U. Vahalia, UNIX Internals: The New Frontiers. Prentice Hall, 1996.  Home/Search Document Not in Database Summary Related Articles Check |
This paper is cited in the following contexts:
First 50 documents Next 50
Alcatraz: Better Living Through Segments - Michael Salib March (Correct)
....is largely unchanged. Just as in the original Beta, the OS scheduler is called as a result of either a timer interrupt or an application call to yield( And just as before, the OS saves the system state (i.e. the registers) into a process structure associated with the previously running process [4]. Alcatraz introduces one new wrinkle into the scheduling protocol. Before the OS can transfer control to the next process scheduled to run, it must reset the segment permission tables. In particular, it must replace them with the initial segment permission tables that only contain entries for the ....
....time slice is complete, the first thing the kernel will do is to save the state of the registers and program counter for that process into it s proc structure. Later on, just before scheduling that process to run again, the kernel will restore the system state from the process s proc structure [4]. A unix like OS might implement process creation in the following manner. Given an executable file to run, the kernel would allocate a new segment. This new segment would be equal in size to the executable. The kernel would then copy the contents of the executable file into the newly allocated ....
[Article contains additional citation context not shown here]
Uresh Vahalia. Unix Internals: The New Frontiers. Prentice Hall, Upper Saddle River, NJ, 1996. 18
Middleware versus Native OS Support: Architectural .. - Shenoy, Hasan.. (2001) (2 citations) (Correct)
....Typically this is done in an incremental manner so as to preserve the semantics and the API provided to best effort applications, ensuring backward compatibility. Examples of this approach include the real time priority class employed by Windows 2000 [16] the real time scheduling class in Solaris [19] and the reservation and proportional share schedulers developed for Linux [18] and FreeBSD [1] Use of a middleware system: In this approach, a middleware layer between the application and the operating system arbitrates access to system resources (see Figure 1(b) Since all requests for system ....
U Vahalia. UNIX Internals: The New Frontiers. Prentice Hall, 1996.
Magazines and Vmem: Extending the Slab Allocator to Many CPUs .. - Bonwick, Adams (2001) (Correct)
....Initially deployed in Solaris 2.4, it has since been adopted in whole or in part by several other operating systems including Linux, FreeBSD, NetBSD, OpenBSD, EROS, and Nemesis. It has also been adapted to applications such as BIRD and Perl. Slab allocation is now described in several OS textbooks [Bovet00, Mauro00, Vahalia96] and is part of the curriculum at major universities worldwide. Meanwhile, the Solaris slab allocator has continued to evolve. It now provides per CPU memory allocation, more general resource allocation, and is available as a user level library. We describe these developments in seven sections as ....
....creating vmem, describe the vmem interfaces, explain the implementation in detail, and discuss vmem s performance (fragmentation, latency, and scalability) under both benchmarks and real world conditions. 4.1. Background Almost all versions of Unix have a resource map allocator called rmalloc( [Vahalia96]. A resource map can be any set of integers, though it s most often an address range like [0xe0000000, 0xf0000000) The interface is simple: rmalloc(map, size) allocates a segment of the specified size from map, and rmfree(map, size, addr) gives it back. Linux s resource allocator and BSD s ....
Uresh Vahalia. UNIX Internals: The New Frontiers. Prentice Hall, 1996.
Middleware versus Native OS Support: Architectural .. - Shenoy, Hasan.. (2001) (2 citations) (Correct)
....Typically this is done in an incremental manner so as to preserve the semantics and the API provided to best effort applications, ensuring backward compatibility. Examples of this approach include the real time priority class employed by Windows 2000 [24] the real time scheduling class in Solaris [27] and the reservation and proportional share schedulers developed for Linux [26] and FreeBSD [2] Use of a middleware: In this approach, a middleware layer between the application and the operating system arbitrates access to system resources (see Figure 1(b) Since all requests for system ....
U Vahalia. UNIX Internals: The New Frontiers. Prentice Hall, 1996.
Using Speculative Execution to Automatically Hide I/O Latency - Chang (2001) (1 citation) (Correct)
....from performing any such actions. For example, on most UNIX operating systems, an execution can produce direct output only by performing specific system calls (for example, system calls that might change the contents of a file system) modifying a mapped file, or accessing a mapped device [64]. Therefore, a design could ensure that speculative execution (or added normal execution) could not produce direct output by guaranteeing that the execution could not perform any system call that might produce direct output, could not modify any mapped files, and could not access any mapped ....
.... these protection boundaries ensure that processes can produce indirect output only by terminating, performing specific system calls (for example, system calls that allow the process to communicate with another process) modifying shared memory segments or mapped file, or accessing mapped devices [64]. An added process could produce indirect output by terminating because, on most operating systems, terminating a process causes a child termination signal (SIGCHLD) to be delivered to the process s current parent, and may also result in the process s exit status being delivered to its parent. An ....
[Article contains additional citation context not shown here]
Uresh Vahalia. UNIX internals - The new frontiers. Prentice Hall, 1996.
Exploiting Gray-Box Knowledge of Buffer-Cache Management - Burnett, Bent.. (2002) (4 citations) (Correct)
.... fingerprints produced for the pure replacement policies of FIFO, LRU, and LFU [23] as well as for other simple replacement policies such as Random and Segmented FIFO [31] To explore the impact of internal state within the replacement policy, we investigate Clock [18] and Two handed Clock [32]. We then demonstrate our ability to identify the use of historical information in the replacement policy, focusing on 2Q [12] and LRU K [19] We conclude this section by showing that Dust is robust to some inaccuracy in its estimate of buffer cache size. 3.1 Simulation methodology Given that ....
U. Vahalia. UNIX Internals: The New Frontiers. Prentice Hall, 1996.
On the Efficient Scheduling of Non-Periodic Tasks in Hard.. - Thomadakis, Liu (1999) (3 citations) (Correct)
....; r i 2 IN (natural numbers) and IN = IN n f0g. Under this assumption time t is slotted , and advances in increments of a fixed size quantum q. In modern operating systems, kernels make scheduling decisions and respond to scheduling events at the integer boundaries of the dispatcher s clock [29, 30, 31, 32, 33]. A discrete time schedule S(t) associates a task index of i 2 with S(t) for every t 2 IN , as with its continuous time counterpart. 2.3 The Serviced Workload and Idle Processor Capacity Processes In this and subsequent sections we will use the following quantities repeatedly, n i (t) ....
U. Vahalia, UNIX Internals: The New Frontiers, Prentice-Hall, Upper Saddle River, NJ, 1996.
Reactive Scheduling For Parallel I/O Systems - Ross (2000) (2 citations) (Correct)
....format between clients and I O daemons is needed. ffl The data transfer system should be implemented as a replaceable module. ffl A better low level client side interface is needed. ffl Low level PVFS operations should be implemented in a way that can be easily integrated with VFS [56]. ffl More disk optimizations are needed in the I O daemons. While PVFS support for simple strided accesses (through partitioning) is a step in the right direction in terms of supporting common I O patterns, more powerful request mech59 anisms are needed to take advantage of the descriptive ....
....with the PVFS daemons are modeled after the system calls that they help perform. When implementing a system such as PVFS at the user level, this is the most obvious method of choosing request types. These calls, however, do not map well into the types of operations seen in the VFS system [56]. This has created problems for us with respect to implementing a client side kernel interface. Mapping VFS like operations into system call operations would be a simpler task. The design of the I O daemon focuses on extracting maximum performance from the network. To this end, when reading a ....
Uresh Vahalia. UNIX Internals: The New Frontiers. Prentice Hall, Upper Saddle River, NJ, 1996. 144
Fine-Grain Distributed Shared Memory on Clusters of Workstations - Schoinas (1997) (3 citations) (Correct)
....region mapped with copy on write semantics. While the kernel can optimize memory usage by keeping the same physical page associated with different virtual addresses, as soon as an operation is performed that would reveal this fact, a new copy the page is created. The SysV shared memory interface [Vah96] exposes memory pages for which the name is associated with the virtual address space but the content is associated with the physical address space. In this case, a modification through one virtual name will be visible through other virtual names. Therefore, the attribute content is associated ....
....for each page in the segment and either a pointer to the anon map structure for anonymous pages (e.g. heap or stack pages) or the vnode structure for mem 1. For the interested reader, an excellent source for further information is Uresh Vahalia s book rifled Unix Internals: The New Frontiers [Vah96] 33 Address Space (struct as) Segment List Segment base address Segment size I Segment Driver Functions[ I Segment Driver Private Dat void ( free) faultcode t (fault) int setprott ( setprot)0; int lockop (lockop)0; Free segment Handle fault in segment page ....
[Article contains additional citation context not shown here]
Uresh Vahalia. Unix Internals': The New Frontiers'. Prentice Hall, 1996.
Process Tracking for Parallel Job Control - Franke, Moreira, Pattnaik (Correct)
....In standard UNIX systems, processes are the units of scheduling. In order to perform job based scheduling one has to use process set mechanisms. Modern UNIX systems, such as SVR4 and 4.4BSD provide two standard notions of process sets. The first one is process group and the second one is session [22]. Process groups are provided to identify a set of related processes that should receive a common signal for certain events. Process groups are the standard UNIX mechanism for implementing job control. Sessions are provided to identify a set of related processes that have a common controlling ....
U. Vahalia. UNIX Internals: The New Frontiers. Prentice-Hall, Inc, 1996.
CAR: Clock with Adaptive Replacement - Bansal, Modha (2004) (3 citations) (Correct)
No context found.
U. Vahalia, UNIX Internals: The New Frontiers. Prentice Hall, 1996.
Parallel Huffman Decoding with Applications to JPEG Files - Klein, Wiseman (2003) (Correct)
No context found.
Vahalia, U. (1996) UNIX Internals---The New Frontiers. Prentice-Hall, Englewood Cliffs, NJ.
Fsmlabs Technical Report - Against Priority Inheritance (Correct)
No context found.
Uresh Vahalia. Unix Internals: The new frontiers. Prentice-Hall, 1996.
Temporal Inventory and Real-Time Synchronization in - Rtlinuxpro Victor Yodaiken (Correct)
No context found.
Uresh Vahalia. Unix Internals: The new frontiers. Prentice-Hall, 1996.
USENIX Association - Th Annual Linux (2000) (Correct)
No context found.
Uresh Vahalia. Unix Internals: The New Frontiers. Prentice-Hall, 1996.
Implementation of Distributed Process Management Protocol Server.. - Agarwal (2000) (5 citations) (Correct)
No context found.
Uresh Vahalia. UNIX Internals: The New Frontiers. Prentice Hall, 1996.
The Fluke Device Driver Framework - Van Maren (1999) (1 citation) (Correct)
No context found.
Uresh Vahalia. UNIX Internals: The New Frontier. Prentice Hall, 1996.
Measuring and Characterizing System Behavior Using.. - Yaghmour, Dagenais (2000) (10 citations) (Correct)
No context found.
Uresh Vahalia. Unix Internals: The New Frontiers. Prentice Hall, 1996.
USENIX Association - Bsdcon Conference San (1992) (2 citations) (Correct)
No context found.
Uresh Vahalia, UNIX Internals: The New Frontiers, Prentice-Hall, Inc. (1996).
Performance of Hardware Compressed Main Memory - Abali, Franke, Shen, Poff, Smith (2000) (10 citations) (Correct)
No context found.
Vahalia, U: "Unix Internals, The New Frontiers", Prentice Hall, ISBN 0-13-101908.
Efficient, Protected Extension of Commodity Operating Systems - Ghormley (1998) (Correct)
No context found.
Uresh Vahalia. UNIX Internals: The New Frontiers. Prentice Hall, 1996.
CAR: Clock with Adaptive Replacement - Bansal, Modha (2004) (3 citations) (Correct)
No context found.
U. Vahalia, UNIX Internals: The New Frontiers. Prentice Hall, 1996.
The Global File System - Soltis, Ruwart, O'Keefe (1996) (12 citations) (Correct)
No context found.
U. Vahalia, Unix Internals: The New Frontiers. Upper Saddle River, NJ 07458: Prentice-Hall, Inc., 1996.
Shell over a Cluster (SHOC): - Towards Achieving Single (2002) (Correct)
No context found.
U. Vahalia. Unix Internals -- The New Frontiers. P rentice Hall, 1996.
Improving Interactive System Performance using TIPME - Endo (2000) (Correct)
No context found.
Uresh Vahalia, UNIX Internals: The New Frontiers, Prentice-Hall Inc., 1996.
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