add stepover command

[therealdreg]

Description

This pull request addresses a limitation in GDB where the "nexti" and "next" commands fail to step over call instructions properly. To resolve this, I have implemented a new feature, "stepover", specifically designed to enhance debugging efficiency by allowing users to step over call instructions seamlessly.

In my bootcamps, it’s a real pain to deal with this situation, especially for a GDB-newbie.

---

@Ordoviz

[Ordoviz](https://github.com/Ordoviz) [1 hour ago](https://github.com/hugsy/gef/pull/1086#discussion_r1571995694)
I also find this very useful for stepping over instructions with rep prefix.

---

POC

From a x86_64 Linux create a x32 binary

hello.asm

section .data
    msg db 'Hello, World!', 0xA
    len equ $ - msg
    testmsg db 'test!', 0xA
    lentestmsg equ $ - testmsg

section .text
    global _start

_start:
    mov eax, 4
    mov ebx, 1
    mov ecx, msg
    mov edx, len
    int 0x80
    call test

    mov eax, 1
    xor ebx, ebx
    int 0x80

test:
    mov eax, 4
    mov ebx, 1
    mov ecx, testmsg
    mov edx, lentestmsg
    int 0x80
    ret

Create executable:

nasm -f elf32 hello.asm -o hello.o
ld -m elf_i386 -s -o hello hello.o
strip hello

its impossible step over a call in GDB :(

The "nexti" and "next" commands do not work as expected :-(

---

btw, my Python skills are a bit rusty, sorry about that! xD

Checklist

• My code follows the code style of this project.
• My change includes a change to the documentation, if required.
• If my change adds new code, adequate tests have been added.
• I have read and agree to the CONTRIBUTING document.

[github-actions]

🤖 Coverage update for 164c242 🟢

	Old	New
Commit	d211658	164c242
Score	71.564%	71.564% (0)

[Ordoviz]

Pwndbg calls this command stepover (with alias so), which is a neater name IMO.

pwndbg> help so
usage: so [-h] [addr]

Breaks on the instruction after this one.

positional arguments:
  addr        The address to break after.

options:
  -h, --help  show this help message and exit

[github-actions]

🤖 Coverage update for 40baf1a 🟢

	Old	New
Commit	d211658	40baf1a
Score	71.564%	71.564% (0)

[github-actions]

🤖 Coverage update for 439b1e4 🟢

	Old	New
Commit	d211658	439b1e4
Score	71.564%	71.564% (0)

[github-actions]

🤖 Coverage update for 0b2a9c2 🟢

	Old	New
Commit	d211658	0b2a9c2
Score	71.564%	71.564% (0)

[github-actions]

🤖 Coverage update for fcd6265 🟢

	Old	New
Commit	d211658	fcd6265
Score	71.564%	71.564% (0)

[therealdreg]

@hugsy @Ordoviz @Grazfather opinion? looks good for you?

[hugsy]

Ok this is wild! This should be a GDB bug, is there any tracking of this?

[therealdreg]

Ok this is wild! This should be a GDB bug, is there any tracking of this?

@hugsy , I don't think it's a bug. I don't believe GDB is designed to step over instructions like call/rep/... for all architectures... Implementing this would be quite a hassle, right? This has been an issue for years.

Also, consider unusual cases like push + ret, polymorphic code, or if it's the last instruction in memory...

[github-actions]

🤖 Coverage update for ccb7424 🟢

	Old	New
Commit	399f457	ccb7424
Score	71.564%	71.564% (0)

[hugsy]

All looks good to me but the name, please change from so to stepover as a name everywhere, and make so an alias

Thanks!

[github-actions]

🤖 Coverage update for 3cbad34 🟢

	Old	New
Commit	399f457	3cbad34
Score	71.5776%	71.5776% (0)

[github-actions]

🤖 Coverage update for 310b71a 🟢

	Old	New
Commit	399f457	310b71a
Score	71.5776%	71.5776% (0)

[github-actions]

🤖 Coverage update for 060517f 🟢

	Old	New
Commit	399f457	060517f
Score	71.5776%	71.5776% (0)

[therealdreg]

All looks good to me but the name, please change from so to stepover as a name everywhere, and make so an alias

Thanks!

done! @hugsy

[hugsy]

LGTM now