# AArch64 (native or cross-compile)
aarch64-linux-gnu-gcc -S -O2 foo.c -o foo.s

# 32-bit ARM Thumb
arm-linux-gnueabihf-gcc -S -O2 -mthumb foo.c -o foo.s

# From objdump
aarch64-linux-gnu-objdump -d -S prog

# From GDB on target
(gdb) disassemble /s main

// Non-leaf function
stp  x29, x30, [sp, #-32]!   // save fp, lr; allocate 32 bytes
mov  x29, sp                  // set frame pointer
stp  x19, x20, [sp, #16]     // save callee-saved registers
// ... body ...
ldp  x19, x20, [sp, #16]     // restore
ldp  x29, x30, [sp], #32     // restore fp, lr; deallocate
ret

// Leaf function (no calls, no callee-saved regs needed)
// Can use red zone (no rsp adjustment) — but AArch64 has no red zone
sub  sp, sp, #16             // allocate locals
// ... body ...
add  sp, sp, #16
ret

// Barrier
__asm__ volatile ("dmb ish" ::: "memory");

// Load acquire
static inline int load_acquire(volatile int *p) {
    int val;
    __asm__ volatile ("ldar %w0, %1" : "=r"(val) : "Q"(*p));
    return val;
}

// Store release
static inline void store_release(volatile int *p, int val) {
    __asm__ volatile ("stlr %w1, %0" : "=Q"(*p) : "r"(val));
}

// Read system counter
static inline uint64_t read_cntvct(void) {
    uint64_t val;
    __asm__ volatile ("mrs %0, cntvct_el0" : "=r"(val));
    return val;
}

#include <arm_neon.h>

// Add 4 floats at once
float32x4_t a = vld1q_f32(arr_a);   // load 4 floats
float32x4_t b = vld1q_f32(arr_b);
float32x4_t c = vaddq_f32(a, b);
vst1q_f32(result, c);

// Horizontal sum
float32x4_t sum = vpaddq_f32(c, c);
sum = vpaddq_f32(sum, sum);
float total = vgetq_lane_f32(sum, 0);