Fix row calculation in Queue::write_texture.

Fix row calculation in `Queue::write_texture`.

The D3D12_SUBRESOURCE_FOOTPRINT was calculated incorrectly by multiplying the height by the block count.

DirectX automatically takes format into account because it's passed into the footprint. Only the block width is necessary to compute the pitch or linear size. This prevents a crash when using `Queue::write_texture` with a block compressed texture.

CHANGELOG.md

@@ -182,6 +182,10 @@ Passing an owned value `window` to `Surface` will return a `Surface<'static>`. S
 - Align `wgpu_types::CompositeAlphaMode` serde serialization to spec. By @littledivy in [#4940](https://github.com/gfx-rs/wgpu/pull/4940)
 - Fix error message of `ConfigureSurfaceError::TooLarge`. By @Dinnerbone in [#4960](https://github.com/gfx-rs/wgpu/pull/4960)
 
+#### DX12
+
+- Fixed D3D12_SUBRESOURCE_FOOTPRINT calculation for block compressed textures which caused a crash with `Queue::write_texture` on DX12. By @DTZxPorter in [#4990](https://github.com/gfx-rs/wgpu/pull/4990)
+
 #### Vulkan
 
 - Use `VK_EXT_robustness2` only when not using an outdated intel iGPU driver. By @TheoDulka in [#4602](https://github.com/gfx-rs/wgpu/pull/4602).

wgpu-core/src/device/queue.rs

@@ -740,7 +740,7 @@ impl<G: GlobalIdentityHandlerFactory> Global<G> {
             // doesn't really matter because we need this only if we copy
             // more than one layer, and then we validate for this being not
             // None
-            size.height,
+            height_blocks,
         );
 
         let block_size = dst
@@ -754,7 +754,7 @@ impl<G: GlobalIdentityHandlerFactory> Global<G> {
             wgt::math::align_to(block_size * width_blocks, bytes_per_row_alignment);
 
         let block_rows_in_copy =
-            (size.depth_or_array_layers - 1) * block_rows_per_image + height_blocks;
+            (size.depth_or_array_layers - 1) * block_rows_per_image + block_rows_per_image;
         let stage_size = stage_bytes_per_row as u64 * block_rows_in_copy as u64;
 
         let mut pending_writes = device.pending_writes.lock();

wgpu-hal/src/dx12/command.rs

@@ -33,7 +33,9 @@ impl crate::BufferTextureCopy {
                 Height: self
                     .buffer_layout
                     .rows_per_image
-                    .map_or(self.size.height, |count| count * block_height),
+                    .map_or(self.size.height, |height_blocks| {
+                        height_blocks * block_height
+                    }),
                 Depth: self.size.depth,
                 RowPitch: {
                     let actual = self.buffer_layout.bytes_per_row.unwrap_or_else(|| {