unable to get Fedora login screen.

[home user]

I tried to upgrade from Fedora-40 (patched this morning) to Fedora-41.  
I followed the instructions in the Fedora web site 
".../upgrading-fedora-offline" document.  After the "dnf system-upgrade 
reboot" step, the boot process never resulted in a login screen.  I 
tried several times, including recycling the power switch on the back of 
the tower.  Trying to boot Fedora-40 also does not result in a login 
screen.  I tried that a few times.

This is a stand-alone dual boot workstation, the other OS being 
windows-7, which I am using now to send out this message.

My windows-7 install does have the tool "ext2explore".  That allows me 
to copy files from the Fedora partitions into windows-7, where I can 
view those files with w7's standard tools.

New hardware is out of the question until early summer.

I've never been able to make a live USB and then successfully boot from 
it; I don't know why.

I do not have a rescue.

Help!!!

[Marco Moock]

Am 10.04.2025 um 12:21:15 Uhr schrieb home user via users:
> I've never been able to make a live USB and then successfully boot
> from it; I don't know why.
>
> I do not have a rescue.

Specify what you see, especially after you disabled the boot option
"quiet", which can be done in GRUB2 by using the e key after selecting
the boot entry.

[home user]

On 04/10/2025 12:36 PM, Marco Moock wrote:
> Am 10.04.2025 um 12:21:15 Uhr schrieb home user via users:
> > I've never been able to make a live USB and then successfully boot
> > from it; I don't know why.
> >
> > I do not have a rescue.
>
> Specify what you see, especially after you disabled the boot option
> "quiet", which can be done in GRUB2 by using the e key after selecting
> the boot entry.

It's already set with "quiet" off.  What I see is a lot of logging, 
mostly gray, some "teal"(?), all flying by too fast to catch what it 
says.  Once the logging stops, all I have is a blank screen.  I've been 
previously told there's no way to temporarily "freeze" the screen or the 
boot process during boot.  I do not have a cell phone with which to 
capture the boot process logs.

I did notice some multi-line log message about USB and it seemed bad, 
but I couldn't get enough of it.

I did see something about falling back to nouveau.  But I've seen that 
before, and it did not prevent reaching a login screen and logging in.

I do also see something about a sendmail failure.  But that too has been 
happening for a while and not causing significant problems.

[Marco Moock]

Am 10.04.2025 um 13:12:07 Uhr schrieb home user via users:
> On 04/10/2025 12:36 PM, Marco Moock wrote:
> > Am 10.04.2025 um 12:21:15 Uhr schrieb home user via users:
> > > I've never been able to make a live USB and then successfully boot
> > > from it; I don't know why.
> > >
> > > I do not have a rescue.
> >
> > Specify what you see, especially after you disabled the boot option
> > "quiet", which can be done in GRUB2 by using the e key after
> > selecting the boot entry.
>
> It's already set with "quiet" off.  What I see is a lot of logging, 
> mostly gray, some "teal"(?), all flying by too fast to catch what it 
> says.  Once the logging stops, all I have is a blank screen.  I've
> been previously told there's no way to temporarily "freeze" the
> screen or the boot process during boot.  I do not have a cell phone
> with which to capture the boot process logs.

I assume an issue with the graphical interface. Change the systemd
target in grub
systemd.unit=multi-user.target

Which GPU do you have and which driver is installed?

[home user]

On 04/10/2025 1:19 PM, Marco Moock wrote:
> Am 10.04.2025 um 13:12:07 Uhr schrieb home user via users:
> > On 04/10/2025 12:36 PM, Marco Moock wrote:
> > > Am 10.04.2025 um 12:21:15 Uhr schrieb home user via users:
> > > > I've never been able to make a live USB and then successfully boot
> > > > from it; I don't know why.
> > > >
> > > > I do not have a rescue.
> > >
> > > Specify what you see, especially after you disabled the boot option
> > > "quiet", which can be done in GRUB2 by using the e key after
> > > selecting the boot entry.
> >
> > It's already set with "quiet" off.  What I see is a lot of logging,
> > mostly gray, some "teal"(?), all flying by too fast to catch what it
> > says.  Once the logging stops, all I have is a blank screen.  I've
> > been previously told there's no way to temporarily "freeze" the
> > screen or the boot process during boot.  I do not have a cell phone
> > with which to capture the boot process logs.
>
> I assume an issue with the graphical interface. Change the systemd
> target in grub
> systemd.unit=multi-user.target
>
> Which GPU do you have and which driver is installed?

I added that to the longest line in the f-41 grub editor presents, and 
booted.

It got me to a console login.

While there, I attempted to find 2 key boot log lines.  Using grep, 
redirection, and windows-7's ext2explore, I brought the results into 
windows-7.  Here are those log entries.  I don't know a way of getting 
all the formatting codes out.

------

[[0;1;38:5:185mDEPEND] Dependency failed for 
mnt-usb\x2dU…001c2b1897558107c2af0beb-0:0-part1.

         Starting nvidia-fallback.service -…o nouveau as 
nvidia did not load...

[  OK  ] Finished nvidia-fallback.service -… to 
nouveau as nvidia did not load.

Starting nvidia-fallback.service -…o nouveau as 
nvidia did not load...

[  OK  ] Finished nvidia-fallback.service -… to 
nouveau as nvidia did not load.

Starting nvidia-fallback.service -…o nouveau as 
nvidia did not load...

[  OK  ] Finished nvidia-fallback.service -… to 
nouveau as nvidia did not load.

         Starting nvidia-fallback.service -…o nouveau as 
nvidia did not load...

[  OK  ] Finished nvidia-fallback.service -… to 
nouveau as nvidia did not load.

         Starting nvidia-fallback.service -…o nouveau as 
nvidia did not load...

[  OK  ] Finished nvidia-fallback.service -… to 
nouveau as nvidia did not load.

Starting nvidia-fallback.service -…o nouveau as 
nvidia did not load...

[  OK  ] Finished nvidia-fallback.service -… to 
nouveau as nvidia did not load.

------

The graphics card is nvidia GeForce GTX 660.

The driver is the one from the RPM Fusion repository.  Last attempt to 
update was this morning, just before trying the f40 to f41 upgrade.

[Jeffrey Walton]

On Thu, Apr 10, 2025 at 4:20 PM home user via users
 wrote:
> On 04/10/2025 1:19 PM, Marco Moock wrote:
> > [...]
> > I assume an issue with the graphical interface. Change the systemd
> > target in grub
> > systemd.unit=multi-user.target
> >
> > Which GPU do you have and which driver is installed?
>
> I added that to the longest line in the f-41 grub editor presents, and
> booted.
>
> It got me to a console login.
>
> While there, I attempted to find 2 key boot log lines.  Using grep,
> redirection, and windows-7's ext2explore, I brought the results into
> windows-7.  Here are those log entries.  I don't know a way of getting
> all the formatting codes out.
>
> [ [0;1;38:5:185mDEPEND [0m] Dependency failed for
>  [0;1;39mmnt-usb\x2dU… [0m001c2b1897558107c2af0beb-0:0-part1.
>
> [K         Starting  [0;1;39mnvidia-fallback.service [0m -…o nouveau as
> nvidia did not load...
>
> [ [0;32m  OK   [0m] Finished  [0;1;39mnvidia-fallback.service [0m -… to
> nouveau as nvidia did not load.
>       Starting  [0;1;39mnvidia-fallback.service [0m -…o nouveau as
>
> nvidia did not load...
>
> [ [0;32m  OK   [0m] Finished  [0;1;39mnvidia-fallback.service [0m -… to
> nouveau as nvidia did not load.
>       Starting  [0;1;39mnvidia-fallback.service [0m -…o nouveau as
>
> nvidia did not load...
>
> [ [0;32m  OK   [0m] Finished  [0;1;39mnvidia-fallback.service [0m -… to
> nouveau as nvidia did not load.
>
> [K         Starting  [0;1;39mnvidia-fallback.service [0m -…o nouveau as
> nvidia did not load...
>
> [ [0;32m  OK   [0m] Finished  [0;1;39mnvidia-fallback.service [0m -… to
> nouveau as nvidia did not load.
>
> [K         Starting  [0;1;39mnvidia-fallback.service [0m -…o nouveau as
> nvidia did not load...
>
> [ [0;32m  OK   [0m] Finished  [0;1;39mnvidia-fallback.service [0m -… to
> nouveau as nvidia did not load.
>       Starting  [0;1;39mnvidia-fallback.service [0m -…o nouveau as
>
> nvidia did not load...
>
> [ [0;32m  OK   [0m] Finished  [0;1;39mnvidia-fallback.service [0m -… to
> nouveau as nvidia did not load.
>
> The graphics card is nvidia GeForce GTX 660.
>
> The driver is the one from the RPM Fusion repository.  Last attempt to
> update was this morning, just before trying the f40 to f41 upgrade.

According to Google, you need rpmfusion's akmod-nvidia-470xx for the
GeForce GTX 660 drivers.

Can you confirm the package is installed?

Jeff

[home user]

On 04/10/2025 2:28 PM, Jeffrey Walton wrote:
> On Thu, Apr 10, 2025 at 4:20 PM home user via users
>  wrote:
> > On 04/10/2025 1:19 PM, Marco Moock wrote:
> > > [...]
> > > I assume an issue with the graphical interface. Change the systemd
> > > target in grub
> > > systemd.unit=multi-user.target
> > >
> > > Which GPU do you have and which driver is installed?
> >
> > I added that to the longest line in the f-41 grub editor presents, and
> > booted.
> >
> > It got me to a console login.
> >
> > While there, I attempted to find 2 key boot log lines.  Using grep,
> > redirection, and windows-7's ext2explore, I brought the results into
> > windows-7.  Here are those log entries.  I don't know a way of getting
> > all the formatting codes out.
> >
> > [ [0;1;38:5:185mDEPEND [0m] Dependency failed for
> >   [0;1;39mmnt-usb\x2dU… [0m001c2b1897558107c2af0beb-0:0-part1.
> >
> > [K         Starting  [0;1;39mnvidia-fallback.service [0m -…o nouveau as
> > nvidia did not load...
> >
> > [ [0;32m  OK   [0m] Finished  [0;1;39mnvidia-fallback.service [0m -… to
> > nouveau as nvidia did not load.
> >        Starting  [0;1;39mnvidia-fallback.service [0m -…o nouveau as
> >
> > nvidia did not load...
> >
> > [ [0;32m  OK   [0m] Finished  [0;1;39mnvidia-fallback.service [0m -… to
> > nouveau as nvidia did not load.
> >        Starting  [0;1;39mnvidia-fallback.service [0m -…o nouveau as
> >
> > nvidia did not load...
> >
> > [ [0;32m  OK   [0m] Finished  [0;1;39mnvidia-fallback.service [0m -… to
> > nouveau as nvidia did not load.
> >
> > [K         Starting  [0;1;39mnvidia-fallback.service [0m -…o nouveau as
> > nvidia did not load...
> >
> > [ [0;32m  OK   [0m] Finished  [0;1;39mnvidia-fallback.service [0m -… to
> > nouveau as nvidia did not load.
> >
> > [K         Starting  [0;1;39mnvidia-fallback.service [0m -…o nouveau as
> > nvidia did not load...
> >
> > [ [0;32m  OK   [0m] Finished  [0;1;39mnvidia-fallback.service [0m -… to
> > nouveau as nvidia did not load.
> >        Starting  [0;1;39mnvidia-fallback.service [0m -…o nouveau as
> >
> > nvidia did not load...
> >
> > [ [0;32m  OK   [0m] Finished  [0;1;39mnvidia-fallback.service [0m -… to
> > nouveau as nvidia did not load.
> >
> > The graphics card is nvidia GeForce GTX 660.
> >
> > The driver is the one from the RPM Fusion repository.  Last attempt to
> > update was this morning, just before trying the f40 to f41 upgrade.
>
> According to Google, you need rpmfusion's akmod-nvidia-470xx for the
> GeForce GTX 660 drivers.
>
> Can you confirm the package is installed?
>
> Jeff

1. akmod-nvidia-470xx

From the f41 console login, I grepped the dnf logs, and then grepped 
the results for todays's date.  Here are the results:

------

/var/log/dnf.librepo.log:2025-04-10T10:09:00-0600 INFO Downloading: 
https://mirror.fcix.net/rpmfusion/nonfree/fedora/updates/41/x86_64/a/akmod-n...
/var/log/dnf.log:2025-04-10T11:28:32-0600 DEBUG Upgraded: 
akmod-nvidia-470xx-3:470.256.02-7.fc41.x86_64
/var/log/dnf.log.1:2025-04-10T10:15:52-0600 DEBUG ---> Package 
akmod-nvidia-470xx.x86_64 3:470.256.02-7.fc40 will be upgraded
/var/log/dnf.log.1:2025-04-10T10:15:52-0600 DEBUG ---> Package 
akmod-nvidia-470xx.x86_64 3:470.256.02-7.fc41 will be an upgrade
/var/log/dnf.log.3:2025-04-10T09:53:49-0600 DEBUG ---> Package 
akmod-nvidia-470xx.x86_64 3:470.256.02-7.fc40 will be upgraded
/var/log/dnf.log.3:2025-04-10T09:53:49-0600 DEBUG ---> Package 
akmod-nvidia-470xx.x86_64 3:470.256.02-7.fc41 will be an upgrade
/var/log/dnf.rpm.log:2025-04-10T10:53:15-0600 SUBDEBUG Upgrade: 
akmod-nvidia-470xx-3:470.256.02-7.fc41.x86_64
/var/log/dnf.rpm.log:2025-04-10T11:08:17-0600 SUBDEBUG Upgraded: 
akmod-nvidia-470xx-3:470.256.02-7.fc40.x86_64

------

But it's not clear to me whether the driver is properly installed.  I 
recall that in a recent thread, George referenced a web page 
"linux-hardware.org".  Based on that, last week, I installed hw-probe 
and ran it.  A short while ago, from the f41 console login, I found that 
this morning's f40 to f41 upgrade lost hw-probe.  I re-installed it and 
ran it, piping output via "tee" to an output text file.  Here is that 
output:

------

Probe for hardware ... Ok
Reading logs ... Ok
WARNING: failed to detect EDID
Uploaded to DB, Thank you!

Probe URL: https://linux-hardware.org/?probe=7fa74ee631

------

I looked at the web page the the output referenced.    It's complaining 
about the driver.

a. dnf logs show the driver installed.

b. hw-probe shows no graphics driver.  but I ran it in the f41 console.  
does that make a difference?

c. The upgrade did lose something: hw-probe.  I wonder what else went wrong.

d. iffor this thread, anyone is wondering what hardware I have, the web 
page referenced in the hw-probe output above should answer.

My final answer is that I don't know if the proper graphics driver is 
properly installed or not.  It was just before this morning's "dnf 
system-upgrade reboot".

[Samuel Sieb]

On 4/10/25 3:43 PM, home user via users wrote:
> My final answer is that I don't know if the proper graphics driver is 
> properly installed or not.  It was just before this morning's "dnf 
> system-upgrade reboot".

Try running "akmods".

[home user]

On 04/10/2025 6:27 PM, Samuel Sieb wrote:
> On 4/10/25 3:43 PM, home user via users wrote:
> > My final answer is that I don't know if the proper graphics driver is 
> > properly installed or not.  It was just before this morning's "dnf 
> > system-upgrade reboot".
>
> Try running "akmods".

booted f41 with "systemd.unit=multi-user.target".

f41 console login.

ran "akmods | tee akmods.txt"; the output file...

-----

Checking kmods exist for 6.13.9-200.fc41.x86_64[  OK ]
Building and installing nvidia-470xx-kmod[FAILED]

Hint: Some kmods were ignored or failed to build or install.
You can try to rebuild and install them by by calling
'/usr/sbin/akmods --force' as root.

-----

ran "akmods --force | tee akmodsf.txt"; the output file...

-----

Checking kmods exist for 6.13.9-200.fc41.x86_64[  OK ]
Building and installing nvidia-470xx-kmod[FAILED]

Hint: Some kmods were ignored or failed to build or install.
You can try to rebuild and install them by by calling
'/usr/sbin/akmods --force' as root.

-----

The output to the screen also suggested looking at a certain log file.  
Here's that log file:

-----

2025/04/10 18:33:41 akmods: Building RPM using the command 
'/usr/sbin/akmodsbuild --kernels 6.13.9-200.fc41.x86_64 
/usr/src/akmods/nvidia-470xx-kmod.latest'
Building target platforms: x86_64
Building for target x86_64
Installing /usr/src/akmods/nvidia-470xx-kmod.latest
setting SOURCE_DATE_EPOCH=1740268800
error: Failed build dependencies:
     kernel-devel-uname-r = 6.13.9-200.fc41.x86_64 is needed by 
nvidia-470xx-kmod-3:470.256.02-7.fc41.x86_64
2025/04/10 18:33:42 akmodsbuild: * Rebuilding 
/usr/src/akmods/nvidia-470xx-kmod.latest for kernel(s) 
6.13.9-200.fc41.x86_64: Building target platforms: x86_64
2025/04/10 18:33:42 akmodsbuild: Building for target x86_64
2025/04/10 18:33:42 akmodsbuild: Installing 
/usr/src/akmods/nvidia-470xx-kmod.latest
2025/04/10 18:33:42 akmodsbuild: setting SOURCE_DATE_EPOCH=1740268800
2025/04/10 18:33:42 akmodsbuild: error: Failed build dependencies:
2025/04/10 18:33:42 akmodsbuild:     kernel-devel-uname-r = 
6.13.9-200.fc41.x86_64 is needed by 
nvidia-470xx-kmod-3:470.256.02-7.fc41.x86_64
2025/04/10 18:33:42 akmodsbuild:
2025/04/10 18:33:42 akmods: Building rpms failed; see 
/var/cache/akmods/nvidia-470xx/470.256.02-7-for-6.13.9-200.fc41.x86_64.failed.log 
for details

-----

It doesn't make sense to me.

[Samuel Sieb]

On 4/10/25 5:59 PM, home user via users wrote:
> 2025/04/10 18:33:41 akmods: Building RPM using the command '/usr/sbin/ 
> akmodsbuild --kernels 6.13.9-200.fc41.x86_64 /usr/src/akmods/ 
> nvidia-470xx-kmod.latest'
> Building target platforms: x86_64
> Building for target x86_64
> Installing /usr/src/akmods/nvidia-470xx-kmod.latest
> setting SOURCE_DATE_EPOCH=1740268800
> error: Failed build dependencies:
>      kernel-devel-uname-r = 6.13.9-200.fc41.x86_64 is needed by 
> nvidia-470xx-kmod-3:470.256.02-7.fc41.x86_64
> 2025/04/10 18:33:42 akmodsbuild: * Rebuilding /usr/src/akmods/ 
> nvidia-470xx-kmod.latest for kernel(s) 6.13.9-200.fc41.x86_64: Building 
> target platforms: x86_64
> 2025/04/10 18:33:42 akmodsbuild: Building for target x86_64
> 2025/04/10 18:33:42 akmodsbuild: Installing /usr/src/akmods/ 
> nvidia-470xx-kmod.latest
> 2025/04/10 18:33:42 akmodsbuild: setting SOURCE_DATE_EPOCH=1740268800
> 2025/04/10 18:33:42 akmodsbuild: error: Failed build dependencies:
> 2025/04/10 18:33:42 akmodsbuild:     kernel-devel-uname-r = 
> 6.13.9-200.fc41.x86_64 is needed by nvidia-470xx- 
> kmod-3:470.256.02-7.fc41.x86_64
> 2025/04/10 18:33:42 akmodsbuild:
> 2025/04/10 18:33:42 akmods: Building rpms failed; see /var/cache/akmods/ 
> nvidia-470xx/470.256.02-7-for-6.13.9-200.fc41.x86_64.failed.log for details
>
> It doesn't make sense to me.

Install "kernel-devel".

[home user]

On 04/10/2025 7:05 PM, Samuel Sieb wrote:
> On 4/10/25 5:59 PM, home user via users wrote:
> > ... snip ...
>
> Install "kernel-devel".

The tee's did not capture everything that showed up on the console.

"dnf install kernel-devel"...

-----

Nothing to do.

-----

The command also showed on-screen only

"Package "kernel-devel-6.10.12-200.fc40.x86_64"   is already installed".

So I tried "dnf upgrade kernel-devel"...

-----

Package       Arch   Version         Repository      Size
Installing:
  kernel-devel x86_64 6.13.9-100.fc40 updates     77.7 MiB

Transaction Summary:
  Installing:         1 package
-----

It upgraded, but I need fc41, not fc40.

"uname -a" did report that I was using fc41.

I re-tried "akmods" and "akmods --force", but no improvement.  I have 
the "tee" from those if you want them.  And I have the ...failed.log if 
you want it.

How do I get kernel-devel up to 6.13.9-200.fc41, which is the top entry 
in the grub menu?

...or is there something more/else needed?

[Marco Moock]

Am Thu, 10 Apr 2025 19:58:33 -0600
schrieb home user via users :
> Package       Arch   Version         Repository      Size
> Installing:
>   kernel-devel x86_64 6.13.9-100.fc40 updates     77.7 MiB
>
> Transaction Summary:
>   Installing:         1 package
>
> It upgraded, but I need fc41, not fc40.
>
> "uname -a" did report that I was using fc41.
>
> I re-tried "akmods" and "akmods --force", but no improvement.  I have 
> the "tee" from those if you want them.  And I have the ...failed.log
> if you want it.
>
> How do I get kernel-devel up to 6.13.9-200.fc41, which is the top
> entry in the grub menu?
>
> ...or is there something more/else needed?

cat /etc/yum.repos.d/fedora*

Then dnf update

Also consider removing the nvidia package at all and at least
temporarily switch to nouveau, which should support your card well.

[Samuel Sieb]

On 4/10/25 6:58 PM, home user via users wrote:
> So I tried "dnf upgrade kernel-devel"...
>
> Package       Arch   Version         Repository      Size
> Installing:
>   kernel-devel x86_64 6.13.9-100.fc40 updates     77.7 MiB
>
> Transaction Summary:
>   Installing:         1 package
>
> It upgraded, but I need fc41, not fc40.
>
> "uname -a" did report that I was using fc41.

Was it you or someone else that got into this same situation recently?

What does "rpm -q fedora-release" show you?

You can try "dnf upgrade --releasever=41 kernel-devel" for now.

[home user]

(replying to 2)

(Marco)
> cat /etc/yum.repos.d/fedora*

(see attached file "repos.txt")
> Then dnf update(I'll come back to this)

On 04/10/2025 11:57 PM, Samuel Sieb wrote:
> On 4/10/25 6:58 PM, home user via users wrote:
> > So I tried "dnf upgrade kernel-devel"...
> >
> > Package       Arch   Version         Repository      Size
> > Installing:
> >   kernel-devel x86_64 6.13.9-100.fc40 updates     77.7 MiB
> >
> > Transaction Summary:
> >   Installing:         1 package
> >
> > It upgraded, but I need fc41, not fc40.
> >
> > "uname -a" did report that I was using fc41.
>
> Was it you or someone else that got into this same situation recently?

Not sure.  In early or mid October, I upgraded from f39 to f40.  There 
were major problems.  There was some similarity to this, but definitely 
not identical.  A key part of the problem was the post-upgrade task to 
clean up retired packages apparently malfunctioned big-time. You were 
one of the main helpers on that.
> What does "rpm -q fedora-release" show you?

"fedora-release-40-39.noarch"
> You can try "dnf upgrade --releasever=41 kernel-devel" for now.

I did that from the f41 console (I assume dnf5), booted to with 
"systemd.unit=multi-user.target".  (See the attached file 
"upgrade0411.txt").

The message "[  700.871624] block sda: the capability attribute has been 
deprecated." showed up on the screen near the end of the upgrade, but 
does not show in the attached log (.txt file).

The upgrade resulted in a third Fedora entry in the grub menu: 
"6.13.10-200.fc41".

Upon rebooting (console login was the only way to log in), "akmods" 
still fails.

Next I tried Marco's dnf suggestion, but with "upgrade", not "update".  
It upgraded nearly 1200 packages.  This suggests to me that yesterday 
morning's system-upgrade had many more problems that I thought until 
now.  Logs of today's upgrade are huge, and full of errors and 
warnings.  I can try to send those privately to whoever asks.

My grub menu now has 4 Fedora entries:
6.13.10-200.fc41
6.13.10-100.fc40
6.13.9-200.fc41
6.13.9-100.fc40
I tried booting each twice; all failed to result in a login screen.  All 
8 boot attempts showed failures involving firewalld.service and 
tuned.service. Trying to boot the top 3 grub Fedora entries but not the 
bottom grub Fedora entry resulted in some error/failure about part1 or 
partd1 (I don't recall which) and something about "usb".

[Samuel Sieb]

On 4/11/25 11:14 AM, home user via users wrote:
> On 04/10/2025 11:57 PM, Samuel Sieb wrote:
> > Was it you or someone else that got into this same situation recently?
>
> Not sure.  In early or mid October, I upgraded from f39 to f40.  There 
> were major problems.  There was some similarity to this, but definitely 
> not identical.  A key part of the problem was the post-upgrade task to 
> clean up retired packages apparently malfunctioned big-time. You were 
> one of the main helpers on that.
> > What does "rpm -q fedora-release" show you?
>
> "fedora-release-40-39.noarch"

I don't know what you're doing, but the upgrade clearly failed again.
> Upon rebooting (console login was the only way to log in), "akmods" 
> still fails.

Your login screen isn't going to work until you get the akmods figured 
out.  What's the log from that now?

[home user]

On 04/11/2025 12:26 PM, Samuel Sieb wrote:
> On 4/11/25 11:14 AM, home user via users wrote:
> > On 04/10/2025 11:57 PM, Samuel Sieb wrote:
> > > Was it you or someone else that got into this same situation recently?
> >
> > Not sure.  In early or mid October, I upgraded from f39 to f40. There 
> > were major problems.  There was some similarity to this, but 
> > definitely not identical.  A key part of the problem was the 
> > post-upgrade task to clean up retired packages apparently 
> > malfunctioned big-time. You were one of the main helpers on that.
> > > What does "rpm -q fedora-release" show you?
> >
> > "fedora-release-40-39.noarch"
>
> I don't know what you're doing, but the upgrade clearly failed again.

I followed the instructions in 
"https://docs.fedoraproject.org/en-US/quick-docs/upgrading-fedora-offline/". 
I did likewise this past October.
> > Upon rebooting (console login was the only way to log in), "akmods" 
> > still fails.
>
> Your login screen isn't going to work until you get the akmods figured 
> out.  What's the log from that now?

2025/04/11 11:15:51 akmods: Building RPM using the command 
'/usr/sbin/akmodsbuild --kernels 6.13.9-200.fc41.x86_64 
/usr/src/akmods/nvidia-470xx-kmod.latest'
Building target platforms: x86_64
Building for target x86_64
Installing /usr/src/akmods/nvidia-470xx-kmod.latest
setting SOURCE_DATE_EPOCH=1740268800
error: Failed build dependencies:
     kernel-devel-uname-r = 6.13.9-200.fc41.x86_64 is needed by 
nvidia-470xx-kmod-3:470.256.02-7.fc40.x86_64
2025/04/11 11:15:52 akmodsbuild: * Rebuilding 
/usr/src/akmods/nvidia-470xx-kmod.latest for kernel(s) 
6.13.9-200.fc41.x86_64: Building target platforms: x86_64
2025/04/11 11:15:52 akmodsbuild: Building for target x86_64
2025/04/11 11:15:52 akmodsbuild: Installing 
/usr/src/akmods/nvidia-470xx-kmod.latest
2025/04/11 11:15:52 akmodsbuild: setting SOURCE_DATE_EPOCH=1740268800
2025/04/11 11:15:52 akmodsbuild: error: Failed build dependencies:
2025/04/11 11:15:52 akmodsbuild:     kernel-devel-uname-r = 
6.13.9-200.fc41.x86_64 is needed by 
nvidia-470xx-kmod-3:470.256.02-7.fc40.x86_64
2025/04/11 11:15:52 akmodsbuild:

2025/04/11 11:15:52 akmods: Building rpms failed; see 
/var/cache/akmods/nvidia-470xx/470.256.02-7-for-6.13.9-200.fc41.x86_64.failed.log 
for details

------

I agree with what you said, but my gut is saying that something is wrong 
at a more fundamental level.  I don't know what, or how to determine 
what.  I doubt that the akmods problem can be solved until the more 
fundamental problem is determined and solved.  I think the same was true 
with the October version upgrade.  There was some other key step that 
preceded getting the akmods thing solved.  I still wonder of some 
on-disk data (rpm database?) was damaged by the October version upgrade, 
and is still now in a damaged state.  But I don't know how to check that.

[Samuel Sieb]

On 4/11/25 12:19 PM, home user via users wrote:
> error: Failed build dependencies:
>      kernel-devel-uname-r = 6.13.9-200.fc41.x86_64 is needed by 
> nvidia-470xx-kmod-3:470.256.02-7.fc40.x86_64

Try doing the system upgrade again.  You're still on F40.

"system-upgrade" is better, but if it isn't working for some reason, you 
can also try:
dnf --releasever=41 distro-sync

[home user]

On 4/11/2025 4:09 PM, Samuel Sieb wrote:
> On 4/11/25 12:19 PM, home user via users wrote:
> > error: Failed build dependencies:
> >      kernel-devel-uname-r = 6.13.9-200.fc41.x86_64 is needed by 
> > nvidia-470xx-kmod-3:470.256.02-7.fc40.x86_64
>
> Try doing the system upgrade again.  You're still on F40.
>
> "system-upgrade" is better, but if it isn't working for some reason, you 
> can also try:
> dnf --releasever=41 distro-sync

I have every intention of trying what you suggest, but I'd like to make 
sure the workstation is ready. System upgrades take a lot of time - half 
a day.

My boot(?) partition has only enough space for 2 kernels.  (That's why I 
have no rescue!)  Yet my grub menu is listing 4 Fedora boot choices:
6.13.10-200.fc41
6.13.10-100.fc40
6.13.9-200.fc41
6.13.9-100.fc40
That in itself has me troubled.

A. What should I do before attempting the system-upgrade, to make sure 
there's enough space on the partition?

B. Yesterday's attempt at the system upgrade upgraded dnf from version 4 
to version 5.  What (if anything) do I need to do about that to maximize 
the chances of success?

[Samuel Sieb]

On 4/11/25 3:25 PM, home user via users wrote:
> On 4/11/2025 4:09 PM, Samuel Sieb wrote:
> > On 4/11/25 12:19 PM, home user via users wrote:
> > > error: Failed build dependencies:
> > >      kernel-devel-uname-r = 6.13.9-200.fc41.x86_64 is needed by 
> > > nvidia-470xx-kmod-3:470.256.02-7.fc40.x86_64
> >
> > Try doing the system upgrade again.  You're still on F40.
> >
> > "system-upgrade" is better, but if it isn't working for some reason, 
> > you can also try:
> > dnf --releasever=41 distro-sync
>
> I have every intention of trying what you suggest, but I'd like to make 
> sure the workstation is ready. System upgrades take a lot of time - half 
> a day.
>
> My boot(?) partition has only enough space for 2 kernels.  (That's why I 
> have no rescue!)  Yet my grub menu is listing 4 Fedora boot choices:
> 6.13.10-200.fc41
> 6.13.10-100.fc40
> 6.13.9-200.fc41
> 6.13.9-100.fc40
> That in itself has me troubled.

What does "rpm -q kernel-core" show?
> A. What should I do before attempting the system-upgrade, to make sure 
> there's enough space on the partition?

df -h /boot
But dnf will tell you if there isn't enough room.
> B. Yesterday's attempt at the system upgrade upgraded dnf from version 4 
> to version 5.  What (if anything) do I need to do about that to maximize 
> the chances of success?

Nothing.  dnf5 is fine.

[WILLIAM MATTISON]

Replying from old iPad…

All 4 grub options…
Console login results in immediate return to console login.  That happens regardless of which user I log in as, including root and the admin account.

Can no longer boot windows-7.  Shows error message for < 1 sec and then bios comes up, followed by grub menu.

Local friend, retired IT professional hopefully comes tomorrow with cell phone and fedora live usb stick.  Maybe he can help.

Something done this morning did damage to both Fedora and windows.

For now, all I have is this old iPad.  It’s slow, tedious, and hard to use.  I have no way of getting the information requested or trying what was suggested.

I saw in Fedora Magazine that there’s a new Anaconda that handles dual boot well.  Hopefully the stick has that and it helps.
> On 04/11/2025 4:29 PM MDT Samuel Sieb  wrote:
>
> On 4/11/25 3:25 PM, home user via users wrote:
> > On 4/11/2025 4:09 PM, Samuel Sieb wrote:
> > > On 4/11/25 12:19 PM, home user via users wrote:
> > > > error: Failed build dependencies:
> > > >      kernel-devel-uname-r = 6.13.9-200.fc41.x86_64 is needed by 
> > > > nvidia-470xx-kmod-3:470.256.02-7.fc40.x86_64
> > >
> > > Try doing the system upgrade again.  You're still on F40.
> > >
> > > "system-upgrade" is better, but if it isn't working for some reason, 
> > > you can also try:
> > > dnf --releasever=41 distro-sync
> >
> > I have every intention of trying what you suggest, but I'd like to make 
> > sure the workstation is ready. System upgrades take a lot of time - half 
> > a day.
> >
> > My boot(?) partition has only enough space for 2 kernels.  (That's why I 
> > have no rescue!)  Yet my grub menu is listing 4 Fedora boot choices:
> > 6.13.10-200.fc41
> > 6.13.10-100.fc40
> > 6.13.9-200.fc41
> > 6.13.9-100.fc40
> > That in itself has me troubled.
>
> What does "rpm -q kernel-core" show?
> > A. What should I do before attempting the system-upgrade, to make sure 
> > there's enough space on the partition?
>
> df -h /boot
> But dnf will tell you if there isn't enough room.
> > B. Yesterday's attempt at the system upgrade upgraded dnf from version 4 
> > to version 5.  What (if anything) do I need to do about that to maximize 
> > the chances of success?
>
> Nothing.  dnf5 is fine.
>
> -- 
> _______________________________________________
> users mailing list -- [e-mail]
> To unsubscribe send an email to [e-mail]
> Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/
> List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
> List Archives: https://lists.fedoraproject.org/archives/list/[e-mail]
> Do not reply to spam, report it: https://pagure.io/fedora-infrastructure/new_issue

[Jeffrey Walton]

On Fri, Apr 11, 2025 at 1:57 AM Samuel Sieb  wrote:
> On 4/10/25 6:58 PM, home user via users wrote:
> > So I tried "dnf upgrade kernel-devel"...
> >
> > Package       Arch   Version         Repository      Size
> > Installing:
> >   kernel-devel x86_64 6.13.9-100.fc40 updates     77.7 MiB
> >
> > Transaction Summary:
> >   Installing:         1 package
> >
> > It upgraded, but I need fc41, not fc40.
> >
> > "uname -a" did report that I was using fc41.
>
> Was it you or someone else that got into this same situation recently?

I was the someone else. I had a failed F41 upgrade on an old Proliant
G5. It was the first failed upgrade I have personally experienced.

The hint about Python or Perl symlinks was correct. But after changing
symlinks the machine had other problems. So I did a fresh install on
new hardware, and restored the webserver files from backup.
> What does "rpm -q fedora-release" show you?
>
> You can try "dnf upgrade --releasever=41 kernel-devel" for now.

Jeff

[Barry Scott]

> On 10 Apr 2025, at 19:21, home user via users  wrote:
>
> I tried to upgrade from Fedora-40 (patched this morning) to Fedora-41.  I followed the instructions in the Fedora web site ".../upgrading-fedora-offline" document.  After the "dnf system-upgrade reboot" step, the boot process never resulted in a login screen.  I tried several times, including recycling the power switch on the back of the tower.

Sounds like you have a GPU that does not have working drivers, often the case with nvidia.

Did you try typing Ctrl-Alt-F3 to get to a console? Does that work?
If so you can start to diagnose what's wrong.
> Trying to boot Fedora-40 also does not result in a login screen.  I tried that a few times.

F40 is gone - all that is left is the kernels that you installed for f40.
> This is a stand-alone dual boot workstation, the other OS being windows-7, which I am using now to send out this message.

This is old hardware I assume.
I wonder if you have upgraded to use Wayland desktop but the GPU drivers only support X11?
> My windows-7 install does have the tool "ext2explore".  That allows me to copy files from the Fedora partitions into windows-7, where I can view those files with w7's standard tools.
>
> New hardware is out of the question until early summer.
>
> I've never been able to make a live USB and then successfully boot from it; I don't know why.

That makes recovery a lot harder...

It would be worth your effort to be able to boot USB live images so you can recover from this type of situation.

Barry
> I do not have a rescue.
>
> Help!!!
>
> -- 
> _______________________________________________
> users mailing list -- [e-mail]
> To unsubscribe send an email to [e-mail]
> Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/
> List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
> List Archives: https://lists.fedoraproject.org/archives/list/[e-mail]
> Do not reply to spam, report it: https://pagure.io/fedora-infrastructure/new_issue

[WILLIAM MATTISON]

Good afternoon,
Resuming from yesterday afternoon...
This afternoon, a friend brought me a F41 live USB.  I am able to boot it.  Firefox works, but seems prone to crashing.  I tried to do the Fedora install from the live USB.  Both I and my friend were very soon lost, and did not know how to proceed.  What I found on the Fedora website was not much help for my situation.

* This is a dual-boot stand-alone workstation: windows-7 + Fedora.
* As of yesterday afternoon, I can no longer log in to either OS.
* I'd like to preserve the windows-7 part of the hard drive.
* I want Fedora-41 workstation installed over the current non-functioning Fedora 40/41 workstation.
* I don't know how to get numbers, but from past experience I know that the boot(?) partition needs to be bigger.  In the past, I had to cut back to current kernel + one old kernel, no rescue.  I also have memtest[whatever] (it's in the grub menu).  I'd like to have current kernel + 3 old kernels + rescue + memtest + windows-7.
* I'd sure like to keep my /home data, but I realize that might not be possible.

You can see what my hardware is here:
"https://linux-hardware.org/?probe=7fa74ee631".

A few of you have been wanting me to re-install.  I seem to have no other option.  I seriously need help with this re-install.

Thank-you in advance.
~

[Jeffrey Walton]

On Sat, Apr 12, 2025 at 7:40 PM WILLIAM MATTISON via users
 wrote:
> Good afternoon,
> Resuming from yesterday afternoon...
> This afternoon, a friend brought me a F41 live USB.  I am able to boot it.  Firefox works, but seems prone to crashing.  I tried to do the Fedora install from the live USB.  Both I and my friend were very soon lost, and did not know how to proceed.  What I found on the Fedora website was not much help for my situation.
> This is a dual-boot stand-alone workstation: windows-7 + Fedora.
> As of yesterday afternoon, I can no longer log in to either OS.
> I'd like to preserve the windows-7 part of the hard drive.
> I want Fedora-41 workstation installed over the current non-functioning Fedora 40/41 workstation.
> I don't know how to get numbers, but from past experience I know that the boot(?) partition needs to be bigger.  In the past, I had to cut back to current kernel + one old kernel, no rescue.  I also have memtest[whatever] (it's in the grub menu).  I'd like to have current kernel + 3 old kernels + rescue + memtest + windows-7.
> I'd sure like to keep my /home data, but I realize that might not be possible.
>
> You can see what my hardware is here:
> "https://linux-hardware.org/?probe=7fa74ee631".
>
> A few of you have been wanting me to re-install.  I seem to have no other option.  I seriously need help with this re-install.

I'm going to take a stab in the dark... I'm guessing the UEFI is not
allowing you to boot to either OS because it does not like either
bootloader (both the WIndows and Linux boot loaders).

Boot to a Live ISO, and then update the Secure Boot DBX database
(Forbidden Signature Database). Use fwupdmgr to update it.

If you still cannot boot after the updates, then disable Secure Boot.

Jeff