Intel D7-P5500 timeouts on Linux 6.17.4

[bga]

I have Intel SSD D7-P5500 Series (SSDPF2KX019T9N) SSDs that I seem to be having problems with when trying to use Proxmox's Kernel 6.17.4-2-pve.  Earlier 6.14 kernels work without issue.  I've upgraded to firmware 2CV10400 without improvement.  The systems are Dell Precision 7820s with Intel Xeon Gold 6130 and 6140 CPUs with BIOS 2.50.0.

With no special options, during early boot I end up with:
KERNEL PANIC! Please reboot your computer. Fatal exception in interrupt.

With pci_aspm=off I end up with:
nvme nvme0: I/O tag 16 (1010) QID 0 timeout, completion polled
watchdog: CPU10: Watchdog detected hard LOCKUP on cpu 10

Any thoughts on what else I could try?

[SolidigmChris]

Have you changed back to the 6.14 kernel to see if it works again?

Based off unanswered Proxmox Forums post: ZFS and NVMe issues with PVE Kernel 6.17 | Proxmox Support Forum.

• User had issue with PVE 9, when first released.  Reverted to PVE 8, issues disappeared.
• User later tried PVE 9 again, issues returned.  "Pined' 6.8.12-17-pve, issues disappeared.

Are you able to install SST or nvme-cli?  If so, we could provide commands for either to see logs of the drive and determine it's status.

• Solidigm Storage Tool™
• GitHub - linux-nvme/nvme-cli: NVMe management command line interface.

[bga]

It does work when changing back to the 6.14 kernel.

I've just upgraded to the in-testing 6.17.9-1-pve kernel and that seems to resolve the problem.

I do have SST and nvme-cli installed.

Thanks!