Metal: Force Compute shaders workgroup max thread count

This forces the max thread count to be inlined with what
blender expects. While this will make some shaders less
eficient it avoid crashes / assert on shaders that
could not compile. The performance impacts can be fixed
by tackling the individual cases.

source/blender/gpu/metal/mtl_shader.mm

@@ -1390,9 +1390,15 @@ bool MTLShader::bake_compute_pipeline_state(MTLContext *ctx)
     }
 
     /* Compile PSO. */
+    MTLComputePipelineDescriptor *desc = [[MTLComputePipelineDescriptor alloc] init];
+    desc.maxTotalThreadsPerThreadgroup = 1024;
+    desc.computeFunction = compute_function;
+
     id<MTLComputePipelineState> pso = [ctx->device
-        newComputePipelineStateWithFunction:compute_function
-                                      error:&error];
+        newComputePipelineStateWithDescriptor:desc
+                                      options:MTLPipelineOptionNone
+                                   reflection:nullptr
+                                        error:&error];
 
     if (error) {
       NSLog(@"Failed to create PSO for compute shader: %s error %@\n", this->name, error);