Bug #16248
open
QLink/Marvell 41000 NIC bug
0%
Description
I'm using pfSense 2.8.0 and I have a QL41164 NIC. Driver isn't enabled by default, after adding if_qlnxe_load="YES" under /boot/loader.conf as per documentation and then rebooting the NIC finally shows in pfSense. When I switch the WAN port from an igb0 NIC to the ql0 NIC I cant ping the interface IP. Doing a reboot of pfSense sometimes causes a crash dump, sometimes the interface wont be pingable, and sometimes when it is pingable, the web GUI fails to load and gives a crash report.
rash report begins. Anonymous machine information:
amd64
15.0-CURRENT
FreeBSD 15.0-CURRENT #1 RELENG_2_8_0-n256081-401ec5f685b9: Wed May 21 23:53:51 UTC 2025 root@freebsd:/var/jenkins/workspace/pfSense-CE-snapshots-2_8_0-main/obj/amd64/0q9vjGjc/var/jenkins/workspace/pfSense-CE-snapshots-2_8_0-main/sources/FreeBSD-src-RE
Crash report details:
No PHP errors found.
No FreeBSD crash data found.
I installed FreeBSD-15.0-CURRENT-amd64-20250606-fa02d9fceab7-277727-memstick.img on the same machine and it has no issues with the QL41164 NIC after activation. I get full 9.4Gbps using iperf and no crash dumps. I seems like you are missing an updated QLNXE NIC driver in the 2.8.0 version
Updated by Steve Wheeler about 1 month ago
Do you see a kernel panic? Any sort of error log?
Updated by Adam Esslinger about 1 month ago
it appears to be a bug with the dashboard. Once the login screen times out and I get that crash dump that I pasted in the ticket, I can navigate around the menus OK, but when I try to get to the dashboard is timeout. Are there logs I can look at for the dashboard?
I did find this in the system logs:
2025/06/09 18:21:50 [error] 98156#100316: *1 upstream timed out (60: Operation timed out) while reading response header from upstream, client: 192.168.200.104, server: , request: "GET / HTTP/1.1", upstream: "fastcgi://unix:/var/run/php-fpm.socket", host: "192.168.200.102", referrer: "http://192.168.200.102/"
On the kernel panic topic, when I have pfSense 2.7 or pfSense+ 24.11-RELEASE, is where I get the kernel panic.
Another related issue, I cant get the any of the QL41164 ports to do DHCP. When the system boots up it wont have a DHCP address, if i wait a bit and press return on the console it will eventually show an IP, however I cant ping it, ssh, web anything to it. Only way the NIC is accessible is if I put a static IP on it.
Updated by Adam Esslinger about 1 month ago
Adam Esslinger wrote in #note-2:
it appears to be a bug with the dashboard. Once the login screen times out and I get that crash dump that I pasted in the ticket, I can navigate around the menus OK, but when I try to get to the dashboard is timeout. Are there logs I can look at for the dashboard?
I did find this in the system logs:
2025/06/09 18:21:50 [error] 98156#100316: *1 upstream timed out (60: Operation timed out) while reading response header from upstream, client: 192.168.200.104, server: , request: "GET / HTTP/1.1", upstream: "fastcgi://unix:/var/run/php-fpm.socket", host: "192.168.200.102", referrer: "http://192.168.200.102/"
On the kernel panic topic, when I have pfSense 2.7 or pfSense+ 24.11-RELEASE, is where I get the kernel panic.
Another related issue, I cant get the any of the QL41164 ports to do DHCP. When the system boots up it wont have a DHCP address, if i wait a bit and press return on the console it will eventually show an IP, however I cant ping it, ssh, web anything to it. Only way the NIC is accessible is if I put a static IP on it.
I actually believe this is a bug in FreeBSD. I have a ticket opened with them and been working on it.
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=287445
Updated by Steve Wheeler about 1 month ago
We may have hit this while working on the other qlnxe bug.
If you run:
ifconfig -v ql0
It fails to return when trying, what appears to be, to read the SFP module data.
The dashboard interfaces widget uses that command and since it's present by default it appears to hang the dashboard.
After attempting that several times I also see logged:
ql1: [ecore_dbg_print_attn:6458(qlnx-0)]interrupt: address 0x00001d00, status 0x00000001, mask 0x00000000 ql1: [ecore_int_deassertion_aeu_bit:938(qlnx-0)]`Vaux PCI core': Fatal attention ql3: [ecore_dbg_print_attn:6458(qlnx-0)]interrupt: address 0x00001d00, status 0x00000001, mask 0x00000000 ql3: [ecore_int_deassertion_aeu_bit:938(qlnx-0)]`Vaux PCI core': Fatal attention ql3: [ecore_dbg_print_attn:6458(qlnx-0)]interrupt: address 0x00001d00, status 0x00000001, mask 0x00000000
Updated by Steve Wheeler 2 days ago
- Status changed from New to Waiting on Merge
- Target version set to 2.8.1
- Plus Target Version set to 25.11
This looks to be fixed by this: https://reviews.freebsd.org/D51351
Local test builds look good.
Updated by Jim Pingle about 21 hours ago
- Target version changed from 2.8.1 to 2.9.0