Hi Jintack, I run into the same issue as you described in https://lists.nongnu.org/archive/html/qemu-devel/2018-02/msg03876.html I try to pass through MLNX VF and NVME to level2-vm, but both these two pci device can display with lspci correctly, but have issue with driver. My env are: Host Linux version : 5.15.131 Qemu: 8.2.0
Since you reported this issue in 2018, does this have any update from there? Level2 VM dmesg error logs while passing through MLNX VF are below: [Mon Oct 14 04:12:47 2024] pci 0000:00:06.0: [15b3:101e] type 00 class 0x020000 [Mon Oct 14 04:12:47 2024] pci 0000:00:06.0: reg 0x10: [mem 0x00000000-0x000fffff 64bit pref] [Mon Oct 14 04:12:47 2024] pci 0000:00:06.0: enabling Extended Tags [Mon Oct 14 04:12:47 2024] pci 0000:00:06.0: 0.000 Gb/s available PCIe bandwidth, limited by Unknown x0 link at 0000:00:06.0 (capable of 126.024 Gb/s with 16.0 GT/s PCIe x8 link) [Mon Oct 14 04:12:47 2024] pci 0000:00:06.0: BAR 0: assigned [mem 0x140000000-0x1400fffff 64bit pref] [Mon Oct 14 04:12:48 2024] mlx5_core 0000:00:06.0: enabling device (0000 -> 0002) [Mon Oct 14 04:12:48 2024] mlx5_core 0000:00:06.0: firmware version: 22.35.2000 [Mon Oct 14 04:13:50 2024] mlx5_core 0000:00:06.0: wait_func:1151:(pid 1193): ENABLE_HCA(0x104) timeout. Will cause a leak of a command resource [Mon Oct 14 04:13:50 2024] mlx5_core 0000:00:06.0: mlx5_function_setup:1164:(pid 1193): enable hca failed [Mon Oct 14 04:13:50 2024] mlx5_core 0000:00:06.0: probe_one:1770:(pid 1193): mlx5_init_one failed with error code -110 [Mon Oct 14 04:13:50 2024] mlx5_core: probe of 0000:00:06.0 failed with error -110 Host dmesg error logs while passing through MLNX VF are below: [Mon Oct 14 08:07:08 2024] DMAR: DRHD: handling fault status reg 2 [Mon Oct 14 08:07:08 2024] DMAR: [DMA Read NO_PASID] Request device [ca:08.2] fault addr 0x104ead000 [fault reason 0x06] PTE Read access is not set Thanks, Ryan
