WTR Pro CPU throttling

well everything I tested failed. Grub updates to force iommu makes passthrough conflict with nvme. Blacklick ahci breaks nvme disk detection, driverctl to set driver override breaks also the nvme access....guess it will finidh in truenas baremetal even if it's too powerful.
What's strange is that everything looks fine regarding IOMMU groups but it doesn't work
 
À new model of aoostar nas is coming…. Aoostar wtr max, is there a way to know if it will have the same pass through issue ?
 
It seems it has something to do with the amd_pstate driver. It seems sensor communication isnt working when forwarding both sata controllers. this might also affect the amd_pstate drive (didn't check the code if it requires temperature sensors)

If you want to get the cpu to boost add the following line to /etc/default/grub
GRUB_CMDLINE_LINUX_DEFAULT="quiet amd_pstate=passive"

after saving the file dont forget to run update-grub and reboot

For an explanation what this does you can check this https://docs.kernel.org/admin-guide/pm/amd-pstate.html

There is a down side, the cpu wont go down to 400 MHz, so power consumtion will go up

Ill do some more tinkering and see If I can get it to go to low idle frequency
 
can confirm the boosting and sensors issues also happen with passing only the first SATA controller on my machine.
Pretty frustrating. the sensors i can get away with some custom fan profile, but having considerable amount of power disabled on the CPU that I pay extra money for is really annoying.
 
It seems it has something to do with the amd_pstate driver. It seems sensor communication isnt working when forwarding both sata controllers. this might also affect the amd_pstate drive (didn't check the code if it requires temperature sensors)

If you want to get the cpu to boost add the following line to /etc/default/grub
GRUB_CMDLINE_LINUX_DEFAULT="quiet amd_pstate=passive"

after saving the file dont forget to run update-grub and reboot

For an explanation what this does you can check this https://docs.kernel.org/admin-guide/pm/amd-pstate.html

There is a down side, the cpu wont go down to 400 MHz, so power consumtion will go up

Ill do some more tinkering and see If I can get it to go to low idle frequency
On my searches online, your comment is the only one who gets closer to a possible solution.
Were you able to confirm if this indeed solves problem (both SATA controllers passthrough without CPU throttling)?
 
Hello, Anyone got any other workarounds to get it to work w/o throttling?
my solution was to individually passthrough each sata HDD with S/N to the vm
down side is you have to manage SMART by yourself in the PVE, not in the VM afaik
performance seems fine for 4 12HDDs
 
Hi, I'm looking for help/advice. I have a wtr pro with AMD and a virtual machine running Win11, but the processor doesn't boost at all, it's stuck at 2GHz, as in the screenshot. This is a fresh install of the host and VM, and there's no passthrough HDD or GPU.
I haven't changed anything in the BIOS.
EDIT:
I've noticed that this only happens in the VM, and it works fine in LXC.
Code:
Every 2.0s: grep "cpu MHz" /proc/cpuinfo
cpu MHz         : 4485.912
cpu MHz         : 4503.866


1754660576907.png
 
Last edited:
If the suggested approaches don't work, would an extreme approach of using a M.2 to 6xSATA adapter and cable up to the drive cage be possible? Yes, it sacrifices one M.2 (presumably on the slide-in set of 4. Several major issues there of course ...
 
If you join the Aoostar Discord, this very topic is being actively discussed. No fix yet but Aoostar have been made aware. The only work around at the moment is to compromise on performance by passing through the individual SATA drives instead of the SATA Controllers themselves.
Late to the party. Could you please tell me where I can find Aoostar on Discord?
 
Did anyone test with grub parameters to separate IOMMU
Code:
GRUB_CMDLINE_LINUX_DEFAULT="quiet amd_iommu=on iommu=pt pcie_acs_override=downstream,multifunction"


Yes i tried that to. it kind of helped a bit. That means it does not crash every hour but still at least every day.
It is very frustrating to have a beefy device and wont be able to use at all.

Also to be let down by the manufacturer but now at least we know why it's been so cheap.