Bug 112298

Summary:

[WebGL] array-bounds-clamping should use less dramatic numbers

Product:

WebKit

Reporter:

Dean Jackson <dino>

Component:

WebGL

Assignee:

Dean Jackson <dino>

Status:

RESOLVED FIXED

Severity:

Normal

CC:

gman, kbr, webkit-bug-importer

Priority:

Keywords:

InRadar

Version:

528+ (Nightly build)

Hardware:

Unspecified

OS:

Unspecified

Attachments:

Description	Flags
Patch	kbr: review+

Description Dean Jackson 2013-03-13 16:24:27 PDT

Some architectures seem unable to handle the cast int(infinity) (where infinity is a float) in GLSL, which causes array-bounds-clamping to fail. From what I've seen, they always return 0 in this case, so we're still doing acceptable clamping. Unfortunately we can't use the isinf() function to test for this because it isn't in GLSLES 1.0.

I suggest replacing the MAX_NUMBER (infinity) values in the test with some large but not huge numbers, although maybe we should also be explicitly testing our casts from infinity as well?

Comment 1 Radar WebKit Bug Importer 2013-03-13 16:25:05 PDT

<rdar://problem/13415829>

Comment 2 Dean Jackson 2013-03-13 16:29:28 PDT

Created attachment 193021 [details]
Patch

Comment 3 Kenneth Russell 2013-03-13 17:13:07 PDT

I noticed this too on Windows platforms using ANGLE. It looks like Direct3D doesn't specify the result of HLSL intrinsics like clamp when the inputs are +/-infinity.

Comment 4 Kenneth Russell 2013-03-13 17:25:20 PDT

Comment on attachment 193021 [details]
Patch

Seems fine for this layout test. It would be good to add more tests verifying out-of-range access behavior works according to spec. Currently conformance/uniforms/out-of-bounds-uniform-array-access.html is the only conformance test in this area. r=me

Comment 5 Dean Jackson 2013-03-13 17:46:54 PDT

It's pretty disappointing how small a number can trigger this :(

Comment 6 Dean Jackson 2013-03-13 19:17:29 PDT

Committed r145780: <http://trac.webkit.org/changeset/145780>