[arch-general] Install problem
Hello, today I did a fresh install using the december release of the install media. Everything done 'to the book' (the install guide wiki page). Bootloader is syslinux. When booting the freshly installed system, I get the syslinux menu, select Archlinux or fallback, things seem to be normal for a few seconds then everything stops. After some time I get a message that /dev/sda3 (which is the / partition) can't be found and I'm dropped into an emergency shell. No LVM or RAID, just a single SATA disk with sda1 = /boot (100M, ext2) sda2 = swap (2G) sda3 = / (~250G, ext4) Partioning done with cfdisk. The mkinitcpio.conf used is the default one, which seems to have all required hooks. Everything in /boot/syslinux looks ok. Any hints ?? Ciao, -- FA A world of exhaustive, reliable metadata would be an utopia. It's also a pipe-dream, founded on self-delusion, nerd hubris and hysterically inflated market opportunities. (Cory Doctorow)
On Thu, Dec 27, 2012 at 8:45 PM, Fons Adriaensen <fons@linuxaudio.org>wrote:
Hello,
today I did a fresh install using the december release of the install media. Everything done 'to the book' (the install guide wiki page). Bootloader is syslinux.
When booting the freshly installed system, I get the syslinux menu, select Archlinux or fallback, things seem to be normal for a few seconds then everything stops. After some time I get a message that /dev/sda3 (which is the / partition) can't be found and I'm dropped into an emergency shell.
No LVM or RAID, just a single SATA disk with
sda1 = /boot (100M, ext2) sda2 = swap (2G) sda3 = / (~250G, ext4)
Partioning done with cfdisk.
The mkinitcpio.conf used is the default one, which seems to have all required hooks.
Everything in /boot/syslinux looks ok.
Any hints ??
Do you see the partitions from the emergency shell, doing a `ls /dev/sd*`? Can you paste the contents of the syslinux.cfg file? -- Rodrigo
On Fri, Dec 28, 2012 at 11:38:15AM +0100, Rodrigo Rivas wrote:
Do you see the partitions from the emergency shell, doing a `ls /dev/sd*`?
There are no /dev/sd* Meanwhile I've noticed something new. Once in the about ten times I tried booting, it actually worked (and systemd seems to be incredibly fast). In the other cases, there's an oops: Booting the kernel. :: running early hook [udev] :: running hook [udev] :: Triggering uevents... [ 0.870155] BUG: unable to handle kernel NULL pointer dereference at 00000010 [ .... ] IP [<f81f1240>] pacpi_set_dmamode+0x50/0xa0 [pata_acpi] [ .... ] *pde = 00000000 etc. AND: The same thing can happen when booting from the install media, but it seems less likely (one time in three or four). Looks like some race condition. I scanned the BIOS for any relevant options, but the only thing I found was an option to enable/disable 'legacy SATA0 support'. It doesn't seem to make any difference.
Can you paste the contents of the syslinux.cfg file?
I could if you really need it, but it's the unmodified one as installed. I checked it as suggested by the install guide, but didn't find anything to change. Ciao, -- FA A world of exhaustive, reliable metadata would be an utopia. It's also a pipe-dream, founded on self-delusion, nerd hubris and hysterically inflated market opportunities. (Cory Doctorow)
Hi
[ .... ] IP [<f81f1240>] pacpi_set_dmamode+0x50/0xa0 [pata_acpi]
I've blacklisted pata_acpi because of random boot failures. Maybe it's your case. ciao Dario
On Sun, Dec 30, 2012 at 9:22 PM, Dario <carotinobg@yahoo.it> wrote:
Hi I've blacklisted pata_acpi because of random boot failures. Maybe it's your case.
Wasn't that hook replaced a week or so ago with the 'block' flag?
On 12/31/12 at 05:39am, Martín Cigorraga wrote:
On Sun, Dec 30, 2012 at 9:22 PM, Dario <carotinobg@yahoo.it> wrote:
Hi I've blacklisted pata_acpi because of random boot failures. Maybe it's your case.
Wasn't that hook replaced a week or so ago with the 'block' flag?
I think that Diaro is referring to the actual pata_acpi module. As the mkinitcpio hook was simply "pata". -- Curtis Shimamoto sugar.and.scruffy@gmail.com
On Mon, Dec 31, 2012 at 01:22:24AM +0100, Dario wrote:
[FA]
[ .... ] IP [<f81f1240>] pacpi_set_dmamode+0x50/0xa0 [pata_acpi]
I've blacklisted pata_acpi because of random boot failures. Maybe it's your case.
It was. Blacklisted pata_acpi, mkinitcpio, and the system boots. So now I have my first pure systemd machine running (and to be clear, systemd has nothing to do with this bug). Ten more to follow if all goes well... A few notes/questions. * I forgot to run mkswap during the initial installation. The 'Partioning' and 'File systems' links from the install guide wiki page don't mention mkswap - maybe it should be mentioned somewhere (it is if you go deeper, e.g. on the 'Swap file' page). Strange thing is that cfdisk showed the unitialised partition as ext3, as did blkid, while fdisk showed it as swap (82). * Installed xdm-archlinux and enabled the corresponding service. This gives me a graphical login on tty2 instead of the usual tty7. Is this as intended ? * What is the purpose of halt.target ? It stops the system but leaves power on. AFAICS, it's not possible to restart without a power cycle... * This machine *never* worked with nouveau, not even recent versions. But now by magic it does ! Ciao, -- FA A world of exhaustive, reliable metadata would be an utopia. It's also a pipe-dream, founded on self-delusion, nerd hubris and hysterically inflated market opportunities. (Cory Doctorow)
On Thu, Jan 3, 2013 at 12:28 AM, Fons Adriaensen <fons@linuxaudio.org> wrote:
* What is the purpose of halt.target ? It stops the system but leaves power on. AFAICS, it's not possible to restart without a power cycle...
Under sysvinit halt and poweroff were the same. Under systemd 'halt' halts the machine, whereas 'poweroff' first halts the machine and then switches off the power. I use 'halt' a lot to debug shutdown bugs, don't know if it has other uses. Manually power cycling a halted machine is safe. -t
On 01/02/2013 10:59 PM, Tom Gundersen wrote:
On Thu, Jan 3, 2013 at 12:28 AM, Fons Adriaensen <fons@linuxaudio.org> wrote:
* What is the purpose of halt.target ? It stops the system but leaves power on. AFAICS, it's not possible to restart without a power cycle... Under sysvinit halt and poweroff were the same. Under systemd 'halt' halts the machine, whereas 'poweroff' first halts the machine and then switches off the power.
I use 'halt' a lot to debug shutdown bugs, don't know if it has other uses.
Manually power cycling a halted machine is safe.
-t I can confirm what Tom is saying is correct, on my old slackware install halt was the same as shutting down. On my arch install that uses systemd halt is like giving your computer a lobotomy. Its still on but nothing going on.
Personally I suspend my computer all the time unless I need to reboot it for an update or need to shut it down because of a thunderstorm. Good question though. -Matt
On Thu, Dec 27, 2012 at 12:45 PM, Fons Adriaensen <fons@linuxaudio.org>wrote:
When booting the freshly installed system, I get the syslinux menu, select Archlinux or fallback, things seem to be normal for a few seconds then everything stops. After some time I get a message that /dev/sda3 (which is the / partition) can't be found and I'm dropped into an emergency shell.
Everything in /boot/syslinux looks ok.
Any hints ??
Ciao,
-- FA
Had a very similar issue to that with my rig. Fairly vanilaa x2 Athlon system with 500 gig sata drive. A power cycle resolved it. A subsequent change...adding a new sata drive...on first post install boot caused this to happen again. Power cycling the rig fixed the problem again. C
On Thu, Dec 27, 2012 at 07:45:19PM +0000, Fons Adriaensen wrote:
Hello,
today I did a fresh install using the december release of the install media. Everything done 'to the book' (the install guide wiki page). Bootloader is syslinux.
When booting the freshly installed system, I get the syslinux menu, select Archlinux or fallback, things seem to be normal for a few seconds then everything stops. After some time I get a message that /dev/sda3 (which is the / partition) can't be found and I'm dropped into an emergency shell.
Use a persistent identifier such as a LABEL or UUID tag, rather than an unreliable kernel name. Not convinced this is the problem, but it's better to let /init figure out what device this really corresponds to in the kernel.
No LVM or RAID, just a single SATA disk with
sda1 = /boot (100M, ext2) sda2 = swap (2G) sda3 = / (~250G, ext4)
Assuming this is native SATA and not setup in compat mode, your image needs to contain the modules 'ahci', 'sd_mod', and 'ext4' (ignoring dependencies which I assume mkinitcpio found, added). Make sure the kernel version for the image matches whatever the installer gave you (3.6.10-1-ARCH if you stuck with [core]). 'lsinitcpio -a' will nicely lay all this out for you. You'll probably find your problem/solution here which will involve rebuilding the initramfs. Make sure /boot is mounted...
Partioning done with cfdisk.
Unrelated, I suggest you don't use cfdisk until util-linux 2.23. cfdisk still expects disk geometry to be measurable in cylinders, heads and sectors -- an idea that's been obsolete for nearly 2 decades. fdisk and parted are both better choices if MBR is sufficient. d
On Fri, Dec 28, 2012 at 10:07:56AM -0500, Dave Reisner wrote:
Assuming this is native SATA and not setup in compat mode, your image needs to contain the modules 'ahci', 'sd_mod', and 'ext4' (ignoring dependencies which I assume mkinitcpio found, added). Make sure the kernel version for the image matches whatever the installer gave you (3.6.10-1-ARCH if you stuck with [core]). 'lsinitcpio -a' will nicely lay all this out for you.
You'll probably find your problem/solution here which will involve rebuilding the initramfs. Make sure /boot is mounted...
Module ahci was missing. Added it to MODULES, remade the images, the problem remains. See also previous post: the same thing happens (but less frequently) when booting from the install media (USB stick). I must just have been lucky when I did the install. When there is the oops, none of the three modules you mentioned are in the list printed by the backtrace. Further action will have to wait until next year (this is at work). Until then, best wishes ! Ciao, -- FA A world of exhaustive, reliable metadata would be an utopia. It's also a pipe-dream, founded on self-delusion, nerd hubris and hysterically inflated market opportunities. (Cory Doctorow)
participants (9)
-
Casey Peter
-
Curtis Shimamoto
-
Dario
-
Dave Reisner
-
Fons Adriaensen
-
Martín Cigorraga
-
Matthew Dey
-
Rodrigo Rivas
-
Tom Gundersen