#memflow #introspection #memory #dma

memflow

core components of the memflow physical memory introspection framework

3 unstable releases

0.1.1 Sep 5, 2020
0.1.0 Sep 1, 2020
0.0.0 Jul 16, 2020

#33 in Memory management


Used in less than 7 crates

MIT license

300KB
6.5K SLoC

memflow

Crates.io build and test codecov MIT licensed Discord

physical memory introspection framework

memflow is a library that allows live memory introspection of running systems and their snapshots. Due to its modular approach it trivial to support almost any scenario where Direct Memory Access is available.

The very core of the library is a PhysicalMemory that provides direct memory access in an abstract environment. This object that can be defined both statically, and dynamically with the use of the inventory feature. If inventory is enabled, it is possible to dynamically load libraries that provide Direct Memory Access.

Through the use of OS abstraction layers, like memflow-win32, user can gain access to virtual memory of individual processes, by creating objects that implement VirtualMemory.

Bridging the two is done by a highly throughput optimized virtual address translation function, which allows for crazy fast memory transfers at scale.

The core is architecture independent (as long as addresses fit in 64-bits), and currently both 32, and 64-bit versions of the x86 family are available to be used.

For non-rust libraries, it is possible to use the FFI to interface with the library.

In the repository you can find various examples available (which use memflow-win32 layer)

Building from source

To build all projects in the memflow workspace:

cargo build --release --workspace

To build all examples:

cargo build --release --workspace --examples

Run all tests:

cargo test --workspace

Execute the benchmarks:

cargo bench

Documentation

Extensive code documentation can be found on docs.rs.

An additional getting started guide as well as higher level explanations of the inner workings of memflow can be found at memflow.github.io.

If u decide to build the latest documentation you can do it by issueing:

cargo doc --workspace --no-deps --open

Basic usage

  • run one of the examples with cargo run --release --example (pass nothing to get a list of them).
  • if ran with qemu_procfs connector, the runner will request root permissions to set 'CAP_SYS_PTRACE=ep' on the executables
  • or run the benchmarks cargo bench (can pass regex filters). Win32 benchmarks currently work only on Linux.

Running Examples

All examples support the memflow connector inventory system. You will have to install at least one connector to use the examples.

To install a connector just head over to the corresponding repository and install them via the install.sh script.

You will find a folder called memflow in any of the following locations:

/opt
/lib
/usr/lib/
/usr/local/lib
/lib32
/lib64
/usr/lib32
/usr/lib64
/usr/local/lib32
/usr/local/lib64

On Windows you can put the connector dll in a folder named memflow that is either in your current PATH or put it in C:\Users\{Username}\.local\lib\memflow.

Now you can just run the examples by providing the appropiate connector name:

Run memflow_win32/read_keys example with a procfs connector:

cargo run --example read_keys -- -vv -c qemu_procfs -a [vmname]

Run memflow_win32/read_bench example with a coredump connector:

cargo run --example read_bench --release -- -vv -c coredump -a coredump_win10_64bit.raw

Compilation support

target build tests benches compiles on stable
linux x86_64 ✔️ ✔️ ✔️ ✔️
mac x86_64 ✔️ ✔️ ✔️ ✔️
win x86_64 ✔️ ✔️ ✔️ ✔️
linux aarch64 ✔️ ✔️ ✔️ ✔️
no-std ✔️ ✔️ ✔️

Target support

memflow-win32 is tested on the latest Windows 10 versions all the way down to Windows NT 4.0. If you found a version that does not work please submit an issue with the major/minor version as well as the build number.

Roadmap / Future Development

  • Provide a rust native connector for PCILeech based hardware
  • Provide an UEFI Demo
  • Linux target support

Acknowledgements

  • CasualX for his wonderful pelite crate
  • ufrisk for his prior work on the subject and many inspirations

Contributing

Please check CONTRIBUTE.md

Dependencies

~1.4–2MB
~35K SLoC