I tried downloading and running the isolate on my Intel GPU laptop and Chrome instantly crashed with the following stack:

        base::debug::StackTrace::StackTrace [0x1032D640+32]
        base::debug::StackTrace::StackTrace [0x10303AAD+13]
        logging::LogMessage::~LogMessage [0x102A1643+83]
        ui::ResourceBundle::ResourceBundleImageSource::GetImageForScale [0x10391C10+446]
        gfx::internal::ImageSkiaStorage::FindRepresentation [0x1061566B+547]
        gfx::ImageSkia::GetRepresentation [0x10616334+120]
        gfx::ImageSkiaOperations::CreateIconWithBadge [0x10623DFC+2894]
        gfx::internal::ImageSkiaStorage::FindRepresentation [0x1061566B+547]
        gfx::internal::ImageSkiaStorage::FindRepresentation [0x106155ED+421]
        gfx::ImageSkia::GetRepresentation [0x10616334+120]
        gfx::Canvas::DrawImageInt [0x1061F219+27]
        gfx::Canvas::DrawImageInt [0x1061F1E3+73]
        TabIcon::OnPaint [0x115B1821+425]
        views::View::Paint [0x101CCD94+968]
        views::View::RecursivePaintHelper [0x101CE980+172]
        views::View::PaintChildren [0x101CE800+92]
        Tab::PaintChildren [0x11594521+131]
        views::View::Paint [0x101CCDBC+1008]
        TabStrip::PaintChildren [0x11578125+781]
        views::View::Paint [0x101CCDBC+1008]
        views::View::RecursivePaintHelper [0x101CE980+172]
        views::View::PaintChildren [0x101CE800+92]

I couldn't get to the point where I could run the test. This is actually a pretty significant regression because it affects the stability of any CQ jobs that run webgl2_conformance on Intel (maybe just ANGLE, but maybe also the optional GPU try servers?).

+bsalomon and fmalita from Skia team

Could this have been caused by recent Skia changes?

Tracking this down and fixing it is extremely urgent.

Nothing springs to mind. +reed@ as the ImageSkia classes in the stack look like they use the CPU backend.

I don't think there is much actual skia involved in the stack. Adding someone who works on ImageSkia itself.

Thanks.. I don't know if that crash stack shows up in the actual test failure, but it was very reproducible from downloading the isolate. FWIW test is being disabled in https://chromium-review.googlesource.com/c/chromium/src/+/956116 at which I think we can lower this to pri-2.

I believe this is caused by crbug.com/820318

Will mark Fail until issue 820318 is fixed

Thanks Yuly, but a CL is already in flight to disable the test, see comment #7.

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/c34306d894e84204efb7893f934416cf4435a191

commit c34306d894e84204efb7893f934416cf4435a191
Author: Luc Ferron <lucferron@chromium.org>
Date: Fri Mar 09 20:53:56 2018

Disable angle-stuck-depth-textures flaky test entirely

Bug:  820419 
Tbr: kbr@chromium.org
Cq-Include-Trybots: master.tryserver.chromium.android:android_optional_gpu_tests_rel;master.tryserver.chromium.linux:linux_optional_gpu_tests_rel;master.tryserver.chromium.mac:mac_optional_gpu_tests_rel;master.tryserver.chromium.win:win_optional_gpu_tests_rel
Change-Id: Ibd95ec218a700b7a9a6e34336f4ddee5442f8b02
Reviewed-on: https://chromium-review.googlesource.com/956116
Reviewed-by: Kenneth Russell <kbr@chromium.org>
Reviewed-by: Jamie Madill <jmadill@chromium.org>
Commit-Queue: Luc Ferron <lucferron@chromium.org>
Cr-Commit-Position: refs/heads/master@{#542209}
[modify] https://crrev.com/c34306d894e84204efb7893f934416cf4435a191/content/test/gpu/gpu_tests/webgl2_conformance_expectations.py

 Issue 820698  has been merged into this issue.

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/a9d705f793dfd55045cb801f80594b852e1e4c82

commit a9d705f793dfd55045cb801f80594b852e1e4c82
Author: Yuly Novikov <ynovikov@chromium.org>
Date: Wed Mar 21 20:50:06 2018

Revert "Disable angle-stuck-depth-textures flaky test entirely"

This reverts commit c34306d894e84204efb7893f934416cf4435a191.

Reason for revert: Win10 driver version was fixed

Original change's description:
> Disable angle-stuck-depth-textures flaky test entirely
> 
> Bug:  820419 
> Tbr: kbr@chromium.org
> Cq-Include-Trybots: master.tryserver.chromium.android:android_optional_gpu_tests_rel;master.tryserver.chromium.linux:linux_optional_gpu_tests_rel;master.tryserver.chromium.mac:mac_optional_gpu_tests_rel;master.tryserver.chromium.win:win_optional_gpu_tests_rel
> Change-Id: Ibd95ec218a700b7a9a6e34336f4ddee5442f8b02
> Reviewed-on: https://chromium-review.googlesource.com/956116
> Reviewed-by: Kenneth Russell <kbr@chromium.org>
> Reviewed-by: Jamie Madill <jmadill@chromium.org>
> Commit-Queue: Luc Ferron <lucferron@chromium.org>
> Cr-Commit-Position: refs/heads/master@{#542209}

TBR=jmadill@chromium.org,kbr@chromium.org,lucferron@chromium.org

# Not skipping CQ checks because original CL landed > 1 day ago.

Bug:  820419 
Change-Id: I3b86e2249440b195353583348b0c1fae010e3965
Cq-Include-Trybots: master.tryserver.chromium.android:android_optional_gpu_tests_rel;master.tryserver.chromium.linux:linux_optional_gpu_tests_rel;master.tryserver.chromium.mac:mac_optional_gpu_tests_rel;master.tryserver.chromium.win:win_optional_gpu_tests_rel
Reviewed-on: https://chromium-review.googlesource.com/973981
Reviewed-by: Yuly Novikov <ynovikov@chromium.org>
Commit-Queue: Yuly Novikov <ynovikov@chromium.org>
Cr-Commit-Position: refs/heads/master@{#544832}
[modify] https://crrev.com/a9d705f793dfd55045cb801f80594b852e1e4c82/content/test/gpu/gpu_tests/webgl2_conformance_expectations.py

No new failures since revert, looks like this was indeed a driver issue.