Make bind group an Option for set_bind_group calls. (#6216)

This is just an API change for all the "set_bind_group" calls. Calls that pass a Some() argument should have unchanged behavior. The None cases are left as TODOs.
gfx-rs · Sep 6, 2024 · 9b36a3e · 9b36a3e
1 parent c87717b
commit 9b36a3e
Show file tree

Hide file tree

Showing 59 changed files with 283 additions and 169 deletions.
diff --git a/CHANGELOG.md b/CHANGELOG.md
@@ -65,6 +65,15 @@ traits that have now been implemented for `wgpu` resources.
 
 By @teoxoy [#6134](https://github.com/gfx-rs/wgpu/pull/6134).
 
+#### `set_bind_group` now takes an `Option` for the bind group argument.
+
+https://gpuweb.github.io/gpuweb/#programmable-passes-bind-groups specifies that bindGroup
+is nullable. This change is the start of implementing this part of the spec. Callers that
+specify a `Some()` value should have unchanged behavior. Handling of `None` values still
+needs to be implemented by backends.
+
+By @bradwerth [#6216](https://github.com/gfx-rs/wgpu/pull/6216).
+
 ### New Features
 
 #### Naga

diff --git a/benches/benches/computepass.rs b/benches/benches/computepass.rs
@@ -389,7 +389,7 @@ impl ComputepassState {
  let end_idx = start_idx + dispatch_per_pass;
  for dispatch_idx in start_idx..end_idx {
  compute_pass.set_pipeline(&self.pipeline);
- compute_pass.set_bind_group(0, &self.bind_groups[dispatch_idx], &[]);
+ compute_pass.set_bind_group(0, Some(&self.bind_groups[dispatch_idx]), &[]);
  compute_pass.dispatch_workgroups(1, 1, 1);
  }
 
@@ -412,7 +412,7 @@ impl ComputepassState {
  });
 
  compute_pass.set_pipeline(self.bindless_pipeline.as_ref().unwrap());
- compute_pass.set_bind_group(0, self.bindless_bind_group.as_ref().unwrap(), &[]);
+ compute_pass.set_bind_group(0, Some(self.bindless_bind_group.as_ref().unwrap()), &[]);
  for _ in 0..dispatch_count_bindless {
  compute_pass.dispatch_workgroups(1, 1, 1);
  }

diff --git a/benches/benches/renderpass.rs b/benches/benches/renderpass.rs
@@ -367,7 +367,7 @@ impl RenderpassState {
  let end_idx = start_idx + draws_per_pass;
  for draw_idx in start_idx..end_idx {
  render_pass.set_pipeline(&self.pipeline);
- render_pass.set_bind_group(0, &self.bind_groups[draw_idx], &[]);
+ render_pass.set_bind_group(0, Some(&self.bind_groups[draw_idx]), &[]);
  for i in 0..VERTEX_BUFFERS_PER_DRAW {
  render_pass.set_vertex_buffer(
  i as u32,
@@ -410,7 +410,7 @@ impl RenderpassState {
  });
 
  render_pass.set_pipeline(self.bindless_pipeline.as_ref().unwrap());
- render_pass.set_bind_group(0, self.bindless_bind_group.as_ref().unwrap(), &[]);
+ render_pass.set_bind_group(0, Some(self.bindless_bind_group.as_ref().unwrap()), &[]);
  for i in 0..VERTEX_BUFFERS_PER_DRAW {
  render_pass.set_vertex_buffer(i as u32, self.vertex_buffers[0].slice(..));
  }

diff --git a/deno_webgpu/bundle.rs b/deno_webgpu/bundle.rs
@@ -150,7 +150,7 @@ pub fn op_webgpu_render_bundle_encoder_set_bind_group(
  wgpu_core::command::bundle_ffi::wgpu_render_bundle_set_bind_group(
  &mut render_bundle_encoder_resource.0.borrow_mut(),
  index,
- bind_group_resource.1,
+ Some(bind_group_resource.1),
  dynamic_offsets_data.as_ptr(),
  dynamic_offsets_data.len(),
  );

diff --git a/deno_webgpu/compute_pass.rs b/deno_webgpu/compute_pass.rs
@@ -136,7 +136,7 @@ pub fn op_webgpu_compute_pass_set_bind_group(
  .compute_pass_set_bind_group(
  &mut compute_pass_resource.0.borrow_mut(),
  index,
- bind_group_resource.1,
+ Some(bind_group_resource.1),
  dynamic_offsets_data,
  )?;
 

diff --git a/deno_webgpu/render_pass.rs b/deno_webgpu/render_pass.rs
@@ -231,7 +231,7 @@ pub fn op_webgpu_render_pass_set_bind_group(
  .render_pass_set_bind_group(
  &mut render_pass_resource.0.borrow_mut(),
  index,
- bind_group_resource.1,
+ Some(bind_group_resource.1),
  dynamic_offsets_data,
  )?;
 

diff --git a/examples/src/boids/mod.rs b/examples/src/boids/mod.rs
@@ -298,7 +298,7 @@ impl crate::framework::Example for Example {
  timestamp_writes: None,
  });
  cpass.set_pipeline(&self.compute_pipeline);
- cpass.set_bind_group(0, &self.particle_bind_groups[self.frame_num % 2], &[]);
+ cpass.set_bind_group(0, Some(&self.particle_bind_groups[self.frame_num % 2]), &[]);
  cpass.dispatch_workgroups(self.work_group_count, 1, 1);
  }
  command_encoder.pop_debug_group();

diff --git a/examples/src/bunnymark/mod.rs b/examples/src/bunnymark/mod.rs
@@ -128,11 +128,11 @@ impl Example {
  occlusion_query_set: None,
  });
  rpass.set_pipeline(&self.pipeline);
- rpass.set_bind_group(0, &self.global_group, &[]);
+ rpass.set_bind_group(0, Some(&self.global_group), &[]);
  for i in 0..self.bunnies.len() {
  let offset =
  (i as wgpu::DynamicOffset) * (uniform_alignment as wgpu::DynamicOffset);
- rpass.set_bind_group(1, &self.local_group, &[offset]);
+ rpass.set_bind_group(1, Some(&self.local_group), &[offset]);
  rpass.draw(0..4, 0..1);
  }
  }

diff --git a/examples/src/conservative_raster/mod.rs b/examples/src/conservative_raster/mod.rs
@@ -305,7 +305,7 @@ impl crate::framework::Example for Example {
  });
 
  rpass.set_pipeline(&self.pipeline_upscale);
- rpass.set_bind_group(0, &self.bind_group_upscale, &[]);
+ rpass.set_bind_group(0, Some(&self.bind_group_upscale), &[]);
  rpass.draw(0..3, 0..1);
 
  if let Some(pipeline_lines) = &self.pipeline_lines {

diff --git a/examples/src/cube/mod.rs b/examples/src/cube/mod.rs
@@ -361,7 +361,7 @@ impl crate::framework::Example for Example {
  });
  rpass.push_debug_group("Prepare data for draw.");
  rpass.set_pipeline(&self.pipeline);
- rpass.set_bind_group(0, &self.bind_group, &[]);
+ rpass.set_bind_group(0, Some(&self.bind_group), &[]);
  rpass.set_index_buffer(self.index_buf.slice(..), wgpu::IndexFormat::Uint16);
  rpass.set_vertex_buffer(0, self.vertex_buf.slice(..));
  rpass.pop_debug_group();

diff --git a/examples/src/hello_compute/mod.rs b/examples/src/hello_compute/mod.rs
@@ -135,7 +135,7 @@ async fn execute_gpu_inner(
  timestamp_writes: None,
  });
  cpass.set_pipeline(&compute_pipeline);
- cpass.set_bind_group(0, &bind_group, &[]);
+ cpass.set_bind_group(0, Some(&bind_group), &[]);
  cpass.insert_debug_marker("compute collatz iterations");
  cpass.dispatch_workgroups(numbers.len() as u32, 1, 1); // Number of cells to run, the (x,y,z) size of item being processed
  }

diff --git a/examples/src/hello_synchronization/mod.rs b/examples/src/hello_synchronization/mod.rs
@@ -128,7 +128,7 @@ async fn execute(
  timestamp_writes: None,
  });
  compute_pass.set_pipeline(&patient_pipeline);
- compute_pass.set_bind_group(0, &bind_group, &[]);
+ compute_pass.set_bind_group(0, Some(&bind_group), &[]);
  compute_pass.dispatch_workgroups(local_patient_workgroup_results.len() as u32, 1, 1);
  }
  queue.submit(Some(command_encoder.finish()));
@@ -150,7 +150,7 @@ async fn execute(
  timestamp_writes: None,
  });
  compute_pass.set_pipeline(&hasty_pipeline);
- compute_pass.set_bind_group(0, &bind_group, &[]);
+ compute_pass.set_bind_group(0, Some(&bind_group), &[]);
  compute_pass.dispatch_workgroups(local_patient_workgroup_results.len() as u32, 1, 1);
  }
  queue.submit(Some(command_encoder.finish()));

diff --git a/examples/src/hello_workgroups/mod.rs b/examples/src/hello_workgroups/mod.rs
@@ -127,7 +127,7 @@ async fn run() {
  timestamp_writes: None,
  });
  compute_pass.set_pipeline(&pipeline);
- compute_pass.set_bind_group(0, &bind_group, &[]);
+ compute_pass.set_bind_group(0, Some(&bind_group), &[]);
  /* Note that since each workgroup will cover both arrays, we only need to
  cover the length of one array. */
  compute_pass.dispatch_workgroups(local_a.len() as u32, 1, 1);

diff --git a/examples/src/mipmap/mod.rs b/examples/src/mipmap/mod.rs
@@ -180,7 +180,7 @@ impl Example {
  );
  }
  rpass.set_pipeline(&pipeline);
- rpass.set_bind_group(0, &bind_group, &[]);
+ rpass.set_bind_group(0, Some(&bind_group), &[]);
  rpass.draw(0..3, 0..1);
  if let Some(ref query_sets) = query_sets {
  rpass.write_timestamp(&query_sets.timestamp, timestamp_query_index_base + 1);
@@ -497,7 +497,7 @@ impl crate::framework::Example for Example {
  occlusion_query_set: None,
  });
  rpass.set_pipeline(&self.draw_pipeline);
- rpass.set_bind_group(0, &self.bind_group, &[]);
+ rpass.set_bind_group(0, Some(&self.bind_group), &[]);
  rpass.draw(0..4, 0..1);
  }
 

diff --git a/examples/src/repeated_compute/mod.rs b/examples/src/repeated_compute/mod.rs
@@ -59,7 +59,7 @@ async fn compute(local_buffer: &mut [u32], context: &WgpuContext) {
  timestamp_writes: None,
  });
  compute_pass.set_pipeline(&context.pipeline);
- compute_pass.set_bind_group(0, &context.bind_group, &[]);
+ compute_pass.set_bind_group(0, Some(&context.bind_group), &[]);
  compute_pass.dispatch_workgroups(local_buffer.len() as u32, 1, 1);
  }
  // We finish the compute pass by dropping it.

diff --git a/examples/src/shadow/mod.rs b/examples/src/shadow/mod.rs
@@ -779,10 +779,10 @@ impl crate::framework::Example for Example {
  occlusion_query_set: None,
  });
  pass.set_pipeline(&self.shadow_pass.pipeline);
- pass.set_bind_group(0, &self.shadow_pass.bind_group, &[]);
+ pass.set_bind_group(0, Some(&self.shadow_pass.bind_group), &[]);
 
  for entity in &self.entities {
- pass.set_bind_group(1, &self.entity_bind_group, &[entity.uniform_offset]);
+ pass.set_bind_group(1, Some(&self.entity_bind_group), &[entity.uniform_offset]);
  pass.set_index_buffer(entity.index_buf.slice(..), entity.index_format);
  pass.set_vertex_buffer(0, entity.vertex_buf.slice(..));
  pass.draw_indexed(0..entity.index_count as u32, 0, 0..1);
@@ -823,10 +823,10 @@ impl crate::framework::Example for Example {
  occlusion_query_set: None,
  });
  pass.set_pipeline(&self.forward_pass.pipeline);
- pass.set_bind_group(0, &self.forward_pass.bind_group, &[]);
+ pass.set_bind_group(0, Some(&self.forward_pass.bind_group), &[]);
 
  for entity in &self.entities {
- pass.set_bind_group(1, &self.entity_bind_group, &[entity.uniform_offset]);
+ pass.set_bind_group(1, Some(&self.entity_bind_group), &[entity.uniform_offset]);
  pass.set_index_buffer(entity.index_buf.slice(..), entity.index_format);
  pass.set_vertex_buffer(0, entity.vertex_buf.slice(..));
  pass.draw_indexed(0..entity.index_count as u32, 0, 0..1);

diff --git a/examples/src/skybox/mod.rs b/examples/src/skybox/mod.rs
@@ -451,7 +451,7 @@ impl crate::framework::Example for Example {
  occlusion_query_set: None,
  });
 
- rpass.set_bind_group(0, &self.bind_group, &[]);
+ rpass.set_bind_group(0, Some(&self.bind_group), &[]);
  rpass.set_pipeline(&self.entity_pipeline);
 
  for entity in self.entities.iter() {

diff --git a/examples/src/srgb_blend/mod.rs b/examples/src/srgb_blend/mod.rs
@@ -202,7 +202,7 @@ impl<const SRGB: bool> crate::framework::Example for Example<SRGB> {
  });
  rpass.push_debug_group("Prepare data for draw.");
  rpass.set_pipeline(&self.pipeline);
- rpass.set_bind_group(0, &self.bind_group, &[]);
+ rpass.set_bind_group(0, Some(&self.bind_group), &[]);
  rpass.set_index_buffer(self.index_buf.slice(..), wgpu::IndexFormat::Uint16);
  rpass.set_vertex_buffer(0, self.vertex_buf.slice(..));
  rpass.pop_debug_group();

diff --git a/examples/src/storage_texture/mod.rs b/examples/src/storage_texture/mod.rs
@@ -117,7 +117,7 @@ async fn run(_path: Option<String>) {
  label: None,
  timestamp_writes: None,
  });
- compute_pass.set_bind_group(0, &bind_group, &[]);
+ compute_pass.set_bind_group(0, Some(&bind_group), &[]);
  compute_pass.set_pipeline(&pipeline);
  compute_pass.dispatch_workgroups(TEXTURE_DIMS.0 as u32, TEXTURE_DIMS.1 as u32, 1);
  }

diff --git a/examples/src/texture_arrays/mod.rs b/examples/src/texture_arrays/mod.rs
@@ -391,12 +391,12 @@ impl crate::framework::Example for Example {
  rpass.set_vertex_buffer(0, self.vertex_buffer.slice(..));
  rpass.set_index_buffer(self.index_buffer.slice(..), self.index_format);
  if self.uniform_workaround {
- rpass.set_bind_group(0, &self.bind_group, &[0]);
+ rpass.set_bind_group(0, Some(&self.bind_group), &[0]);
  rpass.draw_indexed(0..6, 0, 0..1);
- rpass.set_bind_group(0, &self.bind_group, &[256]);
+ rpass.set_bind_group(0, Some(&self.bind_group), &[256]);
  rpass.draw_indexed(6..12, 0, 0..1);
  } else {
- rpass.set_bind_group(0, &self.bind_group, &[0]);
+ rpass.set_bind_group(0, Some(&self.bind_group), &[0]);
  rpass.draw_indexed(0..12, 0, 0..1);
  }
 

diff --git a/examples/src/timestamp_queries/mod.rs b/examples/src/timestamp_queries/mod.rs
@@ -324,7 +324,7 @@ fn compute_pass(
  });
  *next_unused_query += 2;
  cpass.set_pipeline(&compute_pipeline);
- cpass.set_bind_group(0, &bind_group, &[]);
+ cpass.set_bind_group(0, Some(&bind_group), &[]);
  cpass.dispatch_workgroups(1, 1, 1);
  if device
  .features()

diff --git a/examples/src/uniform_values/mod.rs b/examples/src/uniform_values/mod.rs
@@ -327,7 +327,11 @@ async fn run(event_loop: EventLoop<()>, window: Arc<Window>) {
  });
  render_pass.set_pipeline(&wgpu_context_ref.pipeline);
  // (9)
- render_pass.set_bind_group(0, &wgpu_context_ref.bind_group, &[]);
+ render_pass.set_bind_group(
+ 0,
+ Some(&wgpu_context_ref.bind_group),
+ &[],
+ );
  render_pass.draw(0..3, 0..1);
  }
  wgpu_context_ref.queue.submit(Some(encoder.finish()));

diff --git a/examples/src/water/mod.rs b/examples/src/water/mod.rs
@@ -630,7 +630,7 @@ impl crate::framework::Example for Example {
  multiview: None,
  });
  encoder.set_pipeline(&terrain_pipeline);
- encoder.set_bind_group(0, &terrain_flipped_bind_group, &[]);
+ encoder.set_bind_group(0, Some(&terrain_flipped_bind_group), &[]);
  encoder.set_vertex_buffer(0, terrain_vertex_buf.slice(..));
  encoder.draw(0..terrain_vertices.len() as u32, 0..1);
  encoder.finish(&wgpu::RenderBundleDescriptor::default())
@@ -784,7 +784,7 @@ impl crate::framework::Example for Example {
  occlusion_query_set: None,
  });
  rpass.set_pipeline(&self.terrain_pipeline);
- rpass.set_bind_group(0, &self.terrain_normal_bind_group, &[]);
+ rpass.set_bind_group(0, Some(&self.terrain_normal_bind_group), &[]);
  rpass.set_vertex_buffer(0, self.terrain_vertex_buf.slice(..));
  rpass.draw(0..self.terrain_vertex_count as u32, 0..1);
  }
@@ -811,7 +811,7 @@ impl crate::framework::Example for Example {
  });
 
  rpass.set_pipeline(&self.water_pipeline);
- rpass.set_bind_group(0, &self.water_bind_group, &[]);
+ rpass.set_bind_group(0, Some(&self.water_bind_group), &[]);
  rpass.set_vertex_buffer(0, self.water_vertex_buf.slice(..));
  rpass.draw(0..self.water_vertex_count as u32, 0..1);
  }

diff --git a/player/tests/data/bind-group.ron b/player/tests/data/bind-group.ron
@@ -70,7 +70,7 @@
  SetBindGroup(
  index: 0,
  num_dynamic_offsets: 0,
- bind_group_id: Id(0, 1, Empty),
+ bind_group_id: Some(Id(0, 1, Empty)),
  ),
  SetPipeline(Id(0, 1, Empty)),
  ],

diff --git a/player/tests/data/zero-init-buffer.ron b/player/tests/data/zero-init-buffer.ron
@@ -149,7 +149,7 @@
  SetBindGroup(
  index: 0,
  num_dynamic_offsets: 0,
- bind_group_id: Id(0, 1, Empty),
+ bind_group_id: Some(Id(0, 1, Empty)),
  ),
  Dispatch((4, 1, 1)),
  ],

diff --git a/player/tests/data/zero-init-texture-binding.ron b/player/tests/data/zero-init-texture-binding.ron
@@ -150,7 +150,7 @@
  SetBindGroup(
  index: 0,
  num_dynamic_offsets: 0,
- bind_group_id: Id(0, 1, Empty),
+ bind_group_id: Some(Id(0, 1, Empty)),
  ),
  Dispatch((4, 1, 1)),
  ],

diff --git a/tests/src/image.rs b/tests/src/image.rs
@@ -377,7 +377,7 @@ fn copy_via_compute(
  let mut pass = encoder.begin_compute_pass(&ComputePassDescriptor::default());
 
  pass.set_pipeline(&pipeline_copy);
- pass.set_bind_group(0, &bg, &[]);
+ pass.set_bind_group(0, Some(&bg), &[]);
  pass.dispatch_workgroups(1, 1, 1);
  }
 

diff --git a/tests/tests/bgra8unorm_storage.rs b/tests/tests/bgra8unorm_storage.rs
@@ -110,7 +110,7 @@ static BGRA8_UNORM_STORAGE: GpuTestConfiguration = GpuTestConfiguration::new()
  timestamp_writes: None,
  });
 
- pass.set_bind_group(0, &bg, &[]);
+ pass.set_bind_group(0, Some(&bg), &[]);
  pass.set_pipeline(&pipeline);
  pass.dispatch_workgroups(256, 256, 1);
  }