Merge pull request 'Merge compiler and emulator progress from last few months into main.' (#11) from compiler into main

Reviewed-on: #11

.cargo/config.toml

@@ -5,3 +5,7 @@ rustc-wrapper = "sccache"
 
 [future-incompat-report]
 frequency = "always"
+
+[profile.profiling]
+inherits = "release"
+debug = true

.vscode/settings.json

@@ -8,4 +8,8 @@
   "files.trimTrailingWhitespace": true,
   "gitea.owner": "LowLevelDevs",
   "gitea.repo": "damn_simple_architecture",
+  "[markdown]": {
+    "editor.formatOnSave": true,
+    "editor.formatOnPaste": true
+  }
 }

.zed/settings.json

@@ -0,0 +1,15 @@
+// Folder-specific settings
+//
+// For a full list of overridable settings, and general information on folder-specific settings,
+// see the documentation: https://zed.dev/docs/configuring-zed#settings-files
+{
+  "lsp": {
+    "rust-analyzer": {
+      "initialization_options": {
+        "check": {
+          "command": "clippy", // rust-analyzer.check.command (default: "check")
+        },
+      },
+    },
+  },
+}

.zed/tasks.json

@@ -0,0 +1,37 @@
+[
+  {
+    "label": "Run Emulator",
+    "command": "cargo run --bin emulator",
+    "use_new_terminal": true,
+  },
+  {
+    "label": "Run Compiler",
+    "command": "cargo run --bin compiler",
+    "use_new_terminal": true,
+  },
+  {
+    "label": "Run Assembler",
+    "command": "cargo run --bin assembler",
+    "use_new_terminal": true,
+  },
+  {
+    "label": "Run Build System (dsx-build)",
+    "command": "cargo run --bin dsx-build",
+    "use_new_terminal": true,
+  },
+  {
+    "label": "Build All (Release)",
+    "command": "cargo build --release",
+    "use_new_terminal": false,
+  },
+  {
+    "label": "Run Tests",
+    "command": "cargo test",
+    "use_new_terminal": true,
+  },
+  {
+    "label": "Profile Emulator with perf",
+    "command": "cargo build --profile profiling; perf record -g -F 999 target/profiling/emulator; perf script -F +pid | save test.perf",
+    "use_new_terminal": true,
+  },
+]

Cargo.toml

@@ -11,7 +11,11 @@ authors = ["zxq5", "nullndvoid"]
 
 [profile.dev]
 codegen-backend = "cranelift"
-panic = "abort" # Cranelift does not support stack unwinds.
+panic = "abort"               # Cranelift does not support stack unwinds.
 lto = false
 debug = true
-incremental = false # sccache does not support caching incremental crates.
+incremental = false           # sccache does not support caching incremental crates.
+
+[profile.release]
+debug = true
+lto = "fat"

assembler/Cargo.toml

@@ -5,7 +5,7 @@ edition.workspace = true
 authors.workspace = true
 
 [[bin]]
-name = "assembler_runner"
+name = "assembler"
 path = "src/main.rs"
 
 [lib]

assembler/src/assembler/codegen.rs

@@ -223,19 +223,31 @@ fn build_shift_instruction(
     opcode: Opcode,
     args: &[crate::assembler::model::Token],
 ) -> Result<Instruction, AssembleError> {
-    let Some(reg_token) = args.first() else {
+    let Some(src_reg) = args.first() else {
         return Err(AssembleError::MissingArgument(0));
     };
-    let Some(amount_token) = args.get(1) else {
+    let Some(r_shamt) = args.get(1) else {
+        return Err(AssembleError::MissingArgument(0));
+    };
+    let Some(i_shamt) = args.get(2) else {
+        return Err(AssembleError::MissingArgument(1));
+    };
+    let Some(dest_reg) = args.get(3) else {
         return Err(AssembleError::MissingArgument(1));
     };
 
-    let reg = expect_token!(reg_token, Register)?;
-    let amount = expect_token!(amount_token, Immediate)? as u8;
+    let src = expect_token!(src_reg, Register)?;
+    let r_shamt = expect_token!(r_shamt, Register)?;
+    let i_shamt = expect_token!(i_shamt, Immediate)? as u8;
+    let dest = expect_token!(dest_reg, Register)?;
 
     match opcode {
-        Opcode::Shl => Ok(Instruction::ShiftLeft(args!(R, sr1: reg, shamt: amount))),
-        Opcode::Shr => Ok(Instruction::ShiftRight(args!(R, sr1: reg, shamt: amount))),
+        Opcode::Shl => Ok(Instruction::ShiftLeft(
+            args!(R, sr1: src, sr2: r_shamt, shamt: i_shamt, dr: dest),
+        )),
+        Opcode::Shr => Ok(Instruction::ShiftRight(
+            args!(R, sr1: src, sr2: r_shamt, shamt: i_shamt, dr: dest),
+        )),
         _ => unreachable!(),
     }
 }

assembler/src/assembler/macros.rs

@@ -4,7 +4,8 @@ use crate::assembler::model::{Node, Opcode, Symbol, Token};
 /// Parse DSA assembly code with optional formatting
 ///
 /// # Examples
-/// ```
+/// ```rs
+/// use assembler::macros::dsa;
 /// // With formatting:
 /// let nodes = dsa!(hash, "mov r1, {}", 42)?;
 ///

assembler/src/assembler/model.rs

@@ -184,11 +184,11 @@ pub enum Token {
 impl fmt::Display for Token {
     fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
         match self {
-            Self::Symbol(symbol) => write!(f, "{}", symbol),
-            Self::Register(register) => write!(f, "{}", register),
-            Self::Immediate(immediate) => write!(f, "{}", immediate),
-            Self::StringLit(string_lit) => write!(f, "{}", string_lit),
-            Self::Opcode(opcode) => write!(f, "{}", opcode),
+            Self::Symbol(symbol) => write!(f, "{symbol}"),
+            Self::Register(register) => write!(f, "{register}",),
+            Self::Immediate(immediate) => write!(f, "{immediate}",),
+            Self::StringLit(string_lit) => write!(f, "{string_lit}",),
+            Self::Opcode(opcode) => write!(f, "{opcode}",),
         }
     }
 }

assembler/src/assembler/parser.rs

@@ -1,5 +1,6 @@
 use std::path::{Path, PathBuf};
 
+use crate::assembler::TokenType;
 use crate::{assembler::AssembleError, expect_token, expect_type, node};
 
 use crate::assembler::model::{Node, Opcode, Token};
@@ -100,6 +101,7 @@ impl Parser {
         let opcode = expect_token!(self.next()?, Opcode)?;
         let args: Vec<Token>;
 
+        #[allow(clippy::match_same_arms)]
         match opcode {
             // R-type instructions
             Opcode::Mov | Opcode::Movs => {
@@ -112,22 +114,25 @@ impl Parser {
                 let base = expect_type!(self.next()?, Register, Symbol)?;
                 let dest = expect_type!(self.next()?, Register)?;
 
-                let mut offset = Token::Immediate(0);
-                if let Ok(next) = self.peek_next()
-                    && expect_type!(next, Immediate).is_ok() {
-                        offset = self.next()?;
+                let offset = match self.peek_next() {
+                    Ok(next) if expect_type!(next.clone(), Immediate).is_ok() => {
+                        self.next()?
                     }
+                    _ => Token::Immediate(0),
+                };
 
                 args = vec![base, dest, offset];
             }
             Opcode::Stb | Opcode::Sth | Opcode::Stw => {
                 let base = expect_type!(self.next()?, Register)?;
                 let dest = expect_type!(self.next()?, Register, Symbol)?;
-                let mut offset = Token::Immediate(0);
-                if let Ok(next) = self.peek_next()
-                    && expect_type!(next, Immediate).is_ok() {
-                        offset = self.next()?;
+
+                let offset = match self.peek_next() {
+                    Ok(next) if expect_type!(next.clone(), Immediate).is_ok() => {
+                        self.next()?
                     }
+                    _ => Token::Immediate(0),
+                };
                 args = vec![base, dest, offset];
             }
 
@@ -146,15 +151,49 @@ impl Parser {
             }
 
             Opcode::Not | Opcode::Cmp => {
-                let reg1 = expect_type!(self.next()?, Register, Symbol)?;
-                let reg2 = expect_type!(self.next()?, Register, Symbol)?;
-                args = vec![reg1, reg2];
+                let src = expect_type!(self.next()?, Register, Symbol)?;
+                let dest = expect_type!(self.next()?, Register, Symbol)?;
+                args = vec![src, dest];
             }
-
             Opcode::Shl | Opcode::Shr => {
-                let reg = expect_type!(self.next()?, Register, Symbol)?;
-                let num = expect_type!(self.next()?, Immediate)?;
-                args = vec![reg, num];
+                let src = expect_type!(self.next()?, Register, Symbol)?;
+
+                // First operand after src: could be immediate or register
+                let first = self.next()?;
+
+                let (r_shamt, i_shamt) = match first {
+                    Token::Register(_) => (
+                        first,
+                        if let Ok(tok) = self.peek_next() {
+                            if expect_type!(tok, Immediate).is_ok() {
+                                self.next()?
+                            } else {
+                                Token::Immediate(0)
+                            }
+                        } else {
+                            Token::Immediate(0)
+                        },
+                    ),
+                    Token::Immediate(_) => (Token::Register(Register::Zero), first),
+                    _ => {
+                        return Err(AssembleError::UnexpectedToken(
+                            first,
+                            TokenType::Immediate,
+                        ));
+                    }
+                };
+
+                let dest = if let Ok(tok) = self.peek_next() {
+                    if expect_type!(tok, Register).is_ok() {
+                        self.next()?
+                    } else {
+                        src.clone() // Default to src if no dest specified
+                    }
+                } else {
+                    src.clone() // Default to src if no dest specified
+                };
+
+                args = vec![src, r_shamt, i_shamt, dest];
             }
 
             Opcode::Inc | Opcode::Dec => {

assembler/src/lib.rs

@@ -34,7 +34,7 @@ use crate::prelude::CompilerEngine;
 pub fn assemble_file(input: &str, output: &str) -> Result<(), std::io::Error> {
     let mut engine = CompilerEngine::new();
     engine.start_compilation(Path::new(input));
-    let result = engine.wait_for_result().unwrap();
+    let result = engine.wait_for_result().expect("assembler failed.");
 
     let buffer: Vec<u8> = result
         .iter()

clippy.toml

@@ -0,0 +1 @@
+disallowed-types = ["std::collections::HashMap", "std::collections::HashSet"]

common/src/instructions.rs

@@ -40,7 +40,7 @@ pub enum InstructionType {
     Immediate,
 }
 
-#[derive(Copy, Clone, Debug, PartialEq, Eq)]
+#[derive(Copy, Clone, Debug, PartialEq, Eq, Default)]
 #[non_exhaustive]
 pub enum Register {
     // general purpose registers
@@ -69,7 +69,9 @@ pub enum Register {
     Idr,
     Mmr,
     Zero,
-    NoReg,
+
+    #[default]
+    Null, // Invalid - Triggers a fault if accessed
 
     // system registers - can't be written to by instructions.
     Mar,
@@ -104,12 +106,6 @@ impl Register {
     }
 }
 
-impl Default for Register {
-    fn default() -> Self {
-        Self::NoReg
-    }
-}
-
 impl TryFrom<u8> for Register {
     type Error = RegisterParseError;
 
@@ -144,7 +140,7 @@ impl TryFrom<u8> for Register {
             0x14 => Self::Idr,
             0x15 => Self::Mmr,
             0x16 => Self::Zero,
-            0x17 => Self::NoReg,
+            0x17 => Self::Null,
             0x18 => Self::Mar,
             0x19 => Self::Mdr,
             0x1A => Self::Sts,
@@ -183,7 +179,7 @@ impl TryFrom<&str> for Register {
             "idr" => Ok(Self::Idr),
             "mmr" => Ok(Self::Mmr),
             "zero" => Ok(Self::Zero),
-            "null" => Ok(Self::NoReg),
+            "null" => Ok(Self::Null),
             "pcx" => Ok(Self::Pcx),
             _ => Err(RegisterParseError::InvalidName(value.to_string())),
         }
@@ -216,7 +212,7 @@ impl std::fmt::Display for Register {
             Self::Idr => write!(f, "idr"),
             Self::Mmr => write!(f, "mmr"),
             Self::Zero => write!(f, "zero"),
-            Self::NoReg => write!(f, "noreg"),
+            Self::Null => write!(f, "null"),
             Self::Mar => write!(f, "mar"),
             Self::Mdr => write!(f, "mdr"),
             Self::Sts => write!(f, "sts"),

common/src/instructions/encode.rs

@@ -8,9 +8,9 @@ pub trait Encode {
 /// Encodes a zero argument instruction.
 fn encode_no_args(opcode: u8) -> u32 {
     let opcode = u32::from(opcode);
-    let sr1 = Register::NoReg as u32;
-    let sr2 = Register::NoReg as u32;
-    let dr = Register::NoReg as u32;
+    let sr1 = Register::Null as u32;
+    let sr2 = Register::Null as u32;
+    let dr = Register::Null as u32;
     let shamt = 0;
 
     (opcode << 26) | (sr1 << 21) | (sr2 << 16) | (dr << 11) | (shamt << 6)

common/src/instructions/encode/tests.rs

@@ -2,7 +2,7 @@ use crate::prelude::*;
 
 #[test]
 fn test_encode_nop() {
-    let no_reg = Register::NoReg as u32;
+    let no_reg = Register::Null as u32;
     let no_op = u32::from(Instruction::Nop.opcode());
 
     let expected = (no_op << 26) | (no_reg << 21) | (no_reg << 16) | (no_reg << 11);
@@ -15,7 +15,7 @@ fn test_encode_nop() {
 fn test_encode_mov() {
     let rg0 = Register::Rg0 as u32;
     let rg1 = Register::Rg1 as u32;
-    let no_reg = Register::NoReg as u32;
+    let no_reg = Register::Null as u32;
 
     let instruction = Instruction::Mov(RTypeArgs::new(
         Some(Register::Rg0),
@@ -53,7 +53,7 @@ fn test_encode_load_byte() {
 #[test]
 fn test_encode_shift_left_shamt() {
     let rg0 = Register::Rg0 as u32;
-    let no_reg = Register::NoReg as u32;
+    let no_reg = Register::Null as u32;
 
     let shift_amount = 5;
 
@@ -80,7 +80,7 @@ fn test_encode_shift_left_shamt() {
 fn test_encode_shift_left_reg() {
     let rg0 = Register::Rg0 as u32;
     let rg1 = Register::Rg1 as u32;
-    let no_reg = Register::NoReg as u32;
+    let no_reg = Register::Null as u32;
 
     let instruction = Instruction::ShiftLeft(RTypeArgs::new(
         Some(Register::Rg0),

compiler/Cargo.toml

@@ -7,3 +7,4 @@ authors.workspace = true
 [dependencies]
 chrono = "0.4.43"
 common = { path = "../common" }
+uuid = { version = "1.20.0", features = ["v4"] }

compiler/src/backend/dsa/instruction.rs

@@ -0,0 +1,797 @@
+use std::fmt;
+
+use crate::backend::dsa::registers::Register;
+
+pub struct InsBlock {
+    instructions: Vec<Instruction>,
+}
+
+impl InsBlock {
+    pub fn new() -> Self {
+        Self {
+            instructions: vec![],
+        }
+    }
+
+    pub fn insert(&mut self, index: usize, instr: Instruction) {
+        self.instructions.insert(index, instr);
+    }
+
+    pub fn push(&mut self, instr: Instruction) {
+        self.instructions.push(instr);
+    }
+
+    pub fn append(&mut self, mut other: Self) {
+        self.instructions.append(&mut other.instructions);
+    }
+
+    pub fn extend(&mut self, instrs: impl IntoIterator<Item = Instruction>) {
+        self.instructions.extend(instrs);
+    }
+
+    pub fn is_empty(&self) -> bool {
+        self.instructions.is_empty()
+    }
+
+    pub fn len(&self) -> usize {
+        self.instructions.len()
+    }
+
+    pub fn iter(&self) -> impl Iterator<Item = &Instruction> {
+        self.instructions.iter()
+    }
+}
+
+impl From<Vec<Instruction>> for InsBlock {
+    fn from(instructions: Vec<Instruction>) -> Self {
+        Self { instructions }
+    }
+}
+
+impl From<Instruction> for InsBlock {
+    fn from(instr: Instruction) -> Self {
+        Self {
+            instructions: vec![instr],
+        }
+    }
+}
+
+#[derive(Debug, Clone)]
+pub enum Instruction {
+    // Labels and comments
+    Label(Label),
+    Comment {
+        text: String,
+        top_level: bool,
+    },
+    Newline,
+
+    // Data Directives
+    Db {
+        label: String,
+        data: Vec<u8>,
+    },
+    Dh {
+        label: String,
+        data: Vec<u16>,
+    },
+    Dw {
+        label: String,
+        data: Vec<u32>,
+    },
+    DString {
+        // alias for db.
+        label: String,
+        data: String,
+    },
+
+    Resx {
+        label: String,
+        size: u32,
+    },
+
+    // Include
+    Include {
+        name: String,
+        path: String,
+    },
+
+    // Data movement
+    Mov {
+        src: Register,
+        dest: Register,
+    },
+    Movs {
+        src: Register,
+        dest: Register,
+    },
+
+    // Memory operations
+    Ldb {
+        src: MemOperand,
+        dest: Register,
+    },
+    Ldh {
+        src: MemOperand,
+        dest: Register,
+    },
+    Ldw {
+        src: MemOperand,
+        dest: Register,
+    },
+    Stb {
+        src: Register,
+        dest: MemOperand,
+    },
+    Sth {
+        src: Register,
+        dest: MemOperand,
+    },
+    Stw {
+        src: Register,
+        dest: MemOperand,
+    },
+
+    // Immediate loads
+    Lli {
+        imm: Imm,
+        dest: Register,
+    },
+    Lui {
+        imm: Imm,
+        dest: Register,
+    },
+    Lwi {
+        imm: Imm,
+        dest: Register,
+    },
+    LwiLabel {
+        label: String,
+        dest: Register,
+    },
+
+    // Arithmetic
+    Add {
+        src1: Register,
+        src2: Register,
+        dest: Register,
+    },
+    Sub {
+        src1: Register,
+        src2: Register,
+        dest: Register,
+    },
+    IAdd {
+        src: Register,
+        imm: Imm,
+        dest: Option<Register>,
+    },
+    ISub {
+        src: Register,
+        imm: Imm,
+        dest: Option<Register>,
+    },
+    Inc {
+        reg: Register,
+    },
+    Dec {
+        reg: Register,
+    },
+
+    // Bitwise
+    And {
+        src1: Register,
+        src2: Register,
+        dest: Register,
+    },
+    Or {
+        src1: Register,
+        src2: Register,
+        dest: Register,
+    },
+    Xor {
+        src1: Register,
+        src2: Register,
+        dest: Register,
+    },
+    Not {
+        src: Register,
+        dest: Register,
+    },
+    Nand {
+        src1: Register,
+        src2: Register,
+        dest: Register,
+    },
+    Nor {
+        src1: Register,
+        src2: Register,
+        dest: Register,
+    },
+    Xnor {
+        src1: Register,
+        src2: Register,
+        dest: Register,
+    },
+
+    // Shifts
+    Shl {
+        src1: Register,
+        r_shamt: Register,
+        i_shamt: u16,
+        dest: Register,
+    },
+    Shr {
+        src1: Register,
+        r_shamt: Register,
+        i_shamt: u16,
+        dest: Register,
+    },
+
+    // Comparison
+    Cmp {
+        reg1: Register,
+        reg2: Register,
+    },
+
+    // Jumps
+    Jmp {
+        target: Label,
+    },
+    Jeq {
+        target: Label,
+    },
+    Jne {
+        target: Label,
+    },
+    Jgt {
+        target: Label,
+    },
+    Jge {
+        target: Label,
+    },
+    Jlt {
+        target: Label,
+    },
+    Jle {
+        target: Label,
+    },
+
+    // Stack
+    Push {
+        reg: Register,
+    },
+    Pop {
+        reg: Register,
+    },
+
+    // Function calls
+    Call {
+        target: String,
+    }, // namespace::function
+    Return,
+
+    // System
+    Hlt,
+    Nop,
+    Int {
+        code: u8,
+    },
+}
+
+pub enum DataDirective {
+    U8(Vec<u8>),
+    U16(Vec<u16>),
+    U32(Vec<u32>),
+    String(String),
+    Char(char),
+}
+
+impl fmt::Display for Instruction {
+    fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
+        match self {
+            Self::Label(l) => write!(f, "{}:", l),
+            Self::Newline => write!(f, ""), /* empty string as newlines are inserted */
+            // automatically.
+            Self::Comment { text, top_level } => write!(
+                f,
+                "{}",
+                text.lines()
+                    .map(|line| format!(
+                        "{}// {}",
+                        if *top_level { "" } else { "    " },
+                        line.trim(),
+                    ))
+                    .collect::<Vec<String>>()
+                    .join("\n")
+            ),
+
+            Self::Include { name, path } => write!(f, "include {name}: \"{}\"", path),
+
+            Self::Db { label, data } => write!(
+                f,
+                "db {}: {}",
+                label,
+                data.iter()
+                    .map(|&b| format!("{:#04X}", b))
+                    .collect::<Vec<String>>()
+                    .join(", ")
+            ),
+            Self::Dh { label, data } => write!(
+                f,
+                "dh {}: {}",
+                label,
+                data.iter()
+                    .map(|&b| format!("{:#06X}", b))
+                    .collect::<Vec<String>>()
+                    .join(", ")
+            ),
+            Self::Dw { label, data } => write!(
+                f,
+                "dw {}: {}",
+                label,
+                data.iter()
+                    .map(|&b| format!("{:#08X}", b))
+                    .collect::<Vec<String>>()
+                    .join(", ")
+            ),
+            Self::DString { label, data } => write!(f, "db {}: \"{}\"", label, data),
+
+            Self::Resx { label, size } => write!(f, "resx {}: {}", label, size),
+
+            Self::Mov { src, dest } => write!(f, "    mov {}, {}", src, dest),
+            Self::Movs { src, dest } => write!(f, "    movs {}, {}", src, dest),
+
+            Self::Ldb { src: addr, dest } => {
+                let (reg, offset) = reg_and_offset(addr);
+                write!(f, "    ldb {}, {}, {}", reg, dest, offset)
+            }
+            Self::Ldh { src: addr, dest } => {
+                let (reg, offset) = reg_and_offset(addr);
+                write!(f, "    ldh {}, {}, {}", reg, dest, offset)
+            }
+            Self::Ldw { src, dest } => {
+                let (reg, offset) = reg_and_offset(src);
+                write!(f, "    ldw {}, {}, {}", reg, dest, offset)
+            }
+            // Self::Ldbs { addr, dest } => {
+            //     write!(f, "    ldbs {}, {}", format_mem_operand(addr), dest)
+            // }
+            // Self::Ldhs { addr, dest } => {
+            //     write!(f, "    ldhs {}, {}", format_mem_operand(addr), dest)
+            // }
+            // Self::Ldws { addr, dest } => {
+            //     write!(f, "    ldws {}, {}", format_mem_operand(addr), dest)
+            // }
+            Self::Stb { src, dest: addr } => {
+                let (reg, offset) = reg_and_offset(addr);
+                write!(f, "    stb {}, {}, {}", src, reg, offset)
+            }
+            Self::Sth { src, dest: addr } => {
+                let (reg, offset) = reg_and_offset(addr);
+                write!(f, "    sth {}, {}, {}", src, reg, offset)
+            }
+            Self::Stw { src, dest: addr } => {
+                let (reg, offset) = reg_and_offset(addr);
+                write!(f, "    stw {}, {}, {}", src, reg, offset)
+            }
+
+            Self::Lli { imm, dest } => write!(f, "    lli {}, {}", imm, dest),
+            Self::Lui { imm, dest } => write!(f, "    lui {}, {}", imm, dest),
+            Self::Lwi { imm, dest } => write!(f, "    lwi {}, {}", imm, dest),
+            Self::LwiLabel { label, dest } => write!(f, "    lwi {}, {}", label, dest),
+
+            // arithmetic
+            Self::Add { src1, src2, dest } => {
+                write!(f, "    add {}, {}, {}", src1, src2, dest)
+            }
+            Self::Sub { src1, src2, dest } => {
+                write!(f, "    sub {}, {}, {}", src1, src2, dest)
+            }
+            Self::And { src1, src2, dest } => {
+                write!(f, "    and {}, {}, {}", src1, src2, dest)
+            }
+            Self::Or { src1, src2, dest } => {
+                write!(f, "    or {}, {}, {}", src1, src2, dest)
+            }
+            Self::Nand { src1, src2, dest } => {
+                write!(f, "    nand {}, {}, {}", src1, src2, dest)
+            }
+            Self::Xor { src1, src2, dest } => {
+                write!(f, "    xor {}, {}, {}", src1, src2, dest)
+            }
+            Self::Nor { src1, src2, dest } => {
+                write!(f, "    nor {}, {}, {}", src1, src2, dest)
+            }
+            Self::Not { src, dest } => {
+                write!(f, "    not {} {}", src, dest)
+            }
+            Self::Xnor { src1, src2, dest } => {
+                write!(f, "    xnor {}, {}, {}", src1, src2, dest)
+            }
+            Self::IAdd { src, imm, dest } => {
+                if let Some(d) = dest {
+                    write!(f, "    addi {}, {}, {}", src, imm, d)
+                } else {
+                    write!(f, "    addi {}, {}", src, imm)
+                }
+            }
+            Self::ISub { src, imm, dest } => {
+                if let Some(d) = dest {
+                    write!(f, "    subi {}, {}, {}", src, imm, d)
+                } else {
+                    write!(f, "    subi {}, {}", src, imm)
+                }
+            }
+
+            // shift instructions
+            Self::Shl {
+                src1,
+                r_shamt,
+                i_shamt,
+                dest,
+            } => {
+                write!(f, "    shl {}, {}, {}, {}", src1, r_shamt, i_shamt, dest)
+            }
+            Self::Shr {
+                src1,
+                r_shamt,
+                i_shamt,
+                dest,
+            } => {
+                write!(f, "    shl {}, {}, {}, {}", src1, r_shamt, i_shamt, dest)
+            }
+
+            // increment instructions
+            Self::Inc { reg } => write!(f, "    inc {}", reg),
+            Self::Dec { reg } => write!(f, "    dec {}", reg),
+
+            Self::Cmp { reg1, reg2 } => write!(f, "    cmp {}, {}", reg1, reg2),
+
+            // jump instructions
+            Self::Jmp { target } => write!(f, "    jmp {}", target),
+            Self::Jeq { target } => write!(f, "    jeq {}", target),
+            Self::Jne { target } => write!(f, "    jne {}", target),
+            Self::Jgt { target } => write!(f, "    jgt {}", target),
+            Self::Jge { target } => write!(f, "    jge {}", target),
+            Self::Jlt { target } => write!(f, "    jlt {}", target),
+            Self::Jle { target } => write!(f, "    jle {}", target),
+
+            // stack pseudoinstructions
+            Self::Push { reg } => write!(f, "    push {}", reg),
+            Self::Pop { reg } => write!(f, "    pop {}", reg),
+
+            // call & return pseudoinstructions
+            Self::Call { target } => write!(f, "    call {}", target),
+            Self::Return => write!(f, "    return"),
+
+            // misc instructions
+            Self::Int { code } => write!(f, "    int {}", code),
+            Self::Hlt => write!(f, "    hlt"),
+            Self::Nop => write!(f, "    nop"),
+        }
+    }
+}
+
+impl Instruction {
+    // data directives
+    pub fn db_string(label: impl Into<String>, data: impl Into<String>) -> Self {
+        Self::DString {
+            label: label.into(),
+            data: data.into(),
+        }
+    }
+
+    pub fn db_word(label: impl Into<String>, data: u32) -> Self {
+        Self::Dw {
+            label: label.into(),
+            data: vec![data],
+        }
+    }
+
+    pub fn db_bytes(label: impl Into<String>, data: &[u8]) -> Self {
+        Self::Db {
+            label: label.into(),
+            data: data.to_vec(),
+        }
+    }
+
+    // Movement
+    pub fn mov<R1, R2>(src: R1, dest: R2) -> Self
+    where
+        R1: Into<Register>,
+        R2: Into<Register>,
+    {
+        Self::Mov {
+            src: src.into(),
+            dest: dest.into(),
+        }
+    }
+
+    // Memory loads
+    pub fn ldw_reg<R>(base: R, dest: Register) -> Self
+    where
+        R: Into<Register>,
+    {
+        Self::Ldw {
+            src: MemOperand::RegIndirect(base.into()),
+            dest,
+        }
+    }
+
+    pub fn ldw_reg_offset<R>(base: R, dest: Register, offset: i32) -> Self
+    where
+        R: Into<Register>,
+    {
+        Self::Ldw {
+            src: MemOperand::RegOffset(base.into(), offset),
+            dest,
+        }
+    }
+
+    pub fn ldw_label(label: impl Into<Label>, dest: Register) -> Self {
+        Self::Ldw {
+            src: MemOperand::Label(label.into()),
+            dest,
+        }
+    }
+
+    // Memory stores
+    pub fn stw_reg<R>(src: Register, base: R) -> Self
+    where
+        R: Into<Register>,
+    {
+        Self::Stw {
+            src,
+            dest: MemOperand::RegIndirect(base.into()),
+        }
+    }
+
+    pub fn stw_reg_offset<R>(src: Register, base: R, offset: i32) -> Self
+    where
+        R: Into<Register>,
+    {
+        Self::Stw {
+            src,
+            dest: MemOperand::RegOffset(base.into(), offset),
+        }
+    }
+
+    pub fn stw_label(src: Register, label: impl Into<Label>) -> Self {
+        Self::Stw {
+            src,
+            dest: MemOperand::Label(label.into()),
+        }
+    }
+
+    // Arithmetic
+    pub fn add(src1: Register, src2: Register, dest: Register) -> Self {
+        Self::Add { src1, src2, dest }
+    }
+
+    pub fn sub(src1: Register, src2: Register, dest: Register) -> Self {
+        Self::Sub { src1, src2, dest }
+    }
+
+    pub fn and(src1: Register, src2: Register, dest: Register) -> Self {
+        Self::And { src1, src2, dest }
+    }
+
+    pub fn or(src1: Register, src2: Register, dest: Register) -> Self {
+        Self::Or { src1, src2, dest }
+    }
+
+    pub fn xor(src1: Register, src2: Register, dest: Register) -> Self {
+        Self::Xor { src1, src2, dest }
+    }
+
+    pub fn not(src: Register, dest: Register) -> Self {
+        Self::Not { src, dest }
+    }
+
+    pub fn shl(src1: Register, r_shamt: Register, i_shamt: u16, dest: Register) -> Self {
+        Self::Shl {
+            src1,
+            r_shamt,
+            i_shamt,
+            dest,
+        }
+    }
+
+    pub fn shr(src1: Register, r_shamt: Register, i_shamt: u16, dest: Register) -> Self {
+        Self::Shr {
+            src1,
+            r_shamt,
+            i_shamt,
+            dest,
+        }
+    }
+
+    pub fn iadd(src: Register, value: i64) -> Self {
+        let imm = Imm(value.unsigned_abs() as u32);
+
+        if value < 0 {
+            Self::ISub {
+                src,
+                imm,
+                dest: None,
+            }
+        } else {
+            Self::IAdd {
+                src,
+                imm,
+                dest: None,
+            }
+        }
+    }
+
+    pub fn iadd_dest(src: Register, value: i32, dest: Register) -> Self {
+        let imm = Imm(value.unsigned_abs());
+
+        if value < 0 {
+            Self::ISub {
+                src,
+                imm,
+                dest: Some(dest),
+            }
+        } else {
+            Self::IAdd {
+                src,
+                imm,
+                dest: Some(dest),
+            }
+        }
+    }
+
+    pub fn inc(reg: Register) -> Self {
+        Self::Inc { reg }
+    }
+
+    pub fn dec(reg: Register) -> Self {
+        Self::Dec { reg }
+    }
+
+    // Immediate loads
+    pub fn lwi(value: u32, dest: Register) -> Self {
+        if value > 0xFFFF {
+            Self::Lwi {
+                imm: Imm(value),
+                dest,
+            }
+        } else {
+            Self::Lli {
+                imm: Imm(value),
+                dest,
+            }
+        }
+    }
+
+    pub fn lwi_label(label: impl Into<String>, dest: Register) -> Self {
+        Self::LwiLabel {
+            label: label.into(),
+            dest,
+        }
+    }
+
+    // Control flow
+    pub fn label(name: impl Into<String>) -> Self {
+        Self::Label(Label(name.into()))
+    }
+
+    pub fn jmp(target: impl Into<Label>) -> Self {
+        Self::Jmp {
+            target: target.into(),
+        }
+    }
+
+    pub fn jeq(target: impl Into<Label>) -> Self {
+        Self::Jeq {
+            target: target.into(),
+        }
+    }
+
+    pub fn cmp(reg1: Register, reg2: Register) -> Self {
+        Self::Cmp { reg1, reg2 }
+    }
+
+    // Stack
+    pub fn push(reg: Register) -> Self {
+        Self::Push { reg }
+    }
+
+    pub fn pop(reg: Register) -> Self {
+        Self::Pop { reg }
+    }
+
+    // Functions
+    pub fn call(target: impl Into<String>) -> Self {
+        Self::Call {
+            target: target.into(),
+        }
+    }
+
+    pub fn int(code: u8) -> Self {
+        Self::Int { code }
+    }
+
+    pub fn ret() -> Self {
+        Self::Return
+    }
+
+    // Utilities
+    pub fn comment(text: impl Into<String>) -> Self {
+        Self::Comment {
+            text: text.into(),
+            top_level: false,
+        }
+    }
+
+    pub fn global_comment(text: impl Into<String>) -> Self {
+        Self::Comment {
+            text: text.into(),
+            top_level: true,
+        }
+    }
+
+    pub fn include(name: impl Into<String>, path: impl Into<String>) -> Self {
+        Self::Include {
+            name: name.into(),
+            path: path.into(),
+        }
+    }
+}
+
+// Convenience trait for Label conversion
+impl From<String> for Label {
+    fn from(s: String) -> Self {
+        Label(s)
+    }
+}
+
+impl From<&str> for Label {
+    fn from(s: &str) -> Self {
+        Label(s.to_string())
+    }
+}
+
+fn reg_and_offset(op: &MemOperand) -> (String, i32) {
+    match op {
+        MemOperand::RegIndirect(reg) => (reg.to_string(), 0),
+        MemOperand::RegOffset(reg, offset) => (reg.to_string(), *offset),
+        MemOperand::Label(label) => (label.to_string(), 0),
+        MemOperand::LabelOffset(label, offset) => (label.to_string(), *offset),
+    }
+}
+
+/// Memory operand for loads/stores
+#[derive(Debug, Clone)]
+pub enum MemOperand {
+    /// Register indirect: [reg]
+    RegIndirect(Register),
+    /// Register with offset: [reg + offset]
+    RegOffset(Register, i32),
+    /// Label: [label]
+    Label(Label),
+    /// Label with offset: [label + offset]
+    LabelOffset(Label, i32),
+}
+
+/// Immediate value (16-bit or 32-bit)
+#[derive(Debug, Clone, Copy, PartialEq, Eq)]
+pub struct Imm(pub u32);
+
+impl fmt::Display for Imm {
+    fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
+        write!(f, "{}", self.0)
+    }
+}
+
+/// Label reference
+#[derive(Debug, Clone, PartialEq, Eq)]
+pub struct Label(pub String);
+
+impl fmt::Display for Label {
+    fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
+        write!(f, "{}", self.0)
+    }
+}

compiler/src/backend/dsa/mod.rs

@@ -1,7 +1,10 @@
 use crate::model::{CompilerError, Program};
 
 mod codegen;
+mod instruction;
 mod registers;
+mod scope;
+mod variable;
 
 pub fn generate_code(ast: &Program) -> Result<String, CompilerError> {
     let mut codegen = codegen::CodeGenerator::new(ast.clone());

compiler/src/backend/dsa/registers.rs

@@ -1,83 +1,116 @@
-use std::collections::HashMap;
+use std::{collections::HashMap, fmt};
 
-use crate::model::CompilerError;
+use crate::{
+    backend::dsa::instruction::{InsBlock, Instruction},
+    model::CompilerError,
+};
 
 /// Register allocator for DSA assembly generation
 /// Manages general-purpose registers (rg0-rgf) and handles stack spilling
 pub struct RegisterAllocator {
     /// Available general-purpose registers
-    available_registers: Vec<String>,
-
     /// Maps variable names to their current location (register or stack offset)
     variable_locations: HashMap<String, Location>,
 
     /// Maps registers to the variables they currently hold
-    register_contents: HashMap<String, String>,
+    register_contents: HashMap<Register, String>,
 
     /// Current stack offset for local variables (relative to bpr)
     /// Starts at -4 (going downward from base pointer)
     stack_offset: i32,
 
     /// Track which registers are currently in use
-    in_use: HashMap<String, bool>,
+    in_use: Vec<(Register, bool)>,
 }
 
 #[derive(Debug, Clone)]
-pub enum Location {
-    Register(String),
-    Stack(i32), // offset from bpr
+pub struct Location {
+    register: Option<Register>,
+    stack: Option<i32>,
+}
+
+impl Location {
+    pub fn stack(offset: i32) -> Self {
+        Location {
+            register: None,
+            stack: Some(offset),
+        }
+    }
+
+    pub fn register(register: Register) -> Self {
+        Location {
+            register: Some(register),
+            stack: None,
+        }
+    }
 }
 
 impl RegisterAllocator {
     pub fn new() -> Self {
         // Initialize with available GP registers (rg0-rgf = 16 registers)
-        let registers = vec![
-            "rg0", "rg1", "rg2", "rg3", "rg4", "rg5", "rg6", "rg7", "rg8", "rg9", "rga",
-            "rgb", "rgc", "rgd", "rge", "rgf",
+        let in_use = vec![
+            Register::Rg0,
+            Register::Rg1,
+            Register::Rg2,
+            Register::Rg3,
+            Register::Rg4,
+            Register::Rg5,
+            Register::Rg6,
+            Register::Rg7,
+            Register::Rg8,
+            Register::Rg9,
+            Register::Rga,
+            Register::Rgb,
+            Register::Rgc,
+            Register::Rgd,
+            Register::Rge,
+            Register::Rgf,
         ]
-        .into_iter()
-        .map(String::from)
+        .iter()
+        .map(|&reg| (reg, false))
         .collect();
 
         RegisterAllocator {
-            available_registers: registers,
+            // available_registers: registers,
             variable_locations: HashMap::new(),
             register_contents: HashMap::new(),
             stack_offset: -4, // Start at -4 (first local below saved bpr)
-            in_use: HashMap::new(),
+            in_use,
         }
     }
 
     /// Allocate a temporary register for expression evaluation
     /// Returns the register name and optionally assembly code to save it
-    pub fn alloc_temp(&mut self) -> Result<(String, Vec<String>), CompilerError> {
-        let mut code = Vec::new();
-
+    pub fn alloc_temp(&mut self) -> Result<(Register, InsBlock), CompilerError> {
         // Try to find an unused register
-        for reg in &self.available_registers {
-            if !self.in_use.get(reg).unwrap_or(&false) {
-                self.in_use.insert(reg.clone(), true);
-                return Ok((reg.clone(), code));
-            }
+
+        // println!("finding! {:#?}", self.in_use);
+
+        if let Some(reg) = self.find_free_register() {
+            self.in_use[reg as usize].1 = true;
+            return Ok((reg, InsBlock::new()));
         }
 
         // All registers in use - need to spill one
         // Choose the first register with a variable we can spill
         // Find a register to spill
-        let reg_to_spill = self
-            .available_registers
-            .iter()
-            .find(|reg| self.register_contents.contains_key(*reg))
-            .cloned();
 
-        if let Some(reg) = reg_to_spill {
-            // Spill this variable to stack
-            let spill_code = self.spill_register(&reg)?;
-            code.extend(spill_code);
+        // let reg_to_spill = self
+        //     .available_registers
+        //     .iter()
+        //     .find(|reg| self.register_contents.contains_key(*reg))
+        //     .cloned();
 
-            self.in_use.insert(reg.clone(), true);
-            return Ok((reg, code));
-        }
+        // if let Some(reg) = reg_to_spill {
+        //     // Spill this variable to stack
+        //     let spill_code = self.spill_register(&reg)?;
+        //     code.extend(spill_code);
+
+        //     self.in_use.insert(reg.clone(), true);
+        //     return Ok((reg, code));
+        // }
+
+        todo!("an efficient stack spilling algorithm. needs scope awareness.");
 
         Err(CompilerError::Generic(
             "All registers are used up yet there are no variables to spill to the stack"
@@ -85,19 +118,39 @@ impl RegisterAllocator {
         ))
     }
 
+    // fn set_in_use(&mut self, reg: Register, in_use: bool) {
+    //     self.in_use[reg as usize].1 = in_use;
+    // }
+
     /// Free a temporary register after use
     /// NOTE: This will NOT free registers that contain variables!
     /// Variables persist throughout their scope and must not be freed
-    pub fn free_temp(&mut self, reg: &str) {
+    pub fn free_temp(&mut self, reg: Register) {
         // Check if this register contains a variable
-        if self.register_contents.contains_key(reg) {
+        if self.register_contents.contains_key(&reg) {
             // This register holds a variable - don't free it!
             // Variables are only freed when they go out of scope via free_var()
             return;
         }
 
         // This is a true temporary - safe to free
-        self.in_use.insert(reg.to_string(), false);
+        if !matches!(reg, Register::Zero | Register::Null) {
+            self.in_use[reg as usize].1 = false;
+        }
+    }
+
+    pub fn free_var(&mut self, var: &str) {
+        // Check if this variable is in a register
+        if let Some(location) = self.variable_locations.get(var).cloned() {
+            if let Some(reg) = location.register
+                && !matches!(reg, Register::Zero | Register::Null)
+            {
+                self.register_contents.remove(&reg);
+                self.in_use[reg as usize].1 = false;
+            }
+
+            self.variable_locations.remove(var);
+        }
     }
 
     /// Allocate a register for a named variable
@@ -105,43 +158,45 @@ impl RegisterAllocator {
     pub fn alloc_var(
         &mut self,
         var_name: &str,
-    ) -> Result<(String, Vec<String>), CompilerError> {
-        if let Some(location) = self.variable_locations.get(var_name).cloned() {
-            match location {
-                Location::Register(reg) => {
-                    return Ok((reg.clone(), Vec::new()));
-                }
-                Location::Stack(offset) => {
-                    // Variable was pushed, need to calculate actual position
-                    let (reg, mut code) = self.alloc_temp()?;
+    ) -> Result<(Register, InsBlock), CompilerError> {
+        if let Some(mut location) = self.variable_locations.get(var_name).cloned() {
+            // if the var is in a register we can use it already.
+            if let Some(reg) = location.register {
+                return Ok((reg, InsBlock::new()));
+            }
 
-                    // Load from bpr + offset (offset is negative)
-                    code.push(format!("\tsubi bpr {} {}", -(offset + 4), reg));
-                    code.push(format!(
-                        "\tldw {}, {} // bpr{}: {}",
-                        reg,
-                        reg,
-                        offset - 4,
-                        var_name
-                    ));
+            // if the variable is on the stack only, we need to get it in a register.
+            if let Some(offset) = location.stack {
+                // Variable was pushed, need to calculate actual position and update its
+                // location.
+                let (reg, mut code) = self.alloc_temp()?;
 
-                    // Update location to register
-                    self.variable_locations
-                        .insert(var_name.to_string(), Location::Register(reg.clone()));
-                    self.register_contents
-                        .insert(reg.clone(), var_name.to_string());
+                // acknowledge var is now in a reg as well.
+                location.register = Some(reg);
 
-                    return Ok((reg, code));
-                }
+                // Load from bpr + offset (offset is negative)
+                // code.push(format!("\tsubi bpr {} {}", -(offset + 4), reg));
+
+                code.push(Instruction::ldw_reg_offset(
+                    Register::Spr,
+                    reg,
+                    offset - self.stack_offset,
+                ));
+
+                // Update location to register
+                self.variable_locations
+                    .insert(var_name.to_string(), location);
+                self.register_contents.insert(reg, var_name.to_string());
+
+                return Ok((reg, code));
             }
         }
 
         // Variable doesn't have a location yet, allocate a new register
         let (reg, code) = self.alloc_temp()?;
         self.variable_locations
-            .insert(var_name.to_string(), Location::Register(reg.clone()));
-        self.register_contents
-            .insert(reg.clone(), var_name.to_string());
+            .insert(var_name.to_string(), Location::register(reg));
+        self.register_contents.insert(reg, var_name.to_string());
 
         Ok((reg, code))
     }
@@ -156,122 +211,178 @@ impl RegisterAllocator {
     pub fn load_var(
         &mut self,
         var_name: &str,
-    ) -> Result<(String, Vec<String>), CompilerError> {
+    ) -> Result<(Register, InsBlock), CompilerError> {
         self.alloc_var(var_name)
     }
 
     /// Store a value from a register into a variable
     /// Updates tracking and returns any necessary assembly code
-    pub fn store_var(&mut self, var_name: &str, source_reg: &str) -> Vec<String> {
-        let mut code = Vec::new();
+    pub fn store_var(&mut self, var_name: &str, source_reg: &Register) -> InsBlock {
+        let mut block = InsBlock::new();
 
         // Check if variable already has a location
         if let Some(location) = self.variable_locations.get(var_name) {
-            match location {
-                Location::Register(dest_reg) => {
-                    if dest_reg != source_reg {
-                        code.push(format!(
-                            "\tmov {}, {} // var {}",
-                            source_reg, dest_reg, var_name
-                        ));
-                    }
-                }
-                Location::Stack(offset) => {
-                    code.push(format!(
-                        "\tstw {}, bpr, {} // var {}",
-                        source_reg, offset, var_name
-                    ));
+            // if the variable exists in a register we write to that.
+            match location.register {
+                Some(reg) if reg == *source_reg => {
+                    block.push(Instruction::mov(*source_reg, reg));
+                    return block;
                 }
+                _ => (),
             }
-        } else {
-            // Variable doesn't exist yet, we can just use the same reg.
 
-            // self.variable_locations.insert(
-            //     var_name.to_string(),
-            //     Location::Register(source_reg.to_string()),
-            // );
-            // self.register_contents
-            //     .insert(source_reg.to_string(), var_name.to_string());
-            // self.in_use.insert(source_reg.to_string(), true);
-
-            let source_reg = source_reg.to_string();
-
-            // if we can avoid a move, absolutely do that.
-            if self.available_registers.contains(&source_reg) {
-                self.variable_locations
-                    .insert(var_name.to_string(), Location::Register(source_reg.clone()));
-                self.register_contents
-                    .insert(source_reg.clone(), var_name.to_string());
-                self.in_use.insert(source_reg, true);
-            } else if let Some(free_reg) = self.find_free_register() {
-                code.push(format!("\tmov {}, {}", source_reg, free_reg));
-                self.variable_locations
-                    .insert(var_name.to_string(), Location::Register(free_reg.clone()));
-                self.register_contents
-                    .insert(free_reg.clone(), var_name.to_string());
-                self.in_use.insert(free_reg, true);
-            } else {
-                // No free registers - allocate on stack
-                // code.push(format!("\tstw {}, bpr, {}", source_reg, self.stack_offset));
-                // self.variable_locations
-                //     .insert(var_name.to_string(), Location::Stack(self.stack_offset));
-                // self.stack_offset -= 4; // Move to next stack slot
-                //
-                todo!(
-                    "we should spill other registers and keep this variable on the stack as it's more recent!"
-                );
+            // if the variable exists on the stack but not a register we write here.
+            if let Some(offset) = location.stack {
+                block.push(Instruction::stw_reg_offset(
+                    *source_reg,
+                    Register::Spr,
+                    offset - self.stack_offset,
+                ));
+                return block;
             }
         }
 
-        code
+        // Variable doesn't exist yet, we can just use the same reg.
+        // if we can avoid a move, absolutely do that.
+
+        // if this is true then there's no permanent variable here so it's safe to use.
+        if !self.register_contents.contains_key(source_reg) {
+            self.variable_locations
+                .insert(var_name.to_string(), Location::register(*source_reg));
+            self.register_contents
+                .insert(*source_reg, var_name.to_string());
+            self.in_use[*source_reg as usize].1 = true;
+
+            return block;
+        }
+
+        // if current register isn't free, (eg is another variable) we assign somewhere
+        // else.
+        if let Some(free_reg) = self.find_free_register() {
+            self.variable_locations
+                .insert(var_name.to_string(), Location::register(free_reg));
+            self.register_contents
+                .insert(free_reg, var_name.to_string());
+            self.in_use[free_reg as usize].1 = true;
+
+            block.push(Instruction::mov(*source_reg, free_reg));
+            return block;
+        }
+
+        // No free registers - allocate on stack
+        // code.push(format!("\tstw {}, bpr, {}", source_reg, self.stack_offset));
+        // self.variable_locations
+        //     .insert(var_name.to_string(), Location::Stack(self.stack_offset));
+        // self.stack_offset -= 4; // Move to next stack slot
+        //
+        todo!("an efficient stack spilling algorithm. needs scope awareness.");
     }
 
-    /// Spill a register to the stack
-    /// Returns assembly code to perform the spill
-    pub fn spill_register(&mut self, reg: &str) -> Result<Vec<String>, CompilerError> {
-        let mut code = Vec::new();
+    /// spill a register to the stack (WITHOUT FREEING)
+    /// DO NOT USE this if it's for a pointer!!!!
+    pub fn _spill_register(&mut self, reg: &Register) -> Result<InsBlock, CompilerError> {
+        let mut code = InsBlock::new();
 
-        if let Some(var_name) = self.register_contents.get(reg).cloned() {
-            // PUSH register to stack (spr decrements automatically)
-            code.push(format!(
-                "\tpush {} // bpr{}: {}",
-                reg, self.stack_offset, var_name
-            ));
+        // check if the variable is declared.
+        if let Some(var_name) = self.register_contents.get(reg).cloned()
+            && let Some(location) = self.variable_locations.get_mut(&var_name)
+        {
+            // check if var is on the stack
+            if let Some(offset) = location.stack {
+                code.push(Instruction::stw_reg_offset(
+                    *reg,
+                    Register::Spr,
+                    offset - self.stack_offset,
+                ));
+                return Ok(code);
+            }
 
             // Track that we pushed one word
             self.stack_offset -= 4;
 
+            // if the variable is not on the stack:
+            // push register to stack (spr decrements automatically)
+            let offset = self.stack_offset;
+            code.push(Instruction::push(*reg));
+
             // Update variable location - it's now at current spr
             // Note: We track offset from bpr for consistency
-            self.variable_locations
-                .insert(var_name.clone(), Location::Stack(self.stack_offset));
+            location.stack = Some(offset);
 
-            // Remove from register tracking
-            self.register_contents.remove(reg);
+            Ok(code)
+        } else {
+            Err(CompilerError::Generic(format!(
+                "Register {} does not contain a variable to spill!",
+                reg
+            )))
         }
+    }
 
-        Ok(code)
+    /// free a register by spilling it to the stack.
+    /// Returns assembly code to perform the spill
+    pub fn free_register(
+        &mut self,
+        reg: &Register,
+    ) -> Result<(i32, Instruction), CompilerError> {
+        // check if the variable is declared.
+        if let Some(var_name) = self.register_contents.get(reg).cloned()
+            && let Some(location) = self.variable_locations.get_mut(&var_name)
+        {
+            // check if var name is on the stack
+            if let Some(offset) = location.stack {
+                // store current register value in stack location
+                let code = Instruction::stw_reg_offset(
+                    *reg,
+                    Register::Spr,
+                    offset - self.stack_offset,
+                );
+
+                // free the register.
+                location.register = None;
+                self.register_contents.remove(reg);
+                return Ok((offset, code));
+            }
+
+            // Track that we pushed one word
+            self.stack_offset -= 4;
+
+            let offset = self.stack_offset;
+            let code = Instruction::push(*reg);
+
+            // Update variable location
+            // Note: We track offset from bpr for consistency
+            location.stack = Some(offset);
+            location.register = None;
+            self.register_contents.remove(reg);
+
+            Ok((offset, code))
+        } else {
+            Err(CompilerError::Generic(format!(
+                "Register {} does not contain a variable to spill!",
+                reg
+            )))
+        }
     }
 
     /// Find a free register (not currently in use)
-    fn find_free_register(&self) -> Option<String> {
-        for reg in &self.available_registers {
-            if !self.in_use.get(reg).unwrap_or(&false) {
-                return Some(reg.clone());
-            }
-        }
-        None
+    fn find_free_register(&self) -> Option<Register> {
+        self.in_use
+            .iter()
+            .filter(|(_, in_use)| !*in_use)
+            .map(|(reg, _)| *reg)
+            .next()
     }
 
     /// Spill all registers to stack (useful before function calls)
-    pub fn _spill_all(&mut self) -> Vec<String> {
-        let mut code = Vec::new();
+    pub fn _spill_all(&mut self) -> InsBlock {
+        let mut code = InsBlock::new();
 
-        let regs_to_spill: Vec<String> = self.register_contents.keys().cloned().collect();
+        let regs_to_spill: Vec<Register> =
+            self.register_contents.keys().cloned().collect();
 
         for reg in regs_to_spill {
-            if let Ok(spill_code) = self.spill_register(&reg) {
-                code.extend(spill_code);
+            if let Ok(spill_code) = self.free_register(&reg) {
+                code.push(spill_code.1);
             }
         }
 
@@ -293,56 +404,157 @@ impl RegisterAllocator {
         self.variable_locations.clear();
         self.register_contents.clear();
         self.stack_offset = -4;
-        self.in_use.clear();
-    }
-
-    /// Mark a variable as dead (no longer needed)
-    /// Frees its register if it's in one
-    pub fn _free_var(&mut self, var_name: &str) {
-        if let Some(Location::Register(reg)) = self.variable_locations.get(var_name) {
-            let reg = reg.clone();
-            self.register_contents.remove(&reg);
-            self.in_use.insert(reg, false);
-        }
-        self.variable_locations.remove(var_name);
+        self.in_use = vec![
+            Register::Rg0,
+            Register::Rg1,
+            Register::Rg2,
+            Register::Rg3,
+            Register::Rg4,
+            Register::Rg5,
+            Register::Rg6,
+            Register::Rg7,
+            Register::Rg8,
+            Register::Rg9,
+            Register::Rga,
+            Register::Rgb,
+            Register::Rgc,
+            Register::Rgd,
+            Register::Rge,
+            Register::Rgf,
+        ]
+        .iter()
+        .map(|&reg| (reg, false))
+        .collect();
     }
 
     /// Get list of registers that contain variables and are in use
     /// These need to be saved before function calls
-    pub fn get_caller_saved_registers(&self) -> Vec<String> {
+    pub fn get_caller_saved_registers(&self) -> Vec<Register> {
         self.register_contents
             .iter()
-            .filter(|(reg, _)| *self.in_use.get(*reg).unwrap_or(&false))
-            .map(|(reg, _)| reg.clone())
+            .filter(|(reg, _)| {
+                self.in_use
+                    .get(**reg as usize)
+                    .unwrap_or(&(Register::Null, false))
+                    .1
+            })
+            .map(|(reg, _)| *reg)
             .collect()
     }
+}
 
-    /// Save caller-saved registers before a function call
-    /// Returns assembly code to save them
-    pub fn _save_caller_saved(&mut self) -> Vec<String> {
-        let mut code = Vec::new();
+#[derive(Debug, Copy, Clone, PartialEq, Eq, Hash)]
+pub enum Register {
+    // general purpose
+    Rg0 = 0,
+    Rg1 = 1,
+    Rg2 = 2,
+    Rg3 = 3,
+    Rg4 = 4,
+    Rg5 = 5,
+    Rg6 = 6,
+    Rg7 = 7,
+    Rg8 = 8,
+    Rg9 = 9,
+    Rga = 10,
+    Rgb = 11,
+    Rgc = 12,
+    Rgd = 13,
+    Rge = 14,
+    Rgf = 15,
 
-        // For simplicity, save all currently used registers
-        // In a more sophisticated compiler, you'd only save registers that are live
-        for (reg, _) in self.register_contents.clone() {
-            if *self.in_use.get(&reg).unwrap_or(&false) {
-                code.push(format!("\tpush {}", reg));
-            }
-        }
+    // special
+    Bpr,
+    Spr,
+    Ret,
+    Acc,
 
-        code
+    // read only
+    Pcx,
+    Zero,
+
+    // null
+    Null,
+}
+
+impl Register {
+    pub fn get_gp() -> [Register; 16] {
+        [
+            Register::Rg0,
+            Register::Rg1,
+            Register::Rg2,
+            Register::Rg3,
+            Register::Rg4,
+            Register::Rg5,
+            Register::Rg6,
+            Register::Rg7,
+            Register::Rg8,
+            Register::Rg9,
+            Register::Rga,
+            Register::Rgb,
+            Register::Rgc,
+            Register::Rgd,
+            Register::Rge,
+            Register::Rgf,
+        ]
     }
 
-    /// Restore caller-saved registers after a function call
-    /// Returns assembly code to restore them
-    pub fn _restore_caller_saved(&mut self, saved_regs: &[String]) -> Vec<String> {
-        let mut code = Vec::new();
+    pub fn is_gp(&self) -> bool {
+        (*self as u8) < 16
+    }
 
-        // Restore in reverse order (LIFO)
-        for reg in saved_regs.iter().rev() {
-            code.push(format!("\tpop {}", reg));
+    pub fn from_index(idx: usize) -> Register {
+        match idx {
+            0 => Register::Rg0,
+            1 => Register::Rg1,
+            2 => Register::Rg2,
+            3 => Register::Rg3,
+            4 => Register::Rg4,
+            5 => Register::Rg5,
+            6 => Register::Rg6,
+            7 => Register::Rg7,
+            8 => Register::Rg8,
+            9 => Register::Rg9,
+            10 => Register::Rga,
+            11 => Register::Rgb,
+            12 => Register::Rgc,
+            13 => Register::Rgd,
+            14 => Register::Rge,
+            15 => Register::Rgf,
+            _ => unreachable!("this function shouldn't ever be called with idx>15"),
+        }
+    }
+}
+
+impl fmt::Display for Register {
+    fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
+        match self {
+            Self::Rg0 => write!(f, "rg0"),
+            Self::Rg1 => write!(f, "rg1"),
+            Self::Rg2 => write!(f, "rg2"),
+            Self::Rg3 => write!(f, "rg3"),
+            Self::Rg4 => write!(f, "rg4"),
+            Self::Rg5 => write!(f, "rg5"),
+            Self::Rg6 => write!(f, "rg6"),
+            Self::Rg7 => write!(f, "rg7"),
+            Self::Rg8 => write!(f, "rg8"),
+            Self::Rg9 => write!(f, "rg9"),
+            Self::Rga => write!(f, "rga"),
+            Self::Rgb => write!(f, "rgb"),
+            Self::Rgc => write!(f, "rgc"),
+            Self::Rgd => write!(f, "rgd"),
+            Self::Rge => write!(f, "rge"),
+            Self::Rgf => write!(f, "rgf"),
+
+            Self::Acc => write!(f, "acc"),
+            Self::Ret => write!(f, "ret"),
+            Self::Bpr => write!(f, "bpr"),
+            Self::Spr => write!(f, "spr"),
+
+            Self::Zero => write!(f, "zero"),
+            Self::Pcx => write!(f, "pcx"),
+
+            Self::Null => write!(f, "null"),
         }
-
-        code
     }
 }

compiler/src/backend/dsa/scope.rs

@@ -0,0 +1,287 @@
+use std::{cell::RefCell, collections::HashMap, ops::Deref, rc::Rc};
+
+use uuid::Uuid;
+
+use crate::{
+    backend::dsa::{
+        instruction::{InsBlock, Instruction},
+        registers::{Register, RegisterAllocator},
+        variable::Variable,
+    },
+    model::CompilerError,
+};
+
+pub struct Allocator {
+    stack_offset: i32,
+    in_use: [(Register, bool); 16],
+}
+
+pub struct TempReg(Register);
+pub struct AssignedReg(Register);
+pub struct StackSlot(i32);
+
+impl Deref for TempReg {
+    type Target = Register;
+
+    fn deref(&self) -> &Self::Target {
+        &self.0
+    }
+}
+
+impl Deref for AssignedReg {
+    type Target = Register;
+
+    fn deref(&self) -> &Self::Target {
+        &self.0
+    }
+}
+
+impl Deref for StackSlot {
+    type Target = i32;
+
+    fn deref(&self) -> &Self::Target {
+        &self.0
+    }
+}
+
+impl Allocator {
+    pub fn new() -> Self {
+        let mut in_use = [(Register::Null, false); 16];
+        in_use.copy_from_slice(&Register::get_gp().map(|r| (r, false))[0..16]);
+
+        Self {
+            stack_offset: 0,
+            in_use,
+        }
+    }
+
+    pub fn get_stack_offset(&self) -> i32 {
+        self.stack_offset
+    }
+
+    pub fn destroy_scope(&mut self, scope: &mut Scope) {
+        self.stack_offset = scope.entry_stack_offset;
+
+        for var in scope.variables.drain() {
+            if let Some(assigned) = var.1.register {
+                self.free_assigned(&assigned);
+            }
+        }
+    }
+
+    // what we need:
+
+    // - create var in register from temporary register. free temp and use it.
+    //
+    // - create var on stack from struct/array literal. return stack offset to write to.
+    //
+    // - spill var from register to stack. return stack offset to write to.
+    //
+    // - read/write var from stack+offset into register to use while preserving the stack
+    //   slot.
+    //
+    // - read / write bytes from the stack+offset in a larger variable into a register.
+
+    pub fn read_var(&mut self, var: &mut Variable) -> Result<InsBlock, CompilerError> {
+        if let Some(slot) = &mut var.stack_slot {
+            if var.register.is_none() {
+                var.register = Some(self.allocate_var()?);
+            }
+
+            if let Some(reg) = &var.register {
+                return Ok(InsBlock::from(Instruction::ldw_reg_offset(
+                    **reg,
+                    Register::Spr,
+                    **slot - self.stack_offset,
+                )));
+            }
+
+            unreachable!()
+        }
+
+        Err(CompilerError::Generic(format!(
+            "Tried to write var {} to stack but var was not assigned a reg and/or stack slot",
+            var.name
+        )))
+    }
+
+    pub fn write_var(&mut self, var: &mut Variable) -> Result<InsBlock, CompilerError> {
+        if let Some(slot) = &var.stack_slot {
+            if let Some(reg) = &var.register {
+                return Ok(InsBlock::from(Instruction::stw_reg_offset(
+                    **reg,
+                    Register::Spr,
+                    **slot - self.stack_offset,
+                )));
+            }
+        }
+
+        Err(CompilerError::Generic(format!(
+            "Tried to write var {} to stack but var was not assigned a reg and/or stack slot",
+            var.name
+        )))
+    }
+
+    pub fn spill_var(&mut self, var: &mut Variable) -> Result<InsBlock, CompilerError> {
+        if let Some(slot) = &var.stack_slot {
+            let block = self.write_var(var)?;
+            if let Some(reg) = &var.register {
+                self.free_assigned(reg);
+                var.register = None;
+            }
+
+            return Ok(block);
+        }
+
+        // var doesn't have a stack slot so we need to create one
+        if let Some(reg) = &var.register {
+            let slot = self.allocate_stack_slot(var.size);
+            let block = InsBlock::from(Instruction::push(**reg));
+
+            self.free_assigned(reg);
+            var.register = None;
+            var.stack_slot = Some(slot);
+            return Ok(block);
+        }
+
+        return Err(CompilerError::Generic(
+            "spill_var called on a variable without a register".to_string(),
+        ));
+    }
+
+    pub fn allocate_stack_slot(&mut self, size: usize) -> StackSlot {
+        self.stack_offset -= size as i32;
+        let offset = self.stack_offset;
+        StackSlot(offset)
+    }
+
+    pub fn allocate_var(&mut self) -> Result<AssignedReg, CompilerError> {
+        if let Some(reg) = self.find_free_register() {
+            Ok(AssignedReg(reg))
+        } else {
+            Err(CompilerError::Generic(
+                "No free registers available".to_string(),
+            ))
+        }
+    }
+
+    pub fn allocate_temp(&mut self) -> Result<TempReg, CompilerError> {
+        // allocates a temporary register
+        if let Some(reg) = self.find_free_register() {
+            Ok(TempReg(reg))
+        } else {
+            todo!("an efficient stack spilling algorithm. needs scope awareness.");
+        }
+    }
+
+    pub fn free_temp(&mut self, temp: &TempReg) {
+        // frees a temporary register.
+        self.in_use[**temp as usize].1 = false;
+    }
+
+    fn free_assigned(&mut self, reg: &AssignedReg) {
+        // frees a register.
+        self.in_use[**reg as usize].1 = false;
+    }
+
+    // if we have register(s) free, return the first one.
+    fn find_free_register(&mut self) -> Option<Register> {
+        self.in_use.iter_mut().find_map(|(reg, used)| {
+            if !*used {
+                *used = true;
+                Some(*reg)
+            } else {
+                None
+            }
+        })
+    }
+}
+
+pub struct FunctionContext {
+    name: String,
+    allocator: RefCell<Allocator>,
+}
+
+impl FunctionContext {
+    pub fn new(name: String) -> Self {
+        Self {
+            name,
+            allocator: RefCell::new(Allocator::new()),
+        }
+    }
+
+    pub fn get_stack_offset(&self) -> i32 {
+        self.allocator.borrow().get_stack_offset()
+    }
+}
+
+/// scope object
+pub struct Scope<'a> {
+    /// outer scope, for a function this will be the global scope.
+    parent: Option<&'a mut Scope<'a>>,
+
+    context: Rc<FunctionContext>,
+
+    /// is the scope a function body or just a loop?
+    /// depending on the type, ending a scope will have different behaviour
+    r#type: ScopeType,
+
+    /// variables
+    variables: HashMap<Uuid, Variable>,
+
+    entry_stack_offset: i32,
+}
+
+impl<'a> Scope<'a> {
+    pub fn new(parent: &'a mut Scope<'a>, r#type: ScopeType) -> Scope<'a> {
+        Self {
+            entry_stack_offset: parent.context.get_stack_offset(),
+            context: Rc::clone(&parent.context),
+            parent: Some(parent),
+            r#type,
+            variables: HashMap::new(),
+        }
+    }
+
+    pub fn close(&mut self) -> Result<(), CompilerError> {
+        // closing a scope means we need to drop all variables in scope and free
+        // registers.
+        for (name, var) in self.variables.iter() {
+            todo!()
+            // if let Some(reg) = var.allocated_register {}
+
+            // if let Some(offset) = var.bpr_offset {
+            //     self.stack_offset -= offset;
+            // }
+        }
+
+        Ok(())
+    }
+
+    pub fn alloc_temp_reg(&mut self) -> Result<(Register, InsBlock), CompilerError> {
+        todo!()
+    }
+
+    pub fn alloc_var_reg(&mut self) -> Result<(Register, InsBlock), CompilerError> {
+        todo!()
+    }
+
+    pub fn alloc_var_stack(&mut self) -> Result<(Register, InsBlock), CompilerError> {
+        todo!()
+    }
+
+    pub fn free_var_stack(&mut self) -> Result<(Register, InsBlock), CompilerError> {
+        todo!()
+    }
+
+    pub fn free_temp_reg(&mut self) -> Result<(Register, InsBlock), CompilerError> {
+        todo!()
+    }
+}
+
+#[derive(PartialEq, Copy, Clone, Debug)]
+pub enum ScopeType {
+    Function,
+    IfBlock,
+    LoopBlock,
+}

compiler/src/backend/dsa/variable.rs

@@ -0,0 +1,93 @@
+use std::{collections::HashMap, hash::Hash, rc::Rc};
+
+use uuid::Uuid;
+
+use crate::{
+    backend::dsa::{
+        instruction::InsBlock,
+        registers::Register,
+        scope::{AssignedReg, FunctionContext, Scope, StackSlot},
+    },
+    model::{CompilerError, TypeId},
+};
+
+pub struct Variable {
+    pub name: String,
+    pub uuid: Uuid,
+
+    /// the type of the variable.
+    r#type: TypeId,
+
+    /// size taken up in bytes.
+    /// if size > 4, value must be stored on the stack.
+    pub size: usize,
+
+    pub stack_slot: Option<StackSlot>,
+    pub register: Option<AssignedReg>,
+}
+
+impl Variable {
+    pub fn new_uninit(name: String, r#type: TypeId) -> Self {
+        Self {
+            name,
+            uuid: Uuid::new_v4(),
+            size: r#type.size(),
+            r#type,
+            stack_slot: None,
+            register: None,
+        }
+    }
+
+    pub fn new(
+        name: String,
+        r#type: TypeId,
+        scope: &'_ mut Scope,
+    ) -> Result<Self, CompilerError> {
+        let mut var = Self::new_uninit(name, r#type);
+        var.alloc_default(scope);
+
+        Ok(var)
+    }
+
+    fn alloc_default(&mut self, scope: &'_ mut Scope) {
+        if self.size > 4 {
+            self.alloc_stack(scope).unwrap();
+        } else {
+            self.alloc_register(scope).unwrap();
+        }
+    }
+
+    fn alloc_register(
+        &mut self,
+        scope: &'_ mut Scope,
+    ) -> Result<Register, CompilerError> {
+        if self.size > 4 {
+            return Err(CompilerError::Generic(format!(
+                "Type {} cannot be allocated a register as it has a size of {} bytes",
+                self.r#type, self.size
+            )));
+        }
+
+        todo!("integrate with register alloc logic")
+
+        // self.allocated_register = Some(...)
+    }
+
+    fn alloc_stack(&mut self, scope: &'_ mut Scope) -> Result<usize, CompilerError> {
+        todo!("integrate with stack alloc logic")
+
+        // self.bpr_offset = Some(...)
+    }
+
+    pub fn load(&mut self, scope: &'_ mut Scope) -> Result<Register, CompilerError> {
+        todo!("load var from stack to reg (if possible)")
+    }
+
+    pub fn drop(&mut self, scope: &'_ mut Scope) -> Result<(), CompilerError> {
+        Ok(())
+    }
+
+    pub fn spill(&mut self, scope: &'_ mut Scope) -> Result<(), CompilerError> {
+        todo!()
+    }
+}

compiler/src/frontend/c/parser.rs

@@ -351,6 +351,7 @@ impl Parser {
                 op,
                 left: Box::new(expr),
                 right,
+                type_id: None,
             };
         }
 
@@ -371,6 +372,7 @@ impl Parser {
                 op,
                 left: Box::new(expr),
                 right,
+                type_id: None,
             };
         }
 
@@ -391,6 +393,7 @@ impl Parser {
                 op,
                 left: Box::new(expr),
                 right,
+                type_id: None,
             };
         }
 
@@ -407,7 +410,11 @@ impl Parser {
         if let Some(op) = op {
             self.advance();
             let operand = Box::new(self.parse_unary()?);
-            return Ok(Expression::Unary { op, operand });
+            return Ok(Expression::Unary {
+                op,
+                operand,
+                type_id: None,
+            });
         }
 
         self.parse_primary()
@@ -418,7 +425,10 @@ impl Parser {
             TokenType::Number(n) => {
                 let value = *n;
                 self.advance();
-                Ok(Expression::Number(value as isize))
+                Ok(Expression::Number {
+                    value: value as isize,
+                    type_id: None,
+                })
             }
             TokenType::Identifier(name) => {
                 let name = name.clone();
@@ -445,6 +455,7 @@ impl Parser {
                             namespace: None,
                         },
                         args,
+                        type_id: None,
                     })
                 } else {
                     Ok(Expression::Variable {

compiler/src/frontend/dsc/mod.rs

@@ -2,18 +2,18 @@ use common::logging::log;
 
 use crate::model::{CompilerError, Program};
 use parser::{ParseResult, Parser};
-use semantic_analyser::Analyser;
+// use semantic_analyser::Analyser;
 
 pub mod lexer;
 pub mod parser;
-pub mod semantic_analyser;
+// pub mod semantic_analyser;
 
 pub fn generate_ast(input: &str) -> Result<Program, CompilerError> {
     log("Tokenising Input...");
 
     let lexer = lexer::Lexer::new(&input);
     let tokens = lexer.collect::<Vec<_>>();
-    // println!("{tokens:?}");
+    println!("{tokens:#?}");
 
     log(&format!("Parsing {} Tokens...", tokens.len()));
 
@@ -30,8 +30,8 @@ pub fn generate_ast(input: &str) -> Result<Program, CompilerError> {
     log("Analyzing AST...");
     log("Checking Type Information...");
 
-    let analyser = Analyser::new();
-    analyser.analyse(ast.clone()).unwrap();
+    // let mut analyser = Analyser::new();
+    // analyser.analyse(ast.clone()).unwrap();
 
     log("Type Checking Complete...");
     Ok(ast)

compiler/src/frontend/dsc/parser.rs

@@ -1,7 +1,8 @@
 use super::lexer::Token;
 use crate::model::{
-    BinaryOperator, Block, CompilerError, ConstExpr, Declaration, Dependency, Expression,
-    Program, Statement, TypeId, UnaryOperator, Variable,
+    AssignmentOperator, BinaryOperator, Block, Call, CompilerError, ConstExpr,
+    Declaration, Dependency, Expression, Number, Program, Statement, TypeId,
+    UnaryOperator, Variable,
 };
 use crate::{expect_tt, expect_value};
 use std::ops::{ControlFlow, FromResidual, Try};
@@ -38,6 +39,10 @@ impl Parser {
             return self.parse_func();
         }
 
+        if expect_tt!(self.peek_next()?, Struct).accepted() {
+            return self.parse_struct();
+        }
+
         if expect_tt!(self.peek_next()?, Include).accepted() {
             // expect include keyword
             let _ = self.next();
@@ -78,7 +83,8 @@ impl Parser {
             let value = self.next()?;
             let init = match value {
                 Token::String(x) => Some(ConstExpr::String(x)),
-                Token::Integer(x) => Some(ConstExpr::Number(x as i32)),
+                Token::SignedInt(x, _) => Some(ConstExpr::Number(x)),
+                Token::UnsignedInt(x, _) => Some(ConstExpr::Number(x as i32)),
                 _ => {
                     return ParseResult::Reject(CompilerError::UnexpectedToken(
                         value.tt().to_string(),
@@ -98,6 +104,28 @@ impl Parser {
         ParseResult::Reject(CompilerError::UnexpectedEndOfInput)
     }
 
+    fn parse_struct(&mut self) -> ParseResult<Declaration, CompilerError> {
+        let _ = expect_tt!(self.next()?, Struct)?;
+        let name = expect_value!(self.next()?, Identifier)?;
+
+        let _ = expect_tt!(self.next()?, LeftBrace)?;
+
+        let mut fields = Vec::new();
+        while expect_tt!(self.peek_next()?, Identifier).accepted() {
+            let arg = self.parse_var_decl()?;
+            fields.push(arg);
+
+            if expect_tt!(self.peek_next()?, Comma).accepted() {
+                self.next()?;
+            } else {
+                break;
+            }
+        }
+
+        let _ = expect_tt!(self.next()?, RightBrace)?;
+        ParseResult::Accept(Declaration::Struct { name, fields })
+    }
+
     fn parse_func(&mut self) -> ParseResult<Declaration, CompilerError> {
         // expect function keyword
         let _ = expect_tt!(self.next()?, Fn);
@@ -317,60 +345,194 @@ impl Parser {
             });
         }
 
-        // handle assignment without "let"
-        let name = expect_value!(self.peek_next()?, Identifier);
-        if name.accepted() {
-            let varname = name?;
-            if expect_tt!(self.peek(1)?, LeftParen).accepted() {
-                let expr = self.parse_expression()?; // a function call expr
-                let _ = expect_tt!(self.next()?, Semicolon)?;
-                return ParseResult::Accept(Statement::Expression { expr });
-            }
-
+        // handle an in-place function call
+        if let ParseResult::Accept(name) = expect_value!(self.peek_next()?, Identifier)
+            && let ParseResult::Accept(operator) = expect_tt!(
+                self.peek(1)?,
+                Assign,
+                PlusEqual,
+                MinusEqual,
+                StarEqual,
+                SlashEqual,
+                PercentEqual,
+                AndEqual,
+                OrEqual,
+                XorEqual,
+                ShlEqual,
+                ShrEqual
+            )
+        {
+            // consume name token
+            self.next()?;
+
+            // pattern match to find operator
+            let operator = match operator {
+                Token::Assign => AssignmentOperator::Assign,
+                Token::PlusEqual => AssignmentOperator::AddAssign,
+                Token::MinusEqual => AssignmentOperator::SubAssign,
+                Token::StarEqual => AssignmentOperator::MulAssign,
+                Token::SlashEqual => AssignmentOperator::DivAssign,
+                Token::PercentEqual => AssignmentOperator::ModAssign,
+                Token::AndEqual => AssignmentOperator::AndAssign,
+                Token::OrEqual => AssignmentOperator::OrAssign,
+                Token::XorEqual => AssignmentOperator::XorAssign,
+                Token::ShlEqual => AssignmentOperator::LeftShiftAssign,
+                Token::ShrEqual => AssignmentOperator::RightShiftAssign,
+                _ => {
+                    return ParseResult::Reject(CompilerError::UnexpectedToken(
+                        self.peek_next()?.tt().to_string(),
+                    ));
+                }
+            };
+
+            // consume operator token
             self.next()?;
-            let _ = expect_tt!(self.next()?, Assign)?;
 
             let value = self.parse_expression()?;
 
             let _ = expect_tt!(self.next()?, Semicolon);
 
             return ParseResult::Accept(Statement::Assign {
-                varname: varname.name,
+                varname: name.name,
+                operator,
                 value,
             });
         }
 
-        ParseResult::Reject(CompilerError::UnexpectedToken(
-            self.peek_next()?.tt().to_string(),
-        ))
+        // parse an expression and a semicolon
+        let expr = self.parse_expression()?;
+        let _ = expect_tt!(self.next()?, Semicolon)?;
+
+        ParseResult::Accept(Statement::Expression { expr })
     }
 
     fn parse_expression(&mut self) -> ParseResult<Expression, CompilerError> {
-        self.parse_comparison()
+        self.parse_logical_or()
+    }
+
+    fn parse_logical_or(&mut self) -> ParseResult<Expression, CompilerError> {
+        let left = self.parse_logical_and()?;
+
+        let op = match self.peek_next()? {
+            Token::LogicalOr => BinaryOperator::LogicalOr,
+            _ => return ParseResult::Accept(left),
+        };
+
+        self.next()?;
+        ParseResult::Accept(Expression::Binary {
+            op,
+            left: Box::new(left),
+            right: Box::new(self.parse_logical_or()?),
+            type_id: Some(TypeId::U32),
+        })
+    }
+
+    fn parse_logical_and(&mut self) -> ParseResult<Expression, CompilerError> {
+        let left = self.parse_bitwise_or()?;
+
+        let op = match self.peek_next()? {
+            Token::LogicalAnd => BinaryOperator::LogicalAnd,
+            _ => return ParseResult::Accept(left),
+        };
+
+        self.next()?;
+        ParseResult::Accept(Expression::Binary {
+            op,
+            left: Box::new(left),
+            right: Box::new(self.parse_logical_and()?),
+            type_id: Some(TypeId::U32),
+        })
+    }
+
+    fn parse_bitwise_or(&mut self) -> ParseResult<Expression, CompilerError> {
+        let left = self.parse_bitwise_xor()?;
+
+        let op = match self.peek_next()? {
+            Token::Pipe => BinaryOperator::BitwiseOr,
+            _ => return ParseResult::Accept(left),
+        };
+
+        self.next()?;
+        ParseResult::Accept(Expression::Binary {
+            op,
+            left: Box::new(left),
+            right: Box::new(self.parse_bitwise_or()?),
+            type_id: Some(TypeId::U32),
+        })
+    }
+
+    fn parse_bitwise_xor(&mut self) -> ParseResult<Expression, CompilerError> {
+        let left = self.parse_bitwise_and()?;
+
+        let op = match self.peek_next()? {
+            Token::Caret => BinaryOperator::BitwiseXor,
+            _ => return ParseResult::Accept(left),
+        };
+
+        self.next()?;
+        ParseResult::Accept(Expression::Binary {
+            op,
+            left: Box::new(left),
+            right: Box::new(self.parse_bitwise_xor()?),
+            type_id: Some(TypeId::U32),
+        })
+    }
+
+    fn parse_bitwise_and(&mut self) -> ParseResult<Expression, CompilerError> {
+        let left = self.parse_comparison()?;
+
+        let op = match self.peek_next()? {
+            Token::Ampersand => BinaryOperator::BitwiseAnd,
+            _ => return ParseResult::Accept(left),
+        };
+
+        self.next()?;
+        ParseResult::Accept(Expression::Binary {
+            op,
+            left: Box::new(left),
+            right: Box::new(self.parse_bitwise_and()?),
+            type_id: Some(TypeId::U32),
+        })
     }
 
     fn parse_comparison(&mut self) -> ParseResult<Expression, CompilerError> {
-        let mut expr = self.parse_additive()?;
+        let left = self.parse_shift()?;
 
-        while let Some(op) = match self.peek_next()? {
-            Token::EqualEqual => Some(BinaryOperator::Ne),
-            Token::BangEqual => Some(BinaryOperator::Ne),
-            Token::Less => Some(BinaryOperator::Lt),
-            Token::Greater => Some(BinaryOperator::Gt),
-            Token::LessEqual => Some(BinaryOperator::Le),
-            Token::GreaterEqual => Some(BinaryOperator::Ge),
-            _ => None,
-        } {
-            self.next()?;
-            let right = Box::new(self.parse_additive()?);
-            expr = Expression::Binary {
-                op,
-                left: Box::new(expr),
-                right,
-            }
-        }
+        let op = match self.peek_next()? {
+            Token::EqualEqual => BinaryOperator::Equal,
+            Token::BangEqual => BinaryOperator::NotEqual,
+            Token::Less => BinaryOperator::LessThan,
+            Token::Greater => BinaryOperator::GreaterThan,
+            Token::LessEqual => BinaryOperator::LessOrEqual,
+            Token::GreaterEqual => BinaryOperator::GreaterOrEqual,
+            _ => return ParseResult::Accept(left),
+        };
 
-        ParseResult::Accept(expr)
+        self.next()?;
+        ParseResult::Accept(Expression::Binary {
+            op,
+            left: Box::new(left),
+            right: Box::new(self.parse_comparison()?),
+            type_id: Some(TypeId::Bool),
+        })
+    }
+
+    fn parse_shift(&mut self) -> ParseResult<Expression, CompilerError> {
+        let left = self.parse_additive()?;
+
+        let op = match self.peek_next()? {
+            Token::LeftShift => BinaryOperator::LeftShift,
+            Token::RightShift => BinaryOperator::RightShift,
+            _ => return ParseResult::Accept(left),
+        };
+
+        self.next()?;
+        ParseResult::Accept(Expression::Binary {
+            op,
+            left: Box::new(left),
+            right: Box::new(self.parse_shift()?),
+            type_id: Some(TypeId::U32),
+        })
     }
 
     fn parse_additive(&mut self) -> ParseResult<Expression, CompilerError> {
@@ -387,6 +549,7 @@ impl Parser {
             op,
             left: Box::new(left),
             right: Box::new(self.parse_additive()?),
+            type_id: Some(TypeId::U32),
         })
     }
 
@@ -404,59 +567,212 @@ impl Parser {
             op,
             left: Box::new(left),
             right: Box::new(self.parse_multiplicative()?),
+            type_id: None,
         })
     }
 
     fn parse_unary(&mut self) -> ParseResult<Expression, CompilerError> {
         let op = match self.peek_next()? {
+            // prefix inc/dec
+            Token::PlusPlus => UnaryOperator::Increment,
+            Token::MinusMinus => UnaryOperator::Decrement,
+
+            // arithmetic
             Token::Plus => UnaryOperator::Plus,
             Token::Minus => UnaryOperator::Minus,
+
+            // pointer
             Token::Star => UnaryOperator::Dereference,
-            Token::Amphersand => UnaryOperator::Reference,
-            _ => return ParseResult::Accept(self.parse_primary()?),
+            Token::Ampersand => UnaryOperator::AddressOf,
+
+            // boolean
+            Token::Bang => UnaryOperator::LogicalNot,
+            Token::Tilde => UnaryOperator::BitwiseNot,
+
+            Token::SizeOf => UnaryOperator::SizeOf,
+            _ => {
+                let expr = self.parse_primary()?;
+                return self.parse_postfix(expr);
+            }
         };
 
         self.next()?;
         let operand = Box::new(self.parse_unary()?);
-        ParseResult::Accept(Expression::Unary { op, operand })
+        ParseResult::Accept(Expression::Unary {
+            op,
+            operand,
+            type_id: None,
+        })
     }
 
-    fn parse_primary(&mut self) -> ParseResult<Expression, CompilerError> {
-        match self.peek_next()? {
-            Token::Integer(value) => {
-                self.next()?;
-                ParseResult::Accept(Expression::Number(value as isize))
-            }
-            Token::String(value) => {
-                self.next()?;
-                ParseResult::Accept(Expression::StringLiteral(value))
-            }
-            Token::Identifier(_) => {
-                let name = expect_value!(self.next()?, Identifier)?;
+    fn parse_postfix(
+        &mut self,
+        mut expr: Expression,
+    ) -> ParseResult<Expression, CompilerError> {
+        loop {
+            match self.peek_next()? {
+                // Type cast: expr as Type
+                Token::As => {
+                    self.next()?; // consume 'as'
+                    let target_type = self.parse_type()?;
+                    expr = Expression::TypeCast {
+                        expr: Box::new(expr),
+                        target_type,
+                        type_id: None,
+                    };
+                }
 
-                if matches!(self.peek_next()?, Token::LeftParen) {
-                    // Function call
+                // Postfix increment/decrement
+                Token::PlusPlus => {
                     self.next()?;
+                    expr = Expression::UnaryPostfix {
+                        op: UnaryOperator::Increment,
+                        operand: Box::new(expr),
+                        type_id: None,
+                    };
+                }
+                Token::MinusMinus => {
+                    self.next()?;
+                    expr = Expression::UnaryPostfix {
+                        op: UnaryOperator::Decrement,
+                        operand: Box::new(expr),
+                        type_id: None,
+                    };
+                }
+
+                // Array indexing: expr[index]
+                Token::LeftBracket => {
+                    self.next()?; // consume '['
+                    let index = Box::new(self.parse_expression()?);
+
+                    let _ = expect_tt!(self.next()?, RightBracket)?;
+
+                    expr = Expression::IndexAccess {
+                        expr: Box::new(expr),
+                        index,
+                        type_id: None,
+                    };
+                }
+
+                // Function call: expr(args...)
+                Token::LeftParen => {
+                    self.next()?; // consume '('
                     let mut args = Vec::new();
 
                     if !matches!(self.peek_next()?, Token::RightParen) {
-                        args.push(self.parse_expression()?);
-
-                        while matches!(self.peek_next()?, Token::Comma) {
-                            self.next()?;
+                        loop {
                             args.push(self.parse_expression()?);
+                            if !matches!(self.peek_next()?, Token::Comma) {
+                                break;
+                            }
+                            self.next()?; // consume comma
                         }
                     }
 
                     let _ = expect_tt!(self.next()?, RightParen)?;
 
-                    ParseResult::Accept(Expression::Call { name, args })
-                } else {
-                    ParseResult::Accept(Expression::Variable {
+                    if let Expression::Variable { name, .. } = expr {
+                        expr = Expression::Call {
+                            func: Call { name, args },
+                            type_id: None,
+                        };
+                    }
+                }
+
+                // Member access: expr.member (if you support structs)
+                Token::Dot => {
+                    self.next()?;
+                    let field_name = expect_value!(self.next()?, Identifier)?;
+                    expr = Expression::MemberAccess {
+                        expr: Box::new(expr),
+                        field_name,
+                        type_id: None,
+                    };
+                }
+
+                // No more postfix operations
+                _ => break,
+            }
+        }
+
+        ParseResult::Accept(expr)
+    }
+
+    fn parse_primary(&mut self) -> ParseResult<Expression, CompilerError> {
+        match self.peek_next()? {
+            Token::UnsignedInt(value, type_id) => {
+                self.next()?;
+                ParseResult::Accept(Expression::Number(Number::Unsigned(value, type_id)))
+            }
+            Token::SignedInt(value, type_id) => {
+                self.next()?;
+                ParseResult::Accept(Expression::Number(Number::Signed(value, type_id)))
+            }
+            Token::String(value) => {
+                self.next()?;
+                ParseResult::Accept(Expression::StringLiteral(value))
+            }
+            Token::Char(value) => {
+                self.next()?;
+                ParseResult::Accept(Expression::CharLiteral(value))
+            }
+
+            Token::Identifier(name) => {
+                self.next()?;
+
+                // if the next token isn't the beginning of a struct literal this is just
+                // an identifier.
+                if !expect_tt!(self.peek_next()?, LeftBrace).accepted() {
+                    return ParseResult::Accept(Expression::Variable {
                         name,
                         expr_type: None,
-                    })
+                    });
                 }
+
+                let _ = self.next()?;
+
+                let mut fields = Vec::new();
+                while !expect_tt!(self.peek_next()?, RightBrace).accepted() {
+                    let name = expect_value!(self.next()?, Identifier)?;
+                    let _ = expect_tt!(self.next()?, Colon)?;
+                    let expr = self.parse_expression()?;
+
+                    fields.push((name, expr));
+
+                    if expect_tt!(self.peek_next()?, Comma).accepted() {
+                        self.next()?;
+                    } else {
+                        break;
+                    }
+                }
+
+                let _ = expect_tt!(self.next()?, RightBrace)?;
+
+                ParseResult::Accept(Expression::StructLiteral {
+                    name,
+                    fields,
+                    type_id: None,
+                })
+            }
+            Token::LeftBracket => {
+                self.next()?; // consume '['
+                let mut elements = Vec::new();
+
+                if !matches!(self.peek_next()?, Token::RightBracket) {
+                    loop {
+                        elements.push(self.parse_expression()?);
+                        if !matches!(self.peek_next()?, Token::Comma) {
+                            break;
+                        }
+                        self.next()?; // consume comma
+                    }
+                }
+
+                expect_tt!(self.next()?, RightBracket)?;
+                ParseResult::Accept(Expression::ArrayLiteral {
+                    elements,
+                    type_id: None,
+                })
             }
             Token::LeftParen => {
                 self.next()?;
@@ -484,21 +800,88 @@ impl Parser {
     }
 
     fn parse_type(&mut self) -> ParseResult<TypeId, CompilerError> {
-        // get the type name incl namespace
-        let typename = expect_value!(self.next()?, Identifier)?;
+        println!("yes {:?}", self.peek_next()?);
 
-        match typename.name.as_str() {
-            "u32" => ParseResult::Accept(TypeId::U32),
-            "u16" => ParseResult::Accept(TypeId::U16),
-            "u8" => ParseResult::Accept(TypeId::U8),
-            "i32" => ParseResult::Accept(TypeId::I32),
-            "i16" => ParseResult::Accept(TypeId::I16),
-            "i8" => ParseResult::Accept(TypeId::I8),
-            "void" => ParseResult::Accept(TypeId::Void),
-            "char" => ParseResult::Accept(TypeId::Char),
-            "str" => ParseResult::Accept(TypeId::Ptr(Box::new(TypeId::Char))),
-            _ => todo!("Implement parsing for other types!!"),
+        // parse primitive or named type
+        if expect_tt!(self.peek_next()?, Identifier).accepted() {
+            return self.parse_type_identifier();
         }
+
+        // parse array type
+        if expect_tt!(self.peek_next()?, LeftBracket).accepted() {
+            let _ = self.next()?;
+
+            let internal_type = self.parse_type()?;
+            let _ = expect_tt!(self.next()?, Semicolon)?;
+
+            let size = expect_value!(self.next()?, UnsignedInt)?;
+
+            let _ = expect_tt!(self.next()?, RightBracket)?;
+
+            return ParseResult::Accept(TypeId::Array {
+                r#type: Box::new(internal_type),
+                size: size as usize,
+            });
+        }
+
+        // parse tuple type
+        if expect_tt!(self.peek_next()?, LeftParen).accepted() {
+            let _ = self.next()?;
+
+            let mut types = Vec::new();
+            while !expect_tt!(self.peek_next()?, RightParen).accepted() {
+                types.push(self.parse_type()?);
+                if !expect_tt!(self.peek_next()?, Comma).accepted() {
+                    break;
+                }
+                let _ = self.next()?;
+            }
+            let _ = expect_tt!(self.next()?, RightParen)?;
+
+            return ParseResult::Accept(TypeId::Tuple(types));
+        }
+
+        ParseResult::Reject(CompilerError::Generic(format!(
+            "Parsing type but no valid type was detected: {:?}",
+            self.peek_next()?
+        )))
+    }
+
+    fn parse_type_identifier(&mut self) -> ParseResult<TypeId, CompilerError> {
+        // get the type name incl namespace
+        let name = expect_value!(self.next()?, Identifier)?;
+
+        let type_id = match name.name.as_str() {
+            "u32" => TypeId::U32,
+            "u16" => TypeId::U16,
+            "u8" => TypeId::U8,
+            "i32" => TypeId::I32,
+            "i16" => TypeId::I16,
+            "i8" => TypeId::I8,
+            "void" => TypeId::Void,
+            "char" => TypeId::Char,
+            "str" => TypeId::Ptr(Box::new(TypeId::Char)),
+            _ => {
+                let mut generics = Vec::new();
+                if expect_tt!(self.peek_next()?, Less).accepted() {
+                    let _ = self.next()?;
+
+                    // loop until we find the closing '>'
+                    while !expect_tt!(self.peek_next()?, Greater).accepted() {
+                        generics.push(self.parse_type()?);
+                        if !expect_tt!(self.peek_next()?, Comma).accepted() {
+                            break;
+                        }
+                        let _ = self.next()?;
+                    }
+                    let _ = expect_tt!(self.next()?, Greater)?;
+                }
+
+                TypeId::UnknownCustom { name, generics }
+            }
+        };
+
+        ParseResult::Accept(type_id)
     }
 
     fn next(&mut self) -> ParseResult<Token, CompilerError> {
@@ -595,7 +978,7 @@ macro_rules! expect_value {
     ($expr:expr, $variant:ident) => {{
         let tok = $expr;
         match tok.clone() {
-            Token::$variant(value) => ParseResult::Accept(value),
+            Token::$variant(first, ..) => ParseResult::Accept(first),
             _ => {
                 ParseResult::Reject(CompilerError::UnexpectedToken(tok.tt().to_string()))
             }

compiler/src/frontend/dsc/semantic_analyser.rs

@@ -1,13 +1,226 @@
-use crate::model::{CompilerError, Program};
+use std::collections::HashMap;
 
-pub struct Analyser;
+use crate::model::{
+    BinaryOperator, // You'll need to add this to your imports
+    CompilerError,
+    Declaration,
+    Dependency,
+    Expression,
+    Program,
+    TypeId,
+    UnaryOperator,
+};
+
+pub struct Analyser {
+    symbol_table: HashMap<String, Declaration>,
+}
+
+const NUMERIC_TYPES: &[TypeId] = &[
+    TypeId::U32,
+    TypeId::I32,
+    TypeId::I16,
+    TypeId::U16,
+    TypeId::I8,
+    TypeId::U8,
+];
 
 impl Analyser {
     pub fn new() -> Self {
-        Self
+        Self {
+            symbol_table: HashMap::new(),
+        }
     }
 
-    pub fn analyse(&self, _ast: Program) -> Result<(), CompilerError> {
+    pub fn analyse(&mut self, ast: Program) -> Result<(), CompilerError> {
+        // build table of global symbols.
+        for dec in ast.declarations {
+            let name = match dec.clone() {
+                Declaration::Function { name, .. } => name,
+                Declaration::Variable { var, .. } => var.name,
+                Declaration::Dependency(Dependency { name, .. }) => name,
+            };
+
+            self.symbol_table.insert(name, dec);
+        }
+
         Ok(())
     }
+
+    fn match_type(
+        actual: TypeId,
+        expected: Option<TypeId>,
+    ) -> Result<TypeId, CompilerError> {
+        match expected {
+            Some(id) => {
+                if id != actual {
+                    Err(CompilerError::TypeMismatch(id, actual))
+                } else {
+                    Ok(actual)
+                }
+            }
+            None => Ok(actual),
+        }
+    }
+
+    fn get_type(
+        &mut self, // Changed from &self to &mut self since we modify expr
+        expr: &mut Expression,
+        expected_type: Option<TypeId>,
+    ) -> Result<TypeId, CompilerError> {
+        match expr {
+            // Correct IFF we're expecting a void type
+            Expression::Empty => Self::match_type(TypeId::Void, expected_type),
+
+            // Correct IFF we're expecting a char type
+            Expression::CharLiteral(_) => Self::match_type(TypeId::Char, expected_type),
+
+            // Correct IFF we're expecting a string slice type
+            Expression::StringLiteral(_) => {
+                Self::match_type(TypeId::Ptr(Box::new(TypeId::Char)), expected_type)
+            }
+
+            Expression::Variable { name, expr_type } => {
+                let actual = expr_type.clone().ok_or(CompilerError::UnknownType)?;
+                Self::match_type(actual, expected_type)
+            }
+
+            Expression::Number { value, type_id } => {
+                // If we already know the TypeId
+                if let Some(id) = type_id {
+                    return Self::match_type(id.clone(), expected_type);
+                }
+
+                // If we're expecting a type id, check it's numeric.
+                // TODO: add checks to make sure it's valid for its size eg u8 cant be
+                // more than 255
+                if let Some(expected) = expected_type {
+                    if NUMERIC_TYPES.contains(&expected) {
+                        *type_id = Some(expected.clone());
+                        return Ok(expected);
+                    } else {
+                        return Err(CompilerError::TypeMismatch(expected, TypeId::U32));
+                    }
+                }
+
+                // Default to i32 if no type information is available
+                *type_id = Some(TypeId::I32);
+                Ok(TypeId::I32)
+            }
+
+            Expression::Binary {
+                op,
+                left,
+                right,
+                type_id,
+            } => {
+                // For binary operations, both operands should have compatible types
+                // and the result type depends on the operation
+                let left_type = self.get_type(left, None)?;
+                let right_type = self.get_type(right, Some(left_type.clone()))?;
+
+                // For numeric operations, result has the same type as operands
+                if NUMERIC_TYPES.contains(&left_type)
+                    && NUMERIC_TYPES.contains(&right_type)
+                {
+                    *type_id = Some(left_type);
+                    Self::match_type(left_type, expected_type)
+                } else {
+                    Err(CompilerError::TypeMismatch(left_type, right_type))
+                }
+            }
+
+            Expression::Unary {
+                op,
+                operand,
+                type_id,
+            } => {
+                match op {
+                    UnaryOperator::Plus | UnaryOperator::Minus => {
+                        // Unary +/- require numeric operands
+                        let inner_type = self.get_type(operand, None)?;
+
+                        if NUMERIC_TYPES.contains(&inner_type) {
+                            *type_id = Some(inner_type.clone());
+                            Self::match_type(inner_type, expected_type)
+                        } else {
+                            Err(CompilerError::TypeMismatch(inner_type, TypeId::I32))
+                        }
+                    }
+
+                    UnaryOperator::Dereference => {
+                        // For dereference (*ptr), the operand must be a pointer
+                        // and the result type is what the pointer points to
+                        let inner_type = self.get_type(operand, None)?;
+
+                        match inner_type {
+                            TypeId::Ptr(inner) => {
+                                let deref_type = *inner;
+                                *type_id = Some(deref_type.clone());
+                                Self::match_type(deref_type, expected_type)
+                            }
+                            _ => Err(CompilerError::Generic(format!(
+                                "Cannot dereference non-pointer type: {:?}",
+                                inner_type
+                            ))),
+                        }
+                    }
+
+                    UnaryOperator::Reference => {
+                        // For reference (&var), we need to determine what we're taking
+                        // a reference to, then wrap it in a Ptr
+                        // If expected_type is Ptr(T), then operand should have type T
+                        let expected_inner = match expected_type.clone() {
+                            Some(TypeId::Ptr(inner)) => Some(*inner),
+                            _ => None,
+                        };
+
+                        let inner_type = self.get_type(operand, expected_inner)?;
+                        let ref_type = TypeId::Ptr(Box::new(inner_type));
+                        *type_id = Some(ref_type.clone());
+                        Self::match_type(ref_type, expected_type)
+                    }
+                }
+            }
+
+            Expression::Call {
+                name,
+                args,
+                type_id,
+            } => match self.symbol_table.get(&name.name) {
+                Some(Declaration::Function {
+                    params,
+                    return_type,
+                    ..
+                }) => {
+                    // check that we've given the right number of arguments.
+                    if args.len() != params.len() {
+                        return Err(CompilerError::Generic(format!(
+                            "Function {} expected {} arguments but received {}",
+                            name.name,
+                            params.len(),
+                            args.len()
+                        )));
+                    }
+
+                    for (arg, param) in args.iter_mut().zip(params.iter()) {
+                        // check that the argument type matches the parameter type.
+                        let provided_type = self.get_type(arg, Some(param.type_id))?;
+                        if provided_type != param.type_id {
+                            return Err(CompilerError::TypeMismatch(
+                                param.type_id,
+                                provided_type,
+                            ));
+                        }
+                    }
+
+                    *type_id = Some(return_type.clone());
+                    Self::match_type(return_type.clone(), expected_type)
+                }
+                _ => Err(CompilerError::Generic(format!(
+                    "Function {} not found in symbol table",
+                    name.name
+                ))),
+            },
+        }
+    }
 }

compiler/src/frontend/mod.rs

@@ -1,12 +1,12 @@
 use crate::model::{CompilerError, Program};
 
-mod c;
+// mod c;
 mod dsc;
 
 pub fn compiler_frontend(ext: &str, data: &str) -> Result<Program, CompilerError> {
     match ext {
         "dsc" => Ok(dsc::generate_ast(&data)?),
-        "c" => Ok(c::generate_ast(&data)?),
+        // "c" => Ok(c::generate_ast(&data)?),
         _ => Err(CompilerError::Generic(format!(
             "File type {} not supported",
             ext

compiler/src/lib.rs

@@ -46,6 +46,8 @@ pub fn compile_file(
         Err(err) => return Err(format!("Compilation failed: {err:?}").into()),
     };
 
+    println!("Parsed AST: {:#?}", ast);
+
     let output_ext = output_path
         .extension()
         .and_then(|s| s.to_str())

compiler/src/main.rs

@@ -1,7 +1,5 @@
 use std::path::Path;
 
-use compiler;
-
 fn main() {
     // read from input file: syntax "c_compiler <src.c> [output.dsa]"
     let args: Vec<String> = std::env::args().collect();

compiler/src/model.rs

@@ -9,13 +9,24 @@ pub enum CompilerError {
     Undefined(Name),
     InvalidSyntax(String),
     Generic(String),
+    UnknownType,
+    TypeMismatch(TypeId, TypeId),
+    Unimplemented(String),
 }
 
-#[derive(Debug, PartialEq, Clone)]
+#[derive(Debug, PartialEq, Eq, Clone)]
 pub struct Name {
     pub name: String,
     pub namespace: Option<String>,
 }
+impl Name {
+    pub fn new(name: impl Into<String>, namespace: Option<String>) -> Self {
+        Self {
+            name: name.into(),
+            namespace,
+        }
+    }
+}
 
 #[derive(Debug, Clone)]
 pub struct Program {
@@ -37,6 +48,10 @@ pub enum Declaration {
         is_const: bool,
     },
     Dependency(Dependency),
+    Struct {
+        name: Name,
+        fields: Vec<Variable>,
+    },
 }
 
 #[derive(Debug, Clone)]
@@ -46,7 +61,7 @@ pub struct Dependency {
 }
 
 #[allow(unused)]
-#[derive(Debug, Clone)]
+#[derive(Debug, Clone, PartialEq)]
 pub enum TypeId {
     U8,
     U16,
@@ -54,18 +69,113 @@ pub enum TypeId {
     I8,
     I16,
     I32,
+    Bool,
     Char,
     Void,
     Ptr(Box<TypeId>),
     Ref(Box<TypeId>),
-    Array(Box<TypeId>, usize),
-    Struct { name: Name, fields: Vec<Variable> },
+    Tuple(Vec<TypeId>),
+    Array {
+        r#type: Box<TypeId>,
+        size: usize,
+    },
+    UnknownCustom {
+        name: Name,
+        generics: Vec<TypeId>,
+    },
+    Struct {
+        name: Name,
+        fields: Vec<TypeId>,
+        generics: Vec<TypeId>,
+    },
+}
+
+impl TypeId {
+    pub fn size(&self) -> usize {
+        match self {
+            Self::U8 => 1,
+            Self::U16 => 2,
+            Self::U32 => 4,
+            Self::I8 => 1,
+            Self::I16 => 2,
+            Self::I32 => 4,
+            Self::Bool => 1,
+            Self::Char => 1,
+            Self::Void => 0,
+            Self::Ptr(t) => t.size(),
+            Self::Ref(t) => t.size(),
+            Self::Tuple(types) => types.iter().map(|t| t.size()).sum(),
+            Self::Array { r#type, size } => r#type.size() * size,
+            Self::UnknownCustom { .. } => 1, /* TODO: calculate type size during */
+            // semantic analysis
+            Self::Struct { fields, .. } => fields.iter().map(|t| t.size()).sum(),
+        }
+    }
+}
+
+impl fmt::Display for TypeId {
+    fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
+        match self {
+            Self::U8 => write!(f, "u8"),
+            Self::U16 => write!(f, "u16"),
+            Self::U32 => write!(f, "u32"),
+            Self::I8 => write!(f, "i8"),
+            Self::I16 => write!(f, "i16"),
+            Self::I32 => write!(f, "i32"),
+            Self::Bool => write!(f, "bool"),
+            Self::Char => write!(f, "char"),
+            Self::Void => write!(f, "void"),
+            Self::Ptr(t) => write!(f, "*{}", t),
+            Self::Ref(t) => write!(f, "&{}", t),
+            Self::Tuple(elems) => write!(
+                f,
+                "({})",
+                elems
+                    .iter()
+                    .map(|t| t.to_string())
+                    .collect::<Vec<String>>()
+                    .join(", ")
+            ),
+            Self::Array { r#type, size } => write!(f, "[{}; {}]", r#type, size),
+            Self::UnknownCustom { name, generics } => {
+                write!(
+                    f,
+                    "{}<{}>",
+                    name,
+                    generics
+                        .iter()
+                        .map(|t| t.to_string())
+                        .collect::<Vec<String>>()
+                        .join(", ")
+                )
+            }
+            Self::Struct {
+                name,
+                fields,
+                generics,
+            } => write!(
+                f,
+                "struct<{}> {} {{{}}}",
+                generics
+                    .iter()
+                    .map(|t| t.to_string())
+                    .collect::<Vec<String>>()
+                    .join(", "),
+                name,
+                fields
+                    .iter()
+                    .map(|t| t.to_string())
+                    .collect::<Vec<String>>()
+                    .join(", ")
+            ),
+        }
+    }
 }
 
 pub type Block = Vec<Statement>;
 
 #[allow(unused)]
-#[derive(Debug, Clone)]
+#[derive(Debug, Clone, PartialEq)]
 pub struct Variable {
     pub name: String,
     pub type_id: TypeId,
@@ -81,6 +191,7 @@ pub enum Statement {
     },
     Assign {
         varname: String,
+        operator: AssignmentOperator,
         value: Expression,
     },
     PtrWrite {
@@ -100,6 +211,7 @@ pub enum Statement {
         body: Vec<Statement>,
     },
     Loop(Block),
+    Defer(Call),
     Break,
     Continue,
     Return(Option<Expression>),
@@ -128,67 +240,227 @@ pub enum Expression {
         op: BinaryOperator,
         left: Box<Expression>,
         right: Box<Expression>,
+
+        // Post-Semantic Analysis
+        type_id: Option<TypeId>,
     },
     Unary {
         op: UnaryOperator,
         operand: Box<Expression>,
+
+        // Post-Semantic Analysis
+        type_id: Option<TypeId>,
+    },
+    UnaryPostfix {
+        op: UnaryOperator,
+        operand: Box<Expression>,
+
+        // Post-Semantic Analysis
+        type_id: Option<TypeId>,
     },
     Variable {
         name: Name,
         expr_type: Option<TypeId>,
     },
-    Call {
-        name: Name,
-        args: Vec<Expression>,
+    TypeCast {
+        expr: Box<Expression>,
+        target_type: TypeId,
+
+        // Post-Semantic Analysis
+        type_id: Option<TypeId>,
     },
-    Number(isize),
+    IndexAccess {
+        expr: Box<Expression>,
+        index: Box<Expression>,
+
+        // Post-Semantic Analysis
+        type_id: Option<TypeId>,
+    },
+    MemberAccess {
+        expr: Box<Expression>,
+        field_name: Name,
+
+        // Post-Semantic Analysis
+        type_id: Option<TypeId>,
+    },
+    Call {
+        func: Call,
+
+        // Post-Semantic Analysis
+        type_id: Option<TypeId>,
+    },
+    Number(Number),
     StringLiteral(String),
     CharLiteral(char),
+    ArrayLiteral {
+        elements: Vec<Expression>,
+        type_id: Option<TypeId>,
+    },
+    StructLiteral {
+        name: Name,
+        fields: Vec<(Name, Expression)>,
+        type_id: Option<TypeId>,
+    },
+}
+
+#[derive(Debug, Clone)]
+pub enum Number {
+    Signed(i32, Option<TypeId>),
+    Unsigned(u32, Option<TypeId>),
+}
+
+#[derive(Debug, Clone)]
+pub struct Call {
+    pub name: Name,
+    pub args: Vec<Expression>,
 }
 
 impl Expression {
     pub fn is_pure(&self) -> bool {
         match self {
-            Expression::Number(_) => true,
+            Expression::Number { .. } => true,
             Expression::StringLiteral(_) => true,
             Expression::CharLiteral(_) => true,
             Expression::Call { .. } => false,
             Expression::Binary { left, right, .. } => left.is_pure() && right.is_pure(),
             Expression::Unary { operand, .. } => operand.is_pure(),
+            Expression::UnaryPostfix { operand, .. } => operand.is_pure(),
             Expression::Empty => true,
             Expression::Variable { .. } => true,
+            Expression::TypeCast { expr, .. } => expr.is_pure(),
+            Expression::IndexAccess { expr, index, .. } => {
+                expr.is_pure() && index.is_pure()
+            }
+            Expression::MemberAccess { expr, .. } => expr.is_pure(),
+            Expression::ArrayLiteral { elements, .. } => {
+                elements.iter().all(|element| element.is_pure())
+            }
+            Expression::StructLiteral { fields, .. } => {
+                fields.iter().all(|(_, expr)| expr.is_pure())
+            }
         }
     }
+
+    pub fn type_id(&self) -> Result<TypeId, CompilerError> {
+        match self {
+            Expression::Number(
+                Number::Signed(_, type_id) | Number::Unsigned(_, type_id),
+            ) => type_id.clone().ok_or(CompilerError::UnknownType),
+            Expression::StringLiteral(_) => Ok(TypeId::Ptr(Box::new(TypeId::Char))),
+            Expression::CharLiteral(_) => Ok(TypeId::Char),
+            Expression::Call { type_id, .. } => {
+                type_id.clone().ok_or(CompilerError::UnknownType)
+            }
+            Expression::Binary { type_id, .. } => {
+                type_id.clone().ok_or(CompilerError::UnknownType)
+            }
+            Expression::Unary { type_id, .. } => {
+                type_id.clone().ok_or(CompilerError::UnknownType)
+            }
+            Expression::UnaryPostfix { type_id, .. } => {
+                type_id.clone().ok_or(CompilerError::UnknownType)
+            }
+            Expression::Empty => Ok(TypeId::Void),
+            Expression::Variable { expr_type, .. } => {
+                expr_type.clone().ok_or(CompilerError::UnknownType)
+            }
+            Expression::TypeCast { type_id, .. } => {
+                type_id.clone().ok_or(CompilerError::UnknownType)
+            }
+            Expression::IndexAccess { expr, .. } => expr.type_id(),
+            Expression::MemberAccess { expr, .. } => expr.type_id(),
+            Expression::ArrayLiteral { elements, .. } => {
+                let element_type = elements
+                    .first()
+                    .map_or(TypeId::Void, |e| e.type_id().unwrap_or(TypeId::Void));
+                Ok(TypeId::Array {
+                    r#type: Box::new(element_type),
+                    size: elements.len(),
+                })
+            }
+            Expression::StructLiteral { name, fields, .. } => {
+                let fields = fields
+                    .iter()
+                    .map(|(_, expr)| expr.type_id())
+                    .collect::<Result<Vec<_>, _>>()?;
+                Ok(TypeId::Struct {
+                    name: name.clone(),
+                    fields,
+                    generics: Vec::new(),
+                })
+            }
+        }
+    }
+}
+
+#[derive(Debug, Clone, PartialEq)]
+pub enum AssignmentOperator {
+    Assign,
+    AddAssign,
+    SubAssign,
+    MulAssign,
+    DivAssign,
+    ModAssign,
+    AndAssign,
+    OrAssign,
+    XorAssign,
+    LeftShiftAssign,
+    RightShiftAssign,
 }
 
 #[allow(unused)]
 #[derive(Debug, Clone, PartialEq)]
 pub enum BinaryOperator {
+    // arithmetic
     Add,
     Sub,
     Mul,
     Div,
-    Eq,
-    Ne,
-    Lt,
-    Gt,
-    Le,
-    Ge,
+    Mod,
+
+    // comparison
+    Equal,
+    NotEqual,
+    LessThan,
+    GreaterThan,
+    LessOrEqual,
+    GreaterOrEqual,
+
+    // bitwise
+    BitwiseAnd,
+    BitwiseOr,
+    BitwiseXor,
+
+    // logical
+    LogicalAnd,
+    LogicalOr,
+
+    // shift
+    LeftShift,
+    RightShift,
 }
 
 impl fmt::Display for BinaryOperator {
     fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
         match self {
-            BinaryOperator::Add => write!(f, "+"),
-            BinaryOperator::Sub => write!(f, "-"),
-            BinaryOperator::Mul => write!(f, "*"),
-            BinaryOperator::Div => write!(f, "/"),
-            BinaryOperator::Eq => write!(f, "=="),
-            BinaryOperator::Ne => write!(f, "!="),
-            BinaryOperator::Lt => write!(f, "<"),
-            BinaryOperator::Gt => write!(f, ">"),
-            BinaryOperator::Le => write!(f, "<="),
-            BinaryOperator::Ge => write!(f, ">="),
+            Self::Add => write!(f, "+"),
+            Self::Sub => write!(f, "-"),
+            Self::Mul => write!(f, "*"),
+            Self::Div => write!(f, "/"),
+            Self::Mod => write!(f, "%"),
+            Self::Equal => write!(f, "=="),
+            Self::NotEqual => write!(f, "!="),
+            Self::LessThan => write!(f, "<"),
+            Self::GreaterThan => write!(f, ">"),
+            Self::LessOrEqual => write!(f, "<="),
+            Self::GreaterOrEqual => write!(f, ">="),
+            Self::BitwiseAnd => write!(f, "&"),
+            Self::BitwiseOr => write!(f, "|"),
+            Self::BitwiseXor => write!(f, "^"),
+            Self::LogicalAnd => write!(f, "&&"),
+            Self::LogicalOr => write!(f, "||"),
+            Self::LeftShift => write!(f, "<<"),
+            Self::RightShift => write!(f, ">>"),
         }
     }
 }
@@ -197,17 +469,27 @@ impl fmt::Display for BinaryOperator {
 pub enum UnaryOperator {
     Plus,
     Minus,
-    Reference,
+    AddressOf,
     Dereference,
+    BitwiseNot,
+    LogicalNot,
+    Increment,
+    Decrement,
+    SizeOf,
 }
 
 impl fmt::Display for UnaryOperator {
     fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
         match self {
-            UnaryOperator::Plus => write!(f, "+"),
-            UnaryOperator::Minus => write!(f, "-"),
-            UnaryOperator::Dereference => write!(f, "*"),
-            UnaryOperator::Reference => write!(f, "&"),
+            Self::Increment => write!(f, "++"),
+            Self::Decrement => write!(f, "--"),
+            Self::Plus => write!(f, "+"),
+            Self::Minus => write!(f, "-"),
+            Self::Dereference => write!(f, "*"),
+            Self::AddressOf => write!(f, "&"),
+            Self::BitwiseNot => write!(f, "~"),
+            Self::LogicalNot => write!(f, "!"),
+            Self::SizeOf => write!(f, "sizeof"),
         }
     }
 }

docs/DSA_ISA_Specification.md

@@ -12,7 +12,9 @@ The Damn Simple Architecture (DSA) is a 32-bit RISC-style architecture designed
 | Halfword | 16 bits | 2-byte aligned |
 | Word | 32 bits | 4-byte aligned |
 
-All multi-byte values use little-endian byte order.
+**Note on Endianness:**
+- Instructions and numeric data in memory: Little-endian
+- Data defined via `db/dh/dw` directives: Big-endian (assembler-specific)
 
 ## Registers
 
@@ -23,33 +25,32 @@ DSA provides 32 programmer-accessible registers plus several internal system reg
 | Hex | Register | Type | Description |
 |-----|----------|------|-------------|
 | 0x00-0x0F | **rg0-rgf** | General Purpose | 16 general-purpose registers for variables and temporary values |
-| 0x10 | **acc** | Special | Accumulator for calculations and temporary storage<br/>⚠️ May be overwritten by pseudo-instructions |
+| 0x10 | **acc** | Special | Accumulator for calculations and temporary storage<br/>⚠️ Used as scratch by pseudo-instructions - volatile |
 | 0x11 | **spr** | Special | Stack pointer - points to top of stack |
 | 0x12 | **bpr** | Special | Base pointer - used for stack frame management |
-| 0x13 | **ret** | Special | Return address register - stores function return addresses |
+| 0x13 | **ret** | Special | Return address register - used for function returns |
 | 0x14 | **idr** | Privileged | Interrupt descriptor table address<br/>Read/write triggers protection fault in user mode |
 | 0x15 | **mmr** | Privileged | Hardware memory map table address<br/>Read/write triggers protection fault in user mode |
 | 0x16 | **zero** | Read-only | Constant zero value<br/>Reads always return 0, writes are discarded |
-| 0x17 | **noreg** | Placeholder | Indicates unused register field<br/>Read/write triggers illegal instruction fault |
+| 0x17 | **noreg** | Placeholder | Indicates unused register field<br/>Read/write triggers illegal instruction fault<br/>Can also be referenced as **null** |
 | 0x18-0x1F | - | Reserved | Reserved for future use |
 
+**System Registers (indices 0x18-0x1C):**
+These exist in the encoding space but are internal to the CPU implementation:
+
+| Hex | Register | Description |
+|-----|----------|-------------|
+| 0x18 | **mar** | Memory Address Register (CPU internal) |
+| 0x19 | **mdr** | Memory Data Register (CPU internal) |
+| 0x1A | **sts** | Status Register (CPU internal) |
+| 0x1B | **cir** | Current Instruction Register (CPU internal) |
+| 0x1C | **pcx** | Program Counter (read-only, special access) |
+
 **Note on PCX (Program Counter):**
-- PCX is a read-only system register that can be accessed in some contexts
+- PCX can be read in certain contexts (e.g., stored during CALL)
 - Writing to PCX triggers a protection fault
 - PCX is automatically updated by jump and branch instructions
 
-### System Registers (Internal)
-
-These registers are used internally by the CPU and are not directly accessible via assembly instructions:
-
-| Register | Description |
-|----------|-------------|
-| **MAR** | Memory Address Register - holds address for memory operations |
-| **MDR** | Memory Data Register - holds data for memory transfers |
-| **CIR** | Current Instruction Register - holds instruction being executed |
-| **STS** | Status Register - stores comparison and arithmetic flags |
-| **PCX** | Program Counter - stores address of next instruction |
-
 ### Status Register (STS) Layout
 
 The status register is a 32-bit register with the following flag bits:
@@ -108,7 +109,7 @@ Used for operations with a 16-bit immediate value.
 **Usage:**
 - Arithmetic: Immediate is a signed value
 - Memory access: Immediate is a signed byte offset from base address
-- Branches: Immediate is a signed offset from current PCX
+- Branches: Immediate is a signed offset added to base register
 - Literal loads: Immediate is unsigned 16-bit value
 
 ### J-Type (Jump) Instructions
@@ -131,7 +132,7 @@ Used for absolute jumps with large address ranges.
 
 **Jump Range:** 256MB region around current PC (±128MB)
 
-**Note:** J-type instructions are defined but currently unused. Use I-type JMP with register addressing for long jumps.
+**Note:** J-type instructions are defined but currently unused. Use I-type JMP with register addressing for all jumps.
 
 ## Hardware Instructions
 
@@ -197,8 +198,8 @@ In machine code: SrcReg (SrcReg field), BaseReg (DestReg field), Offset (Immedia
 
 | Hex | Mnemonic | Type | Operands | Description |
 |-----|----------|------|----------|-------------|
-| 0x0B | **LLI** | I | DestReg, Value | Load 16-bit value into lower 16 bits<br/>⚠️ **CLEARS upper 16 bits!** |
-| 0x0C | **LUI** | I | DestReg, Value | Load 16-bit value into upper 16 bits<br/>Lower 16 bits unchanged |
+| 0x0B | **LLI** | I | Value, DestReg | Load 16-bit value into lower 16 bits<br/>⚠️ **CLEARS upper 16 bits!** |
+| 0x0C | **LUI** | I | Value, DestReg | Load 16-bit value into upper 16 bits<br/>Lower 16 bits unchanged |
 
 **Usage for 32-bit Values:**
 ```
@@ -208,29 +209,38 @@ LUI 0xABCD, rg0    ; rg0 = 0xABCD1234
 
 **⚠️ CRITICAL:** Always execute LLI before LUI, as LLI clears the upper 16 bits!
 
+**Note on LUI:** The assembler may shift the immediate value right by 16 bits when encoding, so specify the upper 16 bits directly (e.g., `LUI 0xABCD, rg0` not `LUI 0xABCD0000, rg0`).
+
 **Encoding Note:**
-In machine code: Value (Immediate field), DestReg field (SrcReg unused, set to noreg)
+In machine code: Value (Immediate field), DestReg (SrcReg field for LLI, SrcReg field for LUI)
 
 ### Jump and Branch Instructions
 
 | Hex | Mnemonic | Type | Operands | Description |
 |-----|----------|------|----------|-------------|
-| 0x0D | **JMP** | I | DestReg, Offset | Unconditional jump to (DestReg + Offset) |
-| 0x0E | **JEQ** | I | DestReg, Offset | Jump if Equal flag set |
-| 0x0F | **JNE** | I | DestReg, Offset | Jump if Equal flag NOT set |
-| 0x10 | **JGT** | I | DestReg, Offset | Jump if GreaterThan flag set |
-| 0x11 | **JGE** | I | DestReg, Offset | Jump if GreaterThan OR Equal flag set |
-| 0x12 | **JLT** | I | DestReg, Offset | Jump if LessThan flag set |
-| 0x13 | **JLE** | I | DestReg, Offset | Jump if LessThan OR Equal flag set |
+| 0x0D | **JMP** | I | Offset, BaseReg | Unconditional jump to (BaseReg + Offset) |
+| 0x0E | **JEQ** | I | Offset, BaseReg | Jump if Equal flag set |
+| 0x0F | **JNE** | I | Offset, BaseReg | Jump if Equal flag NOT set |
+| 0x10 | **JGT** | I | Offset, BaseReg | Jump if GreaterThan flag set |
+| 0x11 | **JGE** | I | Offset, BaseReg | Jump if GreaterThan OR Equal flag set |
+| 0x12 | **JLT** | I | Offset, BaseReg | Jump if LessThan flag set |
+| 0x13 | **JLE** | I | Offset, BaseReg | Jump if LessThan OR Equal flag set |
 
 **Jump Calculation:**
-- Target address = DestReg + SignExtend(Offset)
-- If DestReg = zero, this becomes absolute addressing with Offset
-- If DestReg = pcx, this becomes PC-relative addressing
+- Target address = BaseReg + SignExtend(Offset)
+- If BaseReg = zero, this becomes absolute addressing with Offset
+- If BaseReg = ret, this becomes return-style addressing
 - Conditional jumps check flags in STS register
 
+**Common Patterns:**
+```
+JMP label, zero     ; Absolute jump to label address
+JMP 0, ret          ; Jump to address in ret register
+JMP 4, ret          ; Jump to (ret + 4)
+```
+
 **Encoding Note:**
-In machine code: DestReg field, Offset (Immediate field) (SrcReg unused, set to noreg)
+In machine code: Offset (Immediate field), BaseReg (SrcReg field) (DestReg unused, set to noreg)
 
 ### Comparison
 
@@ -265,8 +275,8 @@ DestReg and ShiftAmt fields unused (set to noreg and 0)
 - Other flags undefined after arithmetic (use CMP for comparisons)
 
 **Encoding Notes:**
-- INC/DEC: Reg in SrcReg1 field, also copied to DestReg field
-- IADD/ISUB: Immediate is signed 16-bit value
+- INC/DEC: Reg in SrcReg1 field, DestReg set to noreg
+- IADD/ISUB: Immediate is signed 16-bit value, all three operands required
 
 ### Bitwise Logical Operations
 
@@ -285,7 +295,7 @@ DestReg and ShiftAmt fields unused (set to noreg and 0)
 - Other flags undefined
 
 **Encoding Note:**
-NOT uses only Src and Dest; SrcReg2 unused (set to noreg)
+NOT uses only Src (SrcReg1) and Dest (DestReg); SrcReg2 unused (set to noreg)
 
 ### Shift Operations
 
@@ -295,17 +305,22 @@ NOT uses only Src and Dest; SrcReg2 unused (set to noreg)
 | 0x18 | **SHR** | R | Reg, ShiftAmount | Shift Reg right by ShiftAmount bits<br/>Zero-fill from left (logical shift) |
 
 **Shift Amount:**
-- Can be a 5-bit literal (0-31) in ShiftAmt field
-- Can be a register value (low 5 bits used)
-  - If using register: Place in SrcReg2, set ShiftAmt to 0
-  - If using literal: Place in ShiftAmt field, set SrcReg2 to noreg
+- **Literal shifts**: ShiftAmount is a 5-bit literal (0-31) in assembly
+  - Stored in ShiftAmt field of instruction
+  - SrcReg2 set to noreg
+- **Register shifts**: ShiftAmount is a register containing shift value
+  - Register specified in SrcReg2 field
+  - ShiftAmt field must be 0
+  - Only low 5 bits of register value used
+
+**Note:** Current assembler implementation may only support literal shifts. Check assembler documentation.
 
 **Flag Effects:**
 - Zero flag set if result is zero
 
 **Encoding Notes:**
-- Reg in both SrcReg1 and DestReg fields
-- For literal shifts: ShiftAmt field contains shift count
+- Reg in both SrcReg1 and DestReg fields (shifted in place)
+- For literal shifts: ShiftAmt field contains shift count, SrcReg2 = noreg
 - For register shifts: SrcReg2 contains register, ShiftAmt must be 0
 
 ### System and Control Instructions
@@ -331,6 +346,17 @@ NOT uses only Src and Dest; SrcReg2 unused (set to noreg)
 - INT: InterruptCode in low 8 bits of Immediate field
 - IRT/HLT: All register fields set to noreg, ShiftAmt to 0
 
+### Meta Instructions (Assembler/Linker)
+
+These instructions are used by the assembler and linker but may not represent real CPU operations.
+
+| Hex | Mnemonic | Description |
+|-----|----------|-------------|
+| 0x27 | **SEGMENT** | Segment marker (implementation-specific) |
+| 0x3E | **DATA** | Raw data embedding |
+
+**Note:** The SEGMENT instruction opcode may vary between implementations (0x27 in assembler, 0x3F in some contexts). Consult your specific toolchain documentation.
+
 ## Instruction Summary Table
 
 | Opcode | Mnemonic | Type | Category |
@@ -374,6 +400,8 @@ NOT uses only Src and Dest; SrcReg2 unused (set to noreg)
 | 0x24 | HLT | R | System |
 | 0x25 | IADD | I | Arithmetic |
 | 0x26 | ISUB | I | Arithmetic |
+| 0x27 | SEGMENT | - | Meta |
+| 0x3E | DATA | - | Meta |
 
 ## Exception Conditions
 
@@ -393,9 +421,9 @@ See the DSA Assembly Language Reference for the complete calling convention and
 ## Notes on Design
 
 1. **Word Size:** All addresses and general computation is 32-bit
-2. **Endianness:** Little-endian byte order
-3. **Stack Growth:** Stack grows upward (incrementing addresses)
+2. **Endianness:** Little-endian for instructions and runtime data; assembler data directives may use big-endian
+3. **Stack Growth:** Stack grows **downward** (toward lower addresses) - PUSH decrements SPR
 4. **Alignment:** Natural alignment required for halfword and word accesses
 5. **Sign Extension:** All immediate values are sign-extended unless noted
 6. **Zero Register:** Provides constant zero, writes are legal but discarded
-7. **Reserved Encodings:** Opcodes 0x27-0x3F reserved for future use
+7. **Reserved Encodings:** Opcodes 0x27-0x3D and 0x3F reserved or implementation-specific

docs/DSA_Project_Roadmap.md

@@ -263,12 +263,12 @@
   - [ ] Array syntax
   - [ ] Struct syntax
   - [x] Pointer syntax
-  - [ ] Namespaced call syntax
+  - [x] Namespaced call syntax
 - [x] AST node definitions
 - [ ] Error recovery mechanisms
 - [ ] Comprehensive parser tests
 - [ ] Syntax error message quality testing
-- [ ] Implement C frontend by moving lexer/parser from `c_compiler` to the new `compiler` project structure
+- [x] Implement C frontend by moving lexer/parser from `c_compiler` to the new `compiler` project structure
 - [ ] Evaluate possible memory management strategies (e.g., keep all variables on the stack vs spill only when calling functions)
 
 ---
@@ -290,7 +290,7 @@
 - [ ] Optimize register allocation further
 - [x] Implement proper function calling conventions
 - [ ] Add constant folding optimization
-- [ ] Dead code elimination
+- [x] Dead code elimination
 - [ ] Test each feature thoroughly
 
 ---
@@ -376,7 +376,7 @@
 **Dependencies:** None  
 **Deliverable:** `docs/build-system-design.md`
 
-- [ ] Define project structure conventions
+- [x] Define project structure conventions
 - [ ] Design build manifest format (`dsa-project.toml` or similar)
 - [ ] Dependency resolution strategy
 - [ ] Build cache design
@@ -391,12 +391,12 @@
 **Dependencies:** 3.1.1, 1.2.2, 1.1.3, 2.1.3  
 **Deliverable:** `dsa-build` executable
 
-- [ ] Create crate: `dsa-build`
+- [x] Create crate: `dsa-build`
 - [ ] Manifest parser
 - [ ] Dependency graph builder
 - [ ] Task orchestrator
-  - [ ] Compilation tasks
-  - [ ] Assembly tasks
+  - [x] Compilation tasks
+  - [x] Assembly tasks
   - [ ] Linking tasks
 - [ ] Build cache implementation
 - [ ] Parallel build support
@@ -412,11 +412,11 @@
 **Dependencies:** 3.1.2  
 **Deliverable:** Enhanced `dsa-build` with project management
 
-- [ ] `dsa new <project>` — Create new project
-- [ ] `dsa init` — Initialize in existing directory
+- [x] `dsa new <project>` — Create new project
+- [x] `dsa init` — Initialize in existing directory
 - [ ] `dsa add <dependency>` — Add dependency
 - [ ] Binary vs library project types
-- [ ] Template system for project scaffolding
+- [x] Template system for project scaffolding
 - [ ] Documentation for each command
 
 ---

docs/ISA REVISION notes.md

@@ -0,0 +1,4 @@
+
+- we definitely need to be able to use registers for shift operations.
+- we need logical boolean operations in addition to the bitwise ones.
+- better conditionals.

docs/todo.md

@@ -0,0 +1,26 @@
+# General TODO's
+
+# Bugfixes
+- [x] [EASY] Investigate logical and operator not compiling - either a lexer or parser issue.
+  - **note**: this was a parser issue.
+
+# Missing features
+- [x] [MEDIUM] Get shift operations working correctly.
+- [ ] [MEDIUM] proper prefix/postfix inc/dec implementation. slightly more complex as we need to check for a variable and modify it in place
+- [ ] [EASY] Add multiply and divide operations to code generation
+  - **note**: very easy to do but our division algorithm is hopelessly slow so not worth doing for now.
+
+# Performance Improvements
+- [ ] [MEDIUM] implement a proper div/mod library that's not slow af.
+- [ ] [HARD] Immediate operations for values that support it (up to +/- u16::max for addi and subi respectively) 
+  - this requires significant complexity in code generation as we need to traverse down the tree when we come across these operations to prevent additional register allocations.
+
+# Compiler optimisations
+
+# Codegen improvements
+- [ ] [MEDIUM / time consuming] Add scoping to code generation
+- [ ] [MEDIUM / time consuming] Rewrite entire codegen to imrpove code quality and make the code more readable.
+  - [ ] type-safe instruction builder
+    - [ ] Instruction & Register enums
+    - [ ] Instruction builder helper fns eg `fn add(left: &Register, right: &Register, dest: &Register) -> Instruction`
+    - [ ] Instruction Block types.

dsx-build/src/templates.rs

@@ -0,0 +1,589 @@
+pub trait Template {
+    fn lib(project: &str) -> String;
+    fn bin(project: &str) -> String;
+
+    fn create(project: &str, lib: bool) -> String {
+        if lib {
+            Self::lib(project)
+        } else {
+            Self::bin(project)
+        }
+    }
+}
+
+pub struct Dsa;
+pub struct Dsc;
+
+impl Template for Dsa {
+    fn lib(project: &str) -> String {
+        format!(
+            r#"//
+lib.dsa
+// usage:
+//
+// include {project} "<relative path>"
+//
+// usage for {project}_main:
+// push (arg1)
+// push (arg0)
+// call {project}::{project}_main
+// pop (arg0)
+// pop (arg1)
+
+// Example data declarations
+// dw example_data: 0x0000
+
+// Main function template
+{project}_main:
+    // the correct way to start a function as defined by the calling convention
+    push bpr
+    mov spr, bpr
+
+    // explanation of how to access args
+    ldw bpr, rg0, 8  // arg 0
+    ldw bpr, rg0, 12 // arg 1
+
+    // your code goes here
+    // Example: load example_data into rg1
+    // ldw example_data, rg1
+
+    // the correct way to end a function as defined by the calling convention
+    mov bpr, spr
+    pop bpr
+    return
+"#,
+        )
+    }
+
+    fn bin(project: &str) -> String {
+        format!(
+            r#"
+// GENERATED BY DSX-BUILD
+// Generated at: {timestamp}
+// Project name: {project}
+
+// Imports
+include print: "./lib/print.dsa"
+
+// Globals & Reserved Memory
+dw stack: 0x10000
+db message: "Process Exited with code:"
+
+// Entry Point
+_init:
+    ldw stack, bpr
+    mov bpr, spr
+    push zero
+    call main
+    call print::print_newline
+    lwi message, rg0
+    push rg0
+    call print::print
+    pop zero
+    call print::print_hex_word
+    pop zero
+    hlt
+
+main:
+    push bpr
+    mov spr, bpr
+
+    // Your code goes here
+
+    // Return zero
+    stw zero, bpr, 8
+
+    mov bpr, spr
+    pop bpr
+    return"#,
+            timestamp = chrono::Utc::now().format("%Y-%m-%d %H:%M:%S").to_string()
+        )
+    }
+}
+
+impl Template for Dsc {
+    fn lib(project: &str) -> String {
+        format!(
+            r#"
+// GENERATED BY DSX-BUILD
+// Generated at: {timestamp}
+// Project name: {project}
+
+// Imports
+include print: "./lib/print.dsa";
+
+// Main Function
+fn {project}_main() -> u32 {{
+    return 0;
+}}"#,
+            timestamp = chrono::Utc::now().format("%Y-%m-%d %H:%M:%S").to_string()
+        )
+    }
+
+    fn bin(project: &str) -> String {
+        format!(
+            r#"
+// GENERATED BY DSX-BUILD
+// Generated at: {timestamp}
+// Project name: {project}
+
+// Imports
+include print: "./lib/print.dsa";
+
+// Main Function
+fn main() -> u32 {{
+    return 0;
+}}"#,
+            timestamp = chrono::Utc::now().format("%Y-%m-%d %H:%M:%S").to_string()
+        )
+    }
+}
+
+pub fn create_print_lib() -> String {
+    format!(
+        r#"
+// lib:
+// print.dsa
+
+// usage:
+//
+// include print "<relative path>""
+//
+// usage for print:
+//      push (register containing address of string)
+//      push pcx
+//      jmp print::print
+//
+// usage for reset:
+//      push pcx
+//      jmp print::reset
+//
+// usage for clear:
+//      push pcx
+//      jmp print::clear
+//
+// usage for print_byte:
+//      push (register containing byte)
+//      push pcx
+//      jmp print::print_byte
+//
+// usage for print_word:
+//      push (register containing word)
+//      push pcx
+//      jmp print::print_word
+//
+// usage for print_num:
+//      push (register containing number to print in decimal)
+//      push pcx
+//      jmp print::print_num
+//
+
+include maths "./maths.dsa"
+
+dw display: 0x20000
+dw current: 0x20000
+
+// ------------------------------------------
+// prints the string at addr(arg[0]) to the screen. (no trailing whitespace unless explicitly provided)
+print:
+    push bpr
+    mov spr, bpr
+
+    ldw bpr, rg0, 8
+    ldw current, rg1
+
+_print_loop:
+    ldb rg0, acc
+    cmp acc, zero
+    jeq _end
+    stb acc, rg1
+
+    addi rg0, 1
+    addi rg1, 1
+
+    jmp _print_loop
+
+// ------------------------------------------
+println:
+    push bpr
+    mov spr, bpr
+
+    ldw bpr, rg0, 8
+    ldw current, rg1
+
+_println_loop:
+    ldb rg0, acc
+    cmp acc, zero
+    jeq _println_end
+    stb acc, rg1
+
+    addi rg0, 1
+    addi rg1, 1
+
+    jmp _println_loop
+
+_println_end:
+    call print_newline
+    jmp _end
+
+// ------------------------------------------
+// prints the value of arg[0] to the screen.
+print_word:
+    // initialise
+    push bpr
+    mov spr, bpr
+
+    // load byte into acc
+    ldw bpr, rg0, 8
+    ldw current, rg1
+
+    addi rg1, 3
+
+    stb rg0, rg1
+    subi rg1, 1
+    shr rg0, 8
+    stb rg0, rg1
+    subi rg1, 1
+    shr rg0, 8
+    stb rg0, rg1
+    subi rg1, 1
+    shr rg0, 8
+    stb rg0, rg1
+
+    addi rg1, 4
+    jmp _end
+
+// ------------------------------------------
+// prints the last byte of arg[0] to the screen.
+print_byte:
+    push bpr
+    mov spr, bpr
+
+    ldw bpr, rg0, 8
+    ldw current, rg1
+
+    stb rg0, rg1
+    addi rg1, 1
+    jmp _end
+
+// ------------------------------------------
+// prints the value of arg[0] to the screen in hex.
+print_hex_word:
+    push bpr
+    mov spr, bpr
+
+    ldw current, rg1
+
+    ldb bpr, rg0, 8
+    push rg0
+    call _print_hex_byte
+    addi spr, 4
+
+    ldb bpr, rg0, 9
+    push rg0
+    call _print_hex_byte
+    addi spr, 4
+
+    ldb bpr, rg0, 10
+    push rg0
+    call _print_hex_byte
+    addi spr, 4
+
+    ldb bpr, rg0, 11
+    push rg0
+    call _print_hex_byte
+    addi spr, 4
+
+    jmp _end
+
+// ------------------------------------------
+// prints the last byte of arg[0] to the screen in hex.
+print_hex_byte:
+    push bpr
+    mov spr, bpr
+
+    ldw bpr, rg0, 8
+    ldw current, rg1
+
+    call _print_hex_byte
+    jmp _end
+
+// function body
+_print_hex_byte:
+    // mask to get lower nibble
+    lli 0xF, rg2
+    // save rg0 state
+    push rg0
+
+    shr rg0, 4
+    and rg0, rg2, rg0
+    call _print_hex_nibble
+    pop rg0
+
+    and rg0, rg2, rg0
+    call _print_hex_nibble
+    return
+
+// print a hex digit
+_print_hex_nibble:
+    lli 10, rg3
+    cmp rg0, rg3
+    jlt _print_hex_nibble_number
+    addi rg0, 0x37, rg0
+    stb rg0, rg1
+    addi rg1, 1
+    return
+
+// helper function.
+_print_hex_nibble_number:
+    addi rg0, 0x30, rg0
+    stb rg0, rg1
+    addi rg1, 1
+    return
+
+// ------------------------------------------
+// print whitespace
+print_whitespace:
+    push bpr
+    mov spr, bpr
+
+    ldw current, rg1
+    lli 0x20, rg0
+    stb rg0, rg1
+    addi rg1, 1
+    jmp _end
+
+// ------------------------------------------
+// print newline
+print_newline:
+    push bpr
+    mov spr, bpr
+
+    // load variables into registers
+    ldw display, rg0
+    ldw current, rg1
+
+    // get the offset from the display base
+    sub rg1, rg0, rg0
+
+    lwi 80, rg2
+    pusha 3
+    push rg0
+    push rg2
+    call maths::divmod
+    pop zero  // result
+    pop rg3 // remainder
+    popa 3
+
+    sub rg1, rg3, rg2
+    addi rg2, 80, rg1
+
+    // _end saves the display state
+    jmp _end
+
+// ------------------------------------------
+// prints arg[0] as a decimal number to the screen.
+print_num:
+    push bpr
+    mov spr, bpr
+
+    ldw bpr, rg0, 8      // load number to print
+    lli 0, rg5           // rg5 = digit counter
+
+    // check if number is zero
+    cmp rg0, zero
+    jne _print_num_extract_digits
+
+    // special case: print '0' for zero
+    lli 0x30, rg6
+    push rg6             // push digit to stack buffer
+    lli 1, rg5           // we have 1 digit
+    jmp _print_num_output
+
+_print_num_extract_digits:
+    // divide by 10 repeatedly to get digits
+    cmp rg0, zero
+    jeq _print_num_output
+
+    // call divmod(rg0, 10)
+    push rg0             // dividend
+    lli 10, rg1
+    push rg1             // divisor (10)
+    call maths::divmod
+    pop rg0              // quotient (continue dividing this)
+    pop rg1              // remainder (the digit)
+
+    // convert digit to ASCII and push to stack buffer
+    addi rg1, 0x30, rg6  // convert to ASCII
+    push rg6             // push digit to stack
+    inc rg5              // increment digit counter
+
+    jmp _print_num_extract_digits
+
+_print_num_output:
+    // now print digits (pop them off in reverse order)
+    ldw current, rg1     // get display pointer
+
+_print_num_output_loop:
+    // check if we've printed all digits
+    cmp rg5, zero
+    jeq _print_num_done
+
+    // pop digit and print it
+    pop rg6
+    stb rg6, rg1
+    addi rg1, 1
+    dec rg5
+
+    jmp _print_num_output_loop
+
+_print_num_done:
+    jmp _end
+
+// ------------------------------------------
+// resets the cursor position on the screen to 0x20000. (0,0)
+reset:
+    push bpr
+    mov spr, bpr
+    ldw display, rg1
+    jmp _end
+
+// ------------------------------------------
+// clears the screen
+clear:
+    push bpr
+    mov spr, bpr
+    // display size = 2000 bytes / 500 words
+    lli 500 rg0
+    ldw display, rg1
+
+_clear_loop:
+    dec rg0
+    stw zero, rg1
+    addi rg1, 4
+    cmp rg0, zero
+    jgt _clear_loop
+    jmp _end
+
+// ------------------------------------------
+// return
+_end:
+    stw rg1, current
+
+    mov bpr, spr
+    pop bpr
+    return
+"#
+    )
+}
+
+pub fn create_maths_lib() -> String {
+    format!(
+        r#"
+// multiply.dsa
+// usage:
+//
+// include multiply "<relative path>"
+//
+// usage for multiply:
+// push (arg1)
+// push (arg0)
+// call multiply::multiply
+// pop (arg0)
+// pop (arg1)
+
+multiply:
+    push bpr
+    mov spr, bpr
+
+    ldw bpr, rg0, 8  // load op 2
+    ldw bpr, rg1, 12 // load op 1
+    lwi 0, rg2       // initialise rg2 to zero
+
+_multiply_loop:
+    add rg2, rg0, rg2
+    dec rg1
+
+    cmp rg1, zero
+    jgt _multiply_loop
+
+_multiply_end:
+    stw rg2, bpr, 8
+
+    mov bpr, spr
+    pop bpr
+    return
+
+divmod:
+    push bpr
+    mov spr, bpr
+
+    ldw bpr, rg1, 8  // load op 2
+    ldw bpr, rg0, 12 // load op 1
+
+    lli 0, rg3
+
+_divmod_loop:
+    cmp rg0, rg1
+    jlt _divmod_end
+
+    sub rg0, rg1, rg0
+    inc rg3
+
+    jmp _divmod_loop
+
+_divmod_end:
+    // store div in first arg
+    // store mod in second arg
+    stw rg3, bpr, 8
+    stw rg0, bpr, 12
+
+    mov bpr, spr
+    pop bpr
+    return
+
+// multiply.dsa - improved version
+// Multiplies two 32-bit numbers using shift-and-add
+//
+// Usage:
+// push operand2  (multiplier)
+// push operand1  (multiplicand)
+// call multiply::multiply
+// pop result
+// pop zero       (discard second argument)
+
+new_multiply:
+    push bpr
+    mov spr, bpr
+
+    ldw bpr, rg0, 8     // rg0 = multiplicand
+    ldw bpr, rg1, 12    // rg1 = multiplier
+
+    lli 0, rg2          // rg2 = result (accumulator)
+    lli 32, rg3         // rg3 = bit counter
+
+mult_loop:
+    // Check if lowest bit of multiplier is 1
+    lli 1, acc
+    and rg1, acc, acc   // acc = rg1 & 1
+    cmp acc, zero
+    jeq skip_add        // if (rg1 & 1) == 0, skip addition
+
+    // Add multiplicand to result
+    add rg2, rg0, rg2
+
+skip_add:
+    shl rg0, 1          // shift multiplicand left
+    shr rg1, 1          // shift multiplier right
+
+    dec rg3
+    cmp rg3, zero
+    jgt mult_loop
+
+    stw rg2, bpr, 8     // store result
+    mov bpr, spr
+    pop bpr
+    return
+"#
+    )
+}

emulator/Cargo.toml

@@ -20,10 +20,11 @@ compiler = { path = "../compiler" }
 dsa_editor = { path = "../dsa_editor" }
 egui = "0.31.1"
 dirs = "6.0.0"
-discord-presence = { version = "1.6.0", optional = true }
+discord-presence = { version = "2.0.0", optional = true }
 toml = { version = "0.8.23", optional = true }
 serde = { version = "1.0.219", features = ["derive"], optional = true }
 egui_file = "0.22.1"
+rustc-hash = "2.1.1"
 
 [features]
 default = ["config"]

emulator/src/emulator/system/cache.rs

@@ -0,0 +1,53 @@
+use common::prelude::Instruction;
+use rustc_hash::FxHashMap;
+
+#[derive(Debug)]
+pub struct Cache {
+    addr: u32,
+    instruction_block: Option<[u8; 256]>,
+    instruction_lookup: FxHashMap<u32, Instruction>,
+}
+
+impl Cache {
+    #[must_use]
+    pub fn new() -> Self {
+        Self {
+            addr: 0,
+            instruction_block: None,
+            instruction_lookup: FxHashMap::default(),
+        }
+    }
+
+    pub fn lookup_value(&mut self, addr: u32) -> Option<u32> {
+        if addr < self.addr || addr >= self.addr + 256 || self.instruction_block.is_none()
+        {
+            return None;
+        }
+
+        Some(u32::from_be_bytes(
+            self.instruction_block.expect("this should not be none!")
+                [(addr - self.addr) as usize..(addr - self.addr + 4) as usize]
+                .try_into()
+                .expect("Failed to convert bytes to u32"),
+        ))
+    }
+
+    pub const fn set(&mut self, addr: u32, block: &[u8; 256]) {
+        self.addr = addr - addr % 256;
+        self.instruction_block = Some(*block);
+    }
+
+    pub fn lookup_instruction(&mut self, instruction: u32) -> Option<Instruction> {
+        self.instruction_lookup.get(&instruction).copied()
+    }
+
+    pub fn insert(&mut self, value: u32, instruction: Instruction) {
+        self.instruction_lookup.insert(value, instruction);
+    }
+}
+
+impl Default for Cache {
+    fn default() -> Self {
+        Self::new()
+    }
+}

emulator/src/emulator/system/emulator.rs

@@ -25,9 +25,11 @@ pub fn run_emulator(
     let mut running = Running::Paused;
     let mut step = 0;
     let mut addr;
-    let mut history = Vec::<(u32, Instruction)>::new();
+    let mut history = Vec::<(u32, u32)>::with_capacity(32768);
     let size = 256;
 
+    let record_history = true;
+
     state_tx
         .send(StateUpdate::Running(Running::Paused))
         .expect("Failed to send initial state!");
@@ -36,7 +38,9 @@ pub fn run_emulator(
     let mut update = false;
 
     loop {
-        let cmd = if running == Running::Running || step > 0 {
+        let cmd = if step > 0 {
+            None
+        } else if running == Running::Running && step == 0 {
             match cmd_rx.try_recv() {
                 Ok(cmd) => Some(cmd),
                 Err(mpsc::TryRecvError::Empty) => {
@@ -52,10 +56,15 @@ pub fn run_emulator(
             }
         };
 
+        if running == Running::Running && step == 0 {
+            step = 32768;
+        }
+
         if let Some(cmd) = cmd {
             match cmd {
                 Command::Start => {
                     running = Running::Running;
+                    step = 32768;
 
                     // Update RPC with current state. TODO: Make this only occur on state
                     // changes.
@@ -71,9 +80,11 @@ pub fn run_emulator(
                 }
                 Command::Stop => {
                     running = Running::Paused;
+                    step = 0;
                 }
                 Command::Reset(x) => {
                     running = Running::Paused;
+                    step = 0;
 
                     match x {
                         0 => {
@@ -95,20 +106,12 @@ pub fn run_emulator(
                 }
                 Command::Step(x) => {
                     step = x;
+                    running = Running::Paused;
                 }
                 Command::Write(offset, data) => {
                     update = true;
 
-                    processor
-                        .memory
-                        .write_range(offset, data)
-                        .unwrap_or_else(|_| {
-                            report_err(
-                                state_tx,
-                                "Failed to write memory range!",
-                                &mut processor,
-                            );
-                        });
+                    processor.memory.write_range(offset, data);
                 }
                 Command::Interrupt(_interrupt) => {
                     update = true;
@@ -118,14 +121,7 @@ pub fn run_emulator(
                 Command::MemRequest(new, size) if update => {
                     addr = new;
                     let _ = state_tx.send(StateUpdate::MemoryView(
-                        processor.memory.read_range(addr, size).unwrap_or_else(|_| {
-                            report_err(
-                                state_tx,
-                                "Failed to read memory range!",
-                                &mut processor,
-                            );
-                            Vec::new()
-                        }),
+                        processor.memory.read_range(addr, size),
                     ));
                 }
                 Command::DisplayRequest if update => {
@@ -163,50 +159,19 @@ pub fn run_emulator(
                     let _ = state_tx.send(StateUpdate::Instructions(instruction_count));
                 }
                 Command::WriteBlock(addr, block) => {
-                    processor
-                        .memory
-                        .write_range(addr, block.to_vec())
-                        .unwrap_or_else(|_| {
-                            report_err(
-                                state_tx,
-                                "Failed to write memory block!",
-                                &mut processor,
-                            );
-                        });
+                    processor.memory.write_range(addr, block.to_vec());
                 }
 
                 _ => {}
             }
         }
 
-        if step > 0 {
-            step -= 1;
-            update = true;
-            running = Running::Paused;
-
-            // Execute one cycle.
-            match processor.cycle() {
-                Ok((addr, instruction)) => {
-                    history.push((addr, instruction));
-                }
-                Err(why) => {
-                    let pcx = processor
-                        .get(Register::Pcx)
-                        .expect("SPR should never be invalid");
-                    report_err(
-                        state_tx,
-                        &format!(
-                            "Could not decode instruction at {pcx:x}. Reason: {why}"
-                        ),
-                        &mut processor,
-                    );
-                }
-            }
-            instruction_count += 1;
-            continue;
+        if running == Running::Running {
+            step += 1;
         }
 
-        if running == Running::Running {
+        if step > 0 {
+            step -= 1;
             update = true;
 
             // Execute one cycle.
@@ -227,9 +192,18 @@ pub fn run_emulator(
                 }
             };
 
-            history.push(instruction);
-            if matches!(instruction.1, Instruction::Halt) {
+            if record_history {
+                history.push((
+                    instruction.0,
+                    processor
+                        .get(Register::Cir)
+                        .expect("CIR should never be invalid"),
+                ));
+            }
+
+            if matches!(instruction, (_, Instruction::Halt)) {
                 running = Running::Halted;
+                step = 0;
             }
 
             instruction_count += 1;

emulator/src/emulator/system/memory.rs

@@ -1,52 +1,42 @@
-use std::collections::HashMap;
+use rustc_hash::FxHashMap;
 
 use crate::emulator::system::model::ProcessorError;
 
 pub trait MemoryUnit: Send + Sync {
     fn reset(&mut self);
-    fn read_byte(&mut self, addr: u32) -> Result<u8, ProcessorError>;
-    fn write_byte(&mut self, addr: u32, value: u8) -> Result<(), ProcessorError>;
+    fn read_byte(&mut self, addr: u32) -> u8;
+    fn write_byte(&mut self, addr: u32, value: u8);
     fn read_word(&mut self, addr: u32) -> Result<u32, ProcessorError>;
     fn write_word(&mut self, addr: u32, value: u32) -> Result<(), ProcessorError>;
 
-    fn read_range(&mut self, addr: u32, size: u32) -> Result<Vec<u8>, ProcessorError> {
+    fn read_range(&mut self, addr: u32, size: u32) -> Vec<u8> {
         let mut data = Vec::with_capacity(size as usize);
         for i in 0..size {
-            data.push(self.read_byte(addr + i)?);
+            data.push(self.read_byte(addr + i));
         }
-        Ok(data)
+        data
     }
 
-    fn write_range(&mut self, addr: u32, value: Vec<u8>) -> Result<(), ProcessorError> {
+    fn write_range(&mut self, addr: u32, value: Vec<u8>) {
         for (i, byte) in value.into_iter().enumerate() {
-            self.write_byte(addr + i as u32, byte)?;
+            self.write_byte(addr + i as u32, byte);
         }
-        Ok(())
     }
 
-    fn read_block(&mut self, addr: u32) -> Result<[u8; 256], ProcessorError> {
-        let mut data = [0; 256];
-        for (i, byte) in data.iter_mut().enumerate() {
-            *byte = self.read_byte(addr + i as u32)?;
-        }
-        Ok(data)
-    }
+    fn read_block(&mut self, addr: u32) -> &[u8; 256];
 
-    fn write_block(&mut self, addr: u32, data: [u8; 256]) -> Result<(), ProcessorError> {
+    fn write_block(&mut self, addr: u32, data: &[u8; 256]) {
         for (i, byte) in data.iter().enumerate() {
-            self.write_byte(addr + i as u32, *byte)?;
+            self.write_byte(addr + i as u32, *byte);
         }
-        Ok(())
     }
 }
 
 pub struct MainStore {
-    pub data: HashMap<u32, Block>,
+    pub data: FxHashMap<u32, Block>,
 }
 
-pub struct Block {
-    data: [u8; 256],
-}
+pub type Block = [u8; 256];
 
 impl Default for MainStore {
     fn default() -> Self {
@@ -58,113 +48,110 @@ impl MainStore {
     #[must_use]
     pub fn new() -> Self {
         Self {
-            data: HashMap::new(),
+            data: FxHashMap::default(),
         }
     }
 
+    #[inline]
     const fn segment_addr(addr: u32) -> (u32, u8) {
         (addr / 256, (addr % 256) as u8)
     }
 
+    #[inline]
     fn mut_block(&mut self, addr: u32) -> &mut Block {
-        self.data
-            .entry(addr)
-            .or_insert_with(|| Block { data: [0; 256] });
-
-        self.data.get_mut(&addr).map_or_else(
-            || panic!("Could not fetch block with address {addr:x?}"),
-            |block| block,
-        )
+        self.data.entry(addr).or_insert([0; 256])
     }
 
+    #[inline]
     fn block(&mut self, addr: u32) -> &Block {
-        self.data
-            .entry(addr)
-            .or_insert_with(|| Block { data: [0; 256] });
-
-        self.data.get(&addr).map_or_else(
-            || panic!("Could not fetch block with address {addr:x?}"),
-            |block| block,
-        )
+        self.data.entry(addr).or_insert([0; 256])
     }
 }
 
 impl MemoryUnit for MainStore {
+    #[inline]
     fn reset(&mut self) {
         self.data.clear();
     }
 
-    fn read_byte(&mut self, addr: u32) -> Result<u8, ProcessorError> {
+    #[inline]
+    fn read_byte(&mut self, addr: u32) -> u8 {
         let (block_addr, offset) = Self::segment_addr(addr);
         let block = self.block(block_addr);
-        Ok(block.data[offset as usize])
+        block[offset as usize]
     }
 
+    #[inline]
     fn read_word(&mut self, addr: u32) -> Result<u32, ProcessorError> {
-        if addr % 4 != 0 {
+        if !addr.is_multiple_of(4) {
             return Err(ProcessorError::BadMemoryAccess(addr));
         }
 
         let (block_addr, offset) = Self::segment_addr(addr);
-        let block = self.mut_block(block_addr);
-        let mut bytes = [0; 4];
-        bytes[0] = block.data[offset as usize];
-        bytes[1] = block.data[(offset + 1) as usize];
-        bytes[2] = block.data[(offset + 2) as usize];
-        bytes[3] = block.data[(offset + 3) as usize];
-        Ok(u32::from_be_bytes(bytes))
+        let offset = offset as usize;
+        let block = self.block(block_addr);
+        Ok(u32::from_be_bytes(
+            block[offset..=offset + 3]
+                .try_into()
+                .expect("Failed to read word!"),
+        ))
     }
 
-    fn read_range(&mut self, addr: u32, size: u32) -> Result<Vec<u8>, ProcessorError> {
+    #[inline]
+    fn read_range(&mut self, addr: u32, size: u32) -> Vec<u8> {
         let mut data = Vec::with_capacity(size as usize);
         for i in 0..size {
-            data.push(self.read_byte(addr + i)?);
+            data.push(self.read_byte(addr + i));
         }
 
-        Ok(data)
+        data
     }
 
-    fn write_byte(&mut self, addr: u32, value: u8) -> Result<(), ProcessorError> {
+    #[inline]
+    fn write_byte(&mut self, addr: u32, value: u8) {
         let (block_addr, offset) = Self::segment_addr(addr);
         let block = self.mut_block(block_addr);
-        block.data[offset as usize] = value;
-        Ok(())
+        block[offset as usize] = value;
     }
 
+    #[inline]
     fn write_word(&mut self, addr: u32, value: u32) -> Result<(), ProcessorError> {
-        if addr % 4 != 0 {
+        if !addr.is_multiple_of(4) {
             return Err(ProcessorError::BadMemoryAccess(addr));
         }
 
         let (block_addr, offset) = Self::segment_addr(addr);
         let block = self.mut_block(block_addr);
-        block.data[offset as usize] = (value >> 24) as u8;
-        block.data[(offset + 1) as usize] = (value >> 16) as u8;
-        block.data[(offset + 2) as usize] = (value >> 8) as u8;
-        block.data[(offset + 3) as usize] = value as u8;
+        block[offset as usize..=(offset + 3) as usize]
+            .copy_from_slice(&value.to_be_bytes());
         Ok(())
     }
 
-    fn write_range(&mut self, addr: u32, value: Vec<u8>) -> Result<(), ProcessorError> {
-        for (i, byte) in value.into_iter().enumerate() {
-            let (block_addr, offset) = Self::segment_addr(addr + i as u32);
-            let block = self.mut_block(block_addr);
-            block.data[offset as usize] = byte;
+    #[inline]
+    fn write_range(&mut self, addr: u32, value: Vec<u8>) {
+        let mut current_block_addr = addr / 256;
+        let mut current_block = self.mut_block(current_block_addr);
+        let mut offset = addr % 256;
+        for byte in value {
+            current_block[offset as usize] = byte;
+            offset += 1;
+            if offset >= 256 {
+                offset = 0;
+                current_block_addr += 1;
+                current_block = self.mut_block(current_block_addr);
+            }
         }
-
-        Ok(())
     }
 
-    fn read_block(&mut self, addr: u32) -> Result<[u8; 256], ProcessorError> {
+    #[inline]
+    fn read_block(&mut self, addr: u32) -> &[u8; 256] {
         let (block_addr, _) = Self::segment_addr(addr);
-        let block = self.block(block_addr);
-        Ok(block.data)
+        self.block(block_addr)
     }
 
-    fn write_block(&mut self, addr: u32, data: [u8; 256]) -> Result<(), ProcessorError> {
+    #[inline]
+    fn write_block(&mut self, addr: u32, data: &[u8; 256]) {
         let (block_addr, _) = Self::segment_addr(addr);
-        let block = self.mut_block(block_addr);
-        block.data = data;
-        Ok(())
+        let _ = self.data.insert(block_addr, *data);
     }
 }

emulator/src/emulator/system/mod.rs

@@ -1,3 +1,4 @@
+pub mod cache;
 pub mod emulator;
 pub mod memory;
 pub mod model;

emulator/src/emulator/system/model.rs

@@ -78,7 +78,7 @@ pub struct State {
 
     pub error_log: Vec<String>,
 
-    pub instruction_history: Vec<(u32, Instruction)>,
+    pub instruction_history: Vec<(u32, u32)>,
 }
 
 impl State {
@@ -154,7 +154,7 @@ pub enum StateUpdate {
     MemoryView(Vec<u8>),
     DisplayView(Vec<u8>),
     Error(String),
-    InstructionHistory(Vec<(u32, Instruction)>),
+    InstructionHistory(Vec<(u32, u32)>),
 }
 
 #[derive(Default, Debug, Clone, Copy, PartialEq, Eq)]
@@ -286,11 +286,10 @@ impl RegFile {
             Register::Sts => &mut self.sts,
             Register::Cir => &mut self.cir,
             Register::Pcx => &mut self.pcx,
-            _ => return Err(ProcessorError::InvalidRegister(Register::NoReg as u8)),
+            _ => return Err(ProcessorError::InvalidRegister(Register::Null as u8)),
         })
     }
 
-    #[must_use]
     pub const fn get(&self, reg: Register) -> Result<u32, ProcessorError> {
         Ok(match reg {
             Register::Rg0 => self.rg0,
@@ -321,7 +320,7 @@ impl RegFile {
             Register::Cir => self.cir,
             Register::Pcx => self.pcx,
             Register::Zero => 0,
-            _ => return Err(ProcessorError::InvalidRegister(Register::NoReg as u8)),
+            _ => return Err(ProcessorError::InvalidRegister(Register::Null as u8)),
         })
     }
 }

emulator/src/emulator/system/processor/mod.rs

@@ -4,6 +4,7 @@ use std::{
 };
 
 use crate::emulator::system::{
+    cache::Cache,
     memory::MemoryUnit,
     model::{IODevice, ProcessorError, RegFile},
 };
@@ -17,10 +18,7 @@ pub struct Processor {
     pub io_devices: Vec<Arc<dyn IODevice>>,
 
     pub void: u32,
-}
-
-fn log(message: &str) {
-    println!("\x1b[32mINFO:\x1b[0m {message}");
+    pub cache: Cache,
 }
 
 impl Processor {
@@ -32,6 +30,7 @@ impl Processor {
             halted: false,
             io_devices,
             void: 0,
+            cache: Cache::new(),
         }
     }
 
@@ -51,21 +50,35 @@ impl Processor {
         // Get value from PCX.
         let addr = self.fetch()?;
         // Increment PCX.
-        self.advance();
+        self.advance()?;
 
         // Set MAR to the previous value of PCX.
         *self.reg(Register::Mar)? = addr;
-        let val = self.memory.read_word(addr)?;
+
+        let encoded = if let Some(val) = self.cache.lookup_value(addr) {
+            val
+        } else {
+            let block = self.memory.read_block(addr);
+            self.cache.set(addr, block);
+            self.cache
+                .lookup_value(addr)
+                .expect("Failed to lookup value!")
+        };
 
         // Set CIR to the value of RAM[MAR].
-        *self.reg(Register::Mar)? = val;
+        *self.reg(Register::Cir)? = encoded;
 
-        // Decode and execute the instruction.
-        let instruction = Instruction::decode(val)
-            .map_err(|_| ProcessorError::InvalidInstruction(val))?;
+        let decoded = if let Some(val) = self.cache.lookup_instruction(addr) {
+            val
+        } else {
+            let decoded = Instruction::decode(encoded)
+                .map_err(|_| ProcessorError::InvalidInstruction(encoded))?;
+            self.cache.insert(addr, decoded);
+            decoded
+        };
 
-        instruction.execute(self)?;
-        Ok((addr, instruction))
+        decoded.execute(self)?;
+        Ok((addr, decoded))
     }
 
     const fn fetch(&self) -> Result<u32, ProcessorError> {
@@ -84,7 +97,7 @@ impl Processor {
     }
 
     pub fn display(&mut self) -> Result<Vec<u8>, ProcessorError> {
-        self.memory.read_range(0x20000, 2000)
+        Ok(self.memory.read_range(0x20000, 2000))
     }
 
     pub fn cmp(&mut self, a: u32, b: u32) {
@@ -163,10 +176,10 @@ impl Processor {
         let addr = self.get(Register::Spr)?;
         let size = n * 4;
         // returns the stack
-        self.memory.read_range(
+        Ok(self.memory.read_range(
             max(addr, 0),    // ensures that we cannot read from a negative address
             min(size, addr), // ensures we don't read above the top of the stack
-        )
+        ))
     }
 }
 
@@ -209,7 +222,7 @@ impl Executable for Instruction {
             Self::LoadByte(a) => {
                 *cpu.reg(a.r2)? = u32::from(
                     cpu.memory
-                        .read_byte(cpu.get(a.r1)? + u32::from(a.immediate))?,
+                        .read_byte(cpu.get(a.r1)? + u32::from(a.immediate)),
                 );
             }
 
@@ -218,7 +231,7 @@ impl Executable for Instruction {
             Self::LoadByteSigned(a) => {
                 *cpu.reg(a.r2)? = sign_extend(u32::from(
                     cpu.memory
-                        .read_byte(cpu.get(a.r1)? + u32::from(a.immediate))?,
+                        .read_byte(cpu.get(a.r1)? + u32::from(a.immediate)),
                 ));
             }
 
@@ -257,7 +270,7 @@ impl Executable for Instruction {
                 cpu.memory.write_byte(
                     cpu.get(a.r2)? + u32::from(a.immediate),
                     cpu.get(a.r1)? as u8,
-                )?;
+                );
             }
 
             // Stores a half-word from SrcReg in memory address (base + offset) The
@@ -266,9 +279,9 @@ impl Executable for Instruction {
                 // split the value into bytes and then write two bytes
                 let bytes = (cpu.get(a.r1)? as u16).to_le_bytes();
                 cpu.memory
-                    .write_byte(cpu.get(a.r2)? + u32::from(a.immediate), bytes[0])?;
+                    .write_byte(cpu.get(a.r2)? + u32::from(a.immediate), bytes[0]);
                 cpu.memory
-                    .write_byte(cpu.get(a.r2)? + u32::from(a.immediate) + 1, bytes[1])?;
+                    .write_byte(cpu.get(a.r2)? + u32::from(a.immediate) + 1, bytes[1]);
             }
 
             // Stores a word from SrcReg in memory address (base + offset) The effective
@@ -349,17 +362,13 @@ impl Executable for Instruction {
             // Left shifts the value in Reg by the given amount (either a register, or a
             // literal value)
             Self::ShiftLeft(a) => {
-                let reg = cpu.get(a.sr1)?;
-                let val = a.shamt;
-                *cpu.reg(a.sr1)? = shl(reg, val);
+                *cpu.reg(a.dr)? = shl(cpu.get(a.sr1)?, a.shamt + cpu.get(a.sr2)? as u8);
             }
 
             // Right shifts the value in Reg by the given amount (either a register, or a
             // literal value).
             Self::ShiftRight(a) => {
-                let regval = cpu.get(a.sr1)?;
-                let val = a.shamt;
-                *cpu.reg(a.sr1)? = shr(regval, val);
+                *cpu.reg(a.dr)? = shr(cpu.get(a.sr1)?, a.shamt + cpu.get(a.sr2)? as u8);
             }
 
             // Adds the value of Src2 to Src1 and writes the result to a.dr

emulator/src/emulator/system/processor/tests.rs

@@ -81,9 +81,7 @@ fn test_mov_signed_instruction() {
 fn test_load_byte_instruction() {
     let mut cpu = create_test_processor();
     let addr = 0x100;
-    cpu.memory
-        .write_byte(addr, 0xAB)
-        .expect("Failed to write byte to memory");
+    cpu.memory.write_byte(addr, 0xAB);
     *cpu.reg(Register::Rg1).expect("Failed to get register Rg1") = addr - 4;
 
     let load_byte_instr = Instruction::LoadByte(ITypeArgs::new(
@@ -105,9 +103,7 @@ fn test_load_byte_instruction() {
 fn test_load_byte_signed_instruction() {
     let mut cpu = create_test_processor();
     let addr = 0x100;
-    cpu.memory
-        .write_byte(addr, 0xFF)
-        .expect("Failed to write byte to memory");
+    cpu.memory.write_byte(addr, 0xFF);
     *cpu.reg(Register::Rg1).expect("Failed to get register Rg1") = addr;
 
     let load_byte_signed_instr = Instruction::LoadByteSigned(ITypeArgs::new(
@@ -189,7 +185,7 @@ fn test_store_byte_instruction() {
     store_byte_instr.execute(&mut cpu).expect(
         "Emulator was slain by losing the game while attempting to execute instruction",
     );
-    assert_eq!(cpu.memory.read_byte(addr).expect("Emulator was slain by losing the game while attempting to execute instruction"), 0xAB);
+    assert_eq!(cpu.memory.read_byte(addr), 0xAB);
 }
 
 #[test]
@@ -468,7 +464,7 @@ fn test_shift_left_with_shamt() {
     let shl_instr = Instruction::ShiftLeft(RTypeArgs::new(
         Some(Register::Rg1),
         Some(Register::Zero),
-        None,
+        Some(Register::Rg1),
         Some(2),
     ));
 
@@ -489,7 +485,7 @@ fn test_shift_right_with_shamt() {
     let shr_instr = Instruction::ShiftRight(RTypeArgs::new(
         Some(Register::Rg1),
         Some(Register::Zero),
-        None,
+        Some(Register::Rg1),
         Some(2),
     ));
 

emulator/src/emulator/ui/editor.rs

@@ -117,10 +117,7 @@ impl Editor {
                 .file_name()
                 .unwrap_or_else(|| OsStr::new("Unnamed!"))
                 .to_str()
-                .map_or_else(
-                    || unreachable!("File name should be valid UTF-8."),
-                    |ext| ext,
-                );
+                .unwrap_or_else(|| unreachable!("File name should be valid UTF-8."));
         }
         "Unnamed!"
     }
@@ -129,12 +126,9 @@ impl Editor {
         if let Some(path) = &self.path {
             return path
                 .extension()
-                .map_or_else(|| OsStr::new("Unknown!"), |ext| ext)
+                .unwrap_or_else(|| OsStr::new("Unknown!"))
                 .to_str()
-                .map_or_else(
-                    || unreachable!("File name should be valid UTF-8."),
-                    |ext| ext,
-                );
+                .unwrap_or_else(|| unreachable!("File name should be valid UTF-8."));
         }
         "Unknown!"
     }
@@ -398,7 +392,7 @@ impl Editor {
             _ => Syntax::default(),
         };
 
-        let ed = CodeEditor::default()
+        let mut editor = CodeEditor::default()
             .id_source("editor")
             .with_fontsize(12.0)
             .with_rows(0)
@@ -407,8 +401,6 @@ impl Editor {
             .with_numlines(true)
             .desired_width(available_width - 500.0);
 
-        let mut editor = ed.clone();
-
         editor.show(ui, &mut self.text);
     }
 
@@ -451,7 +443,7 @@ impl Editor {
                 Some("dsc") => {
                     let output_path = Path::new(path).with_extension("dsa");
                     if let Err(e) = compiler::compile_file(path, &output_path) {
-                        self.error = Some(format!("Compiler error: {}", e));
+                        self.error = Some(format!("Compiler error: {e}"));
                     }
 
                     let mut compiler = CompilerEngine::new();
@@ -461,7 +453,7 @@ impl Editor {
                     let instructions = match compiler.wait_for_result() {
                         Ok(instructions) => instructions,
                         Err(e) => {
-                            self.error = Some(format!("Assembler error: {}", e));
+                            self.error = Some(format!("Assembler error: {e}"));
                             return;
                         }
                     };

emulator/src/emulator/ui/history.rs

@@ -1,3 +1,4 @@
+use common::prelude::Instruction;
 use egui::{Context, Ui};
 
 use crate::emulator::{
@@ -57,8 +58,11 @@ impl Component for History {
                                     .color(egui::Color32::from_rgb(255, 200, 200)),
                             );
 
+                            let decoded = Instruction::decode(instruction.1)
+                                .unwrap_or(Instruction::Nop);
+
                             ui.label(
-                                egui::RichText::new(instruction.1.to_string())
+                                egui::RichText::new(decoded.to_string())
                                     .font(egui::FontId::monospace(12.0))
                                     .color(egui::Color32::from_rgb(200, 255, 200)),
                             );

emulator/src/emulator/ui/loader.rs

@@ -79,10 +79,7 @@ impl Loader {
                 .file_name()
                 .unwrap_or_else(|| OsStr::new("Unnamed!"))
                 .to_str()
-                .map_or_else(
-                    || unreachable!("File name should be valid UTF-8."),
-                    |ext| ext,
-                );
+                .unwrap_or_else(|| unreachable!("File name should be valid UTF-8."));
         }
         "Unnamed!"
     }

resources/dsa/example.dsa

@@ -1,121 +0,0 @@
-
-// GENERATED BY DSC COMPILER
-// Generated at 2026-02-05 00:42:40
-
-// Imports
-include print: "./lib/io/print.dsa"
-include arena: "./lib/memory/arena_alloc.dsa"
-
-// Globals & Reserved Memory
-
-
-// Entry Point
-dw stack: 0x10000
-db message: "Process Exited with code:"
-_init:
-	ldw stack, bpr
-	mov bpr, spr
-	push zero
-	call main
-	call print::print_newline
-	lwi message, rg0
-	push rg0
-	call print::print
-	pop zero
-	call print::print_hex_word
-	pop zero
-	hlt
-
-
-// Return
-_ret:
-	mov bpr, spr
-	pop bpr
-	return
-
-// Compiled Code Starts...
-main:
-	push bpr
-	mov spr, bpr
-
-	lli 0, rg0
-	push rg0 // bpr-4: x
-	subi bpr 4 rg1
-	lli 512, rg0
-	push rg1 // bpr-8: y
-	push rg0 // push arg 0
-	call arena::new
-	pop rg2
-	lli 32, rg0
-	push rg2 // bpr-12: alloc
-	push rg0 // push arg 1
-	push rg2 // push arg 0
-	call arena::alloc
-	pop rg3
-	pop zero
-	lli 32, rg0
-	subi bpr 12 rg2
-	ldw rg2, rg2 // bpr-20: alloc
-	push rg2 // bpr-16: alloc
-	push rg3 // bpr-20: ptr1
-	push rg0 // push arg 1
-	push rg2 // push arg 0
-	call arena::alloc
-	pop rg4
-	pop zero
-	subi bpr 16 rg0
-	ldw rg0, rg0 // bpr-24: alloc
-	push rg4 // bpr-24: ptr2
-	push rg0 // bpr-28: alloc
-	push rg0 // push arg 0
-	call print::print_hex_word
-	pop zero
-	call print::print_newline
-	subi bpr 20 rg0
-	ldw rg0, rg0 // bpr-28: ptr1
-	push rg0 // bpr-32: ptr1
-	push rg0 // push arg 0
-	call print::print_hex_word
-	pop zero
-	call print::print_newline
-	subi bpr 24 rg0
-	ldw rg0, rg0 // bpr-32: ptr2
-	push rg0 // bpr-36: ptr2
-	push rg0 // push arg 0
-	call print::print_hex_word
-	pop zero
-	call print::print_newline
-	subi bpr 36 rg0
-	ldw rg0, rg0 // bpr-44: ptr2
-	ldw rg0, rg2
-	push rg0 // bpr-40: ptr2
-	push rg2 // push arg 0
-	call print::print_num
-	pop zero
-	call print::print_newline
-	lli 42, rg2
-	subi bpr 40 rg5
-	ldw rg5, rg5 // bpr-48: ptr2
-	stw rg2, rg5
-	push rg5 // bpr-44: ptr2
-	push rg5 // push arg 0
-	call print::print_hex_word
-	pop zero
-	call print::print_newline
-	subi bpr 44 rg2
-	ldw rg2, rg2 // bpr-52: ptr2
-	ldw rg2, rg5
-	push rg2 // bpr-48: ptr2
-	push rg5 // push arg 0
-	call print::print_num
-	pop zero
-	call print::print_newline
-	db str_1: "end"
-	lwi str_1, rg5
-	push rg5 // push arg 0
-	call print::println
-	pop zero
-	lli 0, rg5
-	stw rg5, bpr, 8
-	jmp _ret
-

resources/dsa/lib/error/handlers.dsa

@@ -32,3 +32,13 @@ handle_hard_fault:
 	call print::print
 	pop zero
 	hlt
+
+trigger:
+    push bpr
+    mov spr, bpr
+
+    int 0x01
+
+    mov bpr, spr
+    pop bpr
+    return

resources/dsa/lib/memory/block_alloc.dsa

@@ -0,0 +1,216 @@
+// block_alloc.dsa
+// Fixed-size block allocator
+//
+// Memory layout:
+// [base + 0]: free list head pointer (pointer to first free block, or 0 if none)
+// [base + 4]: block size
+// [base + 8]: total blocks
+// [base + 12]: base address of block pool
+// [base + 16+]: block pool (each block starts with a 4-byte next pointer)
+//
+// Usage:
+// include block_alloc "./lib/memory/block_alloc.dsa"
+//
+// For init:
+//   push num_blocks (e.g., 32)
+//   push block_size (e.g., 64 bytes)
+//   call block_alloc::init
+//   pop block_size
+//   pop num_blocks
+//   ; result in spr+8 (allocator handle)
+//
+// For alloc:
+//   push allocator_handle
+//   call block_alloc::alloc
+//   pop allocator_handle
+//   ; result in spr+8 (pointer to block, or 0 if out of memory)
+//
+// For free:
+//   push block_pointer
+//   push allocator_handle
+//   call block_alloc::free
+//   pop allocator_handle
+//   pop block_pointer
+
+dw heap_start: 0x30000  // Start of our heap area
+
+// Initialize the allocator
+// Args: block_size, num_blocks
+// Returns: allocator handle (pointer to metadata)
+init:
+    push bpr
+    mov spr, bpr
+
+    ldw bpr, rg0, 8   // block_size
+    ldw bpr, rg1, 12  // num_blocks
+
+    // Allocate metadata (16 bytes) + pool space
+    ldw heap_start, rg2  // base address for this allocator
+    mov rg2, rg3         // save base in rg3
+
+    // Calculate total size needed: 16 + (block_size * num_blocks)
+    // We'll use a simple multiplication by repeated addition
+    mov rg0, rg4         // block_size to rg4
+    mov rg1, rg5         // num_blocks to rg5
+    lli 0, acc           // accumulator for multiplication
+
+_multiply_loop:
+    cmp rg5, zero
+    jeq _multiply_done
+    add acc, rg4, acc
+    dec rg5
+    jmp _multiply_loop
+
+_multiply_done:
+    // acc now contains block_size * num_blocks
+    addi acc, 16         // add metadata size
+
+    // Update heap_start for next allocation
+    add rg2, acc, acc
+    stw acc, heap_start
+
+    // Now set up metadata at rg3 (base)
+    // [base + 0]: free list head (will point to first block)
+    // [base + 4]: block_size
+    // [base + 8]: total blocks
+    // [base + 12]: pool base address
+
+    addi rg3, 16, rg6     // rg6 = pool base
+
+    stw rg6, rg3          // store pool base as free list head initially
+    stw rg0, rg3, 4       // store block_size
+    stw rg1, rg3, 8       // store total blocks
+    stw rg6, rg3, 12      // store pool base address
+
+    // Now initialize the free list
+    // Each block's first 4 bytes point to the next block
+    // rg6 = current block pointer
+    // rg0 = block_size
+    // rg1 = num_blocks (counter)
+
+    dec rg1              // we'll count down from num_blocks-1
+
+_init_loop:
+    cmp rg1, zero
+    jeq _init_loop_done
+
+    // Calculate next block address: current + block_size
+    add rg6, rg0, rg7    // rg7 = next block address
+
+    // Store next pointer at current block
+    stw rg7, rg6
+
+    // Move to next block
+    mov rg7, rg6
+    dec rg1
+    jmp _init_loop
+
+_init_loop_done:
+    // Last block points to null (0)
+    lli 0, acc
+    stw acc, rg6
+
+    // Return allocator handle (base address - 16 to get back to metadata start)
+    stw rg3, bpr, 8
+
+    mov bpr, spr
+    pop bpr
+    return
+
+
+// Allocate a block
+// Args: allocator_handle
+// Returns: pointer to block (or 0 if out of memory)
+alloc:
+    push bpr
+    mov spr, bpr
+
+    ldw bpr, rg0, 8      // allocator handle (metadata base)
+
+    // Load free list head
+    ldw rg0, rg1         // rg1 = free list head
+
+    // Check if free list is empty
+    cmp rg1, zero
+    jeq _alloc_fail
+
+    // Free list is not empty, pop the first block
+    // Load the next pointer from the block we're allocating
+    ldw rg1, rg2         // rg2 = next free block
+
+    // Update free list head to point to next block
+    stw rg2, rg0
+
+    // Return the allocated block (rg1)
+    stw rg1, bpr, 8
+    jmp _alloc_done
+
+_alloc_fail:
+    // No free blocks, return 0
+    lli 0, acc
+    stw acc, bpr, 8
+
+_alloc_done:
+    mov bpr, spr
+    pop bpr
+    return
+
+
+// Free a block
+// Args: allocator_handle, block_pointer
+// Returns: nothing (but could return error code if block is invalid)
+free:
+    push bpr
+    mov spr, bpr
+
+    ldw bpr, rg0, 8      // allocator handle
+    ldw bpr, rg6, 12     // pointer to the block pointer to free
+    ldw rg6, rg1         // rg1 = block pointer to free
+
+    // Load current free list head
+    ldw rg0, rg2         // rg2 = current head
+
+    // Set the freed block's next pointer to current head
+    stw rg2, rg1
+
+    // Update free list head to point to freed block
+    stw rg1, rg0
+
+    // Update the freed block's previous pointer to NULL
+    lli 0, rg1
+    stw rg1, rg6
+
+    mov bpr, spr
+    pop bpr
+    return
+
+
+// Debug function: get stats
+// Args: allocator_handle
+// Returns: nothing (but could populate a stats structure)
+get_stats:
+    push bpr
+    mov spr, bpr
+
+    ldw bpr, rg0, 8      // allocator handle
+
+    // Count free blocks by traversing the free list
+    ldw rg0, rg1         // rg1 = free list head
+    lli 0, rg2           // rg2 = counter
+
+count_loop:
+    cmp rg1, zero
+    jeq count_done
+
+    inc rg2
+    ldw rg1, rg1         // move to next block
+    jmp count_loop
+
+count_done:
+    // rg2 now contains number of free blocks
+    // Could store this somewhere or return it
+    stw rg2, bpr, 8
+
+    mov bpr, spr
+    pop bpr
+    return

resources/dsa/main.dsa

@@ -1,50 +1,12 @@
+// program to just test compute power
 
-// GENERATED BY DSC COMPILER
-// Generated at 2026-02-04 01:44:06
+dw large_num: 0x333333 // 333,333 instructions
+start:
+	ldw large_num, rg0
 
-// Imports
-include print: "./lib/io/print.dsa"
-include fib: "./lib/maths/fib.dsa"
-
-// Globals & Reserved Memory
-
-
-// Entry Point
-dw stack: 0x10000
-db message: "Process Exited with code:"
-_init:
-	ldw stack, bpr
-	mov bpr, spr
-	push zero
-	call main
-	call print::print_newline
-	lwi message, rg0
-	push rg0
-	call print::print
-	pop zero
-	call print::print_hex_word
-	pop zero
-	hlt
-
-
-// Return
-_ret:
-	mov bpr, spr
-	pop bpr
-	return
-
-// Compiled Code Starts...
-main:
-	push bpr
-	mov spr, bpr
-
-	lli 6, rg0
-	push rg0 // bpr-4: x
-	push rg0 // push arg 0
-	call fib::fib_n
-	pop rg1
-	push rg1 // bpr-8: y
-	push rg1 // push arg 0
-	call print::print_num
-	pop zero
-	jmp _ret
+// run approx 1m instructions
+loop:
+	dec rg0
+	cmp rg0, zero
+	jgt loop
+    hlt

resources/dsa/main.dsc

@@ -1,9 +1,37 @@
 include print: "./lib/io/print.dsa";
-include fib: "./lib/maths/fib.dsa";
+include alloc: "./lib/memory/block_alloc.dsa";
 
 fn main() -> u32 {
-	let x: u32 = 6;
+	let allocator: u32 = alloc::init(64, 32);
 	
-	let y: u32 = fib::fib_n(x);
-	print::print_num(y);
+	print::print_hex_word(allocator);
+
+	print::print_newline();
+
+	let ptr: u32 = alloc::alloc(allocator);
+	print::print_hex_word(ptr);
+	*ptr = 200;
+
+	print::print_newline();
+
+	let p2: u32 = alloc::alloc(allocator);
+	print::print_hex_word(p2);
+	print::print_newline();
+	print::print_num(*ptr);
+
+	alloc::free(allocator, &ptr);
+	let ptr3: u32 = alloc::alloc(allocator);
+
+	print::print_newline();
+
+	print::print_hex_word(ptr3);
+
+	print::print_newline();
+	print::print_hex_word(ptr);
+
+	if ptr == 0 {
+		print::print("successful free of ptr");
+	}
+
+	return 0;
 }

resources/dsa/output.dsa

@@ -1,214 +0,0 @@
-
-// GENERATED BY DSC COMPILER
-// Generated at 2026-02-03 23:37:16
-
-// Imports
-include print: "./lib/io/print.dsa"
-
-// Globals & Reserved Memory
-dw heap_start: 196608
-dw heap_end: 262144
-dw heap_current: 196608
-
-// Entry Point
-dw stack: 0x10000
-db message: "Process Exited with code:"
-_init:
-	ldw stack, bpr
-	mov bpr, spr
-	push zero
-	call main
-	call print::print_newline
-	lwi message, rg0
-	push rg0
-	call print::print
-	pop zero
-	call print::print_hex_word
-	pop zero
-	hlt
-
-
-// Return
-_ret:
-	mov bpr, spr
-	pop bpr
-	return
-
-// Compiled Code Starts...
-main:
-	push bpr
-	mov spr, bpr
-
-	lli 0, rg0
-	push rg0 // bpr-4: x
-	subi bpr 4 rg1
-	lli 512, rg0
-	push rg1 // bpr-8: y
-	push rg0 // push arg 0
-	call arena_create
-	pop rg2
-	lli 32, rg0
-	push rg2 // bpr-12: alloc
-	push rg0 // push arg 1
-	push rg2 // push arg 0
-	call arena_alloc
-	pop rg3
-	pop zero
-	lli 32, rg0
-	subi bpr 12 rg2
-	ldw rg2, rg2 // bpr-20: alloc
-	push rg3 // bpr-16: ptr1
-	push rg2 // bpr-20: alloc
-	push rg0 // push arg 1
-	push rg2 // push arg 0
-	call arena_alloc
-	pop rg4
-	pop zero
-	subi bpr 20 rg0
-	ldw rg0, rg0 // bpr-28: alloc
-	push rg4 // bpr-24: ptr2
-	push rg0 // bpr-28: alloc
-	push rg0 // push arg 0
-	call print::print_hex_word
-	pop zero
-	call print::print_newline
-	subi bpr 16 rg0
-	ldw rg0, rg0 // bpr-24: ptr1
-	push rg0 // bpr-32: ptr1
-	push rg0 // push arg 0
-	call print::print_hex_word
-	pop zero
-	call print::print_newline
-	subi bpr 24 rg0
-	ldw rg0, rg0 // bpr-32: ptr2
-	push rg0 // bpr-36: ptr2
-	push rg0 // push arg 0
-	call print::print_hex_word
-	pop zero
-	call print::print_newline
-	subi bpr 36 rg0
-	ldw rg0, rg0 // bpr-44: ptr2
-	ldw rg0, rg2
-	push rg0 // bpr-40: ptr2
-	push rg2 // push arg 0
-	call print::print_num
-	pop zero
-	call print::print_newline
-	lli 42, rg2
-	subi bpr 40 rg5
-	ldw rg5, rg5 // bpr-48: ptr2
-	stw rg2, rg5
-	push rg5 // bpr-44: ptr2
-	push rg5 // push arg 0
-	call print::print_hex_word
-	pop zero
-	call print::print_newline
-	subi bpr 44 rg2
-	ldw rg2, rg2 // bpr-52: ptr2
-	ldw rg2, rg5
-	push rg2 // bpr-48: ptr2
-	push rg5 // push arg 0
-	call print::print_num
-	pop zero
-	call print::print_newline
-	db str_1: "end"
-	lwi str_1, rg5
-	push rg5 // push arg 0
-	call print::println
-	pop zero
-	lli 0, rg5
-	stw rg5, bpr, 8
-	jmp _ret
-
-arena_create:
-	push bpr
-	mov spr, bpr
-
-	ldw bpr, rg0, 8
-	lli 12, rg1
-	add rg0, rg1, rg2
-	ldw heap_current, rg1
-	add rg1, rg2, rg3
-	ldw heap_end, rg4
-	cmp rg3, rg4
-	lli 0, rg5
-	jle _cmp_end_2
-	lli 1, rg5
-_cmp_end_2:
-	cmp rg5, zero
-	jeq _else_4
-_then_3:
-	lli 0, rg4
-	stw rg4, bpr, 8
-	jmp _ret
-	jmp _end_5
-_else_4:
-	nop
-_end_5:
-	lli 12, rg4
-	add rg1, rg4, rg5
-	add rg1, rg2, rg4
-	stw rg5, rg1
-	lli 4, rg6
-	add rg1, rg6, rg7
-	stw rg5, rg7
-	lli 8, rg6
-	add rg1, rg6, rg7
-	stw rg4, rg7
-	stw rg3, heap_current
-	stw rg1, bpr, 8
-	jmp _ret
-
-arena_alloc:
-	push bpr
-	mov spr, bpr
-
-	ldw bpr, rg0, 8
-	ldw bpr, rg1, 12
-	lli 4, rg2
-	add rg0, rg2, rg3
-	ldw rg3, rg2
-	lli 8, rg3
-	add rg0, rg3, rg4
-	ldw rg4, rg3
-	add rg2, rg1, rg4
-	cmp rg4, rg3
-	lli 0, rg5
-	jle _cmp_end_6
-	lli 1, rg5
-_cmp_end_6:
-	cmp rg5, zero
-	jeq _else_8
-_then_7:
-	lli 0, rg5
-	stw rg5, bpr, 8
-	jmp _ret
-	jmp _end_9
-_else_8:
-	nop
-_end_9:
-	lli 4, rg5
-	add rg0, rg5, rg6
-	stw rg4, rg6
-	stw rg2, bpr, 8
-	jmp _ret
-
-arena_destroy:
-	push bpr
-	mov spr, bpr
-
-	ldw bpr, rg0, 8
-	lli 0, rg1
-	stw rg1, bpr, 8
-	jmp _ret
-
-reset_all:
-	push bpr
-	mov spr, bpr
-
-	ldw heap_start, rg0
-	stw rg0, heap_current
-	lli 0, rg0
-	stw rg0, bpr, 8
-	jmp _ret
-

resources/dsa/test.dsc

@@ -0,0 +1,20 @@
+include print: "./lib/io/print.dsa";
+
+fn main() -> u32 {
+	let x: u32 = 30;
+	print::print_num(x);
+
+	200 + 5;
+
+	let p: Point = Point {
+		x: 10,
+		y: 20,
+		test: [2, 3, 4]
+	};
+}
+
+struct Point {
+	x: u32,
+	y: u32,
+	test: [u32; 3],
+}

resources/dsa/testprint.dsa

@@ -1,4 +1,4 @@
-include print "./lib/io/print.dsa"
+include print: "./lib/io/print.dsa"
 
 dw idt: 0xFFFF0000
 dw stack: 0x10000
@@ -57,7 +57,7 @@ start:
 
 	// test reset cursor pos
 	call print::reset
-	
+
 	// test print string at reset pos
 	lwi replace, rg0
 	push rg0