Claude review: drm/atomic: track individual colorop updates

[Claude Code Review Bot <claude-review@example.com>]

Patch Review

**Assessment: Good, one design observation**

This is the core patch that wires up `color_mgmt_changed` tracking. The approach is clean:

1. `drm_atomic_set_colorop_for_plane()` now returns `bool` (true if pipeline changed) and the caller uses `|=` to accumulate into `plane_state->color_mgmt_changed`.

2. `drm_atomic_colorop_set_property()` and `drm_atomic_color_set_data_property()` gain a `bool *replaced` output parameter to signal when a property value actually changed.

3. In `drm_atomic_set_property()`, after setting a colorop property:
```c
if (!ret && replaced)
    plane_state->color_mgmt_changed = true;
```

The change detection in `drm_atomic_colorop_set_property()` correctly checks each property individually:
```c
if (state->bypass != val) {
    state->bypass = val;
    *replaced = true;
}
```

**Design observation**: The `replaced` variable is initialized to `false` but only set to `true` — it's never checked before being returned. For `drm_atomic_color_set_data_property()`, the `replaced` pointer is forwarded directly to `drm_property_replace_blob_from_id()`, which handles blob comparison internally. This is correct because `drm_property_replace_blob_from_id()` already sets `*replaced` to indicate whether the blob pointer changed.

**Observation about `drm_atomic_set_colorop_for_plane()`**: The early return `if (plane_state->color_pipeline == colorop) return false` is a nice optimization — it avoids the debug prints for no-op pipeline selections. However, this changes behavior slightly: previously, the debug message would fire even for redundant sets. This is fine since it's debug-only.

The `#include <linux/types.h>` in `drm_atomic_uapi.h` for `bool` is correct and matches the v2 changelog note about fixing the MSM driver build.

---
Generated by Claude Code Patch Reviewer