Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 1 | Read the F-ing Papers! |
| 2 | |
| 3 | |
| 4 | This document describes RCU-related publications, and is followed by |
Paul E. McKenney | dd81eca | 2005-09-10 00:26:24 -0700 | [diff] [blame] | 5 | the corresponding bibtex entries. A number of the publications may |
| 6 | be found at http://www.rdrop.com/users/paulmck/RCU/. |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 7 | |
| 8 | The first thing resembling RCU was published in 1980, when Kung and Lehman |
| 9 | [Kung80] recommended use of a garbage collector to defer destruction |
| 10 | of nodes in a parallel binary search tree in order to simplify its |
| 11 | implementation. This works well in environments that have garbage |
Paul E. McKenney | f85d6c7 | 2008-01-25 21:08:25 +0100 | [diff] [blame] | 12 | collectors, but most production garbage collectors incur significant |
| 13 | overhead. |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 14 | |
| 15 | In 1982, Manber and Ladner [Manber82,Manber84] recommended deferring |
| 16 | destruction until all threads running at that time have terminated, again |
| 17 | for a parallel binary search tree. This approach works well in systems |
| 18 | with short-lived threads, such as the K42 research operating system. |
| 19 | However, Linux has long-lived tasks, so more is needed. |
| 20 | |
| 21 | In 1986, Hennessy, Osisek, and Seigh [Hennessy89] introduced passive |
| 22 | serialization, which is an RCU-like mechanism that relies on the presence |
| 23 | of "quiescent states" in the VM/XA hypervisor that are guaranteed not |
| 24 | to be referencing the data structure. However, this mechanism was not |
| 25 | optimized for modern computer systems, which is not surprising given |
| 26 | that these overheads were not so expensive in the mid-80s. Nonetheless, |
| 27 | passive serialization appears to be the first deferred-destruction |
Paul E. McKenney | 4c54005 | 2010-01-14 16:10:57 -0800 | [diff] [blame] | 28 | mechanism to be used in production. Furthermore, the relevant patent |
| 29 | has lapsed, so this approach may be used in non-GPL software, if desired. |
| 30 | (In contrast, implementation of RCU is permitted only in software licensed |
| 31 | under either GPL or LGPL. Sorry!!!) |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 32 | |
| 33 | In 1990, Pugh [Pugh90] noted that explicitly tracking which threads |
| 34 | were reading a given data structure permitted deferred free to operate |
| 35 | in the presence of non-terminating threads. However, this explicit |
| 36 | tracking imposes significant read-side overhead, which is undesirable |
| 37 | in read-mostly situations. This algorithm does take pains to avoid |
| 38 | write-side contention and parallelize the other write-side overheads by |
| 39 | providing a fine-grained locking design, however, it would be interesting |
| 40 | to see how much of the performance advantage reported in 1990 remains |
| 41 | in 2004. |
| 42 | |
| 43 | At about this same time, Adams [Adams91] described ``chaotic relaxation'', |
| 44 | where the normal barriers between successive iterations of convergent |
| 45 | numerical algorithms are relaxed, so that iteration $n$ might use |
| 46 | data from iteration $n-1$ or even $n-2$. This introduces error, |
| 47 | which typically slows convergence and thus increases the number of |
| 48 | iterations required. However, this increase is sometimes more than made |
| 49 | up for by a reduction in the number of expensive barrier operations, |
| 50 | which are otherwise required to synchronize the threads at the end |
| 51 | of each iteration. Unfortunately, chaotic relaxation requires highly |
| 52 | structured data, such as the matrices used in scientific programs, and |
| 53 | is thus inapplicable to most data structures in operating-system kernels. |
| 54 | |
Paul E. McKenney | 3230075 | 2008-05-12 21:21:05 +0200 | [diff] [blame] | 55 | In 1992, Henry (now Alexia) Massalin completed a dissertation advising |
| 56 | parallel programmers to defer processing when feasible to simplify |
| 57 | synchronization. RCU makes extremely heavy use of this advice. |
| 58 | |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 59 | In 1993, Jacobson [Jacobson93] verbally described what is perhaps the |
| 60 | simplest deferred-free technique: simply waiting a fixed amount of time |
| 61 | before freeing blocks awaiting deferred free. Jacobson did not describe |
| 62 | any write-side changes he might have made in this work using SGI's Irix |
| 63 | kernel. Aju John published a similar technique in 1995 [AjuJohn95]. |
| 64 | This works well if there is a well-defined upper bound on the length of |
| 65 | time that reading threads can hold references, as there might well be in |
| 66 | hard real-time systems. However, if this time is exceeded, perhaps due |
| 67 | to preemption, excessive interrupts, or larger-than-anticipated load, |
| 68 | memory corruption can ensue, with no reasonable means of diagnosis. |
| 69 | Jacobson's technique is therefore inappropriate for use in production |
| 70 | operating-system kernels, except when such kernels can provide hard |
| 71 | real-time response guarantees for all operations. |
| 72 | |
| 73 | Also in 1995, Pu et al. [Pu95a] applied a technique similar to that of Pugh's |
| 74 | read-side-tracking to permit replugging of algorithms within a commercial |
| 75 | Unix operating system. However, this replugging permitted only a single |
| 76 | reader at a time. The following year, this same group of researchers |
| 77 | extended their technique to allow for multiple readers [Cowan96a]. |
| 78 | Their approach requires memory barriers (and thus pipeline stalls), |
| 79 | but reduces memory latency, contention, and locking overheads. |
| 80 | |
| 81 | 1995 also saw the first publication of DYNIX/ptx's RCU mechanism |
| 82 | [Slingwine95], which was optimized for modern CPU architectures, |
| 83 | and was successfully applied to a number of situations within the |
| 84 | DYNIX/ptx kernel. The corresponding conference paper appeared in 1998 |
| 85 | [McKenney98]. |
| 86 | |
| 87 | In 1999, the Tornado and K42 groups described their "generations" |
| 88 | mechanism, which quite similar to RCU [Gamsa99]. These operating systems |
| 89 | made pervasive use of RCU in place of "existence locks", which greatly |
| 90 | simplifies locking hierarchies. |
| 91 | |
| 92 | 2001 saw the first RCU presentation involving Linux [McKenney01a] |
| 93 | at OLS. The resulting abundance of RCU patches was presented the |
| 94 | following year [McKenney02a], and use of RCU in dcache was first |
| 95 | described that same year [Linder02a]. |
| 96 | |
Paul E. McKenney | d19720a | 2006-02-01 03:06:42 -0800 | [diff] [blame] | 97 | Also in 2002, Michael [Michael02b,Michael02a] presented "hazard-pointer" |
| 98 | techniques that defer the destruction of data structures to simplify |
| 99 | non-blocking synchronization (wait-free synchronization, lock-free |
| 100 | synchronization, and obstruction-free synchronization are all examples of |
| 101 | non-blocking synchronization). In particular, this technique eliminates |
| 102 | locking, reduces contention, reduces memory latency for readers, and |
| 103 | parallelizes pipeline stalls and memory latency for writers. However, |
| 104 | these techniques still impose significant read-side overhead in the |
| 105 | form of memory barriers. Researchers at Sun worked along similar lines |
Paul E. McKenney | f85d6c7 | 2008-01-25 21:08:25 +0100 | [diff] [blame] | 106 | in the same timeframe [HerlihyLM02]. These techniques can be thought |
| 107 | of as inside-out reference counts, where the count is represented by the |
| 108 | number of hazard pointers referencing a given data structure (rather than |
| 109 | the more conventional counter field within the data structure itself). |
| 110 | |
| 111 | By the same token, RCU can be thought of as a "bulk reference count", |
| 112 | where some form of reference counter covers all reference by a given CPU |
| 113 | or thread during a set timeframe. This timeframe is related to, but |
| 114 | not necessarily exactly the same as, an RCU grace period. In classic |
| 115 | RCU, the reference counter is the per-CPU bit in the "bitmask" field, |
| 116 | and each such bit covers all references that might have been made by |
| 117 | the corresponding CPU during the prior grace period. Of course, RCU |
| 118 | can be thought of in other terms as well. |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 119 | |
| 120 | In 2003, the K42 group described how RCU could be used to create |
Paul E. McKenney | f85d6c7 | 2008-01-25 21:08:25 +0100 | [diff] [blame] | 121 | hot-pluggable implementations of operating-system functions [Appavoo03a]. |
| 122 | Later that year saw a paper describing an RCU implementation of System |
| 123 | V IPC [Arcangeli03], and an introduction to RCU in Linux Journal |
| 124 | [McKenney03a]. |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 125 | |
| 126 | 2004 has seen a Linux-Journal article on use of RCU in dcache |
| 127 | [McKenney04a], a performance comparison of locking to RCU on several |
| 128 | different CPUs [McKenney04b], a dissertation describing use of RCU in a |
Paul E. McKenney | a83f1fe | 2005-05-01 08:59:05 -0700 | [diff] [blame] | 129 | number of operating-system kernels [PaulEdwardMcKenneyPhD], a paper |
| 130 | describing how to make RCU safe for soft-realtime applications [Sarma04c], |
| 131 | and a paper describing SELinux performance with RCU [JamesMorris04b]. |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 132 | |
Paul E. McKenney | f85d6c7 | 2008-01-25 21:08:25 +0100 | [diff] [blame] | 133 | 2005 brought further adaptation of RCU to realtime use, permitting |
Paul E. McKenney | dd81eca | 2005-09-10 00:26:24 -0700 | [diff] [blame] | 134 | preemption of RCU realtime critical sections [PaulMcKenney05a, |
| 135 | PaulMcKenney05b]. |
| 136 | |
Paul E. McKenney | f85d6c7 | 2008-01-25 21:08:25 +0100 | [diff] [blame] | 137 | 2006 saw the first best-paper award for an RCU paper [ThomasEHart2006a], |
| 138 | as well as further work on efficient implementations of preemptible |
| 139 | RCU [PaulEMcKenney2006b], but priority-boosting of RCU read-side critical |
| 140 | sections proved elusive. An RCU implementation permitting general |
| 141 | blocking in read-side critical sections appeared [PaulEMcKenney2006c], |
| 142 | Robert Olsson described an RCU-protected trie-hash combination |
| 143 | [RobertOlsson2006a]. |
| 144 | |
Paul E. McKenney | 3230075 | 2008-05-12 21:21:05 +0200 | [diff] [blame] | 145 | 2007 saw the journal version of the award-winning RCU paper from 2006 |
| 146 | [ThomasEHart2007a], as well as a paper demonstrating use of Promela |
| 147 | and Spin to mechanically verify an optimization to Oleg Nesterov's |
| 148 | QRCU [PaulEMcKenney2007QRCUspin], a design document describing |
| 149 | preemptible RCU [PaulEMcKenney2007PreemptibleRCU], and the three-part |
| 150 | LWN "What is RCU?" series [PaulEMcKenney2007WhatIsRCUFundamentally, |
| 151 | PaulEMcKenney2008WhatIsRCUUsage, and PaulEMcKenney2008WhatIsRCUAPI]. |
Paul E. McKenney | f85d6c7 | 2008-01-25 21:08:25 +0100 | [diff] [blame] | 152 | |
Paul E. McKenney | 4c54005 | 2010-01-14 16:10:57 -0800 | [diff] [blame] | 153 | 2008 saw a journal paper on real-time RCU [DinakarGuniguntala2008IBMSysJ], |
| 154 | a history of how Linux changed RCU more than RCU changed Linux |
| 155 | [PaulEMcKenney2008RCUOSR], and a design overview of hierarchical RCU |
| 156 | [PaulEMcKenney2008HierarchicalRCU]. |
| 157 | |
| 158 | 2009 introduced user-level RCU algorithms [PaulEMcKenney2009MaliciousURCU], |
| 159 | which Mathieu Desnoyers is now maintaining [MathieuDesnoyers2009URCU] |
| 160 | [MathieuDesnoyersPhD]. TINY_RCU [PaulEMcKenney2009BloatWatchRCU] made |
| 161 | its appearance, as did expedited RCU [PaulEMcKenney2009expeditedRCU]. |
| 162 | The problem of resizeable RCU-protected hash tables may now be on a path |
| 163 | to a solution [JoshTriplett2009RPHash]. |
| 164 | |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 165 | Bibtex Entries |
| 166 | |
| 167 | @article{Kung80 |
| 168 | ,author="H. T. Kung and Q. Lehman" |
| 169 | ,title="Concurrent Maintenance of Binary Search Trees" |
| 170 | ,Year="1980" |
| 171 | ,Month="September" |
| 172 | ,journal="ACM Transactions on Database Systems" |
| 173 | ,volume="5" |
| 174 | ,number="3" |
| 175 | ,pages="354-382" |
| 176 | } |
| 177 | |
| 178 | @techreport{Manber82 |
| 179 | ,author="Udi Manber and Richard E. Ladner" |
| 180 | ,title="Concurrency Control in a Dynamic Search Structure" |
| 181 | ,institution="Department of Computer Science, University of Washington" |
| 182 | ,address="Seattle, Washington" |
| 183 | ,year="1982" |
| 184 | ,number="82-01-01" |
| 185 | ,month="January" |
| 186 | ,pages="28" |
| 187 | } |
| 188 | |
| 189 | @article{Manber84 |
| 190 | ,author="Udi Manber and Richard E. Ladner" |
| 191 | ,title="Concurrency Control in a Dynamic Search Structure" |
| 192 | ,Year="1984" |
| 193 | ,Month="September" |
| 194 | ,journal="ACM Transactions on Database Systems" |
| 195 | ,volume="9" |
| 196 | ,number="3" |
| 197 | ,pages="439-455" |
| 198 | } |
| 199 | |
| 200 | @techreport{Hennessy89 |
| 201 | ,author="James P. Hennessy and Damian L. Osisek and Joseph W. {Seigh II}" |
| 202 | ,title="Passive Serialization in a Multitasking Environment" |
| 203 | ,institution="US Patent and Trademark Office" |
| 204 | ,address="Washington, DC" |
| 205 | ,year="1989" |
| 206 | ,number="US Patent 4,809,168 (lapsed)" |
| 207 | ,month="February" |
| 208 | ,pages="11" |
| 209 | } |
| 210 | |
| 211 | @techreport{Pugh90 |
| 212 | ,author="William Pugh" |
| 213 | ,title="Concurrent Maintenance of Skip Lists" |
| 214 | ,institution="Institute of Advanced Computer Science Studies, Department of Computer Science, University of Maryland" |
| 215 | ,address="College Park, Maryland" |
| 216 | ,year="1990" |
| 217 | ,number="CS-TR-2222.1" |
| 218 | ,month="June" |
| 219 | } |
| 220 | |
| 221 | @Book{Adams91 |
| 222 | ,Author="Gregory R. Adams" |
| 223 | ,title="Concurrent Programming, Principles, and Practices" |
| 224 | ,Publisher="Benjamin Cummins" |
| 225 | ,Year="1991" |
| 226 | } |
| 227 | |
Paul E. McKenney | 3230075 | 2008-05-12 21:21:05 +0200 | [diff] [blame] | 228 | @phdthesis{HMassalinPhD |
| 229 | ,author="H. Massalin" |
| 230 | ,title="Synthesis: An Efficient Implementation of Fundamental Operating |
| 231 | System Services" |
| 232 | ,school="Columbia University" |
| 233 | ,address="New York, NY" |
| 234 | ,year="1992" |
| 235 | ,annotation=" |
| 236 | Mondo optimizing compiler. |
| 237 | Wait-free stuff. |
| 238 | Good advice: defer work to avoid synchronization. |
| 239 | " |
| 240 | } |
| 241 | |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 242 | @unpublished{Jacobson93 |
| 243 | ,author="Van Jacobson" |
| 244 | ,title="Avoid Read-Side Locking Via Delayed Free" |
| 245 | ,year="1993" |
| 246 | ,month="September" |
| 247 | ,note="Verbal discussion" |
| 248 | } |
| 249 | |
| 250 | @Conference{AjuJohn95 |
| 251 | ,Author="Aju John" |
| 252 | ,Title="Dynamic vnodes -- Design and Implementation" |
| 253 | ,Booktitle="{USENIX Winter 1995}" |
| 254 | ,Publisher="USENIX Association" |
| 255 | ,Month="January" |
| 256 | ,Year="1995" |
| 257 | ,pages="11-23" |
| 258 | ,Address="New Orleans, LA" |
| 259 | } |
| 260 | |
Paul E. McKenney | f85d6c7 | 2008-01-25 21:08:25 +0100 | [diff] [blame] | 261 | @conference{Pu95a, |
| 262 | Author = "Calton Pu and Tito Autrey and Andrew Black and Charles Consel and |
| 263 | Crispin Cowan and Jon Inouye and Lakshmi Kethana and Jonathan Walpole and |
| 264 | Ke Zhang", |
| 265 | Title = "Optimistic Incremental Specialization: Streamlining a Commercial |
| 266 | Operating System", |
| 267 | Booktitle = "15\textsuperscript{th} ACM Symposium on |
| 268 | Operating Systems Principles (SOSP'95)", |
| 269 | address = "Copper Mountain, CO", |
| 270 | month="December", |
| 271 | year="1995", |
| 272 | pages="314-321", |
| 273 | annotation=" |
| 274 | Uses a replugger, but with a flag to signal when people are |
| 275 | using the resource at hand. Only one reader at a time. |
| 276 | " |
| 277 | } |
| 278 | |
| 279 | @conference{Cowan96a, |
| 280 | Author = "Crispin Cowan and Tito Autrey and Charles Krasic and |
| 281 | Calton Pu and Jonathan Walpole", |
| 282 | Title = "Fast Concurrent Dynamic Linking for an Adaptive Operating System", |
| 283 | Booktitle = "International Conference on Configurable Distributed Systems |
| 284 | (ICCDS'96)", |
| 285 | address = "Annapolis, MD", |
| 286 | month="May", |
| 287 | year="1996", |
| 288 | pages="108", |
| 289 | isbn="0-8186-7395-8", |
| 290 | annotation=" |
| 291 | Uses a replugger, but with a counter to signal when people are |
| 292 | using the resource at hand. Allows multiple readers. |
| 293 | " |
| 294 | } |
| 295 | |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 296 | @techreport{Slingwine95 |
| 297 | ,author="John D. Slingwine and Paul E. McKenney" |
| 298 | ,title="Apparatus and Method for Achieving Reduced Overhead Mutual |
| 299 | Exclusion and Maintaining Coherency in a Multiprocessor System |
| 300 | Utilizing Execution History and Thread Monitoring" |
| 301 | ,institution="US Patent and Trademark Office" |
| 302 | ,address="Washington, DC" |
| 303 | ,year="1995" |
| 304 | ,number="US Patent 5,442,758 (contributed under GPL)" |
| 305 | ,month="August" |
| 306 | } |
| 307 | |
| 308 | @techreport{Slingwine97 |
| 309 | ,author="John D. Slingwine and Paul E. McKenney" |
| 310 | ,title="Method for maintaining data coherency using thread |
| 311 | activity summaries in a multicomputer system" |
| 312 | ,institution="US Patent and Trademark Office" |
| 313 | ,address="Washington, DC" |
| 314 | ,year="1997" |
| 315 | ,number="US Patent 5,608,893 (contributed under GPL)" |
| 316 | ,month="March" |
| 317 | } |
| 318 | |
| 319 | @techreport{Slingwine98 |
| 320 | ,author="John D. Slingwine and Paul E. McKenney" |
| 321 | ,title="Apparatus and method for achieving reduced overhead |
| 322 | mutual exclusion and maintaining coherency in a multiprocessor |
| 323 | system utilizing execution history and thread monitoring" |
| 324 | ,institution="US Patent and Trademark Office" |
| 325 | ,address="Washington, DC" |
| 326 | ,year="1998" |
| 327 | ,number="US Patent 5,727,209 (contributed under GPL)" |
| 328 | ,month="March" |
| 329 | } |
| 330 | |
| 331 | @Conference{McKenney98 |
| 332 | ,Author="Paul E. McKenney and John D. Slingwine" |
| 333 | ,Title="Read-Copy Update: Using Execution History to Solve Concurrency |
| 334 | Problems" |
| 335 | ,Booktitle="{Parallel and Distributed Computing and Systems}" |
| 336 | ,Month="October" |
| 337 | ,Year="1998" |
| 338 | ,pages="509-518" |
| 339 | ,Address="Las Vegas, NV" |
| 340 | } |
| 341 | |
| 342 | @Conference{Gamsa99 |
| 343 | ,Author="Ben Gamsa and Orran Krieger and Jonathan Appavoo and Michael Stumm" |
| 344 | ,Title="Tornado: Maximizing Locality and Concurrency in a Shared Memory |
| 345 | Multiprocessor Operating System" |
| 346 | ,Booktitle="{Proceedings of the 3\textsuperscript{rd} Symposium on |
| 347 | Operating System Design and Implementation}" |
| 348 | ,Month="February" |
| 349 | ,Year="1999" |
| 350 | ,pages="87-100" |
| 351 | ,Address="New Orleans, LA" |
| 352 | } |
| 353 | |
| 354 | @techreport{Slingwine01 |
| 355 | ,author="John D. Slingwine and Paul E. McKenney" |
| 356 | ,title="Apparatus and method for achieving reduced overhead |
| 357 | mutual exclusion and maintaining coherency in a multiprocessor |
| 358 | system utilizing execution history and thread monitoring" |
| 359 | ,institution="US Patent and Trademark Office" |
| 360 | ,address="Washington, DC" |
| 361 | ,year="2001" |
| 362 | ,number="US Patent 5,219,690 (contributed under GPL)" |
| 363 | ,month="April" |
| 364 | } |
| 365 | |
| 366 | @Conference{McKenney01a |
| 367 | ,Author="Paul E. McKenney and Jonathan Appavoo and Andi Kleen and |
| 368 | Orran Krieger and Rusty Russell and Dipankar Sarma and Maneesh Soni" |
| 369 | ,Title="Read-Copy Update" |
| 370 | ,Booktitle="{Ottawa Linux Symposium}" |
| 371 | ,Month="July" |
| 372 | ,Year="2001" |
| 373 | ,note="Available: |
| 374 | \url{http://www.linuxsymposium.org/2001/abstracts/readcopy.php} |
| 375 | \url{http://www.rdrop.com/users/paulmck/rclock/rclock_OLS.2001.05.01c.pdf} |
| 376 | [Viewed June 23, 2004]" |
| 377 | annotation=" |
| 378 | Described RCU, and presented some patches implementing and using it in |
| 379 | the Linux kernel. |
| 380 | " |
| 381 | } |
| 382 | |
| 383 | @Conference{Linder02a |
| 384 | ,Author="Hanna Linder and Dipankar Sarma and Maneesh Soni" |
| 385 | ,Title="Scalability of the Directory Entry Cache" |
| 386 | ,Booktitle="{Ottawa Linux Symposium}" |
| 387 | ,Month="June" |
| 388 | ,Year="2002" |
| 389 | ,pages="289-300" |
| 390 | } |
| 391 | |
| 392 | @Conference{McKenney02a |
| 393 | ,Author="Paul E. McKenney and Dipankar Sarma and |
| 394 | Andrea Arcangeli and Andi Kleen and Orran Krieger and Rusty Russell" |
| 395 | ,Title="Read-Copy Update" |
| 396 | ,Booktitle="{Ottawa Linux Symposium}" |
| 397 | ,Month="June" |
| 398 | ,Year="2002" |
| 399 | ,pages="338-367" |
| 400 | ,note="Available: |
| 401 | \url{http://www.linux.org.uk/~ajh/ols2002_proceedings.pdf.gz} |
| 402 | [Viewed June 23, 2004]" |
| 403 | } |
| 404 | |
Paul E. McKenney | f85d6c7 | 2008-01-25 21:08:25 +0100 | [diff] [blame] | 405 | @conference{Michael02a |
| 406 | ,author="Maged M. Michael" |
| 407 | ,title="Safe Memory Reclamation for Dynamic Lock-Free Objects Using Atomic |
| 408 | Reads and Writes" |
| 409 | ,Year="2002" |
| 410 | ,Month="August" |
| 411 | ,booktitle="{Proceedings of the 21\textsuperscript{st} Annual ACM |
| 412 | Symposium on Principles of Distributed Computing}" |
| 413 | ,pages="21-30" |
| 414 | ,annotation=" |
| 415 | Each thread keeps an array of pointers to items that it is |
| 416 | currently referencing. Sort of an inside-out garbage collection |
| 417 | mechanism, but one that requires the accessing code to explicitly |
| 418 | state its needs. Also requires read-side memory barriers on |
| 419 | most architectures. |
| 420 | " |
| 421 | } |
| 422 | |
| 423 | @conference{Michael02b |
| 424 | ,author="Maged M. Michael" |
| 425 | ,title="High Performance Dynamic Lock-Free Hash Tables and List-Based Sets" |
| 426 | ,Year="2002" |
| 427 | ,Month="August" |
| 428 | ,booktitle="{Proceedings of the 14\textsuperscript{th} Annual ACM |
| 429 | Symposium on Parallel |
| 430 | Algorithms and Architecture}" |
| 431 | ,pages="73-82" |
| 432 | ,annotation=" |
| 433 | Like the title says... |
| 434 | " |
| 435 | } |
| 436 | |
| 437 | @InProceedings{HerlihyLM02 |
| 438 | ,author={Maurice Herlihy and Victor Luchangco and Mark Moir} |
| 439 | ,title="The Repeat Offender Problem: A Mechanism for Supporting Dynamic-Sized, |
| 440 | Lock-Free Data Structures" |
| 441 | ,booktitle={Proceedings of 16\textsuperscript{th} International |
| 442 | Symposium on Distributed Computing} |
| 443 | ,year=2002 |
| 444 | ,month="October" |
| 445 | ,pages="339-353" |
| 446 | } |
| 447 | |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 448 | @article{Appavoo03a |
| 449 | ,author="J. Appavoo and K. Hui and C. A. N. Soules and R. W. Wisniewski and |
| 450 | D. M. {Da Silva} and O. Krieger and M. A. Auslander and D. J. Edelsohn and |
| 451 | B. Gamsa and G. R. Ganger and P. McKenney and M. Ostrowski and |
| 452 | B. Rosenburg and M. Stumm and J. Xenidis" |
| 453 | ,title="Enabling Autonomic Behavior in Systems Software With Hot Swapping" |
| 454 | ,Year="2003" |
| 455 | ,Month="January" |
| 456 | ,journal="IBM Systems Journal" |
| 457 | ,volume="42" |
| 458 | ,number="1" |
| 459 | ,pages="60-76" |
| 460 | } |
| 461 | |
| 462 | @Conference{Arcangeli03 |
| 463 | ,Author="Andrea Arcangeli and Mingming Cao and Paul E. McKenney and |
| 464 | Dipankar Sarma" |
| 465 | ,Title="Using Read-Copy Update Techniques for {System V IPC} in the |
| 466 | {Linux} 2.5 Kernel" |
| 467 | ,Booktitle="Proceedings of the 2003 USENIX Annual Technical Conference |
| 468 | (FREENIX Track)" |
| 469 | ,Publisher="USENIX Association" |
| 470 | ,year="2003" |
| 471 | ,month="June" |
| 472 | ,pages="297-310" |
| 473 | } |
| 474 | |
| 475 | @article{McKenney03a |
| 476 | ,author="Paul E. McKenney" |
| 477 | ,title="Using {RCU} in the {Linux} 2.5 Kernel" |
| 478 | ,Year="2003" |
| 479 | ,Month="October" |
| 480 | ,journal="Linux Journal" |
| 481 | ,volume="1" |
| 482 | ,number="114" |
| 483 | ,pages="18-26" |
| 484 | } |
| 485 | |
Paul E. McKenney | a83f1fe | 2005-05-01 08:59:05 -0700 | [diff] [blame] | 486 | @techreport{Friedberg03a |
| 487 | ,author="Stuart A. Friedberg" |
| 488 | ,title="Lock-Free Wild Card Search Data Structure and Method" |
| 489 | ,institution="US Patent and Trademark Office" |
| 490 | ,address="Washington, DC" |
| 491 | ,year="2003" |
| 492 | ,number="US Patent 6,662,184 (contributed under GPL)" |
| 493 | ,month="December" |
| 494 | ,pages="112" |
| 495 | } |
| 496 | |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 497 | @article{McKenney04a |
| 498 | ,author="Paul E. McKenney and Dipankar Sarma and Maneesh Soni" |
| 499 | ,title="Scaling dcache with {RCU}" |
| 500 | ,Year="2004" |
| 501 | ,Month="January" |
| 502 | ,journal="Linux Journal" |
| 503 | ,volume="1" |
| 504 | ,number="118" |
| 505 | ,pages="38-46" |
| 506 | } |
| 507 | |
| 508 | @Conference{McKenney04b |
| 509 | ,Author="Paul E. McKenney" |
| 510 | ,Title="{RCU} vs. Locking Performance on Different {CPUs}" |
| 511 | ,Booktitle="{linux.conf.au}" |
| 512 | ,Month="January" |
| 513 | ,Year="2004" |
| 514 | ,Address="Adelaide, Australia" |
| 515 | ,note="Available: |
| 516 | \url{http://www.linux.org.au/conf/2004/abstracts.html#90} |
| 517 | \url{http://www.rdrop.com/users/paulmck/rclock/lockperf.2004.01.17a.pdf} |
| 518 | [Viewed June 23, 2004]" |
| 519 | } |
| 520 | |
| 521 | @phdthesis{PaulEdwardMcKenneyPhD |
| 522 | ,author="Paul E. McKenney" |
| 523 | ,title="Exploiting Deferred Destruction: |
| 524 | An Analysis of Read-Copy-Update Techniques |
| 525 | in Operating System Kernels" |
| 526 | ,school="OGI School of Science and Engineering at |
| 527 | Oregon Health and Sciences University" |
| 528 | ,year="2004" |
Paul E. McKenney | a83f1fe | 2005-05-01 08:59:05 -0700 | [diff] [blame] | 529 | ,note="Available: |
| 530 | \url{http://www.rdrop.com/users/paulmck/RCU/RCUdissertation.2004.07.14e1.pdf} |
| 531 | [Viewed October 15, 2004]" |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 532 | } |
| 533 | |
| 534 | @Conference{Sarma04c |
| 535 | ,Author="Dipankar Sarma and Paul E. McKenney" |
| 536 | ,Title="Making RCU Safe for Deep Sub-Millisecond Response Realtime Applications" |
| 537 | ,Booktitle="Proceedings of the 2004 USENIX Annual Technical Conference |
| 538 | (FREENIX Track)" |
| 539 | ,Publisher="USENIX Association" |
| 540 | ,year="2004" |
| 541 | ,month="June" |
| 542 | ,pages="182-191" |
| 543 | } |
Paul E. McKenney | a83f1fe | 2005-05-01 08:59:05 -0700 | [diff] [blame] | 544 | |
| 545 | @unpublished{JamesMorris04b |
| 546 | ,Author="James Morris" |
| 547 | ,Title="Recent Developments in {SELinux} Kernel Performance" |
| 548 | ,month="December" |
| 549 | ,year="2004" |
| 550 | ,note="Available: |
| 551 | \url{http://www.livejournal.com/users/james_morris/2153.html} |
| 552 | [Viewed December 10, 2004]" |
| 553 | } |
Paul E. McKenney | dd81eca | 2005-09-10 00:26:24 -0700 | [diff] [blame] | 554 | |
| 555 | @unpublished{PaulMcKenney05a |
| 556 | ,Author="Paul E. McKenney" |
| 557 | ,Title="{[RFC]} {RCU} and {CONFIG\_PREEMPT\_RT} progress" |
| 558 | ,month="May" |
| 559 | ,year="2005" |
| 560 | ,note="Available: |
| 561 | \url{http://lkml.org/lkml/2005/5/9/185} |
| 562 | [Viewed May 13, 2005]" |
| 563 | ,annotation=" |
| 564 | First publication of working lock-based deferred free patches |
| 565 | for the CONFIG_PREEMPT_RT environment. |
| 566 | " |
| 567 | } |
| 568 | |
| 569 | @conference{PaulMcKenney05b |
| 570 | ,Author="Paul E. McKenney and Dipankar Sarma" |
| 571 | ,Title="Towards Hard Realtime Response from the Linux Kernel on SMP Hardware" |
| 572 | ,Booktitle="linux.conf.au 2005" |
| 573 | ,month="April" |
| 574 | ,year="2005" |
| 575 | ,address="Canberra, Australia" |
| 576 | ,note="Available: |
| 577 | \url{http://www.rdrop.com/users/paulmck/RCU/realtimeRCU.2005.04.23a.pdf} |
| 578 | [Viewed May 13, 2005]" |
| 579 | ,annotation=" |
| 580 | Realtime turns into making RCU yet more realtime friendly. |
| 581 | " |
| 582 | } |
Paul E. McKenney | f85d6c7 | 2008-01-25 21:08:25 +0100 | [diff] [blame] | 583 | |
| 584 | @conference{ThomasEHart2006a |
| 585 | ,Author="Thomas E. Hart and Paul E. McKenney and Angela Demke Brown" |
| 586 | ,Title="Making Lockless Synchronization Fast: Performance Implications |
| 587 | of Memory Reclamation" |
| 588 | ,Booktitle="20\textsuperscript{th} {IEEE} International Parallel and |
| 589 | Distributed Processing Symposium" |
| 590 | ,month="April" |
| 591 | ,year="2006" |
| 592 | ,day="25-29" |
| 593 | ,address="Rhodes, Greece" |
| 594 | ,annotation=" |
| 595 | Compares QSBR (AKA "classic RCU"), HPBR, EBR, and lock-free |
| 596 | reference counting. |
| 597 | " |
| 598 | } |
| 599 | |
| 600 | @Conference{PaulEMcKenney2006b |
| 601 | ,Author="Paul E. McKenney and Dipankar Sarma and Ingo Molnar and |
| 602 | Suparna Bhattacharya" |
| 603 | ,Title="Extending RCU for Realtime and Embedded Workloads" |
| 604 | ,Booktitle="{Ottawa Linux Symposium}" |
| 605 | ,Month="July" |
| 606 | ,Year="2006" |
| 607 | ,pages="v2 123-138" |
| 608 | ,note="Available: |
Justin P. Mattock | 0ea6e61 | 2010-07-23 20:51:24 -0700 | [diff] [blame] | 609 | \url{http://www.linuxsymposium.org/2006/index_2006.php} |
Paul E. McKenney | f85d6c7 | 2008-01-25 21:08:25 +0100 | [diff] [blame] | 610 | \url{http://www.rdrop.com/users/paulmck/RCU/OLSrtRCU.2006.08.11a.pdf} |
| 611 | [Viewed January 1, 2007]" |
| 612 | ,annotation=" |
| 613 | Described how to improve the -rt implementation of realtime RCU. |
| 614 | " |
| 615 | } |
| 616 | |
| 617 | @unpublished{PaulEMcKenney2006c |
| 618 | ,Author="Paul E. McKenney" |
| 619 | ,Title="Sleepable {RCU}" |
| 620 | ,month="October" |
| 621 | ,day="9" |
| 622 | ,year="2006" |
| 623 | ,note="Available: |
| 624 | \url{http://lwn.net/Articles/202847/} |
| 625 | Revised: |
| 626 | \url{http://www.rdrop.com/users/paulmck/RCU/srcu.2007.01.14a.pdf} |
| 627 | [Viewed August 21, 2006]" |
| 628 | ,annotation=" |
| 629 | LWN article introducing SRCU. |
| 630 | " |
| 631 | } |
| 632 | |
| 633 | @unpublished{RobertOlsson2006a |
| 634 | ,Author="Robert Olsson and Stefan Nilsson" |
| 635 | ,Title="{TRASH}: A dynamic {LC}-trie and hash data structure" |
| 636 | ,month="August" |
| 637 | ,day="18" |
| 638 | ,year="2006" |
| 639 | ,note="Available: |
| 640 | \url{http://www.nada.kth.se/~snilsson/public/papers/trash/trash.pdf} |
| 641 | [Viewed February 24, 2007]" |
| 642 | ,annotation=" |
| 643 | RCU-protected dynamic trie-hash combination. |
| 644 | " |
| 645 | } |
| 646 | |
| 647 | @unpublished{ThomasEHart2007a |
| 648 | ,Author="Thomas E. Hart and Paul E. McKenney and Angela Demke Brown and Jonathan Walpole" |
| 649 | ,Title="Performance of memory reclamation for lockless synchronization" |
| 650 | ,journal="J. Parallel Distrib. Comput." |
| 651 | ,year="2007" |
| 652 | ,note="To appear in J. Parallel Distrib. Comput. |
| 653 | \url{doi=10.1016/j.jpdc.2007.04.010}" |
| 654 | ,annotation={ |
| 655 | Compares QSBR (AKA "classic RCU"), HPBR, EBR, and lock-free |
| 656 | reference counting. Journal version of ThomasEHart2006a. |
| 657 | } |
| 658 | } |
| 659 | |
| 660 | @unpublished{PaulEMcKenney2007QRCUspin |
| 661 | ,Author="Paul E. McKenney" |
| 662 | ,Title="Using Promela and Spin to verify parallel algorithms" |
| 663 | ,month="August" |
| 664 | ,day="1" |
| 665 | ,year="2007" |
| 666 | ,note="Available: |
| 667 | \url{http://lwn.net/Articles/243851/} |
| 668 | [Viewed September 8, 2007]" |
| 669 | ,annotation=" |
| 670 | LWN article describing Promela and spin, and also using Oleg |
| 671 | Nesterov's QRCU as an example (with Paul McKenney's fastpath). |
| 672 | " |
| 673 | } |
| 674 | |
Paul E. McKenney | 3230075 | 2008-05-12 21:21:05 +0200 | [diff] [blame] | 675 | @unpublished{PaulEMcKenney2007PreemptibleRCU |
| 676 | ,Author="Paul E. McKenney" |
| 677 | ,Title="The design of preemptible read-copy-update" |
| 678 | ,month="October" |
| 679 | ,day="8" |
| 680 | ,year="2007" |
| 681 | ,note="Available: |
| 682 | \url{http://lwn.net/Articles/253651/} |
| 683 | [Viewed October 25, 2007]" |
| 684 | ,annotation=" |
| 685 | LWN article describing the design of preemptible RCU. |
| 686 | " |
| 687 | } |
| 688 | |
| 689 | ######################################################################## |
| 690 | # |
| 691 | # "What is RCU?" LWN series. |
| 692 | # |
| 693 | |
| 694 | @unpublished{PaulEMcKenney2007WhatIsRCUFundamentally |
| 695 | ,Author="Paul E. McKenney and Jonathan Walpole" |
| 696 | ,Title="What is {RCU}, Fundamentally?" |
| 697 | ,month="December" |
| 698 | ,day="17" |
| 699 | ,year="2007" |
| 700 | ,note="Available: |
| 701 | \url{http://lwn.net/Articles/262464/} |
| 702 | [Viewed December 27, 2007]" |
| 703 | ,annotation=" |
| 704 | Lays out the three basic components of RCU: (1) publish-subscribe, |
| 705 | (2) wait for pre-existing readers to complete, and (2) maintain |
| 706 | multiple versions. |
| 707 | " |
| 708 | } |
| 709 | |
| 710 | @unpublished{PaulEMcKenney2008WhatIsRCUUsage |
| 711 | ,Author="Paul E. McKenney" |
| 712 | ,Title="What is {RCU}? Part 2: Usage" |
| 713 | ,month="January" |
| 714 | ,day="4" |
| 715 | ,year="2008" |
| 716 | ,note="Available: |
| 717 | \url{http://lwn.net/Articles/263130/} |
| 718 | [Viewed January 4, 2008]" |
| 719 | ,annotation=" |
| 720 | Lays out six uses of RCU: |
| 721 | 1. RCU is a Reader-Writer Lock Replacement |
| 722 | 2. RCU is a Restricted Reference-Counting Mechanism |
| 723 | 3. RCU is a Bulk Reference-Counting Mechanism |
| 724 | 4. RCU is a Poor Man's Garbage Collector |
| 725 | 5. RCU is a Way of Providing Existence Guarantees |
| 726 | 6. RCU is a Way of Waiting for Things to Finish |
| 727 | " |
| 728 | } |
| 729 | |
| 730 | @unpublished{PaulEMcKenney2008WhatIsRCUAPI |
| 731 | ,Author="Paul E. McKenney" |
| 732 | ,Title="{RCU} part 3: the {RCU} {API}" |
| 733 | ,month="January" |
| 734 | ,day="17" |
| 735 | ,year="2008" |
| 736 | ,note="Available: |
| 737 | \url{http://lwn.net/Articles/264090/} |
| 738 | [Viewed January 10, 2008]" |
| 739 | ,annotation=" |
| 740 | Gives an overview of the Linux-kernel RCU API and a brief annotated RCU |
| 741 | bibliography. |
| 742 | " |
| 743 | } |
| 744 | |
Paul E. McKenney | 4c54005 | 2010-01-14 16:10:57 -0800 | [diff] [blame] | 745 | # |
| 746 | # "What is RCU?" LWN series. |
| 747 | # |
| 748 | ######################################################################## |
| 749 | |
Paul E. McKenney | 3230075 | 2008-05-12 21:21:05 +0200 | [diff] [blame] | 750 | @article{DinakarGuniguntala2008IBMSysJ |
| 751 | ,author="D. Guniguntala and P. E. McKenney and J. Triplett and J. Walpole" |
| 752 | ,title="The read-copy-update mechanism for supporting real-time applications on shared-memory multiprocessor systems with {Linux}" |
| 753 | ,Year="2008" |
| 754 | ,Month="April" |
| 755 | ,journal="IBM Systems Journal" |
| 756 | ,volume="47" |
| 757 | ,number="2" |
| 758 | ,pages="@@-@@" |
| 759 | ,annotation=" |
| 760 | RCU, realtime RCU, sleepable RCU, performance. |
| 761 | " |
| 762 | } |
Paul E. McKenney | 240ebbf | 2009-06-25 09:08:18 -0700 | [diff] [blame] | 763 | |
| 764 | @article{PaulEMcKenney2008RCUOSR |
| 765 | ,author="Paul E. McKenney and Jonathan Walpole" |
| 766 | ,title="Introducing technology into the {Linux} kernel: a case study" |
| 767 | ,Year="2008" |
| 768 | ,journal="SIGOPS Oper. Syst. Rev." |
| 769 | ,volume="42" |
| 770 | ,number="5" |
| 771 | ,pages="4--17" |
| 772 | ,issn="0163-5980" |
| 773 | ,doi={http://doi.acm.org/10.1145/1400097.1400099} |
| 774 | ,publisher="ACM" |
| 775 | ,address="New York, NY, USA" |
| 776 | ,annotation={ |
| 777 | Linux changed RCU to a far greater degree than RCU has changed Linux. |
| 778 | } |
| 779 | } |
| 780 | |
| 781 | @unpublished{PaulEMcKenney2008HierarchicalRCU |
| 782 | ,Author="Paul E. McKenney" |
| 783 | ,Title="Hierarchical {RCU}" |
| 784 | ,month="November" |
| 785 | ,day="3" |
| 786 | ,year="2008" |
| 787 | ,note="Available: |
| 788 | \url{http://lwn.net/Articles/305782/} |
| 789 | [Viewed November 6, 2008]" |
| 790 | ,annotation=" |
| 791 | RCU with combining-tree-based grace-period detection, |
| 792 | permitting it to handle thousands of CPUs. |
| 793 | " |
| 794 | } |
| 795 | |
| 796 | @conference{PaulEMcKenney2009MaliciousURCU |
| 797 | ,Author="Paul E. McKenney" |
| 798 | ,Title="Using a Malicious User-Level {RCU} to Torture {RCU}-Based Algorithms" |
| 799 | ,Booktitle="linux.conf.au 2009" |
| 800 | ,month="January" |
| 801 | ,year="2009" |
| 802 | ,address="Hobart, Australia" |
| 803 | ,note="Available: |
| 804 | \url{http://www.rdrop.com/users/paulmck/RCU/urcutorture.2009.01.22a.pdf} |
| 805 | [Viewed February 2, 2009]" |
| 806 | ,annotation=" |
| 807 | Realtime RCU and torture-testing RCU uses. |
| 808 | " |
| 809 | } |
| 810 | |
| 811 | @unpublished{MathieuDesnoyers2009URCU |
| 812 | ,Author="Mathieu Desnoyers" |
| 813 | ,Title="[{RFC} git tree] Userspace {RCU} (urcu) for {Linux}" |
| 814 | ,month="February" |
| 815 | ,day="5" |
| 816 | ,year="2009" |
| 817 | ,note="Available: |
| 818 | \url{http://lkml.org/lkml/2009/2/5/572} |
| 819 | \url{git://lttng.org/userspace-rcu.git} |
| 820 | [Viewed February 20, 2009]" |
| 821 | ,annotation=" |
| 822 | Mathieu Desnoyers's user-space RCU implementation. |
| 823 | git://lttng.org/userspace-rcu.git |
| 824 | " |
| 825 | } |
| 826 | |
| 827 | @unpublished{PaulEMcKenney2009BloatWatchRCU |
| 828 | ,Author="Paul E. McKenney" |
| 829 | ,Title="{RCU}: The {Bloatwatch} Edition" |
| 830 | ,month="March" |
| 831 | ,day="17" |
| 832 | ,year="2009" |
| 833 | ,note="Available: |
| 834 | \url{http://lwn.net/Articles/323929/} |
| 835 | [Viewed March 20, 2009]" |
| 836 | ,annotation=" |
| 837 | Uniprocessor assumptions allow simplified RCU implementation. |
| 838 | " |
| 839 | } |
Paul E. McKenney | 4c54005 | 2010-01-14 16:10:57 -0800 | [diff] [blame] | 840 | |
| 841 | @unpublished{PaulEMcKenney2009expeditedRCU |
| 842 | ,Author="Paul E. McKenney" |
| 843 | ,Title="[{PATCH} -tip 0/3] expedited 'big hammer' {RCU} grace periods" |
| 844 | ,month="June" |
| 845 | ,day="25" |
| 846 | ,year="2009" |
| 847 | ,note="Available: |
| 848 | \url{http://lkml.org/lkml/2009/6/25/306} |
| 849 | [Viewed August 16, 2009]" |
| 850 | ,annotation=" |
| 851 | First posting of expedited RCU to be accepted into -tip. |
| 852 | " |
| 853 | } |
| 854 | |
| 855 | @unpublished{JoshTriplett2009RPHash |
| 856 | ,Author="Josh Triplett" |
| 857 | ,Title="Scalable concurrent hash tables via relativistic programming" |
| 858 | ,month="September" |
| 859 | ,year="2009" |
| 860 | ,note="Linux Plumbers Conference presentation" |
| 861 | ,annotation=" |
| 862 | RP fun with hash tables. |
| 863 | " |
| 864 | } |
| 865 | |
| 866 | @phdthesis{MathieuDesnoyersPhD |
Paul E. McKenney | 998f2ac | 2010-02-22 17:04:58 -0800 | [diff] [blame] | 867 | , title = "Low-Impact Operating System Tracing" |
Paul E. McKenney | 4c54005 | 2010-01-14 16:10:57 -0800 | [diff] [blame] | 868 | , author = "Mathieu Desnoyers" |
| 869 | , school = "Ecole Polytechnique de Montr\'{e}al" |
| 870 | , month = "December" |
| 871 | , year = 2009 |
Paul E. McKenney | 998f2ac | 2010-02-22 17:04:58 -0800 | [diff] [blame] | 872 | ,note="Available: |
Paul E. McKenney | 1bd22e3 | 2010-02-22 17:05:00 -0800 | [diff] [blame] | 873 | \url{http://www.lttng.org/pub/thesis/desnoyers-dissertation-2009-12.pdf} |
Paul E. McKenney | 998f2ac | 2010-02-22 17:04:58 -0800 | [diff] [blame] | 874 | [Viewed December 9, 2009]" |
Paul E. McKenney | 4c54005 | 2010-01-14 16:10:57 -0800 | [diff] [blame] | 875 | } |