You are not logged in.

Pages: 1

#1 2014-05-07 21:46:08

jbelanich
Member
Registered: 2014-05-07
Posts: 9

[SOLVED] Kernel version inconsistancies

I recently decided to upgrade my system (pacman -Syu). A few kernel relate things broke, and I am getting some odd behavior. My original kernel version was 3.14.1-1, and I believe the upgrade was supposed to bump me up to 3.14.2-1. Right now, 'pacman -Qi linux' gives me:

Name           : linux
Version        : 3.14.2-1
Description    : The Linux kernel and modules
Architecture   : x86_64
URL            : [url]http://www.kernel.org/[/url]
Licenses       : GPL2
Groups         : base
Provides       : kernel26=3.14.2
Depends On     : coreutils  linux-firmware  kmod  mkinitcpio>=0.7
Optional Deps  : crda: to set the correct wireless channels of your country
Required By    : virtualbox-host-modules
Optional For   : None
Conflicts With : kernel26
Replaces       : kernel26
Installed Size : 71876.00 KiB
Packager       : Tobias Powalowski <tpowa@archlinux.org>
Build Date     : Sun 27 Apr 2014 05:31:54 AM EDT
Install Date   : Wed 07 May 2014 05:02:27 PM EDT
Install Reason : Explicitly installed
Install Script : Yes
Validated By   : Signature

This, I believe, is expected. On the other hand, 'uname -a' gives me the following:

Linux localhost 3.14.1-1-ARCH #1 SMP PREEMPT Mon Apr 14 20:40:47 CEST 2014 x86_64 GNU/Linux

Also, 'cat /proc/version' gives me:

Linux version 3.14.1-1-ARCH (nobody@var-lib-archbuild-testing-x86_64-tobias) (gcc version 4.8.2 20140206 (prerelease) (GCC) ) #1 SMP PREEMPT Mon Apr 14 20:40:47 CEST 2014

I have already rebooted my system after my upgrade (I know if you don't, uname doesn't update). In fact, rebooting is how I learned about this problem. On reboot, none of my standard modules loaded. In /usr/lib/modules I had a 3.14.2-1-ARCH folder with all my modules, so they exist but weren't loading. But, the 3.14.1-1-ARCH folder was gone, which I suppose is expected after the upgrade. I noticed uname was reporting the wrong version, so I copied the /usr/lib/modules/3.14.2-1-ARCH folder to /usr/lib/modules/3.14.1-1-ARCH, and tried again. This time, everything loaded just fine. It seems that the kernel insists on loading from the previous version's folder!

I thought maybe vmlinuz-linux was not properly modified, so I moved it and reinstalled the linux package. Things didn't change. Similarly I tried creating a new initramfs with 'mkinitcpio -p linux', which reported the following:

==> Building image from preset: /etc/mkinitcpio.d/linux.preset: 'default'
  -> -k /boot/vmlinuz-linux -c /etc/mkinitcpio.conf -g /boot/initramfs-linux.img
==> Starting build: 3.14.2-1-ARCH
  -> Running build hook: [base]
  -> Running build hook: [udev]
  -> Running build hook: [autodetect]
  -> Running build hook: [modconf]
  -> Running build hook: [block]
  -> Running build hook: [filesystems]
  -> Running build hook: [keyboard]
  -> Running build hook: [fsck]
==> Generating module dependencies
==> Creating gzip initcpio image: /boot/initramfs-linux.img
==> Image generation successful
==> Building image from preset: /etc/mkinitcpio.d/linux.preset: 'fallback'
  -> -k /boot/vmlinuz-linux -c /etc/mkinitcpio.conf -g /boot/initramfs-linux-fallback.img -S autodetect
==> Starting build: 3.14.2-1-ARCH
  -> Running build hook: [base]
  -> Running build hook: [udev]
  -> Running build hook: [modconf]
  -> Running build hook: [block]
==> WARNING: Possibly missing firmware for module: aic94xx
==> WARNING: Possibly missing firmware for module: smsmdtv
  -> Running build hook: [filesystems]
  -> Running build hook: [keyboard]
  -> Running build hook: [fsck]
==> Generating module dependencies
==> Creating gzip initcpio image: /boot/initramfs-linux-fallback.img
==> Image generation successful

So, mkinitcpio thinks the kernel version is 3.14.2-1-ARCH!

My system is running fine now that I've recreated the previous version's module folder, but I'm concerned about these inconsistancies. Am I really running version 3.14.2-1, or am I running 3.14.1-1? If I'm running 3.14.1-1, why can't I get to 3.14.2-1? I've been poking around trying to figure this out, but I'm stuck.

Anyone know whats going on here and can lend a hand?

Last edited by jbelanich (2014-05-08 01:09:47)

Offline

#2 2014-05-07 21:51:31

SolarBoyMatt
Member
Registered: 2012-01-07
Posts: 263

Re: [SOLVED] Kernel version inconsistancies

Was the partition containing /boot mounted when you did the upgrade?

Offline

#3 2014-05-07 21:53:47

jbelanich
Member
Registered: 2014-05-07
Posts: 9

Re: [SOLVED] Kernel version inconsistancies

Yes, I have everything under one root partition and it was mounted.

Offline

#4 2014-05-07 21:56:48

Xyne
Forum Fellow
Registered: 2008-08-03
Posts: 6,965
Website

Re: [SOLVED] Kernel version inconsistancies

Have you checked the pacman log (/var/log/pacman.log) for errors during the upgrade?

In the absence of a better idea, I would probably switch to a different (reliable) mirror then issue "pacman -Syyuu linux" to force a database refresh and system upgrade against the current repo packages on the new mirror.

My Arch Linux Stuff • Forum Etiquette • Community Ethos - Arch is not for everyone

Offline

#5 2014-05-07 22:37:34

ewaller
Administrator
From: Pasadena, CA
Registered: 2009-07-13
Posts: 20,722

Re: [SOLVED] Kernel version inconsistancies

Stupid question.  Did you reboot?

Nothing is too wonderful to be true, if it be consistent with the laws of nature -- Michael Faraday
The shortest way to ruin a country is to give power to demagogues.— Dionysius of Halicarnassus
---
How to Ask Questions the Smart Way

Offline

#6 2014-05-07 22:40:25

jbelanich
Member
Registered: 2014-05-07
Posts: 9

Re: [SOLVED] Kernel version inconsistancies

Thanks for the reply, Xyne.

Have you checked the pacman log (/var/log/pacman.log) for errors during the upgrade?

I looked through there and I think everything seems fine. I could post the relevant portion if you think it would help.

In the absence of a better idea, I would probably switch to a different (reliable) mirror then issue "pacman -Syyuu linux" to force a database refresh and system upgrade against the current repo packages on the new mirror.

As per your suggestion I just gave that a shot. I commented out all but a few mirrors in /etc/pacman.d/mirrorlist and ran pacman -Syyuu linux. The problem still persists after reboot sad. I tried a variety of different mirrors.

Offline

#7 2014-05-07 22:41:33

jbelanich
Member
Registered: 2014-05-07
Posts: 9

Re: [SOLVED] Kernel version inconsistancies

@ewaller Unfortunately yes, many times.

Offline

#8 2014-05-07 22:53:57

jasonwryan
Anarchist
From: .nz
Registered: 2009-05-09
Posts: 30,426
Website

Re: [SOLVED] Kernel version inconsistancies

We need more details about your setup. EFI or BIOS? What does your partition table look like? Anything that is related to the issue...

Arch + dwm   •   Mercurial repos  •   Surfraw

Registered Linux User #482438

Offline

#9 2014-05-07 23:30:59

jbelanich
Member
Registered: 2014-05-07
Posts: 9

Re: [SOLVED] Kernel version inconsistancies

@jasonwryan

Sure, here it goes. My system is admittedly rather messy. I am on an efi system with two drives. /dev/sda9 is the partition containing archlinux.

My parition table via fdisk -l:

Disk /dev/sda: 931.5 GiB, 1000204886016 bytes, 1953525168 sectors
Units: sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 4096 bytes
I/O size (minimum/optimal): 4096 bytes / 4096 bytes
Disklabel type: gpt
Disk identifier: 04DF5CE4-4265-4565-BC62-8CC734A3A403

Device           Start          End   Size Type
/dev/sda1         2048      2050047  1000M Windows recovery environment
/dev/sda2      2050048      2582527   260M EFI System
/dev/sda3      2582528      4630527  1000M unknown
/dev/sda4      4630528      4892671   128M Microsoft reserved
/dev/sda5      4892672    940036095 445.9G Microsoft basic data
/dev/sda6   1868679168   1869395967   350M Windows recovery environment
/dev/sda7   1869395968   1921824767    25G Microsoft basic data
/dev/sda8   1921824768   1953523711  15.1G Windows recovery environment
/dev/sda9    940036096   1868679167 442.8G Linux filesystem


Disk /dev/sdb: 232.9 GiB, 250059350016 bytes, 488397168 sectors
Units: sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disklabel type: gpt
Disk identifier: 5B234C5B-7883-11E3-BE7C-00C2C61C71AD

Device           Start          End   Size Type
/dev/sdb1         2048      2050047  1000M Windows recovery environment
/dev/sdb2      2050048      2582527   260M EFI System
/dev/sdb3      2582528      2844671   128M Microsoft reserved
/dev/sdb4      2844672    410234151 194.3G Microsoft basic data
/dev/sdb5    410234880    488397134  37.3G Linux filesystem

I am using grub to boot. I had to modify the default grub.cfg to match my setup. Here is the relevant portion of grub.cfg:

menuentry 'Arch Linux' --class arch --class gnu-linux --class gnu --class os $menuentry_id_option 'gnulinux-simple-a5d6ebe1-a4c0-4e55-9a5e-ba76565af48e' {
	load_video
	set gfxpayload=keep
	insmod gzio
	insmod part_gpt 
	insmod ext2
	set root='hd1,gpt9'
	if [ x$feature_platform_search_hint = xy ]; then
	  search --no-floppy --fs-uuid --set=root --hint-bios=hd0,gpt5 --hint-efi=hd0,gpt5 --hint-baremetal=ahci0,gpt5  a5d6ebe1-a4c0-4e55-9a5e-ba76565af48e
	else
	  search --no-floppy --fs-uuid --set=root a5d6ebe1-a4c0-4e55-9a5e-ba76565af48e
	fi
	echo	'Loading Linux linux ...'
	linux	/boot/vmlinuz-linux root=/dev/sda9 ro
	echo	'Loading initial ramdisk ...'
	initrd	/boot/initramfs-linux.img
}

I've only been having problems since the most recent update. I've upgraded many times before with no issues. I've even upgraded the kernel before without any problems.

Please let me know if any other information is necessary.

Last edited by jbelanich (2014-05-07 23:57:28)

Offline

#10 2014-05-07 23:38:45

HalosGhost
Forum Fellow
From: Twin Cities, MN
Registered: 2012-06-22
Posts: 2,097
Website

Re: [SOLVED] Kernel version inconsistancies

Please, please, please, please don't use quote tags for posting terminal output or code snippets. Please use code tags:

for sake in `all that is holy`; do
   use code tags
done

All the best,

-HG

Code of Conduct

Offline

#11 2014-05-07 23:49:04

jbelanich
Member
Registered: 2014-05-07
Posts: 9

Re: [SOLVED] Kernel version inconsistancies

@HalosGhost

Sorry! For readability I've copied quoted text below.

fdisk -l:

    Disk /dev/sda: 931.5 GiB, 1000204886016 bytes, 1953525168 sectors
    Units: sectors of 1 * 512 = 512 bytes
    Sector size (logical/physical): 512 bytes / 4096 bytes
    I/O size (minimum/optimal): 4096 bytes / 4096 bytes
    Disklabel type: gpt
    Disk identifier: 04DF5CE4-4265-4565-BC62-8CC734A3A403

    Device           Start          End   Size Type
    /dev/sda1         2048      2050047  1000M Windows recovery environment
    /dev/sda2      2050048      2582527   260M EFI System
    /dev/sda3      2582528      4630527  1000M unknown
    /dev/sda4      4630528      4892671   128M Microsoft reserved
    /dev/sda5      4892672    940036095 445.9G Microsoft basic data
    /dev/sda6   1868679168   1869395967   350M Windows recovery environment
    /dev/sda7   1869395968   1921824767    25G Microsoft basic data
    /dev/sda8   1921824768   1953523711  15.1G Windows recovery environment
    /dev/sda9    940036096   1868679167 442.8G Linux filesystem

    Disk /dev/sdb: 232.9 GiB, 250059350016 bytes, 488397168 sectors
    Units: sectors of 1 * 512 = 512 bytes
    Sector size (logical/physical): 512 bytes / 512 bytes
    I/O size (minimum/optimal): 512 bytes / 512 bytes
    Disklabel type: gpt
    Disk identifier: 5B234C5B-7883-11E3-BE7C-00C2C61C71AD

    Device           Start          End   Size Type
    /dev/sdb1         2048      2050047  1000M Windows recovery environment
    /dev/sdb2      2050048      2582527   260M EFI System
    /dev/sdb3      2582528      2844671   128M Microsoft reserved
    /dev/sdb4      2844672    410234151 194.3G Microsoft basic data
    /dev/sdb5    410234880    488397134  37.3G Linux filesystem

grug.cfg:

    menuentry 'Arch Linux' --class arch --class gnu-linux --class gnu --class os $menuentry_id_option 'gnulinux-simple-a5d6ebe1-a4c0-4e55-9a5e-ba76565af48e' {
        load_video
        set gfxpayload=keep
        insmod gzio
        insmod part_gpt
        insmod ext2
        set root='hd1,gpt9'
        if [ x$feature_platform_search_hint = xy ]; then
          search --no-floppy --fs-uuid --set=root --hint-bios=hd0,gpt5 --hint-efi=hd0,gpt5 --hint-baremetal=ahci0,gpt5  a5d6ebe1-a4c0-4e55-9a5e-ba76565af48e
        else
          search --no-floppy --fs-uuid --set=root a5d6ebe1-a4c0-4e55-9a5e-ba76565af48e
        fi
        echo    'Loading Linux linux ...'
        linux    /boot/vmlinuz-linux root=/dev/sda9 ro
        echo    'Loading initial ramdisk ...'
        initrd    /boot/initramfs-linux.img
    }

pacman -Qi linux:

Name           : linux
Version        : 3.14.2-1
Description    : The Linux kernel and modules
Architecture   : x86_64
URL            : http://www.kernel.org/
Licenses       : GPL2
Groups         : base
Provides       : kernel26=3.14.2
Depends On     : coreutils  linux-firmware  kmod  mkinitcpio>=0.7
Optional Deps  : crda: to set the correct wireless channels of your country
Required By    : virtualbox-host-modules
Optional For   : None
Conflicts With : kernel26
Replaces       : kernel26
Installed Size : 71876.00 KiB
Packager       : Tobias Powalowski <tpowa@archlinux.org>
Build Date     : Sun 27 Apr 2014 05:31:54 AM EDT
Install Date   : Wed 07 May 2014 05:02:27 PM EDT
Install Reason : Explicitly installed
Install Script : Yes
Validated By   : Signature

uname -a:

Linux localhost 3.14.1-1-ARCH #1 SMP PREEMPT Mon Apr 14 20:40:47 CEST 2014 x86_64 GNU/Linux

cat /proc/version:

Linux version 3.14.1-1-ARCH (nobody@var-lib-archbuild-testing-x86_64-tobias) (gcc version 4.8.2 20140206 (prerelease) (GCC) ) #1 SMP PREEMPT Mon Apr 14 20:40:47 CEST 2014

mkinitcpio -p linux:

==> Building image from preset: /etc/mkinitcpio.d/linux.preset: 'default'
  -> -k /boot/vmlinuz-linux -c /etc/mkinitcpio.conf -g /boot/initramfs-linux.img
==> Starting build: 3.14.2-1-ARCH
  -> Running build hook: [base]
  -> Running build hook: [udev]
  -> Running build hook: [autodetect]
  -> Running build hook: [modconf]
  -> Running build hook: [block]
  -> Running build hook: [filesystems]
  -> Running build hook: [keyboard]
  -> Running build hook: [fsck]
==> Generating module dependencies
==> Creating gzip initcpio image: /boot/initramfs-linux.img
==> Image generation successful
==> Building image from preset: /etc/mkinitcpio.d/linux.preset: 'fallback'
  -> -k /boot/vmlinuz-linux -c /etc/mkinitcpio.conf -g /boot/initramfs-linux-fallback.img -S autodetect
==> Starting build: 3.14.2-1-ARCH
  -> Running build hook: [base]
  -> Running build hook: [udev]
  -> Running build hook: [modconf]
  -> Running build hook: [block]
==> WARNING: Possibly missing firmware for module: aic94xx
==> WARNING: Possibly missing firmware for module: smsmdtv
  -> Running build hook: [filesystems]
  -> Running build hook: [keyboard]
  -> Running build hook: [fsck]
==> Generating module dependencies
==> Creating gzip initcpio image: /boot/initramfs-linux-fallback.img
==> Image generation successful

.

Offline

#12 2014-05-08 00:02:52

ramkatral
Member
Registered: 2014-04-21
Posts: 29

Re: [SOLVED] Kernel version inconsistancies

Just a wild question... Did you verify vmlinuz is properly linked to the new kernel and not to the old?  Stranger things have happened.

Offline

#13 2014-05-08 00:21:52

jbelanich
Member
Registered: 2014-05-07
Posts: 9

Re: [SOLVED] Kernel version inconsistancies

@ramkatral

I'm not quite sure what you mean. Is vmlinuz-linux linked to something? When I run 'file vmlinuz-linux' I get the following:

vmlinuz-linux: Linux kernel x86 boot executable bzImage, version 3.14.2-1-ARCH (nobody@var-lib-archbuild-testing-x86_64-tobias) , RO-rootFS, swap_dev 0x3, Normal VGA

Which seems to me to mean that it is the correct version.

Offline

#14 2014-05-08 00:28:02

jbelanich
Member
Registered: 2014-05-07
Posts: 9

Re: [SOLVED] Kernel version inconsistancies

I've tried booting my kernel using qemu with 'qemu-system-x86_64 -m 1024 -smp 4 -kernel vmlinuz-linux -initrd initramfs-linux.img'. The kernel, of course, can't load all the modules properly but drops into a rescue shell where I run 'uname -a' and get the following:

Linux (none) 3.14.2-1-ARCH #1 SMP PREEMPT Sun Apr 27 11:28:44 CEST 2014 x86_64 GNU/Linux

Which suggests the kernel knows better in qemu! It seems something must be going wrong when my computer boots, because a virtual boot seems to be working properly...

Offline

#15 2014-05-08 00:50:21

HalosGhost
Forum Fellow
From: Twin Cities, MN
Registered: 2012-06-22
Posts: 2,097
Website

Re: [SOLVED] Kernel version inconsistancies

jbelanich wrote:

Sorry! For readability I've copied quoted text below.

I forgive you. Also, use the edit button; avoid double-posting smile

All the best,

-HG

Code of Conduct

Offline

#16 2014-05-08 00:57:40

ramkatral
Member
Registered: 2014-04-21
Posts: 29

Re: [SOLVED] Kernel version inconsistancies

Eh, i use vmlinuz as a link to my actual kernel file once it's compiled.  After I make a new kernel, I update the link.  I compile my own kernel, so perhaps pacman doesn't use the link method.

Either way, yea, it seems your vmlinuz file is the correct version.

Last edited by ramkatral (2014-05-08 01:02:16)

Offline

#17 2014-05-08 01:08:46

jbelanich
Member
Registered: 2014-05-07
Posts: 9

Re: [SOLVED] Kernel version inconsistancies

@ramkatral Ah that seems really convenient when dealing with many kernels. I'll remember that strategy when I start compiling my own!
@HalosGhost Ah awesome. You are full of good advice smile.

I fixed the problem! I switched to using rEFInd for booting and now everything is loading as the proper version. I really don't understand how this fixed it, but I'm happy its working now. It seems grub must have been booting the wrong kernel, but I don't see how that is possible because it seems to me that /boot/vmlinuz-linux is the only kernel.

Thanks everyone for your help!

Last edited by jbelanich (2014-05-08 01:12:13)

Offline

Pages: 1

Board footer

Powered by FluxBB