F42 KDE: Dolphin crashes browsing SMB shares

Did a fresh Fedora 42 KDE Plasma install.

Brief description of system: Old but still kickin’:

–ASRock Z77 Extreme4 motherboard
–Intel Ivy Bridge CPU (i5)

–16 GB RAM

–multiple hard drives and SSD’s

–Intel AX210 add-in card (6 GHz WiFi works, which is really cool and a reason I am exploring Fedora 42 as a replacement for Windows 10)

–Nvidia GTX 970 GPU

–a DVD-RW drive which powers up but doesn’t work, but I don’t use it anyway.

“Monitor” is a 65" Samsung 4k TV, connected via GPU HDMI port.

Immediately after install, every time I tried to access an SMB share on the local network (Windows-based NAS, basically) through Dolphin, I got as far as entering login credentials, then Dolphin crashed.

OK…

Installed Nvidia drivers (570.144). Same result.

Added option to boot into X11 desktop instead of Wayland, because I didn’t get the full Nvidia control panel under Wayland (a known bug apparently). Using X11/xorg desktop, nvidia-settings app works fully (Yay!) but Dolphin still crashes when trying to access an SMB share (boo!).

SMB access does kinda work in this way however: Zero problems accessing SMB shares and playing media over the network via Kodi.

I don’t know where to go from here. Is there a less-buggy, alternate file system explorer I can use? Is there a way to make Dolphin behave?

Additional note: I had exactly the same problem with a Fedora 42 Workstation install.

What does coredumpctl show you about the crash?
See man coredumpctl if you are not familar with that tool.

With that info you will likely need to report as a bug.
See How to file a bug :: Fedora Docs

1 Like

Here’s some more info.

If I launch Dolphin from a terminal, I see the following message after it crashes (when trying to access an SMB share):

kf.kio.core: Internal error: itemsInUse did not contain QUrl("smb://SMBusername@windowsmachine.local/")
KCrash: Application 'dolphin' crashing... crashRecursionCounter = 2
Segmentation fault (core dumped)

Note that “windowsmachine.local” refers to the LAN SMB share I was attempting to access (which is running on a Windows machine). similarly, “SMBusername” is the account I created on that machine for file sharing.

As for coredumpctl, I was able to find this detailed info regarding the last Dolphin crash (using “coredumpctl info” option):

PID: 3083 (dolphin)
           UID: 1000 (my_username)
           GID: 1000 (my_username)
        Signal: 11 (SEGV)
     Timestamp: Wed 2025-05-14 14:03:13 MDT (1min 49s ago)
  Command Line: /usr/bin/dolphin
    Executable: /usr/bin/dolphin
 Control Group: /user.slice/user-1000.slice/user@1000.service/app.slice/app-\x2fusr\x2fbin\x2>
          Unit: user@1000.service
     User Unit: app-\x2fusr\x2fbin\x2fdolphin@bfa2cf479ea74af1bb55db7166a0b6c6.service
         Slice: user-1000.slice
     Owner UID: 1000 (my_username)
       Boot ID: e32234a73dfa42cba9fd8b8004d5682a
    Machine ID: adb71d6e039a4c1ea08558399c37953b
      Hostname: HostnameRedacted
       Storage: /var/lib/systemd/coredump/core.dolphin.1000.e32234a73dfa42cba9fd8b8004d5682a.>
  Size on Disk: 5.1M
       Package: dolphin/25.04.0-3.fc42
      build-id: e4e62544c4356265d8a02ec8df2f71d562d221b6
       Message: Process 3083 (dolphin) of user 1000 dumped core.
                
                Module /usr/bin/dolphin from rpm dolphin-25.04.0-3.fc42.x86_64
                Module kuriikwsfilter.so from rpm kf6-kio-6.13.0-2.fc42.x86_64
                Module fixhosturifilter.so from rpm kf6-kio-6.13.0-2.fc42.x86_64
                Module localdomainurifilter.so from rpm kf6-kio-6.13.0-2.fc42.x86_64
                Module libkuriikwsfiltereng_private.so.6 from rpm kf6-kio-6.13.0-2.fc42.x86_64
                Module kurisearchfilter.so from rpm kf6-kio-6.13.0-2.fc42.x86_64

Further note: Same crash happens if I manually enter SMB host info in the address bar of Dolphin (e.g. “smb://192.168.x.y” and then enter login info).

Does this help?

You need to get the stack trace as well as the info above.
Try the coredumpctl dump command.

Once you have that information you can report the problem to the KDE developers in their bug tracker.

I have the exact same behavior, but a different error on a fresh install of Fedora 42.

From the shell when it crashes:

kellic@prometheus:~$ kf.kio.core: Connection::send() called with connection not inited
kf.kio.core: An error occurred during write. The worker terminates now.

-I was just on Fedora 41, latest updates, no issues getting to the share.
-I open dolphin.
-Go to smb://xxx.xxx.xxx.xxx, it prompts for creds, I submit them, and it crashes.

Currently working on reporting it in the problem reporting tool. It’s working on a backtrace right now. Just figured I’d post it to let people know this may not be a one off issue.

1 Like
Jun 30 12:58:28 somepc.domain systemd[1923]: Started app-org.kde.dolphin@0cc398070805475181558a03aa15dc4e.service - Dolphin - File Manager.
Jun 30 12:58:30 somepc.domain systemd[1923]: Started dbus-:1.2-com.google.code.AccountsSSO.SingleSignOn@4.service.
Jun 30 12:58:35 somepc.domain dolphin[21096]: org.kde.dolphin: could not find entry for charset= "Other encoding ()"
Jun 30 12:58:35 somepc.domain dolphin[21096]: kf.kio.core: Internal error: itemsInUse did not contain QUrl("smb://user@truenas.local/")
Jun 30 12:58:35 somepc.domain dolphin[21096]: KCrash: Application 'dolphin' crashing... crashRecursionCounter = 2
Jun 30 12:58:35 somepc.domain audit[21096]: ANOM_ABEND auid=1000 uid=1000 gid=1000 ses=3 subj=unconfined_u:unconfined_r:unconfined_t:s0-s0:c0.c1023 pid=21096 comm="dolphin" exe="/usr/>
Jun 30 12:58:35 somepc.domain systemd-coredump[21242]: Process 21096 (dolphin) of user 1000 terminated abnormally with signal 11/SEGV, processing...
Jun 30 12:58:35 somepc.domain audit: BPF prog-id=212 op=LOAD
Jun 30 12:58:35 somepc.domain audit: BPF prog-id=213 op=LOAD
Jun 30 12:58:35 somepc.domain audit: BPF prog-id=214 op=LOAD
Jun 30 12:58:35 somepc.domain systemd[1]: Started systemd-coredump@4-21242-0.service - Process Core Dump (PID 21242/UID 0).
Jun 30 12:58:35 somepc.domain audit[1]: SERVICE_START pid=1 uid=0 auid=4294967295 ses=4294967295 subj=system_u:system_r:init_t:s0 msg='unit=systemd-coredump@4-21242-0 comm="systemd" e>
Jun 30 12:58:35 somepc.domain audit: BPF prog-id=215 op=LOAD
Jun 30 12:58:35 somepc.domain audit: BPF prog-id=216 op=LOAD
Jun 30 12:58:35 somepc.domain audit: BPF prog-id=217 op=LOAD
Jun 30 12:58:35 somepc.domain systemd[1]: Started drkonqi-coredump-processor@4-21242-0.service - Pass systemd-coredump journal entries to relevant user for potential DrKonqi handling.
Jun 30 12:58:35 somepc.domain audit[1]: SERVICE_START pid=1 uid=0 auid=4294967295 ses=4294967295 subj=system_u:system_r:init_t:s0 msg='unit=drkonqi-coredump-processor@4-21242-0 comm=">
Jun 30 12:58:35 somepc.domain abrt-dump-journal-core[1352]: Failed to obtain all required information from journald
Jun 30 12:58:35 somepc.domain systemd-coredump[21243]: [🡕] Process 21096 (dolphin) of user 1000 dumped core.

The only way I found around this issue is my adding a drive to dolphin that points to :
smb://username@truenas.local/ . Which is strange because from the logs you can clearly see that browsing the network and clicking on the TRUENAS server available in the list which then asks for my username and password resolves to the same thing yet Dolphin crashes.

This issue came in after the recent updates, Kernel 6.15 + KDE Plasma 6.4

1 Like

There’s an upstream bug that may be related: 451050 – Dolphin crashing when connecting SMB share.

(The similarity to your logs is more obvious in this ticket, which was marked as a duplicate: 503172 – Dolphin crashes when accessing an SMB share)

The KDE developer proposing a fix says:

Due to QTBUG-35921 we sometimes get a trailing slash even if we expect it to be removed.
This patch adds a method that makes sure the trailing slash is completely removed from the URL and returns the cleaned URL, but only if the file is in remote directoy.
The code was built around the behavior of the trailing slash always being removed, so it was comparing URLs:
Even if the URLs point to same location, smb://aaa and smb://aaa/, the URLs are considered as different ones.

So a fix is on its way, though it’s hard to tell now when it will make its way into a KDE release and into Fedora.

Good to know a fix is on its way. Thanks for letting us know.

Thanks! This worked for me as a temp fix.