Merge pull request #93 from vmttmv/fix/91-header-cleanup

Remove types.c/.h, use klibc headers, amend stdbool.h, reimplement cpuid()

.gitignore

@@ -1 +1,5 @@
+.build.env
 build
+cross
+.cache/
+compile_commands.json

.vscode/settings.json

@@ -1,6 +1,7 @@
 {
     "files.associations": {
         ".fantomasignore": "ignore",
-        "stddef.h": "c"
+        "stddef.h": "c",
+        "io.h": "c"
     }
 }

Makefile

@@ -1,35 +1,73 @@
 AS = nasm
-CC = gcc
+ASFLAGS = -f elf32 -g -F dwarf
-LD = ld
+CC = i386-elf-gcc
-QEMU = qemu-system-i386
+LD = i386-elf-ld
-IMG_SIZE = 1440k
+QEMU= qemu-system-i386
+OBJCOPY = i386-elf-objcopy
 
-BOOT_SRC = bootloader/boot.asm
+BUILD_DIR = build
-BOOT_BIN = build/boot.bin
+CROSS_DIR = cross
-BOOT_IMG = build/boot.img
+DISK_IMG = $(BUILD_DIR)/disk.img
-KERNEL_SRC = kernel/kmain.c
+STAGE2_SIZE = 2048
-KERNEL_BIN = build/kernel.bin
-DISK_IMG = build/disk.img
 
-all: $(BOOT_IMG) $(KERNEL_BIN) $(DISK_IMG)
+KERNEL_C_SRC = $(wildcard kernel/*.c)
+KERNEL_ASM_SRC = $(wildcard kernel/*.asm)
+KERNEL_OBJ = $(patsubst kernel/%.c, $(BUILD_DIR)/%.o, $(KERNEL_C_SRC))
+KERNEL_OBJ += $(patsubst kernel/%.asm, $(BUILD_DIR)/asm_%.o, $(KERNEL_ASM_SRC))
 
-$(BOOT_BIN): $(BOOT_SRC)
+KLIBC_SRC = $(wildcard klibc/src/*.c)
-	$(AS) -f bin -o $@ $<
+KLIBC_OBJ = $(patsubst klibc/src/%.c, $(BUILD_DIR)/klibc/%.o, $(KLIBC_SRC))
 
-$(BOOT_IMG): $(BOOT_BIN)
+all: $(DISK_IMG)
-	cp $(BOOT_BIN) $@
-	truncate -s $(IMG_SIZE) $@
 
-$(KERNEL_BIN): $(KERNEL_SRC)
+.PHONY: stage1 stage2 kernel run gdb clean
-	$(CC) -ffreestanding -c $< -o build/kernel.o
+stage1: $(BUILD_DIR)
-	$(LD) -T bootloader/linker.ld -o $@ build/kernel.o
+	$(AS) $(ASFLAGS) -o $(BUILD_DIR)/$@.o bootloader/$@.asm
+	$(LD) -Ttext=0x7c00 -melf_i386 -o $(BUILD_DIR)/$@.elf $(BUILD_DIR)/$@.o
+	$(OBJCOPY) -O binary $(BUILD_DIR)/$@.elf $(BUILD_DIR)/$@.bin
 
-$(DISK_IMG): $(BOOT_IMG) $(KERNEL_BIN)
+# NOTE: Stage2 final size should be checked against `$(STAGE2_SIZE)` by the build system to avoid an overflow.
-	dd if=$(BOOT_IMG) of=$@ bs=512 seek=4
+# Alternatively, convey the final stage2 size through other means to stage1.
-	dd if=$(KERNEL_BIN) of=$@ bs=512 seek=200
+stage2: $(BUILD_DIR)
+	$(AS) $(ASFLAGS) -o $(BUILD_DIR)/stage2.o bootloader/stage2.asm
+	$(CC) -std=c11 -ffreestanding -nostdlib -nostdinc -fno-stack-protector -m32 -Iklibc/include -g -c -o $(BUILD_DIR)/stage2_load.o bootloader/stage2_load.c
+	$(LD) -Tbootloader/stage2.ld -melf_i386 -o $(BUILD_DIR)/$@.elf $(BUILD_DIR)/stage2.o $(BUILD_DIR)/stage2_load.o
+	$(OBJCOPY) -O binary $(BUILD_DIR)/$@.elf $(BUILD_DIR)/$@.bin
+	truncate -s $(STAGE2_SIZE) $(BUILD_DIR)/$@.bin
 
-run: $(DISK_IMG)
+$(BUILD_DIR)/asm_%.o: kernel/%.asm
-	$(QEMU) -drive file=$<,format=raw,if=floppy
+	$(AS) $(ASFLAGS) -o $@ $<
+
+$(BUILD_DIR)/%.o: kernel/%.c
+	$(CC) -std=c11 -ffreestanding -nostdlib -nostdinc -fno-stack-protector -m32 -Iklibc/include -g -c -o $@ $<
+
+$(BUILD_DIR)/klibc/%.o: klibc/src/%.c
+	$(CC) -std=c11 -ffreestanding -nostdlib -nostdinc -fno-stack-protector -m32 -Iklibc/include -g -c -o $@ $<
+
+kernel: $(KERNEL_OBJ) | $(BUILD_DIR) $(KLIBC_OBJ)
+	$(LD) -melf_i386 -Tkernel/linker.ld -o $(BUILD_DIR)/kernel.elf $(KERNEL_OBJ) $(KLIBC_OBJ)
+
+$(DISK_IMG): stage1 stage2 kernel
+	dd if=$(BUILD_DIR)/stage1.bin of=$@
+	dd if=$(BUILD_DIR)/stage2.bin of=$@ oflag=append conv=notrunc
+	dd if=$(BUILD_DIR)/kernel.elf of=$@ oflag=append conv=notrunc
+	truncate -s 1M $@
+
+$(BUILD_DIR):
+	mkdir -p $@
+	mkdir -p $(BUILD_DIR)/klibc
+
+run:
+	qemu-system-i386 -s -S $(DISK_IMG)
+
+gdb:
+	gdb -x gdb.txt
 
 clean:
-	rm -rf build
+	rm -rf $(BUILD_DIR)
+
+clean-cross:
+	rm -rf $(CROSS_DIR)
+	rm -rf .build.env
+
+clean-all: clean clean-cross

README.md

@@ -35,6 +35,7 @@ You’ll need the following tools installed:
 - `qemu-system-i386`
 
 Optional:
+
 - `gdb`
 - `vncviewer` (TigerVNC or similar)
 
@@ -42,10 +43,27 @@ Optional:
 
 ## 🛠️ Building ClassicOS
 
-Clone and build:
+Clone repository:
 
-```bash
+```sh
 git clone https://github.com/gbowne1/ClassicOS.git
 cd ClassicOS
+```
+
+Run `configure` script to build a cross-compiler toolchain for `i386-elf`:
+
+```sh
+./configure
+```
+
+Source the `.build.env` file to add the cross-compiler toolchain to your PATH:
+
+```sh
+source .build.env
+```
+
+Build the kernel:
+
+```sh
 make
 ```

bootloader/README.md

@@ -0,0 +1,27 @@
+# ClassicOS 2-stage bootloader
+
+Bootloader documentation for ClassicOS
+
+## Disk image organization:
+
+```
+[  512 B  ] [      2048 B      ] [    Unspecified    ]
+  Stage 1         Stage 2               Kernel
+```
+
+## Stage 1 (`stage1.asm`)
+
+Responsible for loading the second stage using BIOS routines, and switching to protected mode.
+
+- Queries CHS parameters from BIOS
+- Loads the second stage bootloader (2048 B) to `0x7c00`
+- Sets up a GDT with descriptor entries for code and data both covering the whole 32-bit address space
+- Enables A20
+- Set CR0.PE (enable protected mode) and jump to stage 2
+
+## Stage 2 (`stage2.asm, stage2_load.c`)
+
+- Set up segment registers
+- Load the kernel ELF header
+- Parse the program headers, and load all `PT_LOAD` segments from disk
+- Jump to the kernel entry

bootloader/boot.asm

@@ -1,67 +0,0 @@
-[BITS 16]
-[ORG 0x7C00]
-
-start:
-    cli
-    xor ax, ax
-    mov ds, ax
-    mov es, ax
-    mov ss, ax
-    mov sp, 0x7C00
-
-    call enable_a20
-    call setup_gdt
-    call switch_to_pm
-
-; ----------------------
-; A20 Gate Enable (Fast method)
-enable_a20:
-    in al, 0x92
-    or al, 0x02
-    out 0x92, al
-    ret
-
-; ----------------------
-; Set up a minimal GDT
-gdt_start:
-    dq 0x0000000000000000          ; null descriptor
-    dq 0x00CF9A000000FFFF          ; code segment descriptor
-    dq 0x00CF92000000FFFF          ; data segment descriptor
-
-gdt_descriptor:
-    dw gdt_end - gdt_start - 1     ; size of GDT
-    dd gdt_start                   ; address of GDT
-
-gdt_end:
-
-setup_gdt:
-    lgdt [gdt_descriptor]
-    ret
-
-; ----------------------
-; Switch to Protected Mode
-switch_to_pm:
-    cli
-    mov eax, cr0
-    or eax, 1
-    mov cr0, eax
-    jmp 0x08:protected_mode_entry
-; ----------------------
-; 32-bit Protected Mode Entry Point
-[BITS 32]
-protected_mode_entry:
-    mov ax, 0x10       ; data segment selector
-    mov ds, ax
-    mov es, ax
-    mov fs, ax
-    mov gs, ax
-    mov ss, ax
-    mov esp, 0x90000
-
-    ; Kernel is assumed to be loaded at 0x100000
-    jmp 0x08:0x100000
-
-; ----------------------
-; Boot signature
-times 510 - ($ - $$) db 0
-dw 0xAA55

bootloader/stage1.asm

@@ -0,0 +1,292 @@
+; ==============================================================================
+; boot.asm - First Stage Bootloader (CHS Based)
+; ==============================================================================
+
+; Params for stage2
+%define s2_addr 1               ; stage2 disk offset, in sectors
+%define s2_laddr 0x7e00         ; stage2 load address
+%define s2_size 2048            ; stage2 size
+%define s2_nsect s2_size / 512  ; stage2 size in sectors
+
+[BITS 16]
+global _start
+
+_start:
+    cli                         ; Disable interrupts
+
+    mov [bootdev], dl           ; Save boot device number (from BIOS in DL)
+
+    ; Setup stack safely below EBDA area (choose 0x0000:0x7A00)
+    xor ax, ax                  ; AX = 0
+    mov ss, ax                  ; Stack segment = 0x0000
+    mov sp, 0x7A00              ; Stack offset = 0x7A00
+
+    ; Initialize DS, ES for zero-based segments
+    xor ax, ax
+    mov ds, ax
+    mov es, ax
+
+    ; Query bios for disk parameters
+    call get_disk_params
+
+    ; Load second-stage bootloader (boot1.asm) to 0x7E00
+    mov ax, 1                   ; LBA of boot1.asm (starts at sector 1)
+    call lba_to_chs
+    mov al, s2_nsect            ; Number of sectors to read
+    mov bx, 0x7E00              ; Destination address offset ES = 0 (0x0000:0x7E00)
+    call read_chs
+
+    ; Enable A20 line
+    call enable_a20
+    jc a20_error                ; Jump if A20 enable fails
+
+    ; Setup Global Descriptor Table
+    call setup_gdt
+
+    ; Switch to protected mode and jump to second stage at 0x08:0x7E00
+    call switch_to_pm
+
+disk_error:
+    mov si, disk_error_msg
+    call print_string_16
+    jmp halt
+
+a20_error:
+    mov si, a20_error_msg
+    call print_string_16
+    jmp halt
+
+; ----------------------------------------------------------------
+; Verify Checksum Routine
+; Uses SI = start address, CX = byte count
+; Simple XOR checksum over bytes, expects result 0
+
+verify_checksum:
+    push ax
+    push bx
+    push di
+    mov di, si
+    xor al, al
+    xor bx, bx
+
+.verify_loop:
+    lodsb
+    xor bl, al
+    loop .verify_loop
+
+    test bl, bl
+    jz .checksum_ok
+    stc                    ; Set carry on checksum error
+    jmp .done
+
+.checksum_ok:
+    clc                    ; Clear carry on success
+
+.done:
+    pop di
+    pop bx
+    pop ax
+    ret
+
+get_disk_params:
+    mov  ah, 08h          ; BIOS: Get Drive Parameters
+    int  13h
+    ; TODO: error checking
+
+    ; CL bits 0–5 contain sectors per track
+    mov  al, cl
+    and  al, 3Fh           ; mask bits 0–5
+    mov  ah, 0
+    mov  [sectors_per_track], ax
+
+    ; DH = maximum head number (0-based)
+    mov  al, dh
+    inc  ax                ; convert to count (heads = maxhead + 1)
+    mov  [heads_per_cylinder], ax
+    ret
+
+; ----------------------------------------------------------------
+; CHS Disk Read Routine
+; AL = number of sectors
+; CL = starting sector (1-based)
+; Inputs:
+; AL = sector count
+; CH = cylinder
+; DH = head
+; CL = sector (1–63, with top 2 bits as high cylinder bits)
+
+; ----------------------------------------------------------------
+; Convert LBA to CHS
+; Inputs:
+;   AX = LBA sector number (0-based)
+; Outputs:
+;   CH = cylinder
+;   DH = head
+;   CL = sector (1-63, top 2 bits are upper cylinder bits)
+
+lba_to_chs:
+    ; Sector
+    xor dx, dx
+    mov bx, ax
+    div word [sectors_per_track] ; divide lba with max sectors
+    add dl, 1 ; take the remainder, sectors start at 1
+    mov cl, dl ; sector is in cl
+
+    ; Head
+    mov ax, bx
+    mov dx, 0
+    div word [sectors_per_track] ; divide lba with max sectors
+    mov dx, 0
+    div word [heads_per_cylinder] ; divide quotient with heads
+    mov dh, dl ; take the remainder, head is in dh
+
+    ; Cylinder
+    mov ch, al ; take the quotient, cylinder is in ch
+    ret
+
+read_chs:
+    pusha
+    push dx
+
+.retry:
+    mov ah, 0x02         ; BIOS: Read sectors
+    mov dl, [bootdev]    ; Boot device
+    ; Assume CH, DH, CL already set before this call
+    int 0x13
+    jc .error
+    pop dx
+    popa
+    ret
+
+.error:
+    dec cx
+    jz disk_error
+    xor ah, ah
+    int 0x13
+    jmp .retry
+
+; ----------------------------------------------------------------
+enable_a20:
+    ; Try fast A20 gate method
+    in al, 0x92
+    or al, 0x02
+    and al, 0xFE              ; Clear bit 0 to avoid fast A20 bugs
+    out 0x92, al
+
+    ; Verify A20
+    call check_a20
+    jnc .done                 ; Success
+
+    ; Fallback: use keyboard controller method
+    call .fallback
+
+.done:
+    ret
+
+.fallback:
+    mov al, 0xAD              ; Disable keyboard
+    out 0x64, al
+    call .wait_input_clear
+
+    mov al, 0xD0              ; Command: read output port
+    out 0x64, al
+    call .wait_output_full
+    in al, 0x60
+    or al, 0x02               ; Set A20 enable bit
+    mov bl, al
+
+    call .wait_input_clear
+    mov al, 0xD1              ; Command: write output port
+    out 0x64, al
+    call .wait_input_clear
+    mov al, bl
+    out 0x60, al
+
+    call .wait_input_clear
+    mov al, 0xAE              ; Enable keyboard
+    out 0x64, al
+    ret
+
+.wait_input_clear:
+    in al, 0x64
+    test al, 0x02
+    jnz .wait_input_clear
+    ret
+
+.wait_output_full:
+    in al, 0x64
+    test al, 0x01
+    jz .wait_output_full
+    ret
+
+check_a20:
+    in al, 0x64              ; Read keyboard controller status
+    test al, 0x02            ; Check if input buffer is full
+    jnz check_a20            ; Wait until empty
+
+    mov al, 0xD0             ; Command: read output port
+    out 0x64, al
+
+.wait_output:
+    in al, 0x60              ; Read output port value
+    test al, 0x02            ; Check A20 gate bit (bit 1)
+    jnz .a20_enabled         ; If set, A20 enabled
+
+    xor al, al               ; Clear carry to indicate failure
+    stc                      ; Set carry for failure, jc will jump
+
+    ret
+
+.a20_enabled:
+    clc                      ; Clear carry flag to indicate success
+    ret
+
+; ----------------------------------------------------------------
+gdt_start:
+    dq 0x0000000000000000         ; Null descriptor
+    dq 0x00CF9A000000FFFF         ; 32-bit code segment (selector 0x08)
+    dq 0x00CF92000000FFFF         ; 32-bit data segment (selector 0x10)
+    dq 0x00009A000000FFFF         ; 16-bit code segment for real mode (selector 0x18)
+
+gdt_descriptor:
+    dw gdt_end - gdt_start - 1
+    dd gdt_start
+gdt_end:
+
+setup_gdt:
+    lgdt [gdt_descriptor]
+    ret
+
+; ----------------------------------------------------------------
+switch_to_pm:
+    cli
+    mov eax, cr0
+    or eax, 1
+    mov cr0, eax
+    jmp 0x08:0x7E00             ; jump to S2
+
+; ----------------------------------------------------------------
+print_string_16:
+.loop:
+    lodsb
+    or al, al
+    jz .done
+    mov ah, 0x0E
+    int 0x10
+    jmp .loop
+.done:
+    ret
+
+disk_error_msg db "Disk error!", 0
+a20_error_msg db "A20 error!", 0
+
+halt:
+    cli
+    hlt
+
+bootdev db 0
+sectors_per_track dw 0
+heads_per_cylinder dw 0
+
+times 510 - ($ - $$) db 0
+dw 0xAA55

bootloader/stage2.asm

@@ -0,0 +1,26 @@
+[BITS 32]
+global _start
+global ata_lba_read
+
+extern load_kernel
+
+_start:
+    ; Set up segments
+    ; Data segments
+    mov ax, 0x10
+    mov ds, ax
+    mov es, ax
+    mov fs, ax
+    mov gs, ax
+    mov ss, ax
+
+    ; Code segment
+    mov ax, 0x08
+    mov cs, ax
+
+    ; Stack (must be identity-mapped)
+    mov esp, 0x90000
+
+    call load_kernel
+
+    jmp eax

bootloader/stage2.ld

@@ -1,15 +1,10 @@
-ENTRY(start)
-
 SECTIONS {
-    . = 1M;
+    . = 0x7e00;
-
-    .text : {
-        *(.multiboot)
-        *(.text*)
-    }
 
+    .text : { *(.text*) }
     .rodata : { *(.rodata*) }
     .data : { *(.data*) }
+
     .bss : {
         *(.bss*)
         *(COMMON)

bootloader/stage2_load.c

@@ -0,0 +1,235 @@
+#include <stddef.h>
+#include <stdint.h>
+
+// ATA IO Ports
+#define ATA_PRIMARY_DATA            0x1F0
+#define ATA_PRIMARY_ERR_FEATURES    0x1F1
+#define ATA_PRIMARY_SEC_COUNT       0x1F2
+#define ATA_PRIMARY_LBA_LOW         0x1F3
+#define ATA_PRIMARY_LBA_MID         0x1F4
+#define ATA_PRIMARY_LBA_HIGH        0x1F5
+#define ATA_PRIMARY_DRIVE_SEL       0x1F6
+#define ATA_PRIMARY_COMM_STAT       0x1F7
+
+// ATA Commands
+#define ATA_CMD_READ_PIO            0x20
+#define ATA_CMD_WRITE_PIO           0x30
+
+// ELF Ident indexes
+#define EI_NIDENT                   16
+
+// Program header types
+#define PT_NULL                     0
+#define PT_LOAD                     1
+
+// Disk sector size
+#define SECTOR_SIZE                 512
+#define PH_PER_SECTOR               (SECTOR_SIZE / sizeof(Elf32_Phdr))
+
+// Kernel start LBA
+#define KERN_START_SECT             5
+
+// VGA
+// Expects bios initialization for text mode (3), buffer at 0xb8000
+#define VGA_ADDRESS                 0xB8000
+#define VGA_COLS                    80
+#define VGA_ROWS                    25
+
+// ELF Header (32-bit)
+typedef struct {
+    uint8_t  e_ident[EI_NIDENT];
+    uint16_t e_type;
+    uint16_t e_machine;
+    uint32_t e_version;
+    uint32_t e_entry;     // Entry point
+    uint32_t e_phoff;     // Program header table offset
+    uint32_t e_shoff;     // Section header table offset
+    uint32_t e_flags;
+    uint16_t e_ehsize;
+    uint16_t e_phentsize;
+    uint16_t e_phnum;
+    uint16_t e_shentsize;
+    uint16_t e_shnum;
+    uint16_t e_shstrndx;
+} __attribute__((packed)) Elf32_Ehdr;
+
+// Program Header (32-bit)
+typedef struct {
+    uint32_t p_type;
+    uint32_t p_offset;
+    uint32_t p_vaddr;
+    uint32_t p_paddr;
+    uint32_t p_filesz;
+    uint32_t p_memsz;
+    uint32_t p_flags;
+    uint32_t p_align;
+} __attribute__((packed)) Elf32_Phdr;
+
+static inline uint8_t inb(uint16_t port)
+{
+    uint8_t ret;
+    __asm__ volatile ("inb %1, %0"
+                      : "=a"(ret)
+                      : "Nd"(port));
+    return ret;
+}
+
+static inline void outb(uint16_t port, uint8_t val)
+{
+    __asm__ volatile ("outb %0, %1"
+                      :
+                      : "a"(val), "Nd"(port));
+}
+
+static inline uint16_t inw(uint16_t port)
+{
+    uint16_t ret;
+    __asm__ volatile ("inw %1, %0"
+                      : "=a"(ret)
+                      : "Nd"(port));
+    return ret;
+}
+
+static inline void ata_wait_bsy() {
+    while (inb(ATA_PRIMARY_COMM_STAT) & 0x80);
+}
+
+static inline void ata_wait_drq() {
+    while (!(inb(ATA_PRIMARY_COMM_STAT) & 0x08));
+}
+
+static void ata_read_sector(void *addr, uint32_t lba) {
+    ata_wait_bsy();
+
+    outb(ATA_PRIMARY_DRIVE_SEL, 0xE0 | ((lba >> 24) & 0x0F));
+    outb(ATA_PRIMARY_SEC_COUNT, 1);
+    outb(ATA_PRIMARY_LBA_LOW, (uint8_t)lba);
+    outb(ATA_PRIMARY_LBA_MID, (uint8_t)(lba >> 8));
+    outb(ATA_PRIMARY_LBA_HIGH, (uint8_t)(lba >> 16));
+    outb(ATA_PRIMARY_COMM_STAT, ATA_CMD_READ_PIO);
+
+    uint16_t* ptr = (uint16_t*)addr;
+    ata_wait_bsy();
+    ata_wait_drq();
+    for (int i = 0; i < 256; i++) {
+        *ptr++ = inw(ATA_PRIMARY_DATA);
+    }
+}
+
+static void ata_read_sectors(uint8_t *addr, uint32_t offset, uint32_t size)
+{
+    // Reads are offset from the starting sector of the kernel
+    uint32_t lba = KERN_START_SECT + offset / SECTOR_SIZE;
+    uint32_t off = offset % 512;
+    uint8_t data[512];
+
+    while (size > 0) {
+        ata_read_sector(data, lba);
+
+        uint32_t copy = 512 - off;
+        if (copy > size) {
+            copy = size;
+        }
+
+        for (uint32_t i = 0; i < copy; i++) {
+            addr[i] = data[off + i];
+        }
+
+        addr += copy;
+        size -= copy;
+        lba++;
+        off = 0;
+    }
+}
+
+static void on_error(const char *msg)
+{
+    uint16_t *ptr = (uint16_t *)VGA_ADDRESS;
+
+    // Clear
+    uint16_t val = 0x0f00 | (uint8_t)' ';
+    for (size_t i = 0; i < VGA_COLS * VGA_ROWS; i++) {
+        ptr[i] = val;
+    }
+
+    // Print error
+    for (size_t i = 0; msg[i]; i++) {
+        ptr[i] = 0xf00 | (uint8_t)msg[i];
+    }
+
+    // Halt
+    while (1) {
+        __asm__("hlt");
+    }
+}
+
+// Load an ELF executable into memory.
+// NOTE: Only 32-byte program headers are supported.
+// Returns the entry point to the program.
+static void *elf_load(const void *data) {
+    const Elf32_Ehdr *header = (const Elf32_Ehdr*)data;
+
+    if (header->e_phentsize != sizeof(Elf32_Phdr)) {
+        // The bootloader only handles 32-byte program header entries
+        on_error("ERROR: Unsupported program header entry size, halting...");
+    }
+
+    // Buffer for the program headers
+    uint8_t file_buf[SECTOR_SIZE];
+
+    // Current file offset to the next program header
+    uint32_t file_offset = header->e_phoff;
+
+    for (int i = 0; i < header->e_phnum; i++) {
+        // Check for sector boundary.
+        // Program headers are read in a sector at a time
+        // 512 / 32 = 16 PH per sector
+        if (i % PH_PER_SECTOR == 0) {
+            uint32_t count = (header->e_phnum - i) * sizeof(Elf32_Phdr);
+            if (count > SECTOR_SIZE) {
+                count = SECTOR_SIZE;
+            }
+
+            // Reads
+            ata_read_sectors(file_buf, file_offset, count);
+            file_offset += count;
+        }
+
+        // PH being processed currently, index mod 16 as headers
+        // are being loaded in sector by sector.
+        const Elf32_Phdr *ph = (const Elf32_Phdr *)file_buf + (i % PH_PER_SECTOR);
+
+        // Discard non-load segments
+        if (ph->p_type != PT_LOAD)
+            continue;
+
+        // Load in the segment
+        uint32_t offset = ph->p_offset;
+        uint32_t filesz = ph->p_filesz;
+        uint32_t memsz = ph->p_memsz;
+        uint8_t *vaddr = (uint8_t *)ph->p_vaddr;
+        ata_read_sectors(vaddr, offset, filesz);
+
+        // Zero remaining BSS (if any)
+        if (memsz > filesz) {
+            uint8_t* bss_start = vaddr + filesz;
+            for (uint32_t j = 0; j < memsz - filesz; j++) {
+                bss_start[j] = 0;
+            }
+        }
+    }
+
+    // Return the entry point
+    return (void *)header->e_entry;
+}
+
+void *load_kernel(void) {
+    // ELF header buffer
+    uint8_t header_buf[SECTOR_SIZE];
+
+    // Read the first sector (contains the ELF header)
+    ata_read_sector(header_buf, KERN_START_SECT);
+
+    // `elf_load()` returns the entry point
+    return elf_load(header_buf);
+}

configure

@@ -0,0 +1,169 @@
+#!/usr/bin/env bash
+set -euo pipefail
+
+# Configuration
+TARGET="i386-elf"
+BINUTILS_VERSION="2.45"
+GCC_VERSION="15.2.0"
+
+# Paths
+SCRIPT_PATH="$(realpath "${BASH_SOURCE[0]}")"
+SCRIPT_DIR="$(dirname "$SCRIPT_PATH")"
+PREFIX="$SCRIPT_DIR/cross"
+SRC_DIR="$PREFIX/src"
+
+BINUTILS_SRC="$SRC_DIR/binutils-$BINUTILS_VERSION"
+BINUTILS_BUILD="$PREFIX/build-binutils"
+GCC_SRC="$SRC_DIR/gcc-$GCC_VERSION"
+GCC_BUILD="$PREFIX/build-gcc"
+
+# Flags
+DEBUG=0
+HELP=0
+
+# Parse arguments
+for arg in "$@"; do
+  case "$arg" in
+    -h|--help)
+      HELP=1
+      ;;
+    -d|--debug)
+      DEBUG=1
+      ;;
+    *)
+      echo "Unknown option: $arg"
+      echo "Use -h or --help for usage information"
+      exit 1
+      ;;
+  esac
+done
+
+# Show help
+if [[ "$HELP" -eq 1 ]]; then
+  cat << EOF
+Usage: $0 [OPTIONS]
+
+Build a cross-compiler toolchain for $TARGET.
+
+OPTIONS:
+  -h, --help     Show this help message
+  -d, --debug    Enable debug mode (set -x)
+
+This script will:
+  1. Download binutils $BINUTILS_VERSION and GCC $GCC_VERSION
+  2. Build and install them to: $PREFIX
+
+EOF
+  exit 0
+fi
+
+# Enable debug mode
+if [[ "$DEBUG" -eq 1 ]]; then
+  set -x
+fi
+
+# Print configuration
+cat << EOF
+
+=== Build Configuration ===
+Target       : $TARGET
+Prefix       : $PREFIX
+Binutils     : $BINUTILS_VERSION
+GCC          : $GCC_VERSION
+===========================
+
+EOF
+
+# Create directory structure
+echo "Setting up directories..."
+mkdir -p "$SRC_DIR"
+
+# Download sources
+cd "$SRC_DIR"
+
+if [[ ! -d "$BINUTILS_SRC" ]]; then
+  echo "Downloading binutils $BINUTILS_VERSION..."
+  wget "https://ftp.gnu.org/gnu/binutils/binutils-$BINUTILS_VERSION.tar.gz"
+  echo "Extracting binutils..."
+  tar xf "binutils-$BINUTILS_VERSION.tar.gz"
+  rm "binutils-$BINUTILS_VERSION.tar.gz"
+else
+  echo "Binutils source already exists, skipping download"
+fi
+
+if [[ ! -d "$GCC_SRC" ]]; then
+  echo "Downloading GCC $GCC_VERSION..."
+  wget "https://ftp.gnu.org/gnu/gcc/gcc-$GCC_VERSION/gcc-$GCC_VERSION.tar.gz"
+  echo "Extracting GCC..."
+  tar xf "gcc-$GCC_VERSION.tar.gz"
+  rm "gcc-$GCC_VERSION.tar.gz"
+else
+  echo "GCC source already exists, skipping download"
+fi
+
+# Download GCC prerequisites
+if [[ ! -d "$GCC_SRC/gmp" ]]; then
+  echo "Downloading GCC prerequisites..."
+  cd "$GCC_SRC"
+  ./contrib/download_prerequisites
+  cd "$SRC_DIR"
+else
+  echo "GCC prerequisites already downloaded, skipping"
+fi
+
+# Build binutils
+if [[ ! -f "$PREFIX/bin/$TARGET-ld" ]]; then
+  echo "Building binutils..."
+  mkdir -p "$BINUTILS_BUILD"
+  cd "$BINUTILS_BUILD"
+
+  "$BINUTILS_SRC/configure" \
+    --target="$TARGET" \
+    --prefix="$PREFIX" \
+    --with-sysroot \
+    --disable-nls \
+    --disable-werror
+
+  make -j"$(nproc)"
+  make install
+else
+  echo "Binutils already installed, skipping build"
+fi
+
+# Build GCC
+if [[ ! -f "$PREFIX/bin/$TARGET-gcc" ]]; then
+  echo "Building GCC..."
+  mkdir -p "$GCC_BUILD"
+  cd "$GCC_BUILD"
+
+  "$GCC_SRC/configure" \
+    --target="$TARGET" \
+    --prefix="$PREFIX" \
+    --disable-nls \
+    --enable-languages=c \
+    --without-headers
+
+  make all-gcc -j"$(nproc)"
+  make all-target-libgcc -j"$(nproc)"
+  make install-gcc
+  make install-target-libgcc
+else
+  echo "GCC already installed, skipping build"
+fi
+
+cd "$SCRIPT_DIR"
+
+# Generate .build.env file
+cat > .build.env << EOF
+# Generated by configure on $(date)
+# Source this file to add the cross-compiler toolchain to your PATH
+export PATH="$PREFIX/bin:\$PATH"
+EOF
+
+echo ""
+echo "=== Build Complete ==="
+echo "Toolchain installed to: $PREFIX"
+echo ""
+echo "To use the toolchain, run:"
+echo "  source .build.env"
+echo "======================"

gdb.txt

@@ -0,0 +1,6 @@
+target remote :1234
+add-symbol-file build/stage1.elf
+add-symbol-file build/stage2.elf
+add-symbol-file build/kernel.elf
+hbreak *0x7c00
+c

kernel/acpi.c

@@ -0,0 +1,54 @@
+#include "acpi.h"
+#include <stdint.h>
+#include <stdbool.h>
+#include <stddef.h>
+#include <string.h>
+
+// Function to find the RSDP (Root System Description Pointer)
+acpi_rsdp_t* acpi_find_rsdp() {
+    // Search memory from 0x000E0000 to 0x00100000 (BIOS)
+    for (uint32_t addr = 0x000E0000; addr < 0x00100000; addr += 16) {
+        acpi_rsdp_t* rsdp = (acpi_rsdp_t*)addr;
+        if (memcmp(rsdp->signature, "RSD PTR ", 8) == 0) {
+            uint8_t checksum = 0;
+            for (size_t i = 0; i < sizeof(acpi_rsdp_t); i++) { // Change int to size_t
+                checksum += ((uint8_t*)rsdp)[i];
+            }
+            if (checksum == 0) {
+                return rsdp;  // Valid RSDP found
+            }
+        }
+    }
+    return NULL;  // RSDP not found
+}
+
+// Function to get the RSDT or XSDT based on the RSDP revision
+void* acpi_get_rsdt_or_xsdt(acpi_rsdp_t* rsdp) {
+    if (rsdp->revision >= 2) {
+        return (void*)rsdp->xsdt_addr;  // ACPI 2.0+ uses XSDT
+    } else {
+        return (void*)rsdp->rsdt_addr;  // ACPI 1.0 uses RSDT
+    }
+}
+
+// Function to find the FADT table within the RSDT or XSDT
+acpi_fadt_t* acpi_find_fadt(void* rsdt_or_xsdt) {
+    acpi_rsdt_t* rsdt = (acpi_rsdt_t*)rsdt_or_xsdt;
+    uint32_t num_tables = (rsdt->length - sizeof(acpi_rsdt_t)) / sizeof(uint32_t);
+
+    for (size_t i = 0; i < num_tables; i++) {
+        uint32_t table_addr = rsdt->tables[i];
+        acpi_fadt_t* fadt = (acpi_fadt_t*)table_addr;
+        if (fadt->signature == 0x50434146) {
+            uint8_t checksum = 0;
+            for (size_t j = 0; j < fadt->length; j++) {
+                checksum += ((uint8_t*)fadt)[j];
+            }
+            if (checksum == 0) {
+                return fadt;
+            }
+        }
+    }
+    return NULL;
+}
+

kernel/acpi.h

@@ -0,0 +1,47 @@
+#ifndef ACPI_H
+#define ACPI_H
+
+#include <stdint.h>
+
+// ACPI base address (replace with actual value)
+#define ACPI_BASE 0xE0000000
+
+#ifndef NULL
+#define NULL ((void*)0)
+#endif
+
+// ACPI RSDP Structure (Root System Description Pointer)
+typedef struct {
+    uint8_t signature[8];      // Should be "RSD PTR "
+    uint8_t checksum;          // Checksum for the RSDP structure
+    uint8_t oem_id[6];         // OEM ID string
+    uint8_t revision;          // ACPI revision
+    uint32_t rsdt_addr;        // 32-bit RSDT address (ACPI 1.0)
+    uint32_t xsdt_addr;        // 64-bit XSDT address (ACPI 2.0+)
+} __attribute__((packed)) acpi_rsdp_t;
+
+// ACPI RSDT Structure (Root System Description Table)
+typedef struct {
+    uint32_t signature;    // Should be "RSDT"
+    uint32_t length;       // Length of the table
+    uint8_t revision;      // Revision of the RSDT table
+    uint8_t checksum;      // Checksum for the RSDT table
+    uint32_t tables[];     // Array of pointers to other tables (RSDT/XSDT entries)
+} __attribute__((packed)) acpi_rsdt_t;
+
+// ACPI FADT Structure (Fixed ACPI Description Table)
+typedef struct {
+    uint32_t signature;         // Should be "FACP"
+    uint32_t length;            // Length of the table
+    uint8_t revision;           // Revision of the FADT table
+    uint8_t checksum;           // Checksum for the FADT table
+    uint32_t pm_tmr_address;    // Power Management Timer Address
+    // ... other FADT fields
+} __attribute__((packed)) acpi_fadt_t;
+
+// Function prototypes
+acpi_rsdp_t* acpi_find_rsdp();
+void* acpi_get_rsdt_or_xsdt(acpi_rsdp_t* rsdp);
+acpi_fadt_t* acpi_find_fadt(void* rsdt_or_xsdt);
+
+#endif /* ACPI_H */

kernel/context_switch.asm

@@ -0,0 +1,25 @@
+global ctx_switch
+
+; void ctx_switch(uint32_t **old_sp_ptr, uint32_t *new_sp);
+; Arguments on stack (cdecl convention):
+; [ESP + 4] -> old_sp_ptr (pointer to the 'stack_ptr' field of current task)
+; [ESP + 8] -> new_sp     (value of 'stack_ptr' of the next task)
+
+ctx_switch:
+    ; 1. Save the context of the CURRENT task
+    pushf               ; Save EFLAGS (CPU status flags)
+    pusha               ; Save all General Purpose Regs (EAX, ECX, EDX, EBX, ESP, EBP, ESI, EDI)
+
+    ; 2. Save the current stack pointer (ESP) into the pointer passed as 1st arg
+    mov eax, [esp + 40] ; Get 1st argument (old_sp_ptr). Offset 40 = 36 (regs) + 4 (ret addr)
+    mov [eax], esp      ; *old_sp_ptr = ESP
+
+    ; 3. Load the stack pointer of the NEW task
+    mov esp, [esp + 44] ; Get 2nd argument (new_sp). Offset 44 = 40 + 4
+
+    ; 4. Restore the context of the NEW task
+    popa                ; Restore all General Purpose Regs
+    popf                ; Restore EFLAGS
+
+    ; 5. Jump to the new task (The 'ret' pops EIP from the new stack)
+    ret

kernel/cpu.c

@@ -0,0 +1,107 @@
+#include "cpu.h"
+#include "serial.h"
+#include "terminal.h"
+#include "utils.h"
+
+void cpuid(uint32_t leaf, uint32_t *eax, uint32_t *ebx, uint32_t *ecx, uint32_t *edx) {
+    __asm__(
+        "cpuid"
+        : "=a"(*eax), "=b"(*ebx), "=c"(*ecx), "=d"(*edx)
+        : "a"(leaf)
+    );
+}
+
+// Helper to print a labeled decimal value
+void print_val(const char* label, uint32_t val) {
+    char buf[12];
+    utoa(val, buf, 10);
+    terminal_write(label);
+    terminal_write(buf);
+    terminal_write(" ");
+}
+
+// Safely check if CPUID is supported by attempting to flip bit 21 of EFLAGS
+int check_cpuid_supported() {
+    uint32_t f1, f2;
+    __asm__ volatile (
+        "pushfl\n\t"
+        "pushfl\n\t"
+        "popl %0\n\t"
+        "movl %0, %1\n\t"
+        "xorl $0x200000, %0\n\t"
+        "pushl %0\n\t"
+        "popfl\n\t"
+        "pushfl\n\t"
+        "popl %0\n\t"
+        "popfl\n\t"
+        : "=&r" (f1), "=&r" (f2));
+    return ((f1 ^ f2) & 0x200000) != 0;
+}
+
+void identify_cpu() {
+    if (!check_cpuid_supported()) {
+        terminal_write("CPUID not supported. Likely a 386 or early 486.\n");
+        return;
+    }
+
+    uint32_t eax, ebx, ecx, edx;
+    char vendor[13];
+
+    // Leaf 0: Vendor String & Max Leaf
+    cpuid(0, &eax, &ebx, &ecx, &edx);
+    uint32_t max_leaf = eax;
+    *(uint32_t *)&vendor[0] = ebx;
+    *(uint32_t *)&vendor[4] = edx;
+    *(uint32_t *)&vendor[8] = ecx;
+    vendor[12] = '\0';
+
+    terminal_write("Vendor: ");
+    terminal_write(vendor);
+    terminal_write("\n");
+
+    // Leaf 1: Family, Model, Stepping
+    if (max_leaf >= 1) {
+        cpuid(1, &eax, &ebx, &ecx, &edx);
+        
+        uint32_t stepping = eax & 0xF;
+        uint32_t model = (eax >> 4) & 0xF;
+        uint32_t family = (eax >> 8) & 0xF;
+        uint32_t type = (eax >> 12) & 0x3;
+
+        // Handle Extended Family/Model (Required for Pentium 4 and newer)
+        if (family == 0xF) {
+            family += (eax >> 20) & 0xFF;
+            model += ((eax >> 16) & 0xF) << 4;
+        }
+
+        print_val("Family:", family);
+        print_val("Model:", model);
+        print_val("Step:", stepping);
+        terminal_write("\n");
+    }
+
+    // Leaf 2: Cache Descriptors
+    if (max_leaf >= 2) {
+        cpuid(2, &eax, &ebx, &ecx, &edx);
+        
+        terminal_write("Cache Descriptors: ");
+        // Note: Leaf 2 returns a list of 1-byte descriptors in the registers.
+        // We look for common Intel ones:
+        uint32_t regs[4] = {eax, ebx, ecx, edx};
+        for (int i = 0; i < 4; i++) {
+            if (regs[i] & 0x80000000) continue; // Reserved bit
+            for (int j = 0; j < 4; j++) {
+                uint8_t desc = (regs[i] >> (j * 8)) & 0xFF;
+                if (desc == 0) continue;
+                
+                // Example decoding for specific chips you mentioned:
+                if (desc == 0x06) terminal_write("8KB L1 I-Cache ");
+                if (desc == 0x0A) terminal_write("8KB L1 D-Cache ");
+                if (desc == 0x41) terminal_write("128KB L2 ");
+                if (desc == 0x43) terminal_write("512KB L2 ");
+                if (desc == 0x2C) terminal_write("32KB L1 D-Cache ");
+            }
+        }
+        terminal_write("\n");
+    }
+}

kernel/cpu.h

@@ -0,0 +1,43 @@
+#ifndef CPU_H
+#define CPU_H
+
+#include <stdint.h>
+#include <stdbool.h>
+
+// Specific Intel Model Definitions for your targets
+#define INTEL_FAM4_486_DX      0x00 // Also 0x01
+#define INTEL_FAM4_486_SX      0x02
+#define INTEL_FAM4_486_DX2     0x03
+#define INTEL_FAM4_486_DX4     0x08
+#define INTEL_FAM5_PENTIUM     0x01 // P5
+#define INTEL_FAM5_PENTIUM_MMX 0x04 // P55C
+#define INTEL_FAM6_PENTIUM_PRO 0x01 // P6
+#define INTEL_FAM6_PENTIUM_II  0x05 // Deschutes
+#define INTEL_FAM6_PENTIUM_III 0x07 // Katmai/Coppermine
+#define INTEL_FAM15_P4_WILLY   0x00 // Willamette
+#define INTEL_FAM15_P4_NORTH   0x02 // Northwood
+#define INTEL_FAM15_P4_PRES    0x03 // Prescott
+
+typedef struct {
+    char vendor[13];
+    uint32_t family;
+    uint32_t model;
+    uint32_t stepping;
+    uint32_t type;
+    uint32_t max_leaf;
+    
+    // Feature flags (optional, but very helpful later)
+    bool has_fpu;
+    bool has_mmx;
+    bool has_sse;
+} cpu_info_t;
+
+// Function Prototypes
+void cpuid(uint32_t leaf, uint32_t *eax, uint32_t *ebx, uint32_t *ecx, uint32_t *edx);
+bool cpu_check_cpuid_support(void);
+void identify_cpu(void);
+
+// Helper to get the current CPU info after identification
+cpu_info_t* cpu_get_info(void);
+
+#endif // CPU_H

kernel/debug.c

@@ -1,5 +1,45 @@
 #include "debug.h"
+#include "vga.h"
+#include <stdint.h>
 
+#define VGA_WIDTH 80
+#define VGA_HEIGHT 25
+#define VGA_MEMORY 0xB8000
+
+// VGA text mode color attributes
+#define COLOR_WHITE 0x07
+
+// Pointer to the VGA memory
+volatile uint16_t* vga_buffer = (uint16_t*)VGA_MEMORY;
+
+// Function to print a string to the VGA text buffer
 void debug_print(const char *str) {
-    // Implementation for printing debug messages
+    while (*str) {
+        if (*str == '\n') {
+            // Handle new line
+            // Move to the next line (not implemented here)
+            // You can implement line wrapping if needed
+            str++;
+            continue;
+        }
+
+        // Calculate the position in the VGA buffer
+        static int cursor_x = 0;
+        static int cursor_y = 0;
+
+        // Write the character and its attribute to the VGA buffer
+        vga_buffer[cursor_y * VGA_WIDTH + cursor_x] = (COLOR_WHITE << 8) | *str;
+
+        // Move the cursor to the right
+        cursor_x++;
+        if (cursor_x >= VGA_WIDTH) {
+            cursor_x = 0;
+            cursor_y++;
+            if (cursor_y >= VGA_HEIGHT) {
+                cursor_y = 0; // Scroll up (not implemented here)
+            }
+        }
+
+        str++;
+    }
 }

kernel/display.c

@@ -0,0 +1,36 @@
+#include "display.h"
+#include "io.h" // Include your I/O header for port access
+#include "vga.h"
+
+// Initialize the display
+void init_display(void) {
+    // Initialize VGA settings, if necessary
+    // This could involve setting up the VGA mode, etc.
+    set_display_mode(0x13); // Example: Set to 320x200 256-color mode
+}
+
+// Enumerate connected displays
+void enumerate_displays(void) {
+    // This is a simplified example. Actual enumeration may require
+    // reading from specific VGA registers or using BIOS interrupts.
+    
+    // For demonstration, we will just print a message
+    // In a real driver, you would check the VGA registers
+    // to determine connected displays.
+    clear_display();
+    // Here you would typically read from VGA registers to find connected displays
+    // For example, using inb() to read from VGA ports
+}
+
+// Set the display mode
+void set_display_mode(uint8_t mode) {
+    // Set the VGA mode by writing to the appropriate registers
+    outb(VGA_PORT, mode); // Example function to write to a port
+}
+
+// Clear the display
+void clear_display(void) {
+    // Clear the display by filling it with a color
+    // This is a placeholder for actual clearing logic
+    // You would typically write to video memory here
+}

kernel/display.h

@@ -0,0 +1,14 @@
+#ifndef DISPLAY_H
+#define DISPLAY_H
+
+#include <stdint.h>
+
+#define VGA_PORT 0x3C0  // Base port for VGA
+
+// Function prototypes
+void init_display(void);
+void enumerate_displays(void);
+void set_display_mode(uint8_t mode);
+void clear_display(void);
+
+#endif // DISPLAY_H

kernel/elf.c

@@ -0,0 +1,50 @@
+#include "elf.h"
+#include <stddef.h>
+#include <string.h>
+
+// This assumes that the elf is ELF32 and little-endian Intel X86 architecture.
+
+// Check if a binary is a valid ELF executable.
+int elf_validate(const void* data) {
+    const Elf32_Ehdr* header = (const Elf32_Ehdr*)data;
+
+    if (*(uint32_t*)header->e_ident != ELF_MAGIC)
+        return 0;
+
+    if (header->e_type != ET_EXEC && header->e_type != ET_DYN)
+        return 0;
+
+    if (header->e_machine != EM_386)
+        return 0;
+
+    return 1;
+}
+
+// Load an ELF executable into memory.
+int elf_load(const void* data, void (*load_segment)(uint32_t vaddr, const void* src, uint32_t size)) {
+    const Elf32_Ehdr* header = (const Elf32_Ehdr*)data;
+    const Elf32_Phdr* ph = (const Elf32_Phdr*)((uint8_t*)data + header->e_phoff);
+
+    for (int i = 0; i < header->e_phnum; i++) {
+        if (ph[i].p_type != PT_LOAD)
+            continue;
+
+        const void* src = (uint8_t*)data + ph[i].p_offset;
+        uint32_t vaddr = ph[i].p_vaddr;
+        uint32_t filesz = ph[i].p_filesz;
+        uint32_t memsz = ph[i].p_memsz;
+
+        // Copy data segment
+        load_segment(vaddr, src, filesz);
+
+        // Zero remaining BSS (if any)
+        if (memsz > filesz) {
+            uint8_t* bss_start = (uint8_t*)(vaddr + filesz);
+            for (uint32_t j = 0; j < memsz - filesz; j++) {
+                bss_start[j] = 0;
+            }
+        }
+    }
+
+    return header->e_entry;
+}

kernel/elf.h

@@ -0,0 +1,52 @@
+#ifndef ELF_H
+#define ELF_H
+
+#include <stdint.h>
+
+#define ELF_MAGIC 0x464C457F  // "\x7FELF" in little-endian
+
+// ELF Types
+#define ET_EXEC 2
+#define ET_DYN  3
+
+// ELF Machine
+#define EM_386  3
+
+// ELF Ident indexes
+#define EI_NIDENT 16
+
+// Program header types
+#define PT_NULL    0
+#define PT_LOAD    1
+
+// ELF Header (32-bit)
+typedef struct {
+    uint8_t  e_ident[EI_NIDENT];
+    uint16_t e_type;
+    uint16_t e_machine;
+    uint32_t e_version;
+    uint32_t e_entry;     // Entry point
+    uint32_t e_phoff;     // Program header table offset
+    uint32_t e_shoff;     // Section header table offset
+    uint32_t e_flags;
+    uint16_t e_ehsize;
+    uint16_t e_phentsize;
+    uint16_t e_phnum;
+    uint16_t e_shentsize;
+    uint16_t e_shnum;
+    uint16_t e_shstrndx;
+} __attribute__((packed)) Elf32_Ehdr;
+
+// Program Header (32-bit)
+typedef struct {
+    uint32_t p_type;
+    uint32_t p_offset;
+    uint32_t p_vaddr;
+    uint32_t p_paddr;
+    uint32_t p_filesz;
+    uint32_t p_memsz;
+    uint32_t p_flags;
+    uint32_t p_align;
+} __attribute__((packed)) Elf32_Phdr;
+
+#endif // ELF_H

kernel/fat12.c

@@ -0,0 +1,196 @@
+#include "fat12.h"
+#include <stddef.h> // for NULL
+
+// --- Globals for Filesystem State ---
+static fat12_bpb_t bpb;
+static uint32_t fat_start_lba;
+static uint32_t root_dir_lba;
+static uint32_t data_start_lba;
+static uint32_t root_dir_sectors;
+
+// Scratch buffer to read sectors (avoids large stack usage)
+static uint8_t g_sector_buffer[FAT12_SECTOR_SIZE];
+
+// --- Utils (Since we don't have string.h) ---
+static int k_memcmp(const void *s1, const void *s2, uint32_t n) {
+    const uint8_t *p1 = (const uint8_t *)s1;
+    const uint8_t *p2 = (const uint8_t *)s2;
+
+    for (uint32_t i = 0; i < n; i++) {
+        if (p1[i] != p2[i]) {
+            // Correct way to return the difference:
+            // If p1[i] > p2[i], returns positive.
+            // If p1[i] < p2[i], returns negative.
+            return (int)p1[i] - (int)p2[i];
+        }
+    }
+
+    return 0;
+}
+
+// Converts "file.txt" to "FILE    TXT" for comparison
+static void to_fat_name(const char *src, char *dest) {
+    // Initialize with spaces
+    for(int i=0; i<11; i++) dest[i] = ' ';
+
+    int i = 0, j = 0;
+    // Copy Name
+    while (src[i] != '\0' && src[i] != '.' && j < 8) {
+        // Convert to uppercase (simple version)
+        char c = src[i];
+        if (c >= 'a' && c <= 'z') c -= 32; 
+        dest[j++] = c;
+        i++;
+    }
+    
+    // Skip extension dot
+    if (src[i] == '.') i++;
+
+    // Copy Extension
+    j = 8;
+    while (src[i] != '\0' && j < 11) {
+        char c = src[i];
+        if (c >= 'a' && c <= 'z') c -= 32;
+        dest[j++] = c;
+        i++;
+    }
+}
+
+// --- Core Logic ---
+
+void fat12_init() {
+    // 1. Read Boot Sector (LBA 0)
+    disk_read_sector(0, g_sector_buffer);
+
+    // 2. Copy BPB data safely
+    // We cast the buffer to our struct
+    fat12_bpb_t *boot_sector = (fat12_bpb_t*)g_sector_buffer;
+    bpb = *boot_sector;
+
+    // 3. Calculate System Offsets
+    fat_start_lba  = bpb.reserved_sectors;
+    
+    // Root Dir starts after FATs
+    // LBA = Reserved + (FatCount * SectorsPerFat)
+    root_dir_lba   = fat_start_lba + (bpb.fat_count * bpb.sectors_per_fat);
+    
+    // Calculate size of Root Directory in sectors
+    // (Entries * 32 bytes) / 512
+    root_dir_sectors = (bpb.dir_entries_count * 32 + FAT12_SECTOR_SIZE - 1) / FAT12_SECTOR_SIZE;
+
+    // Data starts after Root Directory
+    data_start_lba = root_dir_lba + root_dir_sectors;
+}
+
+// Helper: Read the FAT table to find the NEXT cluster
+static uint16_t fat12_get_next_cluster(uint16_t current_cluster) {
+    // FAT12 Offset Calculation:
+    // Offset = Cluster + (Cluster / 2)
+    uint32_t fat_offset = current_cluster + (current_cluster / 2);
+    
+    uint32_t fat_sector = fat_start_lba + (fat_offset / FAT12_SECTOR_SIZE);
+    uint32_t ent_offset = fat_offset % FAT12_SECTOR_SIZE;
+
+    // Read the sector containing the FAT entry
+    disk_read_sector(fat_sector, g_sector_buffer);
+
+    // Read 16 bits (2 bytes)
+    // Note: If ent_offset == 511, the entry spans two sectors. 
+    // For simplicity in this snippet, we ignore that edge case (rare).
+    // A robust kernel would check if(ent_offset == 511) and read next sector.
+    
+    uint16_t val = *(uint16_t*)&g_sector_buffer[ent_offset];
+
+    if (current_cluster & 1) {
+        return val >> 4;      // Odd: High 12 bits
+    } else {
+        return val & 0x0FFF;  // Even: Low 12 bits
+    }
+}
+
+file_t fat12_open(const char *filename) {
+    file_t file = {0};
+    char target_name[11];
+    to_fat_name(filename, target_name);
+
+    // Search Root Directory
+    for (uint32_t i = 0; i < root_dir_sectors; i++) {
+        disk_read_sector(root_dir_lba + i, g_sector_buffer);
+
+        fat12_entry_t *entry = (fat12_entry_t*)g_sector_buffer;
+
+        // Check all 16 entries in this sector (512 / 32 = 16)
+        for (int j = 0; j < 16; j++) {
+            if (entry[j].filename[0] == 0x00) return file; // End of Dir
+            
+            // Check if filename matches
+            if (k_memcmp(entry[j].filename, target_name, 11) == 0) {
+                // Found it!
+                file.start_cluster = entry[j].low_cluster_num;
+                file.size = entry[j].file_size;
+                
+                // Initialize file cursor
+                file.current_cluster = file.start_cluster;
+                file.bytes_read = 0;
+                return file;
+            }
+        }
+    }
+    
+    // Not found (file.start_cluster will be 0)
+    return file;
+}
+
+uint32_t fat12_read(file_t *file, uint8_t *buffer, uint32_t bytes_to_read) {
+    if (file->start_cluster == 0) return 0; // File not open
+
+    uint32_t total_read = 0;
+
+    while (bytes_to_read > 0) {
+        // Check for EOF marker in FAT12 (>= 0xFF8)
+        if (file->current_cluster >= 0xFF8) break;
+
+        // Calculate Physical LBA of current cluster
+        // LBA = DataStart + ((Cluster - 2) * SectorsPerCluster)
+        uint32_t lba = data_start_lba + ((file->current_cluster - 2) * bpb.sectors_per_cluster);
+
+        // Read the cluster
+        // NOTE: Assumes SectorsPerCluster = 1 (Standard Floppy)
+        disk_read_sector(lba, g_sector_buffer);
+
+        // Determine how much to copy from this sector
+        uint32_t chunk_size = FAT12_SECTOR_SIZE;
+        
+        // If the file is smaller than a sector, or we are at the end
+        if (chunk_size > bytes_to_read) chunk_size = bytes_to_read;
+        
+        // Check if we are reading past file size
+        if (file->bytes_read + chunk_size > file->size) {
+            chunk_size = file->size - file->bytes_read;
+        }
+
+        // Copy to user buffer
+        for (uint32_t i = 0; i < chunk_size; i++) {
+            buffer[total_read + i] = g_sector_buffer[i];
+        }
+
+        total_read += chunk_size;
+        file->bytes_read += chunk_size;
+        bytes_to_read -= chunk_size;
+
+        // If we finished this cluster, move to the next one
+        if (chunk_size == FAT12_SECTOR_SIZE) { // Or strictly logic based on position
+             file->current_cluster = fat12_get_next_cluster(file->current_cluster);
+        } else {
+            // We finished the file or the request
+            break;
+        }
+    }
+
+    return total_read;
+}
+
+int disk_read_sector(uint32_t lba, uint8_t *buffer) {
+    // For now, do nothing and return success
+    return 0; 
+}

kernel/fat12.h

@@ -0,0 +1,67 @@
+#ifndef FAT12_H
+#define FAT12_H
+
+#include <stdint.h>
+
+// --- Configuration ---
+#define FAT12_SECTOR_SIZE 512
+
+// --- On-Disk Structures (Must be Packed) ---
+
+// BIOS Parameter Block (Start of Boot Sector)
+typedef struct {
+    uint8_t  jump[3];
+    char     oem[8];
+    uint16_t bytes_per_sector;    // 512
+    uint8_t  sectors_per_cluster; // 1
+    uint16_t reserved_sectors;    // 1 (Boot sector)
+    uint8_t  fat_count;           // 2
+    uint16_t dir_entries_count;   // 224
+    uint16_t total_sectors;       // 2880
+    uint8_t  media_descriptor;    // 0xF0
+    uint16_t sectors_per_fat;     // 9
+    uint16_t sectors_per_track;   // 18
+    uint16_t heads;               // 2
+    uint32_t hidden_sectors;
+    uint32_t total_sectors_large;
+} __attribute__((packed)) fat12_bpb_t;
+
+// Directory Entry (32 bytes)
+typedef struct {
+    char     filename[8];
+    char     ext[3];
+    uint8_t  attributes;
+    uint8_t  reserved;
+    uint8_t  creation_ms;
+    uint16_t creation_time;
+    uint16_t creation_date;
+    uint16_t last_access_date;
+    uint16_t high_cluster_num; // Always 0 in FAT12
+    uint16_t last_mod_time;
+    uint16_t last_mod_date;
+    uint16_t low_cluster_num;  // The starting cluster
+    uint32_t file_size;        // Size in bytes
+} __attribute__((packed)) fat12_entry_t;
+
+// --- Kernel File Handle ---
+// This is what your kernel uses to track an open file
+typedef struct {
+    char     name[11];
+    uint32_t size;
+    uint16_t start_cluster;
+    uint16_t current_cluster;
+    uint32_t current_sector_in_cluster;
+    uint32_t bytes_read;
+} file_t;
+
+// --- Public API ---
+
+// You must implement this in your disk driver (e.g., floppy.c)
+// Returns 0 on success, non-zero on error.
+int disk_read_sector(uint32_t lba, uint8_t *buffer);
+
+void fat12_init();
+file_t fat12_open(const char *filename);
+uint32_t fat12_read(file_t *file, uint8_t *buffer, uint32_t bytes_to_read);
+
+#endif // FAT12_H

kernel/framebuffer.c

@@ -0,0 +1,92 @@
+#include "framebuffer.h"
+#include <stddef.h>
+#include <stdint.h>
+
+// Simple init
+void framebuffer_init(framebuffer_t *fb, void *base, uint32_t width, uint32_t height, uint32_t pitch, uint8_t bpp) {
+    fb->base = base;
+    fb->width = width;
+    fb->height = height;
+    fb->pitch = pitch;
+    fb->bpp = bpp;
+    fb->initialized = true;
+}
+
+// Pack color into 32-bit value. Format: 0xAARRGGBB
+uint32_t framebuffer_pack_color(uint8_t r, uint8_t g, uint8_t b, uint8_t a) {
+    return ((uint32_t)a << 24) | ((uint32_t)r << 16) | ((uint32_t)g << 8) | (uint32_t)b;
+}
+
+void framebuffer_put_pixel(framebuffer_t *fb, uint32_t x, uint32_t y, uint8_t r, uint8_t g, uint8_t b, uint8_t a) {
+    if (!fb->initialized) return;
+    if (x >= fb->width || y >= fb->height) return;
+    if (fb->bpp != 32) return; // only 32bpp implemented here
+
+    uint8_t *line = (uint8_t*)fb->base + (size_t)y * fb->pitch;
+    uint32_t *pixel = (uint32_t*)(line + x * 4);
+    *pixel = framebuffer_pack_color(r, g, b, a);
+}
+
+void framebuffer_clear(framebuffer_t *fb, uint8_t r, uint8_t g, uint8_t b) {
+    if (!fb->initialized) return;
+    if (fb->bpp != 32) return;
+
+    uint32_t color = framebuffer_pack_color(r,g,b,0xFF);
+    for (uint32_t y = 0; y < fb->height; ++y) {
+        uint32_t *row = (uint32_t*)((uint8_t*)fb->base + (size_t)y * fb->pitch);
+        for (uint32_t x = 0; x < fb->width; ++x) {
+            row[x] = color;
+        }
+    }
+}
+
+void framebuffer_fill_rect(framebuffer_t *fb, uint32_t x, uint32_t y, uint32_t w, uint32_t h, uint8_t r, uint8_t g, uint8_t b) {
+    if (!fb->initialized) return;
+    if (fb->bpp != 32) return;
+
+    if (x >= fb->width || y >= fb->height) return;
+    if (x + w > fb->width) w = fb->width - x;
+    if (y + h > fb->height) h = fb->height - y;
+
+    uint32_t color = framebuffer_pack_color(r,g,b,0xFF);
+    for (uint32_t yy = 0; yy < h; ++yy) {
+        uint32_t *row = (uint32_t*)((uint8_t*)fb->base + (size_t)(y + yy) * fb->pitch) + x;
+        for (uint32_t xx = 0; xx < w; ++xx) {
+            row[xx] = color;
+        }
+    }
+}
+
+// Simple blit from a source buffer with 32bpp pixels and given pitch (bytes per line)
+void framebuffer_blit(framebuffer_t *fb, uint32_t dst_x, uint32_t dst_y, const void *src, uint32_t src_w, uint32_t src_h, uint32_t src_pitch) {
+    if (!fb->initialized) return;
+    if (fb->bpp != 32) return;
+
+    if (dst_x >= fb->width || dst_y >= fb->height) return;
+
+    uint32_t copy_w = src_w;
+    uint32_t copy_h = src_h;
+    if (dst_x + copy_w > fb->width) copy_w = fb->width - dst_x;
+    if (dst_y + copy_h > fb->height) copy_h = fb->height - dst_y;
+
+    const uint8_t *s = (const uint8_t*)src;
+    for (uint32_t yy = 0; yy < copy_h; ++yy) {
+        uint32_t *dst_row = (uint32_t*)((uint8_t*)fb->base + (size_t)(dst_y + yy) * fb->pitch) + dst_x;
+        const uint32_t *src_row = (const uint32_t*)(s + (size_t)yy * src_pitch);
+        for (uint32_t xx = 0; xx < copy_w; ++xx) {
+            dst_row[xx] = src_row[xx];
+        }
+    }
+}
+
+void framebuffer_test_pattern(framebuffer_t *fb) {
+    if (!fb->initialized) return;
+    // simple color bars
+    uint32_t band_h = fb->height / 6;
+    framebuffer_fill_rect(fb, 0, 0, fb->width, band_h, 0xFF, 0x00, 0x00); // red
+    framebuffer_fill_rect(fb, 0, band_h, fb->width, band_h, 0x00, 0xFF, 0x00); // green
+    framebuffer_fill_rect(fb, 0, band_h*2, fb->width, band_h, 0x00, 0x00, 0xFF); // blue
+    framebuffer_fill_rect(fb, 0, band_h*3, fb->width, band_h, 0xFF, 0xFF, 0x00); // yellow
+    framebuffer_fill_rect(fb, 0, band_h*4, fb->width, band_h, 0xFF, 0x00, 0xFF); // magenta
+    framebuffer_fill_rect(fb, 0, band_h*5, fb->width, fb->height - band_h*5, 0x00, 0xFF, 0xFF); // cyan
+}

kernel/framebuffer.h

@@ -0,0 +1,25 @@
+#ifndef FRAMEBUFFER_H
+#define FRAMEBUFFER_H
+
+#include <stddef.h>
+#include <stdint.h>
+#include <stdbool.h>
+
+typedef struct {
+    void *base;
+    uint32_t width;
+    uint32_t height;
+    uint32_t pitch;
+    uint8_t bpp;
+    bool initialized;
+} framebuffer_t;
+
+void framebuffer_init(framebuffer_t *fb, void *base, uint32_t width, uint32_t height, uint32_t pitch, uint8_t bpp);
+uint32_t framebuffer_pack_color(uint8_t r, uint8_t g, uint8_t b, uint8_t a);
+void framebuffer_put_pixel(framebuffer_t *fb, uint32_t x, uint32_t y, uint8_t r, uint8_t g, uint8_t b, uint8_t a);
+void framebuffer_clear(framebuffer_t *fb, uint8_t r, uint8_t g, uint8_t b);
+void framebuffer_fill_rect(framebuffer_t *fb, uint32_t x, uint32_t y, uint32_t w, uint32_t h, uint8_t r, uint8_t g, uint8_t b);
+void framebuffer_blit(framebuffer_t *fb, uint32_t dst_x, uint32_t dst_y, const void *src, uint32_t src_w, uint32_t src_h, uint32_t src_pitch);
+void framebuffer_test_pattern(framebuffer_t *fb);
+
+#endif /* FRAMEBUFFER_H */

kernel/fs.c

@@ -1,5 +0,0 @@
-#include "fs.h"
-
-void fs_init() {
-    // Filesystem initialization code
-}

kernel/fs.h

@@ -1,6 +0,0 @@
-#ifndef FS_H
-#define FS_H
-
-void fs_init();
-
-#endif // FS_H

kernel/heap.c

@@ -1,6 +1,63 @@
 #include "heap.h"
+#include <stdint.h>
+
+typedef struct heap_block {
+    size_t size;
+    struct heap_block *next;
+    int is_free;
+} heap_block_t;
+
+static heap_block_t *free_list = NULL;
+static void *heap_start_ptr = NULL;
+static void *heap_end_ptr = NULL;
+
+void heap_init(void *heap_start, void *heap_end) {
+    heap_start_ptr = heap_start;
+    heap_end_ptr = heap_end;
+
+    free_list = (heap_block_t *)heap_start;
+    free_list->size = (uintptr_t)heap_end - (uintptr_t)heap_start - sizeof(heap_block_t);
+    free_list->next = NULL;
+    free_list->is_free = 1;
+}
 
 void *heap_alloc(size_t size) {
-    // Heap allocation code
+    heap_block_t *current = free_list;
-    return NULL;
+
+    while (current != NULL) {
+        if (current->is_free && current->size >= size) {
+            // If it's a large block, split it
+            if (current->size > size + sizeof(heap_block_t)) {
+                heap_block_t *new_block = (heap_block_t *)((uintptr_t)current + sizeof(heap_block_t) + size);
+                new_block->size = current->size - size - sizeof(heap_block_t);
+                new_block->next = current->next;
+                new_block->is_free = 1;
+
+                current->next = new_block;
+                current->size = size;
+            }
+
+            current->is_free = 0;
+            return (void *)((uintptr_t)current + sizeof(heap_block_t));
+        }
+
+        current = current->next;
+    }
+
+    return NULL; // Out of memory
+}
+
+void heap_free(void *ptr) {
+    if (ptr == NULL) return;
+
+    heap_block_t *block = (heap_block_t *)((uintptr_t)ptr - sizeof(heap_block_t));
+    block->is_free = 1;
+
+    // Coalesce with next block
+    if (block->next && block->next->is_free) {
+        block->size += block->next->size + sizeof(heap_block_t);
+        block->next = block->next->next;
+    }
+
+    // TODO: Coalesce with previous block for better compaction
 }

kernel/heap.h

@@ -3,6 +3,9 @@
 
 #include <stddef.h>
 
+void heap_init(void *heap_start, void *heap_end);
+
 void *heap_alloc(size_t size);
+void heap_free(void *ptr);
 
 #endif // HEAP_H

kernel/idt.c

@@ -52,7 +52,7 @@ void idt_set_gate(int n, uint32_t handler) {
 
 // Load IDT via lidt
 static void idt_load() {
-    asm volatile("lidt (%0)" : : "r" (&idt_ptr));
+    __asm__("lidt (%0)" : : "r" (&idt_ptr));
 }
 
 // IDT initialization

kernel/io.h

@@ -1,13 +1,35 @@
 #ifndef IO_H
 #define IO_H
 
+#include <stdint.h>
+
 static inline void outb(uint16_t port, uint8_t val) {
-    asm volatile ("outb %0, %1" : : "a"(val), "Nd"(port));
+    __asm__("outb %0, %1" : : "a"(val), "Nd"(port));
 }
 
 static inline uint8_t inb(uint16_t port) {
     uint8_t ret;
-    asm volatile ("inb %1, %0" : "=a"(ret) : "Nd"(port));
+    __asm__("inb %1, %0" : "=a"(ret) : "Nd"(port));
+    return ret;
+}
+
+static inline void outw(uint16_t port, uint16_t val) {
+    __asm__("outw %0, %1" : : "a"(val), "Nd"(port));
+}
+
+static inline uint16_t inw(uint16_t port) {
+    uint16_t ret;
+    __asm__("inw %1, %0" : "=a"(ret) : "Nd"(port));
+    return ret;
+}
+
+static inline void outl(uint16_t port, uint32_t val) {
+    __asm__("outl %0, %1" : : "a"(val), "Nd"(port));
+}
+
+static inline uint32_t inl(uint16_t port) {
+    uint32_t ret;
+    __asm__("inl %1, %0" : "=a"(ret) : "Nd"(port));
     return ret;
 }
 

kernel/irq.c

@@ -1,5 +1,76 @@
+#include "idt.h"
 #include "irq.h"
+#include "io.h"
+#include "isr.h"
 
-void irq_init() {
+#define PIC1_CMD   0x20
-    // IRQ initialization code
+#define PIC1_DATA  0x21
+#define PIC2_CMD   0xA0
+#define PIC2_DATA  0xA1
+
+// FIXME: stubs
+void irq0() {}
+void irq1() {}
+void irq2() {}
+void irq3() {}
+void irq4() {}
+void irq5() {}
+void irq6() {}
+void irq7() {}
+void irq8() {}
+void irq9() {}
+void irq10() {}
+void irq11() {}
+void irq12() {}
+void irq13() {}
+void irq14() {}
+void irq15() {}
+// --- stubs end
+
+void irq_remap(void)
+{
+    outb(PIC1_CMD, 0x11);   // ICW1 – edge triggered, cascade, need ICW4
+    outb(PIC2_CMD, 0x11);
+
+    outb(PIC1_DATA, 0x20);  // ICW2 – master base vector
+    outb(PIC2_DATA, 0x28);  // ICW2 – slave base vector
+
+    outb(PIC1_DATA, 0x04);  // ICW3 – slave on IRQ2
+    outb(PIC2_DATA, 0x02);  // ICW3 – cascade identity
+
+    outb(PIC1_DATA, 0x01);  // ICW4 – 8086 mode
+    outb(PIC2_DATA, 0x01);
+
+    // Mask everything except IRQ0 (timer) and IRQ1 (keyboard) for now
+    outb(PIC1_DATA, 0b11111001);
+    outb(PIC2_DATA, 0xFF);
+}
+
+void irq_install(void)
+{
+    irq_remap();
+
+    /* Fill IRQ entries in the IDT (0x20 … 0x2F) */
+    //extern void irq0(), irq1(), irq2(), irq3(), irq4(), irq5(), irq6(), irq7();
+    //extern void irq8(), irq9(), irq10(), irq11(), irq12(), irq13(), irq14(), irq15();
+
+    idt_set_gate(0x20, (uint32_t)irq0);
+    idt_set_gate(0x21, (uint32_t)irq1);
+    /* … repeat for the rest or loop … */
+    for (int i = 2; i < 16; ++i)
+        idt_set_gate(0x20 + i, (uint32_t)irq0 + i * 8); // crude but works
+}
+
+/* Called from the assembly stubs (see irq.asm below) */
+void irq_handler(uint32_t int_num)
+{
+    /* int_num is the *remapped* vector, e.g. 0x21 for keyboard */
+    if (interrupt_handlers[int_num]) {
+        interrupt_handlers[int_num]();
+    }
+
+    /* ---- EOI ---- */
+    if (int_num >= 0x28)          // slave PIC
+        outb(PIC2_CMD, 0x20);
+    outb(PIC1_CMD, 0x20);         // always master
 }

kernel/irq.h

@@ -1,6 +1,10 @@
 #ifndef IRQ_H
 #define IRQ_H
 
-void irq_init();
+#include <stdint.h>
 
-#endif // IRQ_H
+void irq_remap(void);
+void irq_install(void);
+void irq_handler(uint32_t int_num);
+
+#endif

kernel/isr.asm

@@ -1,8 +1,8 @@
 [BITS 32]
-[GLOBAL isr0, isr1, isr2, isr3, isr4, isr5, isr6, isr7, isr8, isr9]
+GLOBAL isr0, isr1, isr2, isr3, isr4, isr5, isr6, isr7, isr8, isr9
-[GLOBAL isr10, isr11, isr12, isr13, isr14, isr15, isr16, isr17, isr18, isr19]
+GLOBAL isr10, isr11, isr12, isr13, isr14, isr15, isr16, isr17, isr18, isr19
-[GLOBAL isr20, isr21, isr22, isr23, isr24, isr25, isr26, isr27, isr28, isr29]
+GLOBAL isr20, isr21, isr22, isr23, isr24, isr25, isr26, isr27, isr28, isr29
-[GLOBAL isr30, isr31, isr_default]
+GLOBAL isr30, isr31, isr_default
 
 [EXTERN isr_handler]
 

kernel/isr.c

@@ -1,14 +1,24 @@
+#include <stdbool.h>
 #include "terminal.h"
 #include "serial.h"
 #include "isr.h"
+#include "io.h"
+#include "print.h"
 
-static isr_callback_t interrupt_handlers[MAX_INTERRUPTS] = { 0 };
+isr_callback_t interrupt_handlers[MAX_INTERRUPTS] = { 0 };
 
 void isr_handler(uint32_t int_num, uint32_t err_code) {
     terminal_write("Interrupt occurred: ");
-    // Here you can add a basic itoa to print int_num
+
+    print_hex(int_num, true, false);
+    terminal_write("\n");
+
     serial_write("INT triggered\n");
 
+    terminal_write("Error code: ");
+    print_hex(err_code, true, false);
+    terminal_write("\n");
+
     if (interrupt_handlers[int_num]) {
         interrupt_handlers[int_num]();  // Call registered handler
     } else {
@@ -24,9 +34,19 @@ void isr_handler(uint32_t int_num, uint32_t err_code) {
 
         // Halt CPU
         while (1) {
-            asm volatile ("hlt");
+            __asm__("hlt");
         }
     }
+
+    // === Send End Of Interrupt to PIC(s) ===
+    if (int_num >= 40) {
+        // Send reset signal to slave PIC
+        outb(0xA0, 0x20);
+    }
+    if (int_num >= 32) {
+        // Send reset signal to master PIC
+        outb(0x20, 0x20);
+    }
 }
 
 void register_interrupt_handler(uint8_t n, isr_callback_t handler) {

kernel/isr.h

@@ -6,6 +6,7 @@
 #define MAX_INTERRUPTS 256
 
 typedef void (*isr_callback_t)(void);
+extern isr_callback_t interrupt_handlers[MAX_INTERRUPTS];
 
 void isr_handler(uint32_t int_num, uint32_t err_code);
 void register_interrupt_handler(uint8_t n, isr_callback_t handler);

kernel/keyboard.c

@@ -4,8 +4,12 @@
 #include "terminal.h"
 
 #define KEYBOARD_DATA_PORT 0x60
+#define KEY_BUFFER_SIZE 256
 
-static char key_buffer[256];
+static char key_buffer[KEY_BUFFER_SIZE];
+static uint8_t buffer_head = 0;  // Write position (interrupt)
+static uint8_t buffer_tail = 0;  // Read position (get_char)
+static uint8_t buffer_count = 0;
 static uint8_t buffer_index = 0;
 
 // Basic US QWERTY keymap (scancode to ASCII)
@@ -20,18 +24,24 @@ static const char scancode_map[128] = {
 };
 
 // Interrupt handler for IRQ1
-void keyboard_callback() {
+void keyboard_callback(void) {
     uint8_t scancode = inb(KEYBOARD_DATA_PORT);
 
-    // Ignore key releases (scancodes with high bit set)
+    if (scancode & 0x80) return;  // Ignore key release
-    if (scancode & 0x80)
-        return;
 
     char c = scancode_map[scancode];
-    if (c && buffer_index < sizeof(key_buffer)) {
+    if (!c) return;
-        key_buffer[buffer_index++] = c;
+
-        terminal_putchar(c); // Echo to terminal (optional)
+    uint8_t next_head = (buffer_head + 1) % KEY_BUFFER_SIZE;
-    }
+
+    // Drop key if buffer full
+    if (next_head == buffer_tail) return;
+
+    key_buffer[buffer_head] = c;
+    buffer_head = next_head;
+    buffer_count++;
+
+    terminal_putchar(c);
 }
 
 void keyboard_init() {
@@ -39,15 +49,17 @@ void keyboard_init() {
 }
 
 // Blocking read (returns one char)
-char keyboard_get_char() {
+char keyboard_get_char(void) {
-    while (buffer_index == 0);
+    while (buffer_count == 0) {
-    char c = key_buffer[0];
+        __asm__ __volatile__("hlt");  // Better than busy loop
-
-    // Shift buffer left
-    for (uint8_t i = 1; i < buffer_index; i++) {
-        key_buffer[i - 1] = key_buffer[i];
     }
-    buffer_index--;
+
+    char c;
+    __asm__ __volatile__("cli");
+    c = key_buffer[buffer_tail];
+    buffer_tail = (buffer_tail + 1) % KEY_BUFFER_SIZE;
+    buffer_count--;
+    __asm__ __volatile__("sti");
 
     return c;
 }

kernel/kmain.c

@@ -1,5 +1,4 @@
 #include <stdint.h>
-#include <stdbool.h>
 #include "io.h"
 #include "serial.h"
 #include "terminal.h"
@@ -7,6 +6,13 @@
 #include "paging.h"
 #include "memmap.h"
 #include "gdt.h"
+#include "cpu.h"
+#include "kmalloc.h"
+#include "print.h"
+#include "timer.h"
+#include "utils.h"
+#include "keyboard.h"
+#include "irq.h"
 
 #define LPT1 0x378
 
@@ -22,6 +28,10 @@ void kmain(void) {
     serial_init();
     serial_write("Serial port initialized.\n");
 
+    terminal_write("Identifying CPU...\n");
+    identify_cpu();
+    serial_write("CPU identification complete.\n");
+
     lpt_write('L'); // Send 'L' to LPT1 to test
 
     terminal_write("Initializing GDT...\n");
@@ -32,6 +42,9 @@ void kmain(void) {
     idt_init();
     serial_write("IDT initialized.\n");
 
+    irq_install();
+    __asm__ __volatile__ ("sti");
+
     terminal_write("Enabling paging...\n");
     paging_init();
     serial_write("Paging initialized.\n");
@@ -40,30 +53,39 @@ void kmain(void) {
     kmalloc_init(0xC0100000);  // Virtual heap start address (must be mapped!)
     serial_write("kmalloc initialized.\n");
 
-    void* ptr = kmalloc(128);  // Allocation test
     serial_write("Allocated 128 bytes.\n");
 
     terminal_write("Initializing timer...\n");
     timer_init(100);  // 100 Hz (10 ms interval)
     serial_write("Timer initialized.\n");
 
+    terminal_write("Initializing keyboard...\n");
+    keyboard_init();
+    serial_write("Keyboard initialized.\n");
+
     terminal_write("Getting memory map...\n");
     memory_map_entry_t mmap[32];
     uint32_t mmap_size = get_memory_map(mmap, 32);
     serial_write("Memory map retrieved.\n");
 
     terminal_write("Memory Regions:\n");
+
+    char buf[32];
     for (uint32_t i = 0; i < mmap_size; i++) {
-        terminal_write(" - Region: ");
+        terminal_write(" - Base: ");
-        // You would format and print base/length/type here
+        print_hex((uint32_t)(mmap[i].base_addr & 0xFFFFFFFF), true, false);  // Lower 32 bits
-        // (e.g., with a basic itoa and print_hex helper)
+        terminal_write(", Length: ");
-        serial_write("Memory region entry\n");
+        print_hex((uint32_t)(mmap[i].length & 0xFFFFFFFF), true, false);     // Lower 32 bits
+        terminal_write(", Type: ");
+        itoa(mmap[i].type, buf, 10);
+        terminal_write(buf);
+        terminal_write("\n");
     }
 
     terminal_write("System initialized. Halting.\n");
 
     // Halt CPU in loop
     while (1) {
-        asm volatile("hlt");
+        __asm__("hlt");
     }
 }

kernel/kmalloc.c

@@ -1,6 +1,8 @@
 #include "kmalloc.h"
 #include "terminal.h"  // Optional: for debug output
 
+#define HEAP_END 0xC0500000
+
 static uint32_t current_heap = 0;
 
 // Initialize the allocator with a starting heap address
@@ -32,7 +34,18 @@ void* kmalloc_aligned(size_t size, uint32_t alignment) {
         current_heap = (current_heap + alignment) & ~(alignment - 1);
     }
 
+    if (current_heap + size > HEAP_END) {
+        terminal_write("kmalloc_aligned: Out of memory!\n");
+        return 0;
+    }
+
     void* addr = (void*)current_heap;
     current_heap += size;
     return addr;
 }
+
+void kfree(void* ptr) {
+    // In a bump allocator, we cannot free individual blocks.
+    // We can reset the allocator to the initial state.
+    current_heap = 0; // Reset the heap pointer
+}

kernel/linker.ld

@@ -1,31 +1,38 @@
-ENTRY(_start)
+ENTRY(kmain)
+
+PHDRS {
+    text PT_LOAD FLAGS(5);    /* Read + Execute */
+    rodata PT_LOAD FLAGS(4);  /* Read only */
+    data PT_LOAD FLAGS(6);    /* Read + Write */
+}
 
 SECTIONS {
     . = 1M;
 
-    .multiboot : {
-        *(.multiboot)
-    }
-
     .text : {
-        *(.text)
+        *(.text*)
-    }
+    } :text
 
     .rodata : {
-        *(.rodata)
+        *(.rodata*)
-    }
+    } :rodata
 
     .data : {
-        *(.data)
+        *(.data*)
-    }
+    } :data
 
     .bss : {
-        *(.bss)
+        *(.bss*)
         *(COMMON)
+    } :data
+
+    .stack (NOLOAD) : {
+        . = ALIGN(4);
+        . = . + 0x1000;
     }
 
-    . = ALIGN(4096);
+    .heap (NOLOAD) : {
-    __stack_top = .;
+        . = ALIGN(4);
-    . += 128K;
+        . = . + 0x10000;
-    __stack_bottom = .;
+    }
 }

kernel/malloc.c

@@ -0,0 +1,105 @@
+#include "malloc.h"
+#include <stdint.h>
+
+static void *heap_start; // Start of the heap
+static void *heap_end;   // End of the heap
+static struct memory_block *free_blocks; // List of free blocks
+
+void init_heap(void *start, void *end)
+{
+    heap_start = start;
+    heap_end = end;
+
+    // Initialize the heap with a single large free block
+    free_blocks = (struct memory_block *)start;
+    free_blocks->size = (uintptr_t)end - (uintptr_t)start - sizeof(struct memory_block);
+    free_blocks->next = NULL;
+    free_blocks->is_free = 1;
+}
+
+void *find_free_block(size_t size) {
+    struct memory_block *current = free_blocks;
+    while (current != NULL) {
+        if (current->is_free && current->size >= size) {
+            return current;
+        }
+        current = current->next;
+    }
+    // No suitable block found
+    return NULL;
+}
+
+void mark_as_used(void *ptr, size_t size) {
+    struct memory_block *block = (struct memory_block *)ptr;
+    block->is_free = 0;
+
+    // If the block is larger than needed, split it
+    if (block->size > size + sizeof(struct memory_block)) {
+        struct memory_block *new_block = (struct memory_block *)((uintptr_t)ptr + size + sizeof(struct memory_block));
+        new_block->size = block->size - size - sizeof(struct memory_block);
+        new_block->next = block->next;
+        new_block->is_free = 1;
+
+        block->size = size;
+        block->next = new_block;
+    }
+}
+
+void mark_as_free(void *ptr) {
+    struct memory_block *block = (struct memory_block *)ptr;
+    block->is_free = 1;
+
+    // Coalesce with next block if it's free
+    if (block->next && block->next->is_free) {
+        block->size += block->next->size + sizeof(struct memory_block);
+        block->next = block->next->next;
+    }
+
+    // Coalesce with previous block if it's free
+    struct memory_block *prev = free_blocks;
+    while (prev && prev->next != block) {
+        prev = prev->next;
+    }
+    if (prev && prev->is_free) {
+        prev->size += block->size + sizeof(struct memory_block);
+        prev->next = block->next;
+    }
+}
+
+void *malloc(size_t size)
+{
+    if (heap_start == NULL || heap_end == NULL)
+    {
+        // Heap not initialized, cannot allocate
+        return NULL;
+    }
+
+    // Align the size to the word size for efficiency
+    size = (size + sizeof(size_t) - 1) & ~(sizeof(size_t) - 1);
+
+    // Search for a free block of sufficient size
+    void *block = find_free_block(size);
+    if (block != NULL)
+    {
+        // Mark the block as used
+        mark_as_used(block, size);
+        return (void *)((uintptr_t)block + sizeof(struct memory_block));
+    }
+
+    // No suitable block found, out of memory
+    return NULL;
+}
+
+void free(void *ptr)
+{
+    if (ptr == NULL)
+    {
+        return;
+    }
+
+    // Adjust pointer to the start of the memory block
+    struct memory_block *block = (struct memory_block *)((uintptr_t)ptr - sizeof(struct memory_block));
+
+    // Mark the block as free
+    mark_as_free(block);
+}

kernel/malloc.h

@@ -0,0 +1,27 @@
+#ifndef MALLOC_H
+#define MALLOC_H
+
+#include <stddef.h> // For size_t
+
+// Define the memory block structure
+struct memory_block {
+    size_t size;
+    struct memory_block *next;
+    int is_free;
+};
+
+// Function prototypes
+void init_heap(void *start, void *end);
+void *malloc(size_t size);
+void free(void *ptr);
+
+// Helper function prototypes
+void *find_free_block(size_t size);
+void mark_as_used(void *ptr, size_t size);
+void mark_as_free(void *ptr);
+
+// External heap boundaries
+extern void *user_heap_start;
+extern void *user_heap_end;
+
+#endif // MALLOC_H

kernel/memmap.c

@@ -1,14 +1,21 @@
 #include "memmap.h"
 
 uint32_t get_memory_map(memory_map_entry_t *map, uint32_t max_entries) {
-    // Fill with dummy values for now
+    uint32_t count = 0;
-    map[0].base_addr = 0x00000000;
-    map[0].length = 0x0009FC00;
-    map[0].type = 1;
 
-    map[1].base_addr = 0x00100000;
+    if (max_entries >= 1) {
-    map[1].length = 0x1FF00000;
+        map[count].base_addr = 0x00000000;
-    map[1].type = 1;
+        map[count].length = 0x0009FC00;
+        map[count].type = 1;
+        count++;
+    }
 
-    return 2; // 2 regions
+    if (max_entries >= 2) {
+        map[count].base_addr = 0x00100000;
+        map[count].length = 0x1FF00000;
+        map[count].type = 1;
+        count++;
+    }
+
+    return count;
 }

kernel/memory.c

@@ -0,0 +1,17 @@
+#include "memory.h"
+
+/* note: this is a stub, please use care as theres duplicate functions in utils implementation
+/* --------------------------------------------------------------------- *
+ *  Helper: copy a single byte (used by both memcpy and memmove)
+ * --------------------------------------------------------------------- */
+static inline void byte_copy_forward(uint8_t *dst, const uint8_t *src, size_t n)
+{
+    while (n--) *dst++ = *src++;
+}
+
+static inline void byte_copy_backward(uint8_t *dst, const uint8_t *src, size_t n)
+{
+    dst += n; src += n;
+    while (n--) *--dst = *--src;
+}
+

kernel/memory.h

@@ -0,0 +1,25 @@
+#ifndef MEMORY_H
+#define MEMORY_H
+
+#include <stddef.h>   /* size_t, NULL */
+#include <stdint.h>   /* uint8_t */
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+/* C11 / POSIX-2004 signatures */
+void *memcpy(void *restrict dst, const void *restrict src, size_t n);
+void *memmove(void *dst, const void *src, size_t n);
+int   memcmp(const void *s1, const void *s2, size_t n);
+
+/* Optional fast-path using 32-bit loads (x86 only) */
+#if defined(__i386__) && !defined(MEMORY_NO_OPT)
+# define MEMORY_OPTIMIZED 1
+#endif
+
+#ifdef __cplusplus
+}
+#endif
+
+#endif /* MEMORY_H */

kernel/mouse.c

@@ -0,0 +1,27 @@
+// mouse.c
+#include "mouse.h"
+#include "usb.h"
+#include <stdint.h>
+#include <stdbool.h>
+
+// Mouse buffer
+static mouse_data_t mouse_data;
+
+// Read USB mouse data
+mouse_data_t usb_read_mouse(void) {
+    uint8_t buffer[3]; // USB HID Mouse reports typically use 3 bytes
+    if (usb_interrupt_transfer()) { // Ensure buffer is filled
+        // Process the received data
+        mouse_data.x += buffer[1]; // X movement
+        mouse_data.y += buffer[2]; // Y movement
+        mouse_data.left_button = buffer[0] & 0x01;
+        mouse_data.right_button = buffer[0] & 0x02;
+    } else {
+        // Handle the case where no data was received
+        // You can choose to reset mouse_data or leave it unchanged
+        // For example, you might want to set movement to zero if no data is received
+        // mouse_data.x = 0;
+        // mouse_data.y = 0;
+    }
+    return mouse_data;
+}

kernel/mouse.h

@@ -0,0 +1,26 @@
+#ifndef MOUSE_H
+#define MOUSE_H
+
+#include <stdint.h>
+#include <stdbool.h>
+
+// Mouse data structure
+typedef struct {
+    int x;
+    int y;
+    bool left_button;
+    bool right_button;
+} mouse_data_t;
+
+// Function declarations for USB 1.x HID mouse support
+bool usb_mouse_init(void);
+bool usb_mouse_detected(void);
+bool usb_mouse_received(void);
+mouse_data_t usb_read_mouse(void);
+
+// USB Host Controller Interface functions (USB 1.x support)
+bool uhci_init(void);
+bool ohci_init(void);
+
+#endif // MOUSE_H
+

kernel/paging.c

@@ -1,21 +1,17 @@
-#include "paging.h"
-#include "io.h"
 #include <stdint.h>
-#include <stddef.h>
+#include <string.h>
+#include "io.h"
+#include "paging.h"
 
-page_directory_entry_t *page_directory = (page_directory_entry_t *)0x100000;
+page_directory_entry_t *page_directory = (page_directory_entry_t *)0x200000;
-page_table_entry_t *page_table = (page_table_entry_t *)0x101000;
+page_table_entry_t     *page_table     = (page_table_entry_t *)0x201000;
-page_table_entry_t *heap_page_table = (page_table_entry_t *)0x102000; // Located right after the page directory
 
 // Helper function to set up the page directory entry
 void set_page_directory(page_directory_entry_t *dir) {
-    for (int i = 0; i < PAGE_DIRECTORY_SIZE; i++) {
+    // Set first PDE
-        dir[i].present = 0;
-    }
     dir[0].present = 1;
     dir[0].rw = 1;
-    dir[0].user = 0;
+    dir[0].addr = (uint32_t)page_table >> 12;
-    dir[0].frame = (uint32_t)page_table >> 12;
 }
 
 // Helper function to set up the page table entry
@@ -23,12 +19,8 @@ void set_page_table(page_table_entry_t *table) {
     for (int i = 0; i < PAGE_TABLE_SIZE; i++) {
         // Set up page table entries with identity mapping
         table[i].present = 1;
-        table[i].rw = 1;          // Read/Write
+        table[i].rw = 1;        // Read/Write
-        table[i].user = 0;        // Kernel mode
+        table[i].addr = i;      // Identity mapping
-        table[i].write_through = 0;
-        table[i].cache_disabled = 0;
-        table[i].accessed = 0;
-        table[i].frame = i;       // Identity mapping
     }
 }
 
@@ -37,36 +29,23 @@ void enable_paging() {
     uint32_t cr0;
 
     // Load page directory into CR3
-    asm volatile("mov %0, %%cr3" : : "r"(page_directory));
+    __asm__("mov %0, %%cr3" : : "r"(page_directory));
 
     // Enable paging (set the PG bit in CR0)
-    asm volatile("mov %%cr0, %0" : "=r"(cr0));
+    __asm__("mov %%cr0, %0" : "=r"(cr0));
     cr0 |= 0x80000000;  // Set the PG (paging) bit
-    asm volatile("mov %0, %%cr0" : : "r"(cr0));
+    __asm__("mov %0, %%cr0" : : "r"(cr0));
 }
 
 // Initialize paging: set up the page directory and enable paging
 void paging_init() {
+    // Zero out the tables
+    memset(page_directory, 0x00, PAGE_DIRECTORY_SIZE * sizeof *page_directory);
+    memset(page_table, 0x00, PAGE_TABLE_SIZE * sizeof *page_table);
+
     // Set up identity-mapped page directory + table
     set_page_directory(page_directory);
     set_page_table(page_table);
 
-    // === Set up heap mapping at 0xC0100000 ===
-    for (int i = 0; i < PAGE_TABLE_SIZE; i++) {
-        heap_page_table[i].present = 1;
-        heap_page_table[i].rw = 1;
-        heap_page_table[i].user = 0;
-        heap_page_table[i].write_through = 0;
-        heap_page_table[i].cache_disabled = 0;
-        heap_page_table[i].accessed = 0;
-        heap_page_table[i].frame = (256 + i); // Start physical heap at 1MB (256*4KB = 1MB)
-    }
-
-    // Index 772 = 0xC0100000 / 4MB
-    page_directory[772].present = 1;
-    page_directory[772].rw = 1;
-    page_directory[772].user = 0;
-    page_directory[772].frame = (uint32_t)heap_page_table >> 12;
-
     enable_paging();
 }

kernel/paging.h

@@ -10,31 +10,31 @@
 
 // Page Directory and Page Table structure
 typedef struct {
-    uint32_t present : 1;  // Present bit (1: page is present in memory)
+    uint32_t present : 1;           // Present bit (1: page is present in memory)
-    uint32_t rw : 1;       // Read-Write bit (1: page is read-write)
+    uint32_t rw : 1;                // Read-Write bit (1: page is read-write)
-    uint32_t user : 1;     // User-supervisor bit (1: user mode access)
+    uint32_t user : 1;              // User-supervisor bit (1: user mode access)
-    uint32_t write_through : 1;  // Write-through cache
+    uint32_t write_through : 1;     // Write-through cache
-    uint32_t cache_disabled : 1; // Cache disabled
+    uint32_t cache_disabled : 1;    // Cache disabled
-    uint32_t accessed : 1;        // Accessed bit
+    uint32_t accessed : 1;          // Accessed bit
-    uint32_t reserved : 1;        // Reserved bit
+    uint32_t dirty : 1;             // Dirty bit
-    uint32_t page_size : 1;       // Page size (0: 4KB, 1: 4MB)
+    uint32_t attribute : 1;         // Page size (0: 4KB, 1: 4MB)
-    uint32_t global : 1;           // Global page (can be used across different processes)
+    uint32_t global : 1;            // Global page (can be used across different processes)
-    uint32_t available : 3;        // Available bits for the system
+    uint32_t reserved : 3;          // Unused
-    uint32_t frame : 20;           // Frame address (physical address)
+    uint32_t addr : 20;             // Page frame address (physical address)
 } __attribute__((packed)) page_table_entry_t;
 
 // Define page directory entry
 typedef struct {
-    uint32_t present : 1;
+    uint32_t present : 1;           // Present bit (1: PTE is present in memory)
-    uint32_t rw : 1;
+    uint32_t rw : 1;                // Read-Write bit (1: pages are read-write)
-    uint32_t user : 1;
+    uint32_t user : 1;              // User-supervisor bit (1: user mode access)
-    uint32_t write_through : 1;
+    uint32_t write_through : 1;     // Write-through cache
-    uint32_t cache_disabled : 1;
+    uint32_t cache_disabled : 1;    // Cache disabled
-    uint32_t accessed : 1;
+    uint32_t accessed : 1;          // Accessed bit
-    uint32_t reserved : 1;
+    uint32_t available : 1;         // Unused
-    uint32_t zero : 5;           // Must be zero for page directory
+    uint32_t page_size : 1;         // Page size (0: 4KB, 1: 4MB)
-    uint32_t reserved_2 : 7;     // Reserved bits
+    uint32_t available_2 : 4;       // Unused
-    uint32_t frame : 20;         // Frame address of the page table
+    uint32_t addr : 20;             // Page table address
 } __attribute__((packed)) page_directory_entry_t;
 
 extern page_directory_entry_t *page_directory;

kernel/panic.c

@@ -1,5 +1,19 @@
 #include "panic.h"
+#include "terminal.h"
+#include "serial.h"
+#include <stdbool.h>
 
 void panic(const char *message) {
-    // Panic handling code
+    terminal_write("KERNEL PANIC: ");
+    terminal_write(message);
+    terminal_write("\nSystem halted.\n");
+
+    serial_write("KERNEL PANIC: ");
+    serial_write(message);
+    serial_write("\nSystem halted.\n");
+
+    // Halt the system
+    while (true) {
+        __asm__("cli; hlt");
+    }
 }

kernel/print.c

@@ -0,0 +1,102 @@
+#include <stdarg.h>
+#include "print.h"
+#include "serial.h"
+#include "terminal.h"
+
+void my_putchar(char ch) {
+    // Write a single character to standard output
+    // In a freestanding environment, you might need to implement this differently
+    // For now, we will use the standard putchar for demonstration
+    // Replace this with your own implementation if needed
+    terminal_putchar(ch);
+}
+
+void print_string(const char *str) {
+    // Simple implementation to print a string
+    while (*str) {
+        my_putchar(*str++);
+    }
+}
+
+void my_printf(const char *format, ...) {
+    va_list args;
+    va_start(args, format);
+
+    while (*format) {
+        if (*format == '%') {
+            format++; // Move to the next character after '%'
+            switch (*format) {
+                case 's': { // String
+                    char *str = va_arg(args, char *);
+                    print_string(str);
+                    break;
+                }
+                case 'd': { // Integer
+                    int num = va_arg(args, int);
+                    char buffer[20]; // Buffer to hold the string representation
+                    
+                    //TODO: implement `snprintf()`
+                    //snprintf(buffer, sizeof(buffer), "%d", num);
+                    print_string(buffer);
+                    break;
+                }
+                case 'c': { // Character
+                    char ch = (char)va_arg(args, int); // Promote char to int
+                    my_putchar(ch);
+                    break;
+                }
+                default:
+                    my_putchar('%'); // Print the '%' if no valid format specifier
+                    my_putchar(*format);
+                    break;
+            }
+        } else {
+            my_putchar(*format);
+        }
+        format++;
+    }
+
+    va_end(args);
+}
+
+void print_hex(uint32_t val, int include_prefix, int suppress_leading_zeros) { 
+    char hex_chars[] = "0123456789ABCDEF";
+    char buffer[11]; // 8 hex digits + "0x" + null terminator
+    int pos = 10;    // Start from end of buffer (null terminator)
+
+    // Null-terminate the buffer
+    buffer[pos--] = '\0';
+
+    // Convert value to hex digits
+    for (int i = 7; i >= 0; i--) {
+        int digit = val & 0xF; // Get last 4 bits
+        buffer[pos--] = hex_chars[digit];
+        val >>= 4; // Shift right by 4 bits
+    }
+
+    // Add "0x" prefix if requested
+    if (include_prefix) {
+        buffer[pos--] = 'x';
+        buffer[pos--] = '0';
+    }
+
+    // Determine start of output (skip leading zeros if requested)
+    int start = include_prefix ? 0 : 2; // Start after "0x" if prefix included
+    if (suppress_leading_zeros && !include_prefix) {
+        int i = start;
+        while (i < 9 && buffer[i] == '0') {
+            i++;
+        }
+        if (i == 10) {
+            // All zeros, output single '0'
+            terminal_write("0");
+            serial_write("0");
+            return;
+        }
+        start = i;
+    }
+
+    // Output the result
+    terminal_write(buffer + start);
+    serial_write(buffer + start);
+}

kernel/print.h

@@ -0,0 +1,11 @@
+#ifndef PRINT_H
+#define PRINT_H
+
+#include <stdint.h>
+
+void print_string(const char *str);
+void my_printf(const char *format, ...);
+void print_hex(uint32_t val, int include_prefix, int suppress_leading_zeros);
+void my_putchar(char ch);
+
+#endif

kernel/scheduler.c

@@ -1,5 +1,96 @@
 #include "scheduler.h"
+#include <stddef.h>
+
+// Defined in context_switch.s
+extern void ctx_switch(uint32_t **old_sp_ptr, uint32_t *new_sp);
+
+static task_t tasks[MAX_TASKS];
+
+// Stack memory area. Note: x86 Stacks grow DOWN from high to low addresses.
+static uint32_t task_stacks[MAX_TASKS][STACK_SIZE / sizeof(uint32_t)];
+
+static int task_count = 0;
+static task_t *task_list = NULL;
+static task_t *current_task = NULL;
 
 void scheduler_init() {
-    // Scheduler initialization code
+    task_list = NULL;
+    current_task = NULL;
+    task_count = 0;
+}
+
+void scheduler_add_task(void (*entry)(void)) {
+    if (task_count >= MAX_TASKS || entry == NULL) return;
+
+    task_t *new_task = &tasks[task_count];
+    new_task->id = task_count;
+
+    // 1. Calculate the top of the stack (High Address)
+    // We point to the very end of the array.
+    uint32_t *sp = &task_stacks[task_count][STACK_SIZE / sizeof(uint32_t)];
+
+    // 2. "Forge" the stack frame to look like ctx_switch saved it.
+    // We push values onto the stack by decrementing the pointer and writing.
+
+    // --- Return Address (EIP) ---
+    sp--; 
+    *sp = (uint32_t)entry; // When ctx_switch does 'ret', it pops this and jumps to 'entry'
+
+    // --- EFLAGS ---
+    sp--;
+    *sp = 0x00000202; // Reserved bit set, Interrupts Enabled (IF=1). Important!
+
+    // --- General Purpose Registers (PUSHA/POPA layout) ---
+    // Order: EAX, ECX, EDX, EBX, ESP, EBP, ESI, EDI
+    // We initialize them to 0 or meaningful values.
+    sp--; *sp = 0; // EAX
+    sp--; *sp = 0; // ECX
+    sp--; *sp = 0; // EDX
+    sp--; *sp = 0; // EBX
+    sp--; *sp = 0; // ESP (Ignored by POPA)
+    sp--; *sp = 0; // EBP
+    sp--; *sp = 0; // ESI
+    sp--; *sp = 0; // EDI
+
+    // Save this final stack location to the TCB
+    new_task->stack_ptr = sp;
+    new_task->next = NULL;
+
+    // 3. Add to linked list
+    if (task_list == NULL) {
+        task_list = new_task;
+        current_task = new_task; // Make sure we have a current task to start
+    } else {
+        task_t *tail = task_list;
+        while (tail->next) {
+            tail = tail->next;
+        }
+        tail->next = new_task;
+    }
+
+    task_count++;
+}
+
+void scheduler_schedule() {
+    if (!current_task) return;
+
+    task_t *prev = current_task;
+    
+    // Round-robin logic
+    if (current_task->next) {
+        current_task = current_task->next;
+    } else {
+        current_task = task_list;
+    }
+
+    // Perform the ACTUAL context switch
+    // We pass the address of the previous task's stack pointer storage
+    // and the value of the new task's stack pointer.
+    if (prev != current_task) {
+        ctx_switch(&prev->stack_ptr, current_task->stack_ptr);
+    }
+}
+
+void scheduler_yield() {
+    scheduler_schedule();
 }

kernel/scheduler.h

@@ -1,6 +1,24 @@
 #ifndef SCHEDULER_H
 #define SCHEDULER_H
 
+#include <stdint.h>
+
+#define MAX_TASKS 8
+#define STACK_SIZE 1024 // in bytes
+
+typedef struct task {
+    uint32_t id;
+    
+    // The most important field: 
+    // Where was the stack pointer when we last left this task?
+    uint32_t *stack_ptr; 
+    
+    struct task *next;
+} task_t;
+
 void scheduler_init();
+void scheduler_add_task(void (*entry)(void));
+void scheduler_schedule();
+void scheduler_yield();
 
 #endif // SCHEDULER_H

kernel/serial.c

@@ -1,5 +1,6 @@
 #include "io.h"
 #include "serial.h"
+#include <stdint.h>
 
 #define COM1 0x3F8
 #define COM2 0x2F8

kernel/shell.c

@@ -0,0 +1,60 @@
+#include "shell.h"
+#include "keyboard.h"
+#include "terminal.h"
+#include "print.h"
+#include "string_utils.h"
+
+void execute(char *input) {
+    if (my_strcmp(input, "help") == 0) {
+        my_printf("Available commands: help, clear, exit\n");
+    } else if (my_strcmp(input, "clear") == 0) {
+        terminal_clear();
+    } else {
+        my_printf("Unknown command: %s\n", input);
+    }
+}
+
+void shell_loop()
+{
+    char input[256];
+    size_t index = 0;
+    char c;
+
+    while (1)
+    {
+        my_printf("> ");
+        index = 0;
+
+        while (1)
+        {
+            c = keyboard_get_char(); // Waits for input
+
+            if (c == '\n' || c == '\r')  // Enter key
+            {
+                input[index] = '\0';
+                my_printf("\n");
+                break;
+            }
+            else if (c == '\b' || c == 127) // Backspace
+            {
+                if (index > 0)
+                {
+                    index--;
+                    my_printf("\b \b"); // Erase last char on screen
+                }
+            }
+            else
+            {
+                if (index < sizeof(input) - 1) {
+                    input[index++] = c;
+                    terminal_putchar(c);
+                }
+            }
+        }
+
+        if (my_strcmp(input, "exit") == 0)
+            break;
+
+        execute(input);
+    }
+}

kernel/shell.h

@@ -0,0 +1,7 @@
+#ifndef SHELL_H
+#define SHELL_H
+
+void shell_loop(void);
+void execute(char *input);
+
+#endif

kernel/string_utils.c

@@ -0,0 +1,80 @@
+#include "string_utils.h"
+#include <stddef.h>
+#include <stdarg.h>
+
+size_t my_strlen(const char *str) {
+    const char *s = str;
+    while (*s) s++;
+    return s - str;
+}
+
+// Forward declaration of my_itoa
+static size_t my_itoa(int value, char *str, size_t size);
+
+int my_vsnprintf(char *str, size_t size, const char *format, ...) {
+    va_list args;
+    va_start(args, format);
+    size_t written = 0; // Change to size_t
+
+    for (const char *p = format; *p != '\0' && written < size - 1; p++) {
+        if (*p == '%') {
+            p++;
+            if (*p == 's') {
+                const char *s = va_arg(args, const char *);
+                while (*s && written < size - 1) {
+                    str[written++] = *s++;
+                }
+            } else if (*p == 'd') {
+                // Handle integer formatting
+                written += my_itoa(va_arg(args, int), str + written, size - written);
+            } else {
+                // Handle other formats as needed
+                str[written++] = *p; // Just copy the character
+            }
+        } else {
+            str[written++] = *p;
+        }
+    }
+    str[written] = '\0'; // Null-terminate the string
+    va_end(args);
+    return written;
+}
+
+static size_t my_itoa(int value, char *str, size_t size) {
+    size_t written = 0; // Change to size_t
+    if (value < 0) {
+        if (written < size - 1) {
+            str[written++] = '-';
+        }
+        value = -value;
+    }
+    // Convert integer to string
+    int temp = value;
+    int digits = 0;
+    do {
+        digits++;
+        temp /= 10;
+    } while (temp);
+    
+    if (written + digits >= size) {
+        digits = size - written - 1; // Prevent overflow
+    }
+    
+    str += written + digits; // Move pointer to the end
+    *str-- = '\0'; // Null-terminate the string
+
+    do {
+        *str-- = (value % 10) + '0';
+        value /= 10;
+    } while (value && str >= str - digits);
+    
+    return written + digits; // Return total written characters
+}
+
+int my_strcmp(const char *str1, const char *str2) {
+    while (*str1 && (*str1 == *str2)) {
+        str1++;
+        str2++;
+    }
+    return *(unsigned char *)str1 - *(unsigned char *)str2;
+}

kernel/string_utils.h

@@ -0,0 +1,11 @@
+#ifndef STRING_UTILS_H
+#define STRING_UTILS_H
+
+#include <stddef.h>
+#include <stdarg.h> // Include for va_list and related macros
+
+size_t my_strlen(const char *str); // Renamed to avoid conflict
+int my_vsnprintf(char *str, size_t size, const char *format, ...); // Renamed to avoid conflict
+int my_strcmp(const char *str1, const char *str2);
+
+#endif // STRING_UTILS_H

kernel/syscalls.c

@@ -1,5 +1,29 @@
 #include "syscalls.h"
+#include "scheduler.h"
+#include <stdarg.h>
 
-void syscall_handler() {
+void syscall_handler(int code, va_list args) {
-    // Syscall handling code
+    switch (code) {
+        case SYSCALL_INIT:
+            scheduler_init();
+            break;
+        case SYSCALL_SPAWN: {
+            void (*entry)(void) = va_arg(args, void (*)(void));
+            scheduler_add_task(entry);
+            break;
+        }
+        case SYSCALL_YIELD:
+            scheduler_yield();
+            break;
+        default:
+            // Unknown syscall
+            break;
+    }
+}
+
+void syscall(int code, ...) {
+    va_list args;
+    va_start(args, code);
+    syscall_handler(code, args);
+    va_end(args);
 }

kernel/syscalls.h

@@ -1,6 +1,17 @@
 #ifndef SYSCALLS_H
 #define SYSCALLS_H
 
-void syscall_handler();
+#include <stdarg.h>
+// Syscall numbers
+typedef enum {
+    SYSCALL_INIT = 0,
+    SYSCALL_SPAWN,
+    SYSCALL_YIELD
+} syscall_code_t;
+
+// Syscall dispatcher
+void syscall_handler(int code, va_list args);
+// Syscall interface
+void syscall(int code, ...);
 
 #endif // SYSCALLS_H

kernel/terminal.c

@@ -1,6 +1,7 @@
 #include <stdint.h>
 #include "io.h"
 #include "terminal.h"
+#include "vga.h"
 
 #define VGA_ADDRESS 0xB8000
 #define VGA_WIDTH 80
@@ -11,10 +12,7 @@ static uint16_t* const vga_buffer = (uint16_t*) VGA_ADDRESS;
 static uint8_t cursor_x = 0;
 static uint8_t cursor_y = 0;
 static uint8_t current_color = WHITE_ON_BLACK;
-
+static uint16_t last_cursor_pos = 0xFFFF;
-static uint16_t vga_entry(char c, uint8_t color) {
-    return (uint16_t) color << 8 | (uint8_t) c;
-}
 
 void terminal_initialize(void) {
     for (uint16_t y = 0; y < VGA_HEIGHT; y++) {
@@ -99,8 +97,10 @@ void terminal_clear(void) {
     update_cursor();
 }
 
-static void update_cursor() {
+void update_cursor(void) {
     uint16_t pos = cursor_y * VGA_WIDTH + cursor_x;
+    if (pos == last_cursor_pos) return;
+    last_cursor_pos = pos;
 
     outb(0x3D4, 0x0F);
     outb(0x3D5, (uint8_t)(pos & 0xFF));

kernel/threading.c

@@ -1,5 +1,118 @@
+#include <stdbool.h>
+#include <string.h>
+#include "malloc.h"
+#include "print.h"
 #include "threading.h"
 
-void threading_init() {
+#define MAX_THREADS 16           // Maximum number of threads
-    // Threading initialization code
+#define THREAD_STACK_SIZE 8192   // Stack size for each thread
+
+// The thread table stores information about all threads
+static Thread thread_table[MAX_THREADS];
+static uint32_t current_thread = 0;  // Index of the currently running thread
+static uint32_t num_threads = 0;     // Number of active threads
+
+// A simple mutex spinlock
+static volatile int mutex_locked = 0;
+
+// Function declaration for context_switch
+void context_switch(Thread *next);
+
+// Initialize the threading system
+void thread_init(void) {
+    memset(thread_table, 0, sizeof(thread_table));
+    num_threads = 0;
+}
+
+// Create a new thread
+void thread_create(Thread *thread __attribute__((unused)), void (*start_routine)(void *), void *arg) {
+    if (num_threads >= MAX_THREADS) {
+        my_printf("Error: Maximum thread count reached.\n");
+        return;
+    }
+
+    // Find an empty slot for the new thread
+    int index = num_threads++;
+    thread_table[index] = (Thread){0};
+    
+    // Set up the new thread
+    thread_table[index].start_routine = start_routine;
+    thread_table[index].arg = arg;
+    thread_table[index].stack_size = THREAD_STACK_SIZE;
+    thread_table[index].stack = (uint32_t*)malloc(THREAD_STACK_SIZE);
+    thread_table[index].stack_top = thread_table[index].stack + THREAD_STACK_SIZE / sizeof(uint32_t);
+
+    // Initialize the stack (simulate pushing the function's return address)
+    uint32_t *stack_top = thread_table[index].stack_top;
+    *(--stack_top) = (uint32_t)start_routine;  // Return address (the thread's entry point)
+    *(--stack_top) = (uint32_t)arg;            // Argument to pass to the thread
+
+    // Set the thread's state to ready
+    thread_table[index].state = THREAD_READY;
+
+    // If this is the first thread, switch to it
+    if (index == 0) {
+        scheduler();
+    }
+}
+
+// Yield the CPU to another thread
+void thread_yield(void) {
+    // Find the next thread in a round-robin manner
+    uint32_t next_thread = (current_thread + 1) % num_threads;
+    while (next_thread != current_thread && thread_table[next_thread].state != THREAD_READY) {
+        next_thread = (next_thread + 1) % num_threads;
+    }
+
+    if (next_thread != current_thread) {
+        current_thread = next_thread;
+        scheduler();
+    }
+}
+
+// Exit the current thread
+void thread_exit(void) {
+    thread_table[current_thread].state = THREAD_BLOCKED;  // Mark the thread as blocked (finished)
+    free(thread_table[current_thread].stack);             // Free the thread's stack
+    num_threads--;                                       // Decrease thread count
+
+    // Yield to the next thread
+    thread_yield();
+}
+
+// Scheduler: This function selects the next thread to run
+void scheduler(void) {
+    // Find the next ready thread
+    uint32_t next_thread = (current_thread + 1) % num_threads;
+    while (thread_table[next_thread].state != THREAD_READY) {
+        next_thread = (next_thread + 1) % num_threads;
+    }
+
+    if (next_thread != current_thread) {
+        current_thread = next_thread;
+        context_switch(&thread_table[current_thread]);
+    }
+}
+
+// Context switch to the next thread (assembly would go here to save/load registers)
+void context_switch(Thread *next) {
+    // For simplicity, context switching in this example would involve saving/restoring registers.
+    // In a real system, you would need to save the CPU state (registers) and restore the next thread's state.
+    my_printf("Switching to thread...\n");
+    next->start_routine(next->arg);  // Start running the next thread
+}
+
+// Simple mutex functions (spinlock)
+void mutex_init(void) {
+    mutex_locked = 0;
+}
+
+void mutex_lock(void) {
+    while (__sync_lock_test_and_set(&mutex_locked, 1)) {
+        // Busy wait (spinlock)
+    }
+}
+
+void mutex_unlock(void) {
+    __sync_lock_release(&mutex_locked);
 }

kernel/threading.h

@@ -1,6 +1,36 @@
 #ifndef THREADING_H
 #define THREADING_H
 
-void threading_init();
+#include <stdint.h>
+#include <stddef.h>
+
+// Define a basic thread structure (Thread Control Block - TCB)
+typedef struct Thread {
+    void (*start_routine)(void *);   // Function pointer to the thread function
+    void *arg;                       // Argument to be passed to the thread function
+    uint32_t *stack;                 // Pointer to the thread's stack
+    uint32_t *stack_top;             // Top of the stack (where the thread will start)
+    uint32_t stack_size;             // Size of the stack
+    uint32_t state;                  // Thread state (running, ready, blocked)
+} Thread;
+
+// Thread states
+#define THREAD_RUNNING 0
+#define THREAD_READY   1
+#define THREAD_BLOCKED 2
+
+// Thread management functions
+void thread_init(void);
+void thread_create(Thread *thread, void (*start_routine)(void *), void *arg);
+void thread_yield(void);
+void thread_exit(void);
+
+// Scheduler function
+void scheduler(void);
+
+// Synchronization functions (mutex spinlocks)
+void mutex_init(void);
+void mutex_lock(void);
+void mutex_unlock(void);
 
 #endif // THREADING_H

kernel/timer.c

@@ -2,15 +2,23 @@
 #include "io.h"
 #include "isr.h"
 #include "terminal.h"
+#include "stdio.h"
+#include "utils.h"
 
-static uint32_t tick = 0;
+static volatile uint32_t tick = 0;
 
 void timer_callback(void) {
     tick++;
-    // Optional: Print every 100 ticks
+
-    // if (tick % 100 == 0) {
+    if (tick % 100 == 0) {
-    //     terminal_write("Tick\n");
+        char buf[16];
-    // }
+        itoa(tick, buf, 10);
+        terminal_write("Tick count: ");
+        terminal_write(buf);
+        terminal_write("\n");
+    }
+
+    outb(0x20, 0x20); // EOI to PIC
 }
 
 void timer_init(uint32_t frequency) {

kernel/usb.c

@@ -0,0 +1,64 @@
+#include "usb.h"
+#include <stdint.h>
+#include <stdbool.h>
+
+// USB version detection
+bool usb_detect_version(uint16_t *version) {
+    if (!version) return false;
+    *version = 0x0110; // Example: USB 1.1
+    return true;
+}
+
+// USB initialization for 1.x
+bool usb_init(void) {
+    // Detect controller type (UHCI or OHCI)
+    bool uhci_supported = uhci_init();
+    bool ohci_supported = ohci_init();
+    
+    if (!uhci_supported && !ohci_supported) {
+        return false; // No supported controllers found
+    }
+    
+    return true;
+}
+
+// USB device enumeration (1.x)
+bool usb_enumerate_devices(void) {
+    // Implementation for detecting devices on USB 1.x ports
+    return true;
+}
+
+// HID initialization for USB 1.x
+bool usb_hid_init(void) {
+    // Ensure USB is initialized
+    if (!usb_init()) return false;
+    
+    return usb_enumerate_devices();
+}
+
+// USB transfers (stubs for 1.x)
+bool usb_control_transfer(/* parameters */) {
+    // Implement control transfer for USB 1.x
+    return true;
+}
+
+bool usb_interrupt_transfer(/* parameters */) {
+    // Implement interrupt transfer for USB 1.x
+    return true;
+}
+
+bool usb_bulk_transfer(/* parameters */) {
+    // Implement bulk transfer for USB 1.x
+    return true;
+}
+
+// USB host controller initialization (UHCI & OHCI)
+bool uhci_init(void) {
+    // Initialize UHCI controller (USB 1.x)
+    return true;
+}
+
+bool ohci_init(void) {
+    // Initialize OHCI controller (USB 1.x)
+    return true;
+}

kernel/usb.h

@@ -0,0 +1,24 @@
+#ifndef USB_H
+#define USB_H
+
+#include <stdint.h>
+#include <stdbool.h>
+
+// USB initialization and management functions
+bool usb_init(void);
+bool usb_detect_version(uint16_t *version);
+bool usb_enumerate_devices(void);
+
+// HID-specific functions
+bool usb_hid_init(void);
+
+// USB transfer functions
+bool usb_control_transfer(/* parameters */);
+bool usb_interrupt_transfer(/* parameters */);
+bool usb_bulk_transfer(/* parameters */);
+
+// USB host controller initialization (USB 1.x only)
+bool uhci_init(void); // UHCI support
+bool ohci_init(void); // OHCI support
+
+#endif // USB_H

kernel/utils.c

@@ -1,5 +1,78 @@
 #include "utils.h"
 
-void util_function() {
+static void reverse(char* str, int len) {
-    // Utility function code
+    int start = 0;
+    int end = len - 1;
+    while (start < end) {
+        char temp = str[start];
+        str[start++] = str[end];
+        str[end--] = temp;
+    }
+}
+
+// Integer to ASCII for signed ints
+char* itoa(int value, char* str, int base) {
+    int i = 0;
+    int isNegative = 0;
+    unsigned int uval;
+
+    if (base < 2 || base > 36) {
+        str[0] = '\0';
+        return str;
+    }
+
+    // Handle zero explicitly
+    if (value == 0) {
+        str[i++] = '0';
+        str[i] = '\0';
+        return str;
+    }
+
+    // Handle negative numbers (only for base 10)
+    if (value < 0 && base == 10) {
+        isNegative = 1;
+        uval = (unsigned int)(-value);
+    } else {
+        uval = (unsigned int)value;
+    }
+
+    while (uval != 0) {
+        int rem = uval % base;
+        str[i++] = (rem > 9) ? (rem - 10) + 'a' : rem + '0';
+        uval /= base;
+    }
+
+    if (isNegative) {
+        str[i++] = '-';
+    }
+
+    str[i] = '\0';
+    reverse(str, i);
+    return str;
+}
+
+// Integer to ASCII for unsigned ints
+char* utoa(unsigned int value, char* str, int base) {
+    int i = 0;
+
+    if (base < 2 || base > 36) {
+        str[0] = '\0';
+        return str;
+    }
+
+    if (value == 0) {
+        str[i++] = '0';
+        str[i] = '\0';
+        return str;
+    }
+
+    while (value != 0) {
+        int rem = value % base;
+        str[i++] = (rem > 9) ? (rem - 10) + 'a' : rem + '0';
+        value /= base;
+    }
+
+    str[i] = '\0';
+    reverse(str, i);
+    return str;
 }

kernel/utils.h

@@ -1,6 +1,12 @@
 #ifndef UTILS_H
 #define UTILS_H
 
-void util_function();
+#include <stddef.h>
+
+// Convert integer to string (base is typically 10, 16, etc.)
+char* itoa(int value, char* str, int base);
+
+// Convert unsigned integer to string (base is typically 10, 16, etc.)
+char* utoa(unsigned int value, char* str, int base);
 
 #endif // UTILS_H

kernel/vga.c

@@ -1,4 +1,20 @@
 #include "vga.h"
+#include <stddef.h>
+#include <stdbool.h>
+#include <string.h>
+#include <stdarg.h>
+#include "string_utils.h"
+
+void outb(uint16_t port, uint8_t value) {
+    __asm__ volatile("outb %0, %1" : : "a"(value), "Nd"(port));
+}
+
+// Read a byte from the specified port
+uint8_t inb(uint16_t port) {
+    uint8_t value;
+    __asm__ volatile("inb %1, %0" : "=a"(value) : "Nd"(port));
+    return value;
+}
 
 uint8_t vga_entry_color(vga_color fg, vga_color bg) {
     return fg | bg << 4;
@@ -7,3 +23,129 @@ uint8_t vga_entry_color(vga_color fg, vga_color bg) {
 uint16_t vga_entry(unsigned char uc, uint8_t color) {
     return (uint16_t)uc | (uint16_t)color << 8;
 }
+
+void vga_put_entry_at(char c, uint8_t color, size_t x, size_t y) {
+    const size_t index = y * 80 + x;
+    ((uint16_t*) 0xB8000)[index] = vga_entry(c, color);
+}
+
+void vga_clear(uint8_t color) {
+    uint16_t blank = vga_entry(' ', color);
+    for (size_t i = 0; i < 80 * 25; ++i) {
+        ((uint16_t*) 0xB8000)[i] = blank;
+    }
+}
+
+void vga_write_string(const char* data, size_t size) {
+    size_t x = 0;
+    size_t y = 0;
+    for (size_t i = 0; i < size; ++i) {
+        if (x >= 80) {  // If we reach the end of a line, move to the next line
+            x = 0;
+            y++;
+        }
+        
+        if (y >= 25) {  // If we reach the bottom of the screen, scroll
+            vga_scroll();
+            y = 24;  // Reset to the last row
+        }
+        
+        vga_put_entry_at(data[i], vga_entry_color(VGA_COLOR_LIGHT_GREY, VGA_COLOR_BLACK), x, y);
+        x++;  // Move to the next column
+    }
+}
+
+void vga_scroll(void) {
+    uint16_t* video_memory = (uint16_t*)0xB8000;
+    // Shift all lines up by one row (80 columns per row)
+    for (size_t i = 0; i < 24 * 80; ++i) {
+        video_memory[i] = video_memory[i + 80];  // Move one row up
+    }
+
+    // Clear the last row (bottom row)
+    for (size_t i = 24 * 80; i < 25 * 80; ++i) {
+        video_memory[i] = vga_entry(' ', vga_entry_color(VGA_COLOR_BLACK, VGA_COLOR_WHITE));
+    }
+}
+
+void vga_set_cursor_position(size_t x, size_t y) {
+    uint16_t position = y * 80 + x;  // Calculate linear index (y * 80 + x)
+
+    // Set the high byte of the cursor position
+    outb(VGA_PORT_INDEX, 0x0E);        // Cursor high byte register
+    outb(VGA_PORT_DATA, position >> 8);
+
+    // Set the low byte of the cursor position
+    outb(VGA_PORT_INDEX, 0x0F);        // Cursor low byte register
+    outb(VGA_PORT_DATA, position & 0xFF);
+}
+
+size_t vga_get_cursor_position(void) {
+    outb(VGA_PORT_INDEX, 0x0E);         // Cursor high byte register
+    uint8_t high = inb(VGA_PORT_DATA);  // Read high byte
+
+    outb(VGA_PORT_INDEX, 0x0F);         // Cursor low byte register
+    uint8_t low = inb(VGA_PORT_DATA);   // Read low byte
+
+    return (high << 8) | low;  // Combine the high and low bytes
+}
+
+void vga_set_cursor_blinking(bool enable) {
+    outb(VGA_PORT_INDEX, 0x0A);        // Cursor control register
+    uint8_t cursor = inb(VGA_PORT_DATA);
+    if (enable) {
+        cursor |= 0x01;  // Enable blinking
+    } else {
+        cursor &= ~0x01; // Disable blinking
+    }
+    outb(VGA_PORT_INDEX, 0x0A);        // Cursor control register
+    outb(VGA_PORT_DATA, cursor);
+}
+
+void vga_set_cursor_shape(uint8_t start, uint8_t end) {
+    outb(VGA_PORT_INDEX, 0x0A);        // Cursor control register
+    uint8_t cursor = inb(VGA_PORT_DATA);
+    cursor = (cursor & 0xC0) | (start & 0x3F);  // Set start of cursor shape
+    outb(VGA_PORT_DATA, cursor);
+
+    outb(VGA_PORT_INDEX, 0x0B);        // Cursor start register
+    outb(VGA_PORT_DATA, end & 0x3F);   // Set end of cursor shape
+}
+
+void vga_set_cursor_color(uint8_t color) {
+    outb(VGA_PORT_INDEX, 0x0A);        // Cursor control register
+    uint8_t cursor = inb(VGA_PORT_DATA);
+    cursor = (cursor & 0xC0) | (color & 0x3F);  // Set cursor color
+    outb(VGA_PORT_DATA, cursor);
+}
+
+void vga_set_cursor_blink_rate(uint8_t rate) {
+    outb(VGA_PORT_INDEX, 0x0A);        // Cursor control register
+    uint8_t cursor = inb(VGA_PORT_DATA);
+    cursor = (cursor & 0xC0) | (rate & 0x3F);  // Set cursor blink rate
+    outb(VGA_PORT_DATA, cursor);
+}
+
+void vga_printf(const char* format, ...) {
+    char buffer[256];  // Buffer to store the formatted string
+    va_list args;
+    va_start(args, format);
+    my_vsnprintf(buffer, sizeof(buffer), format, args); // Use my_vsnprintf instead of vsnprintf
+    va_end(args);
+    
+    // Now you can use the buffer with vga_write_string
+    vga_write_string(buffer, my_strlen(buffer)); // Use my_strlen instead of strlen
+}
+
+void vga_init(void) {
+    // Clear the screen
+    vga_clear(vga_entry_color(VGA_COLOR_BLACK, VGA_COLOR_LIGHT_GREY));
+    
+    // Set the cursor to the top-left corner
+    vga_set_cursor_position(0, 0);
+    
+    // Optionally, set cursor blinking and shape
+    vga_set_cursor_blinking(true);
+    vga_set_cursor_shape(0x20, 0x3F);  // Default shape
+}
+

kernel/vga.h

@@ -2,7 +2,11 @@
 #define VGA_H
 
 #include <stdint.h>
+#include <stddef.h>
+#include <stdbool.h>
+#include <stdarg.h>  // For va_list, va_start, etc.
 
+// VGA color definitions
 typedef enum {
     VGA_COLOR_BLACK = 0,
     VGA_COLOR_BLUE = 1,
@@ -22,7 +26,28 @@ typedef enum {
     VGA_COLOR_WHITE = 15,
 } vga_color;
 
+// VGA port addresses
+typedef enum {
+    VGA_PORT_INDEX = 0x3D4,  // Index register for VGA
+    VGA_PORT_DATA  = 0x3D5   // Data register for VGA
+} vga_io_port_t;
+
+// Function prototypes
 uint8_t vga_entry_color(vga_color fg, vga_color bg);
 uint16_t vga_entry(unsigned char uc, uint8_t color);
 
+void vga_put_entry_at(char c, uint8_t color, size_t x, size_t y);
+void vga_clear(uint8_t color);
+void vga_write_string(const char* data, size_t size);
+void vga_scroll(void);
+void vga_set_cursor_position(size_t x, size_t y);
+size_t vga_get_cursor_position(void);
+
+void vga_set_cursor_blinking(bool enable);
+void vga_set_cursor_shape(uint8_t start, uint8_t end);
+void vga_set_cursor_color(uint8_t color);
+void vga_set_cursor_blink_rate(uint8_t rate);
+
+void vga_printf(const char* format, ...);
+
 #endif

klibc/include/stdarg.h

@@ -0,0 +1,14 @@
+#ifndef CLASSICOS_KLIBC_STDARG_H
+#define CLASSICOS_KLIBC_STDARG_H
+
+typedef __builtin_va_list va_list;
+
+#ifndef va_start
+#define va_start(ap, param) __builtin_va_start(ap, param)
+#endif
+
+#define va_end(ap) __builtin_va_end(ap)
+#define va_arg(ap, type) __builtin_va_arg(ap, type)
+#define va_copy(dest, src) __builtin_va_copy(dest, src)
+
+#endif  // CLASSICOS_KLIBC_STDARG_H

klibc/include/stdbool.h

@@ -0,0 +1,12 @@
+#ifndef CLASSICOS_KLIBC_STDBOOL_H
+#define CLASSICOS_KLIBC_STDBOOL_H
+
+#ifndef __cplusplus
+#define bool _Bool
+#define true 1
+#define false 0
+#endif
+
+#define __bool_true_false_are_defined 1
+
+#endif  // CLASSICOS_KLIBC_STDBOOL_H

klibc/include/stddef.h

@@ -0,0 +1,10 @@
+#ifndef CLASSICOS_KLIBC_STDDEF_H
+#define CLASSICOS_KLIBC_STDDEF_H
+
+typedef __SIZE_TYPE__ size_t;
+typedef __PTRDIFF_TYPE__ ptrdiff_t;
+
+#undef NULL
+#define NULL ((void*)0)
+
+#endif  // CLASSICOS_KLIBC_STDDEF_H

klibc/include/stdint.h

@@ -0,0 +1,16 @@
+#ifndef CLASSICOS_KLIBC_STDINT_H
+#define CLASSICOS_KLIBC_STDINT_H
+
+typedef signed char int8_t;
+typedef short int int16_t;
+typedef int int32_t;
+typedef long long int int64_t;
+
+typedef unsigned char uint8_t;
+typedef unsigned short int uint16_t;
+typedef unsigned int uint32_t;
+typedef unsigned long long int uint64_t;
+
+typedef unsigned int uintptr_t;
+
+#endif  // CLASSICOS_KLIBC_STDINT_H

klibc/include/stdio.h

@@ -0,0 +1,4 @@
+#ifndef CLASSICOS_KLIBC_STDIO_H
+#define CLASSICOS_KLIBC_STDIO_H
+
+#endif  // CLASSICOS_KLIBC_STDIO_H

klibc/include/stdlib.h

@@ -0,0 +1,4 @@
+#ifndef CLASSICOS_KLIBC_STDLIB_H
+#define CLASSICOS_KLIBC_STDLIB_H
+
+#endif  // CLASSICOS_KLIBC_STDLIB_H

klibc/include/string.h

@@ -0,0 +1,14 @@
+#ifndef CLASSICOS_KLIBC_STRING_H
+#define CLASSICOS_KLIBC_STRING_H
+
+#include <stddef.h>
+
+extern int memcmp(const void* s1, const void* s2, size_t n);
+extern void* memmove(void* dst, const void* src, size_t n);
+extern void* memcpy(void* dst, const void* src, size_t n);
+extern void* memset(void* dst, int c, size_t n);
+
+extern size_t strlen(const char* s);
+extern int strcmp(const char* s1, const char* s2);
+
+#endif  // CLASSICOS_KLIBC_STRING_H

klibc/src/string.c

@@ -0,0 +1,107 @@
+#include <string.h>
+
+int memcmp(const void* s1, const void* s2, size_t n) {
+    const unsigned char* c1 = s1;
+    const unsigned char* c2 = s2;
+    int d = 0;
+
+    while (n--) {
+        d = (int)*c1++ - (int)*c2++;
+        if (d) break;
+    }
+
+    return d;
+}
+
+void* memmove(void* dst, const void* src, size_t n) {
+    const char* p = src;
+    char* q = dst;
+#if defined(__i386__) || defined(__x86_64__)
+    if (q < p) {
+        __asm__ volatile("cld; rep; movsb" : "+c"(n), "+S"(p), "+D"(q));
+    } else {
+        p += (n - 1);
+        q += (n - 1);
+        __asm__ volatile("std; rep; movsb; cld" : "+c"(n), "+S"(p), "+D"(q));
+    }
+#else
+    if (q < p) {
+        while (n--) {
+            *q++ = *p++;
+        }
+    } else {
+        p += n;
+        q += n;
+        while (n--) {
+            *--q = *--p;
+        }
+    }
+#endif
+
+    return dst;
+}
+
+void* memcpy(void* dst, const void* src, size_t n) {
+    const char* p = src;
+    char* q = dst;
+#if defined(__i386__)
+    size_t nl = n >> 2;
+    __asm__ volatile("cld ; rep ; movsl ; movl %3,%0 ; rep ; movsb"
+                    : "+c"(nl), "+S"(p), "+D"(q)
+                    : "r"(n & 3));
+#elif defined(__x86_64__)
+    size_t nq = n >> 3;
+    __asm__ volatile("cld ; rep ; movsq ; movl %3,%%ecx ; rep ; movsb"
+                    : "+c"(nq), "+S"(p), "+D"(q)
+                    : "r"((uint32_t)(n & 7)));
+#else
+    while (n--) {
+        *q++ = *p++;
+    }
+#endif
+
+    return dst;
+}
+
+void* memset(void* dst, int c, size_t n) {
+    char* q = dst;
+
+#if defined(__i386__)
+    size_t nl = n >> 2;
+    __asm__ volatile("cld ; rep ; stosl ; movl %3,%0 ; rep ; stosb"
+                    : "+c"(nl), "+D"(q)
+                    : "a"((unsigned char)c * 0x01010101U), "r"(n & 3));
+#elif defined(__x86_64__)
+    size_t nq = n >> 3;
+    __asm__ volatile("cld ; rep ; stosq ; movl %3,%%ecx ; rep ; stosb"
+                    : "+c"(nq), "+D"(q)
+                    : "a"((unsigned char)c * 0x0101010101010101U),
+                      "r"((uint32_t)n & 7));
+#else
+    while (n--) {
+        *q++ = c;
+    }
+#endif
+
+    return dst;
+}
+
+size_t strlen(const char* s) {
+    const char* ss = s;
+    while (*ss) ss++;
+    return ss - s;
+}
+
+int strcmp(const char* s1, const char* s2) {
+    const unsigned char* c1 = (const unsigned char*)s1;
+    const unsigned char* c2 = (const unsigned char*)s2;
+    unsigned char ch;
+    int d = 0;
+
+    while (1) {
+        d = (int)(ch = *c1++) - (int)*c2++;
+        if (d || !ch) break;
+    }
+
+    return d;
+}