[Bug 94213] [KBL] System needs long delay to boot with MST hub connected and then no multi stream

[bugzilla-daemon at freedesktop.org]

https://bugs.freedesktop.org/show_bug.cgi?id=94213

--- Comment #2 from Jim Bride <jim.bride at linux.intel.com> ---
(In reply to yann from comment #1)
> Jim, any advice on how to proceed on this issue?

I would stop testing with two external displays on MST for now; that will be
problematic until we have the DP link training rework in place.  From what I'm
seeing in dmesg, there are no MST-specific errors visible.  There is a ton of
power management noise in there that I would be suspicious of.

I don't have a Kaby Lake myself, nor have I read much about it in the B-Spec,
but clock setup is an area that has bitten us before, as is power management. 
There is one other thing that I find odd.

Feb 19 14:29:34 KBLU1 kernel: [    1.350958] [drm:intel_dp_probe_oui] Sink OUI:
0080e1
Feb 19 14:29:34 KBLU1 kernel: [    1.353319] [drm:intel_dp_probe_oui] Branch
OUI: 0080e1
Feb 19 14:29:34 KBLU1 kernel: [    1.355688] [drm:intel_dp_probe_mst] Sink is
MST capable

I've never seen the Sink OUI and Branch OUI the same before, and I find this
particularly odd considering that a hub is being used.  What kind of monitors
are being used?  I'd suggest, for initial testing with MST, making sure that
any external displays being used are 1080p monitors so that we don't run up
against configurations that won't be valid for the given link bandwidth.

Just for my own edification, can you please attach copies of the
i915_display_info and i915_dp_mst_info files from debugfs with a DUT->MST
Hub-><single 1080p monitor> configuration for me to look at?  Please be sure
that the single monitor is plugged into port 1 on the MST hub (many of the hubs
require this.)  Additionally, ensure that the MST hub is powered.  Some use an
external power supply, but some others expect to be powered off of a USB port
that is on the hub.

Please also test this on a more recent kernel; there have been several MST
fixes recently, some in the area of power management, that could be impacting
things here.

-- 
You are receiving this mail because:
You are the QA Contact for the bug.
You are on the CC list for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.freedesktop.org/archives/intel-gfx-bugs/attachments/20160425/00ed9e60/attachment.html>