//! ESP32-P4 peripherals, modelled at comptime.
//!
//! There is no HAL here and no generated 20k-line register header: a `Reg` is a typed pointer to
//! an MMIO word, and a peripheral is a struct of them. Everything is `inline`, so `gpio.setHigh(20)`
//! compiles to the single `sw` instruction it should be, and a wrong bit index is a compile error
//! rather than a silent write.
//!
//! Addresses are from ESP-IDF v6.0.2 `components/soc/esp32p4/register/hw_ver1/soc/` - the pre-v3
//! header set, which is the one that matches this silicon (rev v1.3) - except the GPIO matrix
//! signal index, which lives in `components/soc/esp32p4/include/soc/gpio_sig_map.h`.

const std = @import("std");

/// A 32-bit memory-mapped register.
pub fn Reg(comptime addr: usize) type {
    return struct {
        pub const address = addr;
        const ptr: *volatile u32 = @ptrFromInt(addr);

        pub inline fn read() u32 {
            return ptr.*;
        }
        pub inline fn write(value: u32) void {
            ptr.* = value;
        }
        pub inline fn set(mask: u32) void {
            ptr.* = ptr.* | mask;
        }
        pub inline fn clear(mask: u32) void {
            ptr.* = ptr.* & ~mask;
        }
        /// Read-modify-write a bitfield: `modify(.{ .shift = 12, .width = 3 }, 5)`.
        pub inline fn modify(comptime field: Field, value: u32) void {
            const mask: u32 = ((@as(u32, 1) << field.width) - 1) << field.shift;
            ptr.* = (ptr.* & ~mask) | ((value << field.shift) & mask);
        }
    };
}

pub const Field = struct { shift: u5, width: u5 };

/// A register with a named layout: pass a packed struct whose bit width is 32 and the accessors
/// become typed, so a pad is configured by naming fields instead of shifting bits. Read-modify-
/// write stays explicit - `var v = reg.read(); v.mcu_sel = 1; reg.write(v);` - because that is one
/// load and one store, and hiding it behind a partial-update type buys nothing here.
pub fn Typed(comptime T: type, comptime addr: usize) type {
    comptime std.debug.assert(@bitSizeOf(T) == 32);
    return struct {
        pub const address = addr;
        const ptr: *volatile T = @ptrFromInt(addr);

        pub inline fn read() T {
            return ptr.*;
        }
        pub inline fn write(value: T) void {
            ptr.* = value;
        }
        /// Apply `f` to the current value and write the result back.
        pub inline fn modify(comptime f: fn (T) T) void {
            ptr.* = f(ptr.*);
        }
    };
}

/// An array of identical registers. The index type is narrowed to the array's real range, so an
/// out-of-range access is a compile error in every optimize mode - an `assert` would have been
/// compiled out under ReleaseSmall, which is this project's default.
pub fn RegArray(comptime base: usize, comptime stride: usize, comptime count: usize) type {
    return struct {
        pub const Index = std.math.IntFittingRange(0, count - 1);

        /// comptime, because `IntFittingRange` rounds up to a whole width: for a 57-entry array the
        /// index type is u6, which would happily accept 57..63. Every caller here passes a comptime
        /// pin anyway, so this costs nothing and makes the bound real in all optimize modes.
        pub inline fn at(comptime index: Index) *volatile u32 {
            comptime std.debug.assert(index < count);
            return @ptrFromInt(base + @as(usize, index) * stride);
        }
    };
}

const hp_periph1 = 0x500C0000;

/// GPIO and the IO MUX now live in the HAL, which builds them out of ESP-IDF's own register macros
/// (`hal/gpio.zig`) instead of the hand-transcribed addresses that used to be here. The
/// transcription is exactly the kind of thing that goes quietly wrong: this file's matrix constant
/// said 256 with a comment warning that the S3's is 128, and the first hand-written replacement in
/// the HAL used 128 anyway. It now comes from `SIG_GPIO_OUT_IDX` in IDF's `gpio_sig_map.h`.
pub const gpio = @import("hal").gpio;

/// Mask ROM routines. These are the only "library" a bare image links against: the addresses come
/// from `components/esp_rom/esp32p4/ld/esp32p4.rom.ld` and the linker script re-declares them.
pub const rom = struct {
    pub extern fn ets_printf(fmt: [*:0]const u8, ...) c_int;
    pub extern fn ets_delay_us(us: u32) void;

    /// Tell the ROM the CPU's new frequency, in MHz. `ets_delay_us` and everything else built on
    /// `g_ticks_per_us` busy-waits by a cycle count derived from it, so a clock change without this
    /// makes every ROM delay wrong by exactly the ratio. `esp32p4.rom.ld:32`, 0x4fc00044.
    pub extern fn ets_update_cpu_frequency(mhz: u32) void;

    /// Invalidate the caches. `map` selects which, from `rom/cache.h:228-236`:
    /// L1 ICache0 = 1, ICache1 = 2, L1 DCache = 0x10, L2 = 0x20; `cache_all` is all four.
    ///
    /// This is not optional on a large image, and finding that out took a while. The second-stage
    /// bootloader leaves cache lines behind that do not match the mapping it finally installs, so an
    /// application can read its own `.rodata` and get someone else's bytes - deterministically,
    /// which is what makes it look like anything other than a cache. Measured: a load at 0x40035a1c
    /// returned `93 85 85 0f`, and after touching 512 KiB to evict the line the same load returned
    /// `3c ee 08 40`, which is what the image holds there. `_start` calls this before it clears
    /// `.bss`, so nothing of ours is in flight when the lines are dropped.
    pub extern fn Cache_Invalidate_All(map: u32) c_int;

    /// Every cache this chip has: L1 ICache0 | L1 ICache1 | L1 DCache | L2.
    pub const cache_all: u32 = 0x33;

    pub inline fn print(comptime fmt: [*:0]const u8, args: anytype) void {
        _ = @call(.auto, ets_printf, .{fmt} ++ args);
    }
};

/// Evict every flash-mapped cache line the bootloader left behind, by reading more flash than the
/// caches can hold. Call it before the first byte of `.rodata` is touched.
///
/// This is a workaround for a real defect in the hand-over, not a tidiness measure. The second-stage
/// bootloader leaves lines cached against a mapping it then replaces, so an application reads its
/// own `.rodata` and gets its own `.text` back - **deterministically**, which is exactly what makes
/// it look like anything other than a cache. Measured on this die: a load at `0x40035A1C` returned
/// `93 85 85 0f` before eviction and `3c ee 08 40` after, and the second is what the image holds
/// there. A string literal read before this runs is machine code, so a firmware whose first act is
/// to print a marker prints garbage and looks like it never booted at all.
///
/// Everything cheaper was tried first and every one of them said the hardware was fine, which is
/// why the list is here rather than being rediscovered:
///
///   * The MMU table is correct. Entries 0..9 read `0x1001`..`0x100a` - the valid bit plus physical
///     page N+1 - which is exactly what the image builder's single flash-to-vaddr anchor requires,
///     and 10..11 are unmapped as they should be. Read off the die through
///     `SPI_MEM_C_MMU_ITEM_INDEX_REG`, not inferred.
///   * The flash is correct. `zig build flash` verifies an MD5 of what the ROM stored, and the
///     image matches the ELF byte for byte at the addresses that misread.
///   * The page size is not in question: hardwired to 64 KiB on this chip
///     (`hal/esp32p4/mmu_ll.h:126-130` returns `MMU_PAGE_64KB` and the setter asserts it).
///   * Not fragmentation of the mapping either: the bad bytes arrive in one contiguous run of
///     >= 192 B, not in 64-byte lines, and they are identical across three resets and two
///     reflashes - determinism is what kept this looking like anything but a cache.
///
/// 512 KiB is four times the 128 KiB the L2 measured at (`examples/memprobe.zig` found real RAM
/// stopping at `0x4FFA0000`, the cache taking the rest), with the L1s smaller still.
///
/// `rom.Cache_Invalidate_All` is the instrument that ought to do this and does not: called from an
/// image the ROM did not launch, it faults inside the ROM with its argument stranded in `a2`.
/// Capacity eviction needs no preconditions, which is the whole reason it is what ships. 512 KiB at
/// a 64-byte stride is 8,192 loads, once per boot.
pub fn flushFlashCache() void {
    var sink: u32 = 0;
    var p: u32 = 0x4000_0000;
    while (p < 0x4008_0000) : (p += 64) {
        sink +%= @as(*volatile u32, @ptrFromInt(p)).*;
    }
    // Consumed through a volatile store, or the optimiser drops the whole loop as dead.
    @as(*volatile u32, &flush_sink).* = sink;
}

var flush_sink: u32 = 0;

/// Busy-wait for a number of CPU cycles, using the cycle counter rather than the mask ROM. Useful
/// when an image must not depend on ROM entry points at all, and for delays shorter than the ROM's
/// microsecond granularity.
pub inline fn delayCycles(n: u64) void {
    const start = cycles();
    while (cycles() - start < n) {}
}

/// Cycle counter: CSR 0xC00/0xC80, i.e. `cycle`/`cycleh` - the unprivileged shadows of mcycle, and
/// what ESP-IDF itself reads on this part (`rv_utils.h`: `RV_READ_CSR(cycle)`, because
/// SOC_CPU_HAS_CSR_PC is not defined for the P4).
///
/// Read high-low-high: two separate CSR reads can straddle a wrap of the low word, which would
/// otherwise report a value 2^32 too large roughly every 47 seconds at 90 MHz.
pub inline fn cycles() u64 {
    while (true) {
        var hi0: u32 = undefined;
        var lo: u32 = undefined;
        var hi1: u32 = undefined;
        asm volatile ("csrr %[r], 0xC80"
            : [r] "=r" (hi0),
        );
        asm volatile ("csrr %[r], 0xC00"
            : [r] "=r" (lo),
        );
        asm volatile ("csrr %[r], 0xC80"
            : [r] "=r" (hi1),
        );
        if (hi0 == hi1) return (@as(u64, hi0) << 32) | lo;
    }
}