Re: xhci_hcd: AMD Raphael/Granite Ridge USB 2.0 xHCI [1022:15b8] dies on resume from suspend

[Michal Pecio <michal.pecio@gmail.com>]

On Sat, 9 May 2026 15:51:03 +0100, Martin Alderson wrote:
> Hi, still experiencing this on 7.0.2. I tried to pull the logs
> together to get to the bottom of this (I've tried a few different
> kernels)
> 
> Kernel                          Suspends   xHCI 0f:00.0 deaths   Rate
> ------------------------------  --------   -------------------   -----
> 6.17.1-300.fc43 (March)             ~12             0             0%
> 6.18.16-200.fc43                     10             0             0%
> 6.19.7/8-200.fc43                     5             0             0%
> 7.0-rc4   (build 260320)             13             0             0%
> 7.0-rc5   (build 260328)              7             2            ~28%
> 7.0-rc6   (build 260401)             10             4             40%
> 7.0-rc7   (build 260409)              7             2            ~28%
> 7.0.0-261.vanilla.fc43                7             2            ~28%
> 6.17.1-300.fc43 (April, retry)       10             2             20%
>  <-- same bug, stable kernel

Looks like it's not a regression then, but not sure what else may have
caused it.

Any new USB device that wasn't connected before?
Perhaps a BIOS upgrade?

> 7.0.1-262.vanilla.fc43                7             2            ~28%
> 7.0.2-300.vanilla.fc44                6             4            ~66%
> 
> 
> May 09 15:29:37 fedora kernel: Freezing user space processes completed
> (elapsed 0.001 seconds)
> May 09 15:29:37 fedora kernel: OOM killer disabled.
> May 09 15:29:37 fedora kernel: Freezing remaining freezable tasks
> May 09 15:29:37 fedora kernel: Freezing remaining freezable tasks
> completed (elapsed 0.001 seconds)
> May 09 15:29:37 fedora kernel: printk: Suspending console(s) (use
> no_console_suspend to debug)
> May 09 15:29:37 fedora kernel: sd 6:0:0:0: [sdb] Synchronizing SCSI cache
> May 09 15:29:37 fedora kernel: serial 00:01: disabled
> May 09 15:29:37 fedora kernel: sd 0:0:0:0: [sda] Synchronizing SCSI cache
> May 09 15:29:37 fedora kernel: ata1.00: Entering standby power mode
> May 09 15:29:37 fedora kernel: xhci_hcd 0000:0f:00.0: xHCI host not
> responding to stop endpoint command
> May 09 15:29:37 fedora kernel: xhci_hcd 0000:0f:00.0: xHCI host
> controller not responding, assume dead
> May 09 15:29:37 fedora kernel: xhci_hcd 0000:0f:00.0: HC died; cleaning up
> May 09 15:29:37 fedora kernel: PM: suspend devices took 5.758 seconds

That's not resume, it's during suspend. Are other logs also like that?

Regards,
Michal