Package Details: linux-aufs_friendly 4.6.3-1

Git Clone URL: https://aur.archlinux.org/linux-aufs_friendly.git (read-only)
Package Base: linux-aufs_friendly
Description: The aufs-compatible linux kernel and modules
Upstream URL: http://www.kernel.org/
Licenses: GPL2
Submitter: Calimero
Maintainer: Calimero
Last Packager: Calimero
Votes: 51
Popularity: 0.658897
First Submitted: 2011-08-08 16:32
Last Updated: 2016-06-27 14:51

Pinned Comments

Calimero commented on 2016-05-23 18:42

For some time now, OverlayFS has been able to do everything I've been using AUFS for. I'll continue maintaining this, but I won't be using it myself and suggest that AUFS users take a look at the latest version of OverlayFS (which can do several branches, too).
If you have any actual reasons to use AUFS because of features OverlayFS doesn't have, such as multiple RW branches in round-robin, please do comment here.

Calimero commented on 2015-02-17 16:49

Before marking this out-of-date: please do so only if the build breaks.
The PKGBUILD is dynamically self-updating so that it builds the last version regardless, and a script on my server updates every 6 hours the apparent version on AUR for the convenience of AUR helpers. It will still build the latest version in every case.

sadkisson, Drarrel: this is not my bug. Try to build core/linux and you'll get the same error. So fix your keyring or --skippgpcheck.

Latest Comments

tsuflux commented on 2016-05-28 18:07

@Calimero, I am using AUFS with the create=pmfs policy for multiple RW branches. I would change to OverlayFS but I believe that this feature is not supported.

Calimero commented on 2016-05-26 12:35

I can't reproduce.

justin8 commented on 2016-05-26 00:17

This isn failing to build at the moment:

The next patch would create the file mm/prfile.c,
which already exists! Assume -R? [n] n
Apply anyway? [n]
Skipping patch.
1 out of 1 hunk ignored
==> ERROR: A failure occurred in prepare().

Calimero commented on 2016-05-23 18:42

For some time now, OverlayFS has been able to do everything I've been using AUFS for. I'll continue maintaining this, but I won't be using it myself and suggest that AUFS users take a look at the latest version of OverlayFS (which can do several branches, too).
If you have any actual reasons to use AUFS because of features OverlayFS doesn't have, such as multiple RW branches in round-robin, please do comment here.

Drarrell commented on 2016-04-20 07:50

The build failed on the check of PGP signatures again.

linux-4.5.tar ... FAILED (unknown public key 79BE3E4300411886)
patch-4.5.1 ... FAILED (unknown public key 38DBBDC86092693E)
==> ERROR: One or more PGP signatures could not be verified!
==> ERROR: Makepkg was unable to build linux-aufs_friendly.

WarheadsSE commented on 2015-08-31 18:48

@Mutantx
Arch Linux ARM already has the aufs patch set for many of the kernels.

Mutantx commented on 2015-08-31 17:51

I've installed all dependencies but am encountering when attempting to install:

rsync: change_dir "/armv7h/core/linux" (in abs) failed: No such file or directory (2)
rsync error: some files/attrs were not transferred (see previous errors) (code 23) at main.c(1637) [Receiver=3.1. 1]

I did attempt to add arm7h to PKGBUILD arch.

Any ideas?

FadeMind commented on 2015-08-17 09:37

4.1.6 released https://www.kernel.org/pub/linux/kernel/v4.x/ChangeLog-4.1.6

ValdikSS commented on 2015-04-30 04:53

Please update the package for Linux 4.0 and aufs4.

sadkisson commented on 2015-02-19 00:41

understood, thanks for explaining

Calimero commented on 2015-02-17 16:49

Before marking this out-of-date: please do so only if the build breaks.
The PKGBUILD is dynamically self-updating so that it builds the last version regardless, and a script on my server updates every 6 hours the apparent version on AUR for the convenience of AUR helpers. It will still build the latest version in every case.

sadkisson, Drarrel: this is not my bug. Try to build core/linux and you'll get the same error. So fix your keyring or --skippgpcheck.

sadkisson commented on 2015-02-17 01:41

==> Verifying source file signatures with gpg...
linux-3.18.tar ... FAILED (unknown public key 79BE3E4300411886)
patch-3.18.6 ... FAILED (unknown public key 38DBBDC86092693E)
==> ERROR: One or more PGP signatures could not be verified!
==> ERROR: Makepkg was unable to build linux-aufs_friendly.

Calimero commented on 2014-09-20 11:00

This package is just a bootstrap to the linux PKGBUILD from [core], from which I believe that the AUFS patches should never have been removed (or should have been put back, now that it is maintained flawlessly). Since we don't have an alternative in the mainline kernel today (and archiso has become a silly dmraid-based hack).
If only AUFS could have been merged in mainline!
This being said, please consider this as just building linux from ABS. (that's what it actually does)

First, run `makepkg --verifysource` to fetch the ABS sources. Then edit the 'core-linux/PKGBUILD.core' file to configure the kernel to your liking (uncomment `make nconfig` or whatever else suits you). And run `makepkg`.

dalu commented on 2014-09-20 01:58

What do you think about a non-preempt kernel for servers?

Calimero commented on 2014-08-22 14:27

@delta575 https://wiki.archlinux.org/index.php/Arch_User_Repository#Prerequisites

delta575 commented on 2014-08-21 17:01

I can’t build linux-aufs_friendly, it says:

makepkg -s
receiving file list … done
created directory core-linux
./
PKGBUILD
change-default-console-loglevel.patch
compal-laptop-hwmon-fix.patch
config
config.x86_64
linux.install
linux.preset

sent 160 bytes received 330,238 bytes 60,072.36 bytes/sec
total size is 329,627 speedup is 1.00
/home/delta/linux-aufs_friendly/PKGBUILD: line 42: patch: command not found

hooger commented on 2014-08-13 19:24

I get an error in the prepare() stopping the build process. http://pastebin.com/raw.php?i=zcFqBDru

Calimero commented on 2014-05-23 15:04

The maintainer sometimes posts sha256, sometimes md5, which broke the package. Updated to follow whatever is used.

ValdikSS commented on 2014-05-22 15:35

ERROR: Integrity checks (md5) differ in size from the source array.

GI_Jack commented on 2014-05-22 03:03

==> ERROR: Integrity checks (md5) differ in size from the source array.

GI_Jack commented on 2014-05-02 19:07

ahh thanks

Calimero commented on 2014-05-02 08:52

You won't be able to compile core/linux either for the same reason.
Use gcc 4.9 from testing.

GI_Jack commented on 2014-05-01 23:20

gcc: error: unrecognized command line option ‘-fstack-protector-strong’
/home/jack/archpackages64/linux-aufs_friendly/src/linux-3.14/./Kbuild:35: recipe for target 'kernel/bounds.s' failed
make[1]: *** [kernel/bounds.s] Error 1
Makefile:884: recipe for target 'prepare0' failed
make: *** [prepare0] Error 2
==> ERROR: A failure occurred in prepare().
Aborting...

GI_Jack commented on 2014-05-01 22:55

gcc: error: unrecognized command line option ‘-fstack-protector-strong’

dvant commented on 2014-04-11 21:15

Hi All

I'm getting integrity check errors when i try to install

==> Making package: linux-aufs_friendly 3.14-4 (Fri Apr 11 17:14:37 EDT 2014)
==> Checking runtime dependencies...
==> Checking buildtime dependencies...
==> Retrieving sources...
-> Found linux-3.14.tar.xz
-> Found config
-> Found config.x86_64
-> Found linux.preset
-> Found change-default-console-loglevel.patch
-> Found 0001-Bluetooth-allocate-static-minor-for-vhci.patch
-> Found 0002-module-allow-multiple-calls-to-MODULE_DEVICE_TABLE-p.patch
-> Found 0003-module-remove-MODULE_GENERIC_TABLE.patch
-> Found 0004-fs-Don-t-return-0-from-get_anon_bdev.patch
-> Found 0005-Revert-Bluetooth-Enable-autosuspend-for-Intel-Blueto.patch
-> Found 0006-genksyms-fix-typeof-handling.patch
-> Found 0007-Fix-the-use-of-code32_start-in-the-EFI-boot-stub.patch
-> Found aufs3-base.patch
-> Found aufs3-standalone.patch
-> Found aufs3-mmap.patch
-> Found add-aufs-patches.diff
==> ERROR: Integrity checks (md5) differ in size from the source array.

darude commented on 2014-03-15 18:41

Hey guys,
I tried to instal three evenings in a row but get only timeouts when downloading.
It's really slow at downloading the file, I got 50mbit/s downspeed.

==> Making package: linux-aufs_friendly 3.13.6-1 (Sat Mar 15 18:47:23 CET 2014)
==> Checking runtime dependencies...
==> Checking buildtime dependencies...
==> Retrieving sources...
-> Downloading linux-3.13.tar.xz...
% Total % Received % Xferd Average Speed Time Time Time Current
Dload Upload Total Spent Left Speed
100 178 100 178 0 0 236 0 --:--:-- --:--:-- --:--:-- 236
96 73.6M 96 71.3M 0 0 25802 0 0:49:51 0:48:18 0:01:33 0a
curl: (18) transfer closed with 2387944 bytes remaining to read
==> ERROR: Failure while downloading linux-3.13.tar.xz
Aborting...
==> ERROR: Makepkg was unable to build linux-aufs_friendly.

GI_Jack commented on 2014-02-26 21:04

patching file fs/file_table.c
Hunk #1 FAILED at 36.
Hunk #2 FAILED at 360.
Hunk #3 succeeded at 34

Calimero commented on 2014-02-22 18:18

It's not that the variable doesn't get parsed, sanerb; but rather that there is no 3.13.x branch yet. I assumed that there would be one, but actually the branch to use cannot be inferred directly from the kernel version. (that would at least need to check in the repository to decide)

Temporarily fixed. I'll see if a final solution can be found.

sanerb commented on 2014-02-22 16:15

404 on all of the patches due to incorrect fetch url.

for instance:
http://sourceforge.net/p/aufs/aufs3-standalone/ci/aufs$%7B_kernel%7D/tree/aufs3-base.patch?format=raw

the proper link is, for instance:
http://sourceforge.net/p/aufs/aufs3-standalone/ci/aufs3.13/tree/aufs3-base.patch?format=raw

it appears the _kernel variable isn't being properly parsed. likely because:

_kernel="${pkgver%.*}.x"
pkgver=3.13.4

so instead of trying to fetch from .../ci/aufs3.13/... it ends up trying to fetch from .../ci/aufs3.13.4/... (i presume?).

loper commented on 2014-02-22 13:12

If this couldn't find patch:
_source=("aufs3-base.patch::http://sourceforge.net/p/aufs/aufs3-standalone/ci/aufs${_kernel}/tree/aufs3-base.patch?format=raw"

Simply add this variable:
_kernel=3.13

Please make a bugfix for this, cause it wouldn't compile

WarheadsSE commented on 2014-01-20 19:39

Well, my case allowed for that, but, Okay.

Calimero commented on 2014-01-20 19:38

My bad, I overlooked the branch hierarchy with .x for subversions. Thanks for pointing it out.
The fix you proposed isn't exact though — Arch never releases a *.*.0 kernel, it's always a *.* kernel in such cases. I fixed it the right way!

sanerb commented on 2014-01-20 18:57

confirming that the fix provided by WarheadsSE does work.

for those curious, if the three lines he provided are placed in the PKGBUILD between lines 9 and 10, the patch completes fine. this has been tested and confirmed on two different installs.

WarheadsSE commented on 2014-01-20 15:30

It looks like a small change is needed to the PKGBUILD for detecting the patch set.

if [[ "$pkgver" = *.*.0 ]]; then _kernel="${pkgver%.*}";
elif [[ "$pkgver" = *.*.* ]]; then _kernel="${pkgver%.*}.x";
else _kernel="${pkgver}"; fi

This is based on the comparison of:
http://sourceforge.net/p/aufs/aufs3-standalone/ci/aufs3.12/tree/aufs3-mmap.patch
http://sourceforge.net/p/aufs/aufs3-standalone/ci/aufs3.12.x/tree/aufs3-mmap.patch

This is compiled on my system at this time.

Calimero commented on 2014-01-15 12:07

It was that, or a package I would update manually, possibly not up to date — but that would always build.

This error is with the AUFS patches on the kernel — you now have to wait for upstream to be updated, I can't do anything.

You can still build a 3.12.6 kernel or something, though.
Just fetch its sources manually from packages.archlinux.org, and extract them in a folder "core-linux", that you will put in a folder where the sources for this package are, before running makepkg.

lmuscariello commented on 2014-01-15 10:36

Experiencing the same failure:
patching file mm/fremap.c
Hunk #1 FAILED at 207.
1 out of 1 hunk FAILED

patch-3.12.7-2 modifies fremap.c before AUFS3 patches

WarheadsSE commented on 2014-01-14 19:23

http://ix.io/9Rv

Fails to patch against linux 3.12.7-2

patching file mm/fremap.c
Hunk #1 FAILED at 207.
1 out of 1 hunk FAILED

Brainiarc7 commented on 2014-01-06 10:39

The port has been opened and I can now build the package.

Cheers. Good work.

Brainiarc7 commented on 2013-12-28 08:04

Right.
It seems port 873 (priviledged) is blocked by the firewall in use.

Calimero commented on 2013-12-27 19:41

Not at all, the issue is on your side (proxy blocking rsync, or something). Nothing ever changed with ABS.

Brainiarc7 commented on 2013-12-27 02:33

There's a problem encountered during makepkg -c -s P_KGBUILD and Packer's packer -S linux-aufs_friendly

Code:

packer -S linux-aufs_friendly
rsync: failed to connect to rsync.archlinux.org (89.238.67.251): Connection refused (111)
rsync: failed to connect to rsync.archlinux.org (2a00:1828:2000:547::2): Network is unreachable (101)
rsync error: error in socket IO (code 10) at clientserver.c(128) [Receiver=3.1.0

It seems rsync.archlinux.org requires authentication for a connection to be eshtablished. Please update the PKGBUILD to reflect the same so the package can be built.

synthetic commented on 2013-11-25 23:26

Awesome, thank you very much for obliging us never satisfied users! :-)

dressguard commented on 2013-11-21 18:10

Thanks a bunch, Calimero!

hans_jschmid commented on 2013-11-21 18:05

This is frigging awesome!!!! Thanks Calimero!

Sometimes it pays off to being a pain in the backside. ;-)

Calimero commented on 2013-11-21 17:03

For the convenience of demanding users, my server now dutifully updates the package version if needed, automatically, every 6 hours. :)

Calimero commented on 2013-11-21 16:59

My server now dutifully updates the package version if needed automatically now, every 6 hours.

Calimero commented on 2013-11-21 13:54

Note that the version displays when you start building, at the very beginning. Just stop the build right away if it hasn't increased.
As for versions, you'll have noticed that this PKGBUILD just bootstraps core/linux's build. So if linux in core isn't updated, this isn't either.

Also, hey. All GIT PKGBUILDs on AUR do the same thing, and nobody's complaining. I even said I'd make a cron somewhere to make this perfect when I have some time.

Calimero commented on 2013-11-21 13:54

Note that the version displays when you start building, at the very beginning. Just stop the build right away if it hasn't increased.
As for versions, you'll have noticed that this PKGBUILD just bootstraps core/linux's build. So if linux in core isn't updated, this isn't either.

Also, hey. All GIT PKGBUILDs on AUR do the same thing, and nobody's complaining. I even said I'd make a cron somewhere when I have some time.

Calimero commented on 2013-11-21 13:51

Note that the version displays when you start building, at the very beginning. Just stop the build right away if it hasn't increased.

Anyway I'm quite busy so didn't set up automated updates yet. As for versions, you'll have noticed that this PKGBUILD just bootstraps core/linux's build. So if linux in core isn't updated, this isn't either.

Calimero commented on 2013-11-21 13:47

Hey, wait! The version displays when you start building, at the very beginning.
Just stop the build right away if it hasn't increased…
Anyway I'm quite busy so didn't set up automated updates yet.

dressguard commented on 2013-11-21 08:13

http://www.kernel.org/ is showing a new kernel version 3.12.1 but I'm still getting just 3.12(.0) after an hour of compilation.

So how do I find out when it's time to upgrade? This is really getting messy. Sorry to say that.

Calimero commented on 2013-11-20 22:00

Yes, no need for an explanation, I understand the version issue.
As I said, that would require intervention on my side each time a new kernel minor version is released, which happens very often.

I created the dynamic bootstrap PKGBUILD so the new kernel be available as soon as it is in [core], but it has this downside.
When I have some time, I could set up a cron job on a server to update the pkgver in this PKGBUILD — I can't do it dynamically in AUR, it needs external intervention.

synthetic commented on 2013-11-20 20:23

Thanks Calimero, I didn't know about having to clear /tmp/yaourt-tmp-$USER. It all works now. docker here I come! ;-)

About the upgrade loop, I am avoiding it at the moment by putting linux-aufs_friendly on the IgnorePkg line in pacman.conf, but I agree with Hans that the versioning should be changed to normal non-wildcard numbers to avoid this situation.

Anyway thanks a lot for maintaining this package.

hans_jschmid commented on 2013-11-20 03:09

Hi Calimero, guess I didn't make myself clear enough what the issue is. Let me be more specific: I have the latest version of linux-aufs_friendly installed (3.12-1). Then I do a system upgrade with 'yaourt -Syua' (which is aliased to 'yaupg'). It tells me I'm not up-to-date and wants to upgrade from version 3.12-1 to 3.12.x-1. But upgrading to 3.12.x-1 results actually in version 3.12-1. So there is no need to upgrade. Even if I let it run and continue upgrading the actual installed version will again be 3.12-1 and not 3.12.x-1. So this results in an endless loop.

I upgrade my system on a daily basis. Now I have to tell yaourt every time I don't want to upgrade because I have already the latest version installed. Please get the version numbers right. There is no 3.12.x-1 version. The 'x' is causing the problems. It confuses yaourt and it cannot do its job. It's very annoying.

Thanks for maintaining the package. Very highly appreciated.

dressguard commented on 2013-11-19 17:31

So when I do a system upgrade from now on I have to check for myself if I'm having the current version of linux-aufs_friendly already installed? And I do have to explicitly say no to upgrading linux-aufs_friendly because it is already installed every time I do a system upgrade?

Calimero commented on 2013-11-19 09:06

This convenience would require that I run a cron job somewhere, that polls the version of linux in [core] and changes the PKGBUILD here consequently.

hans_jschmid commented on 2013-11-19 00:47

We need proper versioning. Otherwise yaourt and other package managers will upgrade every time we ask for a system upgrade.

Calimero commented on 2013-11-18 19:17

aufs3-mmap.patch is present in the PKGBUILD since November 6ᵗʰ, and it builds here.
Can't reproduce, can't fix! Make sure you erase the folder where you were building before trying again. (/tmp/yaourt-tmp-$USER, by default)

synthetic commented on 2013-11-18 15:59

I'm still having the issue originally posted by gcr (about vm_prfile). I'm using yaourt to build. Is there some manual intervention I need to take during build?

walterjwhite commented on 2013-11-12 20:15

Correction, packer appears to not work with linux-aufs_friendly. Yaourt appears to build the kernel just fine. I'll use yaourt until I find the problem with packer.

Calimero commented on 2013-11-12 20:05

You're not using AUR properly.
The PKGBUILD is not the only file to download from AUR to build the package. Use the complete source archive, it's the "Download tarball" link.

walterjwhite commented on 2013-11-12 16:44

I am still getting add-aufs-patches.diff : No such file or directory.

Calimero commented on 2013-11-10 14:54

Derp. Fixed.
I messed this up when reorganizing the PKGBUILD to allow manual intervention on the build.

mdevaev commented on 2013-11-10 12:18

# change the package basename to what's defined in this file
sed -i "s/^pkgbase=.*/pkgbase=${pkgbase}/" core-linux/PKGBUILD.core

--> .core <--
Fix it, plz.

Calimero commented on 2013-11-06 20:50

Thanks for your research. Should be fixed!

gcr commented on 2013-11-06 16:08

Digging deeper, it seems that
aufs3-mmap.patch
is not getting applied. This creates the problem -- this patch defines the 'vm_prfile' field.

This is definitely a bug in this package.

gcr commented on 2013-11-06 16:04

This package builds fine, but building the 'aufs3' package for the kernel module fails:

In file included from /store/tmp/aufs3/src/aufs3/fs/aufs/aufs.h:45:0,
from /store/tmp/aufs3/src/aufs3/fs/aufs/module.c:25:
/store/tmp/aufs3/src/aufs3/fs/aufs/file.h: In function ‘au_vm_prfile_set’:
/store/tmp/aufs3/src/aufs3/fs/aufs/file.h:302:5: error: ‘struct vm_area_struct’ has no member named ‘vm_prfile’

This package does not even contain the string 'vm_prfile'. What happened?

c00w commented on 2013-10-21 04:06

I'm getting this bizarre bug where it appears to not actually build anything?

==> Starting build()...
==> Entering fakeroot environment...
==> Starting package_linux-aufs_friendly()...
cp: cannot stat ‘arch/x86/boot/bzImage’: No such file or directory

Driadan commented on 2013-09-20 09:34

uhm, some errors here trying to build it:

It doesn't find some files and make says config is wrong :S

==> Iniciando build()...
/tmp/packerbuild-1000/aufs3/aufs3/PKGBUILD: línea 60: /usr/lib/modules/extramodules-3.11-aufs_friendly/version: No existe el fichero o el directorio
/tmp/packerbuild-1000/aufs3/aufs3/PKGBUILD: línea 61: /usr/lib/modules/extramodules-3.11-aufs_friendly/version: No existe el fichero o el directorio
Makefile:10: could not find kernel config file. internal auto-config may fail
config.mk:55: CONFIG_AUFS_SBILIST is enabled but CONFIG_AUFS_MAGIC_SYSRQ nor CONFIG_PROC_FS
config.mk:66: *** CONFIG_AUFS_HFSNOTIFY requires CONFIG_FSNOTIFY. Alto.
==> ERROR: Se produjo un error en build().
Cancelando...

ValdikSS commented on 2013-09-19 20:13

There is something wrong with this new kernel or with aufs3 module. It doesn't work.
aufs au_loopback_init:136:modprobe[639]: loop_backing_file() is not defined

Calimero commented on 2013-09-18 22:30

Because this PKGBUILD is now nothing but a bootstrap to a core/linux build with AUFS patches.

papanikge commented on 2013-09-17 14:47

@mereghost thanks man... but now I'm curious how does the mkinitcpio get created? I can't find anything related in the PKGBUILD.

mereghost commented on 2013-09-17 14:39

"this could sound stupid, but why isn't this a kernel module and it needs rebuilding the kernel..?"

AUFS support is a patch, so a rebuild is needed.

"wait. how am I gonna boot the new kernel? grub-menu-entry? There should be a message about this somewhere. "

It will create a new preset for mkinitcpio. You just need to regenerate the grub config file. Grub2 autodetection will do everything else. =)

papanikge commented on 2013-09-13 12:31

wait. how am I gonna boot the new kernel? grub-menu-entry? There should be a message about this somewhere.

afader commented on 2013-09-10 19:56

Will apparmor work with this kernel?

papanikge commented on 2013-09-09 18:41

this could sound stupid, but why isn't this a kernel module and it needs rebuilding the kernel..?

Calimero commented on 2013-09-09 11:24

I already answered that it's in the makedepends, but there is no visible solution to the actual issue behind this: the PKGBUILD is a hack, and adding rsync even to 'depends' (instead of 'makedepends') won't fix that.
I could hack more, but this isn't even an issue to me.

mnovick1988 commented on 2013-09-08 11:20

please add rsync to dependency list, please check the PKGBUILD.

Calimero commented on 2013-09-06 16:20

Now come on, this file is provided in the source archive that's on this very page, which you should have used to build. Just as for any other AUR package.

thermionix commented on 2013-09-06 08:08

patch: **** Can't open patch file add-aufs-patches.diff : No such file or directory

thermionix commented on 2013-09-06 07:56

hmm, weird;

$ packer -S lxc-docker
/tmp/packertmp-1000/linux-aufs_friendly.PKGBUILD: line 23: rsync: command not found

Calimero commented on 2013-09-05 20:41

Look at the PKGBUILD, it's in makedepends. Can I do any better than this?

thermionix commented on 2013-09-05 14:09

rsync missing from dependency list?

Calimero commented on 2013-08-30 11:07

What, why, where?
What does grub have to do with this package? For instance, I'm using syslinux.
Does the core/linux package provide what you're asking for, whatever it be?
Anyway, http://blog.mathieui.net/grub-mkconfig-brainlessness-considered-harmful.html

Calimero commented on 2013-08-30 11:04

What, why, where?
Does the core/linux package provide what you're asking for, whatever it be?
Anyway, http://blog.mathieui.net/grub-mkconfig-brainlessness-considered-harmful.html

Anonymous comment on 2013-08-30 00:41

Please add sth like
grub-mkconfig -o /boot/grub/grub.cfg

or any comment regarding this.

Took me a while to find out

Calimero commented on 2013-08-24 10:25

Yes, don't expect anything to compile properly with noexec.
This is no place for this issue; actually you should have noticed that you almost cannot build any AUR package.

Calimero commented on 2013-08-24 10:11

Yes, don't expect anything to compile properly with noexec.

Calimero commented on 2013-08-23 19:31

I precisely used yaourt to test. Works for me, can't reproduce.
The only reason I see is you are builting on NTFS, FAT, or some other non-POSIX filesystem implementation.
As fun would have it, this isn't even related to this package; yaourt -Sb linux will fail as well, in such conditions!

Calimero commented on 2013-08-23 19:30

I precisely used yaourt to test. Works for me, can't reproduce.
The only reason I see is you are builting on NTFS, FAT, or some other non-POSIX filesystem implementation.

Calimero commented on 2013-08-22 23:57

For instance, add this somewhere around "# Hack for AUR":
$EDITOR core-linux/PKGBUILD.core

GI_Jack commented on 2013-08-22 19:05

how do I get the old style PKGBUILD back, I need this because I do my own kernel config, and build it against the latest kernel on kernel.org, not -ARCH

Calimero commented on 2013-08-17 19:02

Introducing today the automagically-updating PKGBUILD that uses ABS.
Tested successfully, but report breakages if any!

Calimero commented on 2013-08-04 20:28

Oh, I see! Well, the AUR User Guidelines [1] clearly tell you that AUR packages implicitly depend on base-devel. That is, don't even think of building from AUR without it installed.
PS: this is an practically unmodified linux PKGBUILD aside from the patches. I won't add (make)deps that linux in [core] doesn't have!
[1] https://wiki.archlinux.org/index.php/AUR_User_Guidelines

Calimero commented on 2013-08-04 20:26

Oh, I see! Well, the AUR User Guidelines [1] clearly tell you that AUR packages implicitly depend on base-devel. That is, don't even think of building from AUR without it installed.
[1] https://wiki.archlinux.org/index.php/AUR_User_Guidelines

tlunter commented on 2013-08-03 17:02

Calimero: Tritlo means the program `patch` was missing. He was asking for it to be added to the make-depends. When installing Arch, patch is part of base-devel. If someone only installs base and not base-devel, then patch isn't included, thus he had issues.

Calimero commented on 2013-07-31 20:46

I can't reproduce that. Well, you didn't say how to produce the error.
Using my AUR helper, it works. Most probably, you downloaded only the PKGBUILD, and of course that won't work!

Tritlo commented on 2013-07-31 08:26

Got an error when building because patch was missing.

Calimero commented on 2013-07-19 11:08

What do you think the suggestion "# You can change pkgbase back to linux if you want to replace the one from [core]." is for? Works for nvidia, at least.

Alternatively, (it's faster) rebuild the nvidia package for this kernel (changing ARCH in the PKGBUILD to aufs_friendly). Also rename the package, to keep the nvidia driver for [core]'s kernel.

dlin commented on 2013-07-19 09:37

After install this kernel, I found I can not run nvidia graphics, is there any tips?

dlin commented on 2013-07-19 06:58

Could you give an example of boot loader settings?
I've tried grub, but failed.

title Arch Linux aufs friendly
kernel /boot/vmlinuz-linux-aufs_friendly root=/dev/sda2 ro
initrd /boot/initramfs-linux.img

mazzolino commented on 2013-07-11 06:48

@GI_Jack: Just created a -stbl package. Would you like to try it out? https://aur.archlinux.org/packages/linux-aufs_friendly-stbl/

Anonymous comment on 2013-07-06 00:42

I compiled the kernel with the default config and copied the files to my /boot/efi/EFI/arch directory. The new rEFInd automatically detects kernels, so I didn't need to write a new menuentry.
However, when I try to boot with the aufs-friendly kernel, it cannot find the root partition (/dev/sda1, though I have it by UUID). Furthermore, my keyboard does not work so I can not even use the recovery console.
I did receive two warnings when I was installing the package - that I am possibly missing firmwire for module bfa and aic94xx
I am probably missing something very obvious as this is my first custom kernel.

I am, of course, looking to use the kernel tor run Docker. I hope they switch to btrfs soon...

cedeon commented on 2013-07-04 09:02

I think this depends on base-devel, right?. I got a build error because I didn't have patch installed.

Calimero commented on 2013-05-18 20:56

Please comment if the build happens to fail!

GI_Jack commented on 2012-10-08 16:25

I understand your busy cal, but all I am asking is we get a package update every stable kernel release.(even numbers). I can update the patch level on my own.

rename the package to linux-aufs-stbl and only update based on PKGBUILDs from even numbered kernels.

fandrieu commented on 2012-07-22 10:34

Thanks for the updates Calimero :)

GI_Jack commented on 2012-07-17 16:41

Calimero, can you please orphan the AUFS packages so some other people who have more of an active intrest in maintaining it can do so?

fandrieu commented on 2012-07-17 13:38

Instead of using dl.free.fr, I'll try to keep the PKGBUILDs updated at this repo:

https://github.com/fandrieu/archlinux-aur/

Get the full zip here:

https://github.com/fandrieu/archlinux-aur/zipball/master

Anonymous comment on 2012-07-09 14:11

Thanks for the updated tar balls miztadux and Calimero, they worked great!

Anonymous comment on 2012-06-16 13:50

Calimero's work, uptaded to linux-3.4.2 & aufs3-3.4_20120611

http://dl.free.fr/mr44jOCRj
http://dl.free.fr/i0gxXKYh7

e990761596712b2239e25346a4900723 aur-linux-aufs_friendly.tar.xz
2df43502c7c051d4c0d1c4b884beb4c2 aur-aufs3.tar.xz

Anonymous comment on 2012-05-24 16:35

If somebody's interested, you can try this PKGBUILD with the newer stable patches:
just udpate the "patch*.xz" file definition (the "pkgver=" line and the second "md5sums" line)

For example, the diff to compile the latest 3.3.7:
17c17
< pkgver=${_basekernel}.4
---
> pkgver=${_basekernel}.7
37c37
< '9c4cc16f10b645fbb90f6c05ad388883'
---
> '622a3b43238559aeb778279969631260'

I have no expertise to ensure that's correct, but it compiles and seems to work. (i tested the aufs module with .5 .6 & .7)
I guess the stable fixes don't conflict with the aufs_friendly patches...at least for now.

Calimero commented on 2012-05-04 12:51

It's initramfs-linux-aufs_friendly.img of course.
You tried with the initramfs for [core]'s kernel!

dlin commented on 2012-05-04 09:26

Could you show me the correct GRUB/syslinux boot line?
mine is as following, but failed:

kernel /boot/vmlinuz-linux-aufs_friendly root=/dev/sda2 ro
initrd /boot/initramfs-linux.img

Calimero commented on 2012-04-30 11:44

Yes, sorry for being so sleepy.

GI_Jack commented on 2012-04-19 16:37

this gonna get maintained?

ImNtReal commented on 2012-01-30 18:15

It would be brilliant if AuFS could be mainlined, but at this point I don't think there's much hope of that. Thanks.

Calimero commented on 2012-01-28 11:12

They do work without a hitch, just checkout the right one.
Sorry for being sleepy and not following linux versions very fast… It wouldn't be like this if AUFS were kept in the [core] repository. But well, I guess that AUFS has to be merged to the linux tree for this to happen!

ImNtReal commented on 2012-01-24 20:39

Are the patches working on 3.2? I seem to be having trouble with the upstream patch for some reason.

Calimero commented on 2012-01-08 22:17

Fixed, I missed a part of the diff from linux's PKGBUILD. Thanks.

Anonymous comment on 2012-01-08 18:10

3.1.8-1 fails to build due to missing referenced patch file usb-add-reset-resume-quirk-for-several-webcams.patch
Since the system I'm using doesn't have a webcam attached, I commented out that patch on line 71 of the PKGBUILD and was able to build and install the kernel.

GI_Jack commented on 2011-11-15 19:07

3.1 support?

pretty please with a cherry on top?

GI_Jack commented on 2011-11-07 20:14

can you please add
provides = linux
to the .pkginfo file, so that drivers that need specific kernel versions work.

GI_Jack commented on 2011-11-07 20:14

can you please add
provides = linux
to the .pkginfo file, so that drivers that need specific kernel versions work.

GI_Jack commented on 2011-11-07 20:14

can you please add
provides = linux
to the .pkginfo file, so that drivers that need specific kernel versions work.

GI_Jack commented on 2011-11-07 20:14

can you please add
provides = linux
to the .pkginfo file, so that drivers that need specific kernel versions work.

Calimero commented on 2011-10-30 14:04

Updated to 3.0.7 to follow [core]'s linux package. Although 3.0.8 seems out.
Additional note: as of today, aufs3 patches are rejected against linux 3.1, but I'm confident this will be fixed soon.

GI_Jack commented on 2011-10-21 02:18

I get this when I try makepkg
make[2]: *** No rule to make target `net/core/secure_seq.o', needed by `net/core/built-in.o'. Stop.
make[1]: *** [net/core] Error 2

GI_Jack commented on 2011-10-21 02:18

I get this when I try makepkg
make[2]: *** No rule to make target `net/core/secure_seq.o', needed by `net/core/built-in.o'. Stop.
make[1]: *** [net/core] Error 2

Calimero commented on 2011-10-09 22:32

done, thanks for pointing that out.

td123 commented on 2011-10-09 22:19

please get rid of the pkgbase or rename it to something other than linux because it conflicts with what is in the repositories. thanks

Calimero commented on 2011-10-09 19:39

kernel.org is still a bit mixed up, so since 3.0.6 is in core I'm just following Arch.

GI_Jack commented on 2011-10-09 12:05

I have no idea why kernel 3.0.6 is mainline arch when the latest stable kernel on kernel.org as of now is 3.0.4

unflagging this, because 3.0.6 is NOT stable.

cuihao commented on 2011-10-09 10:46

core/linux 3.0.6-2 is out.
Update plz.

Calimero commented on 2011-09-26 14:55

Sure thing, I replaced the source url with a mirror of kernel.org.

deimos commented on 2011-09-25 07:29

Would it be wise to use the kernel source on github while kernel.org is on hiatus?

Calimero commented on 2011-09-11 08:53

I used makepkg --source, which named the folder linux, after $pkgbase. (pkgbase is kept like this in order to build this package as "linux" for seamless updates from installable liveCDs)
Corrected manually, thanks.

Calimero commented on 2011-09-11 08:52

I used makepkg --source, which namend it named the folder linux, after $pkgbase. (which is kept to linux in order to build this package as "linux" for seamless updates from installable liveCDs)
Corrected manually.

falconindy commented on 2011-09-11 01:33

Please use `makepkg --source' when creating tarballs for submission. It's completely wrong that the directory packed inside this tarball does not match the name of the package.

Anonymous comment on 2011-08-18 13:32

Sorry, can you make a version 3.0.2?

burdickjp commented on 2011-08-12 14:16

I'm getting this error:
https://bbs.archlinux.org/viewtopic.php?id=103362

When I use the major/minor of my root partition it comes back with "unable to find /lib/modules/3.0-aufs_friendly/modules.dep"

Calimero commented on 2011-08-08 18:51

Yes, this is adjusted now (and the pkgdesc too).

Anonymous comment on 2011-08-08 18:27

Sorry, the lines are 108, 116, 159, and 290.

Anonymous comment on 2011-08-08 18:24

Is it just me, or does it no longer provides "aufs_friendly"? I noticed that it gets replaced in lines 109, 11, 160, and 291 due to it replacing kernel26 and that it also provides such features for other packages.

Anyway, cannot compile (although easily bypassed) and install aufs2-3.0 as a result.