Fix unpackUnorm4x8 and unpackSnorm4x8 (#4033)

Fixes #4031

Each component of unpackU/Snorm4x8 had to be masked for 8bits.

1 files changed, 20 insertions, 6 deletions

diff --git a/source/slang/glsl.meta.slang b/source/slang/glsl.meta.slang
index b4e5cf7ab..508156b8c 100644
--- a/source/slang/glsl.meta.slang
+++ b/source/slang/glsl.meta.slang
@@ -633,28 +633,34 @@ uint packSnorm1x8(float c)
 [ForceInline]
 float unpackUnorm1x16(uint p)
 {
-    return float(p) / 65535.0;
+    const uint wordMask = 0xffff;
+    return float(p & wordMask) / 65535.0;
 }
 
 [__readNone]
 [ForceInline]
 float unpackSnorm1x16(uint p)
 {
-    return clamp((float(p) - 32767.0) / 32767.0, -1.0, 1.0);
+    const uint wordMask = 0xffff;
+    return clamp((float(p & wordMask) - 32767.0) / 32767.0, -1.0, 1.0);
 }
 
+__target_intrinsic(glsl)
 [__readNone]
 [ForceInline]
 float unpackUnorm1x8(uint p)
 {
-    return float(p) / 255.0;
+    const uint byteMask = 0xff;
+    return float(p & byteMask) / 255.0;
 }
 
+__target_intrinsic(glsl)
 [__readNone]
 [ForceInline]
 float unpackSnorm1x8(uint p)
 {
-    return clamp((float(p) - 127.0) / 127.0, -1.0, 1.0);
+    const uint byteMask = 0xff;
+    return clamp((float(p & byteMask) - 127.0) / 127.0, -1.0, 1.0);
 }
 
 [__readNone]
@@ -733,7 +739,11 @@ __target_intrinsic(glsl)
 [require(cpp_cuda_glsl_hlsl_spirv, shader5_sm_4_0)]
 public vec4 unpackUnorm4x8(highp uint p)
 {
-    return vec4(unpackUnorm1x8(p & uint(0xffff)), unpackUnorm1x8(p >> uint(8)), unpackUnorm1x8(p >> uint(16)), unpackUnorm1x8(p >> uint(24)));
+    return vec4(
+        unpackUnorm1x8(p),
+        unpackUnorm1x8(p >> 8),
+        unpackUnorm1x8(p >> 16),
+        unpackUnorm1x8(p >> 24));
 }
 
 __target_intrinsic(glsl)
@@ -742,7 +752,11 @@ __target_intrinsic(glsl)
 [require(cpp_cuda_glsl_hlsl_spirv, shader5_sm_4_0)]
 public vec4 unpackSnorm4x8(highp uint p)
 {
-    return vec4(unpackSnorm1x8(p & uint(0xffff)), unpackSnorm1x8(p >> uint(8)), unpackSnorm1x8(p >> uint(16)), unpackSnorm1x8(p >> uint(24)));
+    return vec4(
+        unpackSnorm1x8(p),
+        unpackSnorm1x8(p >> 8),
+        unpackSnorm1x8(p >> 16),
+        unpackSnorm1x8(p >> 24));
 }
 
 __target_intrinsic(glsl)