sty

@lsp7572 我能搜到的软件解决办法都试过。这台机器在机房，走流程挺麻烦，如果软件层面解决不了，那就报硬件流程去了

@cinlen 正常情况下 lspci -s 01:00.0 -nnDk
0000:01:00.0 VGA compatible controller [0300]: NVIDIA Corporation Device [10de:2203] (rev a1)
Subsystem: Device [7377:2000]
Kernel driver in use: nvidia
libkmod: kmod_config_parse: /etc/modprobe.d/blacklist-nouveau.conf line 1: ignoring bad line starting with 'cklist'
Kernel modules: nouveau, nvidia_drm, nvidia

@rickiey nvidia-smi
```
Mon Aug 12 15:30:33 2024
+-----------------------------------------------------------------------------------------+
| NVIDIA-SMI 550.107.02 Driver Version: 550.107.02 CUDA Version: 12.4 |
|-----------------------------------------+------------------------+----------------------+
| GPU Name Persistence-M | Bus-Id Disp.A | Volatile Uncorr. ECC |
| Fan Temp Perf Pwr:Usage/Cap | Memory-Usage | GPU-Util Compute M. |
| | | MIG M. |
|=========================================+========================+======================|
| 0 NVIDIA GeForce RTX 3090 Ti Off | 00000000:01:00.0 Off | Off |
| 30% 41C P0 N/A / 450W | 1MiB / 24564MiB | 0% Default |
| | | N/A |
+-----------------------------------------+------------------------+----------------------+

+-----------------------------------------------------------------------------------------+
| Processes: |
| GPU GI CI PID Type Process name GPU Memory |
| ID ID Usage |
|=========================================================================================|
| No running processes found |
+-----------------------------------------------------------------------------------------+
```

@cinlen [ 2.018550] [drm] [nvidia-drm] [GPU ID 0x00000100] Loading driver
[ 369.857712] NVRM: GPU 0000:01:00.0: Failed to enable MSI; falling back to PCIe virtual-wire interrupts.
[ 493.216012] NVRM: GPU 0000:01:00.0: Failed to enable MSI; falling back to PCIe virtual-wire interrupts.
[ 1537.808965] NVRM: GPU 0000:01:00.0: Failed to enable MSI; falling back to PCIe virtual-wire interrupts.
[ 1764.689999] NVRM: GPU 0000:01:00.0: Failed to enable MSI; falling back to PCIe virtual-wire interrupts.
[ 1766.588211] NVRM: GPU 0000:01:00.0: Failed to enable MSI; falling back to PCIe virtual-wire interrupts.
[ 1775.551022] NVRM: GPU 0000:01:00.0: Failed to enable MSI; falling back to PCIe virtual-wire interrupts.
老哥帮忙看下

@HojiOShi 3090ti ，买了三年多了不咋用，最近 3 个月开始用的

@DigitalG 坏了之后，driver 我自己更新过了。在使用的时候没报错，比如我跑 3 个小时的训练，能跑完。反而是空闲的时候就报上面这个错。一天 1 到 2 次吧，每天都有

@GoRoad 一周多了，每天都得重启，更新了一下驱动也没用。有没有可能是某些区块坏了，要跑一段时间才能碰到坏的区块

@huoxing9888888 坐标？