[Bradford] Nearly working laptop

Devo Too mike_g at devotoo.org.uk
Thu Mar 28 13:45:41 UTC 2019 

Hi Folks,

For anybody still interested in this sad and sorry extended tale, 
SOLVED!!! The laptop works.

In spite of having only the base system installed so far, so no need for 
graphics or anything silly, the "nouveau" GPU manager is now part of the 
current kernel. It clashes with the Nvidia driver. The laptop has an 
Intel onboard GPU and an added Nvidia GeForce pci GPU.

The solution, which took a lot of digging to find, is to blacklist 
nouveau: https://bbs.archlinux.org/viewtopic.php?id=213042

Thanks to John and Dunc for solving other problems on the way.

Mike


On 26/03/2019 17:37, Devo Too via Bradford wrote:
> More on this kernel panic thing - I had the wired connection drop a 
> couple of times and thought I'd check which driver with lspci -v and the 
> screen hung for a short while, then another kernel panic. The screens 
> ran past too quickly to read but the last line when it hung again was "A 
> stop job is running for user root (1min30secs)" and when that time was 
> up it went to 3 mins and so on until I hit the power button.
> 
> I reckon there's a commonality of panic causes. But I don't know how to 
> track it down. I can't find the system logs and dmesg isn't helping. The 
> only way I can find out is to switch off, and dmesg, it seems, only says 
> what's going on in the current session.
> 
> The instrallation is completely new, so there are no clues by way of 
> changes in behaviour or performance.
> 
> On 26/03/2019 15:02, Devo Too via Bradford wrote:
>> Thanks, Dunc. Well, it's a little bit further on. I'll take a look in 
>> the kernel log and see if dmesg-E sheds any further light, too.
>>
>> On 26/03/2019 14:58, Duncan Hughes wrote:
>>> I’d start watching the kernel log to see what the panic is next. 
>>> Unfortunately, you’re on a laptop and not one of the rackmounts I’m 
>>> used to so a serial console is going to be out of the question.  You 
>>> may have some luck with dmesg -E and maybe turning the log level up 
>>> but I’m going to have to hand the baton to someone that knows laptops.
>>>
>>>
>>>> On 26 Mar 2019, at 14:53, Devo Too via Bradford 
>>>> <bradford at mailman.lug.org.uk> wrote:
>>>>
>>>> Left a bit of time before hitting the power switch on the last 
>>>> shutdown and the kernel is still panicking.
>>>>
>>>> The UUIDs in fstab all match those of the blkid output now.
>>>>
>>>> So we have a partial solution. Where to next?
>>>>
>>>> Also, it has dropped its connection and is failing to pick up an ip 
>>>> through dhcp.
>>>>
>>>> On 26/03/2019 14:36, Devo Too via Bradford wrote:
>>>>> I thought I'd already checked all that. But you may have just hit 
>>>>> the nail on the head. sda3 (swap) shows a different UUID for each 
>>>>> entry and blkid has just told me which to delete (I'd commented the 
>>>>> wrong one out!).
>>>>> So tried with the correct (wrong one) entry commented out, 
>>>>> uncommented the accurate entry and it still hung but starts on 
>>>>> steroids, no questions about the home partition.
>>>>> Deleted all the dual entries and same thing, starts on steroids but 
>>>>> hangs on reboot or shutdown.
>>>>> I can't paste anything yet. No software beyond system tools 
>>>>> installed. I'm having to cross the room to the router to see the 
>>>>> laptop as the box (this one) is on a long cat5 and pretty well 
>>>>> fixed in place but no other long cat5 or 6 to put them closer 
>>>>> together.
>>>>> fstab and blkid will probably come up on the same screen though, so 
>>>>> I can double check now and correct any faults manually.
>>>>> On 26/03/2019 14:08, Duncan Hughes via Bradford wrote:
>>>>>> I would just delete all the duplicates from fstab, take a backup 
>>>>>> of the whole file if you’re worried about losing anything.  Then 
>>>>>> double check the UUIDs against what you actually have on the disk 
>>>>>> (blkid is your friend).  Can you paste your fstab and the output 
>>>>>> of blkid?
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>> On 26 Mar 2019, at 14:05, Devo Too via Bradford 
>>>>>>> <bradford at mailman.lug.org.uk> wrote:
>>>>>>>
>>>>>>> Hi Dunc,
>>>>>>>
>>>>>>> I think only once, but as entries were not sticking, e.g. running 
>>>>>>> mk and mount had to be double checked to ensure the command had 
>>>>>>> been carried out, I may have done. Certainly I ran genfstab at 
>>>>>>> least once when it failed to generate the file whislt still in 
>>>>>>> archiso, pre-chrooting to the system.
>>>>>>>
>>>>>>> Yes, new Arch installation on a brand new laptop. This midi tower 
>>>>>>> also runs on Arch but this installation is five to six years old.
>>>>>>>
>>>>>>> Question is whether to delete, or at least comment out, the extra 
>>>>>>> entries. I can't see that they would be doing anything 
>>>>>>> constructive and they may be the root of some of the problems?
>>>>>>>
>>>>>>> Cheers,
>>>>>>>
>>>>>>> Mike
>>>>>>>
>>>>>>> On 26/03/2019 13:44, Duncan Hughes via Bradford wrote:
>>>>>>>> Hi Mike,
>>>>>>>> Is this your Arch box?  Did you run genfstab multiple times?
>>>>>>>> Dunc.
>>>>>>>>> On 26 Mar 2019, at 13:41, Devo Too via Bradford 
>>>>>>>>> <bradford at mailman.lug.org.uk> wrote:
>>>>>>>>>
>>>>>>>>> Hi Folks,
>>>>>>>>>
>>>>>>>>> With thanks to John Hudson and the author of this blog 
>>>>>>>>> http://averagelinuxuser.com/a-step-by-step-arch-linux-installation-guide/ 
>>>>>>>>> I now have a laptop which kind of works, a huge improvement 
>>>>>>>>> upon where I was on Friday.
>>>>>>>>>
>>>>>>>>> It stutters to boot and will only shut down when its power is 
>>>>>>>>> shut off. On boot, it has problems with a UUID which it targets 
>>>>>>>>> 130 seconds to commit and moves on when it fails, so a slow 
>>>>>>>>> boot. It turns out that UUID is for the fourth and final 
>>>>>>>>> partition, /dev/sda4 /home.
>>>>>>>>>
>>>>>>>>> It goes into kernel panic on reboot or shutdown, so the power 
>>>>>>>>> switch comes into play.
>>>>>>>>>
>>>>>>>>> /etc/fstab has multiple identical entries for three of the four 
>>>>>>>>> partitions. In turn:
>>>>>>>>>
>>>>>>>>> /dev/sda1 /boot/efi 1 entry
>>>>>>>>> /dev/sda2    /   3 entries
>>>>>>>>> /dev/sda3 swap (none) 2 entries
>>>>>>>>> /dev/sda4 /mnt/home 2 entries.
>>>>>>>>>
>>>>>>>>> the order they are listed in is / / efi home swap / home swap
>>>>>>>>>
>>>>>>>>> The layout is
>>>>>>>>> # /dev/sdaX
>>>>>>>>> UUID ~blahblah~
>>>>>>>>> (blank line)
>>>>>>>>> # /dev etcetera
>>>>>>>>>
>>>>>>>>> and the kernel panic happened at each shutdown or reboot. So I 
>>>>>>>>> added a single # to each uncommented repeat and hit reboot. A 
>>>>>>>>> blank screen appeared with a cursor flickering, top left hand 
>>>>>>>>> corner. Power button, same cure as before.
>>>>>>>>>
>>>>>>>>> Booting up again, the system is still looking for /home 
>>>>>>>>> partition. I haven't added any ordinary user yet and there is 
>>>>>>>>> no data in there. Would adding a user with ~/username/ help?
>>>>>>>>>
>>>>>>>>> Any sysadmins or gifted amateurs out there to point me in the 
>>>>>>>>> right direction, please? Is fstab the right place to look? If 
>>>>>>>>> not, what else should I be doing?
>>>>>>>>>
>>>>>>>>> TIA,
>>>>>>>>>
>>>>>>>>> Mike
>>>>>>>>>
>>>>>>>>> -- 
>>>>>>>>> Bradford mailing list
>>>>>>>>> Bradford at mailman.lug.org.uk
>>>>>>>>> https://mailman.lug.org.uk/mailman/listinfo/bradford
>>>>>>>
>>>>>>> -- 
>>>>>>> Bradford mailing list
>>>>>>> Bradford at mailman.lug.org.uk
>>>>>>> https://mailman.lug.org.uk/mailman/listinfo/bradford
>>>>>>
>>>>>>
>>>>
>>>> -- 
>>>> Bradford mailing list
>>>> Bradford at mailman.lug.org.uk
>>>> https://mailman.lug.org.uk/mailman/listinfo/bradford
>>>
>>
>

More information about the Bradford mailing list