// Shared-memory audio ring for the injection render-hook audio path.
//
// The injected hook (coop_hook.dll) captures the game's WASAPI render frames and
// is the sole *producer*; the host (coop_host.exe) is the sole *consumer* and
// re-renders the frames for Steam Remote Play Together. This is a separate,
// larger mapping from the input/status SharedBlock (which is only 20-byte pads
// and can't hold PCM): a header followed by a byte ring of `capacity` bytes.
//
// Lock-free SPSC with free-running 64-bit positions (release on publish, acquire
// on read) — the same cross-process atomic model as the input seqlock. POD and
// version-locked: both modules compile this identical header.
#pragma once

#include <algorithm>
#include <atomic>
#include <cstdint>
#include <cstring>
#include <string>

namespace coop
{

// 'AURG' little-endian; sanity-checks the mapping before either side trusts it.
inline constexpr std::uint32_t kAudioRingMagic = 0x47525541u;

// Bump whenever AudioRingHeader's layout changes.
inline constexpr std::uint32_t kAudioRingVersion = 2;

// Operator commands the host issues per stream (host -> hook), applied via the op_*
// fields in the header. The host writes the fields then bumps op_seq; the hook applies
// the command once per new op_seq. Lets the Audio panel re-measure a stream's rate or
// override its format when detection is wrong/unrecoverable.
enum AudioRingOp : std::uint32_t
{
	AudioRingOp_None = 0,
	AudioRingOp_Remeasure = 1, // re-run the sample-rate measurement for this stream
	AudioRingOp_Override = 2,  // adopt the op_rate/channels/bits/format_tag verbatim
};

// Per-pid mapping name, mirroring kSharedMemoryPrefix: coop_audio_<pid>.
inline constexpr wchar_t kAudioRingPrefix[] = L"Local\\coop_audio_";

// Byte capacity of the PCM ring. 1 MiB is >1 s even at 48 kHz / 2 ch / 32-bit
// float (384 kB/s); the host should always keep up, so this is pure slack.
inline constexpr std::uint32_t kAudioRingCapacity = 1u << 20;

// Header preceding the PCM data. All multi-process-shared counters are atomic;
// the format fields are written once by the producer *before* it publishes
// format_valid (release), and read by the consumer *after* it observes
// format_valid (acquire), so they need no atomicity of their own.
struct AudioRingHeader
{
	std::uint32_t magic;
	std::uint32_t version;

	// Host-owned gate. The hook copies+silences frames only while this is 1;
	// when 0 the game audio passes through locally and nothing is mirrored
	// (stream counting in HookStatus still runs regardless of this flag).
	std::atomic<std::uint32_t> capture_enabled;

	// Producer publishes the captured stream's format once, then sets
	// format_valid=1 (release). format_generation is reserved so a future
	// mid-session device re-init can be made forward-compatible; v1 sets once.
	std::atomic<std::uint32_t> format_valid;
	std::atomic<std::uint32_t> format_generation;

	std::uint32_t sample_rate;
	std::uint32_t channels;
	std::uint32_t bits;
	std::uint32_t format_tag; // WAVE_FORMAT_* (PCM=1, IEEE_FLOAT=3, EXTENSIBLE=0xFFFE)
	std::uint32_t block_align; // bytes per frame (all channels)

	std::uint32_t capacity; // bytes in the trailing data region

	std::atomic<std::uint64_t> write_pos;       // producer cursor, free-running
	std::atomic<std::uint64_t> read_pos;        // consumer cursor, free-running
	std::atomic<std::uint64_t> frames_produced; // cumulative frames pushed
	std::atomic<std::uint64_t> overruns;        // packets dropped on a full ring

	// --- Operator control (host -> hook) ---------------------------------------
	// The host writes op_kind + the op_* fields, then bumps op_seq (release); the hook
	// applies the command once per new op_seq (acquire). See AudioRingOp.
	std::atomic<std::uint32_t> op_seq; // bumped by the host on each new command
	std::uint32_t op_kind;             // AudioRingOp
	std::uint32_t op_rate;             // override: sample rate
	std::uint32_t op_channels;         // override: channel count
	std::uint32_t op_bits;             // override: bits per sample
	std::uint32_t op_format_tag;       // override: WAVE_FORMAT_PCM / _IEEE_FLOAT

	// Host -> hook: format-verification co-capture. While 1, the hook pushes a still-being-measured
	// (guessed) stream's raw pre-mix bytes into the ring WITHOUT silencing the game, so the host can
	// capture both the hook (pre-mix) and a parallel process-loopback (post-mix) of the same audio
	// and cross-correlate them to recover the true sample rate (and, in step b, channels/bit-depth)
	// from ground truth instead of guessing. Inert (0) by default -- normal capture is unaffected,
	// so it never changes the shipping no-echo path. Repurposed from `reserved`, so the layout and
	// size are unchanged (old builds saw it as a zero reserved byte).
	std::atomic<std::uint32_t> verify_capture;

	std::uint8_t reserved[36];

	// std::uint8_t data[capacity] follows immediately in the mapping.
};

static_assert(std::atomic<std::uint64_t>::is_always_lock_free,
              "audio ring needs a lock-free 64-bit atomic for cross-process use");

// Total mapping size for a ring of `capacity` bytes.
inline constexpr std::size_t audio_ring_total_size(std::uint32_t capacity)
{
	return sizeof(AudioRingHeader) + capacity;
}

// Pointer to the PCM data region following the header.
inline std::uint8_t* audio_ring_data(AudioRingHeader* h)
{
	return reinterpret_cast<std::uint8_t*>(h) + sizeof(AudioRingHeader);
}

// Host side: stamp a freshly created mapping into a valid empty ring.
inline void audio_ring_init(AudioRingHeader& h, std::uint32_t capacity)
{
	h.magic = kAudioRingMagic;
	h.version = kAudioRingVersion;
	h.capture_enabled.store(0, std::memory_order_relaxed);
	h.format_valid.store(0, std::memory_order_relaxed);
	h.format_generation.store(0, std::memory_order_relaxed);
	h.sample_rate = 0;
	h.channels = 0;
	h.bits = 0;
	h.format_tag = 0;
	h.block_align = 0;
	h.capacity = capacity;
	h.write_pos.store(0, std::memory_order_relaxed);
	h.read_pos.store(0, std::memory_order_relaxed);
	h.frames_produced.store(0, std::memory_order_relaxed);
	h.overruns.store(0, std::memory_order_relaxed);
	h.op_seq.store(0, std::memory_order_relaxed);
	h.op_kind = 0;
	h.op_rate = 0;
	h.op_channels = 0;
	h.op_bits = 0;
	h.op_format_tag = 0;
	h.verify_capture.store(0, std::memory_order_relaxed);
	std::memset(h.reserved, 0, sizeof(h.reserved));
}

// Validate a mapping the other side created/opened.
inline bool audio_ring_valid(const AudioRingHeader& h)
{
	return h.magic == kAudioRingMagic && h.version == kAudioRingVersion && h.capacity != 0;
}

// Producer (hook): publish the captured stream format, then mark it valid.
inline void audio_ring_set_format(AudioRingHeader& h, std::uint32_t sample_rate, std::uint32_t channels,
                                  std::uint32_t bits, std::uint32_t format_tag, std::uint32_t block_align)
{
	h.sample_rate = sample_rate;
	h.channels = channels;
	h.bits = bits;
	h.format_tag = format_tag;
	h.block_align = block_align;
	h.format_generation.fetch_add(1, std::memory_order_relaxed);
	h.format_valid.store(1, std::memory_order_release);
}

// Consumer (host): true once the producer has published a format.
inline bool audio_ring_format_ready(const AudioRingHeader& h)
{
	return h.format_valid.load(std::memory_order_acquire) != 0;
}

// Producer: push `bytes` of PCM. Returns false (and bumps overruns) if the ring
// can't hold the whole packet, in which case nothing is written — drop the
// packet rather than tear a frame. `frames` is recorded for the diagnostics.
inline bool audio_ring_push(AudioRingHeader& h, const void* src, std::uint32_t bytes, std::uint32_t frames)
{
	const std::uint64_t w = h.write_pos.load(std::memory_order_relaxed);
	const std::uint64_t r = h.read_pos.load(std::memory_order_acquire);
	const std::uint32_t used = static_cast<std::uint32_t>(w - r);
	if (bytes > h.capacity - used)
	{
		h.overruns.fetch_add(1, std::memory_order_relaxed);
		return false;
	}
	std::uint8_t* data = audio_ring_data(&h);
	const std::uint32_t off = static_cast<std::uint32_t>(w % h.capacity);
	const std::uint32_t first = std::min(bytes, h.capacity - off);
	std::memcpy(data + off, src, first);
	if (bytes > first)
	{
		std::memcpy(data, static_cast<const std::uint8_t*>(src) + first, bytes - first);
	}
	h.write_pos.store(w + bytes, std::memory_order_release);
	h.frames_produced.fetch_add(frames, std::memory_order_relaxed);
	return true;
}

// Consumer: bytes currently available to read.
inline std::uint32_t audio_ring_available(const AudioRingHeader& h)
{
	const std::uint64_t w = h.write_pos.load(std::memory_order_acquire);
	const std::uint64_t r = h.read_pos.load(std::memory_order_relaxed);
	return static_cast<std::uint32_t>(w - r);
}

// Producer: bytes currently free (so a multi-part packet can be checked to fit before any of it is
// written -- keeps a self-describing [header][payload] framing from tearing on a full ring).
inline std::uint32_t audio_ring_free_space(const AudioRingHeader& h)
{
	const std::uint64_t w = h.write_pos.load(std::memory_order_relaxed);
	const std::uint64_t r = h.read_pos.load(std::memory_order_acquire);
	return h.capacity - static_cast<std::uint32_t>(w - r);
}

// Consumer: copy up to `bytes` into `dst`; returns the number actually popped.
inline std::uint32_t audio_ring_pop(AudioRingHeader& h, void* dst, std::uint32_t bytes)
{
	const std::uint64_t r = h.read_pos.load(std::memory_order_relaxed);
	const std::uint64_t w = h.write_pos.load(std::memory_order_acquire);
	const std::uint32_t avail = static_cast<std::uint32_t>(w - r);
	bytes = std::min(bytes, avail);
	const std::uint8_t* data = audio_ring_data(&h);
	const std::uint32_t off = static_cast<std::uint32_t>(r % h.capacity);
	const std::uint32_t first = std::min(bytes, h.capacity - off);
	std::memcpy(dst, data + off, first);
	if (bytes > first)
	{
		std::memcpy(static_cast<std::uint8_t*>(dst) + first, data, bytes - first);
	}
	h.read_pos.store(r + bytes, std::memory_order_release);
	return bytes;
}

// Host: post an operator command to the hook for this stream. Writes the fields, then
// bumps op_seq (release) so the hook applies it exactly once. For a re-measure the
// rate/channels/bits are ignored.
inline void audio_ring_post_op(AudioRingHeader& h, std::uint32_t kind, std::uint32_t rate = 0,
                               std::uint32_t channels = 0, std::uint32_t bits = 0,
                               std::uint32_t format_tag = 0)
{
	h.op_kind = kind;
	h.op_rate = rate;
	h.op_channels = channels;
	h.op_bits = bits;
	h.op_format_tag = format_tag;
	h.op_seq.fetch_add(1, std::memory_order_release);
}

// One operator command read back by the hook.
struct AudioRingOpCmd
{
	std::uint32_t kind = AudioRingOp_None;
	std::uint32_t rate = 0;
	std::uint32_t channels = 0;
	std::uint32_t bits = 0;
	std::uint32_t format_tag = 0;
};

// Hook: if a new op was posted since `last_seq`, read it into `out`, advance `last_seq`,
// and return its kind; otherwise returns AudioRingOp_None. Robust to ring re-creation
// (op_seq resets to 0 -> a stale higher last_seq just reads the zeroed None command).
inline std::uint32_t audio_ring_poll_op(AudioRingHeader& h, std::uint32_t& last_seq, AudioRingOpCmd& out)
{
	const std::uint32_t seq = h.op_seq.load(std::memory_order_acquire);
	if (seq == last_seq)
	{
		return AudioRingOp_None;
	}
	last_seq = seq;
	out.kind = h.op_kind;
	out.rate = h.op_rate;
	out.channels = h.op_channels;
	out.bits = h.op_bits;
	out.format_tag = h.op_format_tag;
	return out.kind;
}

// Build the per-pid audio ring name both sides agree on. Stream 0 keeps the bare
// coop_audio_<pid> name (backward compatible / the single-stream case); additional
// streams append _<index> (coop_audio_<pid>_1, _2, ...). The host captures every
// render stream into its own ring and mixes them.
inline std::wstring audio_ring_name(unsigned long target_pid, unsigned index = 0)
{
	std::wstring name = std::wstring(kAudioRingPrefix) + std::to_wstring(target_pid);
	if (index != 0)
	{
		name += L"_" + std::to_wstring(index);
	}
	return name;
}

} // namespace coop