diff --git a/.gitignore b/.gitignore
new file mode 100644
index 0000000..880ac59
--- /dev/null
+++ b/.gitignore
@@ -0,0 +1,2 @@
+build/
+.DS_Store
diff --git a/CMakeLists.txt b/CMakeLists.txt
index 77f5900..8609f68 100644
--- a/CMakeLists.txt
+++ b/CMakeLists.txt
@@ -1,8 +1,7 @@
 cmake_minimum_required(VERSION 3.0)
 
 project(ucore_user)
-
-enable_language(C ASM)
+enable_language(ASM)
 
 # Path
 if (${ARCH} STREQUAL riscv32 OR ${ARCH} STREQUAL riscv64)
@@ -36,6 +35,8 @@ elseif (${ARCH} STREQUAL riscv64)
     set(CMAKE_C_FLAGS       "-march=rv64imac -mabi=lp64 -mcmodel=medany")
 elseif (${ARCH} STREQUAL aarch64)
     set(PREFIX              aarch64-none-elf-)
+    set(CMAKE_C_FLAGS       "-mgeneral-regs-only")
+    set(CMAKE_C_LINK_FLAGS  "-Ttext 0xffff000000000000")
 else()
     message("Unsupported arch: ${ARCH}")
 endif ()
@@ -44,8 +45,8 @@ set(CMAKE_C_COMPILER        ${PREFIX}gcc)
 set(CMAKE_RANLIB            ${PREFIX}ranlib)
 set(CMAKE_C_FLAGS           "${CMAKE_C_FLAGS} -fno-builtin -nostdinc -fno-stack-protector -ggdb -Wall")
 set(CMAKE_ASM_FLAGS         ${CMAKE_C_FLAGS})
-set(CMAKE_C_LINK_FLAGS      "-nostdlib")    # get rid of '-Wl,-search_paths_first -Wl,-headerpad_max_install_names'
-set(CMAKE_SHARED_LIBRARY_LINK_C_FLAGS)      # get rid of '-rdynamic' on Linux
+set(CMAKE_C_LINK_FLAGS      "${CMAKE_C_LINK_FLAGS} -nostdlib")  # get rid of '-Wl,-search_paths_first -Wl,-headerpad_max_install_names'
+set(CMAKE_SHARED_LIBRARY_LINK_C_FLAGS)                          # get rid of '-rdynamic' on Linux
 
 # Library
 add_library(ulib ${LIBS})
diff --git a/src/arch/aarch64/arch.h b/src/arch/aarch64/arch.h
new file mode 100644
index 0000000..4fdfb90
--- /dev/null
+++ b/src/arch/aarch64/arch.h
@@ -0,0 +1,14 @@
+// See LICENSE for license details.
+
+#ifndef __LIBS_AARCH64_H__
+#define __LIBS_AARCH64_H__
+
+#define do_div(n, base)                              \
+    ({                                               \
+        int __res;                                   \
+        __res = ((unsigned long)n) % (unsigned)base; \
+        n = ((unsigned long)n) / (unsigned)base;     \
+        __res;                                       \
+    })
+
+#endif
diff --git a/src/arch/aarch64/atomic.h b/src/arch/aarch64/atomic.h
new file mode 100644
index 0000000..bbbb8da
--- /dev/null
+++ b/src/arch/aarch64/atomic.h
@@ -0,0 +1,75 @@
+#ifndef __LIBS_ATOMIC_H__
+#define __LIBS_ATOMIC_H__
+
+// TODO: implement atomic operations for aarch64
+
+/* Atomic operations that C can't guarantee us. Useful for resource counting etc.. */
+
+static inline void set_bit(int nr, volatile void *addr) __attribute__((always_inline));
+static inline void clear_bit(int nr, volatile void *addr) __attribute__((always_inline));
+static inline void change_bit(int nr, volatile void *addr) __attribute__((always_inline));
+static inline bool test_and_set_bit(int nr, volatile void *addr) __attribute__((always_inline));
+static inline bool test_and_clear_bit(int nr, volatile void *addr) __attribute__((always_inline));
+static inline bool test_bit(int nr, volatile void *addr) __attribute__((always_inline));
+
+/* *
+ * set_bit - Atomically set a bit in memory
+ * @nr:     the bit to set
+ * @addr:   the address to start counting from
+ *
+ * Note that @nr may be almost arbitrarily large; this function is not
+ * restricted to acting on a single-word quantity.
+ * */
+static inline void set_bit(int nr, volatile void *addr) {
+    /* unimplemented */
+}
+
+/* *
+ * clear_bit - Atomically clears a bit in memory
+ * @nr:     the bit to clear
+ * @addr:   the address to start counting from
+ * */
+static inline void clear_bit(int nr, volatile void *addr) {
+    /* unimplemented */
+}
+
+/* *
+ * change_bit - Atomically toggle a bit in memory
+ * @nr:     the bit to change
+ * @addr:   the address to start counting from
+ * */
+static inline void change_bit(int nr, volatile void *addr) {
+    /* unimplemented */
+}
+
+/* *
+ * test_bit - Determine whether a bit is set
+ * @nr:     the bit to test
+ * @addr:   the address to count from
+ * */
+static inline bool test_bit(int nr, volatile void *addr) {
+    /* unimplemented */
+    return 0;
+}
+
+/* *
+ * test_and_set_bit - Atomically set a bit and return its old value
+ * @nr:     the bit to set
+ * @addr:   the address to count from
+ * */
+static inline bool test_and_set_bit(int nr, volatile void *addr) {
+    /* unimplemented */
+    return 0;
+}
+
+/* *
+ * test_and_clear_bit - Atomically clear a bit and return its old value
+ * @nr:     the bit to clear
+ * @addr:   the address to count from
+ * */
+static inline bool test_and_clear_bit(int nr, volatile void *addr) {
+    /* unimplemented */
+    return 0;
+}
+
+#endif /* !__LIBS_ATOMIC_H__ */
diff --git a/src/arch/aarch64/defs.h b/src/arch/aarch64/defs.h
new file mode 100644
index 0000000..1e1def7
--- /dev/null
+++ b/src/arch/aarch64/defs.h
@@ -0,0 +1,79 @@
+#ifndef __LIBS_DEFS_H__
+#define __LIBS_DEFS_H__
+
+#ifndef NULL
+#define NULL ((void *)0)
+#endif
+
+#define __always_inline inline __attribute__((always_inline))
+#define __noinline __attribute__((noinline))
+#define __noreturn __attribute__((noreturn))
+
+#define CHAR_BIT        8
+
+/* Represents true-or-false values */
+typedef long long bool;
+
+/* Explicitly-sized versions of integer types */
+typedef char int8_t;
+typedef unsigned char uint8_t;
+typedef short int16_t;
+typedef unsigned short uint16_t;
+typedef int int32_t;
+typedef unsigned int uint32_t;
+typedef long long int64_t;
+typedef unsigned long long uint64_t;
+
+/* *
+ * Pointers and addresses are 32 bits long.
+ * We use pointer types to represent addresses,
+ * uintptr_t to represent the numerical values of addresses.
+ * */
+typedef int64_t intptr_t;
+typedef uint64_t uintptr_t;
+
+/* size_t is used for memory object sizes */
+typedef uintptr_t size_t;
+
+/* off_t is used for file offsets and lengths */
+typedef intptr_t off_t;
+
+/* used for page numbers */
+typedef size_t ppn_t;
+
+/* *
+ * Rounding operations (efficient when n is a power of 2)
+ * Round down to the nearest multiple of n
+ * */
+#define ROUNDDOWN(a, n) ({                                          \
+            size_t __a = (size_t)(a);                               \
+            (typeof(a))(__a - __a % (n));                           \
+        })
+
+/* Round up to the nearest multiple of n */
+#define ROUNDUP(a, n) ({                                            \
+            size_t __n = (size_t)(n);                               \
+            (typeof(a))(ROUNDDOWN((size_t)(a) + __n - 1, __n));     \
+        })
+
+/* Round up the result of dividing of n */
+#define ROUNDUP_DIV(a, n) ({                                        \
+uint64_t __n = (uint64_t)(n);                           \
+(typeof(a))(((a) + __n - 1) / __n);                     \
+})
+
+/* Return the offset of 'member' relative to the beginning of a struct type */
+#define offsetof(type, member)                                      \
+    ((size_t)(&((type *)0)->member))
+
+/* *
+ * to_struct - get the struct from a ptr
+ * @ptr:    a struct pointer of member
+ * @type:   the type of the struct this is embedded in
+ * @member: the name of the member within the struct
+ * */
+#define to_struct(ptr, type, member)                               \
+    ((type *)((char *)(ptr) - offsetof(type, member)))
+
+#endif /* !__LIBS_DEFS_H__ */
+
diff --git a/src/arch/aarch64/elf.h b/src/arch/aarch64/elf.h
new file mode 100644
index 0000000..867066f
--- /dev/null
+++ b/src/arch/aarch64/elf.h
@@ -0,0 +1,48 @@
+#ifndef __LIBS_ELF_H__
+#define __LIBS_ELF_H__
+
+#include <defs.h>
+
+#define ELF_MAGIC   0x464C457FU         // "\x7FELF" in little endian
+
+/* file header */
+struct elfhdr {
+    uint32_t e_magic;     // must equal ELF_MAGIC
+    uint8_t e_elf[12];
+    uint16_t e_type;      // 1=relocatable, 2=executable, 3=shared object, 4=core image
+    uint16_t e_machine;   // 3=x86, 4=68K, etc.
+    uint32_t e_version;   // file version, always 1
+    uint64_t e_entry;     // entry point if executable
+    uint64_t e_phoff;     // file position of program header or 0
+    uint64_t e_shoff;     // file position of section header or 0
+    uint32_t e_flags;     // architecture-specific flags, usually 0
+    uint16_t e_ehsize;    // size of this elf header
+    uint16_t e_phentsize; // size of an entry in program header
+    uint16_t e_phnum;     // number of entries in program header or 0
+    uint16_t e_shentsize; // size of an entry in section header
+    uint16_t e_shnum;     // number of entries in section header or 0
+    uint16_t e_shstrndx;  // section number that contains section name strings
+};
+
+/* program section header */
+struct proghdr {
+    uint32_t p_type;   // loadable code or data, dynamic linking info,etc.
+    uint32_t p_flags;  // read/write/execute bits
+    uint64_t p_offset; // file offset of segment
+    uint64_t p_va;     // virtual address to map segment
+    uint64_t p_pa;     // physical address, not used
+    uint64_t p_filesz; // size of segment in file
+    uint64_t p_memsz;  // size of segment in memory (bigger if contains bss）
+    uint64_t p_align;  // required alignment, invariably hardware page size
+};
+
+/* values for Proghdr::p_type */
+#define ELF_PT_LOAD                     1
+
+/* flag bits for Proghdr::p_flags */
+#define ELF_PF_X                        1
+#define ELF_PF_W                        2
+#define ELF_PF_R                        4
+
+#endif /* !__LIBS_ELF_H__ */
+
diff --git a/src/arch/aarch64/initcode.S b/src/arch/aarch64/initcode.S
new file mode 100644
index 0000000..91fa2eb
--- /dev/null
+++ b/src/arch/aarch64/initcode.S
@@ -0,0 +1,8 @@
+.text
+.globl _start
+_start:
+    # call user-program function
+    ldr x0, [sp]
+    add x1, sp, #8
+    bl umain
+1:  b 1b
diff --git a/src/libs/dirent.h b/src/libs/dirent.h
index 429e4fe..d56c96a 100755
--- a/src/libs/dirent.h
+++ b/src/libs/dirent.h
@@ -5,7 +5,7 @@
 #include <unistd.h>
 
 struct dirent {
-    off_t offset;
+    int32_t offset;
     char name[FS_MAX_FNAME_LEN + 1];
 };
 
diff --git a/src/libs/stat.h b/src/libs/stat.h
index 41a392d..f17d0d3 100644
--- a/src/libs/stat.h
+++ b/src/libs/stat.h
@@ -5,9 +5,9 @@
 
 struct stat {
     uint32_t st_mode;                   // protection mode and file type
-    size_t st_nlinks;                   // number of hard links
-    size_t st_blocks;                   // number of blocks file is using
-    size_t st_size;                     // file size (bytes)
+    uint32_t st_nlinks;                 // number of hard links
+    uint32_t st_blocks;                 // number of blocks file is using
+    uint32_t st_size;                   // file size (bytes)
 };
 
 #define S_IFMT          070000          // mask for type of file
diff --git a/src/sh.c b/src/sh.c
index 9a652f1..60bf86c 100755
--- a/src/sh.c
+++ b/src/sh.c
@@ -78,7 +78,7 @@ readline(const char *prompt) {
             i --;
         }
         else if (c == '\n' || c == '\r') {
-            putc(c);
+            putc('\n');
             buffer[i] = '\0';
             break;
         }
diff --git a/src/ulibs/syscall.c b/src/ulibs/syscall.c
index 3e4758f..b121008 100644
--- a/src/ulibs/syscall.c
+++ b/src/ulibs/syscall.c
@@ -12,10 +12,10 @@ static inline int
 syscall(int num, ...) {
     va_list ap;
     va_start(ap, num);
-    uint32_t a[MAX_ARGS];
+    size_t a[MAX_ARGS];
     int i, ret;
     for (i = 0; i < MAX_ARGS; i ++) {
-        a[i] = va_arg(ap, uint32_t);
+        a[i] = va_arg(ap, size_t);
     }
     va_end(ap);
 
@@ -50,6 +50,25 @@ syscall(int num, ...) {
     "m" (a[4])
     : "memory"
     );
+#elif defined(__aarch64__)
+    asm volatile (
+        "ldr w8, %1\n"
+        "ldr x0, %2\n"
+        "ldr x1, %3\n"
+        "ldr x2, %4\n"
+        "ldr x3, %5\n"
+        "ldr x4, %6\n"
+        "svc 0\n"
+        "str w0, %0"
+         : "=m" (ret)
+         : "m" (num),
+           "m" (a[0]),
+           "m" (a[1]),
+           "m" (a[2]),
+           "m" (a[3]),
+           "m" (a[4])
+        : "cc", "memory"
+      );
 #endif
     return ret;
 }
diff --git a/src/ulibs/umain.c b/src/ulibs/umain.c
index 9bee9ab..8c45ea3 100644
--- a/src/ulibs/umain.c
+++ b/src/ulibs/umain.c
@@ -7,10 +7,11 @@ int main(int argc, char *argv[]);
 
 static int
 initfd(int fd2, const char *path, uint32_t open_flags) {
-    int fd1, ret;
+    int fd1;
     if ((fd1 = open(path, open_flags)) < 0) {
         return fd1;
     }
+    int ret = fd1;
     if (fd1 != fd2) {
         close(fd2);
         ret = dup2(fd1, fd2);