Add 'bind --force' option

[dorssel]

• Remove 'bind --all' option (obsolete)

Fixes #118

[codecov]

Codecov Report

Merging #164 (052185e) into master (72cb390) will increase coverage by 0.04%.
The diff coverage is 8.99%.

@@            Coverage Diff             @@
##           master     #164      +/-   ##
==========================================
+ Coverage   29.91%   29.95%   +0.04%     
==========================================
  Files          21       23       +2     
  Lines        1882     1906      +24     
==========================================
+ Hits          563      571       +8     
- Misses       1319     1335      +16     

Impacted Files	Coverage Δ
UsbIpServer/CommandHandlers.cs	0.00% <0.00%> (ø)
UsbIpServer/ConfigurationManager.cs	0.00% <0.00%> (ø)
UsbIpServer/ConnectedClient.cs	0.00% <0.00%> (ø)
UsbIpServer/ConsoleTools.cs	31.64% <0.00%> (-14.66%)	⬇️
UsbIpServer/ExportedDevice.cs	0.00% <0.00%> (ø)
UsbIpServer/Interop/VBoxUsb.cs	100.00% <ø> (ø)
UsbIpServer/NewDev.cs	0.00% <0.00%> (ø)
UsbIpServer/RegistryUtils.cs	0.00% <0.00%> (ø)
UsbIpServer/UsbDevice.cs	0.00% <0.00%> (ø)
UsbIpServer/Interop/VBoxUsbMon.cs	100.00% <100.00%> (ø)
... and 8 more

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 72cb390...052185e. Read the comment docs.

[dorssel]

@benmcmorran
Don't worry, not asking to review this monster PR. This is just FYI.

The --force options is done! I had to refactor the entire device management (and registry layout), but all for the better. Not tied to BUSID anymore: now it is completely based on Windows unique Device Instance IDs. Much faster, and when using --force it completely (100%) bypasses VBoxUsbMon. I'll ask some users to test; could you do the same?
PR build artifact at: https://github.com/dorssel/usbipd-win/actions/runs/1705631433

[Josverl]

@dorssel ,
I gave this a quick try and all my tested devices work.

However I have seen a few times that usbipd wsl attach or detach crashes.
See: crash on attempted detach : https://gist.github.com/Josverl/06d2f1e633ae75b25c0290302c26fa4c

even when the ubuntu distro was stopped, the devices remained attached.
a restart of the server solved that

[dorssel]

@Josverl

even when the ubuntu distro was stopped, the devices remained attached.
a restart of the server solved that

Zooming in on this part of your report: the socket connection to the client is set to a 15 second keepalive timeout. When I attach a device and then do wsl --shutdown, 15 seconds later usbipd-win recognizes the lost connection and releases the device. Just to make sure, did you give it that much time?

EDIT: keep-alives sent after 10 seconds, 5 times 1 second apart: total timeout 15 seconds.

[dorssel]

@Josverl

See: crash on attempted detach : https://gist.github.com/Josverl/06d2f1e633ae75b25c0290302c26fa4c

EDIT:
Duh! Pilot error on my part. See fix: 052185e
New artifact at: https://github.com/dorssel/usbipd-win/actions/runs/1710391286

[benmcmorran]

I just tried this on a few embedded dev boards and everything worked well.

[dorssel]

Thank for testing, everyone! Merging...

[Josverl]

When I attach a device and then do wsl --shutdown, 15 seconds later usbipd-win recognizes the lost connection and releases the device. Just to make sure, did you give it that much time?
EDIT: keep-alives sent after 10 seconds, 5 times 1 second apart: total timeout 15 seconds.

hmm , no unlikely that I was that patient after shutting down WSL