PLRU Cache Domino Effects

Berg, Christoph

doi:10.4230/oasics.wcet.2006.672

Cited by 14 publications

(3 citation statements)

References 3 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The presence of timing anomalies [29], [30] can often be traced back to the non-monotonicity of a system's dynamics. In contrast to LRU, other cache replacement policies such as FIFO, NMRU, and PLRU, do not behave monotonically, and have been found to exhibit timing anomalies [31], [32]. For that reason it seems unlikely that our analysis approach can be extended to such policies.…”

Section: Related Workmentioning

confidence: 95%

Cache Persistence Analysis: Finally Exact

Stock

Hahn

Reineke

2019

2019 IEEE Real-Time Systems Symposium (RTSS)

View full text Add to dashboard Cite

Cache persistence analysis is an important part of worst-case execution time (WCET) analysis. It has been extensively studied in the past twenty years. Despite these efforts, all existing persistence analyses are approximative in the sense that they are not guaranteed to find all persistent memory blocks.In this paper, we close this gap by introducing the first exact persistence analysis for caches with least-recently-used (LRU) replacement. To this end, we first introduce an exact abstraction that exploits monotonicity properties of LRU to significantly reduce the information the analysis needs to maintain for exact persistence classifications. We show how to efficiently implement this abstraction using zero-suppressed binary decision diagrams (ZDDs) and introduce novel techniques to deal with uncertainty that arises during the analysis of data caches.The experimental evaluation demonstrates that the new exact analysis is competitive with state-of-the-art inexact analyses in terms of both memory consumption and analysis run time, which is somewhat surprising as we show that persistence analysis is NP-complete. We also observe that while prior analyses are not exact in theory they come close to being exact in practice.

show abstract

Section: Related Workmentioning

confidence: 95%

Cache Persistence Analysis: Finally Exact

Stock

Hahn

Reineke

2019

2019 IEEE Real-Time Systems Symposium (RTSS)

View full text Add to dashboard Cite

show abstract

“…For instance, a PLRU cache may, with a specifically concocted cache access pattern, indefinitely retain some data that was used only once at the beginning of execution and is never accessed again [10], which of course cannot happen with LRU. This results in domino effects: the cache behavior of a loop body may be indefinitely affected by the cache contents before the loop [2]. All of this makes the static analysis of programs over PLRU caches difficult; there are no known precise and fast analyses for this policy and for NMRU.…”

Section: Presentation Of Other Policiesmentioning

confidence: 99%

“…Static analysis tools used to compute bounds on worst-case execution time 1 thus include a static analysis for caches, meant to predict which accesses are always cache hits (data in cache) and which are always misses (data not in cache). 2 For almost all policies (including LRU, FIFO, PLRU and NMRU) found in processors, analysis for exist-hit and exist-miss properties may be performed on each cache set separately, by slicing the program according to the cache set (Figure 1), without loss of precision (the only policy for which this is false is pseudo-round-robin, which we do not consider here) [18]. 3 We shall thus consider in this paper, without loss of generality, that the cache consists of a single cache set.…”

Section: Introductionmentioning

confidence: 99%

The complexity gap in the static analysis of cache accesses grows if procedure calls are added

Monniaux¹

2022

Preprint

View full text Add to dashboard Cite

The static analysis of cache accesses consists in correctly predicting which accesses are hits or misses. While there exist good exact and approximate analyses for caches implementing the least recently used (LRU) replacement policy, such analyses were harder to find for other replacement policies. A theoretical explanation was found: for an appropriate setting of analysis over control-flow graphs, cache analysis is PSPACEcomplete for all common replacement policies (FIFO, PLRU, NMRU) except for LRU, for which it is only NP-complete.In this paper, we show that if procedure calls are added to the control flow, then the gap widens: analysis remains NP-complete for LRU, but becomes EXPTIME-complete for the three other policies. For this, we improve on earlier results on the complexity of reachability problems on Boolean programs with procedure calls.In addition, for the LRU policy we derive a backtracking algorithm as well as an approach for using it as a last resort after other analyses have failed to conclude.

show abstract

The complexity gap in the static analysis of cache accesses grows if procedure calls are added

Monniaux

2021

Form Methods Syst Des

View full text Add to dashboard Cite

PLRU Cache Domino Effects

Cited by 14 publications

References 3 publications

Cache Persistence Analysis: Finally Exact

Cache Persistence Analysis: Finally Exact

The complexity gap in the static analysis of cache accesses grows if procedure calls are added

The complexity gap in the static analysis of cache accesses grows if procedure calls are added

Contact Info

Product

Resources

About