Data size optimizations for java programs

Ananian, C. Scott; Rinard, Martin

doi:10.1145/780731.780741

Cited by 10 publications

(12 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In contrast, in our work locality is irrelevant: the MCUs that we target have flat RAM with a uniform access time. Ananian and Rinard [2003] perform static bitwidth analysis and field packing for Java objects, with the goal of reducing memory usage. Zhang and Gupta [2006] use memory profile data to find limited-bitwidth heap data that can be packed into less space.…”

Section: Related Workmentioning

confidence: 99%

Eliminating the call stack to save RAM

2009

View full text Add to dashboard Cite

Most programming languages support a call stack in the programming model and also in the runtime system. We show that for applications targeting low-power embedded microcontrollers (MCUs), RAM usage can be significantly decreased by partially or completely eliminating the runtime callstack. We present flattening, a transformation that absorbs a function into its caller, replacing function invocations and returns with jumps. Unlike inlining, flattening does not duplicate the bodies of functions that have multiple callsites. Applied aggressively, flattening results in stack elimination. Flattening is most useful in conjunction with a lifting transformation that moves global variables into a local scope.Flattening and lifting can save RAM. However, even more benefit can be obtained by adapting the compiler to cope with properties of flattened code. First, we show that flattening adds false paths that confuse a standard live variables analysis. The resulting problems can be mitigated by breaking spurious live-range conflicts between variables using information from the unflattened callgraph. Second, we show that the impact of high register pressure due to flattened and lifted code, and consequent spills out of the register allocator, can be mitigated by improving a compiler's stack layout optimizations. We have implemented both of these improvements in GCC, and have implemented flattening and lifting as source-tosource transformations. On a collection of applications for the AVR family of 8-bit MCUs, we show that total RAM usage can be reduced by 20% by compiling flattened and lifted programs with our improved GCC.

show abstract

Section: Related Workmentioning

confidence: 99%

Eliminating the call stack to save RAM

2009

View full text Add to dashboard Cite

show abstract

“…For example, Ananian and Rinard [2] use static analysis of Java programs to eliminate fields with constant values and reduce the sizes of fields that can assume a small number of values. Cooprider and Regehr [7] use static analysis of C programs to pack scalars, pointers, structures, and arrays using a compression-table scheme.…”

Section: Related Workmentioning

confidence: 99%

Vertical object layout and compression for fixed heaps

Titzer

Palsberg

2007

Proceedings of the 2007 International Conference on Compilers, Architecture, and Synthesis for Embedded Systems

View full text Add to dashboard Cite

Research into embedded sensor networks has placed increased focus on the problem of developing reliable and flexible software for microcontroller-class devices. Languages such as nesC [8] and Virgil [14] have brought higher-level programming idioms to this lowest layer of software, thereby adding expressiveness. Both languages are marked by the absence of dynamic memory allocation, which removes the need for a runtime system to manage memory. To provide data structures, nesC offers modules, and Virgil offers the application an opportunity to allocate and initialize objects during compilation. This paper explores techniques for compressing fixed object heaps with the goal of reducing the RAM footprint of a program. We explore table-based compression and introduce a novel form of object layout called vertical object layout. We provide experimental results that measure the impact on RAM size, code size, and execution time for a set of Virgil programs. Our results show that compressed vertical layout has better execution time and code size than tablebased compression while achieving more than 20% heap reduction on 6 of 12 benchmark programs.

show abstract

“…This is of importance for embedded real-time systems because it can reduce the amount of memory required to execute the program and therefore reduce the cost of the embedded real-time system. See [19] for more information.…”

Section: Data Size Reductionsmentioning

confidence: 99%

“…Moreover, Flex has been shown to be useful for processing and analyzing components in the context of the BBN OEP. Flex also served as the platform for much of the research performed as part of this contract, see, for example [31,20].…”

Section: Flex and Componentsmentioning

confidence: 99%

Component Composition for Embedded Systems Using Semantic Aspect-Oriented Programming

Rinard¹

2004

Self Cite

View full text Add to dashboard Cite

The public reporting burden for this collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources, searching existing data sources, gathering and maintaining the data needed, and completing and reviewing the collection of information. Send comments regarding this burden estimate or any other aspect of this collection of information, including suggestions for reducing this burden, to Department of Defense, Washington Headquarters Services, Directorate for Information SPONSORING/MONITORING AGENCY REPORT NUMBER(S) AFRL-IF-WP-TR-2004-1568 DISTRIBUTION/AVAILABILITY STATEMENTApproved for public release; distribution is unlimited. SUPPLEMENTARY NOTES ABSTRACTThe goal of our research was to develop technologies and techniques in support of real-time systems for the defense community. Our research focused on Real-Time Java implementation and analysis techniques. Real-Time Java is important for the defense community because it holds out the promise of enabling developers to apply COTS Java technology to specialized military embedded systems. It also promises to allow the defense community to utilize a large Java-literate workforce for building defense systems.Our research has delivered several techniques that may make Real-Time Java a better platform for developing embedded systems. These techniques include ways to implement scoped memories (a key Real-Time Java construct) without the possibility of introducing unexpected and potentially catastrophic delays in the execution of real-time threads, analyses that ensure the correct use of Real-Time Java scoped memories, analyses that compute how much memory is required to execute a given Real-Time Java program (potentially helping developers calculate how much memory must be including in a given system to ensure that the system will execute without running out of memory), and optimizations that reduce the amount of memory required to execute a Real-Time Java program.

show abstract

Data size optimizations for java programs

Cited by 10 publications

References 21 publications

Eliminating the call stack to save RAM

Eliminating the call stack to save RAM

Vertical object layout and compression for fixed heaps

Component Composition for Embedded Systems Using Semantic Aspect-Oriented Programming

Contact Info

Product

Resources

About