diff --git a/.clang-format b/.clang-format
index a4fbf68..1edc97c 100644
--- a/.clang-format
+++ b/.clang-format
@@ -2,8 +2,9 @@ BasedOnStyle: LLVM
 IndentWidth: 4                 # Use 4 spaces for indentation
 TabWidth: 4                    # Tab width is also 4 spaces
 UseTab: Never                  # Always use spaces instead of tabs
-ColumnLimit: 80                # Wrap lines after 80 characters
+ColumnLimit: 120                # Wrap lines after 80 characters
 AllowShortLoopsOnASingleLine: true
+AllowShortFunctionsOnASingleLine: false
 AlwaysBreakTemplateDeclarations: true
 BreakConstructorInitializers: BeforeComma
 AlignConsecutiveDeclarations:
@@ -14,3 +15,4 @@ AlignConsecutiveDeclarations:
   AlignFunctionPointers: false
   PadOperators:    false
 AlignConsecutiveMacros: true
+AllowShortCaseLabelsOnASingleLine: true
diff --git a/Makefile b/Makefile
index 3d29dc2..99d8dfc 100644
--- a/Makefile
+++ b/Makefile
@@ -28,7 +28,7 @@ CFLAGS += -fno-omit-frame-pointer # More reliable backtraces in GDB
 
 all: kernel.elf
 
-kernel.elf: entry.o start.o lib/string.o lib/proc.o lib/spinlock.o lib/proc.o lib/uart.o lib/panic.o kern/kalloc.o lib/memory.o
+kernel.elf: entry.o start.o lib/string.o lib/proc.o lib/proc.o lib/uart.o lib/panic.o kern/kalloc.o lib/memory.o kern/ispinlock.o lib/spinlock.o
 	@echo LD $@
 	@$(LD) $(LDFLAGS) -o $@ $^
 
@@ -48,3 +48,5 @@ clean:
 	rm -f *.o *.elf *.d lib/*.o lib/*.d
 
 -include *.d
+
+.PHONY: all
diff --git a/kern/ispinlock.c b/kern/ispinlock.c
new file mode 100644
index 0000000..e2a9e79
--- /dev/null
+++ b/kern/ispinlock.c
@@ -0,0 +1,45 @@
+#include "ispinlock.h"
+
+void spinlock_init(spinlock_t *l) {
+    l->v = 0;
+}
+
+bool spin_trylock(spinlock_t *l) {
+    uint32_t old;
+    // old = xchg_acquire(&l->v, 1) using AMO
+    __asm__ volatile("amoswap.w.aq %0, %2, (%1)\n" : "=&r"(old) : "r"(&l->v), "r"(1u) : "memory");
+    return old == 0;
+}
+
+void spin_unlock(spinlock_t *l) {
+    // Release: store 0 with .rl ordering.
+    uint32_t dummy;
+    __asm__ volatile("amoswap.w.rl %0, %2, (%1)\n" : "=&r"(dummy) : "r"(&l->v), "r"(0u) : "memory");
+}
+
+// Optional: tiny pause/backoff (works even if Zihintpause isn't present).
+// See: https://github.com/riscv/riscv-isa-manual/blob/main/src/zihintpause.adoc
+void cpu_relax(void) {
+#if defined(__riscv_zihintpause)
+    __asm__ volatile("pause");
+#else
+    __asm__ volatile("nop");
+#endif
+}
+
+// Test-and-test-and-set acquire with polite spinning + exponential backoff.
+void spin_lock(spinlock_t *l) {
+    unsigned backoff = 1;
+    for (;;) {
+        if (spin_trylock(l))
+            return;
+
+        // Contended: spin on plain loads (no AMO) until it looks free.
+        while (__atomic_load_n(&l->v, __ATOMIC_RELAXED) != 0) {
+            for (unsigned i = 0; i < backoff; ++i) cpu_relax();
+            if (backoff < 1u << 12)
+                backoff <<= 1;
+        }
+        // Try again; loop.
+    }
+}
diff --git a/kern/ispinlock.h b/kern/ispinlock.h
new file mode 100644
index 0000000..e0303d6
--- /dev/null
+++ b/kern/ispinlock.h
@@ -0,0 +1,12 @@
+#pragma once
+#include <types.h>
+
+typedef struct {
+    volatile uint32_t v; // 0 = unlocked, 1 = locked
+} spinlock_t;
+
+void spinlock_init(spinlock_t *l);
+bool spin_trylock(spinlock_t *l);
+void spin_unlock(spinlock_t *l);
+void cpu_relax(void);
+void spin_lock(spinlock_t *l);
diff --git a/kern/kalloc.c b/kern/kalloc.c
index 1a034bd..d3d254c 100644
--- a/kern/kalloc.c
+++ b/kern/kalloc.c
@@ -1,8 +1,8 @@
+#include <ispinlock.h>
 #include <kalloc.h>
 #include <memory.h>
 #include <panic.h>
 #include <riscv.h>
-#include <spinlock.h>
 #include <string.h>
 #include <types.h>
 
@@ -23,12 +23,12 @@ struct Run {
 
 /** Kernel memory allocator. */
 struct {
-    struct Spinlock lock;
-    struct Run     *freelist;
+    spinlock_t  lock;
+    struct Run *freelist;
 } kmem;
 
 void kalloc_init() {
-    initlock(&kmem.lock, "kmem");
+    spinlock_init(&kmem.lock);
     freerange(kernel_end, (void *)PHYSTOP);
 }
 
@@ -43,8 +43,7 @@ void kfree(void *pa) {
 
     // Assert that page is a ligned to a page boundary and that its correctly
     // sized
-    if (((u64)pa % PGSIZE) != 0 || (char *)pa < kernel_end ||
-        (u64)pa >= PHYSTOP)
+    if (((u64)pa % PGSIZE) != 0 || (char *)pa < kernel_end || (u64)pa >= PHYSTOP)
         panic("kfree");
 
     // Fill with junk to catch dangling refs.
@@ -52,23 +51,23 @@ void kfree(void *pa) {
 
     r = (struct Run *)pa;
 
-    acquire(&kmem.lock);
+    spin_lock(&kmem.lock);
     r->next = kmem.freelist;
     kmem.freelist = r;
-    release(&kmem.lock);
+    spin_unlock(&kmem.lock);
 }
 
 void *kalloc(void) {
     struct Run *r;
 
-    acquire(&kmem.lock);
+    spin_lock(&kmem.lock);
 
     r = kmem.freelist;
 
     if (r)
         kmem.freelist = r->next;
 
-    release(&kmem.lock);
+    spin_unlock(&kmem.lock);
 
     if (r)
         memset((char *)r, 5, PGSIZE); // fill with junk
diff --git a/kern/kprint.c b/kern/kprint.c
new file mode 100644
index 0000000..01bbac3
--- /dev/null
+++ b/kern/kprint.c
@@ -0,0 +1,103 @@
+#include <stdarg.h>
+#include <stdbool.h>
+
+static void append_char(char **buf, size_t *remaining, char c) {
+    if (*remaining > 1) { // Leave space for null terminator
+        **buf = c;
+        (*buf)++;
+        (*remaining)--;
+    }
+}
+
+static void append_str(char **buf, size_t *remaining, const char *str) {
+    while (*str) {
+        append_char(buf, remaining, *str++);
+    }
+}
+
+static void append_int(char **buf, size_t *remaining, int value, int base) {
+    char        tmp[32];
+    const char *digits = "0123456789abcdef";
+    bool        neg = false;
+    int         i = 0;
+
+    if (base == 10 && value < 0) {
+        neg = true;
+        value = -value;
+    }
+
+    do {
+        tmp[i++] = digits[value % base];
+        value /= base;
+    } while (value && i < (int)sizeof(tmp));
+
+    if (neg) {
+        tmp[i++] = '-';
+    }
+
+    while (i--) {
+        append_char(buf, remaining, tmp[i]);
+    }
+}
+
+int kvsnprintf(char *buf, size_t size, const char *fmt, va_list args) {
+    char  *p = buf;
+    size_t remaining = size;
+
+    while (*fmt) {
+        if (*fmt != '%') {
+            append_char(&p, &remaining, *fmt++);
+            continue;
+        }
+        fmt++; // skip '%'
+
+        switch (*fmt) {
+        case 's':
+            append_str(&p, &remaining, va_arg(args, const char *));
+            break;
+        case 'd':
+            append_int(&p, &remaining, va_arg(args, int), 10);
+            break;
+        case 'x':
+            append_int(&p, &remaining, va_arg(args, unsigned int), 16);
+            break;
+        case 'c':
+            append_char(&p, &remaining, (char)va_arg(args, int));
+            break;
+        case '%':
+            append_char(&p, &remaining, '%');
+            break;
+        default:
+            append_char(&p, &remaining, '?');
+            break;
+        }
+        fmt++;
+    }
+
+    if (remaining > 0) {
+        *p = '\0';
+    } else {
+        buf[size - 1] = '\0';
+    }
+
+    return (int)(p - buf);
+}
+
+int ksnprintf(char *buf, size_t size, const char *fmt, ...) {
+    va_list args;
+    va_start(args, fmt);
+    int ret = kvsnprintf(buf, size, fmt, args);
+    va_end(args);
+    return ret;
+}
+
+void kprintf(const char *fmt, ...) {
+    char    buffer[256];
+    va_list args;
+    va_start(args, fmt);
+    kvsnprintf(buffer, sizeof(buffer), fmt, args);
+    va_end(args);
+
+    // Now send buffer to your console/serial output
+    console_write(buffer);
+}
diff --git a/kern/libkern/string.c b/kern/libkern/string.c
new file mode 100644
index 0000000..99a0ea2
--- /dev/null
+++ b/kern/libkern/string.c
@@ -0,0 +1,127 @@
+#include <string.h>
+
+void *memcpy(void *s1, const void *s2, size_t n) {
+    const char *f = s2;
+    char       *t = s1;
+
+    while (n-- > 0) *t++ = *f++;
+    return s1;
+}
+
+void *memmove(void *s1, const void *s2, size_t n) {
+    const char *f = s2;
+    char       *t = s1;
+
+    if (f < t) {
+        f += n;
+        t += n;
+        while (n-- > 0) *--t = *--f;
+    } else
+        while (n-- > 0) *t++ = *f++;
+    return s1;
+}
+
+void *memset(void *dest, int c, size_t n) {
+    unsigned char *s = dest;
+    size_t         k;
+
+    /* Fill head and tail with minimal branching. Each
+     * conditional ensures that all the subsequently used
+     * offsets are well-defined and in the dest region. */
+
+    if (!n)
+        return dest;
+    s[0] = c;
+    s[n - 1] = c;
+    if (n <= 2)
+        return dest;
+    s[1] = c;
+    s[2] = c;
+    s[n - 2] = c;
+    s[n - 3] = c;
+    if (n <= 6)
+        return dest;
+    s[3] = c;
+    s[n - 4] = c;
+    if (n <= 8)
+        return dest;
+
+    /* Advance pointer to align it at a 4-byte boundary,
+     * and truncate n to a multiple of 4. The previous code
+     * already took care of any head/tail that get cut off
+     * by the alignment. */
+
+    k = -(uintptr_t)s & 3;
+    s += k;
+    n -= k;
+    n &= -4;
+
+#ifdef __GNUC__
+    typedef uint32_t __attribute__((__may_alias__)) u32;
+    typedef uint64_t __attribute__((__may_alias__)) u64;
+
+    u32 c32 = ((u32)-1) / 255 * (unsigned char)c;
+
+    /* In preparation to copy 32 bytes at a time, aligned on
+     * an 8-byte bounary, fill head/tail up to 28 bytes each.
+     * As in the initial byte-based head/tail fill, each
+     * conditional below ensures that the subsequent offsets
+     * are valid (e.g. !(n<=24) implies n>=28). */
+
+    *(u32 *)(s + 0) = c32;
+    *(u32 *)(s + n - 4) = c32;
+    if (n <= 8)
+        return dest;
+    *(u32 *)(s + 4) = c32;
+    *(u32 *)(s + 8) = c32;
+    *(u32 *)(s + n - 12) = c32;
+    *(u32 *)(s + n - 8) = c32;
+    if (n <= 24)
+        return dest;
+    *(u32 *)(s + 12) = c32;
+    *(u32 *)(s + 16) = c32;
+    *(u32 *)(s + 20) = c32;
+    *(u32 *)(s + 24) = c32;
+    *(u32 *)(s + n - 28) = c32;
+    *(u32 *)(s + n - 24) = c32;
+    *(u32 *)(s + n - 20) = c32;
+    *(u32 *)(s + n - 16) = c32;
+
+    /* Align to a multiple of 8 so we can fill 64 bits at a time,
+     * and avoid writing the same bytes twice as much as is
+     * practical without introducing additional branching. */
+
+    k = 24 + ((uintptr_t)s & 4);
+    s += k;
+    n -= k;
+
+    /* If this loop is reached, 28 tail bytes have already been
+     * filled, so any remainder when n drops below 32 can be
+     * safely ignored. */
+
+    u64 c64 = c32 | ((u64)c32 << 32);
+    for (; n >= 32; n -= 32, s += 32) {
+        *(u64 *)(s + 0) = c64;
+        *(u64 *)(s + 8) = c64;
+        *(u64 *)(s + 16) = c64;
+        *(u64 *)(s + 24) = c64;
+    }
+#else
+    /* Pure C fallback with no aliasing violations. */
+    for (; n; n--, s++) *s = c;
+#endif
+
+    return dest;
+}
+
+int memcmp(const void *s1, const void *s2, size_t n) {
+    if (n != 0) {
+        const unsigned char *p1 = s1, *p2 = s2;
+
+        do {
+            if (*p1++ != *p2++)
+                return (*--p1 - *--p2);
+        } while (--n != 0);
+    }
+    return (0);
+}
diff --git a/kern/libkern/string.h b/kern/libkern/string.h
new file mode 100644
index 0000000..01d3853
--- /dev/null
+++ b/kern/libkern/string.h
@@ -0,0 +1,7 @@
+#pragma once
+#include <types.h>
+
+void *memcpy(void *s1, const void *s2, size_t n);
+void *memmove(void *s1, const void *s2, size_t n);
+void *memset(void *dest, int c, size_t n);
+int   memcmp(const void *s1, const void *s2, size_t n);
diff --git a/kern/proc.h b/kern/proc.h
new file mode 100644
index 0000000..b594fc9
--- /dev/null
+++ b/kern/proc.h
@@ -0,0 +1,77 @@
+#include <stdint.h>
+
+typedef enum {
+    UNUSED,
+    USED,
+    SLEEPING,
+    RUNNABLE,
+    RUNNING,
+    ZOMBIE,
+} ProcessState;
+
+/** Saved registers for kernel context switches. */
+struct Context {
+    uint64_t ra;
+    uint64_t sp;
+
+    // callee-saved
+    uint64_t s0;
+    uint64_t s1;
+    uint64_t s2;
+    uint64_t s3;
+    uint64_t s4;
+    uint64_t s5;
+    uint64_t s6;
+    uint64_t s7;
+    uint64_t s8;
+    uint64_t s9;
+    uint64_t s10;
+    uint64_t s11;
+};
+
+/** Per-CPU state. */
+struct Cpu {
+    struct Process *proc;    // The process running on this cpu, or null.
+    struct Context  context; // swtch() here to enter scheduler().
+    int             noff;    // Depth of push_off() nesting.
+    int             intena;  // Were interrupts enabled before push_off()?
+};
+
+typedef struct {
+    /*   0 */ uint64_t kernel_satp;   // kernel page table
+    /*   8 */ uint64_t kernel_sp;     // top of process's kernel stack
+    /*  16 */ uint64_t kernel_trap;   // usertrap()
+    /*  24 */ uint64_t epc;           // saved user program counter
+    /*  32 */ uint64_t kernel_hartid; // saved kernel tp
+    /*  40 */ uint64_t ra;
+    /*  48 */ uint64_t sp;
+    /*  56 */ uint64_t gp;
+    /*  64 */ uint64_t tp;
+    /*  72 */ uint64_t t0;
+    /*  80 */ uint64_t t1;
+    /*  88 */ uint64_t t2;
+    /*  96 */ uint64_t s0;
+    /* 104 */ uint64_t s1;
+    /* 112 */ uint64_t a0;
+    /* 120 */ uint64_t a1;
+    /* 128 */ uint64_t a2;
+    /* 136 */ uint64_t a3;
+    /* 144 */ uint64_t a4;
+    /* 152 */ uint64_t a5;
+    /* 160 */ uint64_t a6;
+    /* 168 */ uint64_t a7;
+    /* 176 */ uint64_t s2;
+    /* 184 */ uint64_t s3;
+    /* 192 */ uint64_t s4;
+    /* 200 */ uint64_t s5;
+    /* 208 */ uint64_t s6;
+    /* 216 */ uint64_t s7;
+    /* 224 */ uint64_t s8;
+    /* 232 */ uint64_t s9;
+    /* 240 */ uint64_t s10;
+    /* 248 */ uint64_t s11;
+    /* 256 */ uint64_t t3;
+    /* 264 */ uint64_t t4;
+    /* 272 */ uint64_t t5;
+    /* 280 */ uint64_t t6;
+} TrapFrame_t;
diff --git a/start.c b/start.c
index d9ca8ef..7f72faf 100644
--- a/start.c
+++ b/start.c
@@ -1,9 +1,9 @@
 #include <config.h>
+#include <ispinlock.h>
 #include <kalloc.h>
 #include <memory.h>
 #include <proc.h>
 #include <riscv.h>
-#include <spinlock.h>
 #include <types.h>
 #include <uart.h>
 
@@ -16,8 +16,8 @@
 char stack0[4096 * NCPU] __attribute__((aligned(16)));
 
 /* Keep this here and sync on it until we have synchronized printf */
-struct Spinlock sl = {0};
-volatile int    greeted = 0;
+spinlock_t   sl = {0};
+volatile int hold = 1;
 
 /* This is where entry.S drops us of. All cores land here */
 void start() {
@@ -29,24 +29,22 @@ void start() {
     // cpu (struct Cpu).
     write_tp(id);
 
-    acquire(&sl);
-
-    if (!greeted) {
-        uart_puts("Hello Neptune!\n");
-        greeted = 1;
-    }
-
-    uart_puts("Hart number: ");
-    uart_putc(id + '0');
-    uart_putc('\n');
-
-    release(&sl);
-
     if (id == 0) {
         /* Here we will do a bunch of initialization steps */
         kalloc_init();
+        uart_puts("Hello Neptune!\n");
+        spinlock_init(&sl);
+        hold = 0;
     }
 
+    while (hold);
+
+    spin_lock(&sl);
+    uart_puts("Hart number: ");
+    uart_putc(id + '0');
+    uart_putc('\n');
+    spin_unlock(&sl);
+
     // We should not arrive here, but if we do, hang in a while on wfi.
     while (1) __asm__ volatile("wfi"); // (Wait For Interrupt)
 }
diff --git a/types.h b/types.h
index cd25ff8..69e74d4 100644
--- a/types.h
+++ b/types.h
@@ -4,4 +4,13 @@ typedef unsigned char  u8;
 typedef unsigned short u16;
 typedef unsigned int   u32;
 typedef unsigned long  u64;
-typedef u64            size_t;
+
+typedef unsigned char  uint8_t;
+typedef unsigned short uint16_t;
+typedef unsigned int   uint32_t;
+typedef unsigned long  uint64_t;
+typedef uint64_t       size_t;
+
+typedef uint64_t uintptr_t;
+
+typedef u8 bool;