NEW 207858
Changed results due to ANGLE use 
https://bugs.webkit.org/show_bug.cgi?id=207858
Summary Changed results due to ANGLE use
Dean Jackson
Reported 2020-02-17 13:46:44 PST
While switching over to ANGLE, the following tests required something more than a rebaseline. macOS: fast/canvas/webgl/readPixels-float.html fast/canvas/webgl/draw-elements-out-of-bounds-uint-index.html fast/canvas/webgl/simulated-vertexAttrib0-invalid-indicies.html fast/canvas/webgl/webgl2-texture-upload-enums.html fast/canvas/webgl/texImage2D-video-flipY-false.html (times out) iOS Simulator: fast/canvas/webgl/context-attributes-alpha.html (ref test) fast/canvas/webgl/match-page-color-space.html (ref test) webgl/smell-test.html (ref test) webgl/webgl-backing-store-size-update.html (ref test) webgl/webgl-border.html (ref test) webgl/webgl-box-shadow.html (ref test) webgl/misc/switch-to-discrete-gpu.html (why is this even running on iOS?) fast/canvas/webgl/oes-texture-half-float-uint16.html fast/canvas/webgl/uninitialized-test.html fast/canvas/webgl/webgl-depth-texture.html (marked as not supported! why?) fast/canvas/webgl/webgl-texture-binding-preserved.html webgl/1.0.3/conformance/canvas/texture-bindings-unaffected-on-resize.html webgl/1.0.3/conformance/context/context-lost-restored.html webgl/1.0.3/conformance/extensions/ext-sRGB.html webgl/1.0.3/conformance/extensions/oes-element-index-uint.html webgl/1.0.3/conformance/glsl/misc/shader-struct-scope.html webgl/1.0.3/conformance/misc/delayed-drawing.html webgl/1.0.3/conformance/misc/uninitialized-test.html webgl/1.0.3/conformance/renderbuffers/framebuffer-object-attachment.html webgl/1.0.3/conformance/state/state-uneffected-after-compositing.html webgl/1.0.3/conformance/textures/default-texture.html webgl/1.0.3/conformance/textures/tex-image-and-sub-image-2d-with-canvas-rgb565.html webgl/1.0.3/conformance/textures/tex-image-and-sub-image-2d-with-canvas-rgba4444.html webgl/1.0.3/conformance/textures/tex-image-and-sub-image-2d-with-canvas-rgba5551.html webgl/1.0.3/conformance/textures/tex-image-and-sub-image-2d-with-canvas.html webgl/1.0.3/conformance/textures/tex-image-and-sub-image-2d-with-webgl-canvas-rgb565.html webgl/1.0.3/conformance/textures/tex-image-and-sub-image-2d-with-webgl-canvas-rgba4444.html webgl/1.0.3/conformance/textures/tex-image-and-sub-image-2d-with-webgl-canvas-rgba5551.html webgl/1.0.3/conformance/textures/tex-image-and-sub-image-2d-with-webgl-canvas.html
Attachments
Add attachment proposed patch, testcase, etc.
Radar WebKit Bug Importer
Comment 1 2020-02-17 13:46:54 PST
<rdar://problem/59524399>
Dean Jackson
Comment 2 2020-02-17 14:53:27 PST
Also fast/canvas/webgl/oes-texture-float-linear.html is saying OES_texture_float is not supported on iOS simulator.
Radar WebKit Bug Importer
Comment 3 2020-02-17 14:53:36 PST
<rdar://problem/59527402>
Jacob Uphoff
Comment 4 2020-02-17 16:43:08 PST
Please add expectations for the following: fast/canvas/webgl/program-test.html webgl/1.0.3/conformance/programs/program-test.html
Dean Jackson
Comment 5 2020-02-17 17:43:49 PST
WebGL 2 macOS: webgl/2.0.0/conformance/extensions/oes-vertex-array-object.html webgl/2.0.0/conformance/programs/program-test.html webgl/2.0.0/conformance2/buffers/buffer-type-restrictions.html (improved overall) webgl/2.0.0/conformance2/buffers/get-buffer-sub-data.html webgl/2.0.0/conformance2/buffers/one-large-uniform-buffer.html webgl/2.0.0/conformance2/extensions/promoted-extensions.html webgl/2.0.0/conformance2/misc/expando-loss-2.html webgl/2.0.0/conformance2/reading/read-pixels-from-fbo-test.html webgl/2.0.0/conformance2/renderbuffers/invalidate-framebuffer.html webgl/2.0.0/conformance2/renderbuffers/multisample-with-full-sample-counts.html webgl/2.0.0/conformance2/renderbuffers/multisampled-renderbuffer-initialization.html webgl/2.0.0/conformance2/rendering/canvas-resizing-with-pbo-bound.html webgl/2.0.0/conformance2/rendering/instanced-arrays.html webgl/2.0.0/conformance2/rendering/instanced-rendering-bug.html webgl/2.0.0/conformance2/rendering/rgb-format-support.html webgl/2.0.0/conformance2/textures/image_bitmap_from_image/* (all timed out) webgl/2.0.0/conformance2/textures/image_bitmap_from_video/* (all timed out) webgl/2.0.0/conformance2/textures/video (timed out) webgl/2.0.0/conformance2/vertex_arrays/vertex-array-object.html (These are just a list of things that started failing more. The results might have been misleading to begin with.)
Ryan Haddad
Comment 6 2020-02-17 17:52:55 PST
(In reply to Jacob Uphoff from comment #4) > Please add expectations for the following: > > fast/canvas/webgl/program-test.html > webgl/1.0.3/conformance/programs/program-test.html https://trac.webkit.org/changeset/256802/webkit
Dean Jackson
Comment 7 2020-02-18 09:48:33 PST
*** Bug 207896 has been marked as a duplicate of this bug. ***
Dean Jackson
Comment 8 2020-02-18 09:48:45 PST
iOS: fast/canvas/webgl/oes-vertex-array-object.html
Ryan Haddad
Comment 9 2020-02-18 09:52:44 PST
rdar://59524399
Jacob Uphoff
Comment 10 2020-02-20 10:33:37 PST
fast/images/exif-orientation-webgl-texture.html This test is also failing with the changes made here
Dean Jackson
Comment 11 2020-02-21 14:38:44 PST
iOS: fast/images/exif-orientation-webgl-texture.html
Jason Lawrence
Comment 12 2020-02-24 14:27:01 PST
This test is flaky failing on Mojave: webgl/1.0.3/conformance/textures/tex-image-and-sub-image-2d-with-webgl-canvas-rgba5551.html
Kenneth Russell
Comment 13 2020-02-24 14:46:48 PST
(In reply to Jason Lawrence from comment #12) > This test is flaky failing on Mojave: > webgl/1.0.3/conformance/textures/tex-image-and-sub-image-2d-with-webgl- > canvas-rgba5551.html Link to the above failures: https://results.webkit.org/?suite=layout-tests&test=webgl%2F1.0.3%2Fconformance%2Ftextures%2Ftex-image-and-sub-image-2d-with-webgl-canvas-rgba5551.html
Kenneth Russell
Comment 15 2020-02-24 14:57:30 PST
The diffs are: 31[ 28: PASS ] shouldBe 0,255,0 32[ 29: PASS ] shouldBe 255,0,0 33[ 30: PASS ] shouldBe 0,255,0 34[ 31: PASS ] shouldBe 255,0,0 35[ 32: PASS ] shouldBe 0,255,0 36[ 33: FAIL ] at (0, 0) expected: 255,0,0 was 0,0,0 37[ 34: FAIL ] at (0, 16) expected: 0,255,0 was 0,0,0 38[ 35: FAIL ] at (0, 16) expected: 255,0,0 was 0,0,0 39[ 36: FAIL ] at (0, 0) expected: 0,255,0 was 0,0,0 40[ 37: FAIL ] at (0, 0) expected: 255,0,0 was 0,0,0 41[ 38: FAIL ] at (0, 16) expected: 0,255,0 was 0,0,0 42[ 39: FAIL ] at (0, 16) expected: 255,0,0 was 0,0,0 43[ 40: FAIL ] at (0, 0) expected: 0,255,0 was 0,0,0 ... The test case that failed in this case is the second whole iteration through the test cases in LayoutTests/webgl/1.0.3/resources/webgl_test_files/conformance/resources/tex-image-and-sub-image-2d-with-webgl-canvas.js , the first test case in the list: Testing texImage2D with flipY=true canvas size: 1x2 with red-green Checking top PASS shouldBe 255,0,0 Checking bottom PASS shouldBe 0,255,0
Kenneth Russell
Comment 16 2020-02-24 15:01:20 PST
The failure of webgl/1.0.3/conformance/textures/tex-image-and-sub-image-2d-with-webgl-canvas-rgba5551.html in https://build.webkit.org/results/Apple%20Mojave%20Debug%20WK2%20(Tests)/r257234%20(7731)/results.html failed a different sub-test, however. We should probably mark this test flaky. Not 100% sure what could be causing this; Chrome doesn't seem to have a suppression for this test, but it is running against a newer conformance suite snapshot.
Dean Jackson
Comment 17 2020-02-29 13:40:44 PST
webgl/1.0.3/conformance/renderbuffers/framebuffer-object-attachment.html fixed by 208331
Dean Jackson
Comment 18 2020-03-30 10:06:32 PDT
webgl/2.0.0/conformance2/rendering/framebuffer-completeness-unaffected.html
Justin Fan
Comment 19 2020-05-22 16:38:14 PDT
These are the remaining WebGL1-based regressions: iOS/Simulator: https://bugs.webkit.org/show_bug.cgi?id=212277 webgl/1.0.3/conformance/extensions/ext-sRGB.html https://bugs.webkit.org/show_bug.cgi?id=212271 fast/canvas/webgl/webgl-depth-texture.html https://bugs.webkit.org/show_bug.cgi?id=209139 [ iOS wk2 ] webgl/1.0.3/conformance/textures/copy-tex-image-2d-formats.html is failing. (209139) https://bugs.webkit.org/show_bug.cgi?id=212249 [ANGLE - iOS] fast/canvas/webgl/uninitialized-test.html is still failing —— macOS and iOS: https://bugs.webkit.org/show_bug.cgi?id=212287 fast/canvas/webgl/draw-elements-out-of-bounds-uint-index.html (Not a breaking failure) macOS only: https://bugs.webkit.org/show_bug.cgi?id=212286 fast/canvas/webgl/simulated-vertexAttrib0-invalid-indicies.html (Not a breaking failure, tested code path will be obsoleted)
Justin Fan
Comment 20 2020-05-22 16:39:16 PDT
final comment about "code path will be obsoleted" was in error, please ignore
Note You need to log in before you can comment on or make changes to this bug.