blob: 39ad8f56783a0faded737756afe6688bb6ae76be [file] [log] [blame]
Linus Torvalds1da177e2005-04-16 15:20:36 -07001Read the F-ing Papers!
2
3
4This document describes RCU-related publications, and is followed by
Paul E. McKenneydd81eca2005-09-10 00:26:24 -07005the corresponding bibtex entries. A number of the publications may
6be found at http://www.rdrop.com/users/paulmck/RCU/.
Linus Torvalds1da177e2005-04-16 15:20:36 -07007
8The first thing resembling RCU was published in 1980, when Kung and Lehman
9[Kung80] recommended use of a garbage collector to defer destruction
10of nodes in a parallel binary search tree in order to simplify its
11implementation. This works well in environments that have garbage
Paul E. McKenneyf85d6c72008-01-25 21:08:25 +010012collectors, but most production garbage collectors incur significant
13overhead.
Linus Torvalds1da177e2005-04-16 15:20:36 -070014
15In 1982, Manber and Ladner [Manber82,Manber84] recommended deferring
16destruction until all threads running at that time have terminated, again
17for a parallel binary search tree. This approach works well in systems
18with short-lived threads, such as the K42 research operating system.
19However, Linux has long-lived tasks, so more is needed.
20
21In 1986, Hennessy, Osisek, and Seigh [Hennessy89] introduced passive
22serialization, which is an RCU-like mechanism that relies on the presence
23of "quiescent states" in the VM/XA hypervisor that are guaranteed not
24to be referencing the data structure. However, this mechanism was not
25optimized for modern computer systems, which is not surprising given
26that these overheads were not so expensive in the mid-80s. Nonetheless,
27passive serialization appears to be the first deferred-destruction
28mechanism to be used in production. Furthermore, the relevant patent has
29lapsed, so this approach may be used in non-GPL software, if desired.
30(In contrast, use of RCU is permitted only in software licensed under
31GPL. Sorry!!!)
32
33In 1990, Pugh [Pugh90] noted that explicitly tracking which threads
34were reading a given data structure permitted deferred free to operate
35in the presence of non-terminating threads. However, this explicit
36tracking imposes significant read-side overhead, which is undesirable
37in read-mostly situations. This algorithm does take pains to avoid
38write-side contention and parallelize the other write-side overheads by
39providing a fine-grained locking design, however, it would be interesting
40to see how much of the performance advantage reported in 1990 remains
41in 2004.
42
43At about this same time, Adams [Adams91] described ``chaotic relaxation'',
44where the normal barriers between successive iterations of convergent
45numerical algorithms are relaxed, so that iteration $n$ might use
46data from iteration $n-1$ or even $n-2$. This introduces error,
47which typically slows convergence and thus increases the number of
48iterations required. However, this increase is sometimes more than made
49up for by a reduction in the number of expensive barrier operations,
50which are otherwise required to synchronize the threads at the end
51of each iteration. Unfortunately, chaotic relaxation requires highly
52structured data, such as the matrices used in scientific programs, and
53is thus inapplicable to most data structures in operating-system kernels.
54
55In 1993, Jacobson [Jacobson93] verbally described what is perhaps the
56simplest deferred-free technique: simply waiting a fixed amount of time
57before freeing blocks awaiting deferred free. Jacobson did not describe
58any write-side changes he might have made in this work using SGI's Irix
59kernel. Aju John published a similar technique in 1995 [AjuJohn95].
60This works well if there is a well-defined upper bound on the length of
61time that reading threads can hold references, as there might well be in
62hard real-time systems. However, if this time is exceeded, perhaps due
63to preemption, excessive interrupts, or larger-than-anticipated load,
64memory corruption can ensue, with no reasonable means of diagnosis.
65Jacobson's technique is therefore inappropriate for use in production
66operating-system kernels, except when such kernels can provide hard
67real-time response guarantees for all operations.
68
69Also in 1995, Pu et al. [Pu95a] applied a technique similar to that of Pugh's
70read-side-tracking to permit replugging of algorithms within a commercial
71Unix operating system. However, this replugging permitted only a single
72reader at a time. The following year, this same group of researchers
73extended their technique to allow for multiple readers [Cowan96a].
74Their approach requires memory barriers (and thus pipeline stalls),
75but reduces memory latency, contention, and locking overheads.
76
771995 also saw the first publication of DYNIX/ptx's RCU mechanism
78[Slingwine95], which was optimized for modern CPU architectures,
79and was successfully applied to a number of situations within the
80DYNIX/ptx kernel. The corresponding conference paper appeared in 1998
81[McKenney98].
82
83In 1999, the Tornado and K42 groups described their "generations"
84mechanism, which quite similar to RCU [Gamsa99]. These operating systems
85made pervasive use of RCU in place of "existence locks", which greatly
86simplifies locking hierarchies.
87
882001 saw the first RCU presentation involving Linux [McKenney01a]
89at OLS. The resulting abundance of RCU patches was presented the
90following year [McKenney02a], and use of RCU in dcache was first
91described that same year [Linder02a].
92
Paul E. McKenneyd19720a2006-02-01 03:06:42 -080093Also in 2002, Michael [Michael02b,Michael02a] presented "hazard-pointer"
94techniques that defer the destruction of data structures to simplify
95non-blocking synchronization (wait-free synchronization, lock-free
96synchronization, and obstruction-free synchronization are all examples of
97non-blocking synchronization). In particular, this technique eliminates
98locking, reduces contention, reduces memory latency for readers, and
99parallelizes pipeline stalls and memory latency for writers. However,
100these techniques still impose significant read-side overhead in the
101form of memory barriers. Researchers at Sun worked along similar lines
Paul E. McKenneyf85d6c72008-01-25 21:08:25 +0100102in the same timeframe [HerlihyLM02]. These techniques can be thought
103of as inside-out reference counts, where the count is represented by the
104number of hazard pointers referencing a given data structure (rather than
105the more conventional counter field within the data structure itself).
106
107By the same token, RCU can be thought of as a "bulk reference count",
108where some form of reference counter covers all reference by a given CPU
109or thread during a set timeframe. This timeframe is related to, but
110not necessarily exactly the same as, an RCU grace period. In classic
111RCU, the reference counter is the per-CPU bit in the "bitmask" field,
112and each such bit covers all references that might have been made by
113the corresponding CPU during the prior grace period. Of course, RCU
114can be thought of in other terms as well.
Linus Torvalds1da177e2005-04-16 15:20:36 -0700115
116In 2003, the K42 group described how RCU could be used to create
Paul E. McKenneyf85d6c72008-01-25 21:08:25 +0100117hot-pluggable implementations of operating-system functions [Appavoo03a].
118Later that year saw a paper describing an RCU implementation of System
119V IPC [Arcangeli03], and an introduction to RCU in Linux Journal
120[McKenney03a].
Linus Torvalds1da177e2005-04-16 15:20:36 -0700121
1222004 has seen a Linux-Journal article on use of RCU in dcache
123[McKenney04a], a performance comparison of locking to RCU on several
124different CPUs [McKenney04b], a dissertation describing use of RCU in a
Paul E. McKenneya83f1fe2005-05-01 08:59:05 -0700125number of operating-system kernels [PaulEdwardMcKenneyPhD], a paper
126describing how to make RCU safe for soft-realtime applications [Sarma04c],
127and a paper describing SELinux performance with RCU [JamesMorris04b].
Linus Torvalds1da177e2005-04-16 15:20:36 -0700128
Paul E. McKenneyf85d6c72008-01-25 21:08:25 +01001292005 brought further adaptation of RCU to realtime use, permitting
Paul E. McKenneydd81eca2005-09-10 00:26:24 -0700130preemption of RCU realtime critical sections [PaulMcKenney05a,
131PaulMcKenney05b].
132
Paul E. McKenneyf85d6c72008-01-25 21:08:25 +01001332006 saw the first best-paper award for an RCU paper [ThomasEHart2006a],
134as well as further work on efficient implementations of preemptible
135RCU [PaulEMcKenney2006b], but priority-boosting of RCU read-side critical
136sections proved elusive. An RCU implementation permitting general
137blocking in read-side critical sections appeared [PaulEMcKenney2006c],
138Robert Olsson described an RCU-protected trie-hash combination
139[RobertOlsson2006a].
140
141
Linus Torvalds1da177e2005-04-16 15:20:36 -0700142Bibtex Entries
143
144@article{Kung80
145,author="H. T. Kung and Q. Lehman"
146,title="Concurrent Maintenance of Binary Search Trees"
147,Year="1980"
148,Month="September"
149,journal="ACM Transactions on Database Systems"
150,volume="5"
151,number="3"
152,pages="354-382"
153}
154
155@techreport{Manber82
156,author="Udi Manber and Richard E. Ladner"
157,title="Concurrency Control in a Dynamic Search Structure"
158,institution="Department of Computer Science, University of Washington"
159,address="Seattle, Washington"
160,year="1982"
161,number="82-01-01"
162,month="January"
163,pages="28"
164}
165
166@article{Manber84
167,author="Udi Manber and Richard E. Ladner"
168,title="Concurrency Control in a Dynamic Search Structure"
169,Year="1984"
170,Month="September"
171,journal="ACM Transactions on Database Systems"
172,volume="9"
173,number="3"
174,pages="439-455"
175}
176
177@techreport{Hennessy89
178,author="James P. Hennessy and Damian L. Osisek and Joseph W. {Seigh II}"
179,title="Passive Serialization in a Multitasking Environment"
180,institution="US Patent and Trademark Office"
181,address="Washington, DC"
182,year="1989"
183,number="US Patent 4,809,168 (lapsed)"
184,month="February"
185,pages="11"
186}
187
188@techreport{Pugh90
189,author="William Pugh"
190,title="Concurrent Maintenance of Skip Lists"
191,institution="Institute of Advanced Computer Science Studies, Department of Computer Science, University of Maryland"
192,address="College Park, Maryland"
193,year="1990"
194,number="CS-TR-2222.1"
195,month="June"
196}
197
198@Book{Adams91
199,Author="Gregory R. Adams"
200,title="Concurrent Programming, Principles, and Practices"
201,Publisher="Benjamin Cummins"
202,Year="1991"
203}
204
205@unpublished{Jacobson93
206,author="Van Jacobson"
207,title="Avoid Read-Side Locking Via Delayed Free"
208,year="1993"
209,month="September"
210,note="Verbal discussion"
211}
212
213@Conference{AjuJohn95
214,Author="Aju John"
215,Title="Dynamic vnodes -- Design and Implementation"
216,Booktitle="{USENIX Winter 1995}"
217,Publisher="USENIX Association"
218,Month="January"
219,Year="1995"
220,pages="11-23"
221,Address="New Orleans, LA"
222}
223
Paul E. McKenneyf85d6c72008-01-25 21:08:25 +0100224@conference{Pu95a,
225Author = "Calton Pu and Tito Autrey and Andrew Black and Charles Consel and
226Crispin Cowan and Jon Inouye and Lakshmi Kethana and Jonathan Walpole and
227Ke Zhang",
228Title = "Optimistic Incremental Specialization: Streamlining a Commercial
229Operating System",
230Booktitle = "15\textsuperscript{th} ACM Symposium on
231Operating Systems Principles (SOSP'95)",
232address = "Copper Mountain, CO",
233month="December",
234year="1995",
235pages="314-321",
236annotation="
237 Uses a replugger, but with a flag to signal when people are
238 using the resource at hand. Only one reader at a time.
239"
240}
241
242@conference{Cowan96a,
243Author = "Crispin Cowan and Tito Autrey and Charles Krasic and
244Calton Pu and Jonathan Walpole",
245Title = "Fast Concurrent Dynamic Linking for an Adaptive Operating System",
246Booktitle = "International Conference on Configurable Distributed Systems
247(ICCDS'96)",
248address = "Annapolis, MD",
249month="May",
250year="1996",
251pages="108",
252isbn="0-8186-7395-8",
253annotation="
254 Uses a replugger, but with a counter to signal when people are
255 using the resource at hand. Allows multiple readers.
256"
257}
258
Linus Torvalds1da177e2005-04-16 15:20:36 -0700259@techreport{Slingwine95
260,author="John D. Slingwine and Paul E. McKenney"
261,title="Apparatus and Method for Achieving Reduced Overhead Mutual
262Exclusion and Maintaining Coherency in a Multiprocessor System
263Utilizing Execution History and Thread Monitoring"
264,institution="US Patent and Trademark Office"
265,address="Washington, DC"
266,year="1995"
267,number="US Patent 5,442,758 (contributed under GPL)"
268,month="August"
269}
270
271@techreport{Slingwine97
272,author="John D. Slingwine and Paul E. McKenney"
273,title="Method for maintaining data coherency using thread
274activity summaries in a multicomputer system"
275,institution="US Patent and Trademark Office"
276,address="Washington, DC"
277,year="1997"
278,number="US Patent 5,608,893 (contributed under GPL)"
279,month="March"
280}
281
282@techreport{Slingwine98
283,author="John D. Slingwine and Paul E. McKenney"
284,title="Apparatus and method for achieving reduced overhead
285mutual exclusion and maintaining coherency in a multiprocessor
286system utilizing execution history and thread monitoring"
287,institution="US Patent and Trademark Office"
288,address="Washington, DC"
289,year="1998"
290,number="US Patent 5,727,209 (contributed under GPL)"
291,month="March"
292}
293
294@Conference{McKenney98
295,Author="Paul E. McKenney and John D. Slingwine"
296,Title="Read-Copy Update: Using Execution History to Solve Concurrency
297Problems"
298,Booktitle="{Parallel and Distributed Computing and Systems}"
299,Month="October"
300,Year="1998"
301,pages="509-518"
302,Address="Las Vegas, NV"
303}
304
305@Conference{Gamsa99
306,Author="Ben Gamsa and Orran Krieger and Jonathan Appavoo and Michael Stumm"
307,Title="Tornado: Maximizing Locality and Concurrency in a Shared Memory
308Multiprocessor Operating System"
309,Booktitle="{Proceedings of the 3\textsuperscript{rd} Symposium on
310Operating System Design and Implementation}"
311,Month="February"
312,Year="1999"
313,pages="87-100"
314,Address="New Orleans, LA"
315}
316
317@techreport{Slingwine01
318,author="John D. Slingwine and Paul E. McKenney"
319,title="Apparatus and method for achieving reduced overhead
320mutual exclusion and maintaining coherency in a multiprocessor
321system utilizing execution history and thread monitoring"
322,institution="US Patent and Trademark Office"
323,address="Washington, DC"
324,year="2001"
325,number="US Patent 5,219,690 (contributed under GPL)"
326,month="April"
327}
328
329@Conference{McKenney01a
330,Author="Paul E. McKenney and Jonathan Appavoo and Andi Kleen and
331Orran Krieger and Rusty Russell and Dipankar Sarma and Maneesh Soni"
332,Title="Read-Copy Update"
333,Booktitle="{Ottawa Linux Symposium}"
334,Month="July"
335,Year="2001"
336,note="Available:
337\url{http://www.linuxsymposium.org/2001/abstracts/readcopy.php}
338\url{http://www.rdrop.com/users/paulmck/rclock/rclock_OLS.2001.05.01c.pdf}
339[Viewed June 23, 2004]"
340annotation="
341Described RCU, and presented some patches implementing and using it in
342the Linux kernel.
343"
344}
345
346@Conference{Linder02a
347,Author="Hanna Linder and Dipankar Sarma and Maneesh Soni"
348,Title="Scalability of the Directory Entry Cache"
349,Booktitle="{Ottawa Linux Symposium}"
350,Month="June"
351,Year="2002"
352,pages="289-300"
353}
354
355@Conference{McKenney02a
356,Author="Paul E. McKenney and Dipankar Sarma and
357Andrea Arcangeli and Andi Kleen and Orran Krieger and Rusty Russell"
358,Title="Read-Copy Update"
359,Booktitle="{Ottawa Linux Symposium}"
360,Month="June"
361,Year="2002"
362,pages="338-367"
363,note="Available:
364\url{http://www.linux.org.uk/~ajh/ols2002_proceedings.pdf.gz}
365[Viewed June 23, 2004]"
366}
367
Paul E. McKenneyf85d6c72008-01-25 21:08:25 +0100368@conference{Michael02a
369,author="Maged M. Michael"
370,title="Safe Memory Reclamation for Dynamic Lock-Free Objects Using Atomic
371Reads and Writes"
372,Year="2002"
373,Month="August"
374,booktitle="{Proceedings of the 21\textsuperscript{st} Annual ACM
375Symposium on Principles of Distributed Computing}"
376,pages="21-30"
377,annotation="
378 Each thread keeps an array of pointers to items that it is
379 currently referencing. Sort of an inside-out garbage collection
380 mechanism, but one that requires the accessing code to explicitly
381 state its needs. Also requires read-side memory barriers on
382 most architectures.
383"
384}
385
386@conference{Michael02b
387,author="Maged M. Michael"
388,title="High Performance Dynamic Lock-Free Hash Tables and List-Based Sets"
389,Year="2002"
390,Month="August"
391,booktitle="{Proceedings of the 14\textsuperscript{th} Annual ACM
392Symposium on Parallel
393Algorithms and Architecture}"
394,pages="73-82"
395,annotation="
396 Like the title says...
397"
398}
399
400@InProceedings{HerlihyLM02
401,author={Maurice Herlihy and Victor Luchangco and Mark Moir}
402,title="The Repeat Offender Problem: A Mechanism for Supporting Dynamic-Sized,
403Lock-Free Data Structures"
404,booktitle={Proceedings of 16\textsuperscript{th} International
405Symposium on Distributed Computing}
406,year=2002
407,month="October"
408,pages="339-353"
409}
410
Linus Torvalds1da177e2005-04-16 15:20:36 -0700411@article{Appavoo03a
412,author="J. Appavoo and K. Hui and C. A. N. Soules and R. W. Wisniewski and
413D. M. {Da Silva} and O. Krieger and M. A. Auslander and D. J. Edelsohn and
414B. Gamsa and G. R. Ganger and P. McKenney and M. Ostrowski and
415B. Rosenburg and M. Stumm and J. Xenidis"
416,title="Enabling Autonomic Behavior in Systems Software With Hot Swapping"
417,Year="2003"
418,Month="January"
419,journal="IBM Systems Journal"
420,volume="42"
421,number="1"
422,pages="60-76"
423}
424
425@Conference{Arcangeli03
426,Author="Andrea Arcangeli and Mingming Cao and Paul E. McKenney and
427Dipankar Sarma"
428,Title="Using Read-Copy Update Techniques for {System V IPC} in the
429{Linux} 2.5 Kernel"
430,Booktitle="Proceedings of the 2003 USENIX Annual Technical Conference
431(FREENIX Track)"
432,Publisher="USENIX Association"
433,year="2003"
434,month="June"
435,pages="297-310"
436}
437
438@article{McKenney03a
439,author="Paul E. McKenney"
440,title="Using {RCU} in the {Linux} 2.5 Kernel"
441,Year="2003"
442,Month="October"
443,journal="Linux Journal"
444,volume="1"
445,number="114"
446,pages="18-26"
447}
448
Paul E. McKenneya83f1fe2005-05-01 08:59:05 -0700449@techreport{Friedberg03a
450,author="Stuart A. Friedberg"
451,title="Lock-Free Wild Card Search Data Structure and Method"
452,institution="US Patent and Trademark Office"
453,address="Washington, DC"
454,year="2003"
455,number="US Patent 6,662,184 (contributed under GPL)"
456,month="December"
457,pages="112"
458}
459
Linus Torvalds1da177e2005-04-16 15:20:36 -0700460@article{McKenney04a
461,author="Paul E. McKenney and Dipankar Sarma and Maneesh Soni"
462,title="Scaling dcache with {RCU}"
463,Year="2004"
464,Month="January"
465,journal="Linux Journal"
466,volume="1"
467,number="118"
468,pages="38-46"
469}
470
471@Conference{McKenney04b
472,Author="Paul E. McKenney"
473,Title="{RCU} vs. Locking Performance on Different {CPUs}"
474,Booktitle="{linux.conf.au}"
475,Month="January"
476,Year="2004"
477,Address="Adelaide, Australia"
478,note="Available:
479\url{http://www.linux.org.au/conf/2004/abstracts.html#90}
480\url{http://www.rdrop.com/users/paulmck/rclock/lockperf.2004.01.17a.pdf}
481[Viewed June 23, 2004]"
482}
483
484@phdthesis{PaulEdwardMcKenneyPhD
485,author="Paul E. McKenney"
486,title="Exploiting Deferred Destruction:
487An Analysis of Read-Copy-Update Techniques
488in Operating System Kernels"
489,school="OGI School of Science and Engineering at
490Oregon Health and Sciences University"
491,year="2004"
Paul E. McKenneya83f1fe2005-05-01 08:59:05 -0700492,note="Available:
493\url{http://www.rdrop.com/users/paulmck/RCU/RCUdissertation.2004.07.14e1.pdf}
494[Viewed October 15, 2004]"
Linus Torvalds1da177e2005-04-16 15:20:36 -0700495}
496
497@Conference{Sarma04c
498,Author="Dipankar Sarma and Paul E. McKenney"
499,Title="Making RCU Safe for Deep Sub-Millisecond Response Realtime Applications"
500,Booktitle="Proceedings of the 2004 USENIX Annual Technical Conference
501(FREENIX Track)"
502,Publisher="USENIX Association"
503,year="2004"
504,month="June"
505,pages="182-191"
506}
Paul E. McKenneya83f1fe2005-05-01 08:59:05 -0700507
508@unpublished{JamesMorris04b
509,Author="James Morris"
510,Title="Recent Developments in {SELinux} Kernel Performance"
511,month="December"
512,year="2004"
513,note="Available:
514\url{http://www.livejournal.com/users/james_morris/2153.html}
515[Viewed December 10, 2004]"
516}
Paul E. McKenneydd81eca2005-09-10 00:26:24 -0700517
518@unpublished{PaulMcKenney05a
519,Author="Paul E. McKenney"
520,Title="{[RFC]} {RCU} and {CONFIG\_PREEMPT\_RT} progress"
521,month="May"
522,year="2005"
523,note="Available:
524\url{http://lkml.org/lkml/2005/5/9/185}
525[Viewed May 13, 2005]"
526,annotation="
527 First publication of working lock-based deferred free patches
528 for the CONFIG_PREEMPT_RT environment.
529"
530}
531
532@conference{PaulMcKenney05b
533,Author="Paul E. McKenney and Dipankar Sarma"
534,Title="Towards Hard Realtime Response from the Linux Kernel on SMP Hardware"
535,Booktitle="linux.conf.au 2005"
536,month="April"
537,year="2005"
538,address="Canberra, Australia"
539,note="Available:
540\url{http://www.rdrop.com/users/paulmck/RCU/realtimeRCU.2005.04.23a.pdf}
541[Viewed May 13, 2005]"
542,annotation="
543 Realtime turns into making RCU yet more realtime friendly.
544"
545}
Paul E. McKenneyf85d6c72008-01-25 21:08:25 +0100546
547@conference{ThomasEHart2006a
548,Author="Thomas E. Hart and Paul E. McKenney and Angela Demke Brown"
549,Title="Making Lockless Synchronization Fast: Performance Implications
550of Memory Reclamation"
551,Booktitle="20\textsuperscript{th} {IEEE} International Parallel and
552Distributed Processing Symposium"
553,month="April"
554,year="2006"
555,day="25-29"
556,address="Rhodes, Greece"
557,annotation="
558 Compares QSBR (AKA "classic RCU"), HPBR, EBR, and lock-free
559 reference counting.
560"
561}
562
563@Conference{PaulEMcKenney2006b
564,Author="Paul E. McKenney and Dipankar Sarma and Ingo Molnar and
565Suparna Bhattacharya"
566,Title="Extending RCU for Realtime and Embedded Workloads"
567,Booktitle="{Ottawa Linux Symposium}"
568,Month="July"
569,Year="2006"
570,pages="v2 123-138"
571,note="Available:
572\url{http://www.linuxsymposium.org/2006/view_abstract.php?content_key=184}
573\url{http://www.rdrop.com/users/paulmck/RCU/OLSrtRCU.2006.08.11a.pdf}
574[Viewed January 1, 2007]"
575,annotation="
576 Described how to improve the -rt implementation of realtime RCU.
577"
578}
579
580@unpublished{PaulEMcKenney2006c
581,Author="Paul E. McKenney"
582,Title="Sleepable {RCU}"
583,month="October"
584,day="9"
585,year="2006"
586,note="Available:
587\url{http://lwn.net/Articles/202847/}
588Revised:
589\url{http://www.rdrop.com/users/paulmck/RCU/srcu.2007.01.14a.pdf}
590[Viewed August 21, 2006]"
591,annotation="
592 LWN article introducing SRCU.
593"
594}
595
596@unpublished{RobertOlsson2006a
597,Author="Robert Olsson and Stefan Nilsson"
598,Title="{TRASH}: A dynamic {LC}-trie and hash data structure"
599,month="August"
600,day="18"
601,year="2006"
602,note="Available:
603\url{http://www.nada.kth.se/~snilsson/public/papers/trash/trash.pdf}
604[Viewed February 24, 2007]"
605,annotation="
606 RCU-protected dynamic trie-hash combination.
607"
608}
609
610@unpublished{ThomasEHart2007a
611,Author="Thomas E. Hart and Paul E. McKenney and Angela Demke Brown and Jonathan Walpole"
612,Title="Performance of memory reclamation for lockless synchronization"
613,journal="J. Parallel Distrib. Comput."
614,year="2007"
615,note="To appear in J. Parallel Distrib. Comput.
616 \url{doi=10.1016/j.jpdc.2007.04.010}"
617,annotation={
618 Compares QSBR (AKA "classic RCU"), HPBR, EBR, and lock-free
619 reference counting. Journal version of ThomasEHart2006a.
620}
621}
622
623@unpublished{PaulEMcKenney2007QRCUspin
624,Author="Paul E. McKenney"
625,Title="Using Promela and Spin to verify parallel algorithms"
626,month="August"
627,day="1"
628,year="2007"
629,note="Available:
630\url{http://lwn.net/Articles/243851/}
631[Viewed September 8, 2007]"
632,annotation="
633 LWN article describing Promela and spin, and also using Oleg
634 Nesterov's QRCU as an example (with Paul McKenney's fastpath).
635"
636}
637