Recovering from E810 MDD

Recovering from E810 MDD

[Coby Schmidt]

[-- Attachment #1: Type: text/plain, Size: 1372 bytes --]

Hello developers. Pardon if this is not the correct mailing list.

We are using Intel E810 with ICE PMD (DPDK 21.11.2). Once every 2 weeks or so we are having a MDD event and  we can see in syslog: messages of this sort: "ice_interrupt_handler(): Malicious Driver Detection event 7 by TCLAN on TX queue 8 PF# 0". As expected the relevant queue is non-active

We are not using VF and all packets are generated on our stack, however, we are looking for some guidance on how to recover  from this. On ethtool there is an option to allow to restart VF upon a failure (kernel patch  reference: https://lore.kernel.org/netdev/20240326164455.735739-1-marcin.szycik@linux.intel.com/T/).

Looking on the ICE PMD I don't see how to directly "translate" the kernel driver code to PMD. I wonder if you could provide some guidance on how to recover. Is there a patch somewhere? Should queue restart be enough while disregarding VF?

Some more details on the different HW,SW we use:
DPDK 21.11.2

DDP ICE OS Default Package version 1.3.30.0
Ethernet controller: Intel Corporation Device 159b (rev 02) firmware-version: 2.50 0x800077a8 1.2960.0
Ethernet controller: Intel Corporation Device 159b (rev 02) firmware-version: 3.20 0x8000d853 1.3146.0
Ethernet controller: Intel Corporation Device 159b (rev 02) firmware-version: 4.10 0x80015188 1.3310.0



Thanks,
Coby

[-- Attachment #2: Type: text/html, Size: 4069 bytes --]