Ubuntu 22.04.1 运行一段时间后死机崩溃的问题求助

Hey Stella, sorry to hear you're stuck with this frustrating random freeze-and-restart issue—let's walk through some practical troubleshooting steps to get to the bottom of this:

• Dig into system logs first
  Since the system restarts after freezing, the logs from the previous boot will hold critical clues. Run journalctl -b -1 to pull up logs from your last session (before the restart). Scan the final few minutes of output for keywords like ERROR, CRITICAL, or amdgpu—these will likely point to whether the issue is tied to your AMD GPU drivers. You can also check the /var/log/syslog file with grep -i "freeze\|reboot\|amdgpu" /var/log/syslog to filter relevant entries quickly.

• Verify and update your AMD GPU drivers
  Ubuntu 22.04 typically uses the open-source amdgpu driver by default. First, confirm which driver your system is using with lspci -k | grep -A 5 -i vga—look for the "Kernel driver in use" line.
  Next, make sure all system packages (including driver components) are up to date: run sudo apt update && sudo apt full-upgrade and reboot. If the issue persists, you could try tweaking driver-specific GRUB parameters (since your initial edit didn't help) – add amdgpu.reset=1 or amdgpu.noretry=0 to the GRUB_CMDLINE_LINUX_DEFAULT line in /etc/default/grub, then run sudo update-grub and restart. For testing, you can also try adding nomodeset (this disables GPU acceleration temporarily) – if the freezes stop, that confirms the problem is driver-related.

• Check for overheating
  Random crashes often tie back to overheating, especially with GPUs. Install the lm-sensors tool with sudo apt install lm-sensors, then run sensors to monitor CPU and GPU temperatures. If you see temps spiking above 90°C before a freeze, you might need to clean dust from your GPU/fans or check if cooling fans are functioning properly.

• Test your hardware (memory and disk)
  Faulty memory or disk errors can cause unpredictable crashes:

  • For memory: Install memtester with sudo apt install memtester, then run sudo memtester 4096 2 (replace 4096 with 1/4 of your total RAM, e.g., 16384 for 16GB) to stress-test your memory. Alternatively, use stress-ng with sudo stress-ng --vm 4 --vm-bytes 8G --timeout 60s to push your RAM hard.
  • For disk: Use sudo smartctl -a /dev/nvme0n1 (replace with your disk device, like /dev/sda for SATA) to check SMART health data—look for any "Failed" or "Warning" statuses. You can also run a file system check with sudo fsck /dev/sda1 (do this from a live USB or single-user mode to avoid mounting issues).

• Rule out user profile issues
  Sometimes corrupted user configs can cause odd system behavior. Create a test user with sudo adduser testuser, log into this new account, and use the system as you normally would. If the freezes don't happen here, you'll know the problem is tied to your main user's configuration files.

备注：内容来源于stack exchange，提问作者Stella