Re: [RFC][PATCH] thunderbolt: Quiesce AMD NHI during suspend

From: Mario Limonciello

Date: Wed Jul 15 2026 - 09:42:24 EST


On 7/15/26 00:42, Chia-Lin Kao (AceLan) wrote:

Thanks for checking. That code ended up not helping the original issue it
was for, a different solution for that issue is being provided.

But I'm wondering, is this issue you're working on something to do with
pm_async? Could you turn that off and see if it improves anything?
1. echo 0 | sudo /sys/power/pm_async
2. sudo systemctl suspend
3. plug in the tbt dock
4. power indicator shows 3A5W lights and hard hang

I'm not sure what 3A5W lights means. You'll probably need to check with ODM if that's a post code or what it is. It certainly sounds useful to someone.

Are you sure the hang is during resume? It /sounds/ like it's during suspend.



Any chance you can get UART output from the system using a debug board over
suspend/resume in case there is a panic happening that doesn't flush to
NVME?
No, there is no UART port on the machine for debugging.

Can you please check with ODM about this? For pre-production systems there should be a debug card that can provide UART output so we can cross reference that.

Also; does this only happen on this system or you can reproduce on a reference system? If you can reproduce on a reference system UART is very easily accessible which would make this easier to debug.