[Firmware Bug]: Overriding NUMA node to 0. Contact your vendor for updates

GPU Unix Graphics Linux
#1

Hi

I’m getting the following on dmesg every time I boot the machine:

[ 17.999692] pci 0000:00:00.0: [Firmware Bug]: Overriding NUMA node to 0. Contact your vendor for updates.
[ 18.000767] pci 0000:00:00.2: [Firmware Bug]: Overriding NUMA node to 0. Contact your vendor for updates.
[ 18.001735] pci 0000:00:01.0: [Firmware Bug]: Overriding NUMA node to 0. Contact your vendor for updates.
[ 18.002819] pcieport 0000:00:01.3: [Firmware Bug]: Overriding NUMA node to 0. Contact your vendor for updates.
[ 18.005214] pci 0000:00:02.0: [Firmware Bug]: Overriding NUMA node to 0. Contact your vendor for updates.
[ 18.006415] pci 0000:00:03.0: [Firmware Bug]: Overriding NUMA node to 0. Contact your vendor for updates.
[ 18.008649] pcieport 0000:00:03.1: [Firmware Bug]: Overriding NUMA node to 0. Contact your vendor for updates.
[ 18.010474] pci 0000:00:04.0: [Firmware Bug]: Overriding NUMA node to 0. Contact your vendor for updates.
[ 18.011412] pci 0000:00:07.0: [Firmware Bug]: Overriding NUMA node to 0. Contact your vendor for updates.
[ 18.012531] pcieport 0000:00:07.1: [Firmware Bug]: Overriding NUMA node to 0. Contact your vendor for updates.
[ 18.013598] pci 0000:00:08.0: [Firmware Bug]: Overriding NUMA node to 0. Contact your vendor for updates.
[ 18.014728] pcieport 0000:00:08.1: [Firmware Bug]: Overriding NUMA node to 0. Contact your vendor for updates.
[ 18.015776] piix4_smbus 0000:00:14.0: [Firmware Bug]: Overriding NUMA node to 0. Contact your vendor for updates.
[ 18.016738] pci 0000:00:14.3: [Firmware Bug]: Overriding NUMA node to 0. Contact your vendor for updates.
[ 18.017720] pci 0000:00:18.0: [Firmware Bug]: Overriding NUMA node to 0. Contact your vendor for updates.
[ 18.018811] pci 0000:00:18.1: [Firmware Bug]: Overriding NUMA node to 0. Contact your vendor for updates.
[ 18.019669] pci 0000:00:18.2: [Firmware Bug]: Overriding NUMA node to 0. Contact your vendor for updates.
[ 18.020536] k10temp 0000:00:18.3: [Firmware Bug]: Overriding NUMA node to 0. Contact your vendor for updates.
[ 18.021496] pci 0000:00:18.4: [Firmware Bug]: Overriding NUMA node to 0. Contact your vendor for updates.
[ 18.022377] pci 0000:00:18.5: [Firmware Bug]: Overriding NUMA node to 0. Contact your vendor for updates.
[ 18.023489] pci 0000:00:18.6: [Firmware Bug]: Overriding NUMA node to 0. Contact your vendor for updates.
[ 18.024396] pci 0000:00:18.7: [Firmware Bug]: Overriding NUMA node to 0. Contact your vendor for updates.
[ 18.025310] xhci_hcd 0000:01:00.0: [Firmware Bug]: Overriding NUMA node to 0. Contact your vendor for updates.
[ 18.026528] ahci 0000:01:00.1: [Firmware Bug]: Overriding NUMA node to 0. Contact your vendor for updates.
[ 18.027698] pcieport 0000:01:00.2: [Firmware Bug]: Overriding NUMA node to 0. Contact your vendor for updates.
[ 18.028610] pcieport 0000:02:00.0: [Firmware Bug]: Overriding NUMA node to 0. Contact your vendor for updates.
[ 18.030812] pcieport 0000:02:01.0: [Firmware Bug]: Overriding NUMA node to 0. Contact your vendor for updates.
[ 18.032081] pcieport 0000:02:04.0: [Firmware Bug]: Overriding NUMA node to 0. Contact your vendor for updates.
[ 18.032980] r8169 0000:03:00.0: [Firmware Bug]: Overriding NUMA node to 0. Contact your vendor for updates.
[ 18.034327] pci 0000:04:00.0: [Firmware Bug]: Overriding NUMA node to 0. Contact your vendor for updates.
[ 18.035157] nvidia 0000:07:00.0: [Firmware Bug]: Overriding NUMA node to 0. Contact your vendor for updates.
[ 18.035923] snd_hda_intel 0000:07:00.1: [Firmware Bug]: Overriding NUMA node to 0. Contact your vendor for updates.
[ 18.036852] pci 0000:08:00.0: [Firmware Bug]: Overriding NUMA node to 0. Contact your vendor for updates.
[ 18.038146] ccp 0000:08:00.2: [Firmware Bug]: Overriding NUMA node to 0. Contact your vendor for updates.
[ 18.039700] xhci_hcd 0000:08:00.3: [Firmware Bug]: Overriding NUMA node to 0. Contact your vendor for updates.
[ 18.043737] pci 0000:09:00.0: [Firmware Bug]: Overriding NUMA node to 0. Contact your vendor for updates.
[ 18.044863] ahci 0000:09:00.2: [Firmware Bug]: Overriding NUMA node to 0. Contact your vendor for updates.
[ 18.045760] snd_hda_intel 0000:09:00.3: [Firmware Bug]: Overriding NUMA node to 0. Contact your vendor for updates.
[ 18.855593] resource sanity check: requesting [mem 0x000c0000-0x000fffff], which spans more than PCI Bus 0000:00 [mem 0x000c0000-0x000dffff window]
[ 18.855759] caller _nv000706rm+0x1af/0x200 [nvidia] mapping multiple BARs

The machine locks up with no trace of anything after 0-x days. The motherboard vendor believes the problem is with the GPU.

Mvh.

Torkilnvidia-bug-report.log.gz (917.9 KB)

#2

Only problem I can see is a 1st gen Ryzen cpu, did you check if it affected by the ryzen-bug? The nvidia driver is prone to trigger it.

#3

I have some c-state workarounds for the kernel but I wasn’t aware of a defect beyond that, thanks. Would you happen to know of a sure fire way to determine if my CPU is affected by the manufacturing bug?

Mvh.

Torkil

#4

You’re a little bit late, most tools to check for the bug (to have it replaced by amd) were built at the time the cpu was released. Like this:
https://github.com/suaefar/ryzen-test
So it’s quite difficult to find one for current distributions.

#5

I ran the test from a live 17.04 stick and it segfaulted and locked up in less than 200s so I guess that’s probably it.

Thanks,

Torkil

#6

I replaced the CPU with a Ryzen 3600 but I’m still getting the same NUMA errors when I boot the machine. Any other suggestions?

Thanks,

Torkil

#7

If it doesn’t freeze, ignore it or try numa=off kernel parameter to disable it.