K. Harty, D.R. Cheriton, Application-controlled physical memory using external page-cache management, in: Proceedings of the 5th International Conference on Architectural Support for Programming Languages and Operating Systems, October 1992, pp. 187--197.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....known problem with pure sharebased approaches [25] enabling efficient memory utilization while still maintaining share based isolation. The use of economic metaphors is also related to more explicit market based approaches designed to facilitate decentralized application level optimization [12]. 9 Conclusions We have presented the core mechanisms and policies used to manage memory resources in ESX Server [29] a commercially available product. Our contributions include several novel techniques and algorithms for allocating memory across virtual machines running unmodified commodity ....

Kieran Harty and David R. Cheriton. "ApplicationControlled Physical Memory using External Page-Cache Management," Proc. Fifth International Conference on Architectural Support for Programming Languages and Operating Systems, October 1992.

.... their own customized shared memory cache coherence protocols [22, 70, 95] Examples of the benefits of carefully chosen primitive interfaces are also common in operating systems research for purposes as diverse as communication protocols for distributed file systems [99] virtual memory management [50], and other kernel services [16, 55] The second component of micro optimization involves using automatic compiler optimizations (e.g. partial redundancy elimination) to leverage the decomposed interface, rather than forcing the application programmer to do the work. This technique has been used ....

Kieran Harty and David R. Cheriton. Applicationcontrolled physical memory using external page-cache management. In Proceedings of the Fifth International Conference on Architectural Support for Programming Languages and Operating Systems, pages 187--197, Boston, MA, October 12--15, 1992.

....Current operating system designs are including more capabilities for applications to improve their own efficiency. Examples of this in sequentia operating systems include the madvise system caJ] in UNIX and the research done at Stanford on externa page cache management for the V system [8]. A good example in paraJ]el file systems is the logica partitioning facility in IBM s VESTA [6] Although file prefetching and caching is done profitably by sequentia systems, new problems occur on multiprocessors. The most paradoxica is that improving the hit rate at the cache may not ....

Kieran Harty and David R. Chefitoh. Application-controlled physical memory using external page-cache manage- ment. In Proceedings of the Fifth International Conference on Architectural Support for Programming Languages and Operating Systems, pages 187-197, Boston, MA, October 1992.

....This makes it difficult to respond to changing load situations or suspected denial ofservice attacks because killing tasks is usually not an option. Finally, related work has shown that applications can benefit significantly from managing their memory resources according to their own policy [15, 19, 20]. This is possible because the application has specific knowledge and can therefore give far more accurate predictions on future resource usage than the kernel. Although this argument has mostly been used for ordinary memory resources, it can be applied to kernel memory as well. 1.2 Approach In ....

.... We consider this overly restrictive because this policy is always a compromise between performance and generality; related work has shown that applications are often ill served by the default operating system policy [1, 50] and can benefit significantly from managing their own memory resources [15, 19, 20, 27, 29, 39]. Also, while the allocation policy can be configured in some of the approaches, other policies, e.g. for placement or replacement, cannot be influenced at all. Furthermore, the effectivity of this approach depends strongly on the policy that is being used. FCFS, which is implemented in L4 and ....

Kieran Harty and David R. Cheriton. Application-controlled physical memory using external page-cache management. In Proceedings of the fifth international conference on Architectural support for programming languages and operating systems, pages 187--197. ACM Press, Oct 1992.

....requests for a particular resource. 2.1 Approaches to customized resource management A number of systems have attempted to remedy this problem by exposing greater resource control to applications. Scheduler activations [7] application specific handlers [143] user level virtual memory managers [58, 93], and operating systems such as SPIN [16] Exokernel [70] and Nemesis [87] are all attempts to augment limited operating system interfaces by giving applications the ability to specialize the policy decisions made by the kernel. Application specific handlers allow applications to push fast ....

K. Harty and D. Cheriton. Application controlled physical memory using external page cache management, October 1992.

....and execute within their own containers; their state is therefore resilient. The concept of a manager is similar to the Mach external pager [32, 44] which has been successfully used to implement a coherent distributed persistent address space [20] In common with Mach and more recent systems [13, 19], managers are responsible for: provision of the pages of data stored in the container, responding to access faults, and receiving data removed from physical memory by the kernel. In addition, Grasshopper managers have the following responsibilities: implementation of a stability ....

Harty, K. and Cheriton, D. R. "Application-Controlled Physical Memory Using External Page-Cache Management", ASPLOS V, Boston, ACM, 1992.

....process. However, kernel level techniques can be crude, and require customization. This trade off led to packet filters being run as interpreted code within the kernel [18, 16] For memory intensive applications such as databases, it can beneficial to involve the application with cache management [15]. In an extensible operating system, the safety of the code in question is a critical issue: a bug or Trojan horse inserted into the kernel could have grave consequences. However, the safety mechanisms should not incur too much cost and thereby negate the intended performance benefits. Active ....

K. Harty and D. Cheriton. Application-controlled physical memory using external page-cache management. In Proceedings of the Fifth International Conference on Architectural Support for Programming Languages and Operating Systems, October 1993.

....the implementation details. We learnt that the design of the content delivery scheduler was critical to the overall performance of the system in the presence of a large amount of information waiting to be scheduled. We modified the memory market scheduling model proposed by Harty and Cheriton [1] to suit our goals of making a scalable and cost effective scheduler. The main challenges posed by large data sets on the scheduling process are highlighted in the later section on content scheduling. We have tested the algorithm to perform satisfactorily under heavy workloads, a common ....

....expires and is removed from the system by a helper thread. We shall now describe the modified memory market scheduler that we have implemented. 3.3. 1 The Scheduling Algorithm Many ideas used in designing the scheduler are borrowed from the memory market model proposed by Harty and Cheriton [1]. Consider a set, S, of active data items in a particular channel. A data item earns money in proportion to its priority, P, while in the active queue. At a given point of time, an active data item might have a balance, B, with which it can purchase additional display slots. The scheduler ....

Kieran Harty, David R. Cheriton. Application -Controlled Physical Memory using External Page-Cache Management. Proceedings of the Fifth International Conference on Architectural Support for Programming Languages and Operating Systems, Boston MA, Vol. 34, No. 5, October 1992.

....a remote le as though it were local and the command transformation which invokes an arbitrary executable (even a shell script) on a data stream. 1 Introduction Over the last decade, research in system software has demonstrated that extensibility o ers increased performance and functionality [5, 10, 28, 8]. However, fully extensible operating systems are little used, primarily due to the high cost associated with adopting a new operating system. Moreover, some argue that the cost of an extensible operating system in terms of complexity and performance does not justify the added bene ts [7] On the ....

Kieran Harty and David R. Cheriton. Application-controlled physical memory using external pagecache management. In Proc. Fifth Intl. Conf. on Architectural Support for Programming Languages and Operating Systems, pages 187-197, Boston, MA, October 1992.

.... their own customized shared memory cache coherence protocols [10, 33, 42] Examples of the benefits of carefully chosen primitive interfaces are also common in operating systems research for purposes as diverse as communication protocols for distributed file systems [43] virtual memory management [24], and other kernel services [7, 16, 27] The second component of micro optimization involves using automatic compiler optimizations (e.g. partial redundancy elimination) to leverage the decomposed interface, rather than forcing the application programmer to do the work. This technique has been ....

K. Harty and D. R. Cheriton. Application-Controlled Physical Memory using External Page-Cache Management. In Proceedings of the Fifth International Conference on Architectural Support for Programming Languages and Operating Systems, pages 187--197, Boston, MA, October 12--15, 1992.

.... the history of computer science there has been a fairly constant opinion that current operating systems are inadequate [4, 7, 9, 11, 15, 18] The literature is rife with specific examples that describe the cost of the inappropriate, inefficient abstractions peddled by operating systems [2, 4, 12, 13, 18, 23, 24]. This situation has persisted for the last three decades, and has survived numerous assaults (object oriented operating systems and micro kernels are two of the more popular movements) As a general rule, a concept that cannot be realized after such a long period of time should be reexamined. ....

....abstractions; unfortunately, such emulation is typically clumsy, complicated, and prohibitively expensive. For example, once the application has no access to the raw disk interface, database records must be emulated on top of files. The list of such examples is painfully long and continues to grow [2, 4, 13, 18, 23, 24]. In short, operating systems are complex, fragile, inflexible, and slow, because they have dabbled in the practice of providing a general purpose virtual machine. The operating system is basically hardware masquerading as software: it cannot be changed, all applications must use it, and the ....

K. Harty and D.R. Cheriton. Application-controlled physical memory using external page-cache management. In Proceedings of the Fifth International Conference on ASPLOS, pages 187--199, October 1992.

....such as Exokernel [12] which promotes the implementation of OS components as libraries under application control. Likewise, a SEDAbased OS should expose a virtual memory interface which makes physical memory availability explicit; this approach is similar to that of application controlled paging [7, 8]. 5. Related Work The SEDA design was derived from approaches to managing high concurrency and unpredictable load in a variety of systems. The Flash web server [21] and the Harvest web cache [4] are based on an asynchronous, event driven model which closely resembles the SEDA architecture. In ....

K. Harty and D. Cheriton. Application controlled physical memory using external page cache management, October 1992.

....many proposals to decrease the number of disk accesses for swapping issues. For instance, some work has been devoted to minimize the number of pages that have to be swapped out. If the contents of a page is irrelevant to the application execution, this page does not need to be kept in the swap [15, 5]. In the same line, software has been developed to study the utilization of the pages and thus improve the programs and reduce the number of pages swapped in out [12] There has also been some work that tried to group pages when swapped out so that larger writes were done [1] Finally, the ....

HARTY, K., AND CHERITON, D. R. Applicationcontrolled physical memory using external pagecache management. In Proccedings of the V Architecture Support for Programming Laguages and Operating System (October 1992).

....we should know the capacity of a block and the internal memory beforehand, 2. we must be able to align elements into blocks, and 3. we must have a full control over the replacement of the blocks in internal memory. There are operating systems that provide support for these facilities (see, e.g. [17, 21, 23]) The tree structure of Arge and the heap structure of Fadel et al. do not give any guarantees for the performance of individual operations. In fact, one Insert or DeleteMin can be extremely expensive, the cost of handling the whole sequence being an upper bound. Therefore, it is risky to use ....

K. Harty and D. R. Cheriton. Application-controlled physical memory using external page-cache management. In Proceedings of the 5th International Conference on Architectural Support for Programming Languages and Operating Systems, ACM SIGPLAN Notices, volume 27, number 9, pages 187--197, 1992.

....research. We can fundamentally separate code reordering approaches into three groups based on the granularity of the code module under consideration: page, procedure and basic block. Traditionally page repositioning algorithms have targeted the improvement of the average memory access time [62, 63, 64, 65]. Some of them require some form of operating system support. Procedure reordering also focuses on improving the memory access time [52, 54, 55, 66, 67, 68] Basic block techniques can be roughly characterized as intra or interprocedural. Intraprocedural rearrange blocks strictly within the ....

K. Harty and D.R. Cheriton. Application-controlled Physical Memory using External Page Cache Management. In Proceedings of the International Conference on Architectural Support for Programming Languages and Operating Systems, pages 187--197, October 1992.

....short of allowing the application full control over the page mappings. V Before publishing the Cache Kernel, the designers of V suggested a powerful kernel interface that allows application to control page replacement, pinning, and prefetching, and to obtain the virtual to physical mappings [HC92] A crucial primitive in this interface allows applications to migrate pages from one segment into another through remapping techniques. For example, if the virtual page v1 is mapped to the page frame p, then the page may be migrated such that some other page v2 is mapped to p and v1 is unmapped. ....

Kieran Harty and David R. Cheriton. Applicationcontrolled physical memory using external page-cache management. Proc. of the Fifth Conf. on Architectural Support for Programming languages and Operating Systems, pages 187--199, October 1992.

....obvious that sequence detection can be used for prefetching as well. How to balance the prefetching and caching, though, is a complicated issue that needs further study [7] Recent research projects on application controlled kernels show the potentials of application specific replacement policies [34, 17, 25, 23]. However, most of those studies focus on letting applications inform kernel of their access patterns, instead of having operating systems detect them at run time (as in the SEQ algorithm) Our study shows that in fact, runtime automatic detection by kernel is feasible and can perform quite well. ....

K. Harty and D. R. Cheriton. Applicationcontrolled physical memory using external pagecache management. In Proc. Fifth International Conf. on Architectural Support for Programming Languages and Operating Systems, SIGOPS Operating Systems Review Special Issue, volume 26, page 187, Boston, MA, October 12-15 1992.

....extensible object oriented virtual memory [14] extended Mach pagers by allowing pagers to implement replacement policies. HiPEC [15] allows applications to control replacement policies by downloading policies written in a restricted language to the kernel. Sechrest [24] and V page cache managers [11] extended pagers even further by allowing pagers to implement replacement and placement policies. Sechrest and V page cache managers also moved allocation policies out of the kernel. SPIN [3] attempts to provide application specific VM through kernel extensions. However, downloading untrusted ....

K. Harty and D. Cheriton. Application-controlled physical memory using external page-cache management. In ASPLOS. ACM, October 1992.

....heuristic. LRU has more page faults) These values were computed by dividing the total number of page faults, summation over all relevant simulation runs, for LRU and our heuristic. 1. 4 Experimentation Several recent experimental studies of alternatives to LRU have appeared in the literature [7, 14, 36, 37, 38]. Many of the new models and algorithms studied are adaptive and try to reflect changing statistical parameterization of the sequences. In this context we present an algorithm that is very similar although there is no explicit statistical assumption used. We consider several heuristics, based on ....

K. Harty and D. Cheriton. Application-controlled physical memory using external page-cache management. In Proceedings of the 5th International Conference on Architectural Support for Programming Languages and Operating Systems, 1992.

....hardware to provide high level abstractions such as processes and file systems that enable application developers to build systems both efficiently and portably. Although high level abstractions like this are suitable for many systems, there are also systems for which they are not appropriate [4, 6, 25, 43, 46]. This would be acceptable if the developers of these systems could simply ignore the abstractions that are unsuitable and create their own from scratch. However, since the operating system defines a high level virtual machine, applications are effectively forced to use the abstractions provided. ....

K. Harty and D.R. Cheriton. "Application-Controlled Physical Memory Using External PageCache Management", in Proceedings of the Fifth Conference on Architectural Support for Programming Languages and Operating Systems, pp. 187-199, 1992.

....maximum flexibility and avoids the kernel making decisions which impact upon performance. For example, the memory management policy can have major 6 effects on the performance of garbage collection. User level virtual memory management, supported on a number of recent operating systems [14], has a similar motivation. Each container has an associated manager, which is an ordinary user level program, held within a container. The manager is responsible for: provision of the pages of data stored in the container, responding to access faults, operation within a limited amount ....

Harty, K. and Cheriton, D. R. "Application-Controlled Physical Memory using External Page-Cache Management", ASPLOS V, ACM, Boston, 1992.

....many proposals to decrease the number of disk accesses for swapping issues. For instance, some work has been devoted to minimize the number of pages that have to be swapped out. If the contents of a page is irrelevant to the application execution, this page does not need to be kept in the swap [26, 27]. In the same line, software has been developed to study the utilization of the pages and thus improve the programs and reduce the number of pages swapped in out [2] There has also been some work that tried to group pages when swapped out so that larger writes were done [28] Finally, the ....

Harty, K., and Cheriton, D. R. Application-controlled physical memory using external page-cache management. In Proccedings of the V Architecture Support for Programming Laguages and Operating System (October 1992).

....pinned pages can not be reclaimed by the operating system. User level page replacement To overcome the expressiveness shortcomings of other approaches, researchers have been investigating provisions for user implemented virtual memory management [McNamee Armstrong 90, Sechrest Park 91, Harty Cheriton 92, Krueger et al. 93] The idea behind these systems is to allow programmers to implement a customized page replacement policy that best suits the needs of each application. This is the most general approach of those investigated; all of the other facilities can be implemented as instances of ....

....management modules. Most of these systems are extensions of the idea of Mach s external pager system. A manager is assigned to handle the page replacement policy for a region of virtual memory, just as external pagers handle the backing storage for their memory objects. The V system [Harty Cheriton 92] incorporated external pagers (as in Mach) along with control over physical memory management into the V operating system [Cheriton 88] In addition 10 to the backing storage operations of external pagers, V Segment Managers 1 can control the mapping of physical pages into an application s ....

Keiran Harty and David Cheriton. Application-Controlled Physical Memory using External Page-Cache Management. In Proceedings of the 5th International Conference on Architectural Support for Programming Languages and Operating Systems, pages 187--197, Boston, Massachusetts, October 1992.

.... interface have been used to implement userlevel page replacement policies [14] and to support discardable pages (i.e. dirty pages that do not have to be written to backing store) 20] More aggressive application control of physical memory was implemented in the V kernel by Harty and Cheriton [10]. In their scheme, the application was given complete control over a cache of physical pages, enabling the implementation of application specific memory management policies. Giving applications more control over physical resources (not just memory) is also a part of the motivation behind ....

K. Harty and D. Cheriton. Application-Controlled Physical Memory Using External Page-Cache Management. In Proc. of the 5th ASPLOS, pages 187--199, Oct. 1992.

....real work. We document this behavior in the current This is an extended version of a paper that appeared in Proceedings of Visualization 97, Phoenix AZ, October 1997. 2 paper, thrashing in CFD visualization has also been documented by Ueng [26] Thrashing is more generally addressed in [2, 10, 14, 16, 18]. One approach to out of core visualization that has been more successfully employed than reliance on OS virtual memory is that of application controlled segmentation. With this approach the application chooses some natural unit (segment) of data and specifically loads a segment when it is ....

....for very limited memory (M=32) remote and local are essentially at parity. 6 Related work Researchers in operating systems have recently explored extensions to standard systems to support more application control over virtual memory. The case for these extensions has been made repeatedly (cf. [2, 10, 14, 18, 21, 27]) Some research prototypes have added more application control [11, 14, 20, 21, 27] but these features have unfortunately not found their way into commercial operating systems. Appel and Li have demonstrated by operating system modification that application control over write back policies can ....

[Article contains additional citation context not shown here]

K. Harty and D. R. Cheriton, "Application-Controlled Physical Memory Using External Page-Cache Management," Fifth International Conference on Architectural Support for Programming Languages and Operating Systems, October 1992, pp. 187-197.

....security reasons, both issues need support by the kernel. IPC relies on trusted sender and receiver IDs, memory management on protection schemes to enforce access control to physical frames; for that, systems with virtual memory uses the translation from virtual into physical memory. In [HWL96, HC92] techniques are presented that enable to manage address spaces on user level. A mapping originally belongs to exactly one address space, but can be transferred to another space with less or same access rights. By use of this, address spaces can be created hierarchically. Figure 1 illustrates ....

K. Harty and D. R. Cheriton. Applicationcontrolled physical memory using external pagecache management. In 5th International Conference on Architectural Support for Programming Languages and Operating Systems (ASPLOS), pages 187-197, Boston, MA, October 1992.

....sparse matrix multiplied by dense vector, are also enhanced by more efficiently feeding the microprocessor with data. All of our applications, however, require some small computations which can not be supported without more generalized computation in the memory system than provided by Impulse. In [23] a scheme is proposed to permit user level control of physical memory management. Future exploration of Active Page memory allocation strategies would clearly benefit from more application guided placement and paging control. Intelligent disks [24] 25] have also been proposed. These systems are ....

K. Harty and D. Cheriton, "Application-controlled physical memory using external page-cache management," in Proceedings of the Fifth International Conference on Architectural Support for Programming Languages and Operating Systems, Oct. 1993.

....Windows NT[11] adopts FIFO replacement. Mach[1] introduced the idea of an external pager[45] software outside the operating system kernel proper, yet capable of implementing paging mechanisms. Others have extended the external pager interface to expose the replacement policy as well[29, 39] V [18] implements a segment based scheme in which user level processes control segments. Recently even more exotic interfaces have been proposed, including one built around a so called metaobject protocol[24] and another that involves applications downloading replacement policies coded in a simple ....

K. Harty and D. R. Cheriton. Application-controlled physical memory using external page-cache management. In Proc. Fifth International Conf. on Architectural Support for Programming Languages and Operating Systems, SIGOPS Operating Systems Review Special Issue, volume 26, page 187, Boston, MA, October 1992.

....of customization and to sim This work was supported in part by the Advanced Research Projects Agency under contracts N00014 94 1 0985 and by a NSF National Young Investigator Award. Sandeep Gupta was also supported in part by an ONR Graduate Fellowship. plify kernel, a number of microkernels [1, 11, 25] have put the policy decisions for page replacement in userlevel servers. Unfortunately, many policies in these systems are still hard coded (e.g. the page table structure is fixed) and the page servers are often complex and have superuser privilege, making them hard to modify or replace. Other ....

....to modify or replace. Other microkernels [12, 17, 21] have chosen to implement a very restricted virtual memory system in order to make the system fast and simple. These systems do not support VM techniques such as copy on write, which can improve applicationperformance by an order of magnitude [2, 11, 14]. In summary, current VM systems are either complicated and unwieldy, or naive and rudimentary; both approaches penalize applications. To remedy the situation, this paper proposes a novel organization of the VM system, application level VM (AVM) which allows applications to set policies by ....

[Article contains additional citation context not shown here]

K. Harty and D.R. Cheriton. Application-controlled physical memory using external page-cache management. In Proceedings of the Fifth International Conference on ASPLOS, pages 187--199, October 1992.

....NSF National Young Investigator award. The views and conclusions contained in this document are those of the authors and should not be interpreted as representing the official policies, either expressed or implied, of the U.S. government. tions are inadequate, their implementation slow, or both [4, 5, 14, 15, 22, 29, 31]. We believe that these problems can be solved by simply lowering the interface to the hardware that is enforced by the kernel: namely, by exporting physical resources to applications directly, management and abstraction of these resources can be specialized for simplicity, efficiency, and ....

....For example, file system and relational database applications often have different storage access patterns and therefore benefit from different buffer and disk management policies. Similar policy conflicts arise in almost every area of resource management (for example, in virtual memory management [15]) Microkernels have pushed some policy decisions out of the kernel into user level servers. However, these policies cannot, generally, be altered or replaced by nonprivileged applications. Additionally, in these microkernels the supervisor mode still defines a highlevel, fixed interface among the ....

Kieran Harty and David R. Cheriton. Applicationcontrolled physical memory using external page-cache management. Proc. of the Fifth Conf. on Architectural Support for Programming languages and Operating Systems, pages 187--199, October 1992.

....attention [452, 88, 396, 106, 515, 448, 447, 397, 550] This interaction has great importance. Systems that use non preemptive partitioning typically do not provide any virtual memory or paging, because they cannot afford the overhead of idling a PE while waiting for a page fault to be serviced [452, 515, 256]. All the memory is dedicated to a single application at a time, and the application is explicitly required to fit into the available memory. In systems that provide dynamic partitioning, this can be as small as the memory associated with a single PE. If memory is shared, the system must know the ....

....given that physical memory is scarce, the application itself should be responsible for its optimal management. The mechanisms to do so are control over what pages of virtual memory will be cached or replaced, and facilities 90 to acquire additional frames of physical memory for bounded durations [256]. Interaction with the architecture and hardware support Architectural features can have a decisive impact on how the machine is shared among multiple users. For example, a dedicated machine (or partition) is obviously needed for SIMD architectures because of the single instruction decoding unit ....

K. Harty and D. R. Cheriton, "Application-controlled physical memory using external pagecache management". In 5th Intl. Conf. Architect. Support for Prog. Lang. & Operating Syst., pp. 187--197, Sep 1992.

.... databases [30] Likewise, generic policies for task scheduling are not well suited for specific control transfers such as cross address space procedure calls, which are typical to any microkernel implementation [31] Such conflicts have been shown to plague almost every area of resource management [3, 14, 17]. The conflict between generality and performance suggests that an operating system cannot achieve efficiency without taking into account the specific applications that use its services. To answer this fundamental problem, different forms of customizable operating systems have been proposed ....

K. Harty and D. Cheriton. Application-controlled physical memory using external pagecache management. In Proceedings of the Fifth International Conference on Architectural Support for Programming Languages and Operating Systems, pages 187--199, Oct. 1993.

....spurious threads will just compete with all the other threads in the system, including the threads that actually do the computation in the same job. This will degrade the service to all jobs, including the one that created the spurious threads. 56 while waiting for a page fault to be serviced [278, 310, 156]. All the memory is dedicated to a single application at a time, and the application is explicitly required to fit into the available memory. In systems that provide dynamic partitioning, this can be as small as the memory associated with a single PE. If memory is shared, the system must know the ....

K. Harty and D. R. Cheriton, "Application-controlled physical memory using external pagecache management". In 5th Intl. Conf. Architect. Support for Prog. Lang. & Operating Syst., pp. 187--197, Sep 1992.

....The Bin Hopping algorithm is used by existing operating systems such as Digital Unix (OSF) for allocating pages. We use Bin Hopping as our default page allocation algorithm for our baseline configuration. 3. 2 Software Guided Page Placement Custom operating systems, such as Exokernal [10] and V [14] have been designed that allow applications to provide their own page replacement and page mapping policies. Bugnion et al. 3] recently examined using compiler directed page coloring for arrays on multiprocessors. Their approach at run time generates a preferred coloring for data pages ....

K. Harty and D.R. Cheriton. Application-controlled physical memory using external page cache management. In Proceedings of the Fifth International Conference on Architectural Support for Programming Languages and Operating Systems (ASPLOS-V), October 1992.

....which guards files with processes, providing fine grain, ordinary user extensibility [4] However, a watchdog process executes outside of the file system. Additionally, there are extensible operating systems that implicitly provide file system extensi3 bility as part of the operating system [13, 10, 8, 23]. Finally, the streams interface provides transformation like extensibility that (like Unix pipes) illustrates the utility of many small filter operations [18] Section 5 discusses these and several related systems in detail. Mona is unique in supporting two levels of transformations and runtime ....

Kieran Harty and David R. Cheriton. Application-controlled physical memory using external page-cache management. In Proc. Fifth Intl. Conf. on Architectural Support for Programming Languages and Operating Systems, pages 187--197, Boston, MA, October 1992.

....the service which creates 10 to 14 overhead in their evaluations. Spring [12] has an external paging interface similar to Mach except it separating the caching object from the pager object. The caching objects are controlled by the kernel without any participation of user applications. V [7] uses the segment manager (SM) to handle page faults and has interfaces to request and migrate page frames to and from different segment managers. It uses a memory market(MM) approach [6] to allocate page frames among segment managers. All the operations and requests involve transferring control ....

K. Harty and D. R. Cheriton, 'Application-Controlled Physical Memory using External Page-Cache Management', Proceedings of Fifth International Conference on Architectural Support for Programming Languages and Operating Systems, Boston, Massachusetts, USA, October 1992, pp. 187-197.

No context found.

K. Harty, D.R. Cheriton, Application-controlled physical memory using external page-cache management, in: Proceedings of the 5th International Conference on Architectural Support for Programming Languages and Operating Systems, October 1992, pp. 187--197.

No context found.

D. Cheriton and K. Harty. Application-controlled physical memory using external page-cache management. Technical report, Department of Computer Science, Stanford University, 1991. Technical Report draft.

No context found.

Kieran Harty and David R. Cheriton. Application-controlled physical memory using external pagecache management. In Proceedings of the Fifth International Conference on Architectural Support for Programming Languages and Operating Systems, pages 187--197. ACM Press, 1992. 22

No context found.

Harty, K. and Cheriton, D. R. Application-Controlled Physical Memory Using External Page-Cache Management. In Proceedings of the 5th ACM Symposium on Architectural Support for Programming Languages and Operating Systems, 1992.

No context found.

K. Harty and D. Cheriton. Application controlled physical memory using external page cache management, October 1992.

No context found.

Kieran Harty and David R. Cheriton. Application-controlled physical memory using external page-cache management. In Proc. 5th ASPLOS, pages 187--197, Oct 1992.

No context found.

Kieran Harty and David R. Cheriton. Application-controlled physical memory using external pagecache management. In Proceedings of the Fifth International Conference on Architectural Support for Programming Languages and Operating Systems, pages 187--197. ACM Press, 1992. 22

No context found.

K. Harty and D. Cheriton. Application controlled physical memory using external page cache management, October 1992.

No context found.

K. Harty and D. R. Cheriton. Application-controlled physical memory using external page-cache management. Technical Report CS-TR-91-1394, Stanford University, 95 Department of Computer Science, Oct. 1991.

No context found.

Kieran Harty and David R. Cheriton. Application-controlled physical memory using external page-cache management. In Proc. 5th ASPLOS, pages 187-197, Oct 1992.

No context found.

Kieran Hart and David R. Cheriton. Application-Controlled Physical Memory using External Page-Cache Management. In ASPLOS, pages 187--197, 1992.

No context found.

Kieran Harty and David R. Cheriton. Application-controlled physical memory using external page-cache management. In Proceedings of the Fifth International Conference on Architectural Support for Programming Languages and Operating Systems (ASPLOS-V), pages 187--199, Boston, MA, U.S.A., October 1992. ACM SIGPLAN. 111.

No context found.

Kieran Harty and David R. Cheriton. Application-controlled physical memory using external page-cache management. In Proc. 5th ASPLOS, pages 187--197, Oct 1992.

No context found.

K. Harty and D. Cheriton. Application-controlled physical memory using external page-cache management. In Proceedings of the 5th International Conference on Architectural Support for Programming Languages and Operating Systems, 1992.

First 50 documents  Next 50

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