//! The detail stage — running `dr-pipeline`'s neighbourhood passes.
//!
//! Where [`crate::AdjustPass`] fuses every point operation into one dispatch,
//! this runs the operations that cannot be fused because they read pixels they
//! are not writing: sharpening, noise reduction, clarity, texture, dehaze,
//! spot removal (FR-DEV-3, FR-DEV-8). `dr_pipeline::detail` decides *what* they
//! are and generates their WGSL; this compiles it, finds it somewhere to
//! write, and dispatches it.
//!
//! # Nothing round-trips
//!
//! Every intermediate here is a `wgpu::Texture` and none of them is ever
//! mapped. The chain is `demosaiced -> fused -> f16 -> f16 -> ... -> rgba8`,
//! all of it on the device, and the last write lands in the same texture the
//! compositor was already being handed. ARCH §6.1 and FR-DEV-4 are satisfied
//! by there being no code here that could violate them, which is the only
//! guarantee worth having.
//!
//! # Following the mask pass rather than inventing a second pattern
//!
//! `mask.rs` established how multi-target work is done in this crate, and this
//! copies it deliberately:
//!
//! - **One encoder for the whole chain.** The mask pass rasterises every layer
//!   into one command buffer and submits once; this does the same for every
//!   pass. Submission order is the only synchronisation either needs, because
//!   both write and then read through the same queue.
//! - **Textures reallocated on size change, never per frame.** `ensure_array`
//!   there, [`Intermediates::ensure`] here. Steady-state rendering at one
//!   viewport size allocates nothing.
//! - **An allocation counter that exists to be asserted on.** Reallocating per
//!   frame instead of per resize costs a great deal of bandwidth and shows up
//!   nowhere in the output, which is exactly the kind of regression that needs
//!   a test that can see it.
//! - **Pipelines cached by structure hash**, as `AdjustPass` caches its own.
//!   Moving a slider re-uploads a uniform buffer; it does not recompile.
//!
//! # The ping-pong, and why there are at most three textures
//!
//! Slot 0 holds what the fused colour pass wrote. It is kept **across frames**,
//! which is what makes [`dr_pipeline::Affects::Detail`] mean something: when
//! only a detail parameter has moved, the colour key is unchanged, the fused
//! dispatch is skipped, and dragging a sharpening slider costs the detail
//! passes alone (FR-DEV-3d).
//!
//! The remaining passes alternate between slots 1 and 2, and the last one
//! writes the display texture directly rather than an intermediate — so a
//! chain of *N* passes costs *N* dispatches and not *N* + 1, and there is no
//! resolve pass to pay for. That leaves the allocation at `1 + min(N-1, 2)`
//! textures: one for a single-pass operation, two for a separable blur, three
//! however long the chain gets after that.
//!
//! # The reduced chain, and why a second one was needed
//!
//! A pass may declare [`dr_pipeline::detail::DetailPass::output_scale`] and
//! write a target a fraction of the render size — clarity's base does, which
//! is what TD-4 bought back. Such a pass cannot be part of the ping-pong
//! above, and the reason is the shape of an unsharp mask rather than anything
//! about textures: the pass that *combines* needs the blur **and** the
//! full-resolution colour, and a colour that has been through a quarter-scale
//! target is no longer full resolution. If the scaled passes wrote into the
//! main chain they would destroy the very thing the last pass is going to
//! subtract from.
//!
//! So there are two chains. The full-resolution one carries the colour and is
//! untouched by a scaled pass; the reduced one carries the base. A scaled pass
//! reads the reduced chain if anything has been written to it and the
//! full-resolution chain otherwise — which is exactly "read what the pass
//! before you wrote", the same rule as before. A full-resolution pass always
//! reads the full-resolution chain, and sees the reduced one through binding 4
//! as `reduced_at()`.
//!
//! One reduced buffer, not one per operation. Two operations both wanting a
//! reduced base in the same frame would need more, and nothing does: clarity
//! is the only caller and texture's band is a decade finer, so it must stay at
//! full resolution. A `debug_assert` in [`DetailRunner::encode`] holds that
//! claim rather than leaving it as a comment.

use std::collections::HashMap;

use dr_pipeline::detail::{ComposedDetail, ComposedDetailPass};
use wgpu::util::DeviceExt as _;

use crate::{GpuContext, GpuError};

/// The format every intermediate carries.
///
/// The same `Rgba16Float` the demosaicer produces and the same one ARCH §5.2
/// names as the working precision (FR-DEV-2). It is not a free choice: the
/// stage exists between the colour pass and the output transform precisely so
/// that a kernel runs on linear values at full internal precision, and an
/// 8-bit intermediate would quantise twice and convolve display-encoded
/// numbers — which is how sharpening comes to band a clear sky.
pub const INTERMEDIATE_FORMAT: wgpu::TextureFormat = wgpu::TextureFormat::Rgba16Float;

/// One linear working texture.
struct Slot {
    #[allow(dead_code)]
    texture: wgpu::Texture,
    view: wgpu::TextureView,
}

/// The pool of linear intermediates, sized to the chain and the viewport.
struct Intermediates {
    slots: Vec<Slot>,
    width: u32,
    height: u32,
    allocations: usize,
}

impl Intermediates {
    fn new() -> Self {
        Self {
            slots: Vec::new(),
            width: 0,
            height: 0,
            allocations: 0,
        }
    }

    /// Make sure `count` textures of this size exist.
    ///
    /// Grows but never shrinks within a size: an edit that briefly had a
    /// three-pass chain and then a one-pass one keeps the spare texture rather
    /// than freeing and reallocating it the next time the user turns the
    /// operation back on. A size change drops the lot, because none of them
    /// fits any more.
    fn ensure(&mut self, ctx: &GpuContext, count: usize, width: u32, height: u32) {
        if self.width != width || self.height != height {
            self.slots.clear();
            self.width = width;
            self.height = height;
        }
        while self.slots.len() < count {
            let texture = ctx.device.create_texture(&wgpu::TextureDescriptor {
                label: Some("detail-intermediate"),
                size: wgpu::Extent3d {
                    width,
                    height,
                    depth_or_array_layers: 1,
                },
                mip_level_count: 1,
                sample_count: 1,
                dimension: wgpu::TextureDimension::D2,
                format: INTERMEDIATE_FORMAT,
                // STORAGE_BINDING to be written by a compute pass and
                // TEXTURE_BINDING to be read by the next one. Nothing else:
                // no RENDER_ATTACHMENT, because unlike the adjust pass's
                // output these are never handed to a compositor, and no
                // COPY_SRC, because nothing reads them back — that is the
                // point (ARCH §6.1).
                usage: wgpu::TextureUsages::STORAGE_BINDING | wgpu::TextureUsages::TEXTURE_BINDING,
                view_formats: &[],
            });
            let view = texture.create_view(&Default::default());
            self.slots.push(Slot { texture, view });
            self.allocations += 1;
        }
    }

    /// TRACES: FR-PLAT-AND-5
    /// Drop the pool, leaving it as [`Intermediates::new`] left it.
    ///
    /// The size is reset along with the slots, not merely because it is tidy:
    /// [`Self::ensure`] only refills when the count is short *or* the size
    /// differs, so a pool cleared while still claiming its old dimensions is
    /// indistinguishable from one that never held anything — which is fine
    /// here, and would stop being fine the moment `ensure` grew a fast path
    /// that trusted the stored size. `allocations` deliberately keeps
    /// counting: it exists so a test can see textures being made, and a
    /// counter reset on eviction would hide a reallocation storm rather than
    /// report one.
    fn release(&mut self) {
        self.slots.clear();
        self.width = 0;
        self.height = 0;
    }
}

/// Runs the detail stage.
///
/// Owned by [`crate::AdjustPass`] rather than standing alone, because the two
/// halves are one render: the fused pass writes slot 0, this reads it, and the
/// last pass writes the adjust pass's own output texture. Splitting them into
/// two objects with two lifetimes would mean a caller could hold a stale
/// intermediate against a fresh colour result and never be told.
pub(crate) struct DetailRunner {
    ctx: GpuContext,
    /// Layout for a pass writing another linear intermediate.
    to_linear: Layout,
    /// Layout for the last pass, which writes the display texture.
    to_output: Layout,
    /// Compiled pipelines by pass structure hash.
    cache: HashMap<u64, wgpu::ComputePipeline>,
    pool: Intermediates,
    /// The reduced chain — see the module documentation. Its own pool rather
    /// than more slots in `pool`, because its textures are a different size
    /// and [`Intermediates::ensure`] drops the lot when the size changes.
    reduced: Intermediates,
    /// See [`placeholder_instances`].
    no_instances: wgpu::Buffer,
    /// See [`placeholder_reduced`].
    no_reduced: wgpu::TextureView,
}

struct Layout {
    bind_group: wgpu::BindGroupLayout,
    pipeline: wgpu::PipelineLayout,
}

/// What binding 3 holds for a pass that declared no instance list.
///
/// One zeroed element, allocated once. Zero-length storage buffers cannot be
/// bound, and the passes that read this binding are exactly the ones that
/// uploaded something of their own, so nothing ever reads the placeholder's
/// contents — it exists to keep one bind group layout serving both kinds of
/// pass.
fn placeholder_instances(ctx: &GpuContext) -> wgpu::Buffer {
    ctx.device
        .create_buffer_init(&wgpu::util::BufferInitDescriptor {
            label: Some("detail-instances-placeholder"),
            contents: bytemuck::cast_slice(&[[0.0f32; 4]]),
            usage: wgpu::BufferUsages::STORAGE,
        })
}

/// What binding 4 holds for a pass that never calls `reduced_at`.
///
/// The same trick as [`placeholder_instances`], for the same reason: one bind
/// group layout has to serve a pass that reads the reduced chain and a pass
/// that has never heard of it, and a binding cannot be left unbound. 1x1 and
/// allocated once, so the cost of the arrangement is four bytes for the life
/// of the runner.
fn placeholder_reduced(ctx: &GpuContext) -> wgpu::TextureView {
    ctx.device
        .create_texture(&wgpu::TextureDescriptor {
            label: Some("detail-reduced-placeholder"),
            size: wgpu::Extent3d {
                width: 1,
                height: 1,
                depth_or_array_layers: 1,
            },
            mip_level_count: 1,
            sample_count: 1,
            dimension: wgpu::TextureDimension::D2,
            format: INTERMEDIATE_FORMAT,
            usage: wgpu::TextureUsages::TEXTURE_BINDING,
            view_formats: &[],
        })
        .create_view(&Default::default())
}

impl DetailRunner {
    pub(crate) fn new(ctx: &GpuContext) -> Self {
        Self {
            ctx: ctx.clone(),
            to_linear: Layout::new(ctx, INTERMEDIATE_FORMAT, "detail-linear"),
            to_output: Layout::new(ctx, crate::AdjustPass::FORMAT, "detail-output"),
            cache: HashMap::new(),
            pool: Intermediates::new(),
            reduced: Intermediates::new(),
            no_instances: placeholder_instances(ctx),
            no_reduced: placeholder_reduced(ctx),
        }
    }

    /// The view the fused colour pass should write, given a chain of `passes`.
    ///
    /// Slot 0, always — it is the one that survives between frames so that a
    /// detail-only change can skip the colour dispatch entirely.
    pub(crate) fn colour_target(
        &mut self,
        passes: usize,
        width: u32,
        height: u32,
    ) -> &wgpu::TextureView {
        // One for the colour pass's result, then one per hand-off between
        // detail passes, capped at two because a ping-pong needs no more: the
        // last pass writes the display texture rather than an intermediate.
        let needed = 1 + passes.saturating_sub(1).min(2);
        self.pool.ensure(&self.ctx, needed, width, height);
        &self.pool.slots[0].view
    }

    /// Encode every pass of `chain`, the last one writing `output`.
    ///
    /// The caller must already have run the fused colour pass into
    /// [`Self::colour_target`] — or established that a previous frame's is
    /// still valid, which is the whole point of keeping slot 0.
    pub(crate) fn encode(
        &mut self,
        encoder: &mut wgpu::CommandEncoder,
        chain: &ComposedDetail,
        output: &wgpu::TextureView,
        width: u32,
        height: u32,
    ) -> Result<usize, GpuError> {
        for pass in &chain.passes {
            self.compile(pass)?;
        }

        // The reduced chain's size, allocated once for the whole chain.
        //
        // One scale per chain, so the first scaled pass names the only scale
        // there is — see the module documentation for why one buffer is
        // enough, and the assertion for what would have to change.
        if let Some(scale) = chain
            .passes
            .iter()
            .map(|p| p.output_scale)
            .find(|&scale| scale > 1)
        {
            debug_assert!(
                chain
                    .passes
                    .iter()
                    .all(|p| p.output_scale == 1 || p.output_scale == scale),
                "two reduced scales in one chain, and the runner holds one \
                 reduced buffer"
            );
            // Two, for the ping-pong the reduce and the two blur halves need.
            // A separable blur cannot read the texture it is writing.
            self.reduced.ensure(
                &self.ctx,
                2,
                width.div_ceil(scale).max(1),
                height.div_ceil(scale).max(1),
            );
        }

        // Where each chain last wrote. `full` starts at slot 0 — what the
        // fused colour pass left there — and `carried` starts empty, which is
        // what makes the first scaled pass read the colour rather than an
        // uninitialised base.
        let mut full = 0usize;
        let mut full_writes = 0usize;
        let mut carried: Option<usize> = None;
        let mut reduced_writes = 0usize;

        for pass in chain.passes.iter() {
            let scaled = pass.output_scale > 1;

            // The last pass carries the output transform into the display
            // texture, which is the render size by definition. A scaled pass
            // there would bind a shader dispatching over a quarter-size grid
            // to a full-size target and write a quarter of the picture — a
            // wrong image rather than a validation failure, so it is caught
            // here and named.
            if scaled && pass.writes_output {
                return Err(GpuError::ShaderCompilation(format!(
                    "detail pass {} declares output_scale {} and is last in \
                     the chain; the output transform is written at the render \
                     size",
                    pass.label, pass.output_scale
                )));
            }

            let (dispatch_w, dispatch_h) = if scaled {
                (
                    width.div_ceil(pass.output_scale).max(1),
                    height.div_ceil(pass.output_scale).max(1),
                )
            } else {
                (width, height)
            };

            // Read what the previous pass in *this pass's own chain* wrote;
            // write the next slot of it, or the display texture if this is the
            // last pass. Alternating slots is what stops a pass reading the
            // texture it is writing — on a compute pass that is not an error
            // the driver reports, merely a picture that depends on scheduling.
            let source = match (scaled, carried) {
                (true, Some(slot)) => &self.reduced.slots[slot].view,
                _ => &self.pool.slots[full].view,
            };
            let destination = if pass.writes_output {
                output
            } else if scaled {
                &self.reduced.slots[reduced_writes % 2].view
            } else {
                &self.pool.slots[1 + (full_writes % 2)].view
            };
            // Binding 4. Present for every pass, because one bind group layout
            // serves both kinds; a pass that never calls `reduced_at` gets the
            // 1x1 placeholder and never reads it.
            let reduced_source = match carried {
                Some(slot) => &self.reduced.slots[slot].view,
                None => &self.no_reduced,
            };
            let layout = if pass.writes_output {
                &self.to_output
            } else {
                &self.to_linear
            };

            let params = self
                .ctx
                .device
                .create_buffer_init(&wgpu::util::BufferInitDescriptor {
                    label: Some("detail-params"),
                    contents: bytemuck::cast_slice(&pass.uniforms),
                    usage: wgpu::BufferUsages::UNIFORM,
                });

            // TRACES: FR-DEV-8
            // The instance list, uploaded only by the passes that have one. A
            // kernel pass — which is every pass that is a convolution — is
            // handed the placeholder allocated once in `new`, because a storage
            // buffer of length zero is not bindable and allocating a fresh
            // sixteen bytes per pass per frame is the per-frame allocation this
            // module's documentation exists to refuse.
            let instances = (!pass.storage.is_empty()).then(|| {
                self.ctx
                    .device
                    .create_buffer_init(&wgpu::util::BufferInitDescriptor {
                        label: Some("detail-instances"),
                        contents: bytemuck::cast_slice(pass.storage.as_slice()),
                        usage: wgpu::BufferUsages::STORAGE,
                    })
            });
            let instances = instances.as_ref().unwrap_or(&self.no_instances);

            let bind_group = self
                .ctx
                .device
                .create_bind_group(&wgpu::BindGroupDescriptor {
                    label: Some("detail-bg"),
                    layout: &layout.bind_group,
                    entries: &[
                        wgpu::BindGroupEntry {
                            binding: 0,
                            resource: wgpu::BindingResource::TextureView(source),
                        },
                        wgpu::BindGroupEntry {
                            binding: 1,
                            resource: params.as_entire_binding(),
                        },
                        wgpu::BindGroupEntry {
                            binding: 2,
                            resource: wgpu::BindingResource::TextureView(destination),
                        },
                        wgpu::BindGroupEntry {
                            binding: 3,
                            resource: instances.as_entire_binding(),
                        },
                        wgpu::BindGroupEntry {
                            binding: 4,
                            resource: wgpu::BindingResource::TextureView(reduced_source),
                        },
                    ],
                });

            let pipeline = self
                .cache
                .get(&pass.structure_hash)
                .expect("compiled above");

            let mut compute = encoder.begin_compute_pass(&wgpu::ComputePassDescriptor {
                label: Some(pass.label.as_str()),
                timestamp_writes: None,
            });
            compute.set_pipeline(pipeline);
            compute.set_bind_group(0, &bind_group, &[]);
            compute.dispatch_workgroups(dispatch_w.div_ceil(8), dispatch_h.div_ceil(8), 1);
            drop(compute);

            if pass.writes_output {
                // Nothing downstream to hand anything to.
            } else if scaled {
                carried = Some(reduced_writes % 2);
                reduced_writes += 1;
            } else {
                full = 1 + (full_writes % 2);
                full_writes += 1;
                // A full-resolution pass consumes the reduced chain. It is the
                // combine — the base has been subtracted and now lives in the
                // colour — so a later operation must not be handed a base
                // belonging to this one.
                carried = None;
            }
        }

        Ok(chain.passes.len())
    }

    /// Compile one pass, or leave the cached pipeline in place.
    ///
    /// A validation error here is a codegen bug rather than anything the user
    /// did, so it is caught in an error scope and returned with the generated
    /// source and the pass's label attached — a line number against code
    /// nobody wrote, from one of several passes, is otherwise close to
    /// unactionable.
    fn compile(&mut self, pass: &ComposedDetailPass) -> Result<(), GpuError> {
        if self.cache.contains_key(&pass.structure_hash) {
            return Ok(());
        }

        let scope = self
            .ctx
            .device
            .push_error_scope(wgpu::ErrorFilter::Validation);

        let module = self
            .ctx
            .device
            .create_shader_module(wgpu::ShaderModuleDescriptor {
                label: Some(pass.label.as_str()),
                source: wgpu::ShaderSource::Wgsl(pass.source.as_str().into()),
            });

        let layout = if pass.writes_output {
            &self.to_output
        } else {
            &self.to_linear
        };

        let pipeline = self
            .ctx
            .device
            .create_compute_pipeline(&wgpu::ComputePipelineDescriptor {
                label: Some(pass.label.as_str()),
                layout: Some(&layout.pipeline),
                module: &module,
                entry_point: Some("main"),
                compilation_options: Default::default(),
                cache: None,
            });

        if let Some(err) = pollster::block_on(scope.pop()) {
            return Err(GpuError::ShaderCompilation(format!(
                "detail pass {}: {err}\n\n--- generated source ---\n{}",
                pass.label,
                crate::adjust::numbered(&pass.source)
            )));
        }

        self.cache.insert(pass.structure_hash, pipeline);
        Ok(())
    }

    /// How many distinct detail pipelines are compiled. For tests asserting
    /// that slider movement does not recompile.
    pub(crate) fn cached_pipelines(&self) -> usize {
        self.cache.len()
    }

    /// TRACES: FR-PLAT-AND-5
    /// Give back everything this stage is only holding to be fast.
    ///
    /// Both pools and the pipeline cache. Nothing here is state: a pool slot
    /// is re-created by the next [`Intermediates::ensure`] and a pipeline by
    /// the next compile-on-miss, so the only cost of this call is the work of
    /// doing both again.
    pub(crate) fn release_caches(&mut self) {
        self.cache.clear();
        self.pool.release();
        self.reduced.release();
    }

    /// How many intermediate textures have been allocated since this pass was
    /// created. For tests — see [`crate::MaskPass::allocations`] for the
    /// regression this shape of counter exists to catch.
    pub(crate) fn allocations(&self) -> usize {
        // Both pools. A reduced buffer reallocated every frame is exactly the
        // regression this counter exists to catch, and counting only the
        // full-resolution one would hide it.
        self.pool.allocations + self.reduced.allocations
    }
}

/// A read-only storage buffer entry, as `mask.rs` declares its strokes.
fn storage_entry(binding: u32) -> wgpu::BindGroupLayoutEntry {
    wgpu::BindGroupLayoutEntry {
        binding,
        visibility: wgpu::ShaderStages::COMPUTE,
        ty: wgpu::BindingType::Buffer {
            ty: wgpu::BufferBindingType::Storage { read_only: true },
            has_dynamic_offset: false,
            min_binding_size: None,
        },
        count: None,
    }
}

impl Layout {
    fn new(ctx: &GpuContext, format: wgpu::TextureFormat, label: &str) -> Self {
        let bind_group = ctx
            .device
            .create_bind_group_layout(&wgpu::BindGroupLayoutDescriptor {
                label: Some(label),
                entries: &[
                    // The previous stage's result.
                    wgpu::BindGroupLayoutEntry {
                        binding: 0,
                        visibility: wgpu::ShaderStages::COMPUTE,
                        ty: wgpu::BindingType::Texture {
                            sample_type: wgpu::TextureSampleType::Float { filterable: true },
                            view_dimension: wgpu::TextureViewDimension::D2,
                            multisampled: false,
                        },
                        count: None,
                    },
                    wgpu::BindGroupLayoutEntry {
                        binding: 1,
                        visibility: wgpu::ShaderStages::COMPUTE,
                        ty: wgpu::BindingType::Buffer {
                            ty: wgpu::BufferBindingType::Uniform,
                            has_dynamic_offset: false,
                            min_binding_size: None,
                        },
                        count: None,
                    },
                    wgpu::BindGroupLayoutEntry {
                        binding: 2,
                        visibility: wgpu::ShaderStages::COMPUTE,
                        ty: wgpu::BindingType::StorageTexture {
                            access: wgpu::StorageTextureAccess::WriteOnly,
                            format,
                            view_dimension: wgpu::TextureViewDimension::D2,
                        },
                        count: None,
                    },
                    // TRACES: FR-DEV-8
                    // The instance list, for a pass whose work is a list rather
                    // than a kernel (`DetailPass::storage`). Every other pass
                    // gets `Intermediates`' placeholder here — one entry on both
                    // layouts rather than two more layouts, since a convolution
                    // that never reads the buffer costs nothing for it being
                    // bound.
                    storage_entry(3),
                    // The reduced chain, for a pass that calls `reduced_at`.
                    // Bound on every layout for the same reason binding 3 is:
                    // a pass that never reads it costs nothing for it being
                    // there, and two more layouts would cost a great deal more
                    // than that.
                    wgpu::BindGroupLayoutEntry {
                        binding: 4,
                        visibility: wgpu::ShaderStages::COMPUTE,
                        ty: wgpu::BindingType::Texture {
                            sample_type: wgpu::TextureSampleType::Float { filterable: true },
                            view_dimension: wgpu::TextureViewDimension::D2,
                            multisampled: false,
                        },
                        count: None,
                    },
                ],
            });

        let pipeline = ctx
            .device
            .create_pipeline_layout(&wgpu::PipelineLayoutDescriptor {
                label: Some(label),
                bind_group_layouts: &[Some(&bind_group)],
                immediate_size: 0,
            });

        Self {
            bind_group,
            pipeline,
        }
    }
}