Package Base Details: linux-mainline

Git Clone URL: https://aur.archlinux.org/linux-mainline.git (read-only)
Submitter: miffe
Maintainer: miffe
Last Packager: miffe
Votes: 172
Popularity: 6.699455
First Submitted: 2011-07-22 06:50
Last Updated: 2016-07-25 22:03

Latest Comments

talonz commented on 2016-07-09 22:49

I got a problem trying to get r8168-dkms working it fails to build the module
Is there something I need to add to the package build to get it working, this fails on both mainline and git kernels

Your kernel headers for kernel cannot be found at
/usr/lib/modules//build or /usr/lib/modules//source.

Edit: i have linux-mainline-headers installed and the git one as well

Edit 2:
$ cat /var/lib/dkms/r8168/8.042/build/make.log
DKMS make.log for r8168-8.042 for kernel 4.7.0-rc7-gfa3a9f574 (x86_64)
Sat Jul 16 08:00:11 AEST 2016
make: Entering directory '/usr/lib/modules/4.7.0-rc7-gfa3a9f574/build'
CC [M] /var/lib/dkms/r8168/8.042/build/src/r8168_n.o
/var/lib/dkms/r8168/8.042/build/src/r8168_n.c: In function ‘rtl8168_init_one’:
/var/lib/dkms/r8168/8.042/build/src/r8168_n.c:22799:28: error: ‘struct net_device’ has no member named ‘gso_min_segs’; did you mean ‘gso_max_segs’?
dev->gso_min_segs = NIC_MIN_PHYS_BUF_COUNT;
^~
/var/lib/dkms/r8168/8.042/build/src/r8168_n.c:22807:28: error: ‘struct net_device’ has no member named ‘gso_min_segs’; did you mean ‘gso_max_segs’?
dev->gso_min_segs = NIC_MIN_PHYS_BUF_COUNT;
^~
/var/lib/dkms/r8168/8.042/build/src/r8168_n.c: In function ‘rtl8168_start_xmit’:
/var/lib/dkms/r8168/8.042/build/src/r8168_n.c:24969:12: error: ‘struct net_device’ has no member named ‘trans_start’; did you mean ‘mem_start’?
dev->trans_start = jiffies;
^~
make[1]: *** [scripts/Makefile.build:290: /var/lib/dkms/r8168/8.042/build/src/r8168_n.o] Error 1
make: *** [Makefile:1457: _module_/var/lib/dkms/r8168/8.042/build/src] Error 2
make: Leaving directory '/usr/lib/modules/4.7.0-rc7-gfa3a9f574/build'

any help would be much appreciated as i am currently tethering off of my phone for a week ;[

kgizdov commented on 2016-07-09 16:47

I'm getting the following warning:

Hunk #1 succeeded at 39 (offset 2 lines).
HOSTCC scripts/basic/fixdep
HOSTCC scripts/kconfig/conf.o
SHIPPED scripts/kconfig/zconf.tab.c
SHIPPED scripts/kconfig/zconf.lex.c
SHIPPED scripts/kconfig/zconf.hash.c
HOSTCC scripts/kconfig/zconf.tab.o
HOSTLD scripts/kconfig/conf
scripts/kconfig/conf --olddefconfig Kconfig
.config:1614:warning: symbol value 'm' invalid for RXKAD
.config:7711:warning: symbol value 'm' invalid for ASYMMETRIC_KEY_TYPE
#
# configuration written to .config
#
scripts/kconfig/conf --silentoldconfig Kconfig

I wonder if it's a misconfiguration some where.

Jhackler commented on 2016-06-11 15:46

I had the same issue as craftyguy and his solution fixed it.

craftyguy commented on 2016-06-06 04:05

FYI, after installing 4.7rc2, I got an "error loading libudev.so.1" error when trying to start X11. Reinstalling systemd and libsystemd resolved the issue. Not sure if there's anything you can do in the PKGBUILD for that, but thought I would mention it here in case others hit this.

oniram commented on 2016-05-25 20:19

Package community/bbswitch-dkms fails to build the module on this kernel (linux-mainline-4.6-1). I've tracked the problem down to a new user-space utility (objtool) that comes with 4.6 and needs to be installed.

In the core/linux package they handle it like this:
https://git.archlinux.org/svntogit/packages.git/commit/trunk?h=packages/linux&id=2ae7d54ecc5cbebdf766134fd68452b011d5cd7a

Jhackler commented on 2016-04-26 19:00

"I just wanted to know if something changed in the package to trigger the bad behaviour in yaourt or if it was an helper update which I didn't notice.

P.S. I really don't know why everybody hates helpers so much. Using makepkg is easy but tedious when you have lots of packages from aur: helpers are a great thing, especially since they allow you to edit the pkgbuilds on the fly."

darkbasic, the issue comes because it is clear that either the process of doing it manually is unknown, or the poster was not willing to put forward the effort to test it manually.

One should make sure it was the AUR package and not the helper first. The correct thing to do when it fails is, do it manually, then if it succeeds file the report with yaourt devs not the perfectly fine Aur package.

I use Yaourt and did not come here asking about the issue. Actually in this case rebooting the system and running yaourt again worked out. I doubt anyone hates people using helpers, they just are not interested in going out of the way to support them in their packages. Don't worry I have been lazy at times as well and got varied degrees of grief over it.

mayurkr commented on 2016-04-18 07:35

Please update the source from www.kernel.org to cdn.kernel.org. cdn.kernel.org is using fastly cdn to speedup downloads.

rcoit commented on 2016-04-08 16:55

There have been some changes in the config for the standard kernel build and now my touchscreen on my tablet is working again. Is there any chance that you could update your config file with these changes.

Thanks

Jhackler commented on 2016-03-29 03:50

Was just coming to mention that powerplay is now on by default in latest rc but was a little slow on the draw. Glad this is the case. Now cant wait till this makes it to the official arch kernel so I can remove this from my regular updates.
For the person with the Yaourt issue make sure it is the latest version and restart your computer I tested it and it was working.

holunder commented on 2016-03-28 15:57

AMDGPU Powerplay is now activated by default with 4.6 RC1! \o/

darkbasic commented on 2016-03-01 21:54

I just wanted to know if something changed in the package to trigger the bad behaviour in yaourt or if it was an helper update which I didn't notice.

P.S. I really don't know why everybody hates helpers so much. Using makepkg is easy but tedious when you have lots of packages from aur: helpers are a great thing, especially since they allow you to edit the pkgbuilds on the fly.

WorMzy commented on 2016-03-01 20:31

Why are you posting that? You're using an AUR helper that is broken, just use makepkg.

darkbasic commented on 2016-03-01 20:28

==> This PKGBUILD provides " linux-mainline linux-mainline-headers linux-mainline-docs "
==> Install only linux-mainline ? [Y/n]
==> -----------------------------------
==>
makepkg: invalid option '--pkg'
==> ERROR: Makepkg was unable to build linux-mainline.
==> Restart building linux-mainline ? [y/N]

darkbasic commented on 2016-02-26 19:07

Guys he will not enable it, if you want powerplay just uncomment "make nconfig" from the pkgbuild and add it by yourself. In the next kernel version it will be probably enabled by default and the problem will not exist anymore :)

Jhackler commented on 2016-02-26 19:04

Any news on the powerplay request?

kvvius commented on 2016-02-15 15:13

+5 for enabling Powerplay !

Tromzy commented on 2016-02-09 11:53

+4 for enabling Powerplay !

jayki commented on 2016-02-08 15:36

+3 for using the testing configs and +3 for also enable AMDGPU Powerplay

holunder commented on 2016-02-06 20:42

+2 to please enable AMDGPU Powerplay! Doesn't do harm to nobody, one still needs to set a special boot parameter to use it.

Jhackler commented on 2016-02-06 09:12

+1 to enabling the powerplay flag. As an r9 380 user this is the main reason I was considering the mainline kernel.

darkbasic commented on 2016-02-02 09:58

@miffe I definitely prefer testing/linux please, I don't need a flag.

prazola commented on 2016-02-01 16:24

@miffe I go for the testing one, or you can add a flag to switch between two configs.

miffe commented on 2016-01-28 12:29

@darkbasic: I merge the core/linux config on every mainline-rc1 release. If enough people want it, I guess i could merge with testing/linux every rc release instead.

darkbasic commented on 2016-01-28 10:28

@miffe: Do you really use the config from core/linux? testing/linux would be a far better candidate. For example my laptop needed some config changes with 4.4, so I filed a bug and asked to fix it, which they did. If you still use the 4.3 config I will miss the improvements which will land in 4.4 once it will hit core. Since everyone expects linux-mainline being much more unstable than testing/linux, then aiming to use the testing config is the natural choice IMHO.

miffe commented on 2016-01-25 22:55

@lordheavy: I won't change the config from whats in core/linux.
@prazola: sed '215,242d' -i PKGBUILD

prazola commented on 2016-01-25 14:07

Header compile error if you exclude dvb configs from .config.

lordheavy commented on 2016-01-25 00:38

Can you enable powerplay with 4.5rc1 kernel ? Thks

# CONFIG_DRM_AMD_POWERPLAY is not set

prazola commented on 2016-01-17 21:38

One question: are compile flags present in "Makefile" of this kernel overrided by the /etc/makepkg.conf flags?
ty

Butterkekstorte commented on 2016-01-13 05:13

It installed fine, but somehow the generated ramdisk missed the nvme and btrfs modules present in the 'real' kernels ramdisk. I needed to manually add them to mkinitcpio.conf.
How are those detected? Is this a problem of this package or a general kernel-building issue?

bradmccormack100 commented on 2016-01-10 23:45

The problem with cp: cannot stat ‘arch//Makefile’: No such file or directory is due to the scope of the KARCH variable

https://aur.archlinux.org/cgit/aur.git/tree/PKGBUILD?h=linux-mainline#n110 is where it is defined.

Its not within scope here https://aur.archlinux.org/cgit/aur.git/tree/PKGBUILD?h=linux-mainline#n160

The quickest way to fix is to just move the KARCH variable to the top so it has global scope.

Perhaps around here https://aur.archlinux.org/cgit/aur.git/tree/PKGBUILD?h=linux-mainline#n39

There is already a patch listed here too https://bugs.archlinux.org/task/43885 that could be adopted.

lesto commented on 2015-12-16 20:13

when compiling linux-mainline-header i get:

==> Starting package_linux-mainline-headers()...
cp: cannot stat ‘arch//Makefile’: No such file or directory

sonny commented on 2015-11-24 17:56

I'm getting the following error

kernel/sys.c: In function ‘override_release’:
kernel/sys.c:1133:28: error: expected expression before ‘>>’ token
v = ((LINUX_VERSION_CODE >> 8) & 0xff) + 60;
^
scripts/Makefile.build:258: recipe for target 'kernel/sys.o' failed
make[1]: *** [kernel/sys.o] Error 1
Makefile:943: recipe for target 'kernel' failed
make: *** [kernel] Error 2

lesto commented on 2015-11-24 00:00

same problem here!
seems it should be only in linux-mainline-docs right?

thevoidnn commented on 2015-11-20 23:43

error: failed to commit transaction (conflicting files)
/usr/lib/modules/4.4.0-rc1-mainline/build/Documentation/kbuild/Kconfig.recursion-issue-01 exists in both 'linux-mainline-docs' and 'linux-mainline-headers'
/usr/lib/modules/4.4.0-rc1-mainline/build/Documentation/kbuild/Kconfig.recursion-issue-02 exists in both 'linux-mainline-docs' and 'linux-mainline-headers'
/usr/lib/modules/4.4.0-rc1-mainline/build/Documentation/kbuild/Kconfig.select-break exists in both 'linux-mainline-docs' and 'linux-mainline-headers'
Errors occurred, no packages were upgraded.


(had to install kernel manually (no yaourt), omitted docs package)

miffe commented on 2015-11-20 17:47

@darkbasic: You're right. A | had gone missing from row 50, so the patch was never applied. Fixed in -2

darkbasic commented on 2015-11-20 17:28

@miffe
I mean that it doesn't really apply the patch: when you uncomment "make nconfig" in the ebuild you get a 4.3 kernel nconfig. I even opened the Makefile inside the src directory and I can see a 4.3 version. I even tried to apply the 4.4-rc1 patch inside the src directory while it was compiling and it applied flawlessly, meaning it didn't apply before.

miffe commented on 2015-11-20 15:17

@darkbasic: What do you mean? It's applied on row 50 in the PKGBUILD.

@BugMaze: That looks like an issue with the linux-docs/headers packages not linux-mainline.

darkbasic commented on 2015-11-20 14:42

It doesn't apply patch-4.4-rc1.xz.

BugMaze commented on 2015-11-17 06:35

Compiled it today , but it won't install :

(3/3) checking for file conflicts [##############################################################] 100%
error: failed to commit transaction (conflicting files)
/usr/lib/modules/4.4.0-rc1-ARCH/build/Documentation/kbuild/Kconfig.recursion-issue-01 exists in both 'linux-docs' and 'linux-headers'
/usr/lib/modules/4.4.0-rc1-ARCH/build/Documentation/kbuild/Kconfig.recursion-issue-02 exists in both 'linux-docs' and 'linux-headers'
/usr/lib/modules/4.4.0-rc1-ARCH/build/Documentation/kbuild/Kconfig.select-break exists in both 'linux-docs' and 'linux-headers'
Errors occurred, no packages were upgraded.

darkbasic commented on 2015-11-16 17:22

Please bump to 4.4-rc1.

FadeMind commented on 2015-11-16 11:29

Should be applied this patch: https://lkml.org/lkml/2015/11/6/987
Reference: https://lkml.org/lkml/2015/11/9/702

BugMaze commented on 2015-11-16 10:52

Using this PKGBUILD but changing it to compile 4.4-rc1 results :

DEPMOD 4.4.0-1-ARCH
depmod: ERROR: Found 2 modules in dependency cycles!
depmod: ERROR: Cycle detected: lnet -> libcfs -> lnet
==> ERROR: A failure occurred in package_linux().
Aborting...

FadeMind commented on 2015-11-16 08:19

4.4rc1 released

SBeaver commented on 2015-09-20 17:27

I get a linking error complaining about a path. It seems somewhere in there there is something that doesn't handle paths containing spaces properly. The first part of the path displayed in the error is actually longer and the part shown is the one after the space only. Should be pretty clear even if it's not in english. Message:
ln: målet ”filer/linux-mainline/pkg/linux-mainline/lib/modules/4.3.0-rc1-mainline/source” är inte en katalog
Makefile:1117: receptet för målet ”_modinst_” misslyckades
make: *** [_modinst_] Fel 1

gokcen commented on 2015-08-30 20:39

This patch would make all bcmwl users happy: https://git.kernel.org/cgit/linux/kernel/git/tj/wq.git/commit/?h=for-4.3&id=1dadafa86a779884f14a6e7a3ddde1a57b0a0a65

WorMzy commented on 2015-08-11 17:40

rachaelbe, this is not the place to seek support for these problems. File bugs upstream and (if you can) the bisect to find the commits that cause them.

rachaelbe commented on 2015-08-11 17:33

More than that, suspend doesn't work with rc6 either... have now downgraded back to rc5 and all is right with the world!

rachaelbe commented on 2015-08-11 16:18

Also i3status bar won't start now, although it will run in the terminal (just not in X11)

rachaelbe commented on 2015-08-11 16:09

Just updated (via the repository) to rc6 from rc5 and I can no longer set the backlight on my Acer laptop... it's being working fine up until now :(

rachaelbe commented on 2015-08-11 16:07

Just updated to rc6 from rc5 and I can no longer set the backlight on my Acer laptop... it's being working fine up until now :(

Vi0L0 commented on 2015-08-08 22:21

Are you also unable to build this packlage after glibc 2.22 & toolchain update [testing]?
It throws something like (translating): "Makefile: 944: commands for target 'drivers' failed".
linux builds fine, though.

colorprint commented on 2015-08-05 07:34

Kernel and modules crashes when booting on MS Surface Pro 3.
Worked fine before with 4.1 linux-mainline kernel.
Also works fine when compiled with default .config file.
Something should be fixed in config.x86_64 file, but I can't find out what exactly...

lrusak commented on 2015-07-20 06:54

I tried bumping to 4.2rc3 but I ended up being unable to boot. Had to reinstall 4.2rc2, any ideas?

Det commented on 2015-06-30 00:18

But don't update to 4.1.1. That's not why I flagged. Otherwise we might get fixes in.

graysky commented on 2015-06-22 21:24

Yeah, Tux on hurrdurr.org doesn't look quite right :)

yurikoles commented on 2015-06-17 15:20

Please use official home page: https://www.kernel.org/

FadeMind commented on 2015-06-01 10:17

Please update config files with lastest changes from official linux arch src.

miffe commented on 2015-05-25 14:46

@taylorchu: I saw your comment too late for rc5. But I'll do it for rc6.

taylorchu commented on 2015-05-25 14:25

@miffe
I noticed that miffe's compiled kernel uses older gcc. Could you update gcc?

miffe commented on 2015-05-10 13:05

@gokcen: It'll be included in rc3.

gokcen commented on 2015-05-10 09:08

@miffe how about this patch?

https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git/commit/?id=de71ad2c97862eae1516aa36528cc3b317c17b2f

FadeMind commented on 2015-04-27 15:17

It is source of issue https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git/commit/?id=24933b82c0d9a711475a5ef7904eb733f561e637

Reference: https://devtalk.nvidia.com/default/topic/828445/linux/nvidia-349-16-linux-4-1-rc1-build-error-gpl-only-symbol-cpu_tss-/

https://lkml.org/lkml/2015/3/5/835

miffe commented on 2015-04-27 15:11

@gokcen: I guess that broadcom hasn't updated their driver yet for 4.1. But it's a problem for the broadcom-wl-dkms package, not this one.

gokcen commented on 2015-04-27 14:12

4.1rc compiles fine but broadcom-wl dkms gives the following error:

$ sudo dkms install broadcom-wl/6.30.223.248
Kernel preparation unnecessary for this kernel. Skipping...

Building module:
cleaning build area....
make KERNELRELEASE=4.1.0-1-mainline -C /usr/lib/modules/4.1.0-1-mainline/build M=/var/lib/dkms/broadcom-wl/6.30.223.248/build....(bad exit status: 2)
Error! Bad return status for module build on kernel: 4.1.0-1-mainline (x86_64)
Consult /var/lib/dkms/broadcom-wl/6.30.223.248/build/make.log for more information.

$ cat /var/lib/dkms/broadcom-wl/6.30.223.248/build/make.log
DKMS make.log for broadcom-wl-6.30.223.248 for kernel 4.1.0-1-mainline (x86_64)
Mon Apr 27 15:57:23 CEST 2015
make: Entering directory '/usr/lib/modules/4.1.0-1-mainline/build'
CFG80211 API is prefered for this kernel version
Using CFG80211 API
LD /var/lib/dkms/broadcom-wl/6.30.223.248/build/built-in.o
CC [M] /var/lib/dkms/broadcom-wl/6.30.223.248/build/src/shared/linux_osl.o
CC [M] /var/lib/dkms/broadcom-wl/6.30.223.248/build/src/wl/sys/wl_linux.o
CC [M] /var/lib/dkms/broadcom-wl/6.30.223.248/build/src/wl/sys/wl_iw.o
CC [M] /var/lib/dkms/broadcom-wl/6.30.223.248/build/src/wl/sys/wl_cfg80211_hybrid.o
LD [M] /var/lib/dkms/broadcom-wl/6.30.223.248/build/wl.o
Building modules, stage 2.
CFG80211 API is prefered for this kernel version
Using CFG80211 API
MODPOST 1 modules
FATAL: modpost: GPL-incompatible module wl.ko uses GPL-only symbol 'cpu_tss'
scripts/Makefile.modpost:90: recipe for target '__modpost' failed
make[1]: *** [__modpost] Error 1
Makefile:1386: recipe for target 'modules' failed
make: *** [modules] Error 2
make: Leaving directory '/usr/lib/modules/4.1.0-1-mainline/build'

miffe commented on 2015-04-13 16:50

@FadeMind: I'll update it for RC1, same as I always do. There is very little change. And if you really need the changes, you can use the kernel in testing.

FadeMind commented on 2015-04-13 16:29

miffe
Please using file config and file config.x86_64 for 4.0 series.
You using 3.19 configs - they are obsolete. TY

https://projects.archlinux.org/svntogit/packages.git/tree/trunk?h=packages/linux

miffe commented on 2015-03-22 23:21

@silentmoose: It does. It's only one patch and all it does is change the default log level.

silentmoose commented on 2015-03-22 22:10

Does this come with the patches arch deploys on the kernel as well?

FadeMind commented on 2015-03-06 15:28

miffe
Please using config and config.x86_64 for lastest stable mainline 3.19 series.
You using 3.18 configs obsolete.

https://projects.archlinux.org/svntogit/packages.git/tree/trunk?h=packages/linux

miffe commented on 2015-03-06 15:11

@emolitor: This package uses the config from the core/linux package, so once it's enabled there it will be here too.

emolitor commented on 2015-03-06 12:04

Relatively new to arch but it would be good to enable baytrail/broadwell+ support by adding a few more defaults.

CONFIG_X86_INTEL_LPSS=y
CONFIG_PINMUX=y
CONFIG_PINCONF=y
CONFIG_PINCTRL_BAYTRAIL=y
CONFIG_I2C=y
CONFIG_ACPI_I2C_OPREGION=y

FadeMind commented on 2015-03-03 18:31

Linux 4.0-rc2 http://lkml.iu.edu/hypermail/linux/kernel/1503.0/01963.html

clfarron4 commented on 2015-02-23 10:28

Oh my days... 4.0-rc1 has arrived. https://www.kernel.org/pub/linux/kernel/v4.x/testing/linux-4.0-rc1.tar.xz

Update: I think the 4MB patch-4.0-rc1.tar.xz is botched, because there's a Kconfig file missing which is there if you use the larger 70MB source.

clfarron4 commented on 2015-02-23 09:27

Oh my days... 4.0-rc1 has arrived. https://www.kernel.org/pub/linux/kernel/v4.x/testing/linux-4.0-rc1.tar.xz

alex.henrie commented on 2015-02-19 20:44

@miffe: You're right. `pacaur -Sy linux-mainline linux-mainline-headers` works.

miffe commented on 2015-02-19 10:59

@alex.henrie: It tries to copy arch/${KARCH}/Makefile in _package-headers(), but the KARCH variable is set in _package(). So if you build everything at once and not just the headers package it should work. I'll open a bug for core/linux.

alex.henrie commented on 2015-02-19 00:18

I can't get this package to install. When I run `pacaur -Sy linux-mainline-headers` the entire kernel compiles, then the installation stops with the message

IHEX firmware/yam/9600.bin
==> Entering fakeroot environment...
==> Starting package_linux-mainline-headers()...
cp: cannot stat 'arch//Makefile': No such file or directory
==> ERROR: A failure occurred in package_linux-mainline-headers().
Aborting...
:: linux-mainline cleaning skipped
:: failed to build linux-mainline-headers package(s)

miffe commented on 2015-02-11 12:22

@hzy199411: I wont change the config from whats in the core/linux package. Also, wont x32 need a different userland thats not available in archlinux?

hzy199411 commented on 2015-02-11 01:03

would you like enable x32 ABI for this kernel?

miffe commented on 2015-02-02 12:29

@clfarron4: See https://lkml.org/lkml/2015/1/26/522

I'll switch the PKGBUILD to using the full tarballs.

clfarron4 commented on 2015-02-02 09:17

Just tried building rc7 (by updating the sources) and I get this:

patching file arch/arm64/boot/dts/cavium/thunder-88xx.dtsi
patching file arch/arm64/boot/dts/foundation-v8.dts
patching symbolic link arch/arm64/boot/dts/include/dt-bindings
symbolic link target '../../../../../include/dt-bindings' is invalid
==> ERROR: A failure occurred in prepare().
Aborting...
makepkg -s 27.05s user 2.89s system 38% cpu 1:18.73 total

Am I the only one?

lockheed commented on 2015-01-15 14:06

@miffe,
"setup" option has been appearantly removed:

$ sudo phc-intel setup
usage: /usr/bin/phc-intel {start|stop|status|set}

miffe commented on 2015-01-15 13:27

@edward_81: Thats the same CFLAGS as me. Please add V=1 to the make command on line 89 and rebuild with MAKEFLAGS="-j1" to get verbose output.

@lockheed: Did you run phc-intel setup? Any errors from that?

lockheed commented on 2015-01-14 22:51

Is it not possible to patch mainline kernel with phc-intel?

$ sudo phc-intel status
PHC status: /usr/bin/phc-intel: line 37: /sys/devices/system/cpu/cpu0/cpufreq/phc_vids: No such file or directory
/usr/bin/phc-intel: line 37: /sys/devices/system/cpu/cpu0/cpufreq/phc_default_vids: No such file or directory
/usr/bin/phc-intel: line 37: /sys/devices/system/cpu/cpu1/cpufreq/phc_vids: No such file or directory
/usr/bin/phc-intel: line 37: /sys/devices/system/cpu/cpu1/cpufreq/phc_default_vids: No such file or directory
inactive

edward_81 commented on 2015-01-14 16:27

@miffe: I don't remember to have touch anything.
This is my actual /etc/makepkg.conf
http://pastebin.com/tBv1xn7n

miffe commented on 2015-01-14 14:26

@edward_81: Are you using nonstandard CFLAGS?

@jtyr: There is a repo at https://arch.miffe.org/

jtyr commented on 2015-01-14 13:39

Please could you create an Arch repo with the builds like they do it for linux-lts310 (https://aur.archlinux.org/packages/linux-lts310/)?

edward_81 commented on 2015-01-12 13:17

rc3 compile fail
HOSTLD scripts/genksyms/genksyms
CC scripts/mod/empty.o
scripts/mod/empty.c:1:0: error: code model kernel does not support PIC mode
/* empty file to figure out endianness / word size */
^

clfarron4 commented on 2014-12-25 17:07

@lockheed: Just built it in a clean chroot here just fine. Still to test if it boots though.

lockheed commented on 2014-12-22 17:47

I see. I am asking because it is the first time linux-mainline fails to build on my machine. I will try again.

miffe commented on 2014-12-22 17:46

@lockheed: It downloads the 3.18 tarball and then uses the 3.19-rc1 patch to update the sources. It saves on bandwitdh over time compared to using the rc tarball directly for every version.

WorMzy commented on 2014-12-22 17:44

Because that is the base for the 3.19rc1 patch to be applied to.

lockheed commented on 2014-12-22 17:36

I am wondering why linux-mainline 3.19rc1 downloads linux-3.18.tar.xz

gutigen commented on 2014-11-18 01:45

It was 3.18rc4 in that quoted part from pkgbuild file, it's ok now ;)

miffe commented on 2014-11-17 20:31

@gutigen: It should be 3.18rc5. Where do you see rc4?

gutigen commented on 2014-11-17 17:45

Still named

_patchname=patch-3.18-rc4
pkgver=3.18rc4

Does it make any difference?

Potomac commented on 2014-11-03 02:23

I have just installedl linux-mainline 3.18rc3 and I notice this message on boot :

hwdb.bin does not exist, please run udevadm hwdb --update

is it normal ?

Potomac commented on 2014-11-03 02:22

linux 3.18rc3 has been released november 2 2014

Anonymo commented on 2014-10-27 23:15

@WorMzy @miffe Thanks for the explanation.

WorMzy commented on 2014-10-27 20:13

Anonymo, BFQ is an out-of-tree scheduler, you need to patch the kernel source for it to even be an option, and the patches do not follow mainline (presumably because it is a moving target, and I think BFQ is basically a one or two-man project).

See linux-pf[1] or linux-ck[2] if you want to use BFQ on your system, or try applying the patches to mainline yourself.

[1] https://wiki.archlinux.org/index.php/Linux-pf
[2] https://wiki.archlinux.org/index.php/Linux-ck

miffe commented on 2014-10-27 19:18

@Anonymo: I merge the core/linux package every rc1, so if it gets added there, it will show up here too.

Anonymo commented on 2014-10-27 19:03

Anyway to get bfq so we can enable in /etc/tmpfiles.d/set_IO_scheduler.conf? Right now, the only options are noop, deadline and [cfq].

jaxxed commented on 2014-10-20 10:05

IP_NF_NAT=m is necessary to maintain iptables syntax. Without it packages like docker fail (confirmed by pablo@netfilter.org)

The testing kernel appears to already have this (their 3.17.1-1 kernel works well.)

Also, I tested this mainline build using 3.17.1 (src and pkgver using 2c9c002cf9618962883d82ba3e1caaea29af4ea37147a82bad422afdc8d21dc2) and it worked for me.

punkrockdude commented on 2014-10-06 14:46

Tack Miffe.

miffe commented on 2014-10-06 14:38

punkrockdude: If you don't want to build it, there are binary packages available at http://arch.miffe.org/

punkrockdude commented on 2014-10-06 14:31

Thank you WorMzy. Using Bleachbit to remove a lot of tmp files and other stuff seem to have help. Now it finally have gotten to where it downloads the linux-3.17.tar.xz file.

WorMzy commented on 2014-10-06 14:25

Looks like your /tmp is full, which probably isn't helping. Clean up some temporary files and see if that helps. The PKGBUILD itself is fine, I've tested it here.

punkrockdude commented on 2014-10-06 14:11

I downloaded the file again but still get errors:
==> ERROR: pkgrel must be a decimal.
/usr/bin/makepkg: line 2214: cannot create temp file for here-document: No space left on device

WorMzy commented on 2014-10-06 13:46

It sounds like there are trailing characters on your pkgrel line. Check the PKGBUILD you downloaded, make sure it's not corrupted.

punkrockdude commented on 2014-10-06 13:40

When I download this package and makepkg I get: ==> ERROR: pkgrel must be a decimal.

miffe commented on 2014-09-10 16:28

@hzy199411: For me the download is faster than the patching. But not by a lot. I'll try the patch method for the next release.

hzy199411 commented on 2014-09-10 00:30

@miffe,because download a whole source tarball is slow and wasteful,I mean you can use like this "linux-3.16.tar.xz+patch-3.16.1.xz"

miffe commented on 2014-09-09 20:46

@beta990: make olddefconfig is uncommented by default. It's based on the .config from core/linux.

francoism commented on 2014-09-09 20:15

Thanks for providing this pkg. :)

Just a stupid question: If I don't uncomment a config line in PKGBUILD, what config is used?

miffe commented on 2014-09-09 13:09

@hzy199411: Why? It's faster without the patching.

hzy199411 commented on 2014-09-09 10:08

I think you should use "A base kernel source tarball + patch" to replace the whole kernel source like linux packages in abs.

Anonymo commented on 2014-08-18 02:31

3.17rc1 netfilter error at startup

Anonymo commented on 2014-08-04 03:49

It's out. https://lkml.org/lkml/2014/8/3/82

miffe commented on 2014-06-14 15:56

@AnAkkk: I rebase against core/linux every rc1 release.

AnAkkk commented on 2014-06-14 07:52

You might want to update the config to the latest one of the official package:
https://projects.archlinux.org/svntogit/packages.git/tree/trunk/config.x86_64?h=packages/linux

miffe commented on 2014-06-13 10:14

@mbirkis: The headers are included in the linux-mainline package. It was made this way since the old AUR didn't support split packages.

For 3.16rc1 I will make it a split package.

mbirkis commented on 2014-06-13 07:55

Hi! Thanks for this maintaining this package.

I just upgraded from 3.14 mainline to the current 3.15-1 and when i build it the linux-mainline-header is missing.

I tried reading the PKBUILD and there seem to be some changes with regards to this, but it is not obvious for me how to "enable" the building of the headers now. Appreciate any help, thanks.

graysky commented on 2014-06-08 19:41

3.15.0 is out!

MoSal commented on 2014-05-22 17:26

The ability to use LZ4 compression in the zram module was added in '3.15'.
That's why core/linux couldn't have enabled it.

I expect it to be enabled once 3.15[.0] is pushed to testing.
If not, I will send a feature request.

Anyway, I understand your conservativeness.
I guess I will compile rc6 myself.

miffe commented on 2014-05-22 16:17

@MoSal: I wont change the config from whats in core/linux.

MoSal commented on 2014-05-20 21:27

May I suggest enabling:
CONFIG_ZRAM_LZ4_COMPRESS

miffe commented on 2014-05-06 15:36

linux-mainline 3.15rc4-1:
- New upstream release

Packages at http://arch.miffe.org/

miffe commented on 2014-04-28 13:06

linux-mainline 3.15rc3-1:
- New upstream release

Packages at http://arch.miffe.org/

miffe commented on 2014-04-21 04:02

linux-mainline 3.15rc2-1:
- New upstream release

Packages at http://arch.miffe.org/

miffe commented on 2014-04-14 21:51

linux-mainline 3.15rc1-1:
- New upstream release
- PKGBUILD synced with core/linux 3.14-5

Packages at http://arch.miffe.org/

miffe commented on 2014-03-31 14:45

linux-mainline 3.14-1:
- New upstream release

Packages at http://arch.miffe.org/

miffe commented on 2014-03-25 15:30

linux-mainline 3.14rc8-1:
- New upstream release

Packages at http://arch.miffe.org/

I accidently removed my gpg directory, so new key for this release.
# pacman-key --recv-keys 313F5ABD
# pacman-key --lsign-key 313F5ABD

miffe commented on 2014-03-17 12:27

linux-mainline 3.14rc7-1:
- New upstream release

Packages at http://arch.miffe.org/

AuHop commented on 2014-03-17 07:40

It's funny that you mention that actually, suspend is broken (with radeon) on my laptop now too. 6620g + 6750m (HP dv6z-6100) using PRIME, fwiw. Hibernate seems to work fine, but suspend fails when it tries to start the screen again.

jtheoof commented on 2014-03-16 10:44

I don't know for you, but on my laptop, the suspend seems broken since 3.14rc6.

miffe commented on 2014-03-10 18:59

linux-mainline 3.14rc6-1:
- New upstream release

Packages at http://arch.miffe.org/

miffe commented on 2014-03-04 16:26

linux-mainline 3.14rc5-1:
- New upstream release

Packages at http://arch.miffe.org/

kleinph commented on 2014-02-27 00:13

The problem with the hook is not that annoying.

Thanks for the bisect. I wanted to write to the mailing list, when I saw that you did that already. Thanks alot!

WorMzy commented on 2014-02-26 21:29

I think you're just not meant to use this kernel series. :P

I can't help with those problems, as I don't use that hook. However, I've almost finished bisecting the commit (I believe it _is_ that patch, commit ref 0b947aff1599afbbd2ec07ada87b05af0f94cf10)

Did you open a bug report? If so, can you post the link?

kleinph commented on 2014-02-25 19:17

Ok I can confirm that I can boot with this workaround, but I found two other issues (unrelated to btrfs):
* I cannot boot with the shutdown hook enabled (the hook seems to hang)
* before the initrd handles over to the kernel it prints a message like "old_root was not a initramfs"

Maybe I should write to the mailinglist?

WorMzy commented on 2014-02-25 18:54

I'm almost positive that it's a btrfs regression. The only other possible cause is mkinitcpio, but I believe that the adding of modules to the initrd is automated, using the information the modules themselves provide (via depmod, presumably). That patch (if it was committed) may be a good starting point for a bisect.

kleinph commented on 2014-02-25 18:38

Ok thanks! Is this a upstream bug or related to the package?

Can this be related to this patch http://permalink.gmane.org/gmane.comp.file-systems.btrfs/32145 ?

WorMzy commented on 2014-02-25 17:49

I humbly retract my earlier statement and eat my hat. I can confirm that something is not right. For some reason, the crc32c module is getting left out of the initrd, and btrfs module needs this module to work.

The reason why I didn't have this problem at home, was because I explicitly install crc32c by adding it to mkinitcpio.conf's MODULE array. This is a workaround. Another is to use the fallback initrd for the time being.

I don't have time right now to bisect the commit that caused this regression, but if you want to open a bug report and post a link to it here, I'll see if I can help out getting this fixed. Cheers.

Det commented on 2014-02-25 09:13

Congrats on the 100th vote for the 2nd time.

WorMzy commented on 2014-02-24 22:59

Still no problem here, and I've not noticed anything on the btrfs mailing list.. Personally, I think it's a problem with your setup. If I get chance, I'll try the mainline kernel on my non-raid btrfs work PC tomorrow and see if that has any problems.

kleinph commented on 2014-02-24 22:53

The same problem occurs with the rc4. Should I file a bug upstream?

miffe commented on 2014-02-24 12:23

linux-mainline 3.14rc4-1:
- New upstream release

hzy199411 commented on 2014-02-24 07:03

3.14rc4 is now available,

kleinph commented on 2014-02-19 17:11

Ok then I will wait for the next RC and try again.

cbowman57 commented on 2014-02-19 16:11

Pretty sure the btrfs problems are coming from upstream & it's development issue.

kleinph commented on 2014-02-19 15:22

Thanks, but that doesn't seem to be the problem. Both have kernel 3.14.0-1-mainline.

WorMzy commented on 2014-02-19 15:05

Depending on how you have your system set up, there's a chance that the kernel or the initrd images can become desynchronised, so you have a newer kernel and an older initrd, or vice versa. You don't get a "no initrd or kernel" error, because the old initrd or kernel is still present.

Checking the kernel and the initrds on the ESP would help identify if this is the problem, e.g.:

lsinitcpio -a /boot/EFI/arch/initramfs-linux-mainline.img | grep Kernel
file /boot/EFI/arch/vmlinux-linux-mainline

kleinph commented on 2014-02-19 14:50

Yes I have a UEFI system, but the mkinitcpio script should copy the kernel and initrd automatically. Also wouldn't the bootloader then complain that there is no initrd or kernel?

I think the initramfs is booted but it can't mount /dev/sda3 because btrfs is a unknown filesystem.

It also worked flawlessly with the 3.13 mainline kernel.

WorMzy commented on 2014-02-19 14:09

Any chance you have a UEFI system and forgot to copy either the initrd or the kernel over to your ESP?

I don't have any problems on my btrfs-raid1-root PC.

kleinph commented on 2014-02-19 13:06

I can't boot with a brtfs root filesystem.

When I try to load the btrfs module in the emergency shell I get an "Unknown symbol in module, or unknown parameter (see dmesg)" error, but when I look into dmesg there are no messages regarding btrfs or kernel modules.

miffe commented on 2014-02-17 10:20

linux-mainline 3.14rc3-1:
- New upstream release

miffe commented on 2014-02-13 19:56

linux-mainline 3.14rc2-1:
- New upstream release

miffe commented on 2014-02-04 14:52

linux-mainline 3.14rc1-1:
- New upstream release
- Rebased against core/linux 3.13.1-2

clfarron4 commented on 2014-02-03 22:35

Looks like we have a new mainline (3.14-rc1)... https://www.kernel.org/

miffe commented on 2014-01-20 20:43

linux-mainline 3.13-1:
- New upstream release

Packages at http://arch.miffe.org/

miffe commented on 2014-01-12 20:41

linux-mainline 3.13rc8-1:
- New upstream release

Packages at http://arch.miffe.org/

@reyncor & phects: I wont change the config from what's in the core/linux package. But you can easily change it yourselfs, just uncomment make menuconfig and enable whatever you want.

reyncor commented on 2014-01-12 14:07

@miffe @phects Glad I'm not the only one missing modules... my trackpad and touchscreen fail without chromeos_laptop compiled. Perhaps the AUR linux-git package has the better appraoch... it has about 20 CLI dialogs asking which modules to compile.

miffe commented on 2014-01-07 16:25

linux-mainline 3.13rc7-1:
- New upstream relase

Packages at http://arch.miffe.org

kitsune commented on 2014-01-01 02:28

No, I am not that kitsune :D

miffe commented on 2014-01-01 02:06

@kitsune: Happey New Year you too!

Are you Kitsune^2?

kitsune commented on 2014-01-01 01:53

Thanks miffe!

Happy New Year! :)

miffe commented on 2013-12-31 19:42

linux-mainline 3.13rc6-1:
- New upstream relase
- Packages at http://arch.miffe.org/

kitsune commented on 2013-12-30 11:15

rc6 is out!

kitsune commented on 2013-12-29 12:36

Works great! (finally I can use my Buffalo usb wireless!)

Waiting for the rc5 ;)

phects commented on 2013-12-28 14:24

Could you activate NF_TABLES as modules by default, please?
https://home.regit.org/netfilter-en/nftables-quick-howto/

phects commented on 2013-12-28 14:23

Could you activate NF_TABLES (and friends) as modules by default, please?

tigrang commented on 2013-12-21 03:40

I figured out why resuming from standby was failing. After I removed tlp, it works fine now. Not sure exactly what was the issue, though. Anyone else have the issue with 3.13 + tlp?

agapito commented on 2013-12-17 14:34

No problem. Thanks for your work.

miffe commented on 2013-12-16 16:33

linux-mainline 3.13rc4-1:
- New upstream release
- Packages at http://arch.miffe.org/

Sorry for the delay

agapito commented on 2013-12-16 09:10

RC4 out!

tigrang commented on 2013-12-06 23:40

RC1 audio works, but standby didn't. RC2 is the opposite. We'll see what rc3 fixes, and breaks ;)

WorMzy commented on 2013-12-06 23:39

No problems here. -rc3 is now out though, so see if that resolves the issue.

If not, downgrade to rc1 and see if the issue is still present.

tigrang commented on 2013-12-06 23:36

Anyone else's audio stop working wit rc2?

teateawhy commented on 2013-12-04 18:50

@ WorMzy @ gee
You are probably right that make config is unnecessary. I removed the line from the PKGBUILD, and the kernel was compiled successfully. ( Currently this kernel is running on my machine without problems. )

miffe commented on 2013-12-04 16:41

linux-mainline 3.13rc2-1:
- New upstream release
- Packages at http://arch.miffe.org/

gee commented on 2013-11-26 00:16

ooh

miffe commented on 2013-11-25 14:48

@gee: I don't know, it comes from the core/linux package.

gee commented on 2013-11-25 03:10

Hmmmm.
You still use make config after make olddefconfig, what is the point of that?

miffe commented on 2013-11-24 22:33

linux-mainline 3.13rc1-1:
- New upstream relase
- Rebased against core/linux 3.12.1-1
- Change yes ""|make oldconfig to make olddefconfig, thanks WorMzy for the suggestion
- Packages at http://arch.miffe.org/

WorMzy commented on 2013-11-08 00:08

You've probably answered this before, but I'm curious why you use "yes "" | make config >/dev/null" opposed to "make olddefconfig". Using the default configuration seems saner to me than enabling all the new config options, but perhaps you've had problems with olddefconfig?

miffe commented on 2013-11-04 12:27

linux-mainline 3.12-1:
- New upstream release
- Binary packages at http://arch.miffe.org/

miffe commented on 2013-10-28 13:29

linux-mainline 3.12rc7-1:
- New upstream release
- Binary packages at http://arch.miffe.org/

miffe commented on 2013-10-22 22:12

linux-mainline 3.12rc6-1:
- New upstream release
- Binary packages at http://arch.miffe.org/

miffe commented on 2013-10-14 16:44

linux-mainline 3.12rc5-1:
- New upstream release
- Binary packages at http://arch.miffe.org/

nac2013 commented on 2013-10-14 05:56

took this package and updated it all for 3.12rc5 , and spread out the compile oer multiple jobs (faster but more cpu)
https://aur.archlinux.org/packages/linux-3.12-rc5/

miffe commented on 2013-10-06 23:01

linux-mainline 3.12rc4-1:
- New upstream release
- Binary packages at http://arch.miffe.org/

miffe commented on 2013-09-30 15:57

linux-mainline 3.12rc3-1:
- New upstream release
- Binary packages at http://arch.miffe.org/

miffe commented on 2013-09-24 13:06

linux-mainline 3.12rc2-1:
- New upstream release
- Binary packages at http://arch.miffe.org/

miffe commented on 2013-09-17 22:30

linux-mainline 3.12rc1-1:
- New upstream release
- Rebased against core/linux 3.11.1-1
- Binary packages at http://arch.miffe.org/

miffe commented on 2013-09-03 11:58

linux-mainline 3.11-1:
- New upstream release
- Binary packages at http://arch.miffe.org/

miffe commented on 2013-08-26 12:00

linux-mainline 3.11rc7-1:
- New upstream release
- Binary packages at http://arch.miffe.org/

miffe commented on 2013-08-19 13:59

linux-mainline 3.11rc6-1:
- New upstream release
- Fixed pkgbase issues reported by sandy8925
- Binary packages at http://arch.miffe.org/

sandy8925 commented on 2013-08-17 19:29

Line 167: Replace pkgname with pkgbase

It tries to search for linux-mainline.preset, while the file is actually called linux.preset

miffe commented on 2013-08-12 03:31

linux-mainline 3.11rc5-1:
- New upstream release
- Binaries at http://arch.miffe.org/

miffe commented on 2013-08-06 02:49

linux-mainline 3.11rc4-1:
- New release
- Binaries at http://arch.miffe.org/

miffe commented on 2013-08-03 19:18

linux-mainline 3.11rc3-1:
- New upstream release
- Binaries at http://arch.miffe.org/

@liubenyuan: The headers are included in the linux-mainline package.

WonderWoofy commented on 2013-07-29 14:54

For those wanting a split package with the headers and docs, there is a linux-git package which pulls from Linus' git tree. If you just specify which commits to use as the source, you could potentially build only at the rc releases.

I am one that actually needs the headers (stoopid broadcom....)

KaiSforza commented on 2013-07-25 22:38

No, there is only one package in this. It is not a split package.

liubenyuan commented on 2013-07-25 22:36

compiled with -rc2 (modify rc1 -> rc2 and run md5sum). However does this version of PKGBUILD yielded only one package 'linux-mainline-3.11rc2-1-x86_64.pkg.tar.xz' ? I found no traditional header package .

sandy8925 commented on 2013-07-24 16:09

Linx kernel 3.11-rc2 has been released. Need to update the package.

miffe commented on 2013-07-17 17:11

linux-mainline 3.11rc1-1:
- New upstream release
- Synced with core/linux 3.10.1-1
- Binaries at http://arch.miffe.org/

It's vacation time, so the releases for the next month or so will be somewhat sporadic.

@ridikulus_rat: The tarballs are small and download fast, so I don't really think we need to use git.

KaiSforza commented on 2013-07-15 19:01

@ridikulus_rat: I have been doing this for a long time in my kernel package, which you can find at http://git.kaictl.net/wgiokas/aur.git/tree/kernel/linux-kaictl .

I have been using the git tags style for quite some time now, since before the VCS work was in the release version of pacman.

However, I would argue against using it in this case. While it is extremely useful for people that constantly build it, the kernel repo is ~700M bare on my system. If you are building it every release, yes, it will save you quite a bit of time, but for the general public it is not an improvement.

ridikulusrat commented on 2013-07-15 18:51

@miffe: Modified PKGBUILD at https://gist.github.com/the-ridikulus-rat/6002378 (actual source files and md5sum not updated).

ridikulusrat commented on 2013-07-15 18:32

miffe: Can you use makepkg's git tag support for source instead of downloading the src tarball, that way we don't need to download an entire source tarball everytime a rc is released, if we compile our own pkg (saves bandwidth).

miffe commented on 2013-07-02 10:24

3.10-1:
- New upstream release
- Synced with core/linux 3.10-1
- Binaries at http://arch.miffe.org/

graysky commented on 2013-06-30 19:15

..shit, didn't read down the page. Looks like ridikulus_rat made the same request.

graysky commented on 2013-06-30 17:56

@miffe - Are you tracking the config.3.10 and config.x86_64.31.0 included in the official package[1]? Might be nice to keep this pkg in sync with the official ARCH one unless you have specific tunings contained in the versions you supply with the package.

1. https://projects.archlinux.org/svntogit/packages.git/tree/trunk?h=packages/linux

miffe commented on 2013-06-28 11:17

@ridikulus_rat: Sure thing!

ridikulusrat commented on 2013-06-28 11:09

@miffe: Can you build the next linux-mainline pkg with https://projects.archlinux.org/svntogit/packages.git/tree/trunk/config.x86_64.3.10?h=packages/linux ?

aorth commented on 2013-06-23 16:06

@Bartalion it's on his website: http://arch.miffe.org/

Barthalion commented on 2013-06-20 10:57

Could you share your fingerprint anywhere? I'm not eager to install signed, although non-verifiable binary packages.

miffe commented on 2013-06-13 09:32

@aorth: Yes, everything is now signed.

aorth commented on 2013-06-13 08:55

@miffe: btw, today I had to add your key (not sure why). If anyone's using SigLevel = PackageOptional and getting key errors:

pacman-key --lsign 4F8FACD7

miffe commented on 2013-06-09 21:39

linux-mainline 3.10rc5-1:
- New upstream release
- Resynced against core/linux 3.9.5-1
- Packages at http://arch.miffe.org/

miffe commented on 2013-06-03 15:39

linux-mainline 3.10rc4-1:
- New upstream release
- Packages at http://arch.miffe.org/

CIB commented on 2013-06-03 12:27

rc4 is out, could you update?

miffe commented on 2013-05-27 08:58

linux-mainline 3.10rc3-1:
- New upstream release
- Binaries at http://arch.miffe.org/ (now hosted on s3)

miffe commented on 2013-05-21 23:16

@ridikulus_rat: Done.

ridikulusrat commented on 2013-05-21 15:43

miffe: Can you please add miffe.files to your repo for use with pkgfile?

ridikulusrat commented on 2013-05-21 15:10

If any of you want to build a missing kernel module without recompiling the whole kernel, you can check my script https://github.com/the-ridikulus-rat/My_Shell_Scripts/blob/master/linux/kernel/compile_kernel_module.sh for steps on how to compile the module from the kernel source.

miffe commented on 2013-05-21 12:51

linux-mainline 3.10rc2-1:
- New upstream release
- Binaires at http://arch.miffe.org/

@markc: I won't change the config from what's in core/linux, it's too much of a maintenence burden.
@haagch: linux-mainline only conflicts/replaces kernel26-mainline. I don't see how that can be a problem with linux.

haagch commented on 2013-05-21 12:24

debug: package linux conflicts with linux-mainline (by kernel26)

I don't think it's necessary to conflict with something the repo kernel provides. Or is it?

markc commented on 2013-05-18 16:17

miffe, thanks for your binaries. I think CONFIG_USER_NS is no longer experimental in 4.10 so is there any chance you could enable it for rc2?

miffe commented on 2013-05-13 09:36

linux-mainline 3.10rc1-1:
- New upstream release
- Binaries at http://arch.miffe.org/
- Synced with core/linux 3.9.2-1

KaiSforza commented on 2013-05-09 16:47

@Det (and everyone else): This is the mainline package, the 3.9.x series is a stable kernel. 3.9.0 is a mainline kernel.

(See http://kernel.org )

Det commented on 2013-05-09 16:15

I'm assuming you'll wait for 3.10rc1 instead of continuing with 3.9.x?

aorth commented on 2013-05-01 10:06

@miffe no problem :). Arch is great because it's easy to take your PKGBUILD and drop a new config in. Anyways, cheers!

miffe commented on 2013-04-30 16:58

@aorth: I use the .config from the linux pkg. I sync every rc1 and rc2. Sent from phone while drunk, sorry for spelling errors.

aorth commented on 2013-04-30 16:45

Thanks for updating, miffe. I noticed you're still using the 3.8 defconfig though, so certain 3.9 features like dm-cache (CONFIG_DM_CACHE) aren't enabled. :)

miffe commented on 2013-04-30 09:51

linux-mainline 3.9-1:
- New upstream release
- Binaries at http://arch.miffe.org/

WonderWoofy commented on 2013-04-29 02:39

Hey, my wireless is working again! (Or maybe the firmware is actually loading now...)

miffe commented on 2013-04-22 09:50

linux-mainline 3.9rc8-1:
- New upstream release
- Binaries at https://arch.miffe.org/

miffe commented on 2013-04-15 09:23

linux-mainline 3.9rc7-1:
- New upstream release
- Binaries at http://arch.miffe.org/

kyak commented on 2013-04-08 15:43

@miffe thanks for updating this and also for the binary packages! It is really useful when i want to test something in a newer kernel.

And sorry for breaking your 3.9rc1-1 -> 3.9rc6-1 monologue below :)

miffe commented on 2013-04-08 10:05

linux-mainline 3.9rc6-1:
- New release upstream
- Binaries at http://arch.miffe.org/

miffe commented on 2013-04-01 22:15

linux-mainline 3.9rc5-1:
- New upstream release
- Binaries at http://arch.miffe.org/

miffe commented on 2013-03-25 09:58

linux-mainline 3.9rc4-1:
- New upstream release.
- Binaries at https://arch.miffe.org/

miffe commented on 2013-03-18 10:49

linux-mainline 3.9rc3-1:
- New upstream release
- Binaries at http://arch.miffe.org/

miffe commented on 2013-03-11 15:24

linux-mainline 3.9rc2-1:
- New upstream release
- Binaries at http://arch.miffe.org/

miffe commented on 2013-03-04 14:06

linux-mainline 3.9rc1-1:
- New upstream release
- Binaries at http://arch.miffe.org/

miffe commented on 2013-03-01 18:27

@KaiSforza: I know, but someone always mark this as out of date as soon as the .1 release is out and I was tired of unmarking it. I also had som requests to enable some config values that where on in core/linux so I released this.

KaiSforza commented on 2013-03-01 17:49

Just nit-picking, but this kernel is not a mainline kernel. It's a stable kernel.

miffe commented on 2013-03-01 11:29

linux-mainline 3.8.1-1:
- New upstream release
- Resynced against linux/core 3.8.1-1
- Binaries at http://arch.miffe.org

miffe commented on 2013-02-20 10:37

linux-mainline 3.8-1:
- New upstream release
- Binaries at http://arch.miffe.org/

@forest76: I won't deviate the config from the core/linux package. I create too much of a maintenence burden. But once it's enabled in the linux package it will be enabled here too.

KaiSforza commented on 2013-02-19 19:05

@dhave

Those are not mainline kernels. Look in the Arch Linux Rollback Machine for old stable kernels.

dhave commented on 2013-02-19 18:59

I'm testing some UEFI-related problems on my Arch system, and I'd like to get binaries for linux-mainline kernel 3.7.7, 3.7.8 and 3.7.9. Does anyone know where I could find those? Thanks.

forest76 commented on 2013-02-15 14:08

Please, add f2fs support (module only) to yours binary build. I know I can build kernel from source according with PKGBUILD, but on my pentium mobile 1.3 device it takes a LONG time (about 4 hours).

miffe commented on 2013-02-09 20:25

linux-mainline 3.8rc7-1:
- New upstream release
- Binaries at http://arch.miffe.org/

@Limoto: ATH9K is enabled, CONFIG_ATH9K=m is in the config.

KaiSforza commented on 2013-02-05 16:44

@Limoto: It's not in the default config. I put it in my 3.8-rc6 if you want to grab that one.
http://sakuscans.com/pacmanpkg/x86_64/

Limoto commented on 2013-02-05 14:11

Why linux-mainline 3.8rc6-1 doesn't include ath9k driver?

miffe commented on 2013-02-02 14:27

linux-mainline 3.8rc6-1:
- New upstream release
- Binaries at http://arch.miffe.org/

Rulatir commented on 2013-02-02 12:57

Prettyvanilla, thanks! Your PKGBUILD works like a charm. I was actually using NVIDIA's original installer downloaded from their drivers page; this used to work just fine in the past. I suck at PKGBUILDs so I didn't try repackaging the driver until you kindly did the work for me :)

prettyvanilla commented on 2013-02-02 01:31

linux-api-headers is not kernel-dependent, so you don't need a special package.
Please make sure that you really have the correct packages KaiSforza mentioned, which correctly contain /usr/src/linux-3.8.0-3-rc6-mainline-dirty as the source directory name.
A quick test on my end based on the nvidia-304xx-ck PKGBUILD compiled just fine.
PKGBUILD: http://pastebin.com/GdQkNkHk
nvidia-304xx-mainline.install: http://pastebin.com/xCCtbGgr

Rulatir commented on 2013-02-02 00:06

Or alternatively, how can I have a working nvidia-304xx driver with linux-mainline?

Rulatir commented on 2013-02-02 00:05

... and how do I obtain the latter for linux-mainline?

Rulatir commented on 2013-02-02 00:02

I swear that when I went to that page it didn't look like a homepage of a repo - it was just a server-generated directory listing.

I set up the repo with pacman and installed the packages, and it still doesn't work, and the reason it doesn't work is that the file is **really, really, really** missing from the packages.

nvidia-installer log file '/var/log/nvidia-installer.log'
creation time: Sat Feb 2 00:56:13 2013
installer version: 304.64

PATH: /usr/lib/ccache/bin/:/usr/local/bin:/usr/bin:/bin:/usr/local/sbin:/usr/sbin:/sbin:/opt/android-sdk/tools:/opt/java/bin:/opt/java/db/bin:/opt/java/jre/bin:/opt/kde/bin:/usr/bin/vendor_perl:/usr/bin/core_perl:/opt/qt/bin

nvidia-installer command line:
./nvidia-installer

Using: nvidia-installer ncurses user interface
-> License accepted.
-> Installing NVIDIA driver version 304.64.
-> There appears to already be a driver installed on your system (version: 304.64). As part of installing this driver (version: 304.64), the existing driver will be uninstalled. Are you sure you want to continue? ('no' will abort installation) (Answer: Yes)
-> Would you like to register the kernel module sources with DKMS? This will allow DKMS to automatically build a new module, if you install a different kernel later. (Answer: No)
-> Performing CC sanity check with CC="cc".
-> Performing CC version check with CC="cc".
ERROR: The kernel header file '/lib/modules/3.8.0-3-rc6-mainline-dirty/build/include/linux/version.h' does not exist. The most likely reason for this is that the kernel source files in '/lib/modules/3.8.0-3-rc6-mainline-dirty/build' have not been configured.
ERROR: Installation has failed. Please see the file '/var/log/nvidia-installer.log' for details. You may find suggestions on fixing installation problems in the README available on the Linux driver download page at www.nvidia.com.

[rulatir@berbelek ~]$ LANG=en_US ls /lib/modules/3.8.0-3-rc6-mainline-dirty/build/include/linux/version.h
ls: cannot access /lib/modules/3.8.0-3-rc6-mainline-dirty/build/include/linux/version.h: No such file or directory

KaiSforza commented on 2013-02-02 00:00

Also, a quick pkgfile reveals that linux/version.h is not in the linux-*headers package. It is in the linux-api-headers package.

KaiSforza commented on 2013-02-01 23:17

wgiokas linux-mainline 3.8.rc6.0.g88b62b9-3
wgiokas linux-mainline-docs 3.8.rc6.0.g88b62b9-3
wgiokas linux-mainline-headers 3.8.rc6.0.g88b62b9-3

Stop using the archive, Rulatir. Go to the link that prettyvanilla was kind enough to give you.

KaiSforza commented on 2013-02-01 23:16

wgiokas linux-mainline 3.8.rc6.0.g88b62b9-3
wgiokas linux-mainline-docs 3.8.rc6.0.g88b62b9-3
wgiokas linux-mainline-headers 3.8.rc6.0.g88b62b9-3

Rulatir commented on 2013-02-01 23:06

Nope, not even trying. I just opened the packages and saw that the main package provides the directory:

/usr/src/linux-3.8.0-3-rc5-mainline-dirty

and the headers package provides the directory:

/usr/src/linux-3.8.0-1-rc5-mainline-dirty

Since they are both rc5 anyway, and only differ in pkgrel, I looked at the headers-rc5-3 package, and it provides the pkgrel 3 directory, but that combination was actually my very first attempt and it didn't work.

And it didn't work because <linux/version.h> is **actually missing** from all these -headers packages.

Rulatir commented on 2013-02-01 22:50

OK, trying again with main package -3 and headers package -2 because there is no -3.

Rulatir commented on 2013-02-01 22:47

There is no pkgrel 3 for headers!

Rulatir commented on 2013-02-01 22:45

pacman -Qi says:

Nazwa : linux-mainline
Wersja : 3.8.rc6.0.g88b62b9-2

Nazwa : linux-mainline-headers
Wersja : 3.8.rc6.0.g88b62b9-2

These are the packages currently installed, and I have just uninstalled all linux-mainline related stuff and installed just these two. I am currently running what can only possibly be the kernel that came in the currently installed linux-mainline package. This kernel however seems to have problems with self-identification:

[rulatir@berbelek ~]$ uname -r
3.8.0-1-rc5-mainline-dirty

KaiSforza commented on 2013-02-01 22:09

Rulitar: Get the 3.8-rc6 packages. you need the same headers and kernel. (pkgrels 1 and 2 in repo were bad. pkgrel 3 is good.)

Rulatir commented on 2013-02-01 22:06

Thanks, installed these, but still:

nvidia-installer log file '/var/log/nvidia-installer.log'

[snip]

-> Performing CC version check with CC="cc".
ERROR: The kernel header file '/lib/modules/3.8.0-1-rc5-mainline-dirty/build/include/linux/version.h' does not exist. The most likely reason for this is that the kernel source files in '/lib/modules/3.8.0-1-rc5-mainline-dirty/build' have not been configured.
ERROR: Installation has failed. Please see the file '/var/log/nvidia-installer.log' for details. You may find suggestions on fixing installation problems in the README available on the Linux driver download page at www.nvidia.com.

Rulatir commented on 2013-02-01 22:04

nvidia-installer log file '/var/log/nvidia-installer.log'
creation time: Fri Feb 1 22:59:12 2013
installer version: 304.64

PATH: /usr/lib/ccache/bin/:/usr/local/bin:/usr/bin:/bin:/usr/local/sbin:/usr/sbin:/sbin:/opt/android-sdk/tools:/opt/java/bin:/opt/java/db/bin:/opt/java/jre/bin:/opt/kde/bin:/usr/bin/vendor_perl:/usr/bin/core_perl:/opt/qt/bin

nvidia-installer command line:
./nvidia-installer

Using: nvidia-installer ncurses user interface
-> License accepted.
-> Installing NVIDIA driver version 304.64.
-> There appears to already be a driver installed on your system (version: 304.64). As part of installing this driver (version: 304.64), the existing driver will be uninstalled. Are you sure you want to continue? ('no' will abort installation) (Answer: Yes)
-> Would you like to register the kernel module sources with DKMS? This will allow DKMS to automatically build a new module, if you install a different kernel later. (Answer: No)
-> Performing CC sanity check with CC="cc".
-> Performing CC version check with CC="cc".
ERROR: The kernel header file '/lib/modules/3.8.0-1-rc5-mainline-dirty/build/include/linux/version.h' does not exist. The most likely reason for this is that the kernel source files in '/lib/modules/3.8.0-1-rc5-mainline-dirty/build' have not been configured.
ERROR: Installation has failed. Please see the file '/var/log/nvidia-installer.log' for details. You may find suggestions on fixing installation problems in the README available on the Linux driver download page at www.nvidia.com.

prettyvanilla commented on 2013-02-01 20:37

They are available in KaiSforza's repo: http://sakuscans.com/pacmanpkg/x86_64/
which is still updated regularly.

Rulatir commented on 2013-02-01 20:30

Need headers too, to build older NVIDIA drivers...

miffe commented on 2013-01-26 20:57

linux-mainline 3.8rc5-1:
- New upstream release
- Binaries at http://arch.miffe.org/

@TheSadClown: This uses the same config as the core/linux package, so your wireless interface should work. If it still doesn't work with rc5, try configuring it manually, uncomment the "make menuconfig" line and check if the driver for your wifi adapter is enabled.

Anonymous comment on 2013-01-26 07:49

Package seems to work fine, but I lose wireless interface when booting it. Anyone know why this may be?

miffe commented on 2013-01-18 13:54

linux-mainline 3.8rc4-1:
- New upstream release
- Fixed script permissions, reported by cian1500ww, thanks!
- Binaries at http://arch.miffe.org/

cian1500ww commented on 2013-01-14 11:43

All files in the /usr/src/linux-3.8.0-1-mainline/scripts/ have no execution permissions. If you look at core/linux, you'll see that some of these should have execution permissions. This is causing the nvidia driver installer to stop working as it needs to execute some of these scripts. I'm using the default kernel config.

miffe commented on 2013-01-10 23:26

linux-mainline 3.8rc3-1:
- New upstream release
- Binaries at http://arch.miffe.org/

miffe commented on 2013-01-04 17:10

linux-mainline 3.8rc2-1:
- New upstream release
- Binaries at the usual place
- Rebased against 3.7.1-2

KaiSforza commented on 2013-01-03 22:08

v3.8-rc2 released here:

http://sakuscans.com/pacmanpkg/x86_64/

(See 00README for more information)

KaiSforza commented on 2012-12-22 20:16

v3.8-rc1 release
Grab linux-mainline, linux-mainline-headers and linux-mainline-docs here:

http://sakuscans.com/pacmanpkg/x86_64/

(See 00README for more information)

KaiSforza commented on 2012-12-22 20:12

v3.8-rc1 release
Grab linux-mainline, linux-mainline-headers and linux-mainline-docs here:

http://sakuscans.com/pacmanpkg/x86_64/

(See 00README for more information)

miffe commented on 2012-12-13 00:07

I rebase against the linux package every rc1, so those new config directives will be added then, if they are enabled there.

If you really need them now, you can use the 3.7 package in testing, or recompile.

KaiSforza commented on 2012-12-12 21:04

Note to those that want a nice config/headers:

Use the testing/linux package. It is at 3.7, and their configs include all the stuff you're looking for. (The pkgbuild also includes headers and docs)

exe2 commented on 2012-12-12 20:59

no webcam support and no support for cgroups memory :(

miffe commented on 2012-12-11 10:58

linux-mainline 3.7-1:
- New upstream release
- Binaries at http://arch.miffe.org/

miffe commented on 2012-12-10 12:23

linux-mainline 3.7rc8-1:
- New upstream release
- Binaries at http://arch.miffe.org

miffe commented on 2012-12-10 12:22

linux-mainline 3.7rc8-1:
- New upstream release
- Binaries at http://arch.miffe.

KaiSforza commented on 2012-12-05 00:28

Binaries for 3.7rc8 released here (64-bit only): http://sakuscans.com/pacmanpkg/x86_64/


Split pkgbuild updated: https://github.com/KaiSforza/archlinux-pkgbuilds/blob/master/kernel/linux-mainline/PKGBUILD (modify for pacman 4.0.3)

miffe commented on 2012-11-26 10:01

linux-mainline 3.7rc7-1:
- New upstream release
- Binaries at http://arch.miffe.org/

KaiSforza commented on 2012-11-24 08:30

I've decided to edit the pkgbuild from here, making it more similar to that of the original linux package. Some things moved around, so I re-located them in the package() function, basically 'un-disabling' the stuff in this pkgbuild. Note, the config in that git directory is customized, so please use your own.

It is also a split package, with linux-mainline{,-headers,-docs} as well. (If you are using pacman-git and feel adventurous, the master branch has a fun pkgbuild you can use utilizing the new vcs abilities. Good if you've already got the linux git repo cloned for some reason.)

https://github.com/KaiSforza/archlinux-pkgbuilds/blob/linux-mainline-non-git/kernel/linux-mainline/PKGBUILD

Thank you,
KaiSforza

miffe commented on 2012-11-18 17:05

linux-mainline 3.7rc6-1:
- New upstream release
- Binaries at http://arch.miffe.org/

bjo commented on 2012-11-17 11:20

@donniezazen:

The repo name is miffe, so it has to be 'miffe' in the brackets.
You can also see this when you look at the db-file in http://arch.miffe.org/x86_64/.

donniezazen commented on 2012-11-16 06:24

I couldn't find information about your repository in Unofficial User Repository wiki. Will it be following?

[linux-mainline]
Server = http://arch.miffe.org/x86_64/

frigaut commented on 2012-11-13 18:22

oh, that may be why I lost my webcam with 3.7-rc5 then...

Anonymous comment on 2012-11-13 02:45

Ditto to no uvcvideo modules.

mereghost commented on 2012-11-12 17:40

The uvcvideo modules don't seem to be in the package, but they are on the source (modules/drivers/usb/video)

miffe commented on 2012-11-12 12:33

linux-mainline 3.7rc5-1:
- New upstream release
- Binaries at http://arch.miffe.org/

miffe commented on 2012-11-05 10:01

linux-mainline 3.7rc4-1:
- New upstream release
- Binaries at http://arch.miffe.org/

ahamilton55 commented on 2012-10-31 18:40

Are there supposed to be two md5sum entries in the PKGBUILD? This makes it confusing when trying to add another patch to the kernel if you only edit the obvious md5sum entry you get "ERROR: Integrity checks (md5) differ in size from the source array."

miffe commented on 2012-10-29 10:00

linux-mainline 3.7rc3-1:
- New upstream release
- Binaries at http://arch.miffe.org/

donniezazen commented on 2012-10-22 18:44

Thank you very much.

miffe commented on 2012-10-22 17:09

@donniezazen: It does. The headers are included in the linux-mainline package.

donniezazen commented on 2012-10-22 07:19

The binary package uploaded on your server, it doesn't have headers, does it?

miffe commented on 2012-10-21 16:58

linux-mainline 3.7rc2-1:
- New upstream release
- Binaries at the usual place

miffe commented on 2012-10-21 15:13

@donniezazen: Since AUR can't do splitted packages, the headers are included in the linux-mainline package.

donniezazen commented on 2012-10-21 07:46

Thanks for the package binaries. Would it be possible to also upload linux-headers for mainline along with linux-mainline?

cian1500ww commented on 2012-10-16 20:24

@miffe

Didn't know that, thanks.

miffe commented on 2012-10-16 20:22

linux-mainline 3.7rc1-2:
- Added uapi headers, thanks cian1500ww.
- Binaries at http://arch.miffe.org/

@cian1500ww: -j can easily be set for all packages in /etc/makepkg.conf

cian1500ww commented on 2012-10-16 17:38

The new 3.7 kernel has cleanups to the uapi include files. The package_headers() function in the current PKGBUILD doesn't copy the new location of the uapi include files to the kernel source directory. I've made a patch that corrects this. I've also added a check for the number of processors which is then passed to make to compile the kernel concurrently on machines with more than one processor. http://pastebin.com/0rpKTVhw

miffe commented on 2012-10-15 18:58

linux-mainline 3.7rc1-1:
- New upstream release
- Based of the [core] 3.6,2 package
- Webcams are enabled again
- Binaries at http://arch.miffe.org/

agapito commented on 2012-10-15 08:20

3.7 rc1

http://www.phoronix.com/scan.php?page=news_item&px=MTIwNjY

agapito commented on 2012-10-02 16:37

I've installed 3.6 from testing. Thanks.

miffe commented on 2012-10-02 14:02

@agapito: Yep, it got lost at 3.6rc6. I'll readd it for 3.7rc1. If you need it before then, either recompile or use linux from [testing] which has it enabled.

agapito commented on 2012-10-01 18:05

Webcams and dvb adapters doesn't work again!

miffe commented on 2012-10-01 15:00

linux-mainline 3.6-1:
- New upstream release
- Binaries at http://arch.miffe.org/

miffe commented on 2012-09-25 11:22

linux-mainline 3.6rc7-1:
- New upstream release
- Binaries at http://arch.miffe.org/

miffe commented on 2012-09-17 14:23

linux-mainline 3.6rc6-1:
- New upstream release
- Rebased against the linux 3.5.4-1 package. This should fix the depmod problems.
- Binaries at http://arch.miffe.org/

frigaut commented on 2012-09-16 23:48

@teekay: me too. Even though I have the latest kmod 10-1 from [core]

also, rc6 is out ! but I'm not flagging out of date yet ;-)

teekay commented on 2012-09-13 17:01

3.6-rc4 was fine, but since 3.6-rc5 the PKGBUILD breaks at running depmod.
I had to use

depmod -b "$pkgdir/usr" ..

instead of just -b $pkgdir

miffe commented on 2012-09-10 11:13

linux-mainline 3.6rc5-1:
- New upstream release
- Binaries at http://arch.miffe.org/

miffe commented on 2012-09-03 09:25

linux-mainline 3.6rc4-1:
- New upstream release
- Enabled CONFIG_MEDIA_CAMERA_SUPPORT
- Binaries at http://arch.miffe.org/

agapito commented on 2012-09-02 13:28

I compiled myself rc4, and my webcam and dvd are working again, thanks to tarkasteve.

Anonymous comment on 2012-09-01 06:54

I can confirm that the current config breaks webcams and TV tuners. The problem is the new flag CONFIG_MEDIA_CAMERA_SUPPORT (and similar tuner flags), which enables/disables all camera support but defaults to N. I've regenerated the config based on the existing one but with new/updated flags set as appropriate. Diff for x86_64 is here: http://pastebin.com/WbYpPrg1

agapito commented on 2012-08-25 03:44

@marijus: From linux-tv wikipedia.

The V4L API is essentially a kernel interface for analog video capture and output drivers, and is applicable to most video streams that are not related to DVB devices (see explanation below) or graphics cards -- though, the V4L API also entails a few oddities such as analog radio and RDS services. Examples of devices which fall within the scope of V4L are, amongst others, webcams, MPEG encoders/decoders, digital video streamers, analog TV tuners and video capture devices.

Anonymous comment on 2012-08-24 07:57

i also have no webcam since 3.6
@agapito: what has v4l-dvb to do with webcam?

agapito commented on 2012-08-23 08:34

Installing v4l-dvb-git solved my webcam problem, but not my dvb tuner.

miffe commented on 2012-08-23 08:28

linux-mainline 3.6rc3-1:
- New upstream release
- Binaries at http://arch.miffe.org/

agapito commented on 2012-08-23 06:18

My webcam and my dvb tuner doesn't work with 3.6 kernel, I've tried rc2 and rc3. Has anyone else this problem?

miffe commented on 2012-08-17 09:21

linux-mainline 3.6rc2-1:
- New upstream release
- Binaries at http://arch.miffe.org/

haagch commented on 2012-08-15 20:05

Hm, bug in fuse...

Can somebody please test encfs? It has a weird issue here.
If I create a file file.txt with
echo "foobar" > encfs_mount/file.txt
I get an I/O error.
But touching and then editing a file works.
Also moving files to an encfs mount with mv gives an I/O error, but not moving directories.

Is this a bug in encfs or is just something wrong with fuse?

Vi0L0 commented on 2012-08-06 22:50

there's a bug in fuse which prevent to create new file on ntfs partition.
here's the patch if someone is interested:
http://sourceforge.net/mailarchive/message.php?msg_id=29630599

miffe commented on 2012-08-03 12:37

linux-mainline 3.6rc1-1:
- New upstream release
- Binaries at http://arch.miffe.org/

miffe commented on 2012-07-31 10:19

linux-mainline 3.5-1:
- New upstream release
- Rebased against arch package 3.5-2
- Binaries at the usual place

Wilco commented on 2012-07-16 07:51

@miffe: thanks for the reply, system was up-to-date but I had to reinstall kmod for some reason. Tested the kernel a few hours and haven't found any issues (yet)

miffe commented on 2012-07-15 17:11

@ridikulus_rat: Should be fixed now.

@Wilco: You need the lastest kmod with the /lib -> /usr/lib stuff.

ridikulusrat commented on 2012-07-15 12:53

@miffe: regarding "repo-add --files" I still get "warning: failed to download: http://arch.miffe.org/x86_64//miffe.files.tar.gz" with nosr.

Wilco commented on 2012-07-15 12:11

makepkg gives an error:
MKDIR /mnt/data/linux/linux-mainline/pkg/lib/firmware/yam/
INSTALL /mnt/data/linux/linux-mainline/pkg/lib/firmware/yam/1200.bin
INSTALL /mnt/data/linux/linux-mainline/pkg/lib/firmware/yam/9600.bin
DEPMOD 3.5.0-rc7-mainline
ERROR: could not open directory /mnt/data/linux/linux-mainline/pkg/lib/modules/3.5.0-rc7-mainline: No such file or directory
FATAL: could not search modules: No such file or directory
==> ERROR: A failure occurred in build().
Aborting...

miffe commented on 2012-07-15 07:32

linux-mainline 3.5rc7-1:
- New upstream release
- Binaries at http://arch.miffe.org/

Samsagax commented on 2012-07-15 00:17

@miffe: glibc with the /lib -> /usr/lib change already hit [core] you should upgrade the PKGBUILD

dhave commented on 2012-07-14 23:04

Once you've finished working through the /lib -> /usr/lib changeover which has now moved out of testing phase, you'll probably need to do as miffe says here:

" ... please change _libdir on row 31 to /usr/lib instead".

In doing the /lib -> /usr/lib changeover, I uninstalled linux-mainline and nvidia-mainline, since linux-mainline was claiming ownership of files in /lib. Then, after I did the changeover successfully (see the Arch home page and the Arch forum for discussion), I reinstalled linux-mainline and nvidia-mainline. But I needed to change the "_libdir" line as miffe describes above.

dcelasun commented on 2012-07-13 12:43

@Cdh: I'm still having the error with net/built-in.o. What exactly did you change to make it work? Using the config from the binary didn't seem to help.

dhave commented on 2012-07-09 06:10

Thanks, miffe. The package built and installed just fine. I'm having some probs with wifi, but I'm pretty sure that has nothing to do with your work.

dhave commented on 2012-07-09 04:41

Thanks, miffe. The package built and installed just fine. I'm having some probs with wifi, but I'm pretty sure that has nothing to do with your work.

miffe commented on 2012-07-09 03:10

linux-mainline 3.5rc6-2:
- Fixed a problem with /lib vs /usr/lib handling

If you use the /lib elimination stuff from [testing], please change _libdir on row 31 to /usr/lib instead

@dhave: This version should work for you.

miffe commented on 2012-07-08 20:27

@dhave: Sorry, it's a bug with the package. I'll upload a newer version later today.

dhave commented on 2012-07-08 19:08

I'm pretty sure this is what you folks are discussing, but I'm having trouble following. Anyway, here's the error I'm getting, which pretty much is saying that the modules aren't being found, I think:

------------

==> Generating module dependencies
install: cannot stat '/usr/lib/modules/3.5.0-rc6-mainline//modules.builtin': No such file or directory
install: cannot stat '/usr/lib/modules/3.5.0-rc6-mainline//modules.order': No such file or directory
WARNING: could not open /tmp/mkinitcpio.nLcbtV/root/lib/modules/3.5.0-rc6-mainline/modules.order: No such file or directory
WARNING: could not open /tmp/mkinitcpio.nLcbtV/root/lib/modules/3.5.0-rc6-mainline/modules.builtin: No such file or directory
==> Creating gzip initcpio image: /boot/initramfs-linux-mainline.img
==> WARNING: errors were encountered during the build. The image may not be complete.

---------------

Two questions: (1) What has changed that this is happening (I've never had problems with linux-mainline builds and installation before)? (2) What can I do to make the problem go away?

Thanks.

miffe commented on 2012-07-08 13:52

@ridikulus_rat: The modules are in /lib/modules. I've added the file lists, thanks for the tip.

ridikulusrat commented on 2012-07-08 13:09

@miffe: Does the linux-mainline 3.5rc6-1 package in http://arch.miffe.org have the modules in /usr/lib/modules/ or /lib/modules/ ?

Also can you create a list of files of all the packages in the [miffe] repo (repo-add --files) for use with pkgfile or nosr ?

ridikulusrat commented on 2012-07-08 13:08

@miffe: Does the package in http://arch.miffe.org have the modules in /usr/lib/modules/ or /lib/modules/ ?

Also can you create a list of files of all the packages in the [miffe] repo (repo-add --files) for use with pkgfile or nosr ?

miffe commented on 2012-07-08 12:23

linux-mainline 3.5rc6-1:
- New upstream release
- Binaries at http://arch.miffe.org

Uncomment line 166 in the PKGBUILD if you use the /lib elimination things from testing.

donniezazen commented on 2012-07-08 04:43

3.5-rc6 is out. Thanks.

haagch commented on 2012-07-05 07:29

mkinitcpio had big problems finding the modules for me.

I first had to put them to where the modules of the repo kernel are:
sudo ln -s /lib/modules/3.5.0-1-mainline /usr/lib/modules/
And then the modules.* were not ready for gzipped kernel modules so I needed to run depmod
sudo depmod -a 3.5.0-1-mainline

Then all was fine.

miffe commented on 2012-07-02 01:43

linux-mainline 3.5rc5-1:
- New upstream release
- Binaries at http://arch.miffe.org/

It is vacation time, so the next release might be late if i'm not at home.

miffe commented on 2012-06-26 08:30

linux-mainline 3.5rc4-1:
- New upstream release
- Binaries at the usual place.

miffe commented on 2012-06-19 22:41

linux-mainline 3.5rc3-1:
- New upstream release
- Binaries at http://arch.miffe.org/

progandy commented on 2012-06-16 17:55

Please include this patch to fix issues with the backlight brightness setting.
I can confirm it resolevs the problem on my laptop.

https://bugzilla.kernel.org/show_bug.cgi?id=43168#c15

haagch commented on 2012-06-14 19:38

With the config from the binary it works.

For the lazy: http://pastebin.com/Ekk4pRCJ
Save as config.x86_64, and comment the line.
# yes ""|make oldconfig # lazy version of the above

haagch commented on 2012-06-14 08:26

I'm getting this with the
yes "" | make config
method:

LD [M] net/netfilter/ipvs/ip_vs.o
LD net/netfilter/built-in.o
LD net/built-in.o
==> ERROR: A failure occurred in build().
Aborting..

No warnings as far as I see... I use the binary, but I am trying to apply two patches to the radeon and as they both apply in drivers/gpu/drm/radeon/ I don't think they cause a failure for net/built-in.o
Does the kernel build create a more verbose log somewhere? I didn't see one...

miffe commented on 2012-06-10 17:38

linux-mainline 3.5rc2-1:
- New upstream release
- Rebased against arch linux 3.4.2-1
- Removed i915 and acer patches, thansk mus for the report
- Binaries at http://arch.miffe.org/

mus commented on 2012-06-08 13:13

the i915 and acer patches have both been merged upstream and are not needed anymore.

miffe commented on 2012-06-04 22:16

@KaNe23: You need to configure your bootloader and add an entry for linux-mainline. Look at the arch wiki page for your bootloader.

miffe commented on 2012-06-03 20:30

linux-mainline 3.5rc1-1:
- New upstream release
- Config based on linux 3.4 arch package
- Binaries at http://arch.miffe.org/ as usual

miffe commented on 2012-05-30 18:08

@Samsagax: When I updated to 3.4, the tarball was missing, so it uses 3.3 and the 3.4 patch instead.

Samsagax commented on 2012-05-30 17:46

Noob question: Shouldn't _basekernel be 3.4 instead of 3.3 ?

miffe commented on 2012-05-21 15:28

linux-mainline 3.4-1:
- New upstream release
- Binaries at http://arch.miffe.org/

miffe commented on 2012-05-13 12:29

linux-mainline 3.4rc7-1:
- New upstream release
- Binaries at http://arch.miffe.org/

zwastik commented on 2012-05-12 12:40

Do you guys have any problem with hibernating? my computer cannot boot after running pm-hibernate :L
The only solution for this is to unplug the power source wait a few seconds and the plug again the power source cable. After this, when you boot the computer there is a warning message from the bios about incorrect setting of the cpu frequency. Then I am prompted to the bios configuration screen but all settings look ok. After restart the computer boots normally and resume successfully from hibernating.

I will also mail this message to aur-general mailing list so I can attach any usefull log file.

zwastik commented on 2012-05-12 12:38

Do you guys have any problem with hibernating? my computer cannot boot after running pm-hibernate :L
The only solution for this is to unplug the power source wait a few seconds and the plug again the power source cable. After this, when you boot the computer there is a warning message from the bios about incorrect setting of the cpu frequency. Then I am prompted to the bios configuration screen but all settings look ok. After restart the computer boots normally and resume successfully from hibernating.

miffe commented on 2012-05-07 11:21

linux-mainline 3.4rc6-1:
- New upstream release
- Binaries at http://arch.miffe.org/ as usual

cian1500ww commented on 2012-05-06 23:07

3.4-rc6 is out: http://lwn.net/Articles/496079/

cian1500ww commented on 2012-05-06 23:04

3.4-rc6 is out: http://lwn.net/Articles/496079/

miffe commented on 2012-04-30 09:31

linux-mainline 3.4rc5-1:
- New upstream release
- Binaries at http://arch.miffe.org/

@waldeck: I don't make changes to the .config, this package uses the same .config as the linux package. So once the change is made there it will show up here to.

Anonymous comment on 2012-04-28 09:36

Please add CONFIG_NETFILTER_XT_TARGET_LOG=m.
This configuration is new in 3.4 and required for iptables -j LOG.
Otherwise you will get "iptables: No chain/target/match by that name".
Thanks!

miffe commented on 2012-04-22 23:32

linux-mainline 3.4rc4-1:
- New upstream release
- Binaries at http://arch.miffe.org/

miffe commented on 2012-04-22 23:30

linux-mainline 3.4rc3-1:
- New upstream release
- Binaries at http://arch.miffe.org/

miffe commented on 2012-04-16 13:17

linux-mainline 3.4rc3-1:
- New upstream release
- Binaries at http://arch.miffe.org
- Shoutout to frigaut :D

frigaut commented on 2012-04-16 10:11

... but I'm sure you will be on top of it very very soon.
Thanks for maintaining this package, miffe.

frigaut commented on 2012-04-16 10:10

I flagged "out of date". rc3 came out today. sorry:-)

miffe commented on 2012-04-08 15:25

linux-mainline 3.4rc2-1:
- New upstream release
- Binaries at http://arch.miffe.org/

haagch commented on 2012-04-03 06:48

Maybe this is my problem, will test later: https://bugzilla.kernel.org/show_bug.cgi?id=43022

haagch commented on 2012-04-03 06:41

Quick question: Does suspend to ram work for anyone? For me it suspends fine but then boots normally when trying to wake up.

miffe commented on 2012-04-02 10:54

linux-mainline 3.4rc1-1:
- New upstream release.
- Rebased against arch linux 3.3-1 package.
- Binaries at http://arch.miffe.org

farnsworth commented on 2012-03-21 20:23

I have a kernel panic during boot with this last release.
going back to rc7 solve the problem (I've just uncommented lines related to rc7 in PKGBUILD and changed the md5).
Am I missing something?

farnsworth commented on 2012-03-21 19:59

I have a kernel panic during boot with this last release.
going back to rc7 solve the problem (I've just uncommented lines related to rc7 in PKGBUILD and changed the md5).
Am I missing something?

Wilco commented on 2012-03-20 13:16

@miffe: yes, I found that out while checking the PKGBUILD. Anyway, thanks for package, it's working fine so far.

miffe commented on 2012-03-20 09:56

@Wilco: I don't make any changes to the config, I just use the config from the linux packages. So once brcmsmac is enabled there it will show up here to.

Wilco commented on 2012-03-20 09:51

linux-mainline misses the brcmsmac module. Can you add it in future versions? There are lots of fixes for broadcom wireless chips in this module.
The following must be enabled in the kernel config:
Device drivers -> Broadcom specific AMBA -> BCMA support
Device drivers -> Broadcom specific AMBA -> BCMA support -> Support for BCMA on PCI-host bus

After enabling this option the brcmsmac driver can be created as a module:
Device drivers -> Network device support -> Wireless LAN -> Broadcom IEEE802.11n PCIe SoftMAC WLAN driver

miffe commented on 2012-03-19 09:29

linux-mainline 3.3-1:
- New upstream release
- Binaries at http://arch.miffe.org/

farnsworth commented on 2012-03-12 19:19

It's working fine, thanks!

miffe commented on 2012-03-12 14:02

linux-mainline 3.3rc7-1:
- New upstream release
- Rebased against 3.2.9-1
- Reenabled autofs4
- Binaries at http://arch.miffe.org/

miffe commented on 2012-03-09 17:43

@farnsworth: It didn't compile a few version ago, so I disabled it. I'll enable it again if it works in the next release.

farnsworth commented on 2012-03-09 17:05

Hello, I have an issue with autofs:
when I launch automount I have this return: "/usr/sbin/automount: test mount forbidden or incorrect kernel protocol version, kernel protocol version 5.00 or above required."
seems like the autofs4 module is missing, in config.gz I've have this:
# CONFIG_AUTOFS4_FS is not set
Is there any problem with it? Is it possible to add it?
thanks for your help

farnsworth commented on 2012-03-09 17:02

I have an issue with autofs since some release:
when I launch automount I have this return: "/usr/sbin/automount: test mount forbidden or incorrect kernel protocol version, kernel protocol version 5.00 or above required."
seems like the autofs4 module is missing
It was working fine in rc3 and maybe 4. any idea?
thanks

farnsworth commented on 2012-03-09 16:56

I have an issue with autofs since some release:
when I launch automount I have this return: "/usr/sbin/automount: test mount forbidden or incorrect kernel protocol version, kernel protocol version 5.00 or above required."
seems like the autofs4 module is missing
It was working fine in rc3 and maybe 4. any idea?
thanks

farnsworth commented on 2012-03-09 16:53

I have an issue with autofs since some release:
when I launch automount I have this return: "/usr/sbin/automount: test mount forbidden or incorrect kernel protocol version, kernel protocol version 5.00 or above required."
It was working fine in rc3 and maybe 4. any idea?
thanks

Anonymous comment on 2012-03-08 11:08

Many thanks for this PKGBUILD and the binaries.

miffe commented on 2012-03-04 21:06

linux-mainline 3.3rc6-1:
- New upstream release
- Binaries at http://arch.miffe.org/

cowai commented on 2012-03-01 15:56

I made a package that fixes rc6 issues and wifi kernel panics tailored for asus zenbooks: https://aur.archlinux.org/packages.php?ID=57197

cowai commented on 2012-02-29 11:05

@Canute

I tried to use the patch you provided.

"patching file drivers/gpu/drm/i915/i915_gem.c
Hunk #1 succeeded at 3084 (offset -3 lines).
patching file drivers/gpu/drm/i915/intel_tv.c
Hunk #1 succeeded at 1153 (offset -114 lines).
patching file kernel/printk.c
patching file drivers/gpu/drm/i915/intel_display.c
patch: **** malformed patch at line 18: *dev_priv)"

Anonymous comment on 2012-02-27 22:45

The patch @cowai is talking about was applied to 3.3rc5, however it contains a small bug due to operator precedence.
The following patch has _not_ been applied to 3.3rc5

http://lists.freedesktop.org/archives/intel-gfx/2012-February/015319.html

miffe commented on 2012-02-26 04:35

linux-mainline 3.3rc5-1:
- New upstream release
- Binaries at http://arch.miffe.org/

cowai commented on 2012-02-20 02:05

This patch I think should be in there: http://lists.freedesktop.org/archives/intel-gfx/2012-February/015131.html
intel RC6 is unstable with deep rc6. This patch disables it.

miffe commented on 2012-02-19 17:11

linux-mainline 3.3rc4-1:
- New upstream release
- Binaries at http://arch.miffe.org/

miffe commented on 2012-02-12 19:32

@vasilito: Once the linux package is updated to include that, it will appear here too.

vasilito commented on 2012-02-12 08:01

As far as I understand bcma and brcsmac can coexist on the most recent kernel.
Please configure the mainline with the following options to support recent MacBook notebooks:

+CONFIG_B43_BCMA=y
+CONFIG_B43_BCMA_PIO=y
+CONFIG_B43_PHY_HT=y
+CONFIG_BCMA=m
+CONFIG_BCMA_BLOCKIO=y
+CONFIG_BCMA_HOST_PCI_POSSIBLE=y
+CONFIG_BCMA_HOST_PCI=y

vasilito commented on 2012-02-12 07:56

As far as I understand bcma and brcsmac can coexist on the most recent kernel.
Please configure the mainline with the following options to support recent MacBook notebooks:

+CONFIG_B43_BCMA=y
+CONFIG_B43_BCMA_PIO=y
+CONFIG_BCMA=m
+CONFIG_BCMA_BLOCKIO=y
+CONFIG_BCMA_HOST_PCI_POSSIBLE=y
+CONFIG_BCMA_HOST_PCI=y

vasilito commented on 2012-02-12 07:55

As far as I understand bcma and brcsmac can coexist on the most recent kernel.
Please configure the mainline with the following options to support recent MacBook notebooks:

+CONFIG_B43_BCMA=y
+CONFIG_B43_BCMA_PIO=y
+CONFIG_BCMA=m
+CONFIG_BCMA_BLOCKIO=y
+CONFIG_BCMA_HOST_PCI_POSSIBLE=y
+CONFIG_BCMA_HOST_PCI=y

miffe commented on 2012-02-09 14:21

linux-mainline 3.3rc3-1:
- New upstream release
- Merged linux 3.2.5-1 package
- Binaries at http://arch.miffe.org/

@taylorchu: I hope this eats less battery for you.

miffe commented on 2012-02-05 19:53

@taylorchu: I do use the linux packages config, but I don't merge it for every release, so if the linux packages was resently updated, it might not have showed up here yet. I'll do a merge for rc3.

taylorchu commented on 2012-02-05 19:28

could you also use default archlinux kernel config?
you dont seem to enable some powersave features, so my laptop battery dies really quickly.

miffe commented on 2012-02-01 00:01

linux-mainline 3.3rc2-1:
- New upstream release
- Binaries at http://arch.miffe.org/

miffe commented on 2012-01-31 11:02

@polslinux: The headers are included in the linux-mainline package. They aren't splitted out since AUR doesn't like splitted packages.

polslinux commented on 2012-01-31 10:17

please could you make your pkgbuild similar to the official so we get also the header?
Some applications need it...thanks :)

ridikulusrat commented on 2012-01-21 14:35

Can anyone compile this kernel with CONFIG_EFI_STUB=y and CONFIG_EXT4_USE_FOR_EXT23=y ? Thanks in advance. I tried compile with localmodconfig but it excludes many modules that I load only when needed.

miffe commented on 2012-01-20 23:53

linux-mainline 3.3rc1-1:
- New upstream release
- Based on arch kernel 3.2.1-1
- Binaries at http://arch.miffe.org/

agapito commented on 2012-01-20 01:15

Linux 3.3-rc1 Released

miffe commented on 2012-01-16 08:55

Unflagging OOD, since 3.3-rc1 isn't out yet.

miffe commented on 2012-01-05 13:46

linux-mainline 3.2-1:
- New upstream release
- Binaries at http://arch.miffe.org/

frigaut commented on 2011-12-30 01:26

Oh, I see: you mean BCMA has to be disabled at kernel config, not just blacklisted...
will try.

frigaut commented on 2011-12-30 01:22

Somehow I don't have brcmsmac either.
It is missing from the module tree:

494:~ $ cd /lib/modules/3.2.0-1-mainline/
495:3.2.0-1-mainline $ find . -name brcmsmac
496:3.2.0-1-mainline $ find . -name mac80211
./kernel/net/mac80211
497:3.2.0-1-mainline $ find . -name b43
./kernel/drivers/net/wireless/b43
498:3.2.0-1-mainline $

I have bcma blacklisted.

it seems enabled as module in my config.gz (assuming it's the one that the mainline PKGCONFIG starts from):
499:3.2.0-1-mainline $ zcat /proc/config.gz | grep BRCMS
CONFIG_BRCMSMAC=m

Radioactiveman commented on 2011-12-29 23:11

@fleet: You have to disable BCMA to get brcmsmac. Read this: http://www.h-online.com/open/features/Kernel-Log-Coming-in-3-2-Part-1-Networking-1379085.html

miffe commented on 2011-12-25 02:33

linux-mainline 3.2rc7-1:
- New upstream release
- Binaries at http://arch.miffe.org/

Vi0L0 commented on 2011-12-19 00:54

Very nice set of patches (working very well here on 3.2rc6) to "Reduce compaction-related stalls and improve asynchronous migration of dirty pages v6", usefull mostly when using USB stick with VFAT:
https://lkml.org/lkml/2011/12/14/254

miffe commented on 2011-12-17 16:55

linux-mainline 3.2rc6-1
- New upstream release.
- Binaries at http://arch.miffe.org/

miffe commented on 2011-12-11 23:34

linux-mainline 3.2rc5-1:
- New upstream release.
- Binaries at http://arch.miffe.org/

fleet commented on 2011-12-11 09:54

brcmsmac is missing in this kernel.

miffe commented on 2011-12-02 21:19

@fooblahblah: I don't make changes to the config, but once that options has been set in the linux package, it will show up here too.

fooblahblah commented on 2011-12-02 21:07

Can we get the following config option added by default? It enabled the changes in B43 to make newer MacbookPro Wifi work.

CONFIG_B43_PHY_HT=y

I've tested it on my Macbookpro8,2 and it seems to work well.

miffe commented on 2011-12-02 11:26

linux-mainline 3.2rc4-1:
- New upstream release
- Binaries at http://arch.miffe.org/

miffe commented on 2011-11-24 09:37

linux-mainline 3.2rc3-1:
- New upstream release
- Binaries at http://arch.miffe.org

mar04 commented on 2011-11-20 21:58

@Cdh: yup, happened to me twice today, htop, top, gnome-system-monitor, all showed constant 100% usage on 2 of 4 cores without reason, and I could only fix it by rebooting

mar04 commented on 2011-11-20 21:52

@Cdh: yup, happened to me twice today, htop, top, gnome-system-monitor, all showed constant 100% usage on 2 of 4 cores without reason, and I could only fix it by rebooting

miffe commented on 2011-11-19 17:05

@tydell: This package uses the same config as the linux package, so once the driver is turned on there it will be on here too. You can easily build with a custom config, just uncomment the make menuconfig line and build as usual.

tydell commented on 2011-11-19 11:20

3.2-rc1 release (http://www.kernel.org/pub/linux/kernel/v3.x/) seems to have a few more drivers included for some devices, dvb tuners etc (like dvb-as102 module, that I am interested)

I also see there is 3.2-rc2 release.

Could You make package with these new modules ? It would be very helpfull for many people.

Thanks in advance.

miffe commented on 2011-11-16 23:42

linux-mainline 3.2rc2-1:
- New upstream release
- Binaries at http://arch.miffe.org/

ridikulusrat commented on 2011-11-15 18:57

Can someone please compile iwlwifi module alone for 3.2rc1-1 and post it somewhere (dropbox maybe)? TIA

ridikulusrat commented on 2011-11-15 18:42

https://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commitdiff;h=3c607d27c818cf4a5d28f2c73b18a88f8fbdfa33 - iwlagn renamed to iwlwifi

ridikulusrat commented on 2011-11-12 16:47

I can't find iwlagn kernel module in 3.2rc1 (x86_64 pkg downloaded from miffe's repo). I can't connect to wifi in 3.2rc1 because of this (no wlan0 device), but can do so in 3.1-4 [core] kernel. Any idea wats wrong? TIA

ridikulusrat commented on 2011-11-12 16:46

I can't find iwlagn kernel module in 3.2rc1 (x86_64 pkg downloaded from miffe's repo). I can't connect to wifi in 3.2rc1 because of this, but can do so in 3.1-4 [core] kernel. Any idea wats wrong? TIA

haagch commented on 2011-11-08 13:47

With 3.2 the utilities that display cpu usage sometimes go crazy. Can somebody confirm?

miffe commented on 2011-11-08 10:21

linux-mainline 3.2rc1-1:
- New upstream release.
- Package rebased against 3.1-4
- Binaries available at http://arch.miffe.org/

miffe commented on 2011-10-24 13:12

linux-mainline 3.1-1:
- New upstream release
- Back to using kernel.org
- Binaries available at http://arch.miffe.org

miffe commented on 2011-10-18 16:50

linux-mainline 3.1rc10-1:
- New upstream release
- Binaries available at the usual place

miffe commented on 2011-10-16 09:39

@Kev: It seems the tarballs downloaded from github aren't always the same, so the md5sum will vary. It will be fixed once kernel.org starts working again.

Kev commented on 2011-10-15 20:50

md5 checksum does fail on linux-3.1-rc9.tar.gz. Tried several times as this shouldn't happen.. Can someone reproduce this issue?

JonnyJD commented on 2011-10-06 09:23

iscsi target is new in linux 3.1 and once the regular linux package reaches 3.1 (which should be soon), I will file a report there (if it is not included then anyways).

miffe commented on 2011-10-06 06:59

@JonnyJD: linux-mainline uses the same config as the regular linux package, so once it's added there it will appear here too.

JonnyJD commented on 2011-10-06 01:31

I included the new iscsi target module in my config (which works with lio/targetcli: https://aur.archlinux.org/packages.php?ID=52905)

CONFIG_TARGET_CORE=m
CONFIG_TCM_IBLOCK=m
CONFIG_TCM_FILEIO=m
CONFIG_TCM_PSCSI=m
+ CONFIG_ISCSI_TARGET=m
CONFIG_LOOPBACK_TARGET=m

Might be a nice thing to include in the pkgbuild.

miffe commented on 2011-10-05 18:03

linux-mainline 3.1rc9-1:
- New upstream release
- Binaries by me at the usuall place

dcelasun commented on 2011-09-26 11:03

Nvm, just realized I've updated the package without rebooting.

dcelasun commented on 2011-09-26 11:00

@miffe: That's weird, since VirtualBox can't find the kernel headers.

miffe commented on 2011-09-26 10:57

@dcelasun: Since AUR doesn't support split packages the headers are included in the linux-mainline package.

dcelasun commented on 2011-09-26 10:28

I've just compiled this but it doesn't seem to create a linux-mainline-headers package. Any ideas?

miffe commented on 2011-09-25 10:18

@markino7200: There seems to be an issue with tarballs from github, they sometimes contain the incorrect version. The one in the PKGBUILD should be the correct one according to https://github.com/torvalds/linux/commits/v3.1-rc7

Anonymous comment on 2011-09-24 10:44

kernel checksum and upack path seem wrong in PKGBUILD?
new checksum is '8d33f2d3cae613c49573be50d16aca7a'
and unpack path is "$srcdir/torvalds-linux-4cf670b"

miffe commented on 2011-09-23 08:58

linux-mainline 3.1rc7-1:
- New upstream release
- Binaries by foutrelis available.

miffe commented on 2011-09-16 08:43

linux-mainline 3.1rc6-1:
- New upstream release
- Binaries by foutrelis
- Using github since kernel.org is still down

miffe commented on 2011-09-06 22:01

linux-mainline 3.1rc5-1:
- New upstream release
- Binaries by foutrelis
- Downloads from Github since kernel.org is down due a security breach

miffe commented on 2011-08-29 19:19

linux-mainline 3.1rc4-1:
- New upstream release
- Binaries from foutrelis

@Det: Looks like they finally applied the patch you needed.

miffe commented on 2011-08-24 15:08

linux-mainline 3.1rc3-1:
- New upstream release
- Binaries provided by foutrelis

Det commented on 2011-08-16 11:27

I still need the devtmpfs.c patch (the second link).

miffe commented on 2011-08-16 07:30

Binaries are up now, thanks to foutrelis!

miffe commented on 2011-08-16 06:44

linux-mainline 3.1rc2-1:
- New upstream release
- No binaries for now

jelly commented on 2011-08-15 19:04

http://www.kernel.org/pub/linux/kernel/v3.x/testing/ url with rc2

Det commented on 2011-08-12 12:20

Btw. If anybody else is having these 'mismatches' with their static 3.1rc1 kernels, as I did, here are patches for them:
http://choon.net/forum/read.php?21,293988
https://patchwork.kernel.org/patch/996132/ (this one was posted even before 3.0.1 but still not in)

Anonymous comment on 2011-08-10 14:07

Works now without problems. Thank you =)

miffe commented on 2011-08-09 20:47

linux-mainline 3.1rc1-2:
- Fixed mkinitcpio preset (thanks Wubbbi for the report)

It seems the linux packager lost the sed line to fix the kernel filename in the mkinitcpio preset in the linux -> kernel26 change and I didn't notice since it still works if you have the linux package installed. I fixed the binaries and shall file a bugreport for the linux package.

Anonymous comment on 2011-08-09 19:15

Hey I have a problem with the new Binarys:

>>> Updating module dependencies. Please wait ...
>>> Generating initial ramdisk, using mkinitcpio. Please wait...
==> Building image from preset: 'default'
-> -k /boot/vmlinuz-linux -c /etc/mkinitcpio.conf -g /boot/initramfs-linux-mainline.img
==> ERROR: invalid kernel specifier: `/boot/vmlinuz-linux'
==> Building image from preset: 'fallback'
-> -k /boot/vmlinuz-linux -c /etc/mkinitcpio.conf -g /boot/initramfs-linux-mainline-fallback.img -S autodetect
==> ERROR: invalid kernel specifier: `/boot/vmlinuz-linux

All versions befor, worked without problems. Any Idea? :)

miffe commented on 2011-08-09 17:37

linux-mainline 3.1rc1-1:
- New upstream release
- Rebased against linux 3.0.1-1
- Binaries by foutrelis

Anonymous comment on 2011-07-24 17:22

It's in testing as 'linux 3.0-1'.
See pacman -Qi linux: http://pastebin.com/F0FeEJhH

miffe commented on 2011-07-22 08:54

@skodabenz: It's already there, in the package_kernel26 function.

Anonymous comment on 2011-07-22 08:38

@miffe: I think its better to add replaces=('kernel26-mainline') to the package for those who use your repo.

foutrelis commented on 2011-07-22 08:08

@skodabenz: The kernel maintainers are preparing a 'linux' package for Linux 3.0:

http://projects.archlinux.org/svntogit/packages.git/tree/linux/trunk

Anonymous comment on 2011-07-22 08:05

[quote]Based on the linux package[/quote]

There seems to be no 'linux'package anywhere. Seems like the devs have not decided a new package name.

miffe commented on 2011-07-22 06:52

linux-mainline 3.0-1:
- New upstream release
- Binaries provided by foutrelis
- Based on the linux package
- Goodbye kernel26-mainline


Repository available at http://arch.miffe.org

[miffe]
Server = http://arch.miffe.org/$arch/