Skip to main content

gdbrpc

🌐 Languages: English | 中文

A Python-based RPC (Remote Procedure Call) framework for GDB (GNU Debugger) that enables programmatic control and automation of debugging sessions.

Table of Contents


Overview

gdbrpc provides a client-server architecture that allows you to control GDB instances remotely through a simple Python API. It's designed to be framework-agnostic and can be used with any GDB-compatible debugger, not limited to any specific operating system or embedded platform.

Features

  • Remote GDB Control: Execute GDB commands remotely via socket communication
  • Bidirectional Communication: Client-server architecture with full duplex support
  • Command Serialization: Uses cloudpickle for robust serialization of Python objects
  • Interactive CLI: Built-in command-line interface for quick debugging sessions
  • Extensible: Easy to integrate into custom debugging workflows and automation scripts

Installation

From PyPI

pip install gdbrpc

From Source

cd gdbrpc
pip install -e .

Requirements

  • Python >= 3.10
  • GDB with Python support
  • cloudpickle >= 0.0.0

Quick Start

Starting the GDB Server

Within a GDB session, use the GDB commands (after importing gdbrpc):

(gdb) py import gdbrpc
(gdb) gdbrpc start
(gdb) gdbrpc start --port 20820 --host 0.0.0.0
(gdb) gdbrpc status
(gdb) gdbrpc stop

Using the Python Client

from gdbrpc import Client
from gdbrpc.utils import ShellExec

# Create and connect to the GDB server
client = Client(host="localhost", port=20819)
client.connect()

# Execute GDB commands
response = client.call(ShellExec("info threads"))
print(response)

# Get backtrace
bt = client.call(ShellExec("backtrace"))
print(bt)

# Evaluate expressions
result = client.call(ShellExec("print my_variable"))
print(result)

# Execute shell commands (prefix with !)
output = client.call(ShellExec("!ls -la"))
print(output)

# Close connection
client.disconnect()

Using the CLI

Load the current source checkout in GDB with the repository initializer:

source /path/to/gdbrpc/gdbinit.py
gdbrpc start --host 127.0.0.1 --port 20819

Run the module or installed gdbrpc command without an action to open the interactive client:

python3 -m gdbrpc --host localhost --port 20819
gdbrpc --host localhost --port 20819

For automation, repeat -c to send complete GDB command units in one RPC. Each unit may itself contain native multiline GDB syntax. A failed unit is reported, and the remaining units still run:

gdbrpc --port 20819 -c 'help' -c 'bt' -c 'info registers'
gdbrpc --port 20819 \
  -c $'define dump_state\n  bt\n  info registers\nend' \
  -c 'dump_state'

-f accepts either a local Python file or inline Python source. An existing file is read locally; otherwise the argument is compiled as source in GDB:

gdbrpc --port 20819 -f inspect.py
gdbrpc --port 20819 -f 'import gdb; print(gdb.newest_frame().name())'
gdbrpc --port 20819 -f $'import gdb\nprint(gdb.selected_thread())'

Add --wait when the script registers a GDB callback. The CLI remains connected until the script calls the injected emit(value) function once:

# wait_stop.py
import gdb


def on_stop(event):
    gdb.events.stop.disconnect(on_stop)
    emit(str(event))


gdb.events.stop.connect(on_stop)
gdbrpc --port 20819 -f wait_stop.py --wait

A waiting script must unregister its GDB callback before calling emit. One CLI invocation receives one emitted value; continuous event streaming is not supported.

TODO

  • make the CLI provide the same experience as the gdb CLI
    • auto-completion
    • command history reading
  • improve network transmission
    • improving security during deserialization

Architecture

Components

  • Server: Runs inside GDB process, listens for incoming connections
  • Client: Python client that connects to the server and sends commands
  • CLI: Interactive command-line interface built on top of the client
  • Protocol: Custom protocol for request/response communication using cloudpickle

Communication Flow

┌─────────────┐         Socket         ┌─────────────┐
│   Client    │◄──────────────────────►│   Server    │
│  (Python)   │    (Port 20819)        │  (In GDB)   │
└─────────────┘                        └─────────────┘
      │                                       │
      │ Send Request                          │
      │──────────────────────────────────────►│
      │                                       │ Execute Command
      │                                       │ in GDB Context
      │                            Response   │
      │◄──────────────────────────────────────│
      │                                       │

API Reference

Client

Client(host="localhost", port=20819, logLevel=logging.INFO)

Create a new client instance.

Parameters:

  • host (str): Server hostname or IP address (default: "localhost")
  • port (int): Server port number (default: 20819)
  • logLevel (int): Logging level (default: logging.INFO)

connect() -> bool

Establish connection to the GDB server.

Returns: True if connection successful, False otherwise

call(request: Request, post_request: Optional[PostRequest] = None, timeout: float = 300) -> Any

Send a request to the GDB server and receive response.

Parameters:

  • request (Request): Request object to send (typically ShellExec for executing commands)
  • post_request (Optional[PostRequest]): Optional callback request for async handling
  • timeout (float): Request timeout in seconds (default: 300)

Returns: Response payload from the server

Example:

from gdbrpc import Client
from gdbrpc.utils import ShellExec

client = Client("localhost", 20819)
client.connect()

# Execute GDB command
result = client.call(ShellExec("info threads"))
print(result)

# Execute shell command (prefix with !)
result = client.call(ShellExec("!ls -la"))
print(result)

disconnect()

Close the connection to the server and cleanup resources.

Request Classes

ShellExec(command: str)

Request to execute a GDB command or shell command on the server.

Parameters:

  • command (str): Command to execute
    • GDB commands: "info threads", "backtrace", "print variable"
    • Shell commands: prefix with ! or shell, e.g., "!ls" or "shell pwd"

Example:

from gdbrpc.utils import ShellExec

# GDB command
gdb_request = ShellExec("backtrace full")

# Shell command
shell_request = ShellExec("!cat /proc/meminfo")

Request

Base class for all request types. Custom requests can be created by subclassing.

Methods:

  • __init__(): Initializes request with unique tag ID
  • __call__(*args, **kwargs): Must be implemented by subclasses

PostRequest

Base class for requests with callbacks. Used for asynchronous request handling.

Methods:

  • __init__(): Initializes with finish event
  • __call__(argument: Any): Must be implemented by subclasses
  • finish (threading.Event): Event to signal completion

Configuration

Server Configuration

The server can be configured when starting:

import logging
import gdbrpc

# Start with debug logging
gdbrpc.start_gdb_socket_server(
    host="0.0.0.0",  # Listen on all interfaces
    port=20819,
    logLevel=logging.DEBUG
)

Client Configuration

import logging
from gdbrpc import Client
from gdbrpc.utils import ShellExec

# Create client with custom log level
client = Client(
    host="localhost",
    port=20819,
    logLevel=logging.DEBUG  # Enable debug logging
)
client.connect()

# Custom timeout for specific requests
result = client.call(
    ShellExec("interrupt"),
    timeout=60  # Wait up to 60 seconds for this command
)

Troubleshooting

Server won't start

  • Ensure GDB has Python support: gdb --configuration | grep python
  • Check if port is already in use: netstat -an | grep 20819
  • Verify firewall settings allow the connection

Connection refused

  • Verify server is running: (gdb) gdbrpc status
  • Check host/port configuration matches between client and server
  • Ensure network connectivity between client and server

Command execution fails

  • Verify GDB is in correct state (e.g., program loaded, running)
  • Check command syntax is correct for your GDB version
  • Review server logs for detailed error messages
  • Make sure python version GDB uses is same as client

Contributing

Contributions are welcome! Please feel free to submit pull requests or open issues for bugs and feature requests.

License

Licensed under the Apache License, Version 2.0. See the LICENSE file for details.

Support

For questions and support:

  • Open an issue on the project repository
  • Consult the GDB Python API documentation
  • Review the examples in the examples/ directory

Metadata

Release files for gdbrpc 0.7.0

For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.

Source distribution (sdist)

Source distribution for gdbrpc 0.7.0
File Size Uploaded
gdbrpc-0.7.0.tar.gz 34.7 kB Details

Built distribution (wheel)

Table of built distributions (wheels) for gdbrpc 0.7.0
File Interpreter ABI Platform
gdbrpc-0.7.0-py3-none-any.whl Python 3 none any Details

Total release size: 70.8 kB

Release files / gdbrpc-0.7.0.tar.gz

Download URL gdbrpc-0.7.0.tar.gz
Size 34.7 kB
Tags Source
SHA-256 checksum
How to use checksums
62e796e5977e9d8473c50ede7462491e5e747b8896f985686106a69a55663fd3
BLAKE2b-256 checksum
How to use checksums
47661d81ada564654b16f60a04f58e980a19e601354cf04ac5590d9d01338f02
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
No
Uploaded via twine/7.0.0 CPython/3.12.14

Release files / gdbrpc-0.7.0-py3-none-any.whl

Download URL gdbrpc-0.7.0-py3-none-any.whl
Size 36.1 kB
Tags Python 3
SHA-256 checksum
How to use checksums
397484cb77321c504625b4b59d9b4fc87be02c779097bcb2bd8da40f0e32588b
BLAKE2b-256 checksum
How to use checksums
912567d2cdf76aaac4aa533e1bce114313f5fa34c5e693cfeab07020fbc957bc
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
No
Uploaded via twine/7.0.0 CPython/3.12.14

Release history Release notifications | RSS feed

This release

0.7.0 This release

2 release files

0.6.0

2 release files

0.5.2

2 release files

0.5.1

2 release files

0.5.0

2 release files

0.1.0

2 release files

0.0.2

2 release files

Anthropic, PBC Visionary sponsor Bloomberg Visionary sponsor Hudson River Trading Visionary sponsor Meta Visionary sponsor NVIDIA Visionary sponsor Microsoft Sustainability sponsor Depot Continuous Integration AWS Cloud computing and Security Sponsor Datadog Monitoring Fastly CDN Google Download Analytics Sentry Error logging StatusPage Status page