/symbolic

Stack trace symbolication library written in Rust

Primary LanguageRustMIT LicenseMIT

Symbolic

Symbolic is a library written in Rust which is used at Sentry to implement symbolication of native stack traces, sourcemap handling for minified JavaScript and more. It consists of multiple largely independent crates which are bundled together into a C and Python library so it can be used independently of Rust.

What's in the package

Symbolic provides the following functionality:

  • Symbolication based on custom cache files (symcache)
  • Symbol cache file generators from:
    • Mach, ELF and PE symbol tables
    • Mach and ELF embedded DWARF data
    • PDB CodeView debug information
    • Breakpad symbol files
  • Demangling support
    • C++ (GCC, clang and MSVC)
    • Objective C / Objective C++
    • Rust
    • Swift
  • JavaScript sourcemap expansion
    • Basic token mapping
    • Heuristics to find original function names based on minified sources
    • Indexed sourcemap to sourcemap merging
  • Proguard function mappings
  • Minidump / Breakpad processing
    • Generate Breakpad symbol files from Mach, ELF and PDBs
    • Process Minidumps to retrieve stack traces
  • Convenient C and Python library
  • Processing of Unreal Engine 4 native crash reports
    • Extract and process minidumps
    • Expose logs and UE4 context information

Rust Usage

Add symbolic as a dependency to your Cargo.toml. You will most likely want to activate some of the features:

  • debuginfo (default): Contains support for various object file formats and debugging information. Currently, this comprises MachO and ELF (with DWARF debugging), PE and PDB, as well as Breakpad symbols.
  • demangle: Demangling for Rust, C++, Swift and Objective C symbols. This feature requires a C++11 compiler on the PATH.
  • minidump: Rust bindings for the Breakpad Minidump processor. Additionally, this includes facilities to extract stack unwinding information (sometimes called CFI) from object files. This feature requires a C++11 compiler on the PATH.
  • proguard: Processing of Proguard mapping files to look up mangled Java function paths.
  • sourcemap: Processing and expansion of JavaScript source maps, as well as lookups for minified function names.
  • symcache: An optimized, platform-independent storage for common debugging information. This allows blazing fast symbolication of instruction addresses to function names and file locations.
  • unreal: Processing of Unreal Engine 4 crash reports.

There are also alternate versions for some of the above features that additionally add implementations for serde::{Deserialize, Serialize} on suitable types:

  • common-serde
  • debuginfo-serde
  • minidump-serde
  • unreal-serde

Python Usage

Symbolic is hosted on PyPI. It comes as a library with prebuilt wheels for linux and macOS. On other operating systems or when using as rust library, you need to build symbolic manually. It should be compatible with both Python 2 and Python 3.

The python library ships all of the above features in a flat module:

from symbolic import Archive

fat = Archive.open('/path/to/object')
obj = fat.get_object(arch = 'x86_64')
print 'object debug id: {}' % obj.debug_id

C Bindings

Symbolic also offers C bindings, which allow for FFI into arbitrary languages. Have a look at the the Symbolic C-ABI readme for more information.

Source Crates

A lot of functionality exposed by this library come from independent Rust crates for better use:

Additionally we use the following C++ libraries to fill in gaps:

Building and Development

To build the Rust crate, we require the latest stable Rust, as well as a C++11 compiler. The crate is split into a workspace with multiple features, so when running building or running tests always make sure to pass the --all and --all-features flags.

# Check whether the crate compiles
cargo check --all --all-features

# Run Rust tests
cargo test --all --all-features

We use rustfmt and clippy from the latest stable channel for code formatting and linting. To make sure that these tools are set up correctly and running with the right configuration, use the following make targets:

# Format the entire codebase
make format

# Run clippy on the entire codebase
make lint

Most likely, new functionality also needs to be added to the Python package. This first requires to expose new functions in the C ABI. For this, refer to the Symbolic C-ABI readme.

We highly recommend to develop and test the python package in a virtual environment. Once the ABI has been updated and tested, ensure the virtualenv is active and install the package, which builds the native library. There are two ways to install this:

# Install the release build, recommended:
pip install --editable ./py

# Install the debug build, faster installation but much slower runtime:
SYMBOLIC_DEBUG=1 pip install --editable ./py

For testing, we use ubiquitous pytest. Again, ensure that your virtualenv is active and the latest version of the native library has been installed. Then, run:

# Run tests manually
pytest ./py/tests

# Creates a new virtualenv, installs the release build and runs tests:
make pytest

License

Symbolic is licensed under the MIT license. It uses some Apache2 licensed code from Apple for the Swift demangling.