1 files changed, 93 insertions, 0 deletions
diff --git a/intern/cycles/kernel/film/cryptomatte_passes.h b/intern/cycles/kernel/film/cryptomatte_passes.h
new file mode 100644
index 00000000000..4765777e7e2
--- /dev/null
+++ b/intern/cycles/kernel/film/cryptomatte_passes.h
@@ -0,0 +1,93 @@
+/* SPDX-License-Identifier: Apache-2.0
+ * Copyright 2018-2022 Blender Foundation */
+
+#pragma once
+
+CCL_NAMESPACE_BEGIN
+
+/* Element of ID pass stored in the render buffers.
+ * It is `float2` semantically, but it must be unaligned since the offset of ID passes in the
+ * render buffers might not meet expected by compiler alignment. */
+typedef struct CryptoPassBufferElement {
+  float x;
+  float y;
+} CryptoPassBufferElement;
+
+ccl_device_inline void film_write_cryptomatte_slots(ccl_global float *buffer,
+                                                    int num_slots,
+                                                    float id,
+                                                    float weight)
+{
+  kernel_assert(id != ID_NONE);
+  if (weight == 0.0f) {
+    return;
+  }
+
+  for (int slot = 0; slot < num_slots; slot++) {
+    ccl_global CryptoPassBufferElement *id_buffer = (ccl_global CryptoPassBufferElement *)buffer;
+#ifdef __ATOMIC_PASS_WRITE__
+    /* If the loop reaches an empty slot, the ID isn't in any slot yet - so add it! */
+    if (id_buffer[slot].x == ID_NONE) {
+      /* Use an atomic to claim this slot.
+       * If a different thread got here first, try again from this slot on. */
+      float old_id = atomic_compare_and_swap_float(buffer + slot * 2, ID_NONE, id);
+      if (old_id != ID_NONE && old_id != id) {
+        continue;
+      }
+      atomic_add_and_fetch_float(buffer + slot * 2 + 1, weight);
+      break;
+    }
+    /* If there already is a slot for that ID, add the weight.
+     * If no slot was found, add it to the last. */
+    else if (id_buffer[slot].x == id || slot == num_slots - 1) {
+      atomic_add_and_fetch_float(buffer + slot * 2 + 1, weight);
+      break;
+    }
+#else  /* __ATOMIC_PASS_WRITE__ */
+    /* If the loop reaches an empty slot, the ID isn't in any slot yet - so add it! */
+    if (id_buffer[slot].x == ID_NONE) {
+      id_buffer[slot].x = id;
+      id_buffer[slot].y = weight;
+      break;
+    }
+    /* If there already is a slot for that ID, add the weight.
+     * If no slot was found, add it to the last. */
+    else if (id_buffer[slot].x == id || slot == num_slots - 1) {
+      id_buffer[slot].y += weight;
+      break;
+    }
+#endif /* __ATOMIC_PASS_WRITE__ */
+  }
+}
+
+ccl_device_inline void film_sort_cryptomatte_slots(ccl_global float *buffer, int num_slots)
+{
+  ccl_global CryptoPassBufferElement *id_buffer = (ccl_global CryptoPassBufferElement *)buffer;
+  for (int slot = 1; slot < num_slots; ++slot) {
+    if (id_buffer[slot].x == ID_NONE) {
+      return;
+    }
+    /* Since we're dealing with a tiny number of elements, insertion sort should be fine. */
+    int i = slot;
+    while (i > 0 && id_buffer[i].y > id_buffer[i - 1].y) {
+      const CryptoPassBufferElement swap = id_buffer[i];
+      id_buffer[i] = id_buffer[i - 1];
+      id_buffer[i - 1] = swap;
+      --i;
+    }
+  }
+}
+
+/* post-sorting for Cryptomatte */
+ccl_device_inline void film_cryptomatte_post(KernelGlobals kg,
+                                             ccl_global float *render_buffer,
+                                             int pixel_index)
+{
+  const int pass_stride = kernel_data.film.pass_stride;
+  const uint64_t render_buffer_offset = (uint64_t)pixel_index * pass_stride;
+  ccl_global float *cryptomatte_buffer = render_buffer + render_buffer_offset +
+                                         kernel_data.film.pass_cryptomatte;
+  film_sort_cryptomatte_slots(cryptomatte_buffer, 2 * kernel_data.film.cryptomatte_depth);
+}
+
+CCL_NAMESPACE_END