source: doc/papers/concurrency/mail2 @ 76d73fc

ADTarm-ehast-experimentalenumforall-pointer-decayjacob/cs343-translationnew-ast-unique-exprpthread-emulationqualifiedEnum
Last change on this file since 76d73fc was 76d73fc, checked in by Peter A. Buhr <pabuhr@…>, 4 years ago

add additional email messages from SP&E about publishing concurrency paper

  • Property mode set to 100644
File size: 80.4 KB
Line 
1
2Date: Wed, 26 Jun 2019 20:12:38 +0000
3From: Aaron Thomas <onbehalfof@manuscriptcentral.com>
4Reply-To: speoffice@wiley.com
5To: tdelisle@uwaterloo.ca, pabuhr@uwaterloo.ca
6Subject: SPE-19-0219 successfully submitted
7
826-Jun-2019
9
10Dear Dr Buhr,
11
12Your manuscript entitled "Advanced Control-flow and Concurrency in Cforall" has been received by Software: Practice and Experience. It will be given full consideration for publication in the journal.
13
14Your manuscript number is SPE-19-0219.  Please mention this number in all future correspondence regarding this submission.
15
16You can view the status of your manuscript at any time by checking your Author Center after logging into https://mc.manuscriptcentral.com/spe.  If you have difficulty using this site, please click the 'Get Help Now' link at the top right corner of the site.
17
18
19Thank you for submitting your manuscript to Software: Practice and Experience.
20
21Sincerely,
22
23Software: Practice and Experience Editorial Office
24
25
26
27Date: Tue, 12 Nov 2019 22:25:17 +0000
28From: Richard Jones <onbehalfof@manuscriptcentral.com>
29Reply-To: R.E.Jones@kent.ac.uk
30To: tdelisle@uwaterloo.ca, pabuhr@uwaterloo.ca
31Subject: Software: Practice and Experience - Decision on Manuscript ID
32 SPE-19-0219
33
3412-Nov-2019
35
36Dear Dr Buhr,
37
38Many thanks for submitting SPE-19-0219 entitled "Advanced Control-flow and Concurrency in Cforall" to Software: Practice and Experience. The paper has now been reviewed and the comments of the referees are included at the bottom of this letter.
39
40The decision on this paper is that it requires substantial further work is required. The referees have a number of substantial concerns. All the reviewers found the submission very hard to read; two of the reviewers state that it needs very substantial restructuring. These concerns must be addressed before your submission can be considered further.
41
42A revised version of your manuscript that takes into account the comments of the referees will be reconsidered for publication.
43
44Please note that submitting a revision of your manuscript does not guarantee eventual acceptance, and that your revision will be subject to re-review by the referees before a decision is rendered.
45
46You have 90 days from the date of this email to submit your revision. If you are unable to complete the revision within this time, please contact me to request an extension.
47
48You can upload your revised manuscript and submit it through your Author Center. Log into https://mc.manuscriptcentral.com/spe  and enter your Author Center, where you will find your manuscript title listed under "Manuscripts with Decisions".
49
50When submitting your revised manuscript, you will be able to respond to the comments made by the referee(s) in the space provided.  You can use this space to document any changes you make to the original manuscript.
51
52If you feel that your paper could benefit from English language polishing, you may wish to consider having your paper professionally edited for English language by a service such as Wiley's at http://wileyeditingservices.com. Please note that while this service will greatly improve the readability of your paper, it does not guarantee acceptance of your paper by the journal.
53 
54Once again, thank you for submitting your manuscript to Software: Practice and Experience and I look forward to receiving your revision.
55
56
57Sincerely,
58
59Prof. Richard Jones
60Software: Practice and Experience
61R.E.Jones@kent.ac.uk
62
63
64Referee(s)' Comments to Author:
65
66Reviewing: 1
67
68Comments to the Author
69This article presents the design and rationale behind the various
70threading and synchronization mechanisms of C-forall, a new low-level
71programming language.  This paper is very similar to a companion paper
72which I have also received: as the papers are similar, so will these
73reviews be --- in particular any general comments from the other
74review apply to this paper also.
75
76As far as I can tell, the article contains three main ideas: an
77asynchronous execution / threading model; a model for monitors to
78provide mutual exclusion; and an implementation.  The first two ideas
79are drawn together in Table 1: unfortunately this is on page 25 of 30
80pages of text. Implementation choices and descriptions are scattered
81throughout the paper - and the sectioning of the paper seems almost
82arbitrary.
83
84The article is about its contributions.  Simply adding feature X to
85language Y isn't by itself a contribution, (when feature X isn't
86already a contribution).  The contribution can be in the design: the
87motivation, the space of potential design options, the particular
88design chosen and the rationale for that choice, or the resulting
89performance.  For example: why support two kinds of generators as well
90as user-level threads?  Why support both low and high level
91synchronization constructs?  Similarly I would have found the article
92easier to follow if it was written top down, presenting the design
93principles, present the space of language features, justify chosen
94language features (and rationale) and those excluded, and then present
95implementation, and performance.
96
97Then the writing of the article is often hard to follow, to say the
98least. Two examples: section 3 "stateful functions" - I've some idea
99what that is (a function with Algol's "own" or C's "static" variables?
100but in fact the paper has a rather more specific idea than that. The
101top of page 3 throws a whole lot of defintions at the reader
102"generator" "coroutine" "stackful" "stackless" "symmetric"
103"asymmetric" without every stopping to define each one --- but then in
104footnote "C" takes the time to explain what C's "main" function is?  I
105cannot imagine a reader of this paper who doesn't know what "main" is
106in C; especially if they understand the other concepts already
107presented in the paper.  The start of section 3 then does the same
108thing: putting up a whole lot of definitions, making distinctions and
109comparisons, even talking about some runtime details, but the critical
110definition of a monitor doesn't appear until three pages later, at the
111start of section 5 on p15, lines 29-34 are a good, clear, description
112of what a monitor actually is.  That needs to come first, rather than
113being buried again after two sections of comparisons, discussions,
114implementations, and options that are ungrounded because they haven't
115told the reader what they are actually talking about.  First tell the
116reader what something is, then how they might use it (as programmers:
117what are the rules and restrictions) and only then start comparison
118with other things, other approaches, other languages, or
119implementations.
120
121The description of the implementation is similarly lost in the trees
122without ever really seeing the wood. Figure 19 is crucial here, but
123it's pretty much at the end of the paper, and comments about
124implementations are threaded throughout the paper without the context
125(fig 19) to understand what's going on.   The protocol for performance
126testing may just about suffice for C (although is N constantly ten
127million, or does it vary for each benchmark) but such evaluation isn't
128appropriate for garbage-collected or JITTed languages like Java or Go.
129
130other comments working through the paper - these are mostly low level
131and are certainly not comprehensive.
132
133p1 only a subset of C-forall extensions?
134
135p1 "has features often associated with object-oriented programming
136languages, such as constructors, destructors, virtuals and simple
137inheritance."   There's no need to quibble about this. Once a language
138has inheritance, it's hard to claim it's not object-oriented.
139
140
141p2 barging? signals-as-hints?
142
143p3 start your discussion of generations with a simple example of a
144C-forall generator.  Fig 1(b) might do: but put it inline instead of
145the python example - and explain the key rules and restrictions on the
146construct.  Then don't even start to compare with coroutines until
147you've presented, described and explained your coroutines...
148p3 I'd probably leave out the various "C" versions unless there are
149key points to make you can't make in C-forall. All the alternatives
150are just confusing.
151
152
153p4 but what's that "with" in Fig 1(B)
154
155p5 start with the high level features of C-forall generators...
156
157p5 why is the paper explaining networking protocols?
158
159p7 lines 1-9 (transforming generator to coroutine - why would I do any
160of this? Why would I want one instead of the other (do not use "stack"
161in your answer!)
162
163p10 last para "A coroutine must retain its last resumer to suspend
164back because the resumer is on a different stack. These reverse
165pointers allow suspend to cycle backwards, "  I've no idea what is
166going on here?  why should I care?  Shouldn't I just be using threads
167instead?  why not?
168
169p16 for the same reasons - what reasons?
170
171p17 if the multiple-monitor entry procedure really is novel, write a
172paper about that, and only about that.
173
174p23 "Loose Object Definitions" - no idea what that means.  in that
175section: you can't leave out JS-style dynamic properties.  Even in
176OOLs that (one way or another) allow separate definitions of methods
177(like Objective-C, Swift, Ruby, C#) at any time a runtime class has a
178fixed definition.  Quite why the detail about bit mask implementation
179is here anyway, I've no idea.
180
181p25 this cluster isn't a CLU cluster then?
182
183* conclusion should conclude the paper, not the related.
184
185
186Reviewing: 2
187
188Comments to the Author
189This paper describes the concurrency features of an extension of C (whose name I will write as "C\/" here, for convenience), including much design-level discussion of the coroutine- and monitor-based features and some microbenchmarks exploring the current implementation's performance. The key message of the latter is that the system's concurrency abstractions are much lighter-weight than the threading found in mainstream C or Java implementations.
190
191There is much description of the system and its details, but nothing about (non-artificial) uses of it. Although the microbenchmark data is encouraging, arguably not enough practical experience with the system has been reported here to say much about either its usability advantages or its performance.
192
193As such, the main contribution of the paper seem to be to document the existence of the described system and to provide a detailed design rationale and (partial) tutorial. I believe that could be of interest to some readers, so an acceptable manuscript is lurking in here somewhere.
194
195Unfortunately, at present the writing style is somewhere between unclear and infuriating. It omits to define terms; it uses needlessly many terms for what are apparently (but not clearly) the same things; it interrupts itself rather than deliver the natural consequent of whatever it has just said; and so on. Section 5 is particularly bad in these regards -- see my detailed comments below. Fairly major additional efforts will be needed to turn the present text into a digestible design-and-tutorial document. I suspect that a shorter paper could do this job better than the present manuscript, which is overwrought in parts.
196
197p2: lines 4--9 are a little sloppy. It is not the languages but their popular implementations which "adopt" the 1:1 kernel threading model.
198
199line 10: "medium work" -- "medium-sized work"?
200
201line 18: "is all sequential to the compiler" -- not true in modern compilers, and in 2004 H-J Boehm wrote a tech report describing exactly why ("Threads cannot be implemented as a library", HP Labs).
202
203line 20: "knows the optimization boundaries" -- I found this vague. What's an example?
204
205line 31: this paragraph has made a lot of claims. Perhaps forward-reference to the parts of the paper that discuss each one.
206
207line 33: "so the reader can judge if" -- this reads rather passive-aggressively. Perhaps better: "... to support our argument that..."
208
209line 41: "a dynamic partitioning mechanism" -- I couldn't tell what this meant
210
211p3. Presenting concept of a "stateful function" as a new language feature seems odd. In C, functions often have local state thanks to static local variables (or globals, indeed). Of course, that has several limitations. Can you perhaps present your contributions by enumerating these limitations? See also my suggestion below about a possible framing centred on a strawman.
212
213line 2: "an old idea that is new again" -- this is too oblique
214
215lines 2--15: I found this to be a word/concept soup. Stacks, closures, generators, stackless stackful, coroutine, symmetric, asymmetric, resume/suspend versus resume/resume... there needs to be a more gradual and structured way to introduce all this, and ideally one that minimises redundancy. Maybe present it as a series of "definitions" each with its own heading, e.g. "A closure is stackless if its local state has statically known fixed size"; "A generator simply means a stackless closure." And so on. Perhaps also strongly introduce the word "activate" as a direct contrast with resume and suspend. These are just a flavour of the sort of changes that might make this paragraph into something readable.
216
217Continuing the thought: I found it confusing that by these definitinos, a stackful closure is not a stack, even though logically the stack *is* a kind of closure (it is a representation of the current thread's continuation).
218
219lines 24--27: without explaining what the boost functor types mean, I don't think the point here comes across.
220
221line 34: "semantically coupled" -- I wasn't surew hat this meant
222
223p4: the point of Figure 1 (C) was not immediately clear. It seem to be showing how one might "compile down" Figure 1 (B). Or is that Figure 1 (A)?
224
225It's right that the incidental language features of the system are not front-and-centre, but I'd appreciate some brief glossing of non-C languages features as they appear. Examples are the square bracket notation, the pipe notation and the constructor syntax. These explanations could go in the caption of the figure which first uses them, perhaps. Overall I found the figure captions to be terse, and a missed opportunity to explain clearly what was going on.
226
227p5 line 23: "This restriction is removed..." -- give us some up-front summary of your contributions and the elements of the language design that will be talked about, so that this isn't an aside. This will reduce the "twisty passages" feeling that characterises much of the paper.
228
229line 40: "a killer asymmetric generator" -- this is stylistically odd, and the sentence about failures doesn't convincigly argue that C\/ will help with them. Have you any experience writing device drivers using C\/? Or any argument that the kinds of failures can be traced to the "stack-ripping" style that one is forced to use without coroutines? Also, a typo on line 41: "device drives". And saying "Windows/Linux" is sloppy... what does the cited paper actually say?
230
231p6 lines 13--23: this paragraph is difficult to understand. It seems to be talking about a control-flow pattern roughly equivalent to tail recursion. What is the high-level point, other than that this is possible?
232
233line 34: "which they call coroutines" -- a better way to make this point is presumably that the C++20 proposal only provides a specialised kind of coroutine, namely generators, despite its use of the more general word.
234
235line 47: "... due to dynamic stack allocation, execution..." -- this sentence doesn't scan. I suggest adding "and for" in the relevant places where currently there are only commas.
236
237p8 / Figure 5 (B) -- the GNU C extension of unary "&&" needs to be explained. The whole figure needs a better explanation, in fact.
238
239p9, lines 1--10: I wasn't sure this stepping-through really added much value. What are the truly important points to note about this code?
240
241p10: similarly, lines 3--27 again are somewhere between tedious and confusing. I'm sure the motivation and details of "starter semantics" can both be stated much more pithily.
242
243line 32: "a self-resume does not overwrite the last resumer" -- is this a hack or a defensible principled decision?
244
245p11: "a common source of errors" -- among beginners or among production code? Presumably the former.
246
247line 23: "with builtin and library" -- not sure what this means
248
249lines 31--36: these can be much briefer. The only important point here seems to be that coroutines cannot be copied.
250
251p12: line 1: what is a "task"? Does it matter?
252
253line 7: calling it "heap stack" seems to be a recipe for confusion. "Stack-and-heap" might be better, and contrast with "stack-and-VLS" perhaps. When "VLS" is glossed, suggest actually expanding its initials: say "length" not "size".
254
255line 21: are you saying "cooperative threading" is the same as "non-preemptive scheduling", or that one is a special case (kind) of the other? Both are defensible, but be clear.
256
257line 27: "mutual exclusion and synchronization" -- the former is a kind of the latter, so I suggest "and other forms of synchronization".
258
259line 30: "can either be a stackless or stackful" -- stray "a", but also, this seems to be switching from generic/background terminology to C\/-specific terminology.
260
261An expositional idea occurs: start the paper with a strawman naive/limited realisation of coroutines -- say, Simon Tatham's popular "Coroutines in C" web page -- and identify point by point what the limitations are and how C\/ overcomes them. Currently the presentation is often flat (lacking motivating contrasts) and backwards (stating solutions before problems). The foregoing approach might fix both of these.
262
263page 13: line 23: it seems a distraction to mention the Python feature here.
264
265p14 line 5: it seems odd to describe these as "stateless" just because they lack shared mutable state. It means the code itself is even more stateful. Maybe the "stack ripping" argument could usefully be given here.
266
267line 16: "too restrictive" -- would be good to have a reference to justify this, or at least give a sense of what the state-of-the-art performance in transactional memory systems is (both software and hardware)
268
269line 22: "simulate monitors" -- what about just *implementing* monitors? isn't that what these systems do? or is the point more about refining them somehow into something more specialised?
270
271p15: sections 4.1 and 4.2 seem adrift and misplaced. Split them into basic parts (which go earlier) and more advanced parts (e.g. barging, which can be explained later).
272
273line 31: "acquire/release" -- misses an opportunity to contrast the monitor's "enter/exit" abstraction with the less structured acquire/release of locks.
274
275p16 line 12: the "implicit" versus "explicit" point is unclear. Is it perhaps about the contract between an opt-in *discipline* and a language-enforced *guarantee*?
276
277line 28: no need to spend ages dithering about which one is default and which one is the explicit qualifier. Tell us what you decided, briefly justify it, and move on.
278
279p17: Figure 11: since the main point seems to be to highlight bulk acquire, include a comment which identifies the line where this is happening.
280
281line 2: "impossible to statically..." -- or dynamically. Doing it dynamically would be perfectly acceptable (locking is a dynamic operation after all)
282
283"guarantees acquisition order is consistent" -- assuming it's done in a single bulk acquire.
284
285p18: section 5.3: the text here is a mess. The explanations of "internal" versus "external" scheduling are unclear, and "signals as hints" is not explained. "... can cause thread starvation" -- means including a while loop, or not doing so? "There are three signalling mechanisms.." but the text does not follow that by telling us what they are. My own scribbled attempt at unpicking the internal/external thing: "threads already in the monitor, albeit waiting, have priority over those trying to enter".
286
287p19: line 3: "empty condition" -- explain that condition variables don't store anything. So being "empty" means that the queue of waiting threads (threads waiting to be signalled that the condition has become true) is empty.
288
289line 6: "... can be transformed into external scheduling..." -- OK, but give some motivation.
290
291p20: line 6: "mechnaism"
292
293lines 16--20: this is dense and can probably only be made clear with an example
294
295p21 line 21: clarify that nested monitor deadlock was describe earlier (in 5.2). (Is the repetition necessary?)
296
297line 27: "locks, and by extension monitors" -- this is true but the "by extension" argument is faulty. It is perfectly possible to use locks as a primitive and build a compositional mechanism out of them, e.g. transactions.
298
299p22 line 2: should say "restructured"
300
301line 33: "Implementing a fast subset check..." -- make clear that the following section explains how to do this. Restructuring the sections themselves could do this, or noting in the text.
302
303p23: line 3: "dynamic member adding, eg, JavaScript" -- needs to say "as permitted in JavaScript", and "dynamically adding members" is stylistically better
304
305p23: line 18: "urgent stack" -- back-reference to where this was explained before
306
307p24 line 7: I did not understand what was more "direct" about "direct communication". Also, what is a "passive monitor" -- just a monitor, given that monitors are passive by design?
308
309line 14 / section 5.9: this table was useful and it (or something like it) could be used much earlier on to set the structure of the rest of the paper. The explanation at present is too brief, e.g. I did not really understand the point about cases 7 and 8.
310
311p25 line 2: instead of casually dropping in a terse explanation for the newly intrdouced term "virtual processor", introduce it properly. Presumably the point is to give a less ambiguous meaning to "thread" by reserving it only for C\/'s green threads.
312
313Table 1: what does "No / Yes" mean?
314
315p26 line 15: "transforms user threads into fibres" -- a reference is needed to explain what "fibres" means... guessing it's in the sense of Adya et al.
316
317line 20: "Microsoft runtime" -- means Windows?
318
319lines 21--26: don't say "interrupt" to mean "signal", especially not without clear introduction. You can use "POSIX signal" to disambiguate from condition variables' "signal".
320
321p27 line 3: "frequency is usually long" -- that's a "time period" or "interval", not a frequency
322
323line 5: the lengthy quotation is not really necessary; just paraphrase the first sentence and move on.
324
325line 20: "to verify the implementation" -- I don't think that means what is intended
326
327Tables in section 7 -- too many significant figures. How many overall runs are described? What is N in each case?
328
329p29 line 2: "to eliminate this cost" -- arguably confusing since nowadays on commodity CPUs most of the benefits of inlining are not to do with call overheads, but from later optimizations enabled as a consequence of the inlining
330
331line 41: "a hierarchy" -- are they a hierarchy? If so, this could be explained earlier. Also, to say these make up "an integrated set... of control-flow features" verges on the tautologous.
332
333p30 line 15: "a common case being web servers and XaaS" -- that's two cases
334
335
336Reviewing: 3
337
338Comments to the Author
339# Cforall review
340
341Overall, I quite enjoyed reading the paper. Cforall has some very interesting ideas. I did have some suggestions that I think would be helpful before final publication. I also left notes on various parts of the paper that I find confusing when reading, in hopes that it may be useful to you.
342
343## Summary
344
345* Expand on the motivations for including both generator and coroutines, vs trying to build one atop the other
346* Expand on the motivations for having Why both symmetric and asymettric coroutines?
347* Comparison to async-await model adopted by other languages
348    * C#, JS
349    * Rust and its async/await model
350* Consider performance comparisons against node.js and Rust frameworks
351* Discuss performance of monitors vs finer-grained memory models and atomic operations found in other languages
352* Why both internal/external scheduling for synchronization?
353
354## Generator/coroutines
355
356In general, this section was clear, but I thought it would be useful to provide a somewhat deeper look into why Cforall opted for the particular combination of features that it offers. I see three main differences from other languages:
357
358* Generators are not exposed as a "function" that returns a generator object, but rather as a kind of struct, with communication happening via mutable state instead of "return values". That is, the generator must be manually resumed and (if I understood) it is expected to store values that can then later be read (perhaps via methods), instead of having a `yield <Expr>` statement that yields up a value explicitly.
359* Both "symmetric" and "asymmetric" generators are supported, instead of only asymmetric.
360* Coroutines (multi-frame generators) are an explicit mechanism.
361
362In most other languages, coroutines are rather built by layering single-frame generators atop one another (e.g., using a mechanism like async-await), and symmetric coroutines are basically not supported. I'd like to see a bit more justification for Cforall including all the above mechanisms -- it seemed like symmetric coroutines were a useful building block for some of the user-space threading and custom scheduler mechanisms that were briefly mentioned later in the paper.
363
364In the discussion of coroutines, I would have expected a bit more of a comparison to the async-await mechanism offered in other languages. Certainly the semantics of async-await in JavaScript implies significantly more overhead (because each async fn is a distinct heap object). [Rust's approach avoids this overhead][zc], however, and might be worthy of a comparison (see the Performance section).
365
366## Locks and threading
367
368### Comparison to atomics overlooks performance
369
370There are several sections in the paper that compare against atomics -- for example, on page 15, the paper shows a simple monitor that encapsulates an integer and compares that to C++ atomics. Later, the paper compares the simplicity of monitors against the `volatile` quantifier from Java. The conclusion in section 8 also revisits this point.
371
372While I agree that monitors are simpler, they are obviously also significantly different from a performance perspective -- the paper doesn't seem to address this at all. It's plausible that (e.g.) the `Aint` monitor type described in the paper can be compiled and mapped to the specialized instructions offered by hardware, but I didn't see any mention of how this would be done. There is also no mention of the more nuanced memory ordering relations offered by C++11 and how one might achieve similar performance characteristics in Cforall (perhaps the answer is that one simply doesn't need to; I think that's defensible, but worth stating explicitly).
373
374### Justification for external scheduling feels lacking
375
376Cforall includes both internal and external scheduling; I found the explanation for the external scheduling mechanism to be lacking in justification. Why include both mechanisms when most languages seem to make do with only internal scheduling? It would be useful to show some scenarios where external scheduling is truly more powerful.
377
378I would have liked to see some more discussion of external scheduling and how it  interacts with software engineering best practices. It seems somewhat similar to AOP in certain regards. It seems to add a bit of "extra semantics" to monitor methods, in that any method may now also become a kind of synchronization point. The "open-ended" nature of this feels like it could easily lead to subtle bugs, particularly when code refactoring occurs (which may e.g. split an existing method into two). This seems particularly true if external scheduling can occur across compilation units -- the paper suggested that this is true, but I wasn't entirely clear.
379
380I would have also appreciated a few more details on how external scheduling is implemented. It seems to me that there must be some sort of "hooks" on mutex methods so that they can detect whether some other function is waiting on them and awaken those blocked threads. I'm not sure how such hooks are inserted, particularly across compilation units. The material in Section 5.6 didn't quite clarify the matter for me. For example, it left me somewhat confused about whether the `f` and `g` functions declared were meant to be local to a translation unit, or shared with other unit.
381
382### Presentation of monitors is somewhat confusing
383
384I found myself confused fairly often in the section on monitors. I'm just going to leave some notes here on places that I got confused in how that it could be useful to you as feedback on writing that might want to be clarified.
385
386To start, I did not realize that the `mutex_opt` notation was a keyword, I thought it was a type annotation. I think this could be called out more explicitly.
387
388Later, in section 5.2, the paper discusses `nomutex` annotations, which initially threw me, as they had not been introduced (now I realize that this paragraph is there to justify why there is no such keyword). The paragraph might be rearranged to make that clearer, perhaps by leading with the choice that Cforall made.
389
390On page 17, the paper states that "acquiring multiple monitors is safe from deadlock", but this could be stated a bit more precisely: acquiring multiple monitors in a bulk-acquire is safe from deadlock (deadlock can still result from nested acquires).
391
392On page 18, the paper states that wait states do not have to be enclosed in loops, as there is no concern of barging. This seems true but there are also other reasons to use loops (e.g., if there are multiple reasons to notify on the same condition). Thus the statement initially surprised me, as barging is only one of many reasons that I typically employ loops around waits.
393
394I did not understand the diagram in Figure 12 for some time. Initially, I thought that it was generic to all monitors, and I could not understand the state space. It was only later that I realized it was specific to your example. Updating the caption from "Monitor scheduling to "Monitor scheduling in the example from Fig 13" might have helped me quite a bit.
395
396I spent quite some time reading the boy/girl dating example (\*) and I admit I found it somewhat confusing. For example, I couldn't tell whether there were supposed to be many "girl" threads executing at once, or if there was only supposed to be one girl and one boy thread executing in a loop. Are the girl/boy threads supposed to invoke the girl/boy methods or vice versa? Surely there is some easier way to set this up? I believe that when reading the paper I convinced myself of how it was supposed to be working, but I'm writing this review some days later, and I find myself confused all over again and not able to easily figure it out.
397
398(\*) as an aside, I would consider modifying the example to some other form of matching, like customers and support personnel.
399
400## Related work
401
402The paper offered a number of comparisons to Go, C#, Scala, and so forth, but seems to have overlooked another recent language, Rust. In many ways, Rust seems to be closest in philosophy to Cforall, so it seems like an odd omission. I already mentioned above that Rust is in the process of shipping [async-await syntax][aa], which is definitely an alternative to the generator/coroutine approach in Cforall (though one with clear pros/cons).
403
404## Performance
405
406In the performance section in particular, you might consider comparing against some of the Rust web servers and threading systems. For example, actix is top of the [single query TechEmpower Framework benchmarks], and tokio is near the top of the [plainthreading benchmarks][pt] (hyper, the top, is more of an HTTP framework, though it is also written in Rust). It would seem worth trying to compare their "context switching" costs as well -- I believe both actix and tokio have a notion of threads that could be readily compared.
407
408Another addition that might be worth considering is to compare against node.js promises, although I think the comparison to process creation is not as clean.
409
410That said, I think that the performance comparison is not a big focus of the paper, so it may not be necessary to add anything to it.
411
412## Authorship of this review
413
414I'm going to sign this review. This review was authored by Nicholas D. Matsakis. In the intrerest of full disclosure, I'm heavily involved in the Rust project, although I dont' think that influenced this review in particular. Feel free to reach out to me for clarifying questions.
415
416## Links
417
418[aa]: https://blog.rust-lang.org/2019/09/30/Async-await-hits-beta.html
419[zc]: https://aturon.github.io/blog/2016/08/11/futures/
420[sq]: https://www.techempower.com/benchmarks/#section=data-r18&hw=ph&test=db
421[pt]: https://www.techempower.com/benchmarks/#section=data-r18&hw=ph&test=plaintext
422
423
424
425Subject: Re: manuscript SPE-19-0219
426To: "Peter A. Buhr" <pabuhr@uwaterloo.ca>
427From: Richard Jones <R.E.Jones@kent.ac.uk>
428Date: Tue, 12 Nov 2019 22:43:55 +0000
429
430Dear Dr Buhr
431
432Your should have received a decision letter on this today. I am sorry that this
433has taken so long. Unfortunately SP&E receives a lot of submissions and getting
434reviewers is a perennial problem.
435
436Regards
437Richard
438
439Peter A. Buhr wrote on 11/11/2019 13:10:
440>     26-Jun-2019
441>     Your manuscript entitled "Advanced Control-flow and Concurrency in Cforall"
442>     has been received by Software: Practice and Experience. It will be given
443>     full consideration for publication in the journal.
444>
445> Hi, it has been over 4 months since submission of our manuscript SPE-19-0219
446> with no response.
447>
448> Currently, I am refereeing a paper for IEEE that already cites our prior SP&E
449> paper and the Master's thesis forming the bases of the SP&E paper under
450> review. Hence our work is apropos and we want to get it disseminates as soon as
451> possible.
452>
453> [3] A. Moss, R. Schluntz, and P. A. Buhr, "Cforall: Adding modern programming
454>      language features to C," Software - Practice and Experience, vol. 48,
455>      no. 12, pp. 2111-2146, 2018.
456>
457> [4] T. Delisle, "Concurrency in C for all," Master's thesis, University of
458>      Waterloo, 2018.  [Online].  Available:
459>      https://uwspace.uwaterloo.ca/bitstream/handle/10012/12888
460
461
462
463Date: Mon, 13 Jan 2020 05:33:15 +0000
464From: Richard Jones <onbehalfof@manuscriptcentral.com>
465Reply-To: R.E.Jones@kent.ac.uk
466To: pabuhr@uwaterloo.ca
467Subject: Revision reminder - SPE-19-0219
468
46913-Jan-2020
470Dear Dr Buhr
471SPE-19-0219
472
473This is a reminder that your opportunity to revise and re-submit your
474manuscript will expire 28 days from now. If you require more time please
475contact me directly and I may grant an extension to this deadline, otherwise
476the option to submit a revision online, will not be available.
477
478I look forward to receiving your revision.
479
480Sincerely,
481
482Prof. Richard Jones
483Editor, Software: Practice and Experience
484https://mc.manuscriptcentral.com/spe
485
486
487
488Date: Wed, 5 Feb 2020 04:22:18 +0000
489From: Aaron Thomas <onbehalfof@manuscriptcentral.com>
490Reply-To: speoffice@wiley.com
491To: tdelisle@uwaterloo.ca, pabuhr@uwaterloo.ca
492Subject: SPE-19-0219.R1 successfully submitted
493
49404-Feb-2020
495
496Dear Dr Buhr,
497
498Your manuscript entitled "Advanced Control-flow and Concurrency in Cforall" has
499been successfully submitted online and is presently being given full
500consideration for publication in Software: Practice and Experience.
501
502Your manuscript number is SPE-19-0219.R1.  Please mention this number in all
503future correspondence regarding this submission.
504
505You can view the status of your manuscript at any time by checking your Author
506Center after logging into https://mc.manuscriptcentral.com/spe.  If you have
507difficulty using this site, please click the 'Get Help Now' link at the top
508right corner of the site.
509
510Thank you for submitting your manuscript to Software: Practice and Experience.
511
512Sincerely,
513Software: Practice and Experience Editorial Office
514
515
516
517Date: Sat, 18 Apr 2020 10:42:13 +0000
518From: Richard Jones <onbehalfof@manuscriptcentral.com>
519Reply-To: R.E.Jones@kent.ac.uk
520To: tdelisle@uwaterloo.ca, pabuhr@uwaterloo.ca
521Subject: Software: Practice and Experience - Decision on Manuscript ID
522 SPE-19-0219.R1
523
52418-Apr-2020
525
526Dear Dr Buhr,
527
528Many thanks for submitting SPE-19-0219.R1 entitled "Advanced Control-flow and Concurrency in Cforall" to Software: Practice and Experience. The paper has now been reviewed and the comments of the referees are included at the bottom of this letter.
529
530I believe that we are making progress here towards a paper that can be published in Software: Practice and Experience.  However the referees still have significant concerns about the paper. The journal's focus is on practice and experience, and one of the the reviewers' concerns remains that your submission should focus the narrative more on the perspective of the programmer than the language designer. I agree that this would strengthen your submission, and I ask you to address this as well as the referees' other comments.
531
532A revised version of your manuscript that takes into account the comments of the referee(s) will be reconsidered for publication.
533
534Please note that submitting a revision of your manuscript does not guarantee eventual acceptance, and that your revision may be subject to re-review by the referees before a decision is rendered.
535
536You have 90 days from the date of this email to submit your revision. If you are unable to complete the revision within this time, please contact me to request a short extension.
537
538You can upload your revised manuscript and submit it through your Author Center. Log into https://mc.manuscriptcentral.com/spe  and enter your Author Center, where you will find your manuscript title listed under "Manuscripts with Decisions".
539
540When submitting your revised manuscript, you will be able to respond to the comments made by the referee(s) in the space provided.  You can use this space to document any changes you make to the original manuscript.
541
542If you would like help with English language editing, or other article preparation support, Wiley Editing Services offers expert help with English Language Editing, as well as translation, manuscript formatting, and figure formatting at www.wileyauthors.com/eeo/preparation. You can also check out our resources for Preparing Your Article for general guidance about writing and preparing your manuscript at www.wileyauthors.com/eeo/prepresources.
543
544Once again, thank you for submitting your manuscript to Software: Practice and Experience and I look forward to receiving your revision.
545
546Sincerely,
547Richard
548
549Prof. Richard Jones
550Software: Practice and Experience
551R.E.Jones@kent.ac.uk
552
553
554Referee(s)' Comments to Author:
555
556Reviewing: 1
557
558Comments to the Author
559(A relatively short second review)
560
561I thank the authors for their revisions and comprehensive response to
562reviewers' comments --- many of my comments have been successfully
563addressed by the revisions.  Here I'll structure my comments around
564the main salient points in that response which I consider would
565benefit from further explanation.
566
567>  Table 1 is moved to the start and explained in detail.
568
569I consider this change makes a significant improvement to the paper,
570laying out the landscape of language features at the start, and thus
571addresses my main concerns about the paper.
572
573I still have a couple of issues --- perhaps the largest is that it's
574still not clear at this point in the paper what some of these options
575are, or crucially how they would be used. I don't know if it's
576possbile to give high-level examples or use cases to be clear about
577these up front - or if that would duplicate too much information from
578later in the paper - either way expanding out the discussion - even if
579just two a couple of sentences for each row - would help me more.  The
580point is not just to define these categories but to ensure the
581readers' understanding of these definitons agrees with that used in
582the paper.
583
584in a little more detail:
585
586 * 1st para section 2 begs the question: why not support each
587   dimension independently, and let the programmer or library designer
588   combiine features?
589
590 * "execution state" seems a relatively low-level description here.
591  I don't think of e.g. the lambda calculus that way. Perhaps it's as
592  good a term as any.
593
594 * Why must there "be language mechanisms to create, block/unblock,
595   and join with a thread"?  There aren't in Smalltalk (although there
596   are in the runtime).  Especially given in Cforall those mechanisms
597   are *implicit* on thread creation and destruction?
598
599 * "Case 1 is a function that borrows storage for its state (stack
600   frame/activation) and a thread from its invoker"
601
602   this much makes perfect sense to me, but I don't understand how a
603   non-stateful, non-theaded function can then retain
604
605   "this state across callees, ie, function local-variables are
606   retained on the stack across calls."
607
608   how can it retain function-local values *across calls* when it
609   doesn't have any functional-local state?
610
611   I'm not sure if I see two separate cases here - rougly equivalent
612   to C functions without static storage, and then C functions *with*
613   static storage. I assumed that was the distinction between cases 1
614   & 3; but perhpas the actual distinction is that 3 has a
615   suspend/resume point, and so the "state" in figure 1 is this
616   component of execution state (viz figs 1 & 2), not the state
617   representing the cross-call variables?
618
619>    but such evaluation isn't appropriate for garbage-collected or JITTed
620   languages like Java or Go.
621
622For JITTed languages in particular, reporting peak performance needs
623to "warm up" the JIT with a number of iterators before beginning
624measurement. Actually for JIT's its even worse: see Edd Barrett et al
625OOPSLA 2017.
626   
627
628
629minor issues:
630
631 * footnote A - I've looked at various other papers & the website to
632   try to understand how "object-oriented" Cforall is - I'm still not
633   sure.  This footnote says Cforall has "virtuals" - presumably
634   virtual functions, i.e. dynamic dispatch - and inheritance: that
635   really is OO as far as I (and most OO people) are concerned.  For
636   example Haskell doesn't have inheritance, so it's not OO; while
637   CLOS (the Common Lisp *Object* System) or things like Cecil and
638   Dylan are considered OO even though they have "multiple function
639   parameters as receivers", lack "lexical binding between a structure
640   and set of functions", and don't have explicit receiver invocation
641   syntax.  Python has receiver syntax, but unlike Java or Smalltalk
642   or C++, method declarations still need to have an explicit "self"
643   receiver parameter.  Seems to me that Go, for example, is
644   more-or-less OO with interfaces, methods, and dynamic dispatch (yes
645   also and an explicit receiver syntax but that's not
646   determiniative); while Rust lacks dynamic dispatch built-in.  C is
647   not OO as a language, but as you say given it supports function
648   pointers with structures, it does support an OO programm style.
649 
650   This is why I again recommend just not buying into this fight: not
651   making any claims about whether Cforall is OO or is not - because
652   as I see it, the rest of the paper doesn't depend on whether
653   Cforall is OO or not.  That said: this is just a recommendation,
654   and I won't quibble over this any further.
655
656 * is a "monitor function" the same as a "mutex function"?
657   if so the paper should pick one term; if not, make the distinction clear.
658
659
660 * "As stated on line 1 because state declarations from the generator
661    type can be moved out of the coroutine type into the coroutine main"
662
663    OK sure, but again: *why* would a programmer want to do that?
664    (Other than, I guess, to show the difference between coroutines &
665    generators?)  Perhaps another way to put this is that the first
666    para of 3.2 gives the disadvantages of coroutines vs-a-vs
667    generators, briefly describes the extended semantics, but never
668    actualy says why a programmer may want those extended semantics,
669    or how they would benefit.  I don't mean to belabour the point,
670    but (generalist?) readers like me would generally benefit from
671    those kinds of discussions about each feature throughout the
672    paper: why might a programmer want to use them?
673   
674
675> p17 if the multiple-monitor entry procedure really is novel, write a paper
676> about that, and only about that.
677
678> We do not believe this is a practical suggestion.
679
680 * I'm honestly not trying to be snide here: I'm not an expert on
681   monitor or concurrent implementations. Brinch Hansen's original
682   monitors were single acquire; this draft does not cite any other
683   previous work that I could see. I'm not suggesting that the brief
684   mention of this mechanism necessarily be removed from this paper,
685   but if this is novel (and a clear advance over a classical OO
686   monitor a-la Java which only acquires the distinguished reciever)
687   then that would be worth another paper in itself.
688 
689> * conclusion should conclude the paper, not the related.
690> We do not understand this comment.if ithis
691
692My typo: the paper's conclusion should come at the end, after the
693future work section.
694
695
696
697
698To encourage accountability, I'm signing my reviews in 2020.
699For the record, I am James Noble, kjx@ecs.vuw.ac.nz.
700
701
702Reviewing: 2
703
704Comments to the Author
705I thank the authors for their detailed response. To respond to a couple of points raised  in response to my review (number 2):
706
707- on the Boehm paper and whether code is "all sequential to the compiler": I now understand the authors' position better and suspect we are in violent agreement, except for whether it's appropriate to use the rather breezy phrase "all sequential to the compiler". It would be straightforward to clarify that code not using the atomics features is optimized *as if* it were sequential, i.e. on the assumption of a lack of data races.
708
709- on the distinction between "mutual exclusion" and "synchronization": the added citation does help, in that it makes a coherent case for the definition the authors prefer. However, the text could usefully clarify that this is a matter of definition not of fact, given especially that in my assessment the authors' preferred definition is not the most common one. (Although the mention of Hoare's apparent use of this definition is one data point, countervailing ones are found in many contemporaneous or later papers, e.g. Habermann's 1972 "Synchronization of Communicating Processes" (CACM 15(3)), Reed & Kanodia's 1979 "Synchronization with eventcounts and sequencers" (CACM (22(2)) and so on.)
710
711I am glad to see that the authors have taken on board most of the straightforward improvements I suggested.
712
713However, a recurring problem of unclear writing still remains through many parts of the paper, including much of sections 2, 3 and 6. To highlight a couple of problem patches (by no means exhaustive):
714
715- section 2 (an expanded version of what was previously section 5.9) lacks examples and is generally obscure and allusory ("the most advanced feature" -- name it! "in triplets" -- there is only one triplet!; what are "execution locations"? "initialize" and "de-initialize" what? "borrowed from the invoker" is a concept in need of explaining or at least a fully explained example -- in what sense does a plain function borrow" its stack frame? "computation only" as opposed to what? in 2.2, in what way is a "request" fundamental to "synchronization"? and the "implicitly" versus "explicitly" point needs stating as elsewhere, with a concrete example e.g. Java built-in mutexes versus java.util.concurrent).
716
717- section 6: 6.2 omits the most important facts in preference for otherwise inscrutable detail: "identify the kind of parameter" (first say *that there are* kinds of parameter, and what "kinds" means!); "mutex parameters are documentation" is misleading (they are also semantically significant!) and fails to say *what* they mean; the most important thing is surely that 'mutex' is a language feature for performing lock/unlock operations at function entry/exit. So say it! The meanings of examples f3 and f4 remain unclear. Meanwhile in 6.3, "urgent" is not introduced (we are supposed to infer its meaning from Figure 12, but that Figure is incomprehensible to me), and we are told of "external scheduling"'s long history in Ada but not clearly what it actually means; 6.4's description of "waitfor" tells us it is different from an if-else chain but tries to use two *different* inputs to tell us that the behavior is different; tell us an instance where *the same* values of C1 and C2 give different behavior (I even wrote out a truth table and still don't see the semantic difference)
718
719The authors frequently use bracketed phrases, and sometimes slashes "/", in ways that are confusing and/or detrimental to readability. Page 13 line 2's "forward (backward)" is one particularly egregious example. In general I would recommend the the authors try to limit their use of parentheses and slashes as a means of forcing a clearer wording to emerge. Also, the use of "eg." is often cursory and does not explain the examples given, which are frequently a one- or two-word phrase of unclear referent.
720
721Considering the revision more broadly, none of the more extensive or creative rewrites I suggested in my previous review have been attempted, nor any equivalent efforts to improve its readability. The hoisting of the former section 5.9 is a good idea, but the newly added material accompanying it (around Table 1) suffers fresh deficiencies in clarity. Overall the paper is longer than before, even though (as my previous review stated), I believe a shorter paper is required in order to serve the likely purpose of publication. (Indeed, the authors' letter implies that a key goal of publication is to build community and gain external users.)
722
723Given this trajectory, I no longer see a path to an acceptable revision of the present submission. Instead I suggest the authors consider splitting the paper in two: one half about coroutines and stack management, the other about mutexes, monitors and the runtime. (A briefer presentation of the runtime may be helpful in the first paper also, and a brief recap of the generator and coroutine support is obviously needed in the second too.) Both of these new papers would need to be written with a strong emphasis on clarity, paying great care to issues of structure, wording, choices of example, and restraint (saying what's important, not everything that could be said). I am confident the authors could benefit from getting early feedback from others at their institution. For the performance experiments, of course these do not split evenly -- most (but not all) belong in the second of these two hypothetical papers. But the first of them would still have plenty of meat to it; for me, a clear and thorough study of the design space around coroutines is the most interesting and tantalizing prospect.
724
725I do not buy the authors' defense of the limited practical experience or "non-micro" benchmarking presented. Yes, gaining external users is hard and I am sympathetic on that point. But building something at least *somewhat* substantial with your own system should be within reach, and without it the "practice and experience" aspects of the work have not been explored. Clearly C\/ is the product of a lot of work over an extended period, so it is a surprise that no such experience is readily available for inclusion.
726
727Some smaller points:
728
729It does not seem right to state that a stack is essential to Von Neumann architectures -- since the earliest Von Neumann machines (and indeed early Fortran) did not use one.
730
731To elaborate on something another reviewer commented on: it is a surprise to find a "Future work" section *after* the "Conclusion" section. A "Conclusions and future work" section often works well.
732
733
734Reviewing: 3
735
736Comments to the Author
737This is the second round of reviewing.
738
739As in the first review, I found that the paper (and Cforall) contains
740a lot of really interesting ideas, but it remains really difficult to
741have a good sense of which idea I should use and when. This applies in
742different ways to different features from the language:
743
744* coroutines/generators/threads: here there is
745  some discussion, but it can be improved.
746* interal/external scheduling: I didn't find any direct comparison
747  between these features, except by way of example.
748
749I requested similar things in my previous review and I see that
750content was added in response to those requests. Unfortunately, I'm
751not sure that I can say it improved the paper's overall read. I think
752in some sense the additions were "too much" -- I would have preferred
753something more like a table or a few paragraphs highlighting the key
754reasons one would pick one construct or the other.
755
756In general, I do wonder if the paper is just trying to do too much.
757The discussion of clusters and pre-emption in particular feels quite
758rushed.
759
760## Summary
761
762I make a number of suggestions below but the two most important
763I think are:
764
765* Recommend to shorten the comparison on coroutine/generator/threads
766  in Section 2 to a paragraph with a few examples, or possibly a table
767  explaining the trade-offs between the constructs
768* Recommend to clarify the relationship between internal/external
769  scheduling -- is one more general but more error-prone or low-level?
770
771## Coroutines/generators/threads
772
773There is obviously a lot of overlap between these features, and in
774particular between coroutines and generators. As noted in the previous
775review, many languages have chosen to offer *only* generators, and to
776build coroutines by stacks of generators invoking one another.
777
778I believe the newly introduced Section 2 of the paper is trying to
779motivate why each of these constructs exist, but I did not find it
780effective. It was dense and difficult to understand. I think the
781problem is that Section 2 seems to be trying to derive "from first
782principles" why each construct exists, but I think that a more "top
783down" approach would be easier to understand.
784
785In fact, the end of Section 2.1 (on page 5) contains a particular
786paragraph that embodies this "top down" approach. It starts,
787"programmers can now answer three basic questions", and thus gives
788some practical advice for which construct you should use and when. I
789think giving some examples of specific applications that this
790paragraph, combined with some examples of cases where each construct
791was needed, would be a better approach.
792
793I don't think this compariosn needs to be very long. It seems clear
794enough that one would
795
796* prefer generators for simple computations that yield up many values,
797* prefer coroutines for more complex processes that have significant
798  internal structure,
799* prefer threads for cases where parallel execution is desired or
800  needed.
801
802I did appreciate the comparison in Section 2.3 between async-await in
803JS/Java and generators/coroutines. I agree with its premise that those
804mechanisms are a poor replacement for generators (and, indeed, JS has
805a distinct generator mechanism, for example, in part for this reason).
806I believe I may have asked for this in a previous review, but having
807read it, I wonder if it is really necessary, since those mechanisms
808are so different in purpose.
809
810## Internal vs external scheduling
811
812I find the motivation for supporting both internal and external
813scheduling to be fairly implicit. After several reads through the
814section, I came to the conclusion that internal scheduling is more
815expressive than external scheduling, but sometimes less convenient or
816clear. Is this correct? If not, it'd be useful to clarify where
817external scheduling is more expressive.
818
819The same is true, I think, of the `signal_block` function, which I
820have not encountered before; it seems like its behavior can be modeled
821with multiple condition variables, but that's clearly more complex.
822
823One question I had about `signal_block`: what happens if one signals
824but no other thread is waiting? Does it block until some other thread
825waits? Or is that user error?
826
827I would find it very interesting to try and capture some of the
828properties that make internal vs external scheduling the better
829choice.
830
831For example, it seems to me that external scheduling works well if
832there are only a few "key" operations, but that internal scheduling
833might be better otherwise, simply because it would be useful to have
834the ability to name a signal that can be referenced by many
835methods. Consider the bounded buffer from Figure 13: if it had
836multiple methods for removing elements, and not just `remove`, then
837the `waitfor(remove)` call in `insert` might not be sufficient.
838
839## Comparison of external scheduling to messaging
840
841I did enjoy the section comparing external scheduling to Go's
842messaging mechanism, which I believe is a new addition.
843
844I believe that one difference between the Go program and the Cforall
845equivalent is that the Goroutine has an associated queue, so that
846multiple messages could be enqueued, whereas the Cforall equivalent is
847effectively a "bounded buffer" of length 1. Is that correct? I think
848this should be stated explicitly. (Presumably, one could modify the
849Cforall program to include an explicit vector of queued messages if
850desired, but you would also be reimplementing the channel
851abstraction.)
852
853Also, in Figure 20, I believe that there is a missing `mutex` keyword.
854The fiugre states:
855
856```
857void main(GoRtn & gortn) with(gortn) {
858```
859
860but I think it should probably be as follows:
861
862```
863void main(GoRtn & mutex gortn) with(gortn) {
864```
865
866Unless there is some implicit `mutex` associated with being a main
867function for a `monitor thread`.
868
869## Atomic operations and race freedom
870
871I was glad to see that the paper acknowledged that Cforall still had
872low-level atomic operations, even if their use is discouraged in favor
873of higher-level alternatives.
874
875However, I still feel that the conclusion overstates the value of the
876contribution here when it says that "Cforall high-level race-free
877monitors and threads provide the core mechanisms for mutual exclusion
878and synchronization, without the need for volatile and atomics". I
879feel confident that Java programmers, for example, would be advised to
880stick with synchronized methods whenever possible, and it seems to me
881that they offer similar advantages -- but they sometimes wind up using
882volatiles for performance reasons.
883
884I was also confused by the term "race-free" in that sentence. In
885particular, I don't think that Cforall has any mechanisms for
886preventing *data races*, and it clearly doesn't prevent "race
887conditions" (which would bar all sorts of useful programs). I suppose
888that "race free" here might be referring to the improvements such as
889removing barging behavior.
890
891## Performance comparisons
892
893In my previous review, I requested comparisons against Rust and
894node.js, and I see that the new version of the paper includes both,
895which is a good addition.
896
897One note on the Rust results: I believe that the results are comparing
898against the threads found in Rust's standard library, which are
899essentially a shallow wrapper around pthreads, and hence the
900performance is quite close to pthread performance (as one would
901expect). It would perhaps be more interesting to see a comparison
902built using [tokio] or [async-std], two of the more prominent
903user-space threading libraries that build on Rust's async-await
904feature (which operates quite differently than Javascript's
905async-await, in that it doesn't cause every aync function call to
906schedule a distinct task).
907
908[tokio]: https://tokio.rs/
909[async-std]: https://async.rs/
910
911That said, I am satisfied with the performance results as they are in
912the current revision.
913
914## Minor notes and typos
915
916Several figures used the `with` keyword. I deduced that `with(foo)`
917permits one to write `bar` instead of `foo.bar`. It seems worth
918introducing. Apologies if this is stated in the paper, if so I missed
919it.
920
921On page 20, section 6.3, "external scheduling and vice versus" should be
922"external scheduling and vice versa".
923
924On page 5, section 2.3, the paper states "we content" but it should be
925"we contend".
926
927Reviewing: Editor
928
929A few small comments in addition to those of the referees.
930
931Page 1. I don't believe that it s fair to imply that Scala is  "research vehicle" as it is used by major players, Twitter being the most prominent example.
932
933Page 15. Must Cforall threads start after construction (e.g. see your example on page 15, line 21)? I can think of examples where it is not desirable that threads start immediately after construction, e.g. a game with N players, each of whom is expensive to create, but all of whom should be started at the same time.
934
935Page 18, line 17: is using
936
937
938
939Date: Tue, 16 Jun 2020 13:45:03 +0000
940From: Aaron Thomas <onbehalfof@manuscriptcentral.com>
941Reply-To: speoffice@wiley.com
942To: tdelisle@uwaterloo.ca, pabuhr@uwaterloo.ca
943Subject: SPE-19-0219.R2 successfully submitted
944
94516-Jun-2020
946
947Dear Dr Buhr,
948
949Your manuscript entitled "Advanced Control-flow and Concurrency in Cforall" has been successfully submitted online and is presently being given full consideration for publication in Software: Practice and Experience.
950
951Your manuscript number is SPE-19-0219.R2.  Please mention this number in all future correspondence regarding this submission.
952
953You can view the status of your manuscript at any time by checking your Author Center after logging into https://mc.manuscriptcentral.com/spe.  If you have difficulty using this site, please click the 'Get Help Now' link at the top right corner of the site.
954
955
956Thank you for submitting your manuscript to Software: Practice and Experience.
957
958Sincerely,
959
960Software: Practice and Experience Editorial Office
961
962
963
964Date: Wed, 2 Sep 2020 20:55:34 +0000
965From: Richard Jones <onbehalfof@manuscriptcentral.com>
966Reply-To: R.E.Jones@kent.ac.uk
967To: tdelisle@uwaterloo.ca, pabuhr@uwaterloo.ca
968Subject: Software: Practice and Experience - Decision on Manuscript ID
969 SPE-19-0219.R2
970
97102-Sep-2020
972
973Dear Dr Buhr,
974
975Many thanks for submitting SPE-19-0219.R2 entitled "Advanced Control-flow and Concurrency in Cforall" to Software: Practice and Experience. The paper has now been reviewed and the comments of the referees are included at the bottom of this letter. I apologise for the length of time it has taken to get these.
976
977Both reviewers consider this paper to be close to acceptance. However, before I can accept this paper, I would like you address the comments of Reviewer 2, particularly with regard to the description of the adaptation Java harness to deal with warmup. I would expect to see a convincing argument that the computation has reached a steady state. I would also like you to provide the values for N for each benchmark run. This should be very straightforward for you to do. There are a couple of papers on steady state that you may wish to consult (though I am certainly not pushing my own work).
978
9791) Barrett, Edd; Bolz-Tereick, Carl Friedrich; Killick, Rebecca; Mount, Sarah and Tratt, Laurence. Virtual Machine Warmup Blows Hot and Cold. OOPSLA 2017. https://doi.org/10.1145/3133876
980Virtual Machines (VMs) with Just-In-Time (JIT) compilers are traditionally thought to execute programs in two phases: the initial warmup phase determines which parts of a program would most benefit from dynamic compilation, before JIT compiling those parts into machine code; subsequently the program is said to be at a steady state of peak performance. Measurement methodologies almost always discard data collected during the warmup phase such that reported measurements focus entirely on peak performance. We introduce a fully automated statistical approach, based on changepoint analysis, which allows us to determine if a program has reached a steady state and, if so, whether that represents peak performance or not. Using this, we show that even when run in the most controlled of circumstances, small, deterministic, widely studied microbenchmarks often fail to reach a steady state of peak performance on a variety of common VMs. Repeating our experiment on 3 different machines, we found that at most 43.5% of pairs consistently reach a steady state of peak performance.
981
9822) Kalibera, Tomas and Jones, Richard. Rigorous Benchmarking in Reasonable Time. ISMM  2013. https://doi.org/10.1145/2555670.2464160
983Experimental evaluation is key to systems research. Because modern systems are complex and non-deterministic, good experimental methodology demands that researchers account for uncertainty. To obtain valid results, they are expected to run many iterations of benchmarks, invoke virtual machines (VMs) several times, or even rebuild VM or benchmark binaries more than once. All this repetition costs time to complete experiments. Currently, many evaluations give up on sufficient repetition or rigorous statistical methods, or even run benchmarks only in training sizes. The results reported often lack proper variation estimates and, when a small difference between two systems is reported, some are simply unreliable.In contrast, we provide a statistically rigorous methodology for repetition and summarising results that makes efficient use of experimentation time. Time efficiency comes from two key observations. First, a given benchmark on a given platform is typically prone to much less non-determinism than the common worst-case of published corner-case studies. Second, repetition is most needed where most uncertainty arises (whether between builds, between executions or between iterations). We capture experimentation cost with a novel mathematical model, which we use to identify the number of repetitions at each level of an experiment necessary and sufficient to obtain a given level of precision.We present our methodology as a cookbook that guides researchers on the number of repetitions they should run to obtain reliable results. We also show how to present results with an effect size confidence interval. As an example, we show how to use our methodology to conduct throughput experiments with the DaCapo and SPEC CPU benchmarks on three recent platforms.
984
985You have 42 days from the date of this email to submit your revision. If you are unable to complete the revision within this time, please contact me to request a short extension.
986
987You can upload your revised manuscript and submit it through your Author Center. Log into https://mc.manuscriptcentral.com/spe and enter your Author Center, where you will find your manuscript title listed under "Manuscripts with Decisions".
988
989When submitting your revised manuscript, you will be able to respond to the comments made by the referee(s) in the space provided.  You can use this space to document any changes you make to the original manuscript.
990
991If you would like help with English language editing, or other article preparation support, Wiley Editing Services offers expert help with English Language Editing, as well as translation, manuscript formatting, and figure formatting at www.wileyauthors.com/eeo/preparation. You can also check out our resources for Preparing Your Article for general guidance about writing and preparing your manuscript at www.wileyauthors.com/eeo/prepresources.
992 
993Once again, thank you for submitting your manuscript to Software: Practice and Experience. I look forward to receiving your revision.
994
995Sincerely,
996Richard
997
998Prof. Richard Jones
999Editor, Software: Practice and Experience
1000R.E.Jones@kent.ac.uk
1001
1002Referee(s)' Comments to Author:
1003
1004Reviewing: 1
1005
1006Comments to the Author
1007Overall, I felt that this draft was an improvement on previous drafts and I don't have further changes to request.
1008
1009I appreciated the new language to clarify the relationship of external and internal scheduling, for example, as well as the new measurements of Rust tokio. Also, while I still believe that the choice between thread/generator/coroutine and so forth could be made crisper and clearer, the current draft of Section 2 did seem adequate to me in terms of specifying the considerations that users would have to take into account to make the choice.
1010
1011
1012Reviewing: 2
1013
1014Comments to the Author
1015First: let me apologise for the delay on this review. I'll blame the global pandemic combined with my institution's senior management's counterproductive decisions for taking up most of my time and all of my energy.
1016
1017At this point, reading the responses, I think we've been around the course enough times that further iteration is unlikely to really improve the paper any further, so I'm happy to recommend acceptance.    My main comments are that there were some good points in the responses to *all* the reviews and I strongly encourage the authors to incorporate those discursive responses into the final paper so they may benefit readers as well as reviewers.   I agree with the recommendations of reviewer #2 that the paper could usefully be split in to two, which I think I made to a previous revision, but I'm happy to leave that decision to the Editor.
1018
1019Finally, the paper needs to describe how the Java harness was adapted to deal with warmup; why the computation has warmed up and reached a steady state - similarly for js and Python. The tables should also give the "N" chosen for each benchmark run.
1020 
1021minor points
1022* don't start sentences with "However"
1023* most downloaded isn't an "Award"
1024
1025
1026
1027Date: Thu, 1 Oct 2020 05:34:29 +0000
1028From: Richard Jones <onbehalfof@manuscriptcentral.com>
1029Reply-To: R.E.Jones@kent.ac.uk
1030To: pabuhr@uwaterloo.ca
1031Subject: Revision reminder - SPE-19-0219.R2
1032
103301-Oct-2020
1034
1035Dear Dr Buhr
1036
1037SPE-19-0219.R2
1038
1039This is a reminder that your opportunity to revise and re-submit your manuscript will expire 14 days from now. If you require more time please contact me directly and I may grant an extension to this deadline, otherwise the option to submit a revision online, will not be available.
1040
1041If your article is of potential interest to the general public, (which means it must be timely, groundbreaking, interesting and impact on everyday society) then please e-mail ejp@wiley.co.uk explaining the public interest side of the research. Wiley will then investigate the potential for undertaking a global press campaign on the article.
1042
1043I look forward to receiving your revision.
1044
1045Sincerely,
1046
1047Prof. Richard Jones
1048Editor, Software: Practice and Experience
1049
1050https://mc.manuscriptcentral.com/spe
1051
1052
1053
1054Date: Tue, 6 Oct 2020 15:29:41 +0000
1055From: Mayank Roy Chowdhury <onbehalfof@manuscriptcentral.com>
1056Reply-To: speoffice@wiley.com
1057To: tdelisle@uwaterloo.ca, pabuhr@uwaterloo.ca
1058Subject: SPE-19-0219.R3 successfully submitted
1059
106006-Oct-2020
1061
1062Dear Dr Buhr,
1063
1064Your manuscript entitled "Advanced Control-flow and Concurrency in Cforall" has been successfully submitted online and is presently being given full consideration for publication in Software: Practice and Experience.
1065
1066Your manuscript number is SPE-19-0219.R3.  Please mention this number in all future correspondence regarding this submission.
1067
1068You can view the status of your manuscript at any time by checking your Author Center after logging into https://mc.manuscriptcentral.com/spe.  If you have difficulty using this site, please click the 'Get Help Now' link at the top right corner of the site.
1069
1070
1071Thank you for submitting your manuscript to Software: Practice and Experience.
1072
1073Sincerely,
1074
1075Software: Practice and Experience Editorial Office
1076
1077
1078
1079Date: Thu, 15 Oct 2020 13:48:52 +0000
1080From: Richard Jones <onbehalfof@manuscriptcentral.com>
1081Reply-To: R.E.Jones@kent.ac.uk
1082To: tdelisle@uwaterloo.ca, pabuhr@uwaterloo.ca
1083Subject: Software: Practice and Experience - Decision on Manuscript ID
1084 SPE-19-0219.R3
1085
108615-Oct-2020
1087
1088Dear Dr Buhr,
1089
1090It is a pleasure to accept your manuscript entitled "Advanced Control-flow and Concurrency in Cforall" in its current form for publication in Software: Practice and Experience. 
1091
1092Please note although the manuscript is accepted the files will now be checked to ensure that everything is ready for publication, and you may be contacted if final versions of files for publication are required.
1093
1094Your article cannot be published until the publisher has received the appropriate signed license agreement. Within the next few days the corresponding author will receive an email from Wiley's Author Services system which will ask them to log in and will present them with the appropriate license for completion.
1095
1096Thank you for your fine contribution.
1097
1098Sincerely,
1099Richard
1100
1101Prof. Richard Jones
1102Editor, Software: Practice and Experience
1103R.E.Jones@kent.ac.uk
1104
1105P.S. - You can help your research get the attention it deserves! Check out Wiley's free Promotion Guide for best-practice recommendations for promoting your work at www.wileyauthors.com/eeo/guide. And learn more about Wiley Editing Services which offers professional video, design, and writing services to create shareable video abstracts, infographics, conference posters, lay summaries, and research news stories for your research at www.wileyauthors.com/eeo/promotion.
1106
1107This journal accepts artwork submissions for Cover Images. This is an optional service you can use to help increase article exposure and showcase your research. For more information, including artwork guidelines, pricing, and submission details, please visit the Journal Cover Image page at www.wileyauthors.com/eeo/covers. If you want help creating an image, Wiley Editing Services offers a professional cover image design service that creates eye-catching images, ready to be showcased on the journal cover at www.wileyauthors.com/eeo/design.
1108
1109
1110
1111Date: Fri, 16 Oct 2020 12:44:42 +0000
1112From: Mayank Roy Chowdhury <onbehalfof@manuscriptcentral.com>
1113Reply-To: speoffice@wiley.com
1114To: pabuhr@uwaterloo.ca
1115Subject: Manuscript Accepted - Please submit final updates to SPE-19-0219.R3 [email ref: ENR-AW-1-c]
1116
111716-Oct-2020
1118
1119Dear Dr. Buhr,
1120
1121Manuscript id: SPE-19-0219.R3
1122Manuscript title: Advanced Control-flow and Concurrency in Cforall
1123
1124Although your manuscript has been accepted for publication it is now being returned to your author center for you to review and make any final adjustments or corrections prior to production and publication.
1125
1126Any special instructions will be listed below:
11271) Funding Information added in ScholorOne but missing in main document, Kindly add the Funding information in main document.
11282) Please provide the clean version of the manuscript without any highlights or tracked changes.
11293) Kindly check and make sure citations for all figures and Tables are present in the main document
1130
1131Please now log back into your Scholar One Author Center and click on the "Manuscripts Accepted for First Look" queue. In order to update the submission, click on the "submit updated manuscript" link in the "Actions" column and follow the steps as you would during a manuscript submission process.
1132
1133On the File Upload screen please upload the FINAL versions of all the files, including print quality image files. For information about image quality requirements, please refer to the guidelines at https://authorservices.wiley.com/asset/photos/electronic_artwork_guidelines.pdf.
1134
1135Instructions for uploading replacement files:
11361. On the "File Upload" step, click on the "edit" button for the file you wish to replace.
11372. In the "Upload a later version" section, browse to locate the replacement final version.
11383. Add any comments concerning the replacement (e.g. "high res image").
11394. Select whether the new file is a minor or major version (we suggest you select minor version)
11405. Click upload.
11416. Click 'Submit' when all the files have been uploaded and you will receive an automated email to say that submission is successful.
1142
1143Please submit your updates within the next 7 days to ensure there are no unnecessary delays in production.
1144
1145Sincerely,
1146Software: Practice and Experience Editorial Office
1147
1148
1149
1150From: SPE Office <speoffice@wiley.com>
1151To: "Peter A. Buhr" <pabuhr@uwaterloo.ca>
1152Subject: Re: Manuscript Accepted - Please submit final updates to SPE-19-0219.R3 [email ref: ENR-AW-1-c]
1153Date: Mon, 19 Oct 2020 17:04:24 +0000
1154
1155Dear Dr. Buhr,
1156
1157Thank you very much for contacting the Editorial Office.
1158
1159I would like to let you know that the files has been found in order and moved to production.
1160
1161Plesae let me know for further assistance in this regard.
1162
1163Best Regards
1164
1165Mayank Roy Chowdhury
1166Editorial Assistant
1167Software practice and Experience
1168________________________________
1169From: Peter A. Buhr <pabuhr@uwaterloo.ca>
1170Sent: Sunday, October 18, 2020 2:00 PM
1171To: SPE Office <speoffice@wiley.com>
1172Cc: Thierry Delisle <tdelisle@uwaterloo.ca>
1173Subject: Re: Manuscript Accepted - Please submit final updates to SPE-19-0219.R3 [email ref: ENR-AW-1-c]
1174
1175       This is an external email.
1176
1177    Mayank Roy Chowdhury <onbehalfof@manuscriptcentral.com> writes:
1178
1179    Instructions for uploading replacement files:
1180    1. On the "File Upload" step, click on the "edit" button for the file you wish to replace.
1181    2. In the "Upload a later version" section, browse to locate the replacement final version.
1182    3. Add any comments concerning the replacement (e.g. "high res image").
1183    4. Select whether the new file is a minor or major version (we suggest you select minor version)
1184    5. Click upload.
1185    6. Click 'Submit' when all the files have been uploaded and you will receive an automated email to say that submission is successful.
1186
1187There was no "edit" button on the "File Upload" page, so I just upload the
1188final version of the PDF and source files using the mechanism on the "File
1189Upload" page and submitted that.
1190
1191
1192
1193Date: Tue, 20 Oct 2020 13:28:37 +0530
1194To: "Dr. Peter Buhr" <pabuhr@uwaterloo.ca>
1195From: jpcms@spi-global.com
1196Subject: Information: Production Editor Contact Software:Practice and Experience  | Advanced Control-flow and Concurrency in C A
1197
1198Dear Dr. Peter Buhr,
1199
1200We are in the process of preparing "Advanced Control-flow and Concurrency in C A" for publication. Your production editor, Joel Pacaanas, will support you and your article throughout the process.
1201
1202Please get in touch with your Production Editor at SPEproofs@wiley.com;EllaMae.Navor@spi-global.com if you have any questions.
1203               
1204Sincerely,
1205Booking-in Team,
1206On behalf of Wiley
1207
1208Article ID: SPE_2925
1209Article DOI: 10.1002/SPE.2925
1210
1211
1212
1213Date: Tue, 20 Oct 2020 10:33:04 +0000
1214From: <cs-author@wiley.com>
1215To: <pabuhr@uwaterloo.ca>
1216Subject: In Production: Your article accepted in Software: Practice and Experience
1217
1218Dear Peter Buhr,
1219
1220Article ID: SPE2925
1221Article DOI: 10.1002/spe.2925
1222Internal Article ID: 16922213
1223Article: Advanced Control-flow and Concurrency in C A
1224Journal: Software: Practice and Experience
1225
1226Congratulations on the acceptance of your article for publication in Software: Practice and Experience.
1227
1228Your article has been received and the production process is now underway. We look forward to working with you and publishing your article. Using Wiley Author Services, you can track your article's progress.
1229
1230Please click below to login - if you are using a different email address than this one, you will need to manually assign this article to your Dashboard (see https://hub.wiley.com/docs/support/assigning-a-missing-article-to-my-dashboard-DOC-11871?utm_source=new%20user%20invitation&utm_medium=email How do I assign a missing article to My Dashboard?):
1231
1232https://authorservices.wiley.com/index.html#login?campaign=email_invitation-new
1233
1234If applicable, a list of available actions will appear below - check out your Author Services Dashboard for all actions related to your articles.
1235
1236Sign your license agreement (REQUIRED)  -- you will receive an email when this task is ready on your dashboard. Track your article's progress to publicationAccess your published articleInvite colleagues to view your published article
1237If you need any assistance, please click http://www.wileyauthors.com/help?utm_source=new%20user%20invitation&utm_medium=email here to view our Help section.
1238
1239Sincerely,
1240Wiley Author Services
1241
1242P.S. - Some journals accept artwork submissions for Cover Images. This is an optional service you can use to help increase article exposure and showcase your research. Pricing and placement options vary by journal. For more information, including artwork guidelines, pricing, and submission details, please visit the https://authorservices.wiley.com/author-resources/Journal-Authors/Promotion/journal-cover-image.html?utm_source=as&utm_medium=email&utm_term=invitation_msg&utm_content=covers&utm_campaign=2019feb?campaign=email_invitation-new" target=_blank">Journal Cover Image page. If you want help creating an image, Wiley Editing Services offers a professional https://wileyeditingservices.com/en/article-promotion/cover-image-design.html?utm_source=as&utm_medium=email&utm_term=ie&utm_content=cid&utm_campaign=prodops" target=_blank">Cover Image Design service that creates eye-catching images, ready to be showcased on the journal cover.
1243
1244
1245
1246Date: Thu, 22 Oct 2020 20:21:49 +0000
1247From: <cs-author@wiley.com>
1248To: <pabuhr@uwaterloo.ca>
1249Subject: You have actions to complete in Author Services
1250
1251Dear Peter Buhr,
1252
1253Article ID: SPE2925
1254Article DOI: 10.1002/spe.2925
1255Internal Article ID: 16922213
1256Article: Advanced Control-flow and Concurrency in C A
1257Journal: Software: Practice and Experience
1258
1259For the above article, you have the following open tasks:
1260
1261Sign your license agreement in order to publish your article. Simply click the Sign License button on your https://authorservices.wiley.com?campaign=email_license-notice1">Wiley Author Services Dashboard.
1262
1263Need any help? Please visit our https://authorsupport.wiley.com/s/">Author Support Center.
1264
1265Sincerely,
1266Wiley Author Services
1267
1268
1269
1270Date: Thu, 22 Oct 2020 23:13:07 +0000
1271From: <cs-author@wiley.com>
1272To: <pabuhr@uwaterloo.ca>
1273Subject: License was successfully submitted! Thank you!
1274
1275Dear Peter Buhr,                                                                 
1276
1277Article ID: SPE2925
1278Article DOI: 10.1002/spe.2925
1279Internal Article ID: 16922213
1280Article: Advanced Control-flow and Concurrency in C A 
1281Journal: Software: Practice and Experience                                                                     
1282                                                                         
1283You've successfully completed license signing for your article - thank you! You can view your signed agreement at any time by visiting your https://authorservices.wiley.com?campaign=email_license-confirm">Wiley Author Services Dashboard.                                                                     
1284
1285Sincerely,                                                                                 
1286
1287Wiley Author Services
Note: See TracBrowser for help on using the repository browser.