Re: [PATCH 0/2] usbip: stability changes

From: John Ernberg

Date: Tue Nov 17 2026 - 07:34:17 EST


Hi Shuah,

On 17/11/2026 00:16, Shuah Khan wrote:
On 11/13/26 04:34, John Ernberg wrote:
Hi Shuah,

On 12/11/2026 23:46, Shuah Khan wrote:
On 11/11/26 04:34, Michael Grzeschik wrote:
While using usbip on kirkwood embedded machine, I came accross
some stability issues. This series is addressing the reference counting
on and usage of the event handler itself.

Tell me more about the stability issues? What doesn't work
without these patches? Do you have reproducers for these
problems?


Since we also had the issue that 2/2 solves: The issue we have is that we see occasional CI failures with ghost USB devices remaining on the client side after our system (server side) had been suspended, even if the device was removed from the server.

Applying a patch 2/2 (or our version of it sent earlier) solved the issue.

Hi John,

Can you send Tested-by for the 2/2 - did you also test patch 1/2?

I have not tested patch 1/2 because we have not seen that problem in our use cases so far, so we do not have a measurement of "good" for that fix. Michael will have to comment on that one.

Regarding patch 2/2, it seems Michael prefers our version [1].
I can provide a tested-by if you rather want to pick this one.

Best regards // John Ernberg

[1]: https://lore.kernel.org/linux-usb/avToeux2DxhBWGN4@xxxxxxxxxxxxxx/