M. Hirzel and T. Chilimbi. Bursty tracing: A framework for low-overhead temporal profiling. In Fourth ACM Workshop on Feedback-Directed and Dynamic Optimization (FDDO4) , 2001.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....for a bounded amount of time. Although the implementation and evaluation in [9] focused on collecting profiles along acyclic paths in the duplicated code, the paper also suggested variants of this approach, such as staying in the duplicated code for multiple loop iterations. Hirzel and Chilimbi [29] implemented and evaluated this variation in the context of x86 binaries, calling it Bursty Tracing . In their system the instrumented method is modified so that upon taking a sample, execution remains in the duplicated code until a fixed number of checks are executed. This approach enables ....

Martin Hirzel and Trishul Chilimbi. Bursty tracing: A framework for low-overhead temporal profiling. In 4th ACM Workshop on Feedback-Directed and Dynamic Optimization (FDDO-4), pages 117--126, December 2001.

....checks executed is to perform analysis of the instrumentation that is inserted in the duplicated code, and remove checks that are unnecessary. For example, if a particular loop contains no instrumentation then no backedge check is needed for that loop. This approach was used by Hirzel and Chilimbi [48] to reduce the checking overhead when collecting memory reference profiles. 3.3 Trigger mechanisms The sampling framework relies on runtime checks which need some kind of trigger mechanism to determine when execution should be transferred from the checking code to the duplicated code. There is a ....

....in the duplicated code, rather than executing a single acyclic path. The technique of remaining in the duplicated code for N checks can be used to collect 28 interprocedural profiles as well. Hirzel Chilimbi used this approach to collect interprocedural memory reference profiles for C programs [48]; their work is described in detail in Section 7. Modifications to the method entry check (see [48] can ensure that execution remains in the duplicated code if a method call is made. Unfortunately the same cannot be said about method returns; if a method returns while the duplicated code is ....

[Article contains additional citation context not shown here]

Martin Hirzel and Trishul Chilimbi. Bursty tracing: A framework for low-overhead temporal profiling. In 4th ACM Workshop on Feedback-Directed and Dynamic Optimization (FDDO-4), pages 117--126, December 2001.

....forward, we cross only a finite number of instrumentation sites before reaching the next threshold check. Thus, we can compute a finite weight for each threshold check. There is some flexibility regarding exactly where a threshold check is placed, but computing an optimal solution is NP hard [18]. For simplicity, our present system places one threshold check at function entry and one along each loop back edge. Weights may be computed in a single bottomup traversal of each function s control flow graph. If any threshold check is found to have zero weight, it is simply discarded. We ....

M. Hirzel and T. Chilimbi. Bursty tracing: A framework for low-overhead temporal profiling. In 4th ACM Workshop on Feedback-Directed and Dynamic Optimization, Austin, Texas, Dec. 1 2001.

....(2) a store profile, i.e. the set of memory addresses that were written in the profiled interval. Thanks to recent advances in sampling based profiling, these profiles can be collected with high accuracy, yet with overheads that are sufficiently low for dynamic optimizers (well below 5 slowdown) [3, 10, 15, 25, 26, 30]. We turn a constant propagator into a specializer by letting it consult the two profiles, essentially as follows: given a program point and an object that is frequently referenced at point , we perform standard constant propagation starting from , with two modifications: 1) we assume ....

Martin Hirzel and Trishul Chilimbi. Bursty tracing: A framework for low-overhead temporal profiling.In 4th ACM Workshop on Feedback-Directed and Dynamic Optimization (FDDO-4), December 2001.

No context found.

Martin Hirzel and Trishul M. Chilimbi. Bursty tracing: a framework for lowoverhead temporal profiling. In Feedback-Directed and Dynamic Optimizations (FDDO), 2001.

No context found.

M. Hirzel and T. Chilimbi. Bursty tracing: A framework for low-overhead temporal profiling. In Fourth ACM Workshop on Feedback-Directed and Dynamic Optimization (FDDO4) , 2001.

No context found.

Martin Hirzel and Trishul M. Chilimbi. Bursty tracing: A framework for low-overhead temporal profiling. In Workshop on Feedback-Directed and Dynamic Optimizations (FDDO), 2001.

No context found.

M. Hirzel and T. Chilimbi. Bursty Tracing: A Framework for Low-Overhead Temporal Profiling. In 4th ACM Workshop on Feedback-Directed and Dynamic Optimization (FDDO-4), Dec. 2001.

No context found.

M. Hirzel and T. Chilimbi, "Bursty tracing: A framework for low-overhead temporal profiling," in 4th ACM Workshop on Feedback-Directed and Dynamic Optimization (FDDO-4), Dec. 2001, pp. 117--126.

No context found.

M. Hirzel and T. M. Chilimbi. Bursty tracing: A framework for low-overhead temporal pro ling. In Proceedings of ACM Workshop on Feedback-Directed and Dynamic Optimization, Dallas, Texas, 2001.

No context found.

M. Hirzel and T. Chilimbi. Bursty tracing: A framework for low-overhead temporal profiling. In Proc. Fourth Workshop on Feedback-Directed and Dynamic Optimization (FDDO-4), December 2001.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC