Section: New Results

Compiler, vectorization, interpretation

Participants : Erven Rohou, Emmanuel Riou, Bharath Narasimha Swamy, Arjun Suresh, André Seznec, Nabil Hallou, Alain Ketterlin, Sylvain Collange.

Compilers for emerging throughput architectures

Participant : Sylvain Collange.

This work is done in collaboration with Fernando Pereira and his team at UFMG, Brasil.

GPU architectures present new challenges for compilers. Their performance characteristics demand SPMD programs with a high control-flow and memory regularity. Such architecture takes advantage of the regularity in programs to exploit data-level parallelism. In addition to the traditional challenges of code parallelization, new compilers for GPU and future throughput architectures face the task of improving the regularity of parallel programs. In particular, compiler analyses that identify control-flow divergence and memory divergence are a stepping stone for many optimizations. These optimizations include traditional code transformation such as loop interchange and tiling, which use divergence as an additional decision criterion, but also new optimizations specific to GPU architectures such as iteration delaying or branch fusion. In addition, the regularity parameter is an important aspect for workload characterization, as it provides a criterion for task scheduling in heterogeneous environments, such as multi-core processors with GPU. Our objectives include both accurate static and dynamic analyses for thread divergence, and the applications that it enables. We propose to combine static analyses with runtime checks, in order to get the best from both complementary approaches.

Improving sequential performance through memoization

Participants : Erven Rohou, André Seznec, Arjun Suresh.

Many applications perform repetitive computations, even when properly programmed and optimized. Performance can be improved by caching results of pure functions, and retrieving them instead of recomputing a result (a technique called memoization).

We proposed a simple technique for enabling software memoization of any dynamically linked pure function and we illustrate our framework using a set of computationally expensive pure functions – the transcendental functions.

Our technique does not need the availability of source code and thus can be applied even to commercial applications as well as applications with legacy codes. As far as users are concerned, enabling memoization is as simple as setting an environment variable.

Our framework does not make any specific assumptions about the underlying architecture or compiler tool-chains, and can work with a variety of current architectures.

We present experimental results for x86-64 platform using both gcc and icc compiler tool-chains, and for ARM cortex-A9 platform using gcc. Our experiments include a mix of real world programs and standard benchmark suites: SPEC and Splash2x. On standard benchmark applications that extensively call the transcendental functions we report memoization benefits of upto 16 %, while much higher gains were realized for programs that call the expensive Bessel functions. Memoization was also able to regain a performance loss of 76 % in bwaves due to a known performance bug in the gcc libm implementation of pow function.

Code Obfuscation

Participant : Erven Rohou.

This research is done in collaboration with the group of Prof. Ahmed El-Mahdy at E-JUST, Alexandria, Egypt.

A new obfuscation technique [27] based of decomposition of CFGs into threads has been proposed. We exploit the mainstream multi-core processing in these systems to substantially increase the complexity of programs, making reverse engineering more complicated. The novel method automatically partitions any serial thread into an arbitrary number of parallel threads, at the basic-block level. The method generates new control-flow graphs, preserving the blocks' serial successor relations and guaranteeing that one basic-block is active at a time through using guards. The method generates $m^n$ different combinations for $m$ threads and $n$ basic-blocks, significantly complicating the execution state. We also provide proof of correctness for the method.

We propose to leverage JIT compilation to make software tamper-proof. The idea is to constantly generate different versions of an application, even while it runs, to make reverse engineering hopeless. More precisely a JIT engine is used to generate new versions of a function each time it is invoked, applying different optimizations, heuristics and parameters to generate diverse binary code. A strong random number generator will guarantee that generated code is not reproducible, though the functionality is the same.

This work has been accepted for publication in January 2015 at the International Workshop on Dynamic Compilation Everywhere (DCE-2015).

Padrone

Participants : Erven Rohou, Alain Ketterlin, Emmanuel Riou.

The objective of the ADT PADRONE is to design and develop a platform for re-optimization of binary executables at run-time. Development is ongoing, and an early prototype is functional. In [30] , we described the infrastructure of Padrone, and showed that its profiling overhead is minimum. We illustrated its use through two examples. The first example shows how a user can easily write a tool to identify hotspots in their application, and how well they perform (for example, by computing the number of executed instructions per cycle). In the second example, we illustrate the replacement of a given function (typically a hotspot) by an optimized version, while the program runs.

We believe PADRONE fills an empty design point in the ecosystem of dynamic binary tools.

Dynamic Binary Re-vectorization

Participants : Erven Rohou, Nabil Hallou, Alain Ketterlin, Emmanuel Riou.

This work is done in collaboration with Philippe Clauss (Inria CAMUS).

Applications are often under-optimized for the hardware on which they run. Several reasons contribute to this unsatisfying situation, including the use of legacy code, commercial code distributed in binary form, or deployment on compute farms. In fact, backward compatibility of instruction sets guarantees only the functionality, not the best exploitation of the hardware. In particular SIMD instruction sets are always evolving.

We proposed a runtime re-vectorization platform that dynamically adapts applications to execution hardware. Programs distributed in binary forms are re-vectorized at runtime for the underlying execution hardware. Focusing on the x86 SIMD extensions, we are able to automatically convert loops vectorized for SSE into the more recent and powerful AVX. A lightweight mechanism leverages the sophisticated technology put in a static vectorizer and adjusts, at minimal cost, the width of vectorized loops. We achieve speedups in line with a native compiler targeting AVX. Our re-vectorizer is implemented inside a dynamic optimization platform; its usage is completely transparent to the user and requires neither access to source code nor rewriting binaries.