LTTng

h1. Google Summer of Code 2013

h1. Project Ideas

h2. Babeltrace python bindings

*Brief explanation*:

Provide Python bindings to the Babeltrace APIs to provide a complete CTF reading/writting infrastructure to Python programmers.

Most of the public Babeltrace API is currently mapped using SWIG. However, a more “Pythonic” set of wrapper classes will make trace reading and writting more accessible than using the current native API. The production and interpretation of CTF traces in a high level language such as Python will ease the development of custom trace analysis solutions. It will also provide a convenient language in which to implement trace conversion plugins.

*Expected results*: TBD

*Complementary information*:

The Common Trace Format (CTF) specifies a trace format based on the requirements of the industry (through collaboration with the Multicore Association) and the Linux community.

Babeltrace is a trace conversion library that aims to provide a convenient to read and convert trace files between different formats. It currently supports CTF, which is the trace format outputted by the LTTng 2.0 tracers and provides the CTF reference implementation.

*Skill level*: Easy - Medium

*Prerequisite Knowledge*: Python, OO development

*Mentors*: Jérémie Galarneau, TBD

h2. Trace probe support in C++

*Brief explanation*:

The LTTng userspace tracing relies on the instrumentation of target applications with tracepoint probes. Currently, the only way to use tracepoints in C++ applications is by compiling a stand-alone C library linked to liblttng-ust which is then linked to the target C++ application. 

While this method is functional, it introduces significant development time overhead and confusion for new users. This project aims at making the UST tracepoint probe code compatible with popular C++ compilers to enable direct compilation of the providers with the application. The necessary modifications should not be too complex but are an excellent introductory project.

*Expected results*: TBD

*Complementary information*:

Issue #338

*Skill level*: Easy - Medium

*Prerequisite Knowledge*: C++ development, C development, build environments

*Mentors*: TBD

h2. Development of an strace-like wrapper

*Brief explanation*:

The LTTng kernel tracer should provide all the functionality required to emulate the behaviour of the strace tracer. The goal of this project is the development of an strace drop-in replacement based on the LTTng tracing toolchain that shall provide the user with the functionality of strace and the performance characteristics of the LTTng kernel tracer. This should ensure backward compatibility with tools based on the strace output format.

This should be implemented as a Babeltrace trace converter plug-in used in “live mode” -- an upcoming Babeltrace feature currently being developped by Julien Desfossez. The resulting plug-in will enable seamless interoperability between CTF and strace-based tools.

*Expected results*: TBD

*Complementary information*:

Issue #15

*Skill level*: Medium

*Prerequisite Knowledge*:  C development, scripting

*Mentors*: TBD

h2. Improvement of liblttng-ust-libc-wrapper''s coverage of libc''s functions

*Brief explanation*:

liblttng-ust-libc-wrapper is a library that currently provides a wrapper around libc to instrument malloc() and free() with UST tracepoint probes. This project aims to provide a full coverage of the standard libc functions.

This library is typically “LD_PRELOAD”-ed when debugging an application to override malloc() and free() symbols. The underlying libc symbol is then looked-up and called. The resulting trace can then be analyzed to identify memory allocation issues. Gaining a better libc coverage will provide LTTng users with even better debugging capabilities.

As an extension to this project, one could also develop a tool to automatically generate static tracepoints for arbitrary libraries from the header files.

*Expected results*: TBD

*Skill level*: Easy - Hard

*Prerequisite Knowledge*: C development, parser knowledge, scripting

*Mentors*: Christian Babeux, TBD

h2. Dynamic instrumentation support in UST

*Brief explanation*:

The current UST tracer relies on static tracepoint probes manually inserted in the traced application’s source code. This project aims at providing dynamic instrumentation capabilities in arbitrary applications.

The Dyninst project already provides a great API to dynamically modify code which may come in handy in this project. This is a great opportunity to initiate a collaboration between both projects.

*Expected results*: TBD

*Skill level*: Hard

*Prerequisite Knowledge*: C++ development, C development, Assembly

*Mentors*: TBD

h2. Android port

*Brief explanation*:

Port the LTTng infrastructure to the Android platform. The exact scope of this project will need to be reassessed based on the conclusions of the team currently working on this project as port of their undergrad project at École Polytechnique de Montréal.

This project presents a number of uncertain technical hurdles but some effort has already been made in most key areas.

The userspace part of the LTTng project relies on glibc features not present in Bionic -- Google''s libc implementation used on Android.

Glibc is now available on Android thanks to the efforts made by the cyborgstack project.

* System V IPC mechanisms are heavily relied-upon.

* The Ericsson port makes use of temporary files.

* Glibc on Android should provide such mechanisms.

* Userspace RCU must be ported and tested on the Android platform. Ported by Ericsson, not mainlined.

*Expected results*: TBD

*Complementary information*:

http://lists.lttng.org/pipermail/lttng-dev/2012-February/017626.html

*Skill level*: Hard

*Prerequisite Knowledge*: C development, kernel internals, Android development

*Mentors*: Christian Babeux, Jérémie Galarneau, TBD