- mpiwg-coll - mailman3.cels.anl.gov

Re: [mpi3-coll] Non-blocking Collectives Proposal Draft
by Bronis R. de Supinski 15 Oct '08

15 Oct '08

Rajeev and Torsten: I believe that Rajeev is correct. The Chicago Manual of Style passge that Torsten cited is not germane. It deals with whether you are supposed to add hyphens to adjective/ noun combinations that are being used as adjectives, e.g., "graduate student" (instead of "graduate-student") in graduate student housing. It has nothing to do with hyphenating compound words or words that are synthesized by adding well defined prefixes to existing words. Bronis On Wed, 15 Oct 2008, Rajeev Thakur wrote: > I will just add that if you go to www. dictionary.com and look up > "nonbelligerent", it doesn't have a hyphen. In the left column, under > "Browse nearby entries" you can scroll through a boat load of "non" words > that don't have a hyphen. If you look up > http:// www. merriam-webster.com/dictionary/nonbelligerent, at the bottom it > shows the MLA style and APA style also without the hyphen. > > Rajeev > > > -----Original Message----- > > From: mpi3-coll-bounces(a)lists.mpi-forum.org > > [mailto:[email protected]] On Behalf Of > > Torsten Hoefler > > Sent: Wednesday, October 15, 2008 3:53 PM > > To: MPI-3 Collective Subgroup Discussions > > Subject: Re: [mpi3-coll] Non-blocking Collectives Proposal Draft > > > > On Wed, Oct 15, 2008 at 03:46:39PM -0500, Rajeev Thakur wrote: > > > We can do whatever we want really. It's not terribly > > important whether > > > nonblocking is hyphenated or not. > > that's right :). I will still add this as an MPI-2.2 ticket so that we > > can have the official 5 minute discussion in front of the forum. > > > > Thanks for your remarks! > > > > Best, > > Torsten > > > > -- > > bash$ :(){ :|:&};: --------------------- http:// www. unixer.de/ ----- > > Torsten Hoefler | Postdoctoral Researcher > > Open Systems Lab | Indiana University > > 150 S. Woodlawn Ave. | Bloomington, IN, 474045, USA > > Lindley Hall Room 135 | +01 (812) 855-3608 > > _______________________________________________ > > mpi3-coll mailing list > > mpi3-coll(a)lists.mpi-forum.org > > http:// lists.mpi-forum.org/mailman/listinfo.cgi/mpi3-coll > > > > _______________________________________________ > mpi3-coll mailing list > mpi3-coll(a)lists.mpi-forum.org > http:// lists.mpi-forum.org/mailman/listinfo.cgi/mpi3-coll > >

1 0

Re: [mpi3-coll] Non-blocking Collectives Proposal Draft
by Torsten Hoefler 15 Oct '08

15 Oct '08

On Wed, Oct 15, 2008 at 04:03:48PM -0500, Rajeev Thakur wrote: > I will just add that if you go to www.dictionary.com and look up > "nonbelligerent", it doesn't have a hyphen. In the left column, under > "Browse nearby entries" you can scroll through a boat load of "non" words > that don't have a hyphen. If you look up > http://www.merriam-webster.com/dictionary/nonbelligerent, at the bottom it > shows the MLA style and APA style also without the hyphen. yes, but this whole thing is very ambiguous. I entered non-blocking at www.dictionary.com which was not found, but the "Reference suggestions" show all three ways of writing it (1 x nonblocking, 1 x non blocking and 3 x non-blocking). And try to click on "non blocking" - it leads to a webpage with that only lists non-blocking. This is very unclear to me. Do you know a way to access MLA or APA online? This is getting interesting! I don't even know a rule how to do this in German. It would be interesting to know if there is a rule in English. Actually, try "cat /usr/share/dict/words | grep blocking" on your machine (not that this would mean anything more). I guess we will have (an amusing) vote on this in Chicago! All the Best, Torsten -- bash$ :(){ :|:&};: --------------------- http://www.unixer.de/ ----- Torsten Hoefler | Postdoctoral Researcher Open Systems Lab | Indiana University 150 S. Woodlawn Ave. | Bloomington, IN, 474045, USA Lindley Hall Room 135 | +01 (812) 855-3608

1 0

Re: [mpi3-coll] Non-blocking Collectives Proposal Draft
by Rajeev Thakur 15 Oct '08

15 Oct '08

I will just add that if you go to www.dictionary.com and look up "nonbelligerent", it doesn't have a hyphen. In the left column, under "Browse nearby entries" you can scroll through a boat load of "non" words that don't have a hyphen. If you look up http://www.merriam-webster.com/dictionary/nonbelligerent, at the bottom it shows the MLA style and APA style also without the hyphen. Rajeev > -----Original Message----- > From: mpi3-coll-bounces(a)lists.mpi-forum.org > [mailto:[email protected]] On Behalf Of > Torsten Hoefler > Sent: Wednesday, October 15, 2008 3:53 PM > To: MPI-3 Collective Subgroup Discussions > Subject: Re: [mpi3-coll] Non-blocking Collectives Proposal Draft > > On Wed, Oct 15, 2008 at 03:46:39PM -0500, Rajeev Thakur wrote: > > We can do whatever we want really. It's not terribly > important whether > > nonblocking is hyphenated or not. > that's right :). I will still add this as an MPI-2.2 ticket so that we > can have the official 5 minute discussion in front of the forum. > > Thanks for your remarks! > > Best, > Torsten > > -- > bash$ :(){ :|:&};: --------------------- http://www.unixer.de/ ----- > Torsten Hoefler | Postdoctoral Researcher > Open Systems Lab | Indiana University > 150 S. Woodlawn Ave. | Bloomington, IN, 474045, USA > Lindley Hall Room 135 | +01 (812) 855-3608 > _______________________________________________ > mpi3-coll mailing list > mpi3-coll(a)lists.mpi-forum.org > http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi3-coll >

1 0

Re: [mpi3-coll] Non-blocking Collectives Proposal Draft
by Torsten Hoefler 15 Oct '08

15 Oct '08

On Wed, Oct 15, 2008 at 03:46:39PM -0500, Rajeev Thakur wrote: > We can do whatever we want really. It's not terribly important whether > nonblocking is hyphenated or not. that's right :). I will still add this as an MPI-2.2 ticket so that we can have the official 5 minute discussion in front of the forum. Thanks for your remarks! Best, Torsten -- bash$ :(){ :|:&};: --------------------- http://www.unixer.de/ ----- Torsten Hoefler | Postdoctoral Researcher Open Systems Lab | Indiana University 150 S. Woodlawn Ave. | Bloomington, IN, 474045, USA Lindley Hall Room 135 | +01 (812) 855-3608

1 0

Re: [mpi3-coll] Non-blocking Collectives Proposal Draft
by Rajeev Thakur 15 Oct '08

15 Oct '08

We can do whatever we want really. It's not terribly important whether nonblocking is hyphenated or not. Rajeev > -----Original Message----- > From: mpi3-coll-bounces(a)lists.mpi-forum.org > [mailto:[email protected]] On Behalf Of > Torsten Hoefler > Sent: Wednesday, October 15, 2008 3:38 PM > To: MPI-3 Collective Subgroup Discussions > Subject: Re: [mpi3-coll] Non-blocking Collectives Proposal Draft > > On Wed, Oct 15, 2008 at 03:24:49PM -0500, Rajeev Thakur wrote: > > I just walked over to our editor, Gail Pieper, and asked > her the reason. She > > said that if the alphabets that the prefix connects are > different, as in > > nonblocking or noncontiguous, the convention nowadays is to > drop the hyphen. > > You would keep the hyphen in words like re-establish and > also in words that > > would read weird without the hyphen, like non-organic (nono > rganic). This is > > the convention followed by the Chicago Manual of Style, the > Associated Press > > Manual, and the MLA Manual (Modern Language Association). > hmm, I just read your cited Chicago Manual of Style (CMS) and found: > http://www.chicagomanualofstyle.org/ch07/ch07_sec085.html > which reads: > "A hyphen can make for easier reading by showing structure and, often, > pronunciation." ... "Where no ambiguity could result, as in public > welfare administration or graduate student housing, hyphenation is not > mandatory, though it is quite acceptable and preferred by many writers > and editors." > > (you should read the complete page) > > > > Also, if you do the <pick your favorit search enigne> > test (I know ... > > > but languages are dynamically changed by their users), > you'll see that > > > non-blocking seems to be the most-used term in the I/O and > > > communication > > > context. > > > > Web searches are not useful for this because most people's > English is not > > that good. > I strongly disagree. A language is determined by the people > who use it, > who defines what the correct language is? Even the CMS says > things like: > "With frequent use" or "have become widely accepted,". See > http://www.chicagomanualofstyle.org/ch07/ch07_sec084.html for more of > this :). I know, especially as a German, that a clearly > defined language > is preferred, but I think our case is not clearly defined (or can you > cite a particular part of CMS?). > > If we don't have a clear definition, I'd like to go with the > unofficial > conventions. Those unofficial conventions can be backed up with web > searches reasonably well (I read the first few results and those were > reasonable hits). > > All the Best, > Torsten > > -- > bash$ :(){ :|:&};: --------------------- http://www.unixer.de/ ----- > Torsten Hoefler | Postdoctoral Researcher > Open Systems Lab | Indiana University > 150 S. Woodlawn Ave. | Bloomington, IN, 474045, USA > Lindley Hall Room 135 | +01 (812) 855-3608 > _______________________________________________ > mpi3-coll mailing list > mpi3-coll(a)lists.mpi-forum.org > http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi3-coll >

1 0

Re: [mpi3-coll] Non-blocking Collectives Proposal Draft
by Torsten Hoefler 15 Oct '08

15 Oct '08

On Wed, Oct 15, 2008 at 03:24:49PM -0500, Rajeev Thakur wrote: > I just walked over to our editor, Gail Pieper, and asked her the reason. She > said that if the alphabets that the prefix connects are different, as in > nonblocking or noncontiguous, the convention nowadays is to drop the hyphen. > You would keep the hyphen in words like re-establish and also in words that > would read weird without the hyphen, like non-organic (nono rganic). This is > the convention followed by the Chicago Manual of Style, the Associated Press > Manual, and the MLA Manual (Modern Language Association). hmm, I just read your cited Chicago Manual of Style (CMS) and found: http://www.chicagomanualofstyle.org/ch07/ch07_sec085.html which reads: "A hyphen can make for easier reading by showing structure and, often, pronunciation." ... "Where no ambiguity could result, as in public welfare administration or graduate student housing, hyphenation is not mandatory, though it is quite acceptable and preferred by many writers and editors." (you should read the complete page) > > Also, if you do the <pick your favorit search enigne> test (I know ... > > but languages are dynamically changed by their users), you'll see that > > non-blocking seems to be the most-used term in the I/O and > > communication > > context. > > Web searches are not useful for this because most people's English is not > that good. I strongly disagree. A language is determined by the people who use it, who defines what the correct language is? Even the CMS says things like: "With frequent use" or "have become widely accepted,". See http://www.chicagomanualofstyle.org/ch07/ch07_sec084.html for more of this :). I know, especially as a German, that a clearly defined language is preferred, but I think our case is not clearly defined (or can you cite a particular part of CMS?). If we don't have a clear definition, I'd like to go with the unofficial conventions. Those unofficial conventions can be backed up with web searches reasonably well (I read the first few results and those were reasonable hits). All the Best, Torsten -- bash$ :(){ :|:&};: --------------------- http://www.unixer.de/ ----- Torsten Hoefler | Postdoctoral Researcher Open Systems Lab | Indiana University 150 S. Woodlawn Ave. | Bloomington, IN, 474045, USA Lindley Hall Room 135 | +01 (812) 855-3608

1 0

Re: [mpi3-coll] Non-blocking Collectives Proposal Draft
by Rajeev Thakur 15 Oct '08

15 Oct '08

I just walked over to our editor, Gail Pieper, and asked her the reason. She said that if the alphabets that the prefix connects are different, as in nonblocking or noncontiguous, the convention nowadays is to drop the hyphen. You would keep the hyphen in words like re-establish and also in words that would read weird without the hyphen, like non-organic (nono rganic). This is the convention followed by the Chicago Manual of Style, the Associated Press Manual, and the MLA Manual (Modern Language Association). > Also, if you do the <pick your favorit search enigne> test (I know ... > but languages are dynamically changed by their users), you'll see that > non-blocking seems to be the most-used term in the I/O and > communication > context. Web searches are not useful for this because most people's English is not that good. Rajeev > -----Original Message----- > From: mpi3-coll-bounces(a)lists.mpi-forum.org > [mailto:[email protected]] On Behalf Of > Torsten Hoefler > Sent: Wednesday, October 15, 2008 3:00 PM > To: MPI-3 Collective Subgroup Discussions > Subject: Re: [mpi3-coll] Non-blocking Collectives Proposal Draft > > On Wed, Oct 15, 2008 at 02:49:03PM -0500, Rajeev Thakur wrote: > > > 1) The MPI-2.1 document has 127 occurrences of "nonblocking" > > > and only 8 > > > occurrences of "non-blocking". What would be the correct > > > term? Should we try to be more consistent? > > > > The technical editor in our division prefers the version > without the hyphen. > > She will remove the hyphen in our papers. > > hmm, I would disagree - I had exactly the same discussion with our > technical editor :). She explained me for 15 minutes why it is > "non-blocking" and not "nonblocking" or "non blocking". > > Also, if you do the <pick your favorit search enigne> test (I know ... > but languages are dynamically changed by their users), you'll see that > non-blocking seems to be the most-used term in the I/O and > communication > context. > > I'm open for arguments! > > Best, > Torsten > > -- > bash$ :(){ :|:&};: --------------------- http://www.unixer.de/ ----- > Torsten Hoefler | Postdoctoral Researcher > Open Systems Lab | Indiana University > 150 S. Woodlawn Ave. | Bloomington, IN, 474045, USA > Lindley Hall Room 135 | +01 (812) 855-3608 > _______________________________________________ > mpi3-coll mailing list > mpi3-coll(a)lists.mpi-forum.org > http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi3-coll >

1 0

Re: [mpi3-coll] Non-blocking Collectives Proposal Draft
by Torsten Hoefler 15 Oct '08

15 Oct '08

On Wed, Oct 15, 2008 at 02:49:03PM -0500, Rajeev Thakur wrote: > > 1) The MPI-2.1 document has 127 occurrences of "nonblocking" > > and only 8 > > occurrences of "non-blocking". What would be the correct > > term? Should we try to be more consistent? > > The technical editor in our division prefers the version without the hyphen. > She will remove the hyphen in our papers. hmm, I would disagree - I had exactly the same discussion with our technical editor :). She explained me for 15 minutes why it is "non-blocking" and not "nonblocking" or "non blocking". Also, if you do the <pick your favorit search enigne> test (I know ... but languages are dynamically changed by their users), you'll see that non-blocking seems to be the most-used term in the I/O and communication context. I'm open for arguments! Best, Torsten -- bash$ :(){ :|:&};: --------------------- http://www.unixer.de/ ----- Torsten Hoefler | Postdoctoral Researcher Open Systems Lab | Indiana University 150 S. Woodlawn Ave. | Bloomington, IN, 474045, USA Lindley Hall Room 135 | +01 (812) 855-3608

1 0

Re: [mpi3-coll] Non-blocking Collectives Proposal Draft
by Rajeev Thakur 15 Oct '08

15 Oct '08

> 1) The MPI-2.1 document has 127 occurrences of "nonblocking" > and only 8 > occurrences of "non-blocking". What would be the correct > term? Should we try to be more consistent? The technical editor in our division prefers the version without the hyphen. She will remove the hyphen in our papers. Rajeev > -----Original Message----- > From: mpi3-coll-bounces(a)lists.mpi-forum.org > [mailto:[email protected]] On Behalf Of > Christian Siebert > Sent: Wednesday, October 15, 2008 9:03 AM > To: MPI-3 Collective Subgroup Discussions > Subject: Re: [mpi3-coll] Non-blocking Collectives Proposal Draft > > @Torsten: > > Thanks a lot for this first draft! I tried to make some > useful comments > and attached the results to this mail (sorry, I can't comment > PDFs so I > did it the "old inconvenient way"). > > @Working Group (something to discuss): > > 1) The MPI-2.1 document has 127 occurrences of "nonblocking" > and only 8 > occurrences of "non-blocking". What would be the correct > term? Should we > try to be more consistent? > > 2) Clarification of MPI_Request_free() for requests from non-blocking > collective operations. > > 3) Better definition/description for "matching" (there is > nothing like > "at the same time" -> logical order?). > > 4) Define "levels of progression"? To be queried (e.g., for > "Synchronous > Progress" MPI_Tests are needed for performance, but for "Asynchronous > Progress" they would only add unnecessary overhead)? UP >= AP >= SP? > > 5) "Unexpected Progress" => "buffered" collectives? > > 6) NBC gives several possible ways for optimizations. With > this "General > advice to implementers" we stick to only one, and might > prevent others. > Can we already fix a decision for optimization strategies at > this stage? > Should we fix it at all? > > 7) Should there be a concrete code example in the proposal (e.g. an > implementation of this double buffering example)? > > Hope this isn't too much for now. > > Best regards, > Christian > > Torsten Hoefler wrote: > > Hello Working Group, > > I finished a first draft of our non-blocking collectives proposal. > > Please read it carefully and send comments to me or the list (e.g., > > marked-up pdf documents). > > > > The draft is attached to this mail. > > > > All the Best, > > Torsten > > -- > Christian Siebert, Dipl.-Inf. Research Associate > > NEC Laboratories Europe, NEC Europe Ltd. > Rathausallee 10, D-53757 Sankt Augustin, Germany > > Phone: +49 (0) 2241 - 92 52 44 Fax: +49 (0) 2241 - 92 52 99 > > (Registered Office: 1 Victoria Road, London W3 6BL, 2832014) >

1 0

Re: [mpi3-coll] Non-blocking Collectives Proposal Draft
by Supalov, Alexander 15 Oct '08

15 Oct '08

Dear Torsten, Thanks for the proposal. A couple of comments in the more old-fashioned way. Generally, some sentences start with lower case letters that should be capitalized. Page 3, top. We probably should not explain here how to ensure progress for nonblocking ops. This matter should be covered by the general progress description elsewhere. Page 3, lower half. Look how progress is defined in this list by referring to the pt2pt progress rules. Ibid. There are already 3 classes of requests: pt2pt, generalized, and file I/O ones. Page 3, bottom. How many nonblocking ops can overlap with a blocking one? I.e., can one have the following situation: Process 0 Process 1 MPI_Ibarrier(req) MPI_Ibarrier(req) MPI_Wait(req) MPI Bcast MPI_Bcast MPI_Wait(req) Page 5, bottom. 32767 simultaneous ops to support is a tall order. I'd say, 1 (one) would be a good lower limit, thus allowing no overlap. Or even 0 (zero), meaning no nonblocking collective support at all. The rest of this passage would probably be better reformulated as an advice to implementors. In the margin: editing a PDF requires creativity (exemplified by the earlier comments) or a pretty expensive Acrobat. Maybe we can find some other way of distributing and commenting on our drafts? Best regards. Alexander -----Original Message----- From: mpi3-coll-bounces(a)lists.mpi-forum.org [mailto:[email protected]] On Behalf Of Torsten Hoefler Sent: Tuesday, October 14, 2008 9:47 PM To: mpi3-coll(a)lists.mpi-forum.org Subject: [mpi3-coll] Non-blocking Collectives Proposal Draft Hello Working Group, I finished a first draft of our non-blocking collectives proposal. Please read it carefully and send comments to me or the list (e.g., marked-up pdf documents). The draft is attached to this mail. All the Best, Torsten -- bash$ :(){ :|:&};: --------------------- http://www.unixer.de/ ----- Torsten Hoefler | Research Assistant Open Systems Lab | Indiana University 150 S. Woodlawn Ave. | Bloomington, IN, 474045, USA Lindley Hall Room 135 | +01 (812) 855-3608 --------------------------------------------------------------------- Intel GmbH Dornacher Strasse 1 85622 Feldkirchen/Muenchen Germany Sitz der Gesellschaft: Feldkirchen bei Muenchen Geschaeftsfuehrer: Douglas Lusk, Peter Gleissner, Hannes Schwaderer Registergericht: Muenchen HRB 47456 Ust.-IdNr. VAT Registration No.: DE129385895 Citibank Frankfurt (BLZ 502 109 00) 600119052 This e-mail and any attachments may contain confidential material for the sole use of the intended recipient(s). Any review or distribution by others is strictly prohibited. If you are not the intended recipient, please contact the sender and delete all copies.

1 0