- Ionuț

Dan McGee

7:28 p.m.

On Wed, Aug 4, 2010 at 6:26 AM, Ionuț Bîru wrote:

On 08/04/2010 02:23 PM, Dan McGee wrote:

...
On Wed, Aug 4, 2010 at 1:03 AM, Ionuț Bîru wrote:

...
On 08/04/2010 05:34 AM, Dan McGee wrote:

...
On Tue, Aug 3, 2010 at 9:24 PM, Dan McGee wrote:

...
On Tue, Aug 3, 2010 at 2:39 AM, Tobias Powalowski wrote:

...
Latest kernel is in testing, please signoff for both arches.

I already have .35 prepared, a fast signoff would be great.

This completely broke wireless for me on my laptop. b43 driver, i686 architecture. The wlan0 device would show up, but then you couldn't do much of anything with it.

03:00.0 Network controller: Broadcom Corporation BCM4312 802.11b/g (rev 01) Subsystem: Hewlett-Packard Company Device 365e

Interesting, because the driver hasn't changed at all since the base 2.6.34 was released:

dmcgee@galway ~/projects/linux-2.6 (master) $ (find -name b43*) | xargs git diff v2.6.34..v2.6.34.2 | cat <no output here>

The diff on drivers/net/wireless didn't look odd either; specific (other) drivers changed but nothing that sticks out as affecting b43.

-Dan

it has ssb changes and b43 relies on it

Thanks, Ionuț. So much for a stable release...I should probably let them know upstream, no? Is this something to send to linux-stable?

-Dan

it was reported and we have a patch available

http://bugs.archlinux.org/task/20353

Awesome, good to see. It might be a pain in the ass but this is going to get reported again and again if we don't 1) build 2.6.34.3 when it comes out, hopefully soon or 2) Rebuild 2.6.34.2 with this patch. Loosing networking really sucks the big one. To Thomas: they are gong to have to given that this better end up in the next stable tree for that kernel. -Dan

Thomas Bächler

7:56 p.m.

Am 04.08.2010 13:28, schrieb Dan McGee:

On Wed, Aug 4, 2010 at 6:26 AM, Ionuț Bîru wrote:

...
it was reported and we have a patch available

http://bugs.archlinux.org/task/20353

Awesome, good to see. It might be a pain in the ass but this is going to get reported again and again if we don't 1) build 2.6.34.3 when it comes out, hopefully soon or 2) Rebuild 2.6.34.2 with this patch. Loosing networking really sucks the big one.

I am already rebuilding both kernels. This screwup sucks a lot.

To Thomas: they are gong to have to given that this better end up in the next stable tree for that kernel.

They should be careful what they merge in the future. The stable releases should never introduce regressions.

Dan McGee

7:58 p.m.

On Wed, Aug 4, 2010 at 6:56 AM, Thomas Bächler wrote:

Am 04.08.2010 13:28, schrieb Dan McGee:

...
On Wed, Aug 4, 2010 at 6:26 AM, Ionuț Bîru wrote:

...
it was reported and we have a patch available

http://bugs.archlinux.org/task/20353

Awesome, good to see. It might be a pain in the ass but this is going to get reported again and again if we don't 1) build 2.6.34.3 when it comes out, hopefully soon or 2) Rebuild 2.6.34.2 with this patch. Loosing networking really sucks the big one.

I am already rebuilding both kernels. This screwup sucks a lot.

...
To Thomas: they are gong to have to given that this better end up in the next stable tree for that kernel.

They should be careful what they merge in the future. The stable releases should never introduce regressions.

I completely agree. I appreciate your effort in doing this, it might go rather silently but it isn't unnoticed. -Dan

Thomas Bächler

7:27 p.m.

Am 04.08.2010 13:23, schrieb Dan McGee:

...
it has ssb changes and b43 relies on it

Thanks, Ionuț. So much for a stable release...I should probably let them know upstream, no? Is this something to send to linux-stable?

This is the second time they introduced a regression within a stable series _that I know of_. I am sure Greg and Linus would both love to know about this.

Ionuț Bîru

2:06 p.m.

On 08/04/2010 05:34 AM, Dan McGee wrote:

On Tue, Aug 3, 2010 at 9:24 PM, Dan McGee wrote:

...
On Tue, Aug 3, 2010 at 2:39 AM, Tobias Powalowski wrote:

...
Latest kernel is in testing, please signoff for both arches.

I already have .35 prepared, a fast signoff would be great.

This completely broke wireless for me on my laptop. b43 driver, i686 architecture. The wlan0 device would show up, but then you couldn't do much of anything with it.

03:00.0 Network controller: Broadcom Corporation BCM4312 802.11b/g (rev 01) Subsystem: Hewlett-Packard Company Device 365e

Interesting, because the driver hasn't changed at all since the base 2.6.34 was released:

dmcgee@galway ~/projects/linux-2.6 (master) $ (find -name b43*) | xargs git diff v2.6.34..v2.6.34.2 | cat <no output here>

The diff on drivers/net/wireless didn't look odd either; specific (other) drivers changed but nothing that sticks out as affecting b43.

-Dan

note to mention that with 2.6.35 kernel broadcom is working. i guess they did a partial merge from 2.6.35 to 2.6.34 -- Ionuț

5017

Age (days ago)

5022

Last active (days ago)

Download

20 comments

9 participants

tags

participants (9)

Dan McGee
Evangelos Foutras
Ionuț Bîru
Lukáš Jirkovský
Mario Figueiredo
Simon Stoakley
Stefano Avallone
Thomas Bächler
Tobias Powalowski