Linux-ck/Changelog

From ArchWiki
< Linux-ck
Revision as of 00:53, 20 September 2011 by Karol (Talk | contribs) (creating changelog subpage)

(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to: navigation, search

31-Aug-2011 linux-ck-3.0.4-1

Bump to latest version

http://projects.archlinux.org/svntogit/packages.git/commit/linux?id=0e64789b388e39515964f046fbaf366de3456d96

20-Aug-2011 linux-ck-3.0.3-4

Bump to 3.0.3-4

Why? Pacman 3.5.4-4 just hit [core] and included a significant change to the CFLAGS. I am simply bumping to compile again with these new flags for the repo. Why no 3.0.3-3? I goofed on it and neglected to make the change to the PKGBUILD aside from just bumping the pkgver AND I got about 80 % through recompiling all 17 repo package sets before I knew. Anyway, -3 also went into the AUR and one of the users (ethail) pointed out the error I made. Anyway, rebuilding and will publish soon.

18-Aug-2011 linux-ck-3.0.3-2

What's different? See line 173 of the PKGBUILD. Added the following into the make step for bzImage modules modules_install: KCFLAGS="-march=native -O2 -pipe" KCPPFLAGS="-march=native -O2 -pipe"

It is my understanding that this will impart CPU-specific optimization in the kernel beyond those selected by the config defining the kernel type. Enjoy!

17-Aug-2011 linux-ck 3.0.3-1

Bump to 3.0.3

16-Aug-2011 linux-ck 3.0.2-1

Bump to 3.0.2

13-Aug-2011 linux-ck 3.0.1-3

Forgot to enable the Arch boot logo in the -2 build. The purpose of the -3 build is to have this enabled. This is the only change (trivial I know but I am somewhat OCD).

11-Aug-2011 linux-ck 3.0.1-2

Officially released incorporating CK's 3.0.0-ck1 patchset with BFS v0.406!

22-Jul-2011 linux-ck 3.0.0-1

It lives! Uploaded a draft PKG as a placeholder waiting for CK and Paolo to release their respective patchsets for the 3,x release.


19-Jul-2011 kernel26-ck 2.6.39.3-3

Bump to 2.6.39.3-3

This release is more or less identical to the previous one with one key difference: the virtualbox modules won't compile bug has been fixed. Full details on page 14 of the discussion thread. Packages should be online around 6:00 EST.

14-Jul-2011 kernel26-ck 2.6.39.3-2

Bump to 2.6.39.3-2

No internal differences between the -1 and -2 release. The changes are to the packages, not their contents. xclaude made me aware of a small bug in my packaging script that didn't get the conflict arrays right for some packages. I fixed it but it required me bumping the package to -2 and rebuilding. So here we are with -2 :)

09-Jul-2011 kernel26-ck 2.6.39.3-1

Bump to 2.6.39.3-1

27-Jun-2011 kernel26-ck 2.6.39.2-1

Bump to 2.6.39.2-1

11-Jun-2011 kernel26-ck 2.6.39.1-3

Bump to 2.6.39.1-3

I had to bump to -3 because I added an additional arch to the repo (k10 optimized) and the only way I would make sure that all packages got an update to their respective conflicts arrays was to rebuild EVERYTHING. Repo users will get the update but AUR users have no reason to build/update unless you like to watch kernel packages compile :)

The repo now hosts the following packages/CPU-specific optimized builds.

kernel, kernel-headers, nvidia, and broadcom-wl in the following flavors:

  • x86_64: generic, atom, core2, k8, k10
  • i686: generic, atom, core2, k7, k8, k10, p4, pentm


09-Jun-2011 kernel26-ck 2.6.39.1-2

Bump to 2.6.39.1-2

The ONLY difference is that this PKGBUILD gives a new option to enable the "classic" blue Arch logo at boot time which will replace the Linux Tux logo we've had since tpowa removed the custom logo in 2.6.38.8. Repo users have this enabled by default by popular demand.

06-Jun-2011 kernel26-ck 2.6.39.1-1

Bump to latest version.

05-Jun-2011 kernel26-ck 2.6.39-10

Bump to latest version bfs which is v0.406 as well the ck2 patchset. Previous instabilities with plugged I/O flushing have been fixed.

http://ck-hack.blogspot.com/2011/06/2639-ck2-bfs-0406.html

02-Jun-2011 kernel26-ck 2.6.38-9

Looks like CK made some progress on the hardlock issues affecting 2.6.39 by releasing test9 which I have added to 2.6.39-9 that just published to the AUR. If you elect to build this kernel, please report your experiences, good and bad, to CK on his blog which will help the community: http://ck-hack.blogspot.com/2011/06/2639-bfs-test-9-is-this-one.html

Note that the repo doesn't get this since the patch is an option. Once CK calls the code "stable" he will iterate the bfs version and I will then include it.

30-May-2011 kernel26-ck 2.6.38-8

Looks like CK made some progress on the hardlock issues affecting 2.6.39 by releasing bfs404-recheck_unplugged which I have added to 2.6.39-8 that just published to the AUR. If you elect to build this kernel, please report your experiences, good and bad, to CK on his blog which will help the community: http://ck-hack.blogspot.com/2011/05/2639-bfs-progress.html

Note that the repo doesn't get this since the patch is an option. Once CK calls the code "stable" he will iterate the bfs version and I will then include it.

26-May-2011 kernel26-ck 2.6.39-7

Bump to 2.6.39-7

Includes an option to compile with ck's new bfs v.404 test7 patch for those of us with these random hardlocks associated with Virtualbox and Chromium. This patch tweaks the new plugged I/O flushing. For more or to discuss this patch, see: http://ck-hack.blogspot.com/2011/05/2639-ck1.html

Note that the repo doesn't get this since the patch is an option. Once CK calls the code "stable" he will iterate the bfs version and I will then include it.

25-May-2011 kernel26-ck 2.6.39-5

Bump to 2.6.39-5

Includes an option to compile with ck's new bfs v.404 test5 patch for those of us with these random hardlocks associated with Virtualbox and Chromium. This patch tweaks the new plugged I/O flushing. For more or to discuss this patch, see: http://ck-hack.blogspot.com/2011/05/2639-ck1.html

Note that the repo doesn't get this since the patch is an option. Once CK calls the code "stable" he will iterate the bfs version and I will then include it.


25-May-2011 kernel26-ck 2.6.39-4

Bump to 2.6.39-4

Includes an option to compile with ck's new bfs v.404 test4 patch for those of us with these random hardlocks associated with Virtualbox and Chromium. This patch tweaks the new plugged I/O flushing. For more or to discuss this patch, see: http://ck-hack.blogspot.com/2011/05/2639-ck1.html

Note that the repo doesn't get this since the patch is an option. Once CK calls the code "stable" he will iterate the bfs version and I will then include it.

24-May-2011 kernel26-ck 2.6.39-3

Bump to 2.6.39-3

Includes an option to compile with ck's new bfs v.404 test3 patch for those of us with these random hardlocks associated with Virtualbox and Chromium. This patch tweaks the new plugged I/O flushing. For more or to discuss this patch, see: http://ck-hack.blogspot.com/2011/05/2639-ck1.html

Note that the repo doesn't get this since the patch is an option. Once CK calls the code "stable" he will iterate the bfs version and I will then include it.

23-May-2011 kernel26-ck 2.6.39-3

Bump to 2.6.39-2

Includes an option to compile with ck's new bfs v.404 test2 patch for those of us with these random hardlocks associated with Virtualbox and Chromium. This patch tweaks the new plugged I/O flushing. For more or to discuss this patch, see: http://ck-hack.blogspot.com/2011/05/2639-ck1.html

Note that the repo doesn't get this since the patch is an option. Once CK calls the code "stable" he will iterate the bfs version and I will then include it.

20-May-2011 kernel26-ck 2.6.39-1

Bump to latest version from mainline. Also includes bfs v0.404 from ck. For details, see: http://ck-hack.blogspot.com/2011/05/2639-ck1.html


13-May-2011 kernel26-ck 2.6.38.6-3

Bump to latest version with a fixed libata regression http://projects.archlinux.org/svntogit/packages.git/commit/?id=82d1e0060dfe93404bd54305da18469d51e4ee13

For repo-only versions, the package now comes with the BFQ I/O scheduler built as a module which users can enable at will. Read the section for instructions to load it and enable it.


11-May-2011 kernel26-ck 2.6.38.6-2

Minor change to the config and config.x86_64: set the CONFIG_FRAME_POINTER option from enabled to disabled. Result will be a slightly smaller and faster kernel image without kernel debugging information (precise oopses/stacktraces/warnings). Note that as of gcc 4.6.0+, the -O2 switch adds the corresponding -fomit-frame-pointer switch to your CFLAGS, so this kernel option seems a bit redundant.

10-May-2011 kernel26-ck 2.6.38.6-1

Bump to latest version. http://projects.archlinux.org/svntogit/packages.git/commit/?id=6d930e1582988016b3f6ac119ffcbfe0bfb4809b


07-May-2011 kernel26-ck 2.6.38.5-2

Added a line to the PKGBUILD that appends user defined CFLAGS to the KCFLAGS since, as falconindy pointed out in this thread (see post #185), the kernel does not use the userland CFLAGS defined in Template:Filename as we thought.


03-May-2011 kernel26-ck 2.6.38.5-1

Bump to latest version. http://projects.archlinux.org/svntogit/packages.git/commit/?id=8a62f9f25cdbe754fa70d448ad7dea6287f2ff71

22-Apr-2011 kernel26-ck 2.6.38.4-1

Bump to latest version. http://projects.archlinux.org/svntogit/packages.git/commit/?id=31611f00a15d7f1ffc3317afcbc63d6158cea457

17-Apr-2011 kernel26-ck 2.6.38.3-1

Bump to latest version. http://projects.archlinux.org/svntogit/packages.git/commit/?id=f13814c4550d17d05d2cc36bd01596aecc24b5aa

Also, updated BFS from v0.400 --> v0.401

15-Apr-2011 kernel26-ck 2.6.38.2-6

Major f*ckup on my part: I included the WRONG config/config.x86_64 when I packaged the -5 release. The 1000 Hz/tickless=n options were still there. This release has the correct settings! Once you update and reboot, you can verify that you have tickless enabled (CONFIG_NO_HZ=y) and that the ARCH default value of 300 Hz is in use (CONFIG_HZ_300=y and CONFIG_HZ=300) from the following command:

$ zcat /proc/config.gz | grep HZ

14-Apr-2011 kernel26-ck 2.6.38.2-5

Reverted to the ARCH default configs (plus BFS) based on several reports from the community regarding significant decreases in battery life for laptop users.

13-Apr-2011 kernel26-ck 2.6.38.2-4

Changed the configs based on CK's suggestions:

  1. Disabled dynamic ticks (ARCH default is enabled).
  2. Tick rate = 1000 Hz (ARCH default is 300 Hz).

All other options are set to the ARCH defaults outlined in the main kernel's config files. Why the changes? See CK's [BFS configuration FAQ]. These settings are based on his recommended configuration for "Desktop."

10-Apr-2011 kernel26-ck 2.6.38.2-3

Added official CK3 patchset which includes the BFS v0.400. Functionally, there is no difference between -3 and -2. http://ck-hack.blogspot.com/2011/04/bfs-0400.html

10-Apr-2011 kernel26-ck 2.6.38.2-2

Added BFS v0.400

30-Mar-2011 kernel26-ck 2.6.38.2-1

Bump to latest version git-svn-id: eb2447ed-0c53-47e4-bac8-5bc4a241df78

29-Mar-2011 kernel26-ck 2.6.38.1-3

Downgraded to ck1 patchset due to regression introduced by the BFS v0.370. For more, see CK's blog: http://ck-hack.blogspot.com/2011/03/2638-ck1-bfs-0363.html

26-Mar-2011 kernel26-ck 2.6.38.1-2

Added CK's bfs363-test2 patch to the standard ck1 patchset. This further refines the bfs on a per-cpu frequency scaling basis and is enabled by default. See CK's blog entry: http://ck-hack.blogspot.com/2011/03/bfs-and-optimising-global-scheduler-for.html

Files affected:

Excerpt from the blog, "Not only will the workload complete faster, but almost certainly battery usage will be reduced. It also has no effect without cpu frequency scaling, and no effect on single CPU machines, but the more cores and threads you have, the greater the benefit. Since some new Linux distributions now no longer even allow you to change the governor and just set it to ondemand by default, this change is something that will be essential."

25-Mar-2011 kernel26-ck 2.6.38.1-1

bump to latest version git-svn-id: eb2447ed-0c53-47e4-bac8-5bc4a241df78

Features included: - latest stable patches - disabled /dev/kmem - added AMD_IOMMU support - kernel image is now xz compressed - NUMA is enabled on x86_64 - AUTOSCHED (aka the wonder patch) is enabled - aufs2.1 latest snapshot - added additional i915 patch - added radeaon kms fix

08-Mar-2011 kernel26-ck 2.6.37.3-1

bump to latest version git-svn-id: eb2447ed-0c53-47e4-bac8-5bc4a241df78


08-Mar-2011 kernel26-ck 2.6.37.3-1

Bump to latest version git-svn-id: eb2447ed-0c53-47e4-bac8-5bc4a241df78

25-Feb-2011 kernel26-ck 2.6.37.2-1

Bump to latest version git-svn-id: eb2447ed-0c53-47e4-bac8-5bc4a241df78

Note: Users will need to have util-linux-2.19-3 and mkinitcpio-0.6.8-2 to use this release.

23-Feb-2011 kernel26-ck 2.6.37.1-2

Bump dependency to avoid util-linux breakage.

Note: Users will need to have util-linux-2.19-2 and mkinitcpio-0.6.8-2 to use this release.

19-Feb-2011 kernel26-ck 2.6.37.1-1

Bump to latest version - thanks to koeleck for providing [this patch] to fix the ck2 patchset!

13-Feb-2011 kernel26-ck 2.6.37-7

Includes ck2 patchset not ck1 patchset. From ck's blog:

"I've put up a small updated -ck version for 2.6.37. There are only 2 changes: 1. The build fix for it not compiling with CPU HOTPLUG disabled, and 2. I've dropped the patch called mm-make_swappiness_really_mean_it.patch . This second patch broke a while back and I never noticed because I had swap disabled, sorry. It works better with it disabled. Note that the ubuntu packages I recently made available include this change which I quietly snuck in, but I will make ck2 packages officially available just to avoid confusion ;) If you've built your own 2.6.37-ck1 then I recommend upgrading only if you have swap enabled on your machine (which most people still do). Alternatively if you built it with CPU_HOTPLUG enabled just to make it build, then you _can_ upgrade to this version to build with it disabled but you won't notice a difference with the feature disabled - your kernel will only be slightly smaller. It might be a few nanoseconds faster.."

12-Feb-2011 kernel26-ck/headers-ck 2.6.37-6

rebuild against latest glibc, binutils git-svn-id: 109724 eb2447ed-0c53-47e4-bac8-5bc4a241df78

I am reading this as glibc 2.13-3 and binutils 2.21-4, both of which are currently in [testing] need to be installed prior to kernel compilation. I can install both of these testing packages on my machine, and compile for the repo, but individual users will have to install those packages, THEN compile their own kernels... therefore both glibc and binuntils have been temporarily added to the deps array in the PKGBUILD. Users will need to snag them either by enabling [testing] in /etc/pacman.conf or simply by downloading via pacman -U.

29-Jan-2011 kernel26-ck/headers-ck 2.6.37-5

fix module-init-tools dependency git-svn-id: eb2447ed-0c53-47e4-bac8-5bc4a241df78

It's a very MINOR change. Same ARCH patch set (-4) and same config files. Other than that change, it's identical to -4.

s/module-init-tools/module-init-tools>=3.12-2/

27-Jan-2011 kernel26-ck/headers-ck 2.6.37-4

added i915 patch git-svn-id: eb2447ed-0c53-47e4-bac8-5bc4a241df78 removed nfs idmapper option git-svn-id: eb2447ed-0c53-47e4-bac8-5bc4a241df78

26-Jan-2011 kernel26-ck/headers-ck 2.6.37-3

rebuild with mkinitcpio bump to 0.6.8 and gzipped modules saves 70MB space git-svn-id: eb2447ed-0c53-47e4-bac8-5bc4a241df78

24-Jan-2011 kernel26-ck/headers-ck 2.6.37-2

fix udev crash, updated aufs, changed ext3 mount default git-svn-id: eb2447ed-0c53-47e4-bac8-5bc4a241df78