Linux Lite Forums - NUC Install Issues

Pages: 1 2 3

Found this post saying problem likely caused by UEFI/(Bios) settings -- possibly solved by updating the firmware, if an update is available. Sounds like these types of systems are meant to boot in UEFI mode and the firmware is not set-up to properly handle Legacy/Bios mode booting. As you can read in rest of that thread, I had the OP run through various things to try getting it to work all to no avail.

Unfortunately, (assuming you don't find another solution), you may end up having to re-install everything (Windows and LL) in UEFI mode to get it working right.

P.s. Just for heck of it, run sudo blkid from terminal in live LL to see if UUID of LL on nvme disk matches that shown in your grub.conf output. If not, try editing it and see if that works. Other than that I'm afraid I have no idea how to get around the problem other than doing a UEFI-mode install.

Hi,

Thanks for all the help,
Yes I had seen that thread(mmm... I recognise that username Smile

) 1 of hundreds of posts I have now looked at.
- The BIOS was the first thing I updated, even before installing any OS's, I read that there was a few issues with older versions of BIOS not recognising the nvme pcie disk.
- TTBOMK I have everything set for "Legacy", and per the BIOS picture the nvme pcie disk shows, and it boots Windows O.K

Here is the sudo blkid:

Code:
it@it:~$ sudo blkid

/dev/loop0: TYPE="squashfs"

/dev/nvme0n1p1: LABEL="OpSys" UUID="8050EF0F50EF0B2C" TYPE="ntfs" 

/dev/nvme0n1p2: LABEL="rootlm" UUID="5459ab50-69f9-4ecb-88ba-3bf56fe65a55" TYPE="ext4" 

/dev/nvme0n1p3: UUID="89ddbfbb-26a0-4476-ba01-736c7aee234b" TYPE="ext4"                             # The LL install partition 

/dev/nvme0n1p5: UUID="c9ee7c1a-581a-4103-9ba4-2984c9051cc1" TYPE="swap"

/dev/sda1: LABEL="Data" UUID="545937484610D8B1" TYPE="ntfs" 

/dev/sda2: LABEL="homelm" UUID="a5a28014-affb-44db-93b1-293b34481633" TYPE="ext4" 

/dev/sda3: LABEL="homell" UUID="9188d164-5929-4ae0-8413-d39e7de47e1f" TYPE="ext4" 

/dev/sda5: UUID="1B3FF3B32DE36FDE" TYPE="ntfs" 

/dev/sdb1: LABEL="MULTIBOOT" UUID="1293-FD4C" TYPE="vfat" 

it@it:~$

So if I understand correctly the grub.conf is pointing by UUID to the partition /dev/nvme0n1p3
The thread firenice03 and myself found seem to imply that pointing directly to the root boot device,
where the grub is /dev/nvme0n1 instead of the UUID for the LL partition is, may solve it..??

I guess given it's not working, do I have much to loose.?? ~ "NOT sure"
Change the root=UUID=89ddbfbb-26a0-4476-ba01-736c7aee234b to: root=/dev/nvme0n1 or should it be [color=rgb(255, 0, 0)]/dev/nvme0n1p3[/color]
From your previous post, I would open the grub.conf as root, edit the file.

Additionally I found there are 2 variants of this M2 ssd,This one, a pcie and also a plain ahci, but with reduced performance.(maybe for legacy scenarios.?)Maybe my friends is going to have to do return/rma and get a ahci version.??Not sure if he was aware of this,or even understood the difference, and also if it would make a difference.

??Thanks again.. much appreciated.

Just an update, still no luck with any of the suggestions.
I have just found this statement for the SM951:

Quote:The UEFI drivers for this M.2 PCIE SSD are built onto it and must be read/loaded into your UEFI BIOS first

I have a couple ideas, so one last try.
Then I think I will have to tell my friend to swap/rma the disk for a regular SATA drive in M.2 size/format.

@Jerry

Having exhausted all my options, a final google led me to this: NVM Express Disk Driver
It's in a Ubuntu 14.04 Man page, but seems to be linked to BSD.
Not sure if it's valid, or usable..??

I have absolutely no idea how to compile a kernel ???

But there is also a "load the driver as module at boot time",, however,
I don't seem to be able to find the loader.conf file , should I have one..??
or is there another way to do this with LL..??

any advice greatly received Smile

(11-22-2015, 03:00 PM)Wirezfree link Wrote: [ -> ][size=1em]So if I understand correctly the grub.conf is pointing by UUID to the partition /dev/nvme0n1p3[/size]
[size=1em]The thread firenice03 and myself found seem to imply that pointing directly to the root boot device,[/size]
[size=1em]where the grub is /dev/nvme0n1 instead of the UUID for the LL partition is, may solve it..??[/size]

I doubt that will help. That's the MBR of the disk, so just like pointing to /dev/sda if it were regular disk. On boot-up (in Legacy mode on MBR disks), the MBR of drive only holds first stage boot loader which then points to rest of boot code on root partition (in this case) to finish booting process. The boot code on root partition is what needs to be specified in grub.conf.

(11-23-2015, 12:45 PM)Wirezfree link Wrote: [ -> ][size=1em]Having exhausted all my options, a final google led me to this: NVM Express Disk Driver
It's in a Ubuntu 14.04 Man page, but seems to be linked to BSD.[/size]
[size=1em]Not sure if it's valid, or usable..??[/size]
[size=1em]
I have absolutely no idea how to compile a kernel ??? [/size]

[size=1em]But there is also a "load the driver as module at boot time",, however,[/size]
[size=1em]I don't seem to be able to find the loader.conf file , should I have one..??[/size]
[size=1em]or is there another way to do this with LL..??[/size]

[size=1em]any advice greatly received [/size]

I'm not sure if this will work, but might as well give it a try.

Boot with live LL dvd/usb and open a terminal.

Mount the root partition on installed disk to live environment

Code:
sudo mount /dev/nvme0n1p3 /mnt

for i in /dev /dev/pts /proc /sys; do sudo mount -B $i /mnt$i; done

Chroot into installed system

Code:
sudo chroot /mnt

Open grub file with nano. (When chrooted into system you're operating as root, so no need to precede commands with "sudo". Just enter commands as written below.)

Code:
nano /etc/default/grub

Use arrow keys to navigate to this line

Code:
GRUB_CMDLINE_LINUX_DEFAULT="quiet splash"

Change it to read as

Code:
GRUB_CMDLINE_LINUX_DEFAULT="quiet splash nvme_load=YES nvd_load=YES"

Hit Ctrl+O to "Write Out"/(Save changes); then Ctrl+X to exit nano and get back to terminal prompt.

Now update grub

Code:
update-grub

Exit out of chroot

Code:
exit

Unmount everything

Code:
for i in /sys /proc /dev/pts /dev; do sudo umount /mnt$i; done

sudo umount /mnt

Reboot (without live dvd/usb), cross your fingers and hopefully that worked.

@gold_finger

Thanks for all your help, unfortunately still the same issue. Sad

I had found the following "Intel nvme page" , and followed the check points, apart from point 4 , it all seemed to check out..??
I did this before your suggestions.
I think the clue maybe in the fact that whilst everything "appears" to checkout..??
If you look in "DISKS", the nvme drive is "not" in the disk section, it appears under "Other devices", see screenshot in an earlier post I dis.

Results from doing the checks on Intel Page:

1.)

Code:
03:00.0 Non-Volatile memory controller: Samsung Electronics Co Ltd Device a802 (rev 01)

it@it:~$ lsblk

NAME        MAJ:MIN RM   SIZE RO TYPE MOUNTPOINT

sda           8:0    0 931.5G  0 disk 

├─sda1        8:1    0    50G  0 part 

├─sda2        8:2    0    20G  0 part 

├─sda3        8:3    0    20G  0 part 

├─sda4        8:4    0     1K  0 part 

└─sda5        8:5    0   820G  0 part 

sdb               8:16   1   7.5G  0 disk 

└─sdb1        8:17   1   7.5G  0 part /cdrom

loop0            7:0    0 747.6M  1 loop /rofs

nvme0n1     259:0    0 119.2G  0 disk 

├─nvme0n1p1 259:1    0    50G  0 part 

├─nvme0n1p2 259:2    0    20G  0 part 

├─nvme0n1p3 259:3    0    20G  0 part 

├─nvme0n1p4 259:4    0       1K  0 part 

└─nvme0n1p5 259:5    0    17G  0 part 

it@it:~$

2.)

Code:

it@it:~$ modinfo nvme

filename:         /lib/modules/3.13.0-62-generic/kernel/drivers/block/nvme.ko

version:           0.8

license:            GPL

author:            Matthew Wilcox <[email protected]>

srcversion:     56D1A65BFAFAC90F460C031

alias:                 pci:v*d*sv*sd*bc01sc08i02*

depends:        

intree:               Y

vermagic:         3.13.0-62-generic SMP mod_unload modversions 

signer:               Magrathea: Glacier signing key

sig_key:              E0:7C:FC:BE:E8:34:59:BA:45:54:3D:39:B2:C2:A5:78:75:74:50:01

sig_hashalgo:  sha512

parm:                 shutdown_timeout:timeout in seconds for controller shutdown (byte)

parm:                 nvme_major:int

parm:                 use_threaded_interrupts:int

it@it:~$

3.)

Code:
it@it:~$  lsmod | grep nvm

nvme                   58970  0 

it@it:~$

4.)

Code:
$ ll /dev/nvme*n1

>> command not found

5.)

Code:
it@it:~$ sudo hdparm -tT --direct /dev/nvme0n1

/dev/nvme0n1:

 Timing O_DIRECT cached reads:     2118 MB in  2.00 seconds = 1058.31 MB/sec

 Timing O_DIRECT disk reads:         3736 MB in  3.00 seconds = 1244.85 MB/sec

it@it:~$

It does appear that everything checks out and that there is a kernel driver for it built-in.

Re: #4, use command "ls -l" in place of "ll". Redhat Linux must have "ll" aliased to "ls -l" and Linux Lite doesn't.

I'm fresh out of ideas, but looks to me like drive should work. Best bet would be to post question on Ubuntu forums. Many more users on that forum and more likely that people there have tackled this problem before.

One final thought:

Can you set the boot priority to boot from the standard HDD first instead of the new drive? If so, maybe try installing grub to MBR on that drive and see if it makes any difference trying to boot from there.

Boot live LL dvd/usb and do following in a terminal:

Code:
sudo mount /dev/nvme0n1p3 /mnt

sudo grub-install boot-directory=/mnt/boot /dev/sda

sudo umount /mnt

Reboot without dvd/usb and with boot priority set to regular hard drive in first position.

If doesn't work, undo the edits to /etc/default/grub made according to my prior post, (run "update-grub" after re-edit, while in chroot); then try rebooting again.

If it works then at least you'll have a solution that makes computer usable. Can keep it like that until find solution to booting with grub on MBR of new drive. Only thing that's happening is the very initial grub boot stage is on MBR of standard drive and passes boot process on to root partition on new drive after that -- so I doubt whole boot process will be significantly different than if it started from MBR of new drive.

Hi,

Many Thanks.... Noted on trying Ubuntu Forum.
Here is the output for point 4 , I did 2 variants:

Code:
it@it:~$ ls -l /dev/nvme*n1

brw-rw---- 1 root disk 259, 0 Nov 24 06:47 /dev/nvme0n1

it@it:~$ ls -l /dev/nvme*

crw------- 1 root root  10, 57 Nov 24 06:47 /dev/nvme0

brw-rw---- 1 root disk 259,  0 Nov 24 06:47 /dev/nvme0n1

brw-rw---- 1 root disk 259,  1 Nov 24 06:47 /dev/nvme0n1p1

brw-rw---- 1 root disk 259,  2 Nov 24 06:47 /dev/nvme0n1p2

brw-rw---- 1 root disk 259,  3 Nov 24 06:47 /dev/nvme0n1p3

brw-rw---- 1 root disk 259,  4 Nov 24 06:47 /dev/nvme0n1p4

brw-rw---- 1 root disk 259,  5 Nov 24 06:47 /dev/nvme0n1p5

Hi,

Many thanks...
So to be clear, go into BIOS, move regular disk to top of list?

Then follow your commands you listed, there is no need to do

Code:
sudo update-grub

at any point..??

It will still know where the Win7 install is..??

Pages: 1 2 3