Created attachment 286265 [details]
dmesg, v5.5-rc1

This is found on connecting usb-storage drives on Dell WD19 dock, and retry s2idle suspend/resume several times by running fwts:

  $ sudo fwts s3 --s3-multiple=100 --s3-sleep-delay=40 \
      --s3-min-delay=30 --s3-max-delay=30

The suspend/resume process may success, but /dev/sd{a,b,c}1 may not be available after resumed. One can find following error messages when this happens:

  PM: noirq resume of devices complete after 2073.542 msecs
  PM: early resume of devices complete after 1.086 msecs
  hub 2-3:1.0: hub_ext_port_status failed (err = -71)
  hub 2-3:1.0: hub_ext_port_status failed (err = -71)
  usb 2-3-port3: cannot disable (err = -71)
  xhci_hcd 0000:00:0d.0: xHCI host not responding to stop endpoint command.
  xhci_hcd 0000:00:0d.0: xHCI host controller not responding, assume dead
  xhci_hcd 0000:00:0d.0: HC died; cleaning up
  usb 2-3-port4: cannot disable (err = -110)
  PM: resume of devices complete after 6399.217 msecs
  PM: Finishing wakeup.
  OOM killer enabled.
  Restarting tasks ... 
  usb 2-3: USB disconnect, device number 62
  usb 2-3.3: USB disconnect, device number 63
  usb 2-3.3.1: USB disconnect, device number 65
  done.
  usb 2-3.3.2: USB disconnect, device number 66
  scsi 1:0:0:0: rejecting I/O to dead device
  PM: suspend exit
  usb 2-3.3.3: USB disconnect, device number 67
  scsi 2:0:0:0: rejecting I/O to dead device
  usb 2-3.4: USB disconnect, device number 64

This can be reproduced with current v5.5-rc1 kernel, and can only be found on Intel Ice Lake(model 126) platforms.