TrueNAS 25.10 - Sporadic macOS Tahoe Kernel Panics in "com.apple.filesystems.smbfs" driver

Hi TrueNAS Community,

I am reaching out to you in hope that someone else might have experienced a similar bug in their setup. I have now been hunting this issue since a couple weeks and originally, I even believed that my MacBook has a hardware issue (e.g. my device could also be a dud). However, recently I found some hints that the issues might be related to my specific TrueNAS set-up. I have also spent a good amount of time searching the web whether anyone else experiences the same issue but I did not find anything helpful so far. My issue is as follows:

When I am connected to my home network, I am experiencing sporadic kernel panics on my MacBook Air M3 15” with macOS Tahoe (currently I am on macOS 26.6.2). My TrueNAS system is on version 25.10.7 - Goldeye. The kernel panics are almost exclusively related to the “com.apple.filesystems.smbfs” driver and often show “[skywalk.reg.bufctl]: element modified after free” as immediate panic cause. They happen randomly during system usage, sometimes even while the device is in hibernation.

I have set-up a timemachine SMB share on TrueNAS and also configured timemachine on the Macbook - this works so far. Below you can find the configuration of my SMB service:

On TrueNAS I have only SMB enabled, no other fileshare service is active. I think it is also worth to note that my macOS username is identical to my TrueNAS SMB user - could this be an issue?

Today I could finally reproduce the issue on my MacBook with the help of Gemini: When I start a timemachine backup and concurrently scan file permissions on the NAS share, the Macbook panics after a couple minutes.

# Start a time machine backup
$ tmutil startbackup
# Scan file permissions on the NAS
$ sudo find "/Volumes/Backups of Macbook" -exec xattr -l {} + 2>/dev/null

I then created a nsmb.conf file and so far I can no longer reproduce the kernel panic:

$ cat << EOF | sudo tee /etc/nsmb.conf
[default]
mc_on=no
signing_required=no
dir_cache_off=yes
EOF

I think it is to early to state that this really fixed the issue, I will need to test this a little longer. However, even if it fixes the issue, it is not a nice fix since it is client side only. This is why I am reaching out to you, maybe I have missed something or someone else might know immediately how to fix this on TrueNAS side.

Thank you very much,
Michael

To clarify, it’s MacOS that’s having kernel panic? If so, then you should file a bug report with Apple as that’s a serious issue.

Yes macOS is having the kernel panic. What held me back from creating a bug report at Apple yet is that it’s hard for me to believe that I am the only person experiencing this issue and that I am still not 100% confident that my hardware is not defective. If the settings in nsmb.conf turn on out to permanently fix the problem, I will file a bug report. In addition, it is quite hard to consistently reproduce the issue since it currently looks like some race condition in the kernel.

Update - the kernel panics still appear even with the nsmb.conf settings. I think my MacBook is just a bad sample…

Maybe.
Still, report it to Apple in case it’s not.