Hello
We created a virtual machine on a machine equipped with 8 RTX 4090 GPUs. When creating a VM with all 8 GPUs attached, some GPUs failed to appear inside the VM. The dmesg output in the VM reported the following error:
Jun 30 09:15:02 0197c01c-86fc-7235-8687-604911c3ba81 kernel: [ 13.851280] NVRM: GPU 0000:f1:00.0: RmInitAdapter failed! (0x31:0x40:2628)
Jun 30 09:15:02 0197c01c-86fc-7235-8687-604911c3ba81 kernel: [ 13.852359] NVRM: GPU 0000:f1:00.0: rm_init_adapter failed, device minor number 0
Jun 30 09:15:02 0197c01c-86fc-7235-8687-604911c3ba81 kernel: [ 13.853582] [drm:nv_drm_load [nvidia_drm]] ERROR [nvidia-drm] [GPU ID 0x0000f100] Failed to allocate NvKmsKapiDevice
Jun 30 09:15:02 0197c01c-86fc-7235-8687-604911c3ba81 kernel: [ 13.855887] [drm:nv_drm_register_drm_device [nvidia_drm]] ERROR [nvidia-drm] [GPU ID 0x0000f100] Failed to register device
I am attaching the nvidia-bug-report output…
nvidia-bug-report.log.gz (2.5 MB)