Firefox and kde wallet on login

[einar.hjortdal]

Hi, I’ve had my girlfriend move to Openmamba for her office computer (given that Windows 10 died off and her system is too old for Windows 11 (X370 and Ryzen 1800X too old smh)).
She is using firefox as the browser right now, and when firefox starts it requests her to enter her user password to be able to use wifi.
She finds this prompt annoying, I also find it annoying.
Is there any way to automatically unlock the wallet on login for firefox?

Attachments:

You must be logged in to view attached files.

[Silvan]

Hi,
this would normally happen because the user has set a password for KDE Wallet. You may want to try to open KWalletManager and change the password by setting an empty one.

[einar.hjortdal]

I don’t think I’ve ever set a password for the KDE wallet, the password is the same as the user login password. Perhaps it is automatically set to that?

[Silvan]

The report you sent (related to another problem) shows a crash of the ksecretd daemon. I don’t know if it is the same installation of this post, in any case it might be worth to see a useful stacktrace of the crash dump. As an example the following console commands might produce the stacktrace as output:

sudo dnf install kf6-kwallet-debug
coredumpctl info -q -r

[einar.hjortdal]

Silvan wrote:

I don’t know if it is the same installation of this post, in any case it might be worth to see a useful stacktrace of the crash dump.

The kwallet behavior I described happens on both this system, with Microsoft Edge, and on my girlfriend’s system with Firefox.

coredumpctl info -q -r
           PID: 6457 (ksecretd)
           UID: 1000 (einar)
           GID: 1001 (einar)
        Signal: 11 (SEGV)
     Timestamp: Tue 2025-11-04 18:39:48 CET (14h ago)
  Command Line: /usr/bin/ksecretd
    Executable: /usr/bin/ksecretd
 Control Group: /user.slice/user-1000.slice/user@1000.service/session.slice/dbus.service
          Unit: user@1000.service
     User Unit: dbus.service
         Slice: user-1000.slice
     Owner UID: 1000 (einar)
       Boot ID: 47ed0ddcb97a447385b37b7837646ae4
    Machine ID: 499600159b1242a489ec6afef12cbf3c
      Hostname: ws00
       Storage: /var/lib/systemd/coredump/core.ksecretd.1000.47ed0ddcb97a447385b37b7837646ae4.6457.1762277988000000.zst (present)
  Size on Disk: 3.1M
       Message: Process 6457 (ksecretd) of user 1000 dumped core.
                
                Stack trace of thread 6457:
                #0  0x00007f38c789e95c n/a (n/a + 0x0)
                #1  0x0000000000000000 n/a (n/a + 0x0)
                ELF object binary architecture: AMD x86-64

Do you want the 3.1M file in /var/lib/systemd/coredump/core.ksecretd.1000.47ed0ddcb97a447385b37b7837646ae4.6457.1762277988000000.zst?
That’s a massive stack trace
I have uploaded it here

• This reply was modified 2 weeks, 4 days ago by einar.hjortdal.
• This reply was modified 2 weeks, 4 days ago by einar.hjortdal.
• This reply was modified 2 weeks, 4 days ago by einar.hjortdal.
• This reply was modified 2 weeks, 4 days ago by einar.hjortdal.

[Silvan]

I cannot get a useful stacktrace.
This and this are reports to a similar problem.
I assume that no system changes have been made and the package kwallet-pam is installed.
If it is a general openmamba problem it should be reproduced by me or reported by other users when updating existing installations or doing a fresh installation but the fact that this is happening only to you and on two installations suggests to investigate on any post installation operation you have made on both systems, for example the installation of any external software might have changed the /etc/pam.d files and as written above it seems that the problem is at PAM level.

[einar.hjortdal]

I believe kwallet-pam is installed by default.
The system I am using right now has some non-default software installed, like nvidia_470, though that shouldn’t touch kwallet.
However, my girlfriend’s secondary system is much simpler: it’s an intel-based laptop with no discrete graphics and no additional software installed.
The issue appeared immediately after install too, before any updates, and persisted after updates.
Should I provide a stack trace using the same method on her laptop? I can do that later today. Let me know if there is more I can provide so I only take over her workstation once today

• This reply was modified 2 weeks, 4 days ago by einar.hjortdal.

[Silvan]

I will do an installation test with the latest livecd ISO image. Did you set automatic login on your installations?

[einar.hjortdal]

No automatic login. Disabled root user login.

[Silvan]

I’ve tried today’s livecd, installed the system on VirtualBox, rebooted, logged in, run firefox and did not get any kde wallet password request. I also tried logging in with plasma-x11 and did not reproduce the problem either.

[einar.hjortdal]

Should I attempt to reinstall?

[einar.hjortdal]

It happens on all the systems I installed openmamba on though 🤔

[einar.hjortdal]

I have reinstalled openmamba using the openmamba-livecd-en-snapshot-20251107.x86_64.iso image and as soon as I opened firefox (for the first time after logging in) I was given the kde wallet password request.
It happens on both Wayland and x11, I use x11 on this one system in particular because of the nvidia_470 package that is not supported by wayland (the nvidia package has its own problems recently according to the reports I’ve submitted).

• This reply was modified 2 weeks ago by einar.hjortdal.

[Silvan]

While the nvidia-470 installation may have a history of its own, I suppose the Ryzen PC is running with standard drivers and you are reporting the kwallet problem as present on both. By the way it seems you only sent reports of the nvidia-470 PC. A report from the Ryzen PC might be useful to try to make a diagnosis of the problem.

[Silvan]

The new received report is about a Intel Core 3 CPU based hardware. The log shows a series of crashed which would suggest investigation:

Nov 10 08:52:26 lptp drkonqi-coredump-processor[5184]: "/usr/bin/gpg-agent" 6821 ""
Nov 10 08:52:26 lptp drkonqi-coredump-processor[5184]: "/usr/bin/gpg-agent" 6922 ""
Nov 10 08:52:26 lptp drkonqi-coredump-processor[5184]: "/usr/bin/gpg-agent" 7134 ""
Nov 10 08:52:26 lptp drkonqi-coredump-processor[5184]: "/usr/bin/kwalletmanager5" 9204 "/var/lib/systemd/coredump/core.kwalletmanager5.1000.2c0861e734744eaca96288c02f4e69c1.9204.1756379708000000.zst"
Nov 10 08:52:26 lptp drkonqi-coredump-processor[5184]: "/usr/libexec64/drkonqi" 9252 "/var/lib/systemd/coredump/core.drkonqi.1000.2c0861e734744eaca96288c02f4e69c1.9252.1756379709000000.zst"
Nov 10 08:52:26 lptp drkonqi-coredump-processor[5184]: "/usr/libexec64/drkonqi" 9251 "/var/lib/systemd/coredump/core.drkonqi.1000.2c0861e734744eaca96288c02f4e69c1.9251.1756379709000000.zst"

A useful backtrace for these coredumps would be necessary. Unfortunately
– the packages installation log shows that the last package was updated on October 20, in order to try to install the needed debug packages they should be aligned with current repositories, so the system should be updated first
– drkonqi itself is crashing so no crash helper would be available on the desktop

I suggest to update the system, try to install some debug package like:
sudo dnf install gnupg2-debug kwalletmanager-debug qt6-qtbase-debug

reboot, reproduce the problem and generate a new report. By having the system up-do-date the more recent coredumps can be shared in order to try to obtain a useful stack trace.
The more useful would be the one related to gpg-agent as is it probably the cause of the subsequent kwalletmanager5 crash, but first it has to be confirmed that after the system upgrade it gets these same processes to crash.

• This reply was modified 1 week, 5 days ago by Silvan.
• This reply was modified 1 week, 5 days ago by Silvan.

[einar.hjortdal]

That will take me a little while because her systems are at her office and I work at another location, but I will do that and send you a new report.

[Author]

Posts