subgroups on mac are broken

[ArthurBrussee]

Subgroups (or SIMD in metal speak) were accidentally disabled for a while on the MSL backend. Recent changes to the feature registration meaning it now does work! However, I ran into

2025-09-22T20:51:16Z ERROR wgpu_core::device::resource] Shader error: MSL: "program_source:123:22: error: no matching function for call to 'simd_sum'\nconst float_4 l_61 = simd_sum(l_60);\n                     ^~~~~~~~\n/System/Library/PrivateFrameworks/GPUCompiler.framework/Versions/32023/Libraries/lib/clang/32023.830/include/metal/metal_simdgroup:274:14: note: candidate template ignored: requirement '__is_valid_simdgroup_type<float_4, void>::value' was not satisfied [with T = float_4]\nMETAL_FUNC T simd_sum(T data)\n             ^\nprogram_source:133:12: warning: unused variable 'l_66' [-Wunused-variable]\nconst uint l_66 = l_arr_0[l_65];\n           ^\n"
[2025-09-22T20:51:16Z ERROR wgpu::backend::wgpu_core] Handling wgpu errors as fatal by default

thread 'tokio-runtime-worker' panicked at /Users/arthurbrussee/.cargo/git/checkouts/wgpu-50febe783b343bfe/bdf430c/wgpu/src/backend/wgpu_core.rs:1055:30:
wgpu error: Validation Error

Caused by:
  In Device::create_shader_module_passthrough, label = 'reduce_kernel'
    Failed to generate the backend-specific code


note: run with `RUST_BACKTRACE=1` environment variable to display a backtrace

it seems that some functions try to operate on the generated float_4 type rather than a native float4 for which the SIMD functions are defined.