I have a hard nut to crack with starting the VM where 2 (the same type) PCIe network cards are connected with PCI passthrough.
VM is running fine with one card, even if the second card is inserted in the PCIe slot.
dmesg shows:
[ 198.679707] vfio-pci 0000:0b:00.0: enabling device (0140 -> 0142)
[ 198.794074] vfio-pci 0000:05:00.0: enabling device (0140 -> 0142)
[ 198.910312] genirq: Flags mismatch irq 16. 00000000 (vfio-intx(0000:05:00.0)) vs. 00000000 (vfio-intx(0000:0b:00.0))
Swapping NICs between PCIe slots doesn't help - they still have the same IRQ (16).
Platform: HP DL380 Gen9 (ROM: P89_3.04_08_04_2022 - latest available)
NIC1: HP 516937-B21 (Mellanox Connect-X2, FW Version: 2.9.1530)
NIC2: HP 516937-B21 (Mellanox Connect-X2, FW Version: 2.9.1530)
05:00.0 Ethernet controller: Mellanox Technologies MT26448 [ConnectX EN 10GigE, PCIe 2.0 5GT/s] (rev b0)
0b:00.0 Ethernet controller: Mellanox Technologies MT26448 [ConnectX EN 10GigE, PCIe 2.0 5GT/s] (rev b0)
## Error visble in GUI:
kvm: -device vfio-pci,host=0000:05:00.0,id=hostpci1,bus=pci.0,addr=0x11: vfio 0000:05:00.0: Failed to set up TRIGGER eventfd signaling for interrupt INTX-0: VFIO_DEVICE_SET_IRQS failure: Device or resource busy
TASK ERROR: start failed: QEMU exited with code 1
## journalctl
Jan 05 21:43:50 prox2 systemd[1]: Started 101.scope.
Jan 05 21:43:53 prox2 kernel: vfio-pci 0000:0b:00.0: enabling device (0140 -> 0142)
Jan 05 21:43:53 prox2 kernel: vfio-pci 0000:05:00.0: enabling device (0140 -> 0142)
Jan 05 21:43:53 prox2 kernel: genirq: Flags mismatch irq 16. 00000000 (vfio-intx(0000:05:00.0)) vs. 00000000 (vfio-intx(0000:0b:00.0))
Jan 05 21:43:53 prox2 pvedaemon[1185]: VM 101 qmp command failed - VM 101 not running
Jan 05 21:43:53 prox2 systemd[1]: 101.scope: Succeeded.
Jan 05 21:43:53 prox2 systemd[1]: 101.scope: Consumed 3.274s CPU time.
Jan 05 21:43:53 prox2 pvedaemon[4912]: start failed: QEMU exited with code 1
Any ide how to solve this problem?
VM is running fine with one card, even if the second card is inserted in the PCIe slot.
dmesg shows:
[ 198.679707] vfio-pci 0000:0b:00.0: enabling device (0140 -> 0142)
[ 198.794074] vfio-pci 0000:05:00.0: enabling device (0140 -> 0142)
[ 198.910312] genirq: Flags mismatch irq 16. 00000000 (vfio-intx(0000:05:00.0)) vs. 00000000 (vfio-intx(0000:0b:00.0))
Swapping NICs between PCIe slots doesn't help - they still have the same IRQ (16).
Platform: HP DL380 Gen9 (ROM: P89_3.04_08_04_2022 - latest available)
NIC1: HP 516937-B21 (Mellanox Connect-X2, FW Version: 2.9.1530)
NIC2: HP 516937-B21 (Mellanox Connect-X2, FW Version: 2.9.1530)
05:00.0 Ethernet controller: Mellanox Technologies MT26448 [ConnectX EN 10GigE, PCIe 2.0 5GT/s] (rev b0)
0b:00.0 Ethernet controller: Mellanox Technologies MT26448 [ConnectX EN 10GigE, PCIe 2.0 5GT/s] (rev b0)
## Error visble in GUI:
kvm: -device vfio-pci,host=0000:05:00.0,id=hostpci1,bus=pci.0,addr=0x11: vfio 0000:05:00.0: Failed to set up TRIGGER eventfd signaling for interrupt INTX-0: VFIO_DEVICE_SET_IRQS failure: Device or resource busy
TASK ERROR: start failed: QEMU exited with code 1
## journalctl
Jan 05 21:43:50 prox2 systemd[1]: Started 101.scope.
Jan 05 21:43:53 prox2 kernel: vfio-pci 0000:0b:00.0: enabling device (0140 -> 0142)
Jan 05 21:43:53 prox2 kernel: vfio-pci 0000:05:00.0: enabling device (0140 -> 0142)
Jan 05 21:43:53 prox2 kernel: genirq: Flags mismatch irq 16. 00000000 (vfio-intx(0000:05:00.0)) vs. 00000000 (vfio-intx(0000:0b:00.0))
Jan 05 21:43:53 prox2 pvedaemon[1185]: VM 101 qmp command failed - VM 101 not running
Jan 05 21:43:53 prox2 systemd[1]: 101.scope: Succeeded.
Jan 05 21:43:53 prox2 systemd[1]: 101.scope: Consumed 3.274s CPU time.
Jan 05 21:43:53 prox2 pvedaemon[4912]: start failed: QEMU exited with code 1
Any ide how to solve this problem?