Disconnect3d

@OBarronCS would u perhaps want to make a review of this PR? :)

@dmur1 RE: non-existent stack frame - run `gdb --quiet /bin/ls` and then: ``` pwndbg> set context-sections '' ... pwndbg> entry ... pwndbg> stack 00:0000│ r13 rsp ret0 0x7fffffffe370 ◂— 0x1...

@dmur1 ping about rebase/conflicts

This is a bit stale and has lots of conflicts. Closing for now.

Here is a patch that fixes this: ```py diff --git a/pwndbg/gdblib/events.py b/pwndbg/gdblib/events.py index 1c18221..3abdd8c 100644 --- a/pwndbg/gdblib/events.py +++ b/pwndbg/gdblib/events.py @@ -220,15 +220,29 @@ gdb.events.new_objfile.connect(log_objfiles) def after_reload(start=True) -> None: + """...

This eventually means that all global variables that were there before reload, are still there afterwards.

I can't reproduce the reload command failing. However, the global vars are not reloaded. I tested this by adding a global var to `pwndbg/__init__.py` called `testvar`: An interesting point here:...

@CptGibbon @lebr0nli pls check this out & help with questions/reviews!

> * Does the way the current implementation detects accesses to freed memory look reasonable? Are there any big pitfalls to it that I've missed? Clearing cache: we probably want...

Maybe the command should be `heap-trace` and it should enable all heap tracing that we will implement and should have a `[--no-uaf-detection]` and some other flags?