On Linux, it looks like we can get color profile information from two places:

1. Parsing the EDID info at
https://cs.chromium.org/chromium/src/ui/display/util/edid_parser.cc?rcl=6a22ae37ea816b3dd750a17fe91d23e562cdb671&l=61

2. XGetWindowProperty for _ICC_PROFILE as in ICCProfile::FromBestMonitor, somewhere in DesktopScreenX11::BuildDisplaysFromXRandRInfo().

We should also be parsing_ICC_PROFILE_N to get data for secondary monitors.

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/0269cae2f81d9920c4371e9fab0af5b569ddbe91

commit 0269cae2f81d9920c4371e9fab0af5b569ddbe91
Author: ccameron <ccameron@chromium.org>
Date: Fri Jun 23 22:49:02 2017

color: Use ICCProfile from display::Display instead of FromBestMonitor

Remove aura::WindowTreeHost::GetICCProfileForCurrentDisplay, since
this can be inferred from display::Screen::GetDisplayNearestWindow.

This sets us up to notify the renderer and window of their new color
space when moved across displays.

Still rely on ICCProfile::FromBestMonitor for now, since removing that
will require attention to detail on all platforms.

R=hubbe
TBR=skyostil (headless/)
BUG= 735613 
CQ_INCLUDE_TRYBOTS=master.tryserver.blink:linux_trusty_blink_rel

Review-Url: https://codereview.chromium.org/2950183002
Cr-Commit-Position: refs/heads/master@{#482056}

[modify] https://crrev.com/0269cae2f81d9920c4371e9fab0af5b569ddbe91/cc/surfaces/surface_aggregator.cc
[modify] https://crrev.com/0269cae2f81d9920c4371e9fab0af5b569ddbe91/cc/surfaces/surface_aggregator.h
[modify] https://crrev.com/0269cae2f81d9920c4371e9fab0af5b569ddbe91/content/browser/renderer_host/browser_compositor_view_mac.h
[modify] https://crrev.com/0269cae2f81d9920c4371e9fab0af5b569ddbe91/content/browser/renderer_host/browser_compositor_view_mac.mm
[modify] https://crrev.com/0269cae2f81d9920c4371e9fab0af5b569ddbe91/content/browser/renderer_host/render_widget_host_view_mac.mm
[modify] https://crrev.com/0269cae2f81d9920c4371e9fab0af5b569ddbe91/content/browser/web_contents/web_contents_view_aura.cc
[modify] https://crrev.com/0269cae2f81d9920c4371e9fab0af5b569ddbe91/content/browser/web_contents/web_contents_view_mac.mm
[modify] https://crrev.com/0269cae2f81d9920c4371e9fab0af5b569ddbe91/headless/lib/browser/headless_window_tree_host.cc
[modify] https://crrev.com/0269cae2f81d9920c4371e9fab0af5b569ddbe91/headless/lib/browser/headless_window_tree_host.h
[modify] https://crrev.com/0269cae2f81d9920c4371e9fab0af5b569ddbe91/ui/aura/mus/window_tree_host_mus.cc
[modify] https://crrev.com/0269cae2f81d9920c4371e9fab0af5b569ddbe91/ui/aura/mus/window_tree_host_mus.h
[modify] https://crrev.com/0269cae2f81d9920c4371e9fab0af5b569ddbe91/ui/aura/window_tree_host.cc
[modify] https://crrev.com/0269cae2f81d9920c4371e9fab0af5b569ddbe91/ui/aura/window_tree_host.h
[modify] https://crrev.com/0269cae2f81d9920c4371e9fab0af5b569ddbe91/ui/compositor/compositor.cc
[modify] https://crrev.com/0269cae2f81d9920c4371e9fab0af5b569ddbe91/ui/compositor/compositor.h
[modify] https://crrev.com/0269cae2f81d9920c4371e9fab0af5b569ddbe91/ui/display/display.cc
[modify] https://crrev.com/0269cae2f81d9920c4371e9fab0af5b569ddbe91/ui/display/display.h
[modify] https://crrev.com/0269cae2f81d9920c4371e9fab0af5b569ddbe91/ui/display/mac/screen_mac.mm
[modify] https://crrev.com/0269cae2f81d9920c4371e9fab0af5b569ddbe91/ui/display/win/screen_win.cc
[modify] https://crrev.com/0269cae2f81d9920c4371e9fab0af5b569ddbe91/ui/gfx/color_space.cc
[modify] https://crrev.com/0269cae2f81d9920c4371e9fab0af5b569ddbe91/ui/gfx/color_space.h
[modify] https://crrev.com/0269cae2f81d9920c4371e9fab0af5b569ddbe91/ui/gfx/icc_profile.cc
[modify] https://crrev.com/0269cae2f81d9920c4371e9fab0af5b569ddbe91/ui/gfx/icc_profile.h
[modify] https://crrev.com/0269cae2f81d9920c4371e9fab0af5b569ddbe91/ui/views/widget/desktop_aura/desktop_screen_x11.cc

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/ab0aab82705d56bc09d6a4fe0a91eb43a610f30e

commit ab0aab82705d56bc09d6a4fe0a91eb43a610f30e
Author: ccameron <ccameron@chromium.org>
Date: Sat Jun 24 17:21:54 2017

color: Replace more instances of ICCProfile with ColorSpace

The gfx::ColorSpace representation is more compact and can communicate
information such as HDR status.

TBR=hubbe
BUG= 735613 

Review-Url: https://codereview.chromium.org/2954793002
Cr-Commit-Position: refs/heads/master@{#482155}

[modify] https://crrev.com/ab0aab82705d56bc09d6a4fe0a91eb43a610f30e/content/browser/renderer_host/render_widget_host_impl.cc
[modify] https://crrev.com/ab0aab82705d56bc09d6a4fe0a91eb43a610f30e/content/browser/web_contents/web_contents_view_aura.cc
[modify] https://crrev.com/ab0aab82705d56bc09d6a4fe0a91eb43a610f30e/content/browser/web_contents/web_contents_view_mac.mm
[modify] https://crrev.com/ab0aab82705d56bc09d6a4fe0a91eb43a610f30e/content/common/view_messages.h
[modify] https://crrev.com/ab0aab82705d56bc09d6a4fe0a91eb43a610f30e/content/public/common/screen_info.cc
[modify] https://crrev.com/ab0aab82705d56bc09d6a4fe0a91eb43a610f30e/content/public/common/screen_info.h
[modify] https://crrev.com/ab0aab82705d56bc09d6a4fe0a91eb43a610f30e/content/renderer/render_view_impl.cc
[modify] https://crrev.com/ab0aab82705d56bc09d6a4fe0a91eb43a610f30e/content/renderer/render_widget.cc
[modify] https://crrev.com/ab0aab82705d56bc09d6a4fe0a91eb43a610f30e/third_party/WebKit/Source/platform/graphics/ColorSpaceGamut.cpp
[modify] https://crrev.com/ab0aab82705d56bc09d6a4fe0a91eb43a610f30e/third_party/WebKit/public/platform/WebScreenInfo.h

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/942e74b0d721b0090cd4b30f8a31f04adf005fff

commit 942e74b0d721b0090cd4b30f8a31f04adf005fff
Author: ccameron <ccameron@chromium.org>
Date: Tue Jun 27 02:27:16 2017

color: Add NSScreenColorSpaceDidChangeNotification observer

Change RenderWidgetHostViewMac to not query the display color profile
directly, rather, use the display::Display screen information.

This exposed a bug whereby we weren't updating the display::Display
color space when NSScreens changed color spaces.

Clean up RenderWidgetHostViewMac to make UpdateBackingStoreProperties
call UpdateScreenInfo and update color space information, and change
calls to UpdateScreenInfo to go through UpdateBackingStoreProperties
appropriately.

Also make BrowserCompositorMac query the current display scale factor
and color space from the display::Display when activating.

BUG= 735613 

Review-Url: https://codereview.chromium.org/2959873002
Cr-Commit-Position: refs/heads/master@{#482518}

[modify] https://crrev.com/942e74b0d721b0090cd4b30f8a31f04adf005fff/content/browser/renderer_host/browser_compositor_view_mac.mm
[modify] https://crrev.com/942e74b0d721b0090cd4b30f8a31f04adf005fff/content/browser/renderer_host/render_widget_host_view_base.cc
[modify] https://crrev.com/942e74b0d721b0090cd4b30f8a31f04adf005fff/content/browser/renderer_host/render_widget_host_view_base.h
[modify] https://crrev.com/942e74b0d721b0090cd4b30f8a31f04adf005fff/content/browser/renderer_host/render_widget_host_view_mac.mm
[modify] https://crrev.com/942e74b0d721b0090cd4b30f8a31f04adf005fff/ui/display/mac/screen_mac.mm

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/8f608ad3382a72a4d11489fccbe895c473ae9992

commit 8f608ad3382a72a4d11489fccbe895c473ae9992
Author: ccameron <ccameron@chromium.org>
Date: Thu Jul 06 00:07:48 2017

color: Read all displays' color profiles on Windows

This restores code that was unused and removed in crrev.com/334309.

Enumerate all monitors, and read all of their ICC profiles, instead of
just reading the profile for the "best monitor".

Remove the ScreenWin::Initialize virtual method because, being called
from the constructor, being virtual had no effect. Replace with a
constructor argument to call initialize.

Don't actually hook any of this up yet -- still only return the profile
for the "best monitor", but do that by looking it up in the full cache
that has been built.

TBR=hubbe
BUG= 735613 

Review-Url: https://codereview.chromium.org/2959423003
Cr-Commit-Position: refs/heads/master@{#484376}

[modify] https://crrev.com/8f608ad3382a72a4d11489fccbe895c473ae9992/ui/display/BUILD.gn
[add] https://crrev.com/8f608ad3382a72a4d11489fccbe895c473ae9992/ui/display/win/color_profile_reader.cc
[add] https://crrev.com/8f608ad3382a72a4d11489fccbe895c473ae9992/ui/display/win/color_profile_reader.h
[modify] https://crrev.com/8f608ad3382a72a4d11489fccbe895c473ae9992/ui/display/win/screen_win.cc
[modify] https://crrev.com/8f608ad3382a72a4d11489fccbe895c473ae9992/ui/display/win/screen_win.h
[modify] https://crrev.com/8f608ad3382a72a4d11489fccbe895c473ae9992/ui/display/win/screen_win_unittest.cc

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/39b1cbc91fa0f3e57eb4f1c2f3807e24af82657a

commit 39b1cbc91fa0f3e57eb4f1c2f3807e24af82657a
Author: ccameron <ccameron@chromium.org>
Date: Thu Jul 06 00:23:55 2017

color: Use ColorSpace for plumbing

Clear up more instances where ICCProfile was used for plumbing when a
ColorSpace would be better.

TBR=hubbe
BUG= 735613 

Review-Url: https://codereview.chromium.org/2970183002
Cr-Commit-Position: refs/heads/master@{#484409}

[modify] https://crrev.com/39b1cbc91fa0f3e57eb4f1c2f3807e24af82657a/content/public/test/layouttest_support.h
[modify] https://crrev.com/39b1cbc91fa0f3e57eb4f1c2f3807e24af82657a/content/renderer/render_view_impl.cc
[modify] https://crrev.com/39b1cbc91fa0f3e57eb4f1c2f3807e24af82657a/content/renderer/render_view_impl.h
[modify] https://crrev.com/39b1cbc91fa0f3e57eb4f1c2f3807e24af82657a/content/shell/renderer/layout_test/blink_test_runner.cc
[modify] https://crrev.com/39b1cbc91fa0f3e57eb4f1c2f3807e24af82657a/content/shell/renderer/layout_test/blink_test_runner.h
[modify] https://crrev.com/39b1cbc91fa0f3e57eb4f1c2f3807e24af82657a/content/shell/test_runner/test_runner.cc
[modify] https://crrev.com/39b1cbc91fa0f3e57eb4f1c2f3807e24af82657a/content/shell/test_runner/test_runner_for_specific_view.cc
[modify] https://crrev.com/39b1cbc91fa0f3e57eb4f1c2f3807e24af82657a/content/shell/test_runner/web_test_delegate.h
[modify] https://crrev.com/39b1cbc91fa0f3e57eb4f1c2f3807e24af82657a/content/test/layouttest_support.cc

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/3394d6ecfb7bf7742a2b9b65fbf6be79f41a1bf5

commit 3394d6ecfb7bf7742a2b9b65fbf6be79f41a1bf5
Author: Christopher Cameron <ccameron@chromium.org>
Date: Wed Jul 12 04:12:32 2017

color: Use ColorProfileReader on Windows

This was already present, now use it to populate the information read
by the compositor and renderers.

Bug:  735613 
Change-Id: Id5c85b2bc3a0891e35760dd27b65d85c898d9176
Reviewed-on: https://chromium-review.googlesource.com/564392
Reviewed-by: Fredrik Hubinette <hubbe@chromium.org>
Commit-Queue: ccameron chromium <ccameron@chromium.org>
Cr-Commit-Position: refs/heads/master@{#485841}
[modify] https://crrev.com/3394d6ecfb7bf7742a2b9b65fbf6be79f41a1bf5/ui/display/win/screen_win.cc

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/a4adb41bd80688eb747c0c8b8c4cfc8e37348966

commit a4adb41bd80688eb747c0c8b8c4cfc8e37348966
Author: Christopher Cameron <ccameron@chromium.org>
Date: Thu Jul 13 06:44:25 2017

color: Clean up ICC approxiation code

This was too many long-and-nested if-else blocks. Turn the bulk of the
work into a separate function.

The impetus of this was to ensure that when we parse an ICC profile, the
gfx::ColorSpace that we return is gfx::ColorSpace::CreateSRGB. The
reason for being concerned about this is that we want to make sure
that, on Windows, when we read the ICC profile from the system and find
that it was sRGB all along, we waste time re-sending that to renderers.

Bug:  735613 
Change-Id: I2977e951c198eda6cb28bb1cf1524a8ef99eacfe
Reviewed-on: https://chromium-review.googlesource.com/568244
Reviewed-by: Fredrik Hubinette <hubbe@chromium.org>
Commit-Queue: ccameron chromium <ccameron@chromium.org>
Cr-Commit-Position: refs/heads/master@{#486295}
[modify] https://crrev.com/a4adb41bd80688eb747c0c8b8c4cfc8e37348966/ui/gfx/color_space.cc
[modify] https://crrev.com/a4adb41bd80688eb747c0c8b8c4cfc8e37348966/ui/gfx/color_space.h
[modify] https://crrev.com/a4adb41bd80688eb747c0c8b8c4cfc8e37348966/ui/gfx/icc_profile.cc
[modify] https://crrev.com/a4adb41bd80688eb747c0c8b8c4cfc8e37348966/ui/gfx/icc_profile_unittest.cc

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/1f3a4e3e534888197094bb0f62f80e4449d17fe9

commit 1f3a4e3e534888197094bb0f62f80e4449d17fe9
Author: Christopher Cameron <ccameron@chromium.org>
Date: Tue Jul 18 06:16:01 2017

color: Make ui::Compositor receive color profile updates

RenderWidgetHostViewAura currently listens appropriately, but
the message is not sent to the compositor, so we always composite
back into whatever the default color profile was when the window
was created. Make WindowTreeHost be a DisplayObserver to fix this.

Bug:  735613 
Change-Id: Id560d5ec0e2f2083f3218b6b68ed7fb4ac9dea75
Reviewed-on: https://chromium-review.googlesource.com/569140
Commit-Queue: ccameron chromium <ccameron@chromium.org>
Reviewed-by: Fredrik Hubinette <hubbe@chromium.org>
Cr-Commit-Position: refs/heads/master@{#487395}
[modify] https://crrev.com/1f3a4e3e534888197094bb0f62f80e4449d17fe9/ui/aura/window_tree_host.cc
[modify] https://crrev.com/1f3a4e3e534888197094bb0f62f80e4449d17fe9/ui/aura/window_tree_host.h
[modify] https://crrev.com/1f3a4e3e534888197094bb0f62f80e4449d17fe9/ui/views/widget/desktop_aura/desktop_window_tree_host_x11.cc
[modify] https://crrev.com/1f3a4e3e534888197094bb0f62f80e4449d17fe9/ui/views/widget/desktop_aura/desktop_window_tree_host_x11.h

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/2bc5e180f592aace72a4b65f83b462265634d4c6

commit 2bc5e180f592aace72a4b65f83b462265634d4c6
Author: Christopher Cameron <ccameron@chromium.org>
Date: Fri Jul 21 01:37:54 2017

cc: Use sRGB rasterization when software compositing

Color conversion is not efficient enough to be done during software
compositing. Rasterize everything into sRGB for consistency.

Bug:  735613 
Change-Id: Ib977ace62f6e4b560312e5322afc4d1f6c45601d
Reviewed-on: https://chromium-review.googlesource.com/576829
Reviewed-by: Fredrik Hubinette <hubbe@chromium.org>
Commit-Queue: ccameron chromium <ccameron@chromium.org>
Cr-Commit-Position: refs/heads/master@{#488541}
[modify] https://crrev.com/2bc5e180f592aace72a4b65f83b462265634d4c6/cc/trees/layer_tree_host_impl.cc
[modify] https://crrev.com/2bc5e180f592aace72a4b65f83b462265634d4c6/cc/trees/layer_tree_host_impl_unittest.cc

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/284086f9b8dc0a7aa5a411166c74822a97133400

commit 284086f9b8dc0a7aa5a411166c74822a97133400
Author: Christopher Cameron <ccameron@chromium.org>
Date: Thu Aug 03 05:37:13 2017

color: Ensure window tree host gets updates

This logic was incorrectly inverted.

Bug:  735613 
Change-Id: Ib07302612c02e860c006f0cc2e9929a04f5de009
Reviewed-on: https://chromium-review.googlesource.com/596710
Commit-Queue: ccameron chromium <ccameron@chromium.org>
Reviewed-by: Fredrik Hubinette <hubbe@chromium.org>
Cr-Commit-Position: refs/heads/master@{#491655}
[modify] https://crrev.com/284086f9b8dc0a7aa5a411166c74822a97133400/ui/aura/test/test_screen.cc
[modify] https://crrev.com/284086f9b8dc0a7aa5a411166c74822a97133400/ui/aura/test/test_screen.h
[modify] https://crrev.com/284086f9b8dc0a7aa5a411166c74822a97133400/ui/aura/window_tree_host.cc
[modify] https://crrev.com/284086f9b8dc0a7aa5a411166c74822a97133400/ui/aura/window_tree_host_unittest.cc
[modify] https://crrev.com/284086f9b8dc0a7aa5a411166c74822a97133400/ui/compositor/compositor.h

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/fd90d276245b7552cfd1bf3855ea668c62362a6a

commit fd90d276245b7552cfd1bf3855ea668c62362a6a
Author: Christopher Cameron <ccameron@chromium.org>
Date: Fri Sep 22 08:11:33 2017

Update color profile when moving window tree

To DesktopWindowTreeHostWin, add CheckForMonitorChange, a
lightweight check to see if the HWND has changed the
HMONITOR it is on. Call this method when resizing or
moving the window.

If this method sees that the HMONITOR has changed, call
the method aura::WindowTreeHost:OnHostDisplayChanged,
which will update the compositor color space based on the
more expensive call to query the current display::Display.

Bug:  735613 
Change-Id: I59e4a79cad599f70e63e898f4a7eacc2dbadc6d2
Reviewed-on: https://chromium-review.googlesource.com/674048
Reviewed-by: Scott Violet <sky@chromium.org>
Commit-Queue: ccameron chromium <ccameron@chromium.org>
Cr-Commit-Position: refs/heads/master@{#503684}
[modify] https://crrev.com/fd90d276245b7552cfd1bf3855ea668c62362a6a/ui/aura/window_tree_host.cc
[modify] https://crrev.com/fd90d276245b7552cfd1bf3855ea668c62362a6a/ui/aura/window_tree_host.h
[modify] https://crrev.com/fd90d276245b7552cfd1bf3855ea668c62362a6a/ui/views/widget/desktop_aura/desktop_window_tree_host_win.cc
[modify] https://crrev.com/fd90d276245b7552cfd1bf3855ea668c62362a6a/ui/views/widget/desktop_aura/desktop_window_tree_host_win.h

The bug is marked as P3 or Feature. It should not be merged as M62 is in beta. 
Please contact the approriate milestone owner if you have questions.
Owners: amineer@(Android), cmasso@(iOS), bhthompson@(ChromeOS), abdulsyed@(Desktop)

For more details visit https://www.chromium.org/issue-tracking/autotriage - Your friendly Sheriffbot

This was a pretty big omission for Windows color rendering and should be in 62.

To be clear, the only patch that needs merge to 62 is #13 -- the other parts are already on the branch.

This bug requires manual review: M62 has already been promoted to the beta branch, so this requires manual review
Please contact the milestone owner if you have questions.
Owners: amineer@(Android), cmasso@(iOS), bhthompson@(ChromeOS), abdulsyed@(Desktop)

For more details visit https://www.chromium.org/issue-tracking/autotriage - Your friendly Sheriffbot

Confirmed with ccameron@, merges are safe and tested in Canary. Approving merge to M62. branch:3202

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/c9f60e2e1e45304339011375e37dc23144f4ef7c

commit c9f60e2e1e45304339011375e37dc23144f4ef7c
Author: Christopher Cameron <ccameron@chromium.org>
Date: Fri Sep 29 05:49:43 2017

Update color profile when moving window tree

To DesktopWindowTreeHostWin, add CheckForMonitorChange, a
lightweight check to see if the HWND has changed the
HMONITOR it is on. Call this method when resizing or
moving the window.

If this method sees that the HMONITOR has changed, call
the method aura::WindowTreeHost:OnHostDisplayChanged,
which will update the compositor color space based on the
more expensive call to query the current display::Display.

TBR=ccameron@chromium.org

(cherry picked from commit fd90d276245b7552cfd1bf3855ea668c62362a6a)

Bug:  735613 
Change-Id: I59e4a79cad599f70e63e898f4a7eacc2dbadc6d2
Reviewed-on: https://chromium-review.googlesource.com/674048
Reviewed-by: Scott Violet <sky@chromium.org>
Commit-Queue: ccameron chromium <ccameron@chromium.org>
Cr-Original-Commit-Position: refs/heads/master@{#503684}
Reviewed-on: https://chromium-review.googlesource.com/691439
Reviewed-by: ccameron chromium <ccameron@chromium.org>
Cr-Commit-Position: refs/branch-heads/3202@{#503}
Cr-Branched-From: fa6a5d87adff761bc16afc5498c3f5944c1daa68-refs/heads/master@{#499098}
[modify] https://crrev.com/c9f60e2e1e45304339011375e37dc23144f4ef7c/ui/aura/window_tree_host.cc
[modify] https://crrev.com/c9f60e2e1e45304339011375e37dc23144f4ef7c/ui/aura/window_tree_host.h
[modify] https://crrev.com/c9f60e2e1e45304339011375e37dc23144f4ef7c/ui/views/widget/desktop_aura/desktop_window_tree_host_win.cc
[modify] https://crrev.com/c9f60e2e1e45304339011375e37dc23144f4ef7c/ui/views/widget/desktop_aura/desktop_window_tree_host_win.h

Could this cause high cpu usage? Recently I bisected high CPU usage in google maps to several commits, one of them https://chromium.googlesource.com/chromium/src/+/3394d6ecfb7bf7742a2b9b65fbf6be79f41a1bf5. I've reviewed the commits in the blame range and none of them really seems possible to cause the problem except for the V8 rollup, or maybe this. Can someone look at https://bugs.chromium.org/p/chromium/issues/detail?id=770054

 Issue 779510  has been merged into this issue.