Gdb-plus

* Check libc.address and elf.address correctly aligned.
* allow Debugger(silent=True) to hide pwntools default logs
* Execute from_entry only if inside the loader

GDB+ 7.1.0 QEMU support
* Find libc used by process under QEMU
* Warn users for missing qemu or gdb-multiarch
* support for arm 32 architecture

GDB+ 7.0.1: catchpoints
* support gdb catchpoints
* correct implementation of load_libc
* allow different register calling conventions for call()
* try to trace function calls
* breakpoint callbacks are executed in LIFO order
* gdbserver starts from entry point by default
* dbg.elf_address and dbg.libc_address warn user if the exploit sets them to the wrong value

GDB+ 7.0.0: Fast GDB
* increased speed up to 10x
* removed the logs to debug the library

* hotfix migrate to gdb while emulate ptrace is on
* Args support slices
* Update heap analysis functions for pwndbg
* Support waitpid(-1) before process traced
* Allow to setup ptrace_emulation without call to ptrace
* make Debugger.canary writeable
* Debugger(binary=...) support ELF object too
* support direct connection to gdbserver

* handle NOPTRACE for emulate_ptrace and split_on_fork
* NOP instructions
* hotfix disable lock.log with NOPTRACE
* hotfix Removed buffer overflow in `__convert_args`
* Assert that malloc is present before trying to allocate on heap
* Option to connect to remote server while debugging locally
* Migrate to logging library
* read pointers based on context
* context manager

GDB+ 6.4.1: syscalls
* Set handlers for syscalls
* Emulate ptrace through syscalls instead of function calls
* Support signals with libdebug
* Return self when possible to simplify debugging from Ipython
* Remove advanced_continue_and_wait_split

GDB+ 6.4.0: **ARM**
* Can "partially" debug aarch64 programs running under qemu. With all the limitations that come from the emulator
- For now qemu has problems identifying the pid of the process and when it forks, so you can only debug single-thread, single-process applications for now.
- you can not reattach to a process, so no `debug_from`
- You can not find the base of your binary or overwrite the code so no patches nor syscalls
* hardware breakpoints

GDB+ 6.3.4: Bruteforce util
* single thread bruteforce
* multithread bruteforce
* support new version of libdebug

GDB+ 6.3.3: Expand ptrace emulation
* single function for ptrace emulation that handles both tracee and tracer.
- process isn't limited at being a slave or a master
* signal handler

GDB+ 6.3.2: hotfix ptrace emulation
* standardise how inferiors are handled
- inferiors can be specified only to read and write memory. Split the debugger or use switch_inferior if you have to perform other actions.
* util functions to read and write ints, longs, and strings
- write functions take an pointer and a value or list of values. If pointer is None the address is allocated in the heap (or in the bss if heap=False). The function returns a pointer to the data
- by default read functions return a single value. If the argument n > 1 the function returns an array
- the null bytes for the strings are added when writing and removed when reading
* set status for waitpid while emulating ptrace
* Define gdbscript to execute when splitting a new Debugger
* inner_debugger write and read from /proc/pid/mem
* util step_until_call

GDB+ 6.3.1: hotfix libdebug
* Make libdebug optional since it can't be added as a dependency (thanks PyPI)

GDB+ 6.3.0: **Let's go faster**
* Can use gdb or libdebug ad debugger
- gdb can be used manually, but will be slower to script
- libdebug has less features and more bugs