Linux Kernel 3.10 RCX has Been Released To Test - Post Your Comments Here!

So tonight, Saturday May 11, 2013 we find our next kernel 3.10-rc1 has been released for us to start testing: https://www.kernel.org/pub/linux/kernel/v3.x/testing/linux-3.10-rc1.tar.xz

And here is what Linus had to say about it: https://lkml.org/lkml/2013/5/11/155

I compiled it OK using S.A.K.C. and it loaded OK on restart and VirtualBox 4.2.12 compiled OK using DKMS, but alas, the latest nVIDIA proprietary video driver 319.17 does not compile with the new kernel and thus will not load.

Error message from dkms said at the end (/var/lib/dkms/nvidia/319.17/build/make.log):

/var/lib/dkms/nvidia/319.17/build/nv-i2c.c: In function ‘nv_i2c_del_adapter’:
/var/lib/dkms/nvidia/319.17/build/nv-i2c.c:327:14: error: void value not ignored as it ought to be
make[3]: *** [/var/lib/dkms/nvidia/319.17/build/nv-i2c.o] Error 1
make[2]: *** [_module_/var/lib/dkms/nvidia/319.17/build] Error 2
NVIDIA: left KBUILD.
nvidia.ko failed to build!
make[1]: *** [module] Error 1
make: *** [module] Error 2

For more information on compiling your own kernel, please have a look here: openSUSE and Installing New Linux Kernel Versions - Blogs - openSUSE Forums
For more info on using DKMS to reinstall VirtualBox that works with openSUSE 12.3, please look here: DKMS & systemd - How to get Dynamic Kernel Module Support to work in openSUSE 12.3 - Blogs - openSUSE Forums

Thank You,

http://www.h-online.com/imgs/43/1/0/2/2/1/8/1/kernel-internals-184499e31fabd567.jpeg

Version 3.10 of the Linux kernel will include the bcache block layer cache that allows SSD cards to be used for caching significantly slower but higher capacity hard disks. Developed by a Google employee and used at Google, bcache is the second such feature to be integrated into the Linux kernel; the first one was dm-cache, which the Linux kernel has offered since version 3.9 was released two weeks ago.

The Radeon drivers for Linux 3.10 will include interfaces to address the Unified Video Decoder (UVD) that has been part of Radeon HD graphics cards since the HD 4000 generation; suitable userspace drivers for the video accelerator are planned for the next major Mesa 3D release, which will be given version number 9.2 or 10.0.

“Full dynamic ticks” allow certain Linux 3.10 configurations to reduce the timer interrupt’s trigger frequency from between a hundred and a thousand (depending on kernel configuration) to one interrupt per second – this will reduce the kernel’s interrupt processing load. In its current form, this feature is mainly relevant for real-time systems and in High Performance Computing (HPC); further changes that are being planned will allow desktop systems to benefit from the feature in the medium term.

Linux 3.10 is expected to arrive in late June/early July, as Linus Torvalds made the first release candidate of Linux 3.10 available last weekend. This release, as usual, marked the end of the merge window phase of the development cycle during which most of the major changes for each new version are integrated into the main code base. With 11,963 commits, Torvalds and his co-developers have integrated more changes than have ever been processed in a merge window before, prompting Torvalds to declare: “this is the biggest -rc1 in the last several years (perhaps ever)”. However, diffstat’s number of new and removed lines of code is only slightly higher than in Linux 3.9 and is far from being a new record.

In addition to Linux 3.10-rc1, the developers also released long-term kernel versions 3.0.78 and 3.4.45, as well as stable kernel versions 3.8.13 and 3.9.2, last weekend. These releases offer the usual bug fixes and minor, harmless improvements.

The maintenance of Linux 3.8 will be discontinued with version 3.8.13; in the release announcement, Greg Kroah-Hartman writes that users should now migrate to kernel version 3.9. In his email, he again points out that kernel version 3.8 will not be a long-term kernel. Which particular version will be given this status will likely only emerge in the second half of 2013: if all goes according to plan, Kroah-Hartman will discontinue the maintenance of Linux 3.0 in October and award long-term status to a stable kernel that is current at that time; he will then maintain this kernel for two years.

         ([EMAIL="fab@h-online.com"]fab[/EMAIL])

Original Document: Feature set of Linux 3.10 defined - The H Open: News and Features

Copyright © 2013 Heise Media UK Ltd.

Thank You,

Tonight I have found our next kernel version to test in rc2 that you can find here: https://www.kernel.org/pub/linux/kernel/v3.x/testing/linux-3.10-rc2.tar.xz

And this is what Linus has said about it here: https://lkml.org/lkml/2013/5/20/577

I have not yet tried it out as I have been working on my new bash script Zapit, adding in RPM Database repair ability. But I will give this new RC a try very shortly. Of course, if you have anything to add, please let us hear what it is.

Thank You,


This is my experience with rc2 (from kernel:HEAD) lol!

So what percent of your PC is working well with kernel 3.10 and what percent is not and does that exceed 100% of the time you have allotted to run kernel 3.10? I can say breaking 50,000 % has been one of my goals for a while, but I can’t say just of what.

Thank You,

Compiling kernel 3.10-rc3 https://lh5.googleusercontent.com/-FsC-MCb7E9Q/UaMx_ujGspI/AAAAAAAACGk/88ga1QTnTVk/w1499-h843-no/snapshot6.png

Thanks Dale. I do see we have our next release candidate rc3 you can find here: https://www.kernel.org/pub/linux/kernel/v3.x/testing/linux-3.10-rc3.tar.xz

This is what Linus said about the release: https://lkml.org/lkml/2013/5/26/212

I have installed it without incident on my test rig and it has been loaded. I did find an error in my SAKR bash script, not expecting a kernel verion number to be 10 and I have posted a new fix for this bash script used to remove kernels no longer needed but compiled and installed using SAKC.

S.A.K.R. - SUSE Automated Kernel Remover - Version 1.0.5 - Blogs - openSUSE Forums

Thank You,

Well tonight we find our next release candidate #4 for kernel 3.10: https://www.kernel.org/pub/linux/kernel/v3.x/testing/linux-3.10-rc4.tar.xz

I had got it from the git tree, but did not see it online. Linus even indicated as much, but then tonight, we find it is online as normal. Here is what Linus has to say: https://lkml.org/lkml/2013/6/2/23

I have it compiling now, even though I could have done so sooner. I have been working on more bash scripts including a new one called mmig. You will just have to wait to find out what it is for.

Thank You,

hi,

loaded kernel linux-3.10-rc4 yesterday

anomaly noticed, only if mtt is run (part of motv package),
(even without decoding anything)
then a poweroff/reboot is not possible, error message includes,

  • Kernel panic - not syncing: Watchdog detected hard LOCKUP on cpu 3 …
  • drm_kms_helper: panic occurred, switching back to text console

this anomaly does not appear to prohibit other os operation

a similar event occurred when using any 3.9 kernel

mtt source motv-3.102-5.1.2.x86_64 package

cheers

ps. maybe no one else is using mtt

ASRock model: A780GM-LE, American Megatrends version: P1.10 date: 02/13/2009
AMD Phenom II X4 940, kernel 3.10.0-rc4-1.g51bf0ff-desktop x86_64 (64 bit),
openSUSE 12.3 (x86_64), KDE 4.10.3,
Advanced Micro Devices [AMD] nee ATI RS780 [Radeon HD 3200]
X.Org: 1.13.2 drivers: ati,radeon (unloaded: fbdev,vesa)
Resolution: 1680x1050@60.0hz, 1920x1080@60.0hz
GLX Renderer: Gallium 0.4 on AMD RS780 GLX Version: 3.0 Mesa 9.0.2

hi, FWIW the fglrx driver (amd-catalyst-13.4-linux-x86.x86_64.run) would not install with linux-3.10-rc4, it failed with the following output SUSE Paste no problems found installing with kernel 3.9.4-1.g51bf0ff-desktop cheers HP Pavilion g7-2248sg Notebook PC Hewlett-Packard model: 184B version: 57.50 Bios: Insyde version: F.17 date: 11/02/2012 AMD A8-4500M APU with Radeon HD Graphics, cpu: 1: 1400.00 MHz 2: 1400.00 MHz 3: 1400.00 MHz 4: 1400.00 MHz Advanced Micro Devices [AMD] nee ATI Trinity [Radeon HD 7640G] X.Org: 1.13.2 driver: fglrx Resolution: 1600x900@60.0hz, 1280x1024@60.0hz GLX Renderer: AMD Radeon HD 7640G GLX Version: 4.2.12217 - CPC 12.104

On 06/05/2013 10:16 AM, keellambert wrote:
>
> hi, FWIW the fglrx driver (amd-catalyst-13.4-linux-x86.x86_64.run)
> would not install with linux-3.10-rc4, it failed with the following
> output ‘SUSE Paste’ (http://paste.opensuse.org/14485775) no problems
> found installing with kernel 3.9.4-1.g51bf0ff-desktop cheers HP
> Pavilion g7-2248sg Notebook PC Hewlett-Packard model: 184B version:
> 57.50 Bios: Insyde version: F.17 date: 11/02/2012 AMD A8-4500M APU with
> Radeon HD Graphics, cpu: 1: 1400.00 MHz 2: 1400.00 MHz 3: 1400.00 MHz 4:
> 1400.00 MHz Advanced Micro Devices [AMD] nee ATI Trinity [Radeon HD
> 7640G] X.Org: 1.13.2 driver: fglrx Resolution: 1600x900@60.0hz,
> 1280x1024@60.0hz GLX Renderer: AMD Radeon HD 7640G GLX Version:
> 4.2.12217 - CPC 12.104

The proc file system changed its API between 3.9 and 3.10, which is why fglrx is
failing to compile. Someone needs to create a patch that makes the necessary
changes. If I needed that driver, I would probably do it, but my graphics
hardware is nVidia.

On 06/05/2013 11:11 AM, Larry Finger wrote:
> On 06/05/2013 10:16 AM, keellambert wrote:
>>
>> hi, FWIW the fglrx driver (amd-catalyst-13.4-linux-x86.x86_64.run)
>> would not install with linux-3.10-rc4, it failed with the following
>> output ‘SUSE Paste’ (http://paste.opensuse.org/14485775) no problems
>> found installing with kernel 3.9.4-1.g51bf0ff-desktop cheers HP
>> Pavilion g7-2248sg Notebook PC Hewlett-Packard model: 184B version:
>> 57.50 Bios: Insyde version: F.17 date: 11/02/2012 AMD A8-4500M APU with
>> Radeon HD Graphics, cpu: 1: 1400.00 MHz 2: 1400.00 MHz 3: 1400.00 MHz 4:
>> 1400.00 MHz Advanced Micro Devices [AMD] nee ATI Trinity [Radeon HD
>> 7640G] X.Org: 1.13.2 driver: fglrx Resolution: 1600x900@60.0hz,
>> 1280x1024@60.0hz GLX Renderer: AMD Radeon HD 7640G GLX Version:
>> 4.2.12217 - CPC 12.104
>
> The proc file system changed its API between 3.9 and 3.10, which is why fglrx is
> failing to compile. Someone needs to create a patch that makes the necessary
> changes. If I needed that driver, I would probably do it, but my graphics
> hardware is nVidia.

It turns out that the 304.60/4 nVidia drivers, which I need to use, have the
same problem, and I will need to fix it. Unfortunately, I don’t have time now,
and the have no need to use 3.10 at the moment.

http://www.h-online.com/open/imgs/45/1/0/3/4/3/0/2/comingin310_1_kicker-244bb9fc38876720.png

Networking

by Thorsten Leemhuis

Linux 3.10 sees improvements in the way lost packets at the end of TCP transactions are handled, speeding up HTTP data transfer. It also sees the addition of support for VLAN stacking and Realtek’s RTL8188EE wireless chip.

Last weekend, Linus Torvalds released the fourth pre-release version of Linux 3.10. Technical problems meant that it was initially only available via the git source code management system, though Torvalds notes in the email announcing the release that he considers git to be the easiest way of getting hold of it anyway. Linux 3.10-rc4 is now also available as a source code archive via kernel.org.

As usual, Torvalds and his fellow developers merged all of the major changes for the 3.10 kernel in the two weeks following the release of version 3.9. Linux 3.10 thus now finds itself in the stabilisation phase, where large-scale changes are undertaken only in exceptional cases. Therefore, we are now able to offer a fuller overview of the most important changes to be anticipated in the 3.10 kernel, which is due for release in late June/early July. We will be presenting a series of articles looking at various areas of the kernel.
http://www.h-online.com/open/imgs/45/1/0/3/4/3/0/2/184499e31fabd567.jpeg

The series opens with a description of new features relating to network support. Over the next few weeks, further articles will deal with storage technology, filesystems, graphics drivers, kernel infrastructure, processor/platform support and drivers for other hardware.

Copyright © 2013 Heise Media UK Ltd.

Read More: Kernel Log: Coming in 3.10 (Part 1) - Networking - The H Open: News and Features

Thank You,

Tonight and a little earlier than normal, we find our next kernel version to test: https://www.kernel.org/pub/linux/kernel/v3.x/testing/linux-3.10-rc5.tar.xz

And here is what our leader Linus has said about the release: https://lkml.org/lkml/2013/6/8/108

I have not had a chance to compile it yet. I was shocked to find it posted on Saturday night. In the meantime I just got a new motherboard/cpu and my main PC is well, in pieces at the moment and so I am afraid to take down my test system just yet, Perhaps in the morning if all goes well tonight.

Thank You,

So today, Sunday Afternoon June 9th, I have successfully compiled (using SAKC) and installed kernel 3.10-rc5. It is loading just fine and I can report that VirtualBox 4.2.12 is also working just fine with kernel 3.10. I have not tried out the nVIDIA driver since its update with kernel 3.10, but I will be giving that a try this week.

Thank You,

hi,

loaded kernel linux-3.10-rc5 yesterday

running mtt shows same results,

  • it can only be run one time
  • sometimes os crashes with
    — Kernel panic - not syncing: Watchdog detected hard LOCKUP on cpu X
  • after running, os can not poweroff/reboot
  • can only manually reset or unplug

all still ok running with kernel 3.8.13-1-desktop

cheers

Kernel: 3.10.0-rc5-1.g51bf0ff-desktop x86_64 (64 bit), KDE 4.10.4,
openSUSE 12.3+TW, Mobo: ASRock model: A780GM-LE, AMD Phenom II X4 940,
ATI RS780 [Radeon HD 3200], X.Org: 1.13.2 drivers: ati,radeon (unloaded: fbdev,vesa),
1680x1050@60.0hz, 1920x1080@60.0hz, Gallium 0.4 on AMD RS780, GLX 3.0 Mesa 9.1.1

Thanks for your comments. When you say mtt, did you mean mutt or just what is not working with kernel 3.10?

Thank You,

With the rate at which they are churning out kernel versions how can the DEs/Distro and other allied GNU/Linux projects keep up?

by mtt i mean mtt, is part of the motv package

motv - Video4Linux TV application (Motif)
/usr/bin/motv
/usr/bin/mtt
/usr/share/X11/app-defaults/MoTV
/usr/share/X11/app-defaults/mtt
/usr/share/X11/de_DE.UTF-8
/usr/share/X11/de_DE.UTF-8/app-defaults
/usr/share/X11/de_DE.UTF-8/app-defaults/MoTV
/usr/share/X11/fr_FR.UTF-8
/usr/share/X11/fr_FR.UTF-8/app-defaults
/usr/share/X11/fr_FR.UTF-8/app-defaults/MoTV
/usr/share/X11/it_IT.UTF-8
/usr/share/X11/it_IT.UTF-8/app-defaults
/usr/share/X11/it_IT.UTF-8/app-defaults/MoTV
/usr/share/applications/motv.desktop
/usr/share/doc/packages/motv
/usr/share/doc/packages/motv/COPYING
/usr/share/doc/packages/motv/Changes
/usr/share/doc/packages/motv/README
/usr/share/doc/packages/motv/README.bttv
/usr/share/doc/packages/motv/README.lirc
/usr/share/doc/packages/motv/README.network
/usr/share/doc/packages/motv/README.recording
/usr/share/doc/packages/motv/README.translate
/usr/share/doc/packages/motv/README.xfree4
/usr/share/doc/packages/motv/TODO
/usr/share/doc/packages/motv/frequencies-europe
/usr/share/doc/packages/motv/frequencies-japan
/usr/share/doc/packages/motv/vdr.config
/usr/share/man/man1/motv.1.gz
/usr/share/man/man1/mtt.1.gz
/usr/share/pixmaps/motv.xpm
31 files total

mtt is a tv teletext decoder
it runs ok but only once,
then it cannot be restarted again for the session

after its been run once, poweroff/reboot is not possible for the os, only manual reset or unplug

somewhere it was seen that v4l is being depreciated, possibly it has some connection?

cheeers

To answer your question you need to understand why the kernel team constantly upgrades the kernel.

  1. Handle New Hardware
  2. Add New Software Features
  3. Make the kernel and all its components run Faster
  4. To Fix Regressions and Bugs
  5. Due to support changes, as in who is handling what kernel task
  6. GPL changes of one sort or another
  7. Other undefined reasons only the Master Linus knows for sure…

That just means that at some point, sooner or later, your old hardware no longer works with the latest software. Unless you are stuck with a problem that never got fixed, that old kernel that worked for you before, still does. Do not needlessly upgrade everything if it is working OK. That even means with openSUSE on an old machine. At some point security issues or the the need for new hardware will make you dump the old stuff, but that should be up to the user as well as the user must understand old hardware will at some point, stop working if you feel the need to upgrade your software all of the time.

Thank You,