jmadil@, geofflang@,

I am really surprised that the newly added conformance test for this issue (https://www.khronos.org/registry/webgl/sdk/tests/conformance2/rendering/framebuffer-unsupported.html?webglVersion=2&quiet=0) failed on Windows, i.e., when we attach the same 2D texture mip level to two color attachment points, the checkFramebufferStatus returns FRAMEBUFFER_COMPLETE.

I thought D3D does not allow that.  Does that mean the failure returns somewhere else (draw call time) rather than in checkFramebufferStatus?

Please advise.

Reassigning to jmadill@ to investigate.

I'll grab this, I was working on it recently.

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/db975cd18716681f50db94d36120ba50148604fe

commit db975cd18716681f50db94d36120ba50148604fe
Author: qiankun.miao <qiankun.miao@intel.com>
Date: Wed Jul 20 18:00:19 2016

Attaching an image to many color attachment points is unsupported

See Section 5.34 Framebuffer color attachments in WebGL 2.0 spec.

BUG= 628861 
TEST=gpu_unittests conformance2/rendering/framebuffer-unsupported.html
CQ_INCLUDE_TRYBOTS=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

Review-Url: https://codereview.chromium.org/2159183002
Cr-Commit-Position: refs/heads/master@{#406613}

[modify] https://crrev.com/db975cd18716681f50db94d36120ba50148604fe/gpu/command_buffer/service/feature_info.cc
[modify] https://crrev.com/db975cd18716681f50db94d36120ba50148604fe/gpu/command_buffer/service/feature_info.h
[modify] https://crrev.com/db975cd18716681f50db94d36120ba50148604fe/gpu/command_buffer/service/framebuffer_manager.cc
[modify] https://crrev.com/db975cd18716681f50db94d36120ba50148604fe/gpu/command_buffer/service/framebuffer_manager.h
[modify] https://crrev.com/db975cd18716681f50db94d36120ba50148604fe/gpu/command_buffer/service/framebuffer_manager_unittest.cc