Re: [gentoo-user] clone root from HDD to SSD causes no video with NVIDIA driver

[Ashley Dixon]

On Tue, Jun 09, 2020 at 05:43:33AM +0200, tu...@posteo.de wrote:
> I would try this:
> Boot your PC, ssh into the PC and download the according
> nvidia-drivers directly from NVIDIA of the same version.

> On 06/08 06:20, Raffaele BELARDI wrote:
> > No console except SSH. I'm not sure I can invoke startx from an SSH.

Irrelevant aside:

The  kernel  loads  the  graphics  drivers  on  boot;  it  is  no   longer   the
responsibility of X under normal circumstances.  Assuming you can get access  to
the kernel command line arguments (with grub, this can be done from  the  bootup
menu [1]), passing the `nomodeset` option will prevent the NVIDIA  drivers  from
loading until you start the X server. There is no need for SSH here.

I had a very similar situation with my AMD drivers a while ago, and  being  able
to use the local (pseudo-)TTY made debugging an order of magnitude less painful.

[1] https://ubuntuforums.org/showthread.php?t=1613132

-- 

Ashley Dixon
suugaku.co.uk

2A9A 4117
DA96 D18A
8A7B B0D2
A30E BF25
F290 A8AA



signature.asc
Description: PGP signature

Re: [gentoo-user] clone root from HDD to SSD causes no video with NVIDIA driver

[tuxic]

Hi,

if even displaying the console login failed, then the hole display
system has gone nuts...but since the boot process as such (that is:
the bios prompt right after POSTing) is visible, I would say, that
there is no physical problem (that is: cable connected to port 2 of 
the monitor while the monitor is switched to port 1 and such).

I would try this:
Boot your PC, ssh into the PC and download the according
nvidia-drivers directly from NVIDIA of the same version.

quickpkg the installed drivers and remove them

Check whether /usr/src/linux links to the kernel
sources of the kernel version you are booting.

Install the NVIDIA-drivers you have downloaded.

Reboot.

Background:
The portage package does not install nvidia-drivers correctly -
in my case, X and such works fine but RTX/Optix which is used
by Blender was defunc. After installing the original package
and masked the one which came with portage everything works
fine.

Cheers!
Meino





On 06/08 06:20, Raffaele BELARDI wrote:
> > -Original Message-
> > From: tu...@posteo.de 
> > Sent: Monday, June 8, 2020 18:14
> > To: gentoo-user@lists.gentoo.org
> > Subject: Re: [gentoo-user] clone root from HDD to SSD causes no video with
> > NVIDIA driver
> > 
> > You said, you are able to ssh into your PC.
> > I would try the following: Boot the PC, ssh into it and disable the start 
> > of X.
> > Boot again: Are you getting the console login successfully?
> 
> X is started by lxdm, which is started by an /etc/local.d/ script. I removed 
> that, after reboot I no longer see X processes, but no conole except for SSH.
> Syslog still shows the nvidia module being loaded. I removed 'modules' from 
> boot runlevel, nvidia is still loaded. I unmerged nvidia-drivers, nvidia 
> still loaded. This is puzzling me.
> 
> > Can you check, whether /dev , /proc , /sys and other directories of a 
> > special
> > function are created and filled correctly?
> > Are the permissions ok?
> > Is /run available and setup correctly?
> 
> To the best of my knowledge yes, they look fine.
> 
> > Are there any leftovers from the root@hd in /etc/fstab?
> 
> I rewrote fstab using UUID instead of /dev/sdx, there shouldn't be problems 
> there.
>  
> > If you get to console successfully, is it possible to start X from the
> > commandline? What is printed on the terminal?
> > What does X.log say?
> 
> No console except SSH. I'm not sure I can invoke startx from an SSH.
> 
> Thanks,
> 
> raffaele
> 

Re: [gentoo-user] Raspberry Pi with 8GB

[Alexandru N. Barloiu]

On Mon, 2020-06-08 at 20:16 -0400, james wrote:
> Any pointers to codes that create a cluster and run on 64Bit arm low 
> power boards is welcome to post to this thread, or drop me a private
> note.

There is no such thing as cluster for arm. It's just daemons. You equip
each pi with the things it's going to need. You treat them as normal
computers.

If you don't know how to start... 
https://gentoo.osuosl.org/experimental/arm64/

there are actually some modern stage3 images. I suggest you google how
to emulate arm64 using qemu-static. google crossdev as well. There are
wonderful resources on the forums, some of which I participated in. 

Goo
d luck and happy hacking. 

Re: [gentoo-user] Raspberry Pi with 8GB

[james]

On 6/8/20 2:36 PM, Michael Jones wrote:


On Sun, Jun 7, 2020 at 8:40 PM james > wrote:


Gentoo Folks,

Has anyone ported gentoo to the newest Raspberry Pi
with 8 gig of ram?


https://www.admin-magazine.com/News/Raspberry-Pi-with-8GB-of-RAM-Now-Available

If so, I'd be curious as to your performance and using it as a
workstation or mobile/laptop.

https://projects.raspberrypi.org/en/projects/raspberry-pi-setting-up

None of the usual gentoo embedded sites I have search list gentoo on
this device.

All input is welcome. 8GB on a 64 bit arm low power (embedded)
board, or
a cluster of 4+ such boards is
of keen interest to my new, gentoo centric low power goals.

TIA,
James


This is the current state of the art, as far as I am aware.

https://github.com/sakaki-/gentoo-on-rpi-64bit



Very cool link !

I'm going to try to buy 4; with the 8 gig of ram. I want to build a 
cluster out of them, for an accelerated portable gentoo system(s).


Any pointers to codes that create a cluster and run on 64Bit arm low 
power boards is welcome to post to this thread, or drop me a private note.



thx,
James

Re: [gentoo-user] clone root from HDD to SSD causes no video with NVIDIA driver

[Neil Bothwick]

On Mon, 8 Jun 2020 18:07:37 +, Raffaele BELARDI wrote:

> > -Original Message-
> > From: Neil Bothwick 
> > Sent: Monday, June 8, 2020 18:07
> > To: gentoo-user@lists.gentoo.org
> > Subject: Re: [gentoo-user] clone root from HDD to SSD causes no video
> > with NVIDIA driver
> > 
> > You're missing device nodes in /dev that are needed before udev
> > populates it. Simply excluding the virtual filesystems as you have
> > results in these files not being copied as they are hidden behind the
> > virtual /dev/ mounted from used. The way I clone a root device is
> > 
> > mkdir -p /mnt/root
> > mount --bind / /mnt/root
> > rsync -a /mnt/root/ /mnt/new/device/
> > 
> > That preserves files that are otherwise hidden by mounted filesystems.
> >   
> I tried this:
> # mkdir -p /mnt/root
> # mount --bind / /mnt/root
> 
> The bind-mounted dev/ does not contain all devices, in particular the
> nvidia ones are missing. BTW, I thought these were created dynamically
> during the module load. Then, for a quick test: # rsync -a
> /mnt/root/dev/* /mnt/new/device/dev

It won't contain all devices, just the bare minimum needed to open a
console etc. The rest are created dynamically when udev starts, all you
need before then is enough to get you to that point.


> But it did not fix the issue.

Then you have a different problem, since the dev directory of the root
partition (which is not the same as /dev on the booted system) should now
contain the same on both disks.
 

-- 
Neil Bothwick

Isn't it a bit unnerving that doctors call what they do "practice?"


pgpsNBeb0NmuZ.pgp
Description: OpenPGP digital signature

Re: [gentoo-user] Raspberry Pi with 8GB

[Michael Jones]

On Sun, Jun 7, 2020 at 8:40 PM james  wrote:

> Gentoo Folks,
>
> Has anyone ported gentoo to the newest Raspberry Pi
> with 8 gig of ram?
>
>
> https://www.admin-magazine.com/News/Raspberry-Pi-with-8GB-of-RAM-Now-Available
>
> If so, I'd be curious as to your performance and using it as a
> workstation or mobile/laptop.
>
> https://projects.raspberrypi.org/en/projects/raspberry-pi-setting-up
>
> None of the usual gentoo embedded sites I have search list gentoo on
> this device.
>
> All input is welcome. 8GB on a 64 bit arm low power (embedded) board, or
> a cluster of 4+ such boards is
> of keen interest to my new, gentoo centric low power goals.
>
>
> TIA,
> James
>
>

This is the current state of the art, as far as I am aware.

https://github.com/sakaki-/gentoo-on-rpi-64bit

RE: [gentoo-user] clone root from HDD to SSD causes no video with NVIDIA driver

[Raffaele BELARDI]

> -Original Message-
> From: tu...@posteo.de 
> Sent: Monday, June 8, 2020 18:14
> To: gentoo-user@lists.gentoo.org
> Subject: Re: [gentoo-user] clone root from HDD to SSD causes no video with
> NVIDIA driver
> 
> You said, you are able to ssh into your PC.
> I would try the following: Boot the PC, ssh into it and disable the start of 
> X.
> Boot again: Are you getting the console login successfully?

X is started by lxdm, which is started by an /etc/local.d/ script. I removed 
that, after reboot I no longer see X processes, but no conole except for SSH.
Syslog still shows the nvidia module being loaded. I removed 'modules' from 
boot runlevel, nvidia is still loaded. I unmerged nvidia-drivers, nvidia still 
loaded. This is puzzling me.

> Can you check, whether /dev , /proc , /sys and other directories of a special
> function are created and filled correctly?
> Are the permissions ok?
> Is /run available and setup correctly?

To the best of my knowledge yes, they look fine.

> Are there any leftovers from the root@hd in /etc/fstab?

I rewrote fstab using UUID instead of /dev/sdx, there shouldn't be problems 
there.
 
> If you get to console successfully, is it possible to start X from the
> commandline? What is printed on the terminal?
> What does X.log say?

No console except SSH. I'm not sure I can invoke startx from an SSH.

Thanks,

raffaele

RE: [gentoo-user] clone root from HDD to SSD causes no video with NVIDIA driver

[Raffaele BELARDI]

> -Original Message-
> From: Neil Bothwick 
> Sent: Monday, June 8, 2020 18:07
> To: gentoo-user@lists.gentoo.org
> Subject: Re: [gentoo-user] clone root from HDD to SSD causes no video with
> NVIDIA driver
> 
> You're missing device nodes in /dev that are needed before udev populates
> it. Simply excluding the virtual filesystems as you have results in these 
> files
> not being copied as they are hidden behind the virtual /dev/ mounted from
> used. The way I clone a root device is
> 
> mkdir -p /mnt/root
> mount --bind / /mnt/root
> rsync -a /mnt/root/ /mnt/new/device/
> 
> That preserves files that are otherwise hidden by mounted filesystems.
> 
I tried this:
# mkdir -p /mnt/root
# mount --bind / /mnt/root

The bind-mounted dev/ does not contain all devices, in particular the nvidia 
ones are missing. BTW, I thought these were created dynamically during the 
module load.
Then, for a quick test:
# rsync -a /mnt/root/dev/* /mnt/new/device/dev

But it did not fix the issue.

raffaele

Re: [gentoo-user] clone root from HDD to SSD causes no video with NVIDIA driver

[tuxic]

Hi Raffaele,

may be this could give some insight of was is happening.

You said, you are able to ssh into your PC. 
I would try the following: Boot the PC, ssh into it and disable the
start of X. Boot again: Are you getting the console login
successfully?

Can you check, whether /dev , /proc , /sys and other directories of a
special function are created and filled correctly?
Are the permissions ok?
Is /run available and setup correctly?

Are there any leftovers from the root@hd in /etc/fstab?

If you get to console successfully, is it possible to start X from
the commandline? What is printed on the terminal?
What does X.log say?

Hopefully we will find something to fix that problem!

Cheers!
Meino




On 06/08 04:02, Raffaele BELARDI wrote:
> Hello,
> 
> I am trying to switch an existing and fully working HDD, MBR-based ~amd64 
> LXDE/Kodi setup to a 500Gb SSD. I also took the opportunity to dual boot 
> Gentoo with Win10 on the same SSD. As suggested by the Wiki [1] I first 
> installed Win letting it GPT-partition the SSD and leaving some empty space 
> for a Gentoo ext4 root partition. Then I copied the Gentoo root from the HDD 
> to the SSD ('cp -a' except /dev, /sys and /proc which I recreated empty on 
> the SSD), UEFI-booted a SysRescue USB and installed GRUB2 to the 
> Windows-created ESP. Now, while the chain-loaded Windows boots happily, I'm 
> not able to get any GUI or terminal interface for SSD-booted Linux. The old, 
> HDD-based installation is still available and running fine.
> 
> In details:
> 
>   1.  GRUB starts and shows the linux/windows choices, select linux
>   2.  The screen changes to 'loading linux-x.x.x', the disk light shows 
> activity, I have indirect evidence that the OpenRC init reaches conclusion, 
> but the screen does not change
>   3.  I am able to SSH into the PC, all partitions are mounted, everything 
> looks fine
>   4.  Xorg.0.log shows NVIDIA driver was loaded but probably not executed, 
> all the NVIDIA lines are missing
>   5.  Only /dev/nvidiactl is present, /dev/nvidia0 and /dev/nvidia-modeset 
> are missing
>   6.  Syslog shows that the NVIDIA driver was loaded by the kernel, no crash
>   7.  CTRL-ALT-Fn  apparently has no effect
>   8.  I'm sure the kernel is fine for this NVIDIA driver because the HDD 
> installation works with it, nevertheless I recompiled and reinstalled the 
> kernel and NVIDIA drivers to the SSD from the SSH, no change
> 
> Before I give up and reinstall Gentoo from scratch on the SSD what else could 
> I try? I can share the logs if needed.
> 
> Thanks,
> 
> Raffaele
> 
> [1] https://wiki.gentoo.org/wiki/UEFI_Dual_boot_with_Windows_7/8

Re: [gentoo-user] clone root from HDD to SSD causes no video with NVIDIA driver

[Neil Bothwick]

On Mon, 8 Jun 2020 16:02:35 +, Raffaele BELARDI wrote:

> I am trying to switch an existing and fully working HDD, MBR-based
> ~amd64 LXDE/Kodi setup to a 500Gb SSD. I also took the opportunity to
> dual boot Gentoo with Win10 on the same SSD. As suggested by the Wiki
> [1] I first installed Win letting it GPT-partition the SSD and leaving
> some empty space for a Gentoo ext4 root partition. Then I copied the
> Gentoo root from the HDD to the SSD ('cp -a' except /dev, /sys and
> /proc which I recreated empty on the SSD), UEFI-booted a SysRescue USB
> and installed GRUB2 to the Windows-created ESP. Now, while the
> chain-loaded Windows boots happily, I'm not able to get any GUI or
> terminal interface for SSD-booted Linux. The old, HDD-based
> installation is still available and running fine.

You're missing device nodes in /dev that are needed before udev populates
it. Simply excluding the virtual filesystems as you have results in
these files not being copied as they are hidden behind the virtual
/dev/ mounted from used. The way I clone a root device is

mkdir -p /mnt/root
mount --bind / /mnt/root
rsync -a /mnt/root/ /mnt/new/device/

That preserves files that are otherwise hidden by mounted filesystems.


-- 
Neil Bothwick

Things are more like they are today than they ever have been before.


pgpSRCLRd_hTY.pgp
Description: OpenPGP digital signature

[gentoo-user] clone root from HDD to SSD causes no video with NVIDIA driver

[Raffaele BELARDI]

Hello,

I am trying to switch an existing and fully working HDD, MBR-based ~amd64 
LXDE/Kodi setup to a 500Gb SSD. I also took the opportunity to dual boot Gentoo 
with Win10 on the same SSD. As suggested by the Wiki [1] I first installed Win 
letting it GPT-partition the SSD and leaving some empty space for a Gentoo ext4 
root partition. Then I copied the Gentoo root from the HDD to the SSD ('cp -a' 
except /dev, /sys and /proc which I recreated empty on the SSD), UEFI-booted a 
SysRescue USB and installed GRUB2 to the Windows-created ESP. Now, while the 
chain-loaded Windows boots happily, I'm not able to get any GUI or terminal 
interface for SSD-booted Linux. The old, HDD-based installation is still 
available and running fine.

In details:

  1.  GRUB starts and shows the linux/windows choices, select linux
  2.  The screen changes to 'loading linux-x.x.x', the disk light shows 
activity, I have indirect evidence that the OpenRC init reaches conclusion, but 
the screen does not change
  3.  I am able to SSH into the PC, all partitions are mounted, everything 
looks fine
  4.  Xorg.0.log shows NVIDIA driver was loaded but probably not executed, all 
the NVIDIA lines are missing
  5.  Only /dev/nvidiactl is present, /dev/nvidia0 and /dev/nvidia-modeset are 
missing
  6.  Syslog shows that the NVIDIA driver was loaded by the kernel, no crash
  7.  CTRL-ALT-Fn  apparently has no effect
  8.  I'm sure the kernel is fine for this NVIDIA driver because the HDD 
installation works with it, nevertheless I recompiled and reinstalled the 
kernel and NVIDIA drivers to the SSD from the SSH, no change

Before I give up and reinstall Gentoo from scratch on the SSD what else could I 
try? I can share the logs if needed.

Thanks,

Raffaele

[1] https://wiki.gentoo.org/wiki/UEFI_Dual_boot_with_Windows_7/8

Re: [gentoo-user] Gentoo-sources 5.7.x

[Neil Bothwick]

On Mon, 8 Jun 2020 11:38:03 -0400, Rich Freeman wrote:

> > # cat /boot/loader/entries/30-gentoo-5.7.1.conf
> > title Gentoo Linux 5.7.1
> > linux /vmlinuz-5.7.1-gentoo
> > options root=/dev/nvme0n1p4 initrd=/intel-uc.img net.ifnames=0
> > raid=noautodetect
> >
> > That file hasn't changed, other than the kernel version.
> >
> > On booting with that setup, I got an error immediately (from the UEFI
> > BIOS, I assume) complaining of an invalid parameter. A few seconds
> > later the kernel panicked because it couldn't find the root device.  
> 
> Can you have your initrd launch a shell?  I'm not sure if the
> nomenclature for the root device has changed.  It is also possible
> you're missing some kernel option needed (maybe an NVME option got
> renamed and you answered the new option No?).
> 
> If it is running the initrd then you probably can get it to launch a
> shell and then you can poke around and see what it is doing.

That's not an initrd with a shell though, it's just the intel microcode
update. You would need an additional dracut (or similar) generated initrd
to do what you suggest.


-- 
Neil Bothwick

I am McCoy of Bo...Damnit! I'm a doctor, not a collective!


pgprnm7xtMMip.pgp
Description: OpenPGP digital signature

Re: [gentoo-user] Gentoo-sources 5.7.x

[Rich Freeman]

On Mon, Jun 8, 2020 at 11:06 AM Peter Humphrey  wrote:
>
> Afternoon all,
>
> Is there something special for me to set in the kernel config to enable it to
> find the root partition? I copied the config from 5.4.38, ran oldconfig and
> followed most of the suggested answers; but it won't boot.
>
> # cat /boot/loader/entries/30-gentoo-5.7.1.conf
> title Gentoo Linux 5.7.1
> linux /vmlinuz-5.7.1-gentoo
> options root=/dev/nvme0n1p4 initrd=/intel-uc.img net.ifnames=0
> raid=noautodetect
>
> That file hasn't changed, other than the kernel version.
>
> On booting with that setup, I got an error immediately (from the UEFI BIOS, I
> assume) complaining of an invalid parameter. A few seconds later the kernel
> panicked because it couldn't find the root device.

Can you have your initrd launch a shell?  I'm not sure if the
nomenclature for the root device has changed.  It is also possible
you're missing some kernel option needed (maybe an NVME option got
renamed and you answered the new option No?).

If it is running the initrd then you probably can get it to launch a
shell and then you can poke around and see what it is doing.

Offhand though those are my two guesses:
1.  Driver config option got renamed and the new option is disabled.
2.  Some kernel change alters the naming of the root device so you're
giving it the wrong name now.

If you're using an initrd you should consider using a UUID/lable/etc
to identify the root device if possible.  That tends to be more robust
when stuff like this happens, though it won't help in #1.

-- 
Rich

Re: [gentoo-user] Gentoo-sources 5.7.x

[Andrew Udvare]

On 08/06/2020 11:06, Peter Humphrey wrote:
> Afternoon all,
> 
> Is there something special for me to set in the kernel config to enable it to 
> find the root partition? I copied the config from 5.4.38, ran oldconfig and 
> followed most of the suggested answers; but it won't boot.

Sounds like missing drivers. oldconfig didn't do everything it was
supposed to. Moving across multiple major versions, this is to be
expected. A lot of names of things have changed.

Do a comparison of your configuration between old and new.

diff -uN old-config-file /usr/src/linux/.config

Make sure to at least enable NVME with CONFIG_BLK_DEV_NVME=y and try
booting 5.7 again. Other than that, the naming scheme may have changed
but I don't know about this. For better future-proofing, use a UUID of
your root partition rather than a device name.

root=UUID=...

You can get this UUID with the blkid command.

Andrew



signature.asc
Description: OpenPGP digital signature

[gentoo-user] Gentoo-sources 5.7.x

[Peter Humphrey]

Afternoon all,

Is there something special for me to set in the kernel config to enable it to 
find the root partition? I copied the config from 5.4.38, ran oldconfig and 
followed most of the suggested answers; but it won't boot.

# cat /boot/loader/entries/30-gentoo-5.7.1.conf
title Gentoo Linux 5.7.1
linux /vmlinuz-5.7.1-gentoo
options root=/dev/nvme0n1p4 initrd=/intel-uc.img net.ifnames=0 
raid=noautodetect

That file hasn't changed, other than the kernel version.

On booting with that setup, I got an error immediately (from the UEFI BIOS, I 
assume) complaining of an invalid parameter. A few seconds later the kernel 
panicked because it couldn't find the root device.

So I chopped off everything after the 1p4 and booted again. No invalid-
parameter error this time, but still it couldn't find the boot device.

Something has changed, and I can't see what it is.

I've tried compiling the kernel (and @system) with gcc-9.3.0 and with 
gcc-10.1.0, but I saw no difference.

-- 
Regards,
Peter.

Re: [gentoo-user] What's the best way to force a particular version of a dependency

[Peter Humphrey]

On Monday, 8 June 2020 10:18:30 BST Neil Bothwick wrote:
> On Mon, 08 Jun 2020 10:12:30 +0100, Peter Humphrey wrote:
> > > @world includes @system.  It doesn't necessarily include everything
> > > installed on your system though.
> > 
> > Eh? Do you mean we might have things installed other than via portage?
> > Otherwise, isn't everything on the system represented in either the
> > world file or sets?
> 
> You may have installed something with --oneshot for testing, or to cover
> a missing dependency.

Ah. Of course.

-- 
Regards,
Peter.

Re: [gentoo-user] What's the best way to force a particular version of a dependency

[Neil Bothwick]

On Mon, 08 Jun 2020 10:12:30 +0100, Peter Humphrey wrote:

> > @world includes @system.  It doesn't necessarily include everything
> > installed on your system though.  
> 
> Eh? Do you mean we might have things installed other than via portage? 
> Otherwise, isn't everything on the system represented in either the
> world file or sets?

You may have installed something with --oneshot for testing, or to cover
a missing dependency.


-- 
Neil Bothwick

With free advice you often get what you pay for.


pgpKynhA2JY4Z.pgp
Description: OpenPGP digital signature

Re: [gentoo-user] What's the best way to force a particular version of a dependency

[Peter Humphrey]

On Sunday, 7 June 2020 22:16:39 BST Rich Freeman wrote:

> @world includes @system.  It doesn't necessarily include everything
> installed on your system though.

Eh? Do you mean we might have things installed other than via portage? 
Otherwise, isn't everything on the system represented in either the world file 
or sets?

-- 
Regards,
Peter.