Attachment 349456 Details for Bug 189502 – Patch

[patch] Patch

bug-189502-20180911143054.patch (text/plain), 55.86 KB, created by Thomas Denney on 2018-09-11 14:30:55 PDT

(hide)

Description:

Filename:

MIME Type:

Creator: Thomas Denney

Created: 2018-09-11 14:30:55 PDT

Size: 55.86 KB

patch

obsolete

>Subversion Revision: 235894
>diff --git a/Tools/ChangeLog b/Tools/ChangeLog
>index e60abee4189f1442afc52c598099305a20175ed1..43ce9e17d565d5c9f2d3d22d5baae83a06298d86 100644
>--- a/Tools/ChangeLog
>+++ b/Tools/ChangeLog
>@@ -1 +1,28 @@
>+2018-09-11  Thomas Denney  <tdenney@apple.com>
>+
>+        [WHLSL] Improve test suite type safety
>+        https://bugs.webkit.org/show_bug.cgi?id=189502
>+
>+        Reviewed by NOBODY (OOPS!).
>+
>+        Each of the 'makeT' functions now call the relevant cast function on the
>+        value before hand. The checkNumber function has also been removed and
>+        its uses have been replaced with functions that also check the type.
>+        Some of the arithmetic checks have been updated to reflect that the
>+        casts happen outside of WHLSL evaluation. Other tests have also been
>+        updated to reflect that some values cannot be precisely expressed in
>+        32-bit floating point.
>+
>+        * WebGPUShadingLanguageRI/Casts.js:
>+        (castToBool): Added.
>+        (castAndCheckValue): Added.
>+        (isBitwiseEquivalent): Moved from Intrinsics.js.
>+        * WebGPUShadingLanguageRI/Intrinsics.js: Ditto.
>+        * WebGPUShadingLanguageRI/SPIRV.html: Update depdencies.
>+        * WebGPUShadingLanguageRI/Test.html: Ditto.
>+        * WebGPUShadingLanguageRI/Test.js: Update makeT functions to do a cast
>+        and check the result. Some tests were also updated to reflect the change
>+        in the behavior of these functions.
>+        * WebGPUShadingLanguageRI/index.html: Update dependencies.
>+
> == Rolled over to ChangeLog-2018-09-11 ==
>diff --git a/Tools/WebGPUShadingLanguageRI/Casts.js b/Tools/WebGPUShadingLanguageRI/Casts.js
>index acbb2ea028535b4670423097f7a5b57caf6b8b1c..c8f08ce69a7f32e8a11a0504352a4052df164d3f 100644
>--- a/Tools/WebGPUShadingLanguageRI/Casts.js
>+++ b/Tools/WebGPUShadingLanguageRI/Casts.js
>@@ -39,6 +39,11 @@ function bitwiseCast(typedArrayConstructor1, typedArrayConstructor2, value)
>     return typedArray2[0];
> }
> 
>+function castToBool(value)
>+{
>+    return !!value;
>+}
>+
> function castToUchar(number)
> {
>     return number & 0xFF;
>@@ -113,3 +118,25 @@ function castToFloat(number)
> {
>     return Math.fround(number);
> }
>+
>+function castAndCheckValue(castFunction, value)
>+{
>+    const castedValue = castFunction(value);
>+    if (!isBitwiseEquivalent(castedValue, value))
>+        throw new Error(`${value} was casted and yielded ${castedValue}, which is not bitwise equivalent.`);
>+    return castedValue;
>+}
>+
>+function isBitwiseEquivalent(left, right)
>+{
>+    let doubleArray = new Float64Array(1);
>+    let intArray = new Int32Array(doubleArray.buffer);
>+    doubleArray[0] = left;
>+    let leftInts = Int32Array.from(intArray);
>+    doubleArray[0] = right;
>+    for (let i = 0; i < 2; ++i) {
>+        if (leftInts[i] != intArray[i])
>+            return false;
>+    }
>+    return true;
>+}
>diff --git a/Tools/WebGPUShadingLanguageRI/Intrinsics.js b/Tools/WebGPUShadingLanguageRI/Intrinsics.js
>index 1b0a12af4002a90be635e738f1d5e7d90dcb04d3..68b0dd573fc02ea38517955d8388a646aa90a596 100644
>--- a/Tools/WebGPUShadingLanguageRI/Intrinsics.js
>+++ b/Tools/WebGPUShadingLanguageRI/Intrinsics.js
>@@ -34,20 +34,6 @@ class Intrinsics {
>         // For example, if a native function is declared using "int" rather than "int", then we must
>         // use "int" here, since we don't yet know that they are the same type.
> 
>-        function isBitwiseEquivalent(left, right)
>-        {
>-            let doubleArray = new Float64Array(1);
>-            let intArray = new Int32Array(doubleArray.buffer);
>-            doubleArray[0] = left;
>-            let leftInts = Int32Array.from(intArray);
>-            doubleArray[0] = right;
>-            for (let i = 0; i < 2; ++i) {
>-                if (leftInts[i] != intArray[i])
>-                    return false;
>-            }
>-            return true;
>-        }
>-
>         this._map.set(
>             "native typedef void",
>             type => {
>diff --git a/Tools/WebGPUShadingLanguageRI/Test.js b/Tools/WebGPUShadingLanguageRI/Test.js
>index 2bcea7d7da3cb382f1aabe2167ef3a7891392333..302d6d9a439836134d475d0978501b076de20c4b 100644
>--- a/Tools/WebGPUShadingLanguageRI/Test.js
>+++ b/Tools/WebGPUShadingLanguageRI/Test.js
>@@ -55,32 +55,37 @@ function doLex(code)
> 
> function makeInt(program, value)
> {
>-    return TypedValue.box(program.intrinsics.int, value);
>+    return TypedValue.box(program.intrinsics.int, castAndCheckValue(castToInt, value));
> }
> 
> function makeUint(program, value)
> {
>-    return TypedValue.box(program.intrinsics.uint, value);
>+    return TypedValue.box(program.intrinsics.uint, castAndCheckValue(castToUint, value));
> }
> 
> function makeUchar(program, value)
> {
>-    return TypedValue.box(program.intrinsics.uchar, value);
>+    return TypedValue.box(program.intrinsics.uchar, castAndCheckValue(castToUchar, value));
> }
> 
> function makeBool(program, value)
> {
>-    return TypedValue.box(program.intrinsics.bool, value);
>+    return TypedValue.box(program.intrinsics.bool, castAndCheckValue(castToBool, value));
> }
> 
> function makeFloat(program, value)
> {
>-    return TypedValue.box(program.intrinsics.float, value);
>+    return TypedValue.box(program.intrinsics.float, castAndCheckValue(castToFloat, value));
>+}
>+
>+function makeCastedFloat(program, value)
>+{
>+    return TypedValue.box(program.intrinsics.float, castToFloat(value));
> }
> 
> function makeHalf(program, value)
> {
>-    return TypedValue.box(program.intrinsics.half, value);
>+    return TypedValue.box(program.intrinsics.half, castAndCheckValue(castToHalf, value));
> }
> 
> function makeEnum(program, enumName, value)
>@@ -221,19 +226,12 @@ function makeRW2DDepthTextureArray(program, array, elementType)
>     return TypedValue.box(program.intrinsics[`RWTextureDepth2DArray<${elementType}>`], new TextureDepth2DArrayRW(elementType, array));
> }
> 
>-function checkNumber(program, result, expected)
>-{
>-    if (!result.type.unifyNode.isNumber)
>-        throw new Error("Wrong result type; result: " + result);
>-    if (result.value != expected)
>-        throw new Error("Wrong result: " + result.value + " (expected " + expected + ")");
>-}
>-
> function checkInt(program, result, expected)
> {
>     if (!result.type.equals(program.intrinsics.int))
>         throw new Error("Wrong result type; result: " + result);
>-    checkNumber(program, result, expected);
>+    if (result.value != expected)
>+        throw new Error(`Wrong result: ${result.value} (expected ${expected})`);
> }
> 
> function checkEnum(program, result, expected)
>@@ -1039,13 +1037,13 @@ tests.passNullToPtrMonomorphicArrayRef = function()
> tests.returnIntLiteralUint = function()
> {
>     let program = doPrep("test uint foo() { return 42; }");
>-    checkNumber(program, callFunction(program, "foo", []), 42);
>+    checkUint(program, callFunction(program, "foo", []), 42);
> }
> 
> tests.returnIntLiteralFloat = function()
> {
>     let program = doPrep("test float foo() { return 42; }");
>-    checkNumber(program, callFunction(program, "foo", []), 42);
>+    checkFloat(program, callFunction(program, "foo", []), 42);
> }
> 
> tests.badIntLiteralForInt = function()
>@@ -2552,7 +2550,6 @@ tests.uintBitAnd = function()
>     `);
>     checkUint(program, callFunction(program, "foo", [makeUint(program, 1), makeUint(program, 7)]), 1);
>     checkUint(program, callFunction(program, "foo", [makeUint(program, 65535), makeUint(program, 42)]), 42);
>-    checkUint(program, callFunction(program, "foo", [makeUint(program, -1), makeUint(program, -7)]), 4294967289);
>     checkUint(program, callFunction(program, "foo", [makeUint(program, 0), makeUint(program, 85732)]), 0);
> }
> 
>@@ -2566,7 +2563,6 @@ tests.uintBitOr = function()
>     `);
>     checkUint(program, callFunction(program, "foo", [makeUint(program, 1), makeUint(program, 7)]), 7);
>     checkUint(program, callFunction(program, "foo", [makeUint(program, 65535), makeUint(program, 42)]), 65535);
>-    checkUint(program, callFunction(program, "foo", [makeUint(program, -1), makeUint(program, -7)]), 4294967295);
>     checkUint(program, callFunction(program, "foo", [makeUint(program, 0), makeUint(program, 85732)]), 85732);
> }
> 
>@@ -2580,7 +2576,6 @@ tests.uintBitXor = function()
>     `);
>     checkUint(program, callFunction(program, "foo", [makeUint(program, 1), makeUint(program, 7)]), 6);
>     checkUint(program, callFunction(program, "foo", [makeUint(program, 65535), makeUint(program, 42)]), 65493);
>-    checkUint(program, callFunction(program, "foo", [makeUint(program, -1), makeUint(program, -7)]), 6);
>     checkUint(program, callFunction(program, "foo", [makeUint(program, 0), makeUint(program, 85732)]), 85732);
> }
> 
>@@ -2594,7 +2589,6 @@ tests.uintBitNot = function()
>     `);
>     checkUint(program, callFunction(program, "foo", [makeUint(program, 1)]), 4294967294);
>     checkUint(program, callFunction(program, "foo", [makeUint(program, 65535)]), 4294901760);
>-    checkUint(program, callFunction(program, "foo", [makeUint(program, -1)]), 0);
>     checkUint(program, callFunction(program, "foo", [makeUint(program, 0)]), 4294967295);
> }
> 
>@@ -2608,7 +2602,6 @@ tests.uintLShift = function()
>     `);
>     checkUint(program, callFunction(program, "foo", [makeUint(program, 1), makeUint(program, 7)]), 128);
>     checkUint(program, callFunction(program, "foo", [makeUint(program, 65535), makeUint(program, 2)]), 262140);
>-    checkUint(program, callFunction(program, "foo", [makeUint(program, -1), makeUint(program, 5)]), 4294967264);
>     checkUint(program, callFunction(program, "foo", [makeUint(program, 0), makeUint(program, 3)]), 0);
> }
> 
>@@ -2622,7 +2615,6 @@ tests.uintRShift = function()
>     `);
>     checkUint(program, callFunction(program, "foo", [makeUint(program, 1), makeUint(program, 7)]), 0);
>     checkUint(program, callFunction(program, "foo", [makeUint(program, 65535), makeUint(program, 2)]), 16383);
>-    checkUint(program, callFunction(program, "foo", [makeUint(program, -1), makeUint(program, 5)]), 134217727);
>     checkUint(program, callFunction(program, "foo", [makeUint(program, 0), makeUint(program, 3)]), 0);
> }
> 
>@@ -2635,9 +2627,8 @@ tests.ucharBitAnd = function()
>         }
>     `);
>     checkUchar(program, callFunction(program, "foo", [makeUchar(program, 1), makeUchar(program, 7)]), 1);
>-    checkUchar(program, callFunction(program, "foo", [makeUchar(program, 65535), makeUchar(program, 42)]), 42);
>-    checkUchar(program, callFunction(program, "foo", [makeUchar(program, -1), makeUchar(program, -7)]), 249);
>-    checkUchar(program, callFunction(program, "foo", [makeUchar(program, 0), makeUchar(program, 85732)]), 0);
>+    checkUchar(program, callFunction(program, "foo", [makeUchar(program, 255), makeUchar(program, 42)]), 42);
>+    checkUchar(program, callFunction(program, "foo", [makeUchar(program, 0), makeUchar(program, 255)]), 0);
> }
> 
> tests.ucharBitOr = function()
>@@ -2649,9 +2640,8 @@ tests.ucharBitOr = function()
>         }
>     `);
>     checkUchar(program, callFunction(program, "foo", [makeUchar(program, 1), makeUchar(program, 7)]), 7);
>-    checkUchar(program, callFunction(program, "foo", [makeUchar(program, 65535), makeUchar(program, 42)]), 255);
>-    checkUchar(program, callFunction(program, "foo", [makeUchar(program, -1), makeUchar(program, -7)]), 255);
>-    checkUchar(program, callFunction(program, "foo", [makeUchar(program, 0), makeUchar(program, 85732)]), 228);
>+    checkUchar(program, callFunction(program, "foo", [makeUchar(program, 255), makeUchar(program, 42)]), 255);
>+    checkUchar(program, callFunction(program, "foo", [makeUchar(program, 0), makeUchar(program, 228)]), 228);
> }
> 
> tests.ucharBitXor = function()
>@@ -2663,9 +2653,8 @@ tests.ucharBitXor = function()
>         }
>     `);
>     checkUchar(program, callFunction(program, "foo", [makeUchar(program, 1), makeUchar(program, 7)]), 6);
>-    checkUchar(program, callFunction(program, "foo", [makeUchar(program, 65535), makeUchar(program, 42)]), 213);
>-    checkUchar(program, callFunction(program, "foo", [makeUchar(program, -1), makeUchar(program, -7)]), 6);
>-    checkUchar(program, callFunction(program, "foo", [makeUchar(program, 0), makeUchar(program, 85732)]), 228);
>+    checkUchar(program, callFunction(program, "foo", [makeUchar(program, 255), makeUchar(program, 42)]), 213);
>+    checkUchar(program, callFunction(program, "foo", [makeUchar(program, 0), makeUchar(program, 255)]), 255);
> }
> 
> tests.ucharBitNot = function()
>@@ -2677,8 +2666,7 @@ tests.ucharBitNot = function()
>         }
>     `);
>     checkUchar(program, callFunction(program, "foo", [makeUchar(program, 1)]), 254);
>-    checkUchar(program, callFunction(program, "foo", [makeUchar(program, 65535)]), 0);
>-    checkUchar(program, callFunction(program, "foo", [makeUchar(program, -1)]), 0);
>+    checkUchar(program, callFunction(program, "foo", [makeUchar(program, 255)]), 0);
>     checkUchar(program, callFunction(program, "foo", [makeUchar(program, 0)]), 255);
> }
> 
>@@ -2691,8 +2679,7 @@ tests.ucharLShift = function()
>         }
>     `);
>     checkUchar(program, callFunction(program, "foo", [makeUchar(program, 1), makeUint(program, 7)]), 128);
>-    checkUchar(program, callFunction(program, "foo", [makeUchar(program, 65535), makeUint(program, 2)]), 252);
>-    checkUchar(program, callFunction(program, "foo", [makeUchar(program, -1), makeUint(program, 5)]), 224);
>+    checkUchar(program, callFunction(program, "foo", [makeUchar(program, 255), makeUint(program, 2)]), 252);
>     checkUchar(program, callFunction(program, "foo", [makeUchar(program, 0), makeUint(program, 3)]), 0);
> }
> 
>@@ -2705,8 +2692,8 @@ tests.ucharRShift = function()
>         }
>     `);
>     checkUchar(program, callFunction(program, "foo", [makeUchar(program, 1), makeUint(program, 7)]), 0);
>-    checkUchar(program, callFunction(program, "foo", [makeUchar(program, 65535), makeUint(program, 2)]), 255);
>-    checkUchar(program, callFunction(program, "foo", [makeUchar(program, -1), makeUint(program, 5)]), 255);
>+    checkUchar(program, callFunction(program, "foo", [makeUchar(program, 255), makeUint(program, 2)]), 63);
>+    checkUchar(program, callFunction(program, "foo", [makeUchar(program, 255), makeUint(program, 5)]), 7);
>     checkUchar(program, callFunction(program, "foo", [makeUchar(program, 0), makeUint(program, 3)]), 0);
> }
> 
>@@ -7293,14 +7280,14 @@ tests.textureSample = function() {
>     checkFloat(program, callFunction(program, "foo14", [texture2D, makeSampler(program, {magFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, 0)]), (12 + 13 + 20 + 21) / 4);
>     checkFloat(program, callFunction(program, "foo14", [texture2D, makeSampler(program, {magFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 1 / texture2D.ePtr.loadValue().width), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, 1 / texture2D.ePtr.loadValue().height)]), (12 + 13 + 20 + 21) / 4);
>     checkFloat(program, callFunction(program, "foo14", [texture2D, makeSampler(program, {minFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 2 / texture2D.ePtr.loadValue().width), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, 2 / texture2D.ePtr.loadValue().height)]), (34 + 35 + 38 + 39) / 4);
>-    checkFloat(program, callFunction(program, "foo14", [texture2D, makeSampler(program, {minFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, Math.pow(2, 0.25) / texture2D.ePtr.loadValue().width), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, Math.pow(2, 0.25) / texture2D.ePtr.loadValue().height)]), (12 + 13 + 20 + 21) / 4);
>-    checkFloat(program, callFunction(program, "foo14", [texture2D, makeSampler(program, {minFilter: "linear", mipmapFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, Math.pow(2, 0.5) / texture2D.ePtr.loadValue().width), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, Math.pow(2, 0.5) / texture2D.ePtr.loadValue().height)]), (((12 + 13 + 20 + 21) / 4) + ((34 + 35 + 38 + 39) / 4)) / 2);
>+    checkFloat(program, callFunction(program, "foo14", [texture2D, makeSampler(program, {minFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeCastedFloat(program, Math.pow(2, 0.25) / texture2D.ePtr.loadValue().width), makeFloat(program, 0), makeFloat(program, 0), makeCastedFloat(program, Math.pow(2, 0.25) / texture2D.ePtr.loadValue().height)]), (12 + 13 + 20 + 21) / 4);
>+    checkFloat(program, callFunction(program, "foo14", [texture2D, makeSampler(program, {minFilter: "linear", mipmapFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeCastedFloat(program, Math.pow(2, 0.5) / texture2D.ePtr.loadValue().width), makeFloat(program, 0), makeFloat(program, 0), makeCastedFloat(program, Math.pow(2, 0.5) / texture2D.ePtr.loadValue().height)]), (((12 + 13 + 20 + 21) / 4) + ((34 + 35 + 38 + 39) / 4)) / 2);
>     checkFloat(program, callFunction(program, "foo15", [texture2D, makeSampler(program, {magFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, 0), makeInt(program, 0), makeInt(program, 0)]), (12 + 13 + 20 + 21) / 4);
>     checkFloat(program, callFunction(program, "foo15", [texture2D, makeSampler(program, {magFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, 0), makeInt(program, 1), makeInt(program, 0)]), (13 + 14 + 21 + 22) / 4);
>     checkFloat(program, callFunction(program, "foo15", [texture2D, makeSampler(program, {magFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 1 / texture2D.ePtr.loadValue().width), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, 1 / texture2D.ePtr.loadValue().height), makeInt(program, 1), makeInt(program, 0)]), (13 + 14 + 21 + 22) / 4);
>     checkFloat(program, callFunction(program, "foo15", [texture2D, makeSampler(program, {minFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 2 / texture2D.ePtr.loadValue().width), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, 2 / texture2D.ePtr.loadValue().height), makeInt(program, 1), makeInt(program, 0)]), (35 + 36 + 39 + 40) / 4);
>-    checkFloat(program, callFunction(program, "foo15", [texture2D, makeSampler(program, {minFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, Math.pow(2, 0.25) / texture2D.ePtr.loadValue().width), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, Math.pow(2, 0.25) / texture2D.ePtr.loadValue().height), makeInt(program, 1), makeInt(program, 0)]), (13 + 14 + 21 + 22) / 4);
>-    checkFloat(program, callFunction(program, "foo15", [texture2D, makeSampler(program, {minFilter: "linear", mipmapFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, Math.pow(2, 0.5) / texture2D.ePtr.loadValue().width), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, Math.pow(2, 0.5) / texture2D.ePtr.loadValue().height), makeInt(program, 1), makeInt(program, 0)]), (((13 + 14 + 21 + 22) / 4) + ((35 + 36 + 39 + 40) / 4)) / 2);
>+    checkFloat(program, callFunction(program, "foo15", [texture2D, makeSampler(program, {minFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeCastedFloat(program, Math.pow(2, 0.25) / texture2D.ePtr.loadValue().width), makeFloat(program, 0), makeFloat(program, 0), makeCastedFloat(program, Math.pow(2, 0.25) / texture2D.ePtr.loadValue().height), makeInt(program, 1), makeInt(program, 0)]), (13 + 14 + 21 + 22) / 4);
>+    checkFloat(program, callFunction(program, "foo15", [texture2D, makeSampler(program, {minFilter: "linear", mipmapFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeCastedFloat(program, Math.pow(2, 0.5) / texture2D.ePtr.loadValue().width), makeFloat(program, 0), makeFloat(program, 0), makeCastedFloat(program, Math.pow(2, 0.5) / texture2D.ePtr.loadValue().height), makeInt(program, 1), makeInt(program, 0)]), (((13 + 14 + 21 + 22) / 4) + ((35 + 36 + 39 + 40) / 4)) / 2);
>     checkFloat(program, callFunction(program, "foo16", [texture2D, makeSampler(program, {magFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 0)]), (12 + 13 + 20 + 21) / 4);
>     checkFloat(program, callFunction(program, "foo16", [texture2D, makeSampler(program, {minFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 1)]), (34 + 35 + 38 + 39) / 4);
>     checkFloat(program, callFunction(program, "foo16", [texture2D, makeSampler(program, {minFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 0.5)]), (12 + 13 + 20 + 21) / 4);
>@@ -7326,25 +7313,25 @@ tests.textureSample = function() {
>     checkFloat(program, callFunction(program, "foo20", [texture2DArray, makeSampler(program, {magFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, 0)]), (12 + 13 + 20 + 21) / 4);
>     checkFloat(program, callFunction(program, "foo20", [texture2DArray, makeSampler(program, {magFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 0), makeFloat(program, 1 / texture2D.ePtr.loadValue().width), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, 1 / texture2D.ePtr.loadValue().height)]), (12 + 13 + 20 + 21) / 4);
>     checkFloat(program, callFunction(program, "foo20", [texture2DArray, makeSampler(program, {minFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 0), makeFloat(program, 2 / texture2D.ePtr.loadValue().width), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, 2 / texture2D.ePtr.loadValue().height)]), (34 + 35 + 38 + 39) / 4);
>-    checkFloat(program, callFunction(program, "foo20", [texture2DArray, makeSampler(program, {minFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 0), makeFloat(program, Math.pow(2, 0.25) / texture2D.ePtr.loadValue().width), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, Math.pow(2, 0.25) / texture2D.ePtr.loadValue().height)]), (12 + 13 + 20 + 21) / 4);
>-    checkFloat(program, callFunction(program, "foo20", [texture2DArray, makeSampler(program, {minFilter: "linear", mipmapFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 0), makeFloat(program, Math.pow(2, 0.5) / texture2D.ePtr.loadValue().width), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, Math.pow(2, 0.5) / texture2D.ePtr.loadValue().height)]), (((12 + 13 + 20 + 21) / 4) + ((34 + 35 + 38 + 39) / 4)) / 2);
>+    checkFloat(program, callFunction(program, "foo20", [texture2DArray, makeSampler(program, {minFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 0), makeCastedFloat(program, Math.pow(2, 0.25) / texture2D.ePtr.loadValue().width), makeFloat(program, 0), makeFloat(program, 0), makeCastedFloat(program, Math.pow(2, 0.25) / texture2D.ePtr.loadValue().height)]), (12 + 13 + 20 + 21) / 4);
>+    checkFloat(program, callFunction(program, "foo20", [texture2DArray, makeSampler(program, {minFilter: "linear", mipmapFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 0), makeCastedFloat(program, Math.pow(2, 0.5) / texture2D.ePtr.loadValue().width), makeFloat(program, 0), makeFloat(program, 0), makeCastedFloat(program, Math.pow(2, 0.5) / texture2D.ePtr.loadValue().height)]), (((12 + 13 + 20 + 21) / 4) + ((34 + 35 + 38 + 39) / 4)) / 2);
>     checkFloat(program, callFunction(program, "foo20", [texture2DArray, makeSampler(program, {magFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 1), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, 0)]), (54 + 55 + 62 + 63) / 4);
>     checkFloat(program, callFunction(program, "foo20", [texture2DArray, makeSampler(program, {magFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 1), makeFloat(program, 1 / texture2D.ePtr.loadValue().width), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, 1 / texture2D.ePtr.loadValue().height)]), (54 + 55 + 62 + 63) / 4);
>     checkFloat(program, callFunction(program, "foo20", [texture2DArray, makeSampler(program, {minFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 1), makeFloat(program, 2 / texture2D.ePtr.loadValue().width), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, 2 / texture2D.ePtr.loadValue().height)]), (76 + 77 + 80 + 81) / 4);
>-    checkFloat(program, callFunction(program, "foo20", [texture2DArray, makeSampler(program, {minFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 1), makeFloat(program, Math.pow(2, 0.25) / texture2D.ePtr.loadValue().width), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, Math.pow(2, 0.25) / texture2D.ePtr.loadValue().height)]), (54 + 55 + 62 + 63) / 4);
>-    checkFloat(program, callFunction(program, "foo20", [texture2DArray, makeSampler(program, {minFilter: "linear", mipmapFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 1), makeFloat(program, Math.pow(2, 0.5) / texture2D.ePtr.loadValue().width), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, Math.pow(2, 0.5) / texture2D.ePtr.loadValue().height)]), (((54 + 55 + 62 + 63) / 4) + ((76 + 77 + 80 + 81) / 4)) / 2);
>+    checkFloat(program, callFunction(program, "foo20", [texture2DArray, makeSampler(program, {minFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 1), makeCastedFloat(program, Math.pow(2, 0.25) / texture2D.ePtr.loadValue().width), makeFloat(program, 0), makeFloat(program, 0), makeCastedFloat(program, Math.pow(2, 0.25) / texture2D.ePtr.loadValue().height)]), (54 + 55 + 62 + 63) / 4);
>+    checkFloat(program, callFunction(program, "foo20", [texture2DArray, makeSampler(program, {minFilter: "linear", mipmapFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 1), makeCastedFloat(program, Math.pow(2, 0.5) / texture2D.ePtr.loadValue().width), makeFloat(program, 0), makeFloat(program, 0), makeCastedFloat(program, Math.pow(2, 0.5) / texture2D.ePtr.loadValue().height)]), (((54 + 55 + 62 + 63) / 4) + ((76 + 77 + 80 + 81) / 4)) / 2);
>     checkFloat(program, callFunction(program, "foo21", [texture2DArray, makeSampler(program, {magFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, 0), makeInt(program, 0), makeInt(program, 0)]), (12 + 13 + 20 + 21) / 4);
>     checkFloat(program, callFunction(program, "foo21", [texture2DArray, makeSampler(program, {magFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, 0), makeInt(program, 1), makeInt(program, 0)]), (13 + 14 + 21 + 22) / 4);
>     checkFloat(program, callFunction(program, "foo21", [texture2DArray, makeSampler(program, {magFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 0), makeFloat(program, 1 / texture2D.ePtr.loadValue().width), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, 1 / texture2D.ePtr.loadValue().height), makeInt(program, 1), makeInt(program, 0)]), (13 + 14 + 21 + 22) / 4);
>     checkFloat(program, callFunction(program, "foo21", [texture2DArray, makeSampler(program, {minFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 0), makeFloat(program, 2 / texture2D.ePtr.loadValue().width), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, 2 / texture2D.ePtr.loadValue().height), makeInt(program, 1), makeInt(program, 0)]), (35 + 36 + 39 + 40) / 4);
>-    checkFloat(program, callFunction(program, "foo21", [texture2DArray, makeSampler(program, {minFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 0), makeFloat(program, Math.pow(2, 0.25) / texture2D.ePtr.loadValue().width), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, Math.pow(2, 0.25) / texture2D.ePtr.loadValue().height), makeInt(program, 1), makeInt(program, 0)]), (13 + 14 + 21 + 22) / 4);
>-    checkFloat(program, callFunction(program, "foo21", [texture2DArray, makeSampler(program, {minFilter: "linear", mipmapFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 0), makeFloat(program, Math.pow(2, 0.5) / texture2D.ePtr.loadValue().width), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, Math.pow(2, 0.5) / texture2D.ePtr.loadValue().height), makeInt(program, 1), makeInt(program, 0)]), (((13 + 14 + 21 + 22) / 4) + ((35 + 36 + 39 + 40) / 4)) / 2);
>+    checkFloat(program, callFunction(program, "foo21", [texture2DArray, makeSampler(program, {minFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 0), makeCastedFloat(program, Math.pow(2, 0.25) / texture2D.ePtr.loadValue().width), makeFloat(program, 0), makeFloat(program, 0), makeCastedFloat(program, Math.pow(2, 0.25) / texture2D.ePtr.loadValue().height), makeInt(program, 1), makeInt(program, 0)]), (13 + 14 + 21 + 22) / 4);
>+    checkFloat(program, callFunction(program, "foo21", [texture2DArray, makeSampler(program, {minFilter: "linear", mipmapFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 0), makeCastedFloat(program, Math.pow(2, 0.5) / texture2D.ePtr.loadValue().width), makeFloat(program, 0), makeFloat(program, 0), makeCastedFloat(program, Math.pow(2, 0.5) / texture2D.ePtr.loadValue().height), makeInt(program, 1), makeInt(program, 0)]), (((13 + 14 + 21 + 22) / 4) + ((35 + 36 + 39 + 40) / 4)) / 2);
>     checkFloat(program, callFunction(program, "foo21", [texture2DArray, makeSampler(program, {magFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 1), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, 0), makeInt(program, 0), makeInt(program, 0)]), (54 + 55 + 62 + 63) / 4);
>     checkFloat(program, callFunction(program, "foo21", [texture2DArray, makeSampler(program, {magFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 1), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, 0), makeInt(program, 1), makeInt(program, 0)]), (55 + 56 + 63 + 64) / 4);
>     checkFloat(program, callFunction(program, "foo21", [texture2DArray, makeSampler(program, {magFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 1), makeFloat(program, 1 / texture2D.ePtr.loadValue().width), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, 1 / texture2D.ePtr.loadValue().height), makeInt(program, 1), makeInt(program, 0)]), (55 + 56 + 63 + 64) / 4);
>     checkFloat(program, callFunction(program, "foo21", [texture2DArray, makeSampler(program, {minFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 1), makeFloat(program, 2 / texture2D.ePtr.loadValue().width), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, 2 / texture2D.ePtr.loadValue().height), makeInt(program, 1), makeInt(program, 0)]), (77 + 78 + 81 + 82) / 4);
>-    checkFloat(program, callFunction(program, "foo21", [texture2DArray, makeSampler(program, {minFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 1), makeFloat(program, Math.pow(2, 0.25) / texture2D.ePtr.loadValue().width), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, Math.pow(2, 0.25) / texture2D.ePtr.loadValue().height), makeInt(program, 1), makeInt(program, 0)]), (55 + 56 + 63 + 64) / 4);
>-    checkFloat(program, callFunction(program, "foo21", [texture2DArray, makeSampler(program, {minFilter: "linear", mipmapFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 1), makeFloat(program, Math.pow(2, 0.5) / texture2D.ePtr.loadValue().width), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, Math.pow(2, 0.5) / texture2D.ePtr.loadValue().height), makeInt(program, 1), makeInt(program, 0)]), (((55 + 56 + 63 + 64) / 4) + ((77 + 78 + 81 + 82) / 4)) / 2);
>+    checkFloat(program, callFunction(program, "foo21", [texture2DArray, makeSampler(program, {minFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 1), makeCastedFloat(program, Math.pow(2, 0.25) / texture2D.ePtr.loadValue().width), makeFloat(program, 0), makeFloat(program, 0), makeCastedFloat(program, Math.pow(2, 0.25) / texture2D.ePtr.loadValue().height), makeInt(program, 1), makeInt(program, 0)]), (55 + 56 + 63 + 64) / 4);
>+    checkFloat(program, callFunction(program, "foo21", [texture2DArray, makeSampler(program, {minFilter: "linear", mipmapFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 1), makeCastedFloat(program, Math.pow(2, 0.5) / texture2D.ePtr.loadValue().width), makeFloat(program, 0), makeFloat(program, 0), makeCastedFloat(program, Math.pow(2, 0.5) / texture2D.ePtr.loadValue().height), makeInt(program, 1), makeInt(program, 0)]), (((55 + 56 + 63 + 64) / 4) + ((77 + 78 + 81 + 82) / 4)) / 2);
>     checkFloat(program, callFunction(program, "foo22", [texture2DArray, makeSampler(program, {magFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 0), makeFloat(program, 0)]), (12 + 13 + 20 + 21) / 4);
>     checkFloat(program, callFunction(program, "foo22", [texture2DArray, makeSampler(program, {minFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 0), makeFloat(program, 1)]), (34 + 35 + 38 + 39) / 4);
>     checkFloat(program, callFunction(program, "foo22", [texture2DArray, makeSampler(program, {minFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 0), makeFloat(program, 0.5)]), (12 + 13 + 20 + 21) / 4);
>@@ -7372,7 +7359,7 @@ tests.textureSample = function() {
>     checkFloat(program, callFunction(program, "foo16", [texture2D, makeSampler(program, {minFilter: "linear", lodMinClamp: 1, lodMaxClamp: 1}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 2)]), (34 + 35 + 38 + 39) / 4);
>     checkFloat(program, callFunction(program, "foo16", [texture2D, makeSampler(program, {minFilter: "linear", lodMinClamp: 1, lodMaxClamp: 1}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 0)]), (34 + 35 + 38 + 39) / 4);
>     checkFloat(program, callFunction(program, "foo14", [texture2D, makeSampler(program, {minFilter: "linear", mipmapFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 2 / texture2D.ePtr.loadValue().width), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, 1 / texture2D.ePtr.loadValue().height)]), (34 + 35 + 38 + 39) / 4);
>-    checkFloat(program, callFunction(program, "foo14", [texture2D, makeSampler(program, {minFilter: "linear", mipmapFilter: "linear", maxAnisotropy: 2}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 4 / texture2D.ePtr.loadValue().width), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, (4/3) / texture2D.ePtr.loadValue().height)]), (34 + 35 + 38 + 39) / 4);
>+    checkFloat(program, callFunction(program, "foo14", [texture2D, makeSampler(program, {minFilter: "linear", mipmapFilter: "linear", maxAnisotropy: 2}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 4 / texture2D.ePtr.loadValue().width), makeFloat(program, 0), makeFloat(program, 0), makeCastedFloat(program, (4/3) / texture2D.ePtr.loadValue().height)]), (34 + 35 + 38 + 39) / 4);
>     let texture1DInt4 = make1DTexture(program, [[[1, 2, 3, 4], [100, 200, 300, 400], [101, 202, 301, 401], [13, 14, 15, 16]], [[17, 18, 19, 20], [21, 22, 23, 24]], [[25, 26, 27, 28]]], "int4");
>     checkInt(program, callFunction(program, "foo24", [texture1DInt4, makeSampler(program, {magFilter: "linear"}), makeFloat(program, 0.5)]), 100);
>     checkInt(program, callFunction(program, "foo25", [texture1DInt4, makeSampler(program, {magFilter: "linear"}), makeFloat(program, 0.5)]), 201);
>@@ -7414,14 +7401,14 @@ tests.textureSample = function() {
>     checkFloat(program, callFunction(program, "foo35", [textureDepth2D, makeSampler(program, {magFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, 0)]), (12 + 13 + 20 + 21) / 4);
>     checkFloat(program, callFunction(program, "foo35", [textureDepth2D, makeSampler(program, {magFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 1 / texture2D.ePtr.loadValue().width), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, 1 / texture2D.ePtr.loadValue().height)]), (12 + 13 + 20 + 21) / 4);
>     checkFloat(program, callFunction(program, "foo35", [textureDepth2D, makeSampler(program, {minFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 2 / texture2D.ePtr.loadValue().width), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, 2 / texture2D.ePtr.loadValue().height)]), (34 + 35 + 38 + 39) / 4);
>-    checkFloat(program, callFunction(program, "foo35", [textureDepth2D, makeSampler(program, {minFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, Math.pow(2, 0.25) / texture2D.ePtr.loadValue().width), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, Math.pow(2, 0.25) / texture2D.ePtr.loadValue().height)]), (12 + 13 + 20 + 21) / 4);
>-    checkFloat(program, callFunction(program, "foo35", [textureDepth2D, makeSampler(program, {minFilter: "linear", mipmapFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, Math.pow(2, 0.5) / texture2D.ePtr.loadValue().width), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, Math.pow(2, 0.5) / texture2D.ePtr.loadValue().height)]), (((12 + 13 + 20 + 21) / 4) + ((34 + 35 + 38 + 39) / 4)) / 2);
>+    checkFloat(program, callFunction(program, "foo35", [textureDepth2D, makeSampler(program, {minFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeCastedFloat(program, Math.pow(2, 0.25) / texture2D.ePtr.loadValue().width), makeFloat(program, 0), makeFloat(program, 0), makeCastedFloat(program, Math.pow(2, 0.25) / texture2D.ePtr.loadValue().height)]), (12 + 13 + 20 + 21) / 4);
>+    checkFloat(program, callFunction(program, "foo35", [textureDepth2D, makeSampler(program, {minFilter: "linear", mipmapFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeCastedFloat(program, Math.pow(2, 0.5) / texture2D.ePtr.loadValue().width), makeFloat(program, 0), makeFloat(program, 0), makeCastedFloat(program, Math.pow(2, 0.5) / texture2D.ePtr.loadValue().height)]), (((12 + 13 + 20 + 21) / 4) + ((34 + 35 + 38 + 39) / 4)) / 2);
>     checkFloat(program, callFunction(program, "foo36", [textureDepth2D, makeSampler(program, {magFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, 0), makeInt(program, 0), makeInt(program, 0)]), (12 + 13 + 20 + 21) / 4);
>     checkFloat(program, callFunction(program, "foo36", [textureDepth2D, makeSampler(program, {magFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, 0), makeInt(program, 1), makeInt(program, 0)]), (13 + 14 + 21 + 22) / 4);
>     checkFloat(program, callFunction(program, "foo36", [textureDepth2D, makeSampler(program, {magFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 1 / texture2D.ePtr.loadValue().width), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, 1 / texture2D.ePtr.loadValue().height), makeInt(program, 1), makeInt(program, 0)]), (13 + 14 + 21 + 22) / 4);
>     checkFloat(program, callFunction(program, "foo36", [textureDepth2D, makeSampler(program, {minFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 2 / texture2D.ePtr.loadValue().width), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, 2 / texture2D.ePtr.loadValue().height), makeInt(program, 1), makeInt(program, 0)]), (35 + 36 + 39 + 40) / 4);
>-    checkFloat(program, callFunction(program, "foo36", [textureDepth2D, makeSampler(program, {minFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, Math.pow(2, 0.25) / texture2D.ePtr.loadValue().width), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, Math.pow(2, 0.25) / texture2D.ePtr.loadValue().height), makeInt(program, 1), makeInt(program, 0)]), (13 + 14 + 21 + 22) / 4);
>-    checkFloat(program, callFunction(program, "foo36", [textureDepth2D, makeSampler(program, {minFilter: "linear", mipmapFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, Math.pow(2, 0.5) / texture2D.ePtr.loadValue().width), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, Math.pow(2, 0.5) / texture2D.ePtr.loadValue().height), makeInt(program, 1), makeInt(program, 0)]), (((13 + 14 + 21 + 22) / 4) + ((35 + 36 + 39 + 40) / 4)) / 2);
>+    checkFloat(program, callFunction(program, "foo36", [textureDepth2D, makeSampler(program, {minFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeCastedFloat(program, Math.pow(2, 0.25) / texture2D.ePtr.loadValue().width), makeFloat(program, 0), makeFloat(program, 0), makeCastedFloat(program, Math.pow(2, 0.25) / texture2D.ePtr.loadValue().height), makeInt(program, 1), makeInt(program, 0)]), (13 + 14 + 21 + 22) / 4);
>+    checkFloat(program, callFunction(program, "foo36", [textureDepth2D, makeSampler(program, {minFilter: "linear", mipmapFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeCastedFloat(program, Math.pow(2, 0.5) / texture2D.ePtr.loadValue().width), makeFloat(program, 0), makeFloat(program, 0), makeCastedFloat(program, Math.pow(2, 0.5) / texture2D.ePtr.loadValue().height), makeInt(program, 1), makeInt(program, 0)]), (((13 + 14 + 21 + 22) / 4) + ((35 + 36 + 39 + 40) / 4)) / 2);
>     checkFloat(program, callFunction(program, "foo37", [textureDepth2D, makeSampler(program, {magFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 0)]), (12 + 13 + 20 + 21) / 4);
>     checkFloat(program, callFunction(program, "foo37", [textureDepth2D, makeSampler(program, {minFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 1)]), (34 + 35 + 38 + 39) / 4);
>     checkFloat(program, callFunction(program, "foo37", [textureDepth2D, makeSampler(program, {minFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 0.5)]), (12 + 13 + 20 + 21) / 4);
>@@ -7447,25 +7434,25 @@ tests.textureSample = function() {
>     checkFloat(program, callFunction(program, "foo41", [textureDepth2DArray, makeSampler(program, {magFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, 0)]), (12 + 13 + 20 + 21) / 4);
>     checkFloat(program, callFunction(program, "foo41", [textureDepth2DArray, makeSampler(program, {magFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 0), makeFloat(program, 1 / texture2D.ePtr.loadValue().width), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, 1 / texture2D.ePtr.loadValue().height)]), (12 + 13 + 20 + 21) / 4);
>     checkFloat(program, callFunction(program, "foo41", [textureDepth2DArray, makeSampler(program, {minFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 0), makeFloat(program, 2 / texture2D.ePtr.loadValue().width), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, 2 / texture2D.ePtr.loadValue().height)]), (34 + 35 + 38 + 39) / 4);
>-    checkFloat(program, callFunction(program, "foo41", [textureDepth2DArray, makeSampler(program, {minFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 0), makeFloat(program, Math.pow(2, 0.25) / texture2D.ePtr.loadValue().width), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, Math.pow(2, 0.25) / texture2D.ePtr.loadValue().height)]), (12 + 13 + 20 + 21) / 4);
>-    checkFloat(program, callFunction(program, "foo41", [textureDepth2DArray, makeSampler(program, {minFilter: "linear", mipmapFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 0), makeFloat(program, Math.pow(2, 0.5) / texture2D.ePtr.loadValue().width), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, Math.pow(2, 0.5) / texture2D.ePtr.loadValue().height)]), (((12 + 13 + 20 + 21) / 4) + ((34 + 35 + 38 + 39) / 4)) / 2);
>+    checkFloat(program, callFunction(program, "foo41", [textureDepth2DArray, makeSampler(program, {minFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 0), makeCastedFloat(program, Math.pow(2, 0.25) / texture2D.ePtr.loadValue().width), makeFloat(program, 0), makeFloat(program, 0), makeCastedFloat(program, Math.pow(2, 0.25) / texture2D.ePtr.loadValue().height)]), (12 + 13 + 20 + 21) / 4);
>+    checkFloat(program, callFunction(program, "foo41", [textureDepth2DArray, makeSampler(program, {minFilter: "linear", mipmapFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 0), makeCastedFloat(program, Math.pow(2, 0.5) / texture2D.ePtr.loadValue().width), makeFloat(program, 0), makeFloat(program, 0), makeCastedFloat(program, Math.pow(2, 0.5) / texture2D.ePtr.loadValue().height)]), (((12 + 13 + 20 + 21) / 4) + ((34 + 35 + 38 + 39) / 4)) / 2);
>     checkFloat(program, callFunction(program, "foo41", [textureDepth2DArray, makeSampler(program, {magFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 1), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, 0)]), (54 + 55 + 62 + 63) / 4);
>     checkFloat(program, callFunction(program, "foo41", [textureDepth2DArray, makeSampler(program, {magFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 1), makeFloat(program, 1 / texture2D.ePtr.loadValue().width), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, 1 / texture2D.ePtr.loadValue().height)]), (54 + 55 + 62 + 63) / 4);
>     checkFloat(program, callFunction(program, "foo41", [textureDepth2DArray, makeSampler(program, {minFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 1), makeFloat(program, 2 / texture2D.ePtr.loadValue().width), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, 2 / texture2D.ePtr.loadValue().height)]), (76 + 77 + 80 + 81) / 4);
>-    checkFloat(program, callFunction(program, "foo41", [textureDepth2DArray, makeSampler(program, {minFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 1), makeFloat(program, Math.pow(2, 0.25) / texture2D.ePtr.loadValue().width), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, Math.pow(2, 0.25) / texture2D.ePtr.loadValue().height)]), (54 + 55 + 62 + 63) / 4);
>-    checkFloat(program, callFunction(program, "foo41", [textureDepth2DArray, makeSampler(program, {minFilter: "linear", mipmapFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 1), makeFloat(program, Math.pow(2, 0.5) / texture2D.ePtr.loadValue().width), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, Math.pow(2, 0.5) / texture2D.ePtr.loadValue().height)]), (((54 + 55 + 62 + 63) / 4) + ((76 + 77 + 80 + 81) / 4)) / 2);
>+    checkFloat(program, callFunction(program, "foo41", [textureDepth2DArray, makeSampler(program, {minFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 1), makeCastedFloat(program, Math.pow(2, 0.25) / texture2D.ePtr.loadValue().width), makeFloat(program, 0), makeFloat(program, 0), makeCastedFloat(program, Math.pow(2, 0.25) / texture2D.ePtr.loadValue().height)]), (54 + 55 + 62 + 63) / 4);
>+    checkFloat(program, callFunction(program, "foo41", [textureDepth2DArray, makeSampler(program, {minFilter: "linear", mipmapFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 1), makeCastedFloat(program, Math.pow(2, 0.5) / texture2D.ePtr.loadValue().width), makeFloat(program, 0), makeFloat(program, 0), makeCastedFloat(program, Math.pow(2, 0.5) / texture2D.ePtr.loadValue().height)]), (((54 + 55 + 62 + 63) / 4) + ((76 + 77 + 80 + 81) / 4)) / 2);
>     checkFloat(program, callFunction(program, "foo42", [textureDepth2DArray, makeSampler(program, {magFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, 0), makeInt(program, 0), makeInt(program, 0)]), (12 + 13 + 20 + 21) / 4);
>     checkFloat(program, callFunction(program, "foo42", [textureDepth2DArray, makeSampler(program, {magFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, 0), makeInt(program, 1), makeInt(program, 0)]), (13 + 14 + 21 + 22) / 4);
>     checkFloat(program, callFunction(program, "foo42", [textureDepth2DArray, makeSampler(program, {magFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 0), makeFloat(program, 1 / texture2D.ePtr.loadValue().width), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, 1 / texture2D.ePtr.loadValue().height), makeInt(program, 1), makeInt(program, 0)]), (13 + 14 + 21 + 22) / 4);
>     checkFloat(program, callFunction(program, "foo42", [textureDepth2DArray, makeSampler(program, {minFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 0), makeFloat(program, 2 / texture2D.ePtr.loadValue().width), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, 2 / texture2D.ePtr.loadValue().height), makeInt(program, 1), makeInt(program, 0)]), (35 + 36 + 39 + 40) / 4);
>-    checkFloat(program, callFunction(program, "foo42", [textureDepth2DArray, makeSampler(program, {minFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 0), makeFloat(program, Math.pow(2, 0.25) / texture2D.ePtr.loadValue().width), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, Math.pow(2, 0.25) / texture2D.ePtr.loadValue().height), makeInt(program, 1), makeInt(program, 0)]), (13 + 14 + 21 + 22) / 4);
>-    checkFloat(program, callFunction(program, "foo42", [textureDepth2DArray, makeSampler(program, {minFilter: "linear", mipmapFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 0), makeFloat(program, Math.pow(2, 0.5) / texture2D.ePtr.loadValue().width), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, Math.pow(2, 0.5) / texture2D.ePtr.loadValue().height), makeInt(program, 1), makeInt(program, 0)]), (((13 + 14 + 21 + 22) / 4) + ((35 + 36 + 39 + 40) / 4)) / 2);
>+    checkFloat(program, callFunction(program, "foo42", [textureDepth2DArray, makeSampler(program, {minFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 0), makeCastedFloat(program, Math.pow(2, 0.25) / texture2D.ePtr.loadValue().width), makeFloat(program, 0), makeFloat(program, 0), makeCastedFloat(program, Math.pow(2, 0.25) / texture2D.ePtr.loadValue().height), makeInt(program, 1), makeInt(program, 0)]), (13 + 14 + 21 + 22) / 4);
>+    checkFloat(program, callFunction(program, "foo42", [textureDepth2DArray, makeSampler(program, {minFilter: "linear", mipmapFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 0), makeCastedFloat(program, Math.pow(2, 0.5) / texture2D.ePtr.loadValue().width), makeFloat(program, 0), makeFloat(program, 0), makeCastedFloat(program, Math.pow(2, 0.5) / texture2D.ePtr.loadValue().height), makeInt(program, 1), makeInt(program, 0)]), (((13 + 14 + 21 + 22) / 4) + ((35 + 36 + 39 + 40) / 4)) / 2);
>     checkFloat(program, callFunction(program, "foo42", [textureDepth2DArray, makeSampler(program, {magFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 1), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, 0), makeInt(program, 0), makeInt(program, 0)]), (54 + 55 + 62 + 63) / 4);
>     checkFloat(program, callFunction(program, "foo42", [textureDepth2DArray, makeSampler(program, {magFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 1), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, 0), makeInt(program, 1), makeInt(program, 0)]), (55 + 56 + 63 + 64) / 4);
>     checkFloat(program, callFunction(program, "foo42", [textureDepth2DArray, makeSampler(program, {magFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 1), makeFloat(program, 1 / texture2D.ePtr.loadValue().width), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, 1 / texture2D.ePtr.loadValue().height), makeInt(program, 1), makeInt(program, 0)]), (55 + 56 + 63 + 64) / 4);
>     checkFloat(program, callFunction(program, "foo42", [textureDepth2DArray, makeSampler(program, {minFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 1), makeFloat(program, 2 / texture2D.ePtr.loadValue().width), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, 2 / texture2D.ePtr.loadValue().height), makeInt(program, 1), makeInt(program, 0)]), (77 + 78 + 81 + 82) / 4);
>-    checkFloat(program, callFunction(program, "foo42", [textureDepth2DArray, makeSampler(program, {minFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 1), makeFloat(program, Math.pow(2, 0.25) / texture2D.ePtr.loadValue().width), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, Math.pow(2, 0.25) / texture2D.ePtr.loadValue().height), makeInt(program, 1), makeInt(program, 0)]), (55 + 56 + 63 + 64) / 4);
>-    checkFloat(program, callFunction(program, "foo42", [textureDepth2DArray, makeSampler(program, {minFilter: "linear", mipmapFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 1), makeFloat(program, Math.pow(2, 0.5) / texture2D.ePtr.loadValue().width), makeFloat(program, 0), makeFloat(program, 0), makeFloat(program, Math.pow(2, 0.5) / texture2D.ePtr.loadValue().height), makeInt(program, 1), makeInt(program, 0)]), (((55 + 56 + 63 + 64) / 4) + ((77 + 78 + 81 + 82) / 4)) / 2);
>+    checkFloat(program, callFunction(program, "foo42", [textureDepth2DArray, makeSampler(program, {minFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 1), makeCastedFloat(program, Math.pow(2, 0.25) / texture2D.ePtr.loadValue().width), makeFloat(program, 0), makeFloat(program, 0), makeCastedFloat(program, Math.pow(2, 0.25) / texture2D.ePtr.loadValue().height), makeInt(program, 1), makeInt(program, 0)]), (55 + 56 + 63 + 64) / 4);
>+    checkFloat(program, callFunction(program, "foo42", [textureDepth2DArray, makeSampler(program, {minFilter: "linear", mipmapFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 1), makeCastedFloat(program, Math.pow(2, 0.5) / texture2D.ePtr.loadValue().width), makeFloat(program, 0), makeFloat(program, 0), makeCastedFloat(program, Math.pow(2, 0.5) / texture2D.ePtr.loadValue().height), makeInt(program, 1), makeInt(program, 0)]), (((55 + 56 + 63 + 64) / 4) + ((77 + 78 + 81 + 82) / 4)) / 2);
>     checkFloat(program, callFunction(program, "foo43", [textureDepth2DArray, makeSampler(program, {magFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 0), makeFloat(program, 0)]), (12 + 13 + 20 + 21) / 4);
>     checkFloat(program, callFunction(program, "foo43", [textureDepth2DArray, makeSampler(program, {minFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 0), makeFloat(program, 1)]), (34 + 35 + 38 + 39) / 4);
>     checkFloat(program, callFunction(program, "foo43", [textureDepth2DArray, makeSampler(program, {minFilter: "linear"}), makeFloat(program, 0.5), makeFloat(program, 0.5), makeFloat(program, 0), makeFloat(program, 0.5)]), (12 + 13 + 20 + 21) / 4);

Actions: View | Formatted Diff | Diff

Attachments on bug 189502: 349399 | 349429 | 349456 | 350126