[naga] Allow Scalar for MathFunction::Distance (gfx-rs#7530)

Co-authored-by: Erich Gubler <[email protected]>

Author: 2 people

CHANGELOG.md

@@ -87,6 +87,12 @@ Naga now infers the correct binding layout when a resource appears only in an as
 
 - Fix validation error when creating a non-mappable buffer using the committed allocation scheme. By @cwfitzgerald and @ErichDonGubler in [#7519](https://github.com/gfx-rs/wgpu/pull/7519).
 
+### Bug Fixes
+
+#### Naga
+
+- Allow scalars as the first argument of the `distance` built-in function. By @bernhl in [#7530](https://github.com/gfx-rs/wgpu/pull/7530).
+
 ## v25.0.1 (2025-04-11)
 
 ### Bug Fixes

naga/src/proc/overloads/mathfunction.rs

@@ -92,7 +92,9 @@ impl ir::MathFunction {
             Mf::Ldexp => ldexp().into(),
             Mf::Outer => outer().into(),
             Mf::Cross => regular!(2, VEC3 of FLOAT).into(),
-            Mf::Distance => regular!(2, VECN of FLOAT_ABSTRACT_UNIMPLEMENTED -> Scalar).into(),
+            Mf::Distance => {
+                regular!(2, SCALAR|VECN of FLOAT_ABSTRACT_UNIMPLEMENTED -> Scalar).into()
+            }
             Mf::Length => regular!(1, SCALAR|VECN of FLOAT_ABSTRACT_UNIMPLEMENTED -> Scalar).into(),
             Mf::Normalize => regular!(1, VECN of FLOAT_ABSTRACT_UNIMPLEMENTED).into(),
             Mf::FaceForward => regular!(3, VECN of FLOAT_ABSTRACT_UNIMPLEMENTED).into(),

naga/tests/naga/wgsl_errors.rs

@@ -3419,6 +3419,9 @@ fn too_many_arguments_2() {
   │                    ^^^^^^^^              ^^ argument #2 has type `i32`
   │
   = note: `distance` accepts the following types for argument #2:
+  = note: allowed type: f32
+  = note: allowed type: f16
+  = note: allowed type: f64
   = note: allowed type: vec2<f32>
   = note: allowed type: vec2<f16>
   = note: allowed type: vec2<f64>