BCM57504 NetXtreme-E not being picked up by VPP

[alexcrow1974]

I've got a new HP DL380 Gen11 server with the 4x25g "BCM57504 NetXtreme-E 10Gb/25Gb/40Gb/50Gb/100Gb/200Gb Ethernet [14e4:1751] (rev 12)" NIC, and I cannot get the interfaces to show up in VPP. I have intel_iommu=on in my linux command line, and dpdk shows the card as bound to the vfio-pci driver:

# dpdk-devbind.py -s

Network devices using DPDK-compatible driver
============================================
0000:11:00.0 'BCM57504 NetXtreme-E 10Gb/25Gb/40Gb/50Gb/100Gb/200Gb Ethernet 1751' drv=vfio-pci unused=bnxt_en,uio_pci_generic
0000:11:00.1 'BCM57504 NetXtreme-E 10Gb/25Gb/40Gb/50Gb/100Gb/200Gb Ethernet 1751' drv=vfio-pci unused=bnxt_en,uio_pci_generic
0000:11:00.2 'BCM57504 NetXtreme-E 10Gb/25Gb/40Gb/50Gb/100Gb/200Gb Ethernet 1751' drv=vfio-pci unused=bnxt_en,uio_pci_generic
0000:11:00.3 'BCM57504 NetXtreme-E 10Gb/25Gb/40Gb/50Gb/100Gb/200Gb Ethernet 1751' drv=vfio-pci unused=bnxt_en,uio_pci_generic

However, VPP logs show errors trying to connect to the card:

2025/09/25 12:06:11:534 notice     dpdk           EAL: Detected CPU lcores: 128
2025/09/25 12:06:11:534 notice     dpdk           EAL: Detected NUMA nodes: 2
2025/09/25 12:06:11:534 notice     dpdk           EAL: Detected static linkage of DPDK
2025/09/25 12:06:11:534 notice     dpdk           EAL: Selected IOVA mode 'VA'
2025/09/25 12:06:11:534 notice     dpdk           EAL: VFIO support initialized
2025/09/25 12:06:11:534 notice     dpdk           EAL: Using IOMMU type 1 (Type 1)
2025/09/25 12:06:11:534 notice     dpdk           BNXT: bnxt_hwrm_send_message(): Error(timeout) sending msg 0x0193, seq_id 8
2025/09/25 12:06:11:534 notice     dpdk           BNXT: bnxt_hwrm_func_backing_store_cfg(): failed rc:-110
2025/09/25 12:06:11:534 notice     dpdk           BNXT: bnxt_alloc_ctx_mem(): Failed to configure context mem: rc = -110
2025/09/25 12:06:11:534 notice     dpdk           ETHDEV: ethdev initialisation failed
2025/09/25 12:06:11:534 notice     dpdk           PCI_BUS: Requested device 0000:11:00.0 cannot be used
2025/09/25 12:06:11:534 notice     dpdk           EAL: Bus (pci) probe failed.

and dmesg also shows DMAR errors:

[  442.551771] DMAR: DRHD: handling fault status reg 2
[  442.551796] DMAR: [DMA Write NO_PASID] Request device [11:00.0] fault addr 0x0 [fault reason 0x79] SM: Read/Write permission error in second-level paging entry

I've tried various options on the kernel command line, but none seem to make any difference. I've looked at Broadcom's niccli utility but I can't find anything useful to do with it that matches this issue.

I'd rather not have to order a new Intel/Nvidia card so any ideas or tests I can perform would be very userful.

Best regards

Alex