Actually, I'll add the suppression in https://codereview.chromium.org/2551753002/ .

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/73a3e000076a895925192693c423a838cfad3da1

commit 73a3e000076a895925192693c423a838cfad3da1
Author: kbr <kbr@chromium.org>
Date: Sun Dec 04 23:50:16 2016

Suppress virtual/color_space/fast/canvas/color-space/display_linear-rgb.html on Windows.

It's failing on most try jobs, causing retries and slowing down the CQ.

BUG= 671048 
TBR=dpranke@chromium.org, zakerinasab@chromium.org
NOTRY=true

Review-Url: https://codereview.chromium.org/2551753002
Cr-Commit-Position: refs/heads/master@{#436207}

[modify] https://crrev.com/73a3e000076a895925192693c423a838cfad3da1/third_party/WebKit/LayoutTests/TestExpectations

Assuming the above suppression worked, this is no longer P1.

It's failing intermittently on Linux too.

https://build.chromium.org/p/tryserver.chromium.linux/builders/linux_chromium_rel_ng/builds/352265

Layout test results are here:

https://storage.googleapis.com/chromium-layout-test-archives/linux_chromium_rel_ng/352265/layout-test-results/results.html

It seems clear that the actual image is occasionally coming up blank. I'm marking this as an expected failure on Linux too.

Since the test will be an expected failure on all desktop platforms, it isn't offering any coverage. CC'ing more people in case quick action should be taken about this.

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/e5f335df7cb9720bd2d0be138d1a42e591e635de

commit e5f335df7cb9720bd2d0be138d1a42e591e635de
Author: kbr <kbr@chromium.org>
Date: Thu Dec 08 08:00:10 2016

Suppress virtual/color_space/fast/canvas/color-space/display_linear-rgb.html on Linux.

BUG= 671048 
TBR=dpranke@chromium.org

Review-Url: https://codereview.chromium.org/2559103002
Cr-Commit-Position: refs/heads/master@{#437207}

[modify] https://crrev.com/e5f335df7cb9720bd2d0be138d1a42e591e635de/third_party/WebKit/LayoutTests/TestExpectations

This might be flaky as different people are working on different parts of color correction in Chrome and it is possible that someone has changed something that has affected this test. Thanks for taking care of suppressing the test. I'll look into it.

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/0934442edc29b70e892d6800fd640ce2427c355c

commit 0934442edc29b70e892d6800fd640ce2427c355c
Author: qyearsley <qyearsley@chromium.org>
Date: Fri Mar 17 00:19:55 2017

Update test expectations of tests that unexpectedly pass.

BUG= 671048 , 417782 ,626703

Review-Url: https://codereview.chromium.org/2754763003
Cr-Commit-Position: refs/heads/master@{#457621}

[modify] https://crrev.com/0934442edc29b70e892d6800fd640ce2427c355c/third_party/WebKit/LayoutTests/TestExpectations

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/1e85d5d4d0841f935a43b3b3dd60b517139a5b39

commit 1e85d5d4d0841f935a43b3b3dd60b517139a5b39
Author: dimich <dimich@chromium.org>
Date: Sat Mar 18 00:18:15 2017

Add new mode of failure for this test. It is known to be flaky, now it also crashes sometimes.
Attaching to existing bug.

Snippet from https://uberchromegw.corp.google.com/i/chromium.webkit/builders/WebKit%20Mac10.11%20%28dbg%29/builds/7916:

virtual/color_space/fast/canvas/color-space/display_linear-rgb.html [ Crash ]
[6080:775:0317/135733.280810:251425167413:INFO:SkImage_Raster.cpp(385)] ../../third_party/skia/src/image/SkImage_Raster.cpp:385: fatal error: "assert(dst.writePixels(src))"

BUG= 671048 
TBR=zakerinasab@chromium.org

Review-Url: https://codereview.chromium.org/2755243002
Cr-Commit-Position: refs/heads/master@{#457909}

[modify] https://crrev.com/1e85d5d4d0841f935a43b3b3dd60b517139a5b39/third_party/WebKit/LayoutTests/TestExpectations

This call fails when color space is enabled (i.e. in the virtual-color_space test originally reported for this bug):

getContext('2d', {colorSpace: 'srgb', pixelFormat: 'float16'})

More precisely, you might have to draw something as the pixelFormat does appear in some other tests.

Skia asserts as listed in comment #10. This is 100% reproducible locally for me on linux. It appears to be a real bug revealed by the test failure.

I cannot reproduce this on Linux + ToT, also the behavior is still flaky. Earlier past week the rectangles were visible when the context2d was created with float16 CanvasPixelFormat. Now the rectangles disappeared again. However, neither chrome nor content_shell crash on my machine and I can't see any assertion from Skia. In fact, this assertion is not in SkImage_Raster.cpp anymore. I guess the flaky behavior of this test will be around for some time.

@Matt: Do you have a guess why the rectangles have disappeared again? Can it be related to some recent change in Skia?

After https://codereview.chromium.org/2797213002/ is landed, getImageData from wide gamut 2D context must work fine, but we still do not have an end-to-end pixel test or something. A proper way to do this can be as follows:
* Draw the colored rectangles on a color managed canvas, let's say in linear SRGB.
* Load a png image of the same rectangles on another canvas in the same color space.
* Call getImageData and check the F32 values of the pixels.

This requires fixing drawImage. I'm not sure what is the current state of drawImage, but I'll come back to this bug after figuring that out.

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/7d22b143fe3412d6423de6bdb8c7afa6d67bb1b1

commit 7d22b143fe3412d6423de6bdb8c7afa6d67bb1b1
Author: Reza.Zakerinasab <zakerinasab@chromium.org>
Date: Fri Aug 25 20:16:09 2017

Reactivate display_linear-rgb.html layout test

This change reactivtes display_linaer-rgb.html test.

Bug:  671048 
Change-Id: Id224474593dcd307a3c90f301cc500a42d0e3901
Reviewed-on: https://chromium-review.googlesource.com/636036
Reviewed-by: Xida Chen <xidachen@chromium.org>
Reviewed-by: Justin Novosad <junov@chromium.org>
Commit-Queue: Justin Novosad <junov@chromium.org>
Cr-Commit-Position: refs/heads/master@{#497515}
[modify] https://crrev.com/7d22b143fe3412d6423de6bdb8c7afa6d67bb1b1/third_party/WebKit/LayoutTests/TestExpectations