On Tue, Apr 23, 2024 at 01:54:27PM -0500, Mario Limonciello wrote:

Hi,

We've got a collection of bug reports about how if a Thunderbolt device is connected at bootup it doesn't behave the same as if it were hotplugged by the user after bootup. Issues range from non-functional devices or display devices working at lower performance.

All of the issues stem from a pre-OS firmware initializing the USB4 controller and the OS re-using those tunnels. This has been fixed in 6.9-rc1 by resetting the controller to a fresh state and discarding whatever firmware has done. I'd like to bring it back to 6.6.y LTS and 6.8.y stable.

01da6b99d49f6 thunderbolt: Introduce tb_port_reset() b35c1d7b11da8 thunderbolt: Introduce tb_path_deactivate_hop() ec8162b3f0683 thunderbolt: Make tb_switch_reset() support Thunderbolt 2, 3 and USB4 routers 9a54c5f3dbde thunderbolt: Reset topology created by the boot firmware

The last one seems wrong, it's really 59a54c5f3dbde00b8ad30aef27fe35b1fe07bf5c, right?

thanks,

greg k-h

On Tue, Apr 23, 2024 at 01:54:27PM -0500, Mario Limonciello wrote:

Hi,

We've got a collection of bug reports about how if a Thunderbolt device is connected at bootup it doesn't behave the same as if it were hotplugged by the user after bootup. Issues range from non-functional devices or display devices working at lower performance.

All of the issues stem from a pre-OS firmware initializing the USB4 controller and the OS re-using those tunnels. This has been fixed in 6.9-rc1 by resetting the controller to a fresh state and discarding whatever firmware has done. I'd like to bring it back to 6.6.y LTS and 6.8.y stable.

01da6b99d49f6 thunderbolt: Introduce tb_port_reset() b35c1d7b11da8 thunderbolt: Introduce tb_path_deactivate_hop() ec8162b3f0683 thunderbolt: Make tb_switch_reset() support Thunderbolt 2, 3 and USB4 routers 9a54c5f3dbde thunderbolt: Reset topology created by the boot firmware

All now queued up, thanks.

greg k-h