fix: no GPU allocation path maps, however small the buffer

The previous commit converted the allocations that scale with the
drawing and deliberately left the small fixed-size ones -- uniforms, the
viewcube's static geometry, per-image and per-wipeout quads -- reasoning
that an allocation failure was not a realistic outcome for a few dozen
bytes.

That reasoning was wrong, and the log says so precisely. On the
reproducer the vertex buffers degraded exactly as intended:

  [gpu] uncaptured ... Out of Memory
  [gpu] uncaptured ... In Queue::write_buffer
          Buffer with 'block_wire.vertices' label is invalid
  ... errors #4 #8 #16 #32 #64 #128, session kept alive

and then the process died here:

  thread 'main' panicked at wgpu_core.rs:2253:
  Error in Buffer::get_mapped_range: Validation Error
  Caused by: Buffer with 'block_wire.const' label is invalid

'block_wire.const' is a single uniform struct. Once the device is out of
memory every allocation fails, including a 32-byte one, and
create_buffer_init maps it and panics. Size was never the relevant
property -- whether the path maps is.

So every remaining allocation goes through gpu_upload too: block-wire
const, the viewcube's seven buffers, the blit and clip-boundary buffers,
image instances/params/vertices, mesh material params and surface and the
empty index stub, the texture-hatch uniform, and all five wipeout
buffers. `grep -r create_buffer_init src` now finds only a comment, and
no mapped_at_creation:true remains outside gpu_upload's own docs.

`queue` is threaded through the last holdouts: upload_wipeouts,
upload_clip_boundary, MeshBatchStubs::new and
create_material_bind_group_from_resources.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

src/scene/pipeline/mesh_gpu.rs

@@ -2,7 +2,6 @@
 
 use crate::scene::model::mesh_model::{MeshLodSet, MeshModel};
 use iced::wgpu;
-use iced::wgpu::util::DeviceExt;
 
 // ── Vertex layout ─────────────────────────────────────────────────────────
 
@@ -377,13 +376,15 @@ struct MeshBatchStubs {
 }
 
 impl MeshBatchStubs {
-    fn new(device: &wgpu::Device) -> Self {
+    fn new(device: &wgpu::Device, queue: &wgpu::Queue) -> Self {
         Self {
-            index: device.create_buffer_init(&wgpu::util::BufferInitDescriptor {
-                label: Some("mesh.batch.empty_ibuf"),
-                contents: bytemuck::cast_slice(&[0u32]),
-                usage: wgpu::BufferUsages::INDEX,
-            }),
+            index: super::gpu_upload::upload_buffer(
+                device,
+                queue,
+                "mesh.batch.empty_ibuf",
+                &[0u32],
+                wgpu::BufferUsages::INDEX,
+            ),
             vertex: device.create_buffer(&wgpu::BufferDescriptor {
                 label: Some("mesh.batch.empty_vbuf"),
                 size: std::mem::size_of::<MeshVertex>() as u64,
@@ -784,11 +785,13 @@ pub fn create_material_resources(
             [material.color_bleed_scale, 0.0, 0.0, 0.0]
         }),
     };
-    let params_buffer = device.create_buffer_init(&wgpu::util::BufferInitDescriptor {
-        label: Some("mesh.material.params"),
-        contents: bytemuck::bytes_of(&params),
-        usage: wgpu::BufferUsages::UNIFORM,
-    });
+    let params_buffer = super::gpu_upload::upload_buffer(
+        device,
+        queue,
+        "mesh.material.params",
+        std::slice::from_ref(&params),
+        wgpu::BufferUsages::UNIFORM,
+    );
     MeshMaterialResources {
         diffuse_view,
         specular_view,
@@ -810,6 +813,7 @@ pub fn create_material_resources(
 
 pub fn create_material_bind_group_from_resources(
     device: &wgpu::Device,
+    queue: &wgpu::Queue,
     layout: &wgpu::BindGroupLayout,
     resources: &MeshMaterialResources,
     material: Option<&crate::scene::model::material_model::MeshMaterial>,
@@ -825,11 +829,13 @@ pub fn create_material_bind_group_from_resources(
         advanced,
         flags,
     };
-    let surface_buffer = device.create_buffer_init(&wgpu::util::BufferInitDescriptor {
-        label: Some("mesh.material.surface"),
-        contents: bytemuck::bytes_of(&surface),
-        usage: wgpu::BufferUsages::UNIFORM,
-    });
+    let surface_buffer = super::gpu_upload::upload_buffer(
+        device,
+        queue,
+        "mesh.material.surface",
+        std::slice::from_ref(&surface),
+        wgpu::BufferUsages::UNIFORM,
+    );
     device.create_bind_group(&wgpu::BindGroupDescriptor {
         label: Some("mesh.material.bind_group"),
         layout,
@@ -912,6 +918,7 @@ pub fn create_material_bind_group(
     let color = material.map_or([0.8, 0.8, 0.8, 1.0], |material| material.diffuse);
     create_material_bind_group_from_resources(
         device,
+        queue,
         layout,
         &resources,
         material,
@@ -936,6 +943,7 @@ pub fn upload_chunk_material_bind_groups(
                     let resources = create_material_resources(device, queue, Some(material));
                     chunk.material_bind_group = Some(create_material_bind_group_from_resources(
                         device,
+                        queue,
                         layout,
                         &resources,
                         chunk.material.as_ref(),
@@ -950,6 +958,7 @@ pub fn upload_chunk_material_bind_groups(
         };
         chunk.material_bind_group = Some(create_material_bind_group_from_resources(
             device,
+            queue,
             layout,
             resources,
             chunk.material.as_ref(),
@@ -1586,7 +1595,7 @@ pub fn build_mesh_batch_filtered(
     // vertices). Bound chunks by the largest buffer produced for wire meshes.
     let max_tris =
         (budget / (6 * std::mem::size_of::<MeshEdgeVertex>())).max(1);
-    let stubs = MeshBatchStubs::new(device);
+    let stubs = MeshBatchStubs::new(device, queue);
 
     let mut chunks = Vec::new();
     let mut verts: Vec<MeshVertex> = Vec::new();