☰

NVMe M.2“口香糖棒”SSD

tag-icon • tag-icon nvidia suspend performance 19.04
NVMe M.2“口香糖棒”SSD

我使用的是 precision 5530、i9、32GB RAM、SSD，使用 nvidia 驱动程序。启动时我没有遇到性能问题。但如果我暂停然后恢复，我会看到性能和延迟显著下降。

我在网上看到这可能是 CPU 节流，但运行cpupower frequency-info结果

analyzing CPU 0:
  driver: intel_pstate
  CPUs which run at the same hardware frequency: 0
  CPUs which need to have their frequency coordinated by software: 0
  maximum transition latency:  Cannot determine or is not supported.
  hardware limits: 800 MHz - 4.80 GHz
  available cpufreq governors: performance powersave
  current policy: frequency should be within 800 MHz and 4.80 GHz.
              The governor "powersave" may decide which speed to use
              within this range.
  current CPU frequency: Unable to call hardware
  current CPU frequency: 3.80 GHz (asserted by call to kernel)
  boost state support:
    Supported: yes
    Active: yes

它似乎没有受到限制，但一切显然都很慢。

我确实看到 chrome gpu 进程占用了 50-100% 的 CPU，这很有启发性。

我在 18.10 上没有遇到这个问题。

这可能是什么原因造成的？我该如何调试和修复？

答案1

使用sudo -H gedit /lib/systemd/system-sleep/custom-xhci_hcd

将这些行复制到编辑器中：

#!/bin/bash

# Original script was using /bin/sh but shellcheck reporting warnings.

# NAME: custom-xhci_hcd
# PATH: /lib/systemd/system-sleep
# CALL: Called from SystemD automatically
# DESC: Suspend broken for USB3.0 as of Oct 25/2018 various kernels all at once

# DATE: Oct 28 2018.

# NOTE: From comment #61 at: https://bugs.launchpad.net/ubuntu/+source/linux/+bug/522998

TMPLIST=/tmp/xhci-dev-list

# Original script was: case "${1}" in hibernate|suspend)

case $1/$2 in
  pre/*)
    echo "$0: Going to $2..."
    echo -n '' > $TMPLIST
          for i in `ls /sys/bus/pci/drivers/xhci_hcd/ | egrep '[0-9a-z]+\:[0-9a-z]+\:.*$'`; do
              # Unbind xhci_hcd for first device XXXX:XX:XX.X:
               echo -n "$i" | tee /sys/bus/pci/drivers/xhci_hcd/unbind
           echo "$i" >> $TMPLIST
          done
        ;;
  post/*)
    echo "$0: Waking up from $2..."
    for i in `cat $TMPLIST`; do
              # Bind xhci_hcd for first device XXXX:XX:XX.X:
              echo -n "$i" | tee /sys/bus/pci/drivers/xhci_hcd/bind
    done
    rm $TMPLIST
        ;;
esac

保存文件并退出编辑器。然后使用：

sudo chmod a+x /lib/systemd/system-sleep/custom-xhci_hcd

重启后，您的挂起/恢复问题有望得到解决。如果没有解决，请告诉我，我会删除此答案。

NVMe M.2“口香糖棒”SSD

这些微型 SSD 性能强大，但需要特别注意暂停/恢复，如下所示错误报告. 编辑文件/etc/default/grub并搜索此行：

GRUB_CMDLINE_LINUX_DEFAULT="quiet splash ... acpiphp.disable=1 pcie_aspm=force"

不要添加 ... 但要在其他参数后面的某个位置添加。除非单独添加第一个选项不起作用，否则acpiphp.disable=1不要添加。acpi_aspm=force

保存文件并使用sudo update-grub。然后重新启动进行测试。

答案2

这可能与这个错误有关：https://bugs.launchpad.net/ubuntu/+source/linux/+bug/671932。

每次重现此情况时我都重新启动服务thermald：

sudo systemctl restart thermald.service

相关内容