Before start the VM:
```
02:00.0 Non-Volatile memory controller: Samsung Electronics Co Ltd NVMe SSD Controller SM981/PM981 (prog-if 02 [NVM Express])
Interrupt: pin A routed to IRQ 17
NUMA node: 0
Kernel driver in use: nvme
07:00.0 VGA compatible controller: NVIDIA Corporation GF119 [GeForce GT 520] (rev a1) (prog-if 00 [VGA controller])
Interrupt: pin A routed to IRQ 11
Kernel driver in use: vfio-pci
Kernel modules: nvidiafb, nouveau, nvidia_drm, nvidia
07:00.1 Audio device: NVIDIA Corporation GF119 HDMI Audio Controller (rev a1)
Interrupt: pin B routed to IRQ 10
Kernel driver in use: vfio-pci
Kernel modules: snd_hda_intel
```
The IRQ of the 02:00 device and the IRQ of the 07:00 device are different.
However, when starting VM, the VM dies with an error(in the host) as shown below
```
[ 918.137315] irq 17: nobody cared (try booting with the "irqpoll" option)
[ 918.137339] CPU: 2 PID: 1979 Comm: kvm Tainted: P O 5.4.73-1-pve #1
[ 918.137340] Hardware name: Supermicro X10SLL+-F/X10SLL+-F, BIOS 3.3 03/28/2020
[ 918.137340] Call Trace:
[ 918.137342] <IRQ>
[ 918.137347] dump_stack+0x6d/0x9a
[ 918.137350] __report_bad_irq+0x3c/0xb6
[ 918.137352] note_interrupt.cold.10+0xb/0x5d
[ 918.137353] handle_irq_event_percpu+0x6f/0x80
[ 918.137354] handle_irq_ev
```
At this time, the IRQ of the passthrouth device is changed.
```
07:00.0 VGA compatible controller: NVIDIA Corporation GF119 [GeForce GT 520] (rev a1) (prog-if 00 [VGA controller])
Interrupt: pin A routed to IRQ 16
Kernel driver in use: vfio-pci
Kernel modules: nvidiafb, nouveau, nvidia_drm, nvidia
07:00.1 Audio device: NVIDIA Corporation GF119 HDMI Audio Controller (rev a1)
Interrupt: pin B routed to IRQ 17
Kernel driver in use: vfio-pci
Kernel modules: snd_hda_intel
```
The 07.00.1 device's IRQ is changed to 17, which collision the existing 02:00 device.
Before starting the VM, the IRQ was unique, not shared with other devices.
The irqpoll option does not help me.
```
02:00.0 Non-Volatile memory controller: Samsung Electronics Co Ltd NVMe SSD Controller SM981/PM981 (prog-if 02 [NVM Express])
Interrupt: pin A routed to IRQ 17
NUMA node: 0
Kernel driver in use: nvme
07:00.0 VGA compatible controller: NVIDIA Corporation GF119 [GeForce GT 520] (rev a1) (prog-if 00 [VGA controller])
Interrupt: pin A routed to IRQ 11
Kernel driver in use: vfio-pci
Kernel modules: nvidiafb, nouveau, nvidia_drm, nvidia
07:00.1 Audio device: NVIDIA Corporation GF119 HDMI Audio Controller (rev a1)
Interrupt: pin B routed to IRQ 10
Kernel driver in use: vfio-pci
Kernel modules: snd_hda_intel
```
The IRQ of the 02:00 device and the IRQ of the 07:00 device are different.
However, when starting VM, the VM dies with an error(in the host) as shown below
```
[ 918.137315] irq 17: nobody cared (try booting with the "irqpoll" option)
[ 918.137339] CPU: 2 PID: 1979 Comm: kvm Tainted: P O 5.4.73-1-pve #1
[ 918.137340] Hardware name: Supermicro X10SLL+-F/X10SLL+-F, BIOS 3.3 03/28/2020
[ 918.137340] Call Trace:
[ 918.137342] <IRQ>
[ 918.137347] dump_stack+0x6d/0x9a
[ 918.137350] __report_bad_irq+0x3c/0xb6
[ 918.137352] note_interrupt.cold.10+0xb/0x5d
[ 918.137353] handle_irq_event_percpu+0x6f/0x80
[ 918.137354] handle_irq_ev
```
At this time, the IRQ of the passthrouth device is changed.
```
07:00.0 VGA compatible controller: NVIDIA Corporation GF119 [GeForce GT 520] (rev a1) (prog-if 00 [VGA controller])
Interrupt: pin A routed to IRQ 16
Kernel driver in use: vfio-pci
Kernel modules: nvidiafb, nouveau, nvidia_drm, nvidia
07:00.1 Audio device: NVIDIA Corporation GF119 HDMI Audio Controller (rev a1)
Interrupt: pin B routed to IRQ 17
Kernel driver in use: vfio-pci
Kernel modules: snd_hda_intel
```
The 07.00.1 device's IRQ is changed to 17, which collision the existing 02:00 device.
Before starting the VM, the IRQ was unique, not shared with other devices.
The irqpoll option does not help me.
Last edited: