Bug 142119
Summary: | fast/canvas/canvas-ellipse-zero-lineto.html failing on Apple Windows | ||
---|---|---|---|
Product: | WebKit | Reporter: | Dean Jackson <dino> |
Component: | Canvas | Assignee: | Nobody <webkit-unassigned> |
Status: | RESOLVED MOVED | ||
Severity: | Normal | CC: | bfulgham, dino, mmaxfield, webkit-bug-importer |
Priority: | P2 | Keywords: | InRadar |
Version: | 528+ (Nightly build) | ||
Hardware: | Unspecified | ||
OS: | Unspecified |
Dean Jackson
This test is reproducibly failing on Windows.
fast/canvas/canvas-ellipse-zero-lineto.html
Attachments | ||
---|---|---|
Add attachment proposed patch, testcase, etc. |
Radar WebKit Bug Importer
<rdar://problem/19993919>
Myles C. Maxfield
*** Bug 142120 has been marked as a duplicate of this bug. ***
Brent Fulgham
Failure looks like the following:
--- /cygdrive/c/Projects/WebKit/OpenSource/WebKitBuild/Release/bin32/layout-test-results/fast/canvas/canvas-ellipse-zero-lineto-expected.txt
+++ /cygdrive/c/Projects/WebKit/OpenSource/WebKitBuild/Release/bin32/layout-test-results/fast/canvas/canvas-ellipse-zero-lineto-actual.txt
@@ -57,7 +57,7 @@
PASS imageData.data[1] is 0
PASS imageData.data[1] is 0
2. sweepAngle < PI
-PASS imageData.data[1] is 0
+FAIL imageData.data[1] should be 0. Was 7.
PASS imageData.data[1] is 255
PASS imageData.data[1] is 0
PASS imageData.data[1] is 0
Brent Fulgham
Here is the failing sequence of commands:
debug(" 2. sweepAngle < PI");
ctx.translate(0, 30);
ctx.save();
ctx.beginPath();
ctx.lineTo(10, 0);
ctx.ellipse(20, 0, 20, zero, Math.PI / 6, -Math.PI / 4, Math.PI / 2, false);
ctx.lineTo(80, 0);
ctx.stroke();
ctx.restore();
imageData = ctx.getImageData(232, 69, 1, 1);
shouldBe("imageData.data[1]", "0");
imageData = ctx.getImageData(238, 72, 1, 1);
shouldBe("imageData.data[1]", "255");
imageData = ctx.getImageData(228, 65, 1, 1);
shouldBe("imageData.data[1]", "0");
imageData = ctx.getImageData(242, 61, 1, 1);
shouldBe("imageData.data[1]", "0");
Brent Fulgham
See the screenshot "Comparison" to see the slight differences between the two drawing libraries. The top "arrow" is the Mac version, the bottom is the WIndows one.
Antialiasing seems different on the two platforms.
The drawing commands are identical:
ctx.beginPath();
ctx.lineTo(10, 0);
ctx.ellipse(20, 0, 20, zero, Math.PI / 6, -Math.PI / 4, Math.PI / 2, false);
ctx.lineTo(80, 0);
ctx.stroke();
ctx.restore();
imageData = ctx.getImageData(232, 69, 1, 1); // Should be black, but WIndows is '7' (not '0')
imageData = ctx.getImageData(238, 72, 1, 1); // Should be White, and both are.
imageData = ctx.getImageData(228, 65, 1, 1); // Should be black, and both are.
imageData = ctx.getImageData(242, 61, 1, 1); // Should be black, and both are.
This might be a drawing difference in the rendering libraries used on the two platforms.
Brent Fulgham
Unskipped and rebaselined test in r180895 <https://trac.webkit.org/changeset/180895>.
Brent Fulgham
The fix for this issue was needed outside the WebKit project, therefore this is being resolved as 'Moved'.
This should now be fixed in shipping software.