NAK: [Utopic SRU][PATCH] Backport of DisplayPort 1.2 MST support

[Tim Gardner]

On 11/03/2014 06:00 AM, Dariusz Gadomski wrote:
> BugLink: https://bugs.launchpad.net/bugs/1104230
> 
> This patch series is a cherry-pick of David Airlie's work on backporting
> support for the DisplayPort 1.2 feature to 3.16 kernel.
> 
> This work is available in his tree git://people.freedesktop.org/~airlied/linux
> in the drm-i915-mst-v3.16 branch. All the changes are already included in the
> mainline kernel v3.17, cherry-picking from David's branch was chosen because
> he has resolved some conflicts caused by differences between v3.16 and v3.17.
> 
> SRU Justification:
> 
> Impact: Many users expect the possibility to use the DP1.2 MST feature which
> enables them to use a number of different display setups: daisy-chaining
> DP1.2 displays, some 4K displays appearing as 2 separate displays, 3D displays.
> This however is not supported in Ubuntu stable kernel.
> 
> Fix: A series of 18 cherry-picks from v3.17 kernel backported to v3.16 by
> the original author of the feature: David Airlie.
> 
> Regression Potential: These patches were merged upstream in v3.17 without any
> reported regressions.
> 
> 
> Chris Wilson (1):
>       drm/i915: only hook up hpd pulse for DP outputs
> 
> Dave Airlie (17):
>       drm/i915: add some registers need for displayport MST support.
>       drm: fix memory leak around mode_group (v2)
>       drm/crtc: add interface to reinitialise the legacy mode group
>       drm/fb_helper: allow adding/removing connectors later
>       drm: add a path blob property
>       drm/i915: rework digital port IRQ handling
>       drm/helper: add Displayport multi-stream helper (v0.6)
>       i915: split some DP modesetting code into a separate function
>       drm/i915: check connector->encoder before using it.
>       drm/i915: split some conversion functions out into separate functions.
>       i915: add DP 1.2 MST support (v0.6)
>       i915: mst topology dumper in debugfs (v0.2)
>       backport v0.7 mst core fixes
>       i915: fixup
>       put back missing modeset
>       drm/i915: lock around link status and link training.
>       fix warning
> 
>  Documentation/DocBook/drm.tmpl        |   6 +
>  drivers/gpu/drm/i915/Makefile         |   1 +
>  drivers/gpu/drm/i915/i915_debugfs.c   |  23 ++
>  drivers/gpu/drm/i915/i915_dma.c       |  10 +
>  drivers/gpu/drm/i915/i915_drv.c       |  15 +-
>  drivers/gpu/drm/i915/i915_drv.h       |  14 +
>  drivers/gpu/drm/i915/i915_irq.c       | 146 ++++++++-
>  drivers/gpu/drm/i915/i915_reg.h       |  11 +-
>  drivers/gpu/drm/i915/intel_ddi.c      | 155 +++++++---
>  drivers/gpu/drm/i915/intel_display.c  |  61 +++-
>  drivers/gpu/drm/i915/intel_dp.c       | 246 +++++++++++++++-
>  drivers/gpu/drm/i915/intel_dp_mst.c   | 536 ++++++++++++++++++++++++++++++++++
>  drivers/gpu/drm/i915/intel_drv.h      |  47 ++-
>  drivers/gpu/drm/i915/intel_opregion.c |   1 +
>  14 files changed, 1192 insertions(+), 80 deletions(-)
> 

Of the 5 patches that I spot checked, none were from Linus' repo, and
are therefore _not_ upstream. Released distro kernels (such as Utopic)
largely adhere to upstream stable rules
(Documentation/stable_kernel_rules.txt). Even if these patches were
clean cherry-picks I'd be loathe to accept a patch set of this magnitude
into a stable kernel without extensive regression testing. Test results
from the bug report don't look encouraging.

rtg
-- 
Tim Gardner tim.gardner at canonical.com