From: Evgenii Akentev Date: Sat, 1 Nov 2025 16:17:50 +0000 (+0400) Subject: Little Computer 3 in Zig X-Git-Url: https://git.xn--bdkaa.com/?a=commitdiff_plain;ds=sidebyside;p=lc3vm.zig.git Little Computer 3 in Zig --- 2fd0f08e93eeead29de9ee224b5c4737c5ce5e74 diff --git a/.gitignore b/.gitignore new file mode 100644 index 0000000..bc0b345 --- /dev/null +++ b/.gitignore @@ -0,0 +1,23 @@ +# This file is for zig-specific build artifacts. +# If you have OS-specific or editor-specific files to ignore, +# such as *.swp or .DS_Store, put those in your global +# ~/.gitignore and put this in your ~/.gitconfig: +# +# [core] +# excludesfile = ~/.gitignore +# +# Cheers! +# -andrewrk + +.zig-cache/ +zig-out/ +/release/ +/debug/ +/build/ +/build-*/ +/docgen_tmp/ + +# Although this was renamed to .zig-cache, let's leave it here for a few +# releases to make it less annoying to work with multiple branches. +zig-cache/ + diff --git a/build.zig b/build.zig new file mode 100644 index 0000000..454216b --- /dev/null +++ b/build.zig @@ -0,0 +1,156 @@ +const std = @import("std"); + +// Although this function looks imperative, it does not perform the build +// directly and instead it mutates the build graph (`b`) that will be then +// executed by an external runner. The functions in `std.Build` implement a DSL +// for defining build steps and express dependencies between them, allowing the +// build runner to parallelize the build automatically (and the cache system to +// know when a step doesn't need to be re-run). +pub fn build(b: *std.Build) void { + // Standard target options allow the person running `zig build` to choose + // what target to build for. Here we do not override the defaults, which + // means any target is allowed, and the default is native. Other options + // for restricting supported target set are available. + const target = b.standardTargetOptions(.{}); + // Standard optimization options allow the person running `zig build` to select + // between Debug, ReleaseSafe, ReleaseFast, and ReleaseSmall. Here we do not + // set a preferred release mode, allowing the user to decide how to optimize. + const optimize = b.standardOptimizeOption(.{}); + // It's also possible to define more custom flags to toggle optional features + // of this build script using `b.option()`. All defined flags (including + // target and optimize options) will be listed when running `zig build --help` + // in this directory. + + // This creates a module, which represents a collection of source files alongside + // some compilation options, such as optimization mode and linked system libraries. + // Zig modules are the preferred way of making Zig code available to consumers. + // addModule defines a module that we intend to make available for importing + // to our consumers. We must give it a name because a Zig package can expose + // multiple modules and consumers will need to be able to specify which + // module they want to access. + const mod = b.addModule("lc3-vm_zig", .{ + // The root source file is the "entry point" of this module. Users of + // this module will only be able to access public declarations contained + // in this file, which means that if you have declarations that you + // intend to expose to consumers that were defined in other files part + // of this module, you will have to make sure to re-export them from + // the root file. + .root_source_file = b.path("src/root.zig"), + // Later on we'll use this module as the root module of a test executable + // which requires us to specify a target. + .target = target, + }); + + // Here we define an executable. An executable needs to have a root module + // which needs to expose a `main` function. While we could add a main function + // to the module defined above, it's sometimes preferable to split business + // business logic and the CLI into two separate modules. + // + // If your goal is to create a Zig library for others to use, consider if + // it might benefit from also exposing a CLI tool. A parser library for a + // data serialization format could also bundle a CLI syntax checker, for example. + // + // If instead your goal is to create an executable, consider if users might + // be interested in also being able to embed the core functionality of your + // program in their own executable in order to avoid the overhead involved in + // subprocessing your CLI tool. + // + // If neither case applies to you, feel free to delete the declaration you + // don't need and to put everything under a single module. + const exe = b.addExecutable(.{ + .name = "lc3-vm_zig", + .root_module = b.createModule(.{ + // b.createModule defines a new module just like b.addModule but, + // unlike b.addModule, it does not expose the module to consumers of + // this package, which is why in this case we don't have to give it a name. + .root_source_file = b.path("src/main.zig"), + // Target and optimization levels must be explicitly wired in when + // defining an executable or library (in the root module), and you + // can also hardcode a specific target for an executable or library + // definition if desireable (e.g. firmware for embedded devices). + .target = target, + .optimize = optimize, + // List of modules available for import in source files part of the + // root module. + .imports = &.{ + // Here "lc3-vm_zig" is the name you will use in your source code to + // import this module (e.g. `@import("lc3-vm_zig")`). The name is + // repeated because you are allowed to rename your imports, which + // can be extremely useful in case of collisions (which can happen + // importing modules from different packages). + .{ .name = "lc3-vm_zig", .module = mod }, + }, + }), + }); + + // This declares intent for the executable to be installed into the + // install prefix when running `zig build` (i.e. when executing the default + // step). By default the install prefix is `zig-out/` but can be overridden + // by passing `--prefix` or `-p`. + b.installArtifact(exe); + + // This creates a top level step. Top level steps have a name and can be + // invoked by name when running `zig build` (e.g. `zig build run`). + // This will evaluate the `run` step rather than the default step. + // For a top level step to actually do something, it must depend on other + // steps (e.g. a Run step, as we will see in a moment). + const run_step = b.step("run", "Run the app"); + + // This creates a RunArtifact step in the build graph. A RunArtifact step + // invokes an executable compiled by Zig. Steps will only be executed by the + // runner if invoked directly by the user (in the case of top level steps) + // or if another step depends on it, so it's up to you to define when and + // how this Run step will be executed. In our case we want to run it when + // the user runs `zig build run`, so we create a dependency link. + const run_cmd = b.addRunArtifact(exe); + run_step.dependOn(&run_cmd.step); + + // By making the run step depend on the default step, it will be run from the + // installation directory rather than directly from within the cache directory. + run_cmd.step.dependOn(b.getInstallStep()); + + // This allows the user to pass arguments to the application in the build + // command itself, like this: `zig build run -- arg1 arg2 etc` + if (b.args) |args| { + run_cmd.addArgs(args); + } + + // Creates an executable that will run `test` blocks from the provided module. + // Here `mod` needs to define a target, which is why earlier we made sure to + // set the releative field. + const mod_tests = b.addTest(.{ + .root_module = mod, + }); + + // A run step that will run the test executable. + const run_mod_tests = b.addRunArtifact(mod_tests); + + // Creates an executable that will run `test` blocks from the executable's + // root module. Note that test executables only test one module at a time, + // hence why we have to create two separate ones. + const exe_tests = b.addTest(.{ + .root_module = exe.root_module, + }); + + // A run step that will run the second test executable. + const run_exe_tests = b.addRunArtifact(exe_tests); + + // A top level step for running all tests. dependOn can be called multiple + // times and since the two run steps do not depend on one another, this will + // make the two of them run in parallel. + const test_step = b.step("test", "Run tests"); + test_step.dependOn(&run_mod_tests.step); + test_step.dependOn(&run_exe_tests.step); + + // Just like flags, top level steps are also listed in the `--help` menu. + // + // The Zig build system is entirely implemented in userland, which means + // that it cannot hook into private compiler APIs. All compilation work + // orchestrated by the build system will result in other Zig compiler + // subcommands being invoked with the right flags defined. You can observe + // these invocations when one fails (or you pass a flag to increase + // verbosity) to validate assumptions and diagnose problems. + // + // Lastly, the Zig build system is relatively simple and self-contained, + // and reading its source code will allow you to master it. +} diff --git a/build.zig.zon b/build.zig.zon new file mode 100644 index 0000000..9a43c8a --- /dev/null +++ b/build.zig.zon @@ -0,0 +1,81 @@ +.{ + // This is the default name used by packages depending on this one. For + // example, when a user runs `zig fetch --save `, this field is used + // as the key in the `dependencies` table. Although the user can choose a + // different name, most users will stick with this provided value. + // + // It is redundant to include "zig" in this name because it is already + // within the Zig package namespace. + .name = .arithvm_zig, + // This is a [Semantic Version](https://semver.org/). + // In a future version of Zig it will be used for package deduplication. + .version = "0.0.0", + // Together with name, this represents a globally unique package + // identifier. This field is generated by the Zig toolchain when the + // package is first created, and then *never changes*. This allows + // unambiguous detection of one package being an updated version of + // another. + // + // When forking a Zig project, this id should be regenerated (delete the + // field and run `zig build`) if the upstream project is still maintained. + // Otherwise, the fork is *hostile*, attempting to take control over the + // original project's identity. Thus it is recommended to leave the comment + // on the following line intact, so that it shows up in code reviews that + // modify the field. + .fingerprint = 0x79da8ce1ed863de0, // Changing this has security and trust implications. + // Tracks the earliest Zig version that the package considers to be a + // supported use case. + .minimum_zig_version = "0.15.1", + // This field is optional. + // Each dependency must either provide a `url` and `hash`, or a `path`. + // `zig build --fetch` can be used to fetch all dependencies of a package, recursively. + // Once all dependencies are fetched, `zig build` no longer requires + // internet connectivity. + .dependencies = .{ + // See `zig fetch --save ` for a command-line interface for adding dependencies. + //.example = .{ + // // When updating this field to a new URL, be sure to delete the corresponding + // // `hash`, otherwise you are communicating that you expect to find the old hash at + // // the new URL. If the contents of a URL change this will result in a hash mismatch + // // which will prevent zig from using it. + // .url = "https://example.com/foo.tar.gz", + // + // // This is computed from the file contents of the directory of files that is + // // obtained after fetching `url` and applying the inclusion rules given by + // // `paths`. + // // + // // This field is the source of truth; packages do not come from a `url`; they + // // come from a `hash`. `url` is just one of many possible mirrors for how to + // // obtain a package matching this `hash`. + // // + // // Uses the [multihash](https://multiformats.io/multihash/) format. + // .hash = "...", + // + // // When this is provided, the package is found in a directory relative to the + // // build root. In this case the package's hash is irrelevant and therefore not + // // computed. This field and `url` are mutually exclusive. + // .path = "foo", + // + // // When this is set to `true`, a package is declared to be lazily + // // fetched. This makes the dependency only get fetched if it is + // // actually used. + // .lazy = false, + //}, + }, + // Specifies the set of files and directories that are included in this package. + // Only files and directories listed here are included in the `hash` that + // is computed for this package. Only files listed here will remain on disk + // when using the zig package manager. As a rule of thumb, one should list + // files required for compilation plus any license(s). + // Paths are relative to the build root. Use the empty string (`""`) to refer to + // the build root itself. + // A directory listed here means that all files within, recursively, are included. + .paths = .{ + "build.zig", + "build.zig.zon", + "src", + // For example... + //"LICENSE", + //"README.md", + }, +} diff --git a/src/main.zig b/src/main.zig new file mode 100644 index 0000000..f3079ce --- /dev/null +++ b/src/main.zig @@ -0,0 +1,20 @@ +const std = @import("std"); +const lc3vm_zig = @import("lc3-vm_zig"); + +pub fn main() !void { + var arena = std.heap.ArenaAllocator.init(std.heap.page_allocator); + defer arena.deinit(); + + var memory = &lc3vm_zig.memory; + + memory[0x3000] = 0xF026; // 1111 0000 0010 0110 TRAP trp_in_u16 ;read an uint16_t from stdin and put it in R0 + memory[0x3002] = 0x1220; // 0001 0010 0010 0000 ADD R1,R0,x0 ;add contents of R0 to R1 + memory[0x3003] = 0xF026; // 1111 0000 0010 0110 TRAP trp_in_u16 ;read an uint16_t from stdin and put it in R0 + memory[0x3004] = 0x1240; // 0001 0010 0010 0000 ADD R1,R1,R0 ;add contents of R0 to R1 + memory[0x3006] = 0x1060; // 0001 0000 0110 0000 ADD R0,R1,x0 ;add contents of R1 to R0 + memory[0x3007] = 0xF027; // 1111 0000 0010 0111 TRAP trp_out_u16;show the contents of R0 to stdout + memory[0x3008] = 0xF025; // 1111 0000 0010 0101 HALT ;halt + + try lc3vm_zig.start(arena.allocator(), 0); +} + diff --git a/src/root.zig b/src/root.zig new file mode 100644 index 0000000..cc4e676 --- /dev/null +++ b/src/root.zig @@ -0,0 +1,221 @@ +//! By convention, root.zig is the root source file when making a library. +const std = @import("std"); + +// Implementation of the Little Computer 3 following +// https://www.andreinc.net/2021/12/01/writing-a-simple-vm-in-less-than-125-lines-of-c + +const pcStart: u16 = 0x3000; +const u16Max: u32 = 65535; +pub var memory: [u16Max + 1]u16 = .{0} ** (u16Max + 1); +var running: bool = true; + +inline fn mr(address: u16) u16 { return memory[address]; } + +inline fn mw(address: u16, value: u16) void { + memory[address] = value; +} + +const Register = enum(u16) { + R0 = 0, + R1 = 1, + R2 = 2, + R3 = 3, + R4 = 4, + R5 = 5, + R6 = 6, + R7 = 7, + RPC = 8, + RCND = 9, + RCNT = 10 +}; + +var registers: [11]u16 = .{0} ** 11; + +// The instruction fits into u16: +// +// [ OP CODE ][ PARAM1 ][ PARAM2 ][ ][ PARAM3 ] +// [ 4 bits ][ 12 bits ] + +const Opcode = enum { + br, // conditional + add, // addition + ld, // load RPC + offset + st, // store + jsr, // jump to subroutine + andop, // bitwise and + ldr, // load base + offset + str, // store base + offset + rti, // return from interrupt + not, // bitwise complement + ldi, // load indirect + sti, // store indirect + jmp, // jump/return to subroutine + unused, // unused opcode + lea, // load effective address + trap, // system trap/call + + pub fn parse(instruction: u16) Opcode { + // shift 12 bits to the right to get first 4 bits + return @enumFromInt(instruction >> 12); + } +}; + +const TrapFunc = enum { + tgetc, tout, tputs, tin, tputsp, thalt, tinu16, toutu16, + + pub fn parse(n: u16) TrapFunc { + return @enumFromInt(n); + } +}; + +const Flag = enum(u16) { + FP = (1 << 0), + FZ = (1 << 1), + FN = (1 << 2), +}; + +inline fn uf(r: u16) void { + if (registers[r] == 0) { registers[@intFromEnum(Register.RCND)] = @intFromEnum(Flag.FZ); } + else if ((registers[r] >> 15) == 1) { registers[@intFromEnum(Register.RCND)] = @intFromEnum(Flag.FN); } + else registers[@intFromEnum(Register.RCND)] = @intFromEnum(Flag.FP); +} + +inline fn fcnd(instruction: u16) u16 { return (instruction >> 9) & 0x7; } +inline fn dr(instruction: u16) u16 { return (instruction >> 9) & 0x7; } +inline fn sr1(instruction: u16) u16 { return(instruction >> 6) & 0x7; } +inline fn sr2(instruction: u16) u16 { return (instruction) & 0x7; } +inline fn imm(instruction: u16) u16 { return (instruction) & 0x1F; } +inline fn fimm(instruction: u16) u16 { return (instruction >> 5) & 1; } + +inline fn sext(n: u16, b: u32) u16 { + return if ((n >> (b - 1)) & 1 == 1) @truncate(@as(u32, n) | (0xFFFF << b)) else n; +} + +// 0000 0000 0001 0110 <--- a in binary +// 1111 1111 1111 0110 <--- sextimm(a) in binary +inline fn sextimm(instruction: u16) u16 { return sext(imm(instruction), 5); } + +inline fn poff(instruction: u16) u16 { return sext(instruction & 0x3F, 6); } +inline fn poff9(instruction: u16) u16 { return sext(instruction & 0x1FF, 9); } +inline fn poff11(instruction: u16) u16 { return sext(instruction & 0x7FF, 11); } +inline fn fl(instruction: u16) u16 { return (instruction >> 11) & 1; } +inline fn br(instruction: u16) u16 { return (instruction >> 6) & 0x7; } +inline fn trp(instruction: u16) u16 { return instruction & 0xFF; } + +pub fn evaluate(allocator: std.mem.Allocator, i: u16) !void { + switch (Opcode.parse(i)) { + inline .add => { + registers[dr(i)] = registers[sr1(i)] + + if (fimm(i) == 1) + sextimm(i) + else registers[sr2(i)]; + uf(dr(i)); + }, + inline .andop => { + registers[dr(i)] = registers[sr1(i)] & (if (fimm(i) == 1) (sextimm(i)) else registers[sr2(i)]); + uf(dr(i)); + }, + inline .ld => { + registers[dr(i)] = mr(registers[@intFromEnum(Register.RPC)] + poff9(i)); + uf(dr(i)); + }, + inline .ldi => { + registers[dr(i)] = mr(mr(registers[@intFromEnum(Register.RPC)] + poff9(i))); + uf(dr(i)); + }, + inline .ldr => { + registers[dr(i)] = mr(registers[sr1(i)] + poff(i)); + uf(dr(i)); + }, + inline .lea => { + registers[dr(i)] = registers[@intFromEnum(Register.RPC)] + poff9(i); + uf(dr(i)); + }, + inline .not => { + registers[dr(i)] = ~ registers[sr1(i)]; + uf(dr(i)); + }, + inline .st => { + mw(registers[@intFromEnum(Register.RPC)] + poff9(i), registers[dr(i)]); + }, + inline .sti => { + mw(mr(registers[@intFromEnum(Register.RPC)] + poff9(i)), registers[dr(i)]); + }, + inline .str => { + mw(registers[sr1(i)] + poff(i), registers[dr(i)]); + }, + inline .jmp => { + registers[@intFromEnum(Register.RPC)] = registers[br(i)]; + }, + inline .jsr => { + registers[@intFromEnum(Register.R7)] = registers[@intFromEnum(Register.RPC)]; + registers[@intFromEnum(Register.RPC)] = if (fl(i) == 1) (registers[@intFromEnum(Register.RPC)] + poff11(i)) else registers[br(i)]; + }, + inline .br => { + if (registers[@intFromEnum(Register.RCND)] & fcnd(i) == 1) { + registers[@intFromEnum(Register.RPC)] += poff9(i); + } + }, + inline .unused, .rti => { + }, + inline .trap => { + switch (TrapFunc.parse(trp(i) - 0x20)) { + inline .tgetc => { + const dest: []u8 = try allocator.alloc(u8, 1); + _ = try std.fs.File.stdin().read(dest); + registers[@intFromEnum(Register.R0)] = dest[0]; + }, + inline .tout => { + std.debug.print("{}", .{ registers[@intFromEnum(Register.R0)] }); + }, + inline .tputs => { + }, + inline .tin => { + const buffer: []u8 = try allocator.alloc(u8, 1024); + defer allocator.free(buffer); + + var stdin = std.fs.File.stdin().reader(buffer).interface; + const m = try stdin.takeInt(u8, .little); + registers[@intFromEnum(Register.R0)] = m; + }, + inline .thalt => { + running = false; + }, + inline .tinu16 => { + const buffer: []u8 = try allocator.alloc(u8, 1024); + defer allocator.free(buffer); + + var line_buffer: []u8 = try allocator.alloc(u8, 1024); + defer allocator.free(line_buffer); + var writer: std.io.Writer = .fixed(line_buffer); + + var stdin = std.fs.File.stdin().reader(buffer).interface; + + const line_length = try stdin.streamDelimiterLimit(&writer, '\n', .unlimited); + const v: u16 = try std.fmt.parseInt(u16, line_buffer[0..line_length], 10); + registers[@intFromEnum(Register.R0)] = v; + }, + inline .tputsp => { + }, + inline .toutu16 => { + const buffer: []u8 = try allocator.alloc(u8, 1024); + defer allocator.free(buffer); + + const v: u16 = registers[@intFromEnum(Register.R0)]; + std.debug.print("{}\n", .{ v }); + }, + } + } + } +} + +pub fn start(allocator: std.mem.Allocator, offset: u16) !void { + registers[@intFromEnum(Register.RPC)] = pcStart + offset; + while (running) { + const instruction = mr(registers[@intFromEnum(Register.RPC)]); + registers[@intFromEnum(Register.RPC)] += 1; + + try evaluate(allocator, instruction); + } +} +