Build display tunnels on the upstream T6034 work or independently #9

Part of #1

Question

Should the spec's DisplayPort-tunnel slices build on the upstream T6034 work in draft omacom#406 (and the T6031 work in omacom#292 it overlaps), or specify T6034 support independently?

Facts from Existing Apple USB4 work and what applies to the T6034:

• M3 Max (J516C): external displays over USB-C and HDMI omacom/linux-aurora#292's T6031 DP-tunnel support is already on usb4-m3max but stays inert on the J516M. It keys on apple,t6031; this machine's root compatible is apple,j516m, apple,t6034.
• T6034 (J516M): external displays over Thunderbolt, tested with a Studio Display at 5K (stacked on #405, overlaps #292) omacom/linux-aurora#406, from another J516M owner, adds apple,t6034 qualification, the dcpext0 handoff and device-tree nodes. It was tested at 5K on the right port, but it's stacked on J516M: drive the 16" M3 Max MacBook Pro's display and GPU by default (stacked on #372) omacom/linux-aurora#405 and Add guarded J514M/T6034 native graphics bring-up omacom/linux-aurora#372 and needs kboot_display_handoff: hand off the T6034 dcpext0 for external displays (stacked on #28, overlaps #21) omacom/m1n1#29.
• Only one tunneled display is possible for now: the other three display processors are marked failed.

Decide: depend on omacom#406 (and on which of omacom#405, omacom#372 and m1n1#29), take its approach but write it ourselves, or wait for omacom#292 and omacom#406 to land. Also decide whether to contact omacom#406's author before PR time, which the map currently rules out.