NVIDIA A100, H200 NVL GPU Reports gpuClearFbhubPoisonIntrForBug2924523_GA100 in Linux Kernel
Summary: The NVRM message "NVRM: gpuClearFbhubPoisonIntrForBug2924523_GA100: FBHUB Interrupt detected. Clearing it" on systems equipped with NVIDIA A100 or H200 NVL GPUs running NVIDIA Driver may be observed in the Linux kernel log such as dmesg. ...
Symptoms
After installing NVIDIA Driver for A100 or H200 NVL (Dual-GPU NVLink configuration), the following message may appear repeatedly when reviewing kernel logs as seen in Figure 1:
NVRM: gpuClearFbhubPoisonIntrForBug2924523_GA100: FBHUB Interrupt detected. Clearing it.
Figure 1: Error Message Example in Linux dmesg Log
Observed behavior:
- System boots normally.
- GPU is detected correctly.
- NVIDIA driver loads successfully.
- No GPU functionality loss observed.
- NVLINK diagnostics pass without errors.
- No additional PCIe, GPU, or NVLINK errors are reported.
Cause
The message is generated by the NVIDIA kernel driver and corresponds to an internal driver routine function:
gpuClearFbhubPoisonIntr
This routine function is designed to clear a poison interrupt reported by the Frame Buffer Hub (FBHUB).
According to NVIDIA, this handling is associated with NVIDIA internal Issue ID 2924523.
The NVIDIA driver contains proactive mitigation logic to automatically detect and clear this interrupt condition. The log entry serves as an informational notification that the workaround has been executed.
The message has previously been observed on NVIDIA A100 and H100 NVL platforms and does not indicate a GPU hardware failure.
Resolution
- No corrective action is required.
- This message can be safely treated as an informational warning generated by the NVIDIA driver.