Memory Allocator - Under The Hood

Memory Allocator in Different OS

Linux

ptmalloc2

FreeBSD

jemalloc

System Call for Malloc

Testing

testing program :

// test_malloc.c

#include <stdio.h>
#include <stdlib.h>

int main() {
    puts("===========");
    free(malloc(sizeof(int)));
    puts("===========");
    return 0;
}

Linux

compile, execute, trace system call :

$ gcc test_malloc.c -o test_malloc
$ strace ./test_malloc  # Linux

在 trace system call 時, 這邊觀察到兩個 system call, brk(0)brk(0x14e5000) (數值可能會不一樣), 所以我們可以確定的是, 在 Linux 上目前去 call malloc 時, 會用到 brk 這個 system call

FreeBSD

compile, execute, trace system call :

$ clang test_malloc.c -o test_malloc
$ truss ./test_malloc   # FreeBSD

在 trace system call 時, 這邊只觀察到一個 system call, 可以發現使用的 system call 是 madvise , 和 Linux 上不同 (Linux 上是 brk)。

madvise(0x2c07000,0x1000,0x5,0xaaaaaaaaaaaaaaab,0x2c00048,0x24c4be0) , 這是用於 free

Explain

所以到底這些 system call (brk, madvise) 都做了些什麼勒? 為什麼 call 完後 heap 就會多出空間可以使用?

我們先來看看 Linux 上的 Memory 都放了些什麼東西 :

Linux Address Layout

sbrk

sbrk 是 glibc 的 brk wrapper

glibc

glibc/misc/sbrk.c :

/* Extend the process's data space by INCREMENT.
   If INCREMENT is negative, shrink data space by - INCREMENT.
   Return start of new space allocated, or -1 for errors.  */
void *
__sbrk (intptr_t increment)
{
  void *oldbrk;

  /* If this is not part of the dynamic library or the library is used
     via dynamic loading in a statically linked program update
     __curbrk from the kernel's brk value.  That way two separate
     instances of __brk and __sbrk can share the heap, returning
     interleaved pieces of it.  */
  if (__curbrk == NULL || __libc_multiple_libcs)
    if (__brk (0) < 0)    /* Initialize the break.  */
      return (void *) -1;

  if (increment == 0)
    return __curbrk;

  oldbrk = __curbrk;
  if ((increment > 0
       ? ((uintptr_t) oldbrk + (uintptr_t) increment < (uintptr_t) oldbrk)
       : ((uintptr_t) oldbrk < (uintptr_t) -increment))
      || __brk (oldbrk + increment) < 0)
    return (void *) -1;

  return oldbrk;
}

glibc/sysdeps/unix/sysv/linux/generic/brk.c

int
__brk (void *addr)
{
  INTERNAL_SYSCALL_DECL (err);

  __curbrk = (void *) INTERNAL_SYSCALL (brk, err, 1, addr);
  if (__curbrk < addr)
    {
      __set_errno (ENOMEM);
      return -1;
    }

  return 0;
}

glibc/sysdeps/unix/sysv/linux/x86_64/sysdep.h

# define INTERNAL_SYSCALL_NCS(name, err, nr, args...)         \
({                                                            \
  unsigned long int resultvar;                                \
  LOAD_ARGS_##nr (args)                                       \
  LOAD_REGS_##nr                                              \
  asm volatile (                                              \
  "syscall\n\t"                                               \
  : "=a" (resultvar)                                          \
  : "0" (name) ASM_ARGS_##nr : "memory", "cc", "r11", "cx");  \
  (long int) resultvar; })

# define INTERNAL_SYSCALL(name, err, nr, args...) \
  INTERNAL_SYSCALL_NCS (__NR_##name, err, nr, ##args)

C - Inline Assembly

關鍵字 : asm

asm("movl %ecx %eax");      /* moves the contents of ecx to eax */
__asm__("movb %bh (%eax)"); /*moves the byte from bh to the memory pointed by eax */

asm__asm__ 都可以用,當 asm 在程式中會 conflict 時就可以使用 __asm__ , 如果 assembly code 有很多行的話,每行結尾加上 “nt”, 如此一來在把 assembly code 字串丟給後面的 as (ex: GAS (GNU Assembler)) 處理時, 格式才會是正確的。

asm ("movl %eax, %ebx\n\t"
     "movl $56, %esi\n\t"
     "movl %ecx, $label(%edx,%ebx,$4)\n\t"
     "movb %ah, (%ebx)");

Extended Asm :

asm ( assembler template
   : output operands                  /* optional */
   : input operands                   /* optional */
   : list of clobbered registers      /* optional */
   );

asm volatile (“syscallnt” : “=a” (result) : “0” (SYS_BRK) : “memory”, “cc”, “r11”, “cx”);


GCC 用 AT&T/UNIX assembly syntax

Source-Destination Ordering

“Op-code src dst”

Register Naming

prefix by %, %eax

Immediate Operand

Operand Size

Memory Operands

Intel Code

AT&T Code

mov eax,1 mov ebx,0ffh int 80h mov ebx, eax mov eax,[ecx] mov eax,[ebx+3] mov eax,[ebx+20h] add eax,[ebx+ecx*2h] lea eax,[ebx+ecx] sub eax,[ebx+ecx*4h-20h]

movl $1,%eax movl $0xff,%ebx int $0x80 movl %eax, %ebx movl (%ecx),%eax movl 3(%ebx),%eax movl 0x20(%ebx),%eax addl (%ebx,%ecx,0x2),%eax leal (%ebx,%ecx),%eax subl -0x20(%ebx,%ecx,0x4),%eax

Misc

Reference