ableos/hbvm/src/lib.rs

//! HoleyBytes Virtual Machine
//!
//! # Alloc feature
//! - Enabled by default
//! - Provides mapping / unmapping, as well as [`Default`] and [`Drop`]
//!   implementations for soft-paged memory implementation

// # General safety notice:
// - Validation has to assure there is 256 registers (r0 - r255)
// - Instructions have to be valid as specified (values and sizes)
// - Mapped pages should be at least 4 KiB

#![no_std]
#![cfg_attr(feature = "nightly", feature(fn_align))]
#![warn(missing_docs, clippy::missing_docs_in_private_items)]

#[cfg(feature = "alloc")]
extern crate alloc;

pub mod mem;
pub mod value;

mod bmc;

use {
    bmc::BlockCopier,
    core::{cmp::Ordering, mem::size_of, ops, slice::SliceIndex},
    derive_more::Display,
    hbbytecode::{OpParam, ParamBB, ParamBBB, ParamBBBB, ParamBBD, ParamBBDH, ParamBBW, ParamBD},
    value::{Value, ValueVariant},
};

/// HoleyBytes Virtual Machine
pub struct Vm<Mem, const TIMER_QUOTIENT: usize> {
    /// Holds 256 registers
    ///
    /// Writing to register 0 is considered undefined behaviour
    /// in terms of HoleyBytes program execution
    pub registers: [Value; 256],

    /// Memory implementation
    pub memory: Mem,

    /// Program counter
    pub pc: usize,

    /// Program timer
    timer: usize,

    /// Saved block copier
    copier: Option<BlockCopier>,
}

impl<Mem, const TIMER_QUOTIENT: usize> Vm<Mem, TIMER_QUOTIENT>
where
    Mem: Memory,
{
    /// Create a new VM with program and trap handler
    ///
    /// # Safety
    /// Program code has to be validated
    pub unsafe fn new(memory: Mem, entry: u64) -> Self {
        Self {
            registers: [Value::from(0_u64); 256],
            memory,
            pc: entry as _,
            timer: 0,
            copier: None,
        }
    }

    /// Execute program
    ///
    /// Program can return [`VmRunError`] if a trap handling failed
    #[cfg_attr(feature = "nightly", repr(align(4096)))]
    pub fn run(&mut self) -> Result<VmRunOk, VmRunError> {
        use hbbytecode::opcode::*;
        loop {
            // Big match
            //
            // Contribution guide:
            // - Zero register shall never be overwitten. It's value has to always be 0.
            //     - Prefer `Self::read_reg` and `Self::write_reg` functions
            // - Extract parameters using `param!` macro
            // - Prioritise speed over code size
            //     - Memory is cheap, CPUs not that much
            // - Do not heap allocate at any cost
            //     - Yes, user-provided trap handler may allocate,
            //       but that is not our »fault«.
            // - Unsafe is kinda must, but be sure you have validated everything
            //     - Your contributions have to pass sanitizers and Miri
            // - Strictly follow the spec
            //     - The spec does not specify how you perform actions, in what order,
            //       just that the observable effects have to be performed in order and
            //       correctly.
            // - Yes, we assume you run 64 bit CPU. Else ?conradluget a better CPU
            //   sorry 8 bit fans, HBVM won't run on your Speccy :(
            unsafe {
                match *self
                    .memory
                    .load_prog(self.pc)
                    .ok_or(VmRunError::ProgramFetchLoadEx(self.pc as _))?
                {
                    UN => {
                        self.decode::<()>();
                        return Err(VmRunError::Unreachable);
                    }
                    TX => {
                        self.decode::<()>();
                        return Ok(VmRunOk::End);
                    }
                    NOP => self.decode::<()>(),
                    ADD => self.binary_op(u64::wrapping_add),
                    SUB => self.binary_op(u64::wrapping_sub),
                    MUL => self.binary_op(u64::wrapping_mul),
                    AND => self.binary_op::<u64>(ops::BitAnd::bitand),
                    OR => self.binary_op::<u64>(ops::BitOr::bitor),
                    XOR => self.binary_op::<u64>(ops::BitXor::bitxor),
                    SL => self.binary_op(|l, r| u64::wrapping_shl(l, r as u32)),
                    SR => self.binary_op(|l, r| u64::wrapping_shr(l, r as u32)),
                    SRS => self.binary_op(|l, r| i64::wrapping_shl(l, r as u32)),
                    CMP => {
                        // Compare a0 <=> a1
                        // < → -1
                        // > →  1
                        // = →  0

                        let ParamBBB(tg, a0, a1) = self.decode();
                        self.write_reg(
                            tg,
                            self.read_reg(a0)
                                .cast::<i64>()
                                .cmp(&self.read_reg(a1).cast::<i64>())
                                as i64,
                        );
                    }
                    CMPU => {
                        // Unsigned comparsion
                        let ParamBBB(tg, a0, a1) = self.decode();
                        self.write_reg(
                            tg,
                            self.read_reg(a0)
                                .cast::<u64>()
                                .cmp(&self.read_reg(a1).cast::<u64>())
                                as i64,
                        );
                    }
                    NOT => {
                        // Logical negation
                        let ParamBB(tg, a0) = self.decode();
                        self.write_reg(tg, !self.read_reg(a0).cast::<u64>());
                    }
                    NEG => {
                        // Bitwise negation
                        let ParamBB(tg, a0) = self.decode();
                        self.write_reg(
                            tg,
                            match self.read_reg(a0).cast::<u64>() {
                                0 => 1_u64,
                                _ => 0,
                            },
                        );
                    }
                    DIR => {
                        // Fused Division-Remainder
                        let ParamBBBB(dt, rt, a0, a1) = self.decode();
                        let a0 = self.read_reg(a0).cast::<u64>();
                        let a1 = self.read_reg(a1).cast::<u64>();
                        self.write_reg(dt, a0.checked_div(a1).unwrap_or(u64::MAX));
                        self.write_reg(rt, a0.checked_rem(a1).unwrap_or(u64::MAX));
                    }
                    ADDI => self.binary_op_imm(u64::wrapping_add),
                    MULI => self.binary_op_imm(u64::wrapping_sub),
                    ANDI => self.binary_op_imm::<u64>(ops::BitAnd::bitand),
                    ORI => self.binary_op_imm::<u64>(ops::BitOr::bitor),
                    XORI => self.binary_op_imm::<u64>(ops::BitXor::bitxor),
                    SLI => self.binary_op_ims(u64::wrapping_shl),
                    SRI => self.binary_op_ims(u64::wrapping_shr),
                    SRSI => self.binary_op_ims(i64::wrapping_shr),
                    CMPI => {
                        let ParamBBD(tg, a0, imm) = self.decode();
                        self.write_reg(
                            tg,
                            self.read_reg(a0)
                                .cast::<i64>()
                                .cmp(&Value::from(imm).cast::<i64>())
                                as i64,
                        );
                    }
                    CMPUI => {
                        let ParamBBD(tg, a0, imm) = self.decode();
                        self.write_reg(tg, self.read_reg(a0).cast::<u64>().cmp(&imm) as i64);
                    }
                    CP => {
                        let ParamBB(tg, a0) = self.decode();
                        self.write_reg(tg, self.read_reg(a0));
                    }
                    SWA => {
                        // Swap registers
                        let ParamBB(r0, r1) = self.decode();
                        match (r0, r1) {
                            (0, 0) => (),
                            (dst, 0) | (0, dst) => self.write_reg(dst, 0_u64),
                            (r0, r1) => {
                                core::ptr::swap(
                                    self.registers.get_unchecked_mut(usize::from(r0)),
                                    self.registers.get_unchecked_mut(usize::from(r1)),
                                );
                            }
                        }
                    }
                    LI => {
                        let ParamBD(tg, imm) = self.decode();
                        self.write_reg(tg, imm);
                    }
                    LD => {
                        // Load. If loading more than register size, continue on adjecent registers
                        let ParamBBDH(dst, base, off, count) = self.decode();
                        let n: u8 = match dst {
                            0 => 1,
                            _ => 0,
                        };

                        self.memory.load(
                            self.ldst_addr_uber(dst, base, off, count, n)?,
                            self.registers
                                .as_mut_ptr()
                                .add(usize::from(dst) + usize::from(n))
                                .cast(),
                            usize::from(count).saturating_sub(n.into()),
                        )?;
                    }
                    ST => {
                        // Store. Same rules apply as to LD
                        let ParamBBDH(dst, base, off, count) = self.decode();
                        self.memory.store(
                            self.ldst_addr_uber(dst, base, off, count, 0)?,
                            self.registers.as_ptr().add(usize::from(dst)).cast(),
                            count.into(),
                        )?;
                    }
                    BMC => {
                        // Block memory copy
                        match if let Some(copier) = &mut self.copier {
                            // There is some copier, poll.
                            copier.poll(&mut self.memory)
                        } else {
                            // There is none, make one!
                            let ParamBBD(src, dst, count) = self.decode();

                            // So we are still on BMC on next cycle
                            self.pc -= size_of::<ParamBBD>() + 1;

                            self.copier = Some(BlockCopier::new(
                                self.read_reg(src).cast(),
                                self.read_reg(dst).cast(),
                                count as _,
                            ));

                            self.copier
                                .as_mut()
                                .unwrap_unchecked() // SAFETY: We just assigned there
                                .poll(&mut self.memory)
                        } {
                            // We are done, shift program counter
                            core::task::Poll::Ready(Ok(())) => {
                                self.copier = None;
                                self.pc += size_of::<ParamBBD>() + 1;
                            }
                            // Error, shift program counter (for consistency)
                            // and yield error
                            core::task::Poll::Ready(Err(e)) => {
                                self.pc += size_of::<ParamBBD>() + 1;
                                return Err(e.into());
                            }
                            // Not done yet, proceed to next cycle
                            core::task::Poll::Pending => (),
                        }
                    }
                    BRC => {
                        // Block register copy
                        let ParamBBB(src, dst, count) = self.decode();
                        if src.checked_add(count).is_none() || dst.checked_add(count).is_none() {
                            return Err(VmRunError::RegOutOfBounds);
                        }

                        core::ptr::copy(
                            self.registers.get_unchecked(usize::from(src)),
                            self.registers.get_unchecked_mut(usize::from(dst)),
                            usize::from(count),
                        );
                    }
                    JAL => {
                        // Jump and link. Save PC after this instruction to
                        // specified register and jump to reg + offset.
                        let ParamBBD(save, reg, offset) = self.decode();
                        self.write_reg(save, self.pc as u64);
                        self.pc =
                            (self.read_reg(reg).cast::<u64>().saturating_add(offset)) as usize;
                    }
                    // Conditional jumps, jump only to immediates
                    JEQ => self.cond_jmp::<u64>(Ordering::Equal),
                    JNE => {
                        let ParamBBD(a0, a1, jt) = self.decode();
                        if self.read_reg(a0).cast::<u64>() != self.read_reg(a1).cast::<u64>() {
                            self.pc = jt as usize;
                        }
                    }
                    JLT => self.cond_jmp::<u64>(Ordering::Less),
                    JGT => self.cond_jmp::<u64>(Ordering::Greater),
                    JLTU => self.cond_jmp::<i64>(Ordering::Less),
                    JGTU => self.cond_jmp::<i64>(Ordering::Greater),
                    ECALL => {
                        self.decode::<()>();

                        // So we don't get timer interrupt after ECALL
                        if TIMER_QUOTIENT != 0 {
                            self.timer = self.timer.wrapping_add(1);
                        }
                        return Ok(VmRunOk::Ecall);
                    }
                    ADDF => self.binary_op::<f64>(ops::Add::add),
                    SUBF => self.binary_op::<f64>(ops::Sub::sub),
                    MULF => self.binary_op::<f64>(ops::Mul::mul),
                    DIRF => {
                        let ParamBBBB(dt, rt, a0, a1) = self.decode();
                        let a0 = self.read_reg(a0).cast::<f64>();
                        let a1 = self.read_reg(a1).cast::<f64>();
                        self.write_reg(dt, a0 / a1);
                        self.write_reg(rt, a0 % a1);
                    }
                    FMAF => {
                        let ParamBBBB(dt, a0, a1, a2) = self.decode();
                        self.write_reg(
                            dt,
                            self.read_reg(a0).cast::<f64>() * self.read_reg(a1).cast::<f64>()
                                + self.read_reg(a2).cast::<f64>(),
                        );
                    }
                    NEGF => {
                        let ParamBB(dt, a0) = self.decode();
                        self.write_reg(dt, -self.read_reg(a0).cast::<f64>());
                    }
                    ITF => {
                        let ParamBB(dt, a0) = self.decode();
                        self.write_reg(dt, self.read_reg(a0).cast::<i64>() as f64);
                    }
                    FTI => {
                        let ParamBB(dt, a0) = self.decode();
                        self.write_reg(dt, self.read_reg(a0).cast::<f64>() as i64);
                    }
                    ADDFI => self.binary_op_imm::<f64>(ops::Add::add),
                    MULFI => self.binary_op_imm::<f64>(ops::Mul::mul),
                    op => return Err(VmRunError::InvalidOpcode(op)),
                }
            }

            if TIMER_QUOTIENT != 0 {
                self.timer = self.timer.wrapping_add(1);
                if self.timer % TIMER_QUOTIENT == 0 {
                    return Ok(VmRunOk::Timer);
                }
            }
        }
    }

    /// Decode instruction operands
    #[inline(always)]
    unsafe fn decode<T: OpParam>(&mut self) -> T {
        let pc1 = self.pc + 1;
        let data = self
            .memory
            .load_prog_unchecked(pc1..pc1 + size_of::<T>())
            .as_ptr()
            .cast::<T>()
            .read();

        self.pc += 1 + size_of::<T>();
        data
    }

    /// Perform binary operating over two registers
    #[inline(always)]
    unsafe fn binary_op<T: ValueVariant>(&mut self, op: impl Fn(T, T) -> T) {
        let ParamBBB(tg, a0, a1) = self.decode();
        self.write_reg(
            tg,
            op(self.read_reg(a0).cast::<T>(), self.read_reg(a1).cast::<T>()),
        );
    }

    /// Perform binary operation over register and immediate
    #[inline(always)]
    unsafe fn binary_op_imm<T: ValueVariant>(&mut self, op: impl Fn(T, T) -> T) {
        let ParamBBD(tg, reg, imm) = self.decode();
        self.write_reg(
            tg,
            op(self.read_reg(reg).cast::<T>(), Value::from(imm).cast::<T>()),
        );
    }

    /// Perform binary operation over register and shift immediate
    #[inline(always)]
    unsafe fn binary_op_ims<T: ValueVariant>(&mut self, op: impl Fn(T, u32) -> T) {
        let ParamBBW(tg, reg, imm) = self.decode();
        self.write_reg(tg, op(self.read_reg(reg).cast::<T>(), imm));
    }

    /// Jump at `#3` if ordering on `#0 <=> #1` is equal to expected
    #[inline(always)]
    unsafe fn cond_jmp<T: ValueVariant + Ord>(&mut self, expected: Ordering) {
        let ParamBBD(a0, a1, ja) = self.decode();
        if self
            .read_reg(a0)
            .cast::<T>()
            .cmp(&self.read_reg(a1).cast::<T>())
            == expected
        {
            self.pc = ja as usize;
        }
    }

    /// Read register
    #[inline(always)]
    unsafe fn read_reg(&self, n: u8) -> Value {
        *self.registers.get_unchecked(n as usize)
    }

    /// Write a register.
    /// Writing to register 0 is no-op.
    #[inline(always)]
    unsafe fn write_reg(&mut self, n: u8, value: impl Into<Value>) {
        if n != 0 {
            *self.registers.get_unchecked_mut(n as usize) = value.into();
        }
    }

    /// Load / Store Address check-computation überfunction
    #[inline(always)]
    unsafe fn ldst_addr_uber(
        &self,
        dst: u8,
        base: u8,
        offset: u64,
        size: u16,
        adder: u8,
    ) -> Result<u64, VmRunError> {
        let reg = dst.checked_add(adder).ok_or(VmRunError::RegOutOfBounds)?;

        if usize::from(reg) * 8 + usize::from(size) > 2048 {
            Err(VmRunError::RegOutOfBounds)
        } else {
            self.read_reg(base)
                .cast::<u64>()
                .checked_add(offset)
                .and_then(|x| x.checked_add(adder.into()))
                .ok_or(VmRunError::AddrOutOfBounds)
        }
    }
}

/// Virtual machine halt error
#[derive(Copy, Clone, Debug, PartialEq, Eq)]
#[repr(u8)]
pub enum VmRunError {
    /// Tried to execute invalid instruction
    InvalidOpcode(u8),

    /// Unhandled load access exception
    LoadAccessEx(u64),

    /// Unhandled instruction load access exception
    ProgramFetchLoadEx(u64),

    /// Unhandled store access exception
    StoreAccessEx(u64),

    /// Register out-of-bounds access
    RegOutOfBounds,

    /// Address out-of-bounds
    AddrOutOfBounds,

    /// Reached unreachable code
    Unreachable,
}

/// Virtual machine halt ok
#[derive(Copy, Clone, Debug, PartialEq, Eq)]
pub enum VmRunOk {
    /// Program has eached its end
    End,

    /// Program was interrupted by a timer
    Timer,

    /// Environment call
    Ecall,
}

/// Load-store memory access
pub trait Memory {
    /// Load data from memory on address
    ///
    /// # Safety
    /// - Shall not overrun the buffer
    unsafe fn load(&mut self, addr: u64, target: *mut u8, count: usize) -> Result<(), LoadError>;

    /// Store data to memory on address
    ///
    /// # Safety
    /// - Shall not overrun the buffer
    unsafe fn store(
        &mut self,
        addr: u64,
        source: *const u8,
        count: usize,
    ) -> Result<(), StoreError>;

    /// Fetch bytes from program section
    ///
    /// # Why?
    /// Even Holey Bytes programs operate with
    /// single address space, the actual implementation
    /// may be different, so for these reasons there is a
    /// separate function.
    ///
    /// Also if your memory implementation differentiates between
    /// readable and executable memory, this is the way to distinguish
    /// the loads.
    ///
    /// # Notice for implementors
    /// This is a hot function. This is called on each opcode fetch
    /// and instruction decode. Inlining the implementation is highly
    /// recommended!
    ///
    /// If you utilise some more heavy memory implementation, consider
    /// performing caching as HBVM does not do that for you.
    ///
    /// Has to return all the requested data. If cannot fetch data of requested
    /// length, return [`None`].
    fn load_prog<I>(&mut self, index: I) -> Option<&I::Output>
    where
        I: SliceIndex<[u8]>;

    /// Fetch bytes from program section, unchecked.
    ///
    /// # Safety
    /// You really have to be sure you get the bytes, got me?
    unsafe fn load_prog_unchecked<I>(&mut self, index: I) -> &I::Output
    where
        I: SliceIndex<[u8]>;
}

/// Unhandled load access trap
#[derive(Clone, Copy, Display, Debug, PartialEq, Eq)]
#[display(fmt = "Load access error at address {_0:#x}")]
pub struct LoadError(pub u64);

/// Unhandled store access trap
#[derive(Clone, Copy, Display, Debug, PartialEq, Eq)]
#[display(fmt = "Store access error at address {_0:#x}")]
pub struct StoreError(pub u64);

/// Reason to access memory
#[derive(Clone, Copy, Display, Debug, PartialEq, Eq)]
pub enum MemoryAccessReason {
    /// Memory was accessed for load (read)
    Load,
    /// Memory was accessed for store (write)
    Store,
}

impl From<LoadError> for VmRunError {
    fn from(value: LoadError) -> Self {
        Self::LoadAccessEx(value.0)
    }
}

impl From<StoreError> for VmRunError {
    fn from(value: StoreError) -> Self {
        Self::StoreAccessEx(value.0)
    }
}