//! HoleyBytes Virtual Machine //! //! All unsafe code here should be sound, if input bytecode passes validation. // # 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 use self::mem::HandlePageFault; pub mod mem; pub mod value; use { crate::validate, core::ops, hbbytecode::{OpParam, ParamBB, ParamBBB, ParamBBBB, ParamBBD, ParamBBDH, ParamBD}, mem::Memory, static_assertions::assert_impl_one, value::Value, }; /// Extract a parameter from program macro_rules! param { ($self:expr, $ty:ty) => {{ assert_impl_one!($ty: OpParam); let data = $self .program .as_ptr() .add($self.pc + 1) .cast::<$ty>() .read(); $self.pc += 1 + core::mem::size_of::<$ty>(); data }}; } /// Perform binary operation `#0 ← #1 OP #2` macro_rules! binary_op { ($self:expr, $ty:ident, $handler:expr) => {{ let ParamBBB(tg, a0, a1) = param!($self, ParamBBB); $self.write_reg( tg, $handler( Value::$ty(&$self.read_reg(a0)), Value::$ty(&$self.read_reg(a1)), ), ); }}; ($self:expr, $ty:ident, $handler:expr, $con:ty) => {{ let ParamBBB(tg, a0, a1) = param!($self, ParamBBB); $self.write_reg( tg, $handler( Value::$ty(&$self.read_reg(a0)), Value::$ty(&$self.read_reg(a1)) as $con, ), ); }}; } /// Perform binary operation with immediate `#0 ← #1 OP imm #2` macro_rules! binary_op_imm { ($self:expr, $ty:ident, $handler:expr) => {{ let ParamBBD(tg, a0, imm) = param!($self, ParamBBD); $self.write_reg( tg, $handler(Value::$ty(&$self.read_reg(a0)), Value::$ty(&imm.into())), ); }}; ($self:expr, $ty:ident, $handler:expr, $con:ty) => {{ let ParamBBD(tg, a0, imm) = param!($self, ParamBBD); $self.write_reg( tg, $handler(Value::$ty(&$self.read_reg(a0)), Value::$ty(&imm.into()) as $con), ); }}; } /// Jump at `#3` if ordering on `#0 <=> #1` is equal to expected macro_rules! cond_jump { ($self:expr, $ty:ident, $expected:ident) => {{ let ParamBBD(a0, a1, jt) = param!($self, ParamBBD); if core::cmp::Ord::cmp(&$self.read_reg(a0).as_u64(), &$self.read_reg(a1).as_u64()) == core::cmp::Ordering::$expected { $self.pc = jt as usize; } }}; } /// HoleyBytes Virtual Machine pub struct Vm<'a, PfHandler, 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: Memory, /// Trap handler pub pfhandler: PfHandler, /// Program counter pc: usize, /// Program program: &'a [u8], /// Cached program length (without unreachable end) program_len: usize, /// Program timer timer: usize, } impl<'a, PfHandler: HandlePageFault, const TIMER_QUOTIENT: usize> Vm<'a, PfHandler, TIMER_QUOTIENT> { /// Create a new VM with program and trap handler /// /// # Safety /// Program code has to be validated pub unsafe fn new_unchecked(program: &'a [u8], traph: PfHandler) -> Self { Self { registers: [Value::from(0_u64); 256], memory: Default::default(), pfhandler: traph, pc: 0, program_len: program.len() - 12, program, timer: 0, } } /// Create a new VM with program and trap handler only if it passes validation pub fn new_validated(program: &'a [u8], traph: PfHandler) -> Result { validate::validate(program)?; Ok(unsafe { Self::new_unchecked(program, traph) }) } /// Execute program /// /// Program can return [`VmRunError`] if a trap handling failed pub fn run(&mut self) -> Result { use hbbytecode::opcode::*; loop { // Check instruction boundary if self.pc >= self.program_len { return Ok(VmRunOk::End); } // 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.program.get_unchecked(self.pc) { UN => { param!(self, ()); return Err(VmRunError::Unreachable); } NOP => param!(self, ()), ADD => binary_op!(self, as_u64, u64::wrapping_add), SUB => binary_op!(self, as_u64, u64::wrapping_sub), MUL => binary_op!(self, as_u64, u64::wrapping_mul), AND => binary_op!(self, as_u64, ops::BitAnd::bitand), OR => binary_op!(self, as_u64, ops::BitOr::bitor), XOR => binary_op!(self, as_u64, ops::BitXor::bitxor), SL => binary_op!(self, as_u64, u64::wrapping_shl, u32), SR => binary_op!(self, as_u64, u64::wrapping_shr, u32), SRS => binary_op!(self, as_i64, i64::wrapping_shr, u32), CMP => { // Compare a0 <=> a1 // < → -1 // > → 1 // = → 0 let ParamBBB(tg, a0, a1) = param!(self, ParamBBB); self.write_reg( tg, self.read_reg(a0).as_i64().cmp(&self.read_reg(a1).as_i64()) as i64, ); } CMPU => { // Unsigned comparsion let ParamBBB(tg, a0, a1) = param!(self, ParamBBB); self.write_reg( tg, self.read_reg(a0).as_u64().cmp(&self.read_reg(a1).as_u64()) as i64, ); } NOT => { // Logical negation let param = param!(self, ParamBB); self.write_reg(param.0, !self.read_reg(param.1).as_u64()); } NEG => { // Bitwise negation let param = param!(self, ParamBB); self.write_reg( param.0, match self.read_reg(param.1).as_u64() { 0 => 1_u64, _ => 0, }, ); } DIR => { // Fused Division-Remainder let ParamBBBB(dt, rt, a0, a1) = param!(self, ParamBBBB); let a0 = self.read_reg(a0).as_u64(); let a1 = self.read_reg(a1).as_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 => binary_op_imm!(self, as_u64, ops::Add::add), MULI => binary_op_imm!(self, as_u64, ops::Mul::mul), ANDI => binary_op_imm!(self, as_u64, ops::BitAnd::bitand), ORI => binary_op_imm!(self, as_u64, ops::BitOr::bitor), XORI => binary_op_imm!(self, as_u64, ops::BitXor::bitxor), SLI => binary_op_imm!(self, as_u64, u64::wrapping_shl, u32), SRI => binary_op_imm!(self, as_u64, u64::wrapping_shr, u32), SRSI => binary_op_imm!(self, as_i64, i64::wrapping_shr, u32), CMPI => { let ParamBBD(tg, a0, imm) = param!(self, ParamBBD); self.write_reg( tg, self.read_reg(a0).as_i64().cmp(&Value::from(imm).as_i64()) as i64, ); } CMPUI => { let ParamBBD(tg, a0, imm) = param!(self, ParamBBD); self.write_reg(tg, self.read_reg(a0).as_u64().cmp(&imm) as i64); } CP => { let param = param!(self, ParamBB); self.write_reg(param.0, self.read_reg(param.1)); } SWA => { // Swap registers let ParamBB(r0, r1) = param!(self, ParamBB); 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 param = param!(self, ParamBD); self.write_reg(param.0, param.1); } LD => { // Load. If loading more than register size, continue on adjecent registers let ParamBBDH(dst, base, off, count) = param!(self, ParamBBDH); let n: usize = match dst { 0 => 1, _ => 0, }; self.memory.load( self.read_reg(base).as_u64() + off + n as u64, self.registers.as_mut_ptr().add(usize::from(dst) + n).cast(), usize::from(count).saturating_sub(n), &mut self.pfhandler, )?; } ST => { // Store. Same rules apply as to LD let ParamBBDH(dst, base, off, count) = param!(self, ParamBBDH); self.memory.store( self.read_reg(base).as_u64() + off, self.registers.as_ptr().add(usize::from(dst)).cast(), count.into(), &mut self.pfhandler, )?; } BMC => { // Block memory copy let ParamBBD(src, dst, count) = param!(self, ParamBBD); self.memory.block_copy( self.read_reg(src).as_u64(), self.read_reg(dst).as_u64(), count as _, &mut self.pfhandler, )?; } BRC => { // Block register copy let ParamBBB(src, dst, count) = param!(self, ParamBBB); 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) = param!(self, ParamBBD); self.write_reg(save, self.pc as u64); self.pc = (self.read_reg(reg).as_u64() + offset) as usize; } // Conditional jumps, jump only to immediates JEQ => cond_jump!(self, int, Equal), JNE => { let ParamBBD(a0, a1, jt) = param!(self, ParamBBD); if self.read_reg(a0).as_u64() != self.read_reg(a1).as_u64() { self.pc = jt as usize; } } JLT => cond_jump!(self, int, Less), JGT => cond_jump!(self, int, Greater), JLTU => cond_jump!(self, sint, Less), JGTU => cond_jump!(self, sint, Greater), ECALL => { param!(self, ()); // 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 => binary_op!(self, as_f64, ops::Add::add), SUBF => binary_op!(self, as_f64, ops::Sub::sub), MULF => binary_op!(self, as_f64, ops::Mul::mul), DIRF => { let ParamBBBB(dt, rt, a0, a1) = param!(self, ParamBBBB); let a0 = self.read_reg(a0).as_f64(); let a1 = self.read_reg(a1).as_f64(); self.write_reg(dt, a0 / a1); self.write_reg(rt, a0 % a1); } FMAF => { let ParamBBBB(dt, a0, a1, a2) = param!(self, ParamBBBB); self.write_reg( dt, self.read_reg(a0).as_f64() * self.read_reg(a1).as_f64() + self.read_reg(a2).as_f64(), ); } NEGF => { let ParamBB(dt, a0) = param!(self, ParamBB); self.write_reg(dt, -self.read_reg(a0).as_f64()); } ITF => { let ParamBB(dt, a0) = param!(self, ParamBB); self.write_reg(dt, self.read_reg(a0).as_i64() as f64); } FTI => { let ParamBB(dt, a0) = param!(self, ParamBB); self.write_reg(dt, self.read_reg(a0).as_f64() as i64); } ADDFI => binary_op_imm!(self, as_f64, ops::Add::add), MULFI => binary_op_imm!(self, as_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); } } } } /// Read register #[inline] 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] unsafe fn write_reg(&mut self, n: u8, value: impl Into) { if n != 0 { *self.registers.get_unchecked_mut(n as usize) = value.into(); } } } /// 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 store access exception StoreAccessEx(u64), /// 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, }