Thanks! Nevertheless, I do not understand how procedure partitions can run correctly in two concurrent processes. cach-size and cash are module top-level variables in math/number-theory. These variables may be assigned new values when calling procedures partitions and set-partitions-cash. Apparently I do not understand the docs on the caveats of futures and threads in this respect. I can imagine that process A retrieves the current cach-size, sees that it is large enough while process B shrinks the cash before process A refers to the cash. I am puzzled here.
Thanks again, Jos PS IIRC in version 6.0.1 there is a global cash that never shrinks and a local one. This I understand to run well in concurrent processes. > -----Original Message----- > From: Matthew Flatt [mailto:mfl...@cs.utah.edu] > Sent: jueves, 17 de julio de 2014 18:02 > To: Jos Koot > Cc: 'Jens Axel Søgaard'; 'Vincent St-Amour'; 'Eric Dobson'; > 'Racket Users List' > Subject: RE: [racket] FW: q about code for partitions > > Your example exposed two different bugs in `future`. I've pushed > repairs. > > Thanks! > > At Thu, 17 Jul 2014 13:56:48 +0200, "Jos Koot" wrote: > > Version 6.0.1 shows the same problem, that is, no problem > with racket, only > > with DrRacket. > > Jos > > > > > -----Original Message----- > > > From: Matthew Flatt [mailto:mfl...@cs.utah.edu] > > > Sent: jueves, 17 de julio de 2014 13:48 > > > To: Jos Koot > > > Cc: 'Jens Axel Søgaard'; 'Vincent St-Amour'; 'Eric Dobson'; > > > 'Racket Users List' > > > Subject: RE: [racket] FW: q about code for partitions > > > > > > I'll work on this. Thanks! > > > > > > At Thu, 17 Jul 2014 13:41:28 +0200, "Jos Koot" wrote: > > > > Strange: I do not get a segfault when running racket, only > > > when running from > > > > DrRacket. > > > > Jos > > > > > > > > > -----Original Message----- > > > > > From: jensaxelsoega...@gmail.com > > > > > [mailto:jensaxelsoega...@gmail.com] On Behalf Of Jens > Axel Søgaard > > > > > Sent: jueves, 17 de julio de 2014 13:12 > > > > > To: Jos Koot > > > > > Cc: Vincent St-Amour; Eric Dobson; Matthew Flatt; Racket > > > Users List > > > > > Subject: Re: [racket] FW: q about code for partitions > > > > > > > > > > What you describe seems more like a bug in partitions than a > > > > > documentation bug. > > > > > Do you have a small example that provoke a segfault? > > > > > > > > > > /Jens Axel > > > > > > > > > > > > > > > 2014-07-17 12:41 GMT+02:00 Jos Koot <jos.k...@gmail.com>: > > > > > > The documentation of partitions has not yet been updated, > > > > > as far as I can > > > > > > see in version 6.0.1.13--2014-07-08(7735dd0/a) [3m]. I > > > > > don't want to hurry > > > > > > up things (there may be other priorities) but when the docs > > > > > will be updated, > > > > > > I think they should include a warning against calling > > > > > function partitions > > > > > > from two concurrently running futures or threads. Reason is > > > > > the non atomic > > > > > > memoization. (let alone calling set-partitions-cache in one > > > > > process while > > > > > > another one is consulting/updating the cache) > > > > > > > > > > > > Calling partitions from two concurrent futures gives me a > > > > > segfault. Not a > > > > > > surprise. > > > > > > Up to now calling from two concurrent threads gives me > > > > > correct results, but > > > > > > slows down vvveeerrryyy much. > > > > > > I assume concurrent threads can lead to the same problems > > > > > as concurrent > > > > > > futures, though. > > > > > > If I understand places correctly, calling from concurrent > > > > > places should go > > > > > > well. Do I understand correctly? > > > > > > > > > > > > Jos > > > > > > > > > > > >> -----Original Message----- > > > > > >> From: Vincent St-Amour [mailto:stamo...@ccs.neu.edu] > > > > > >> Sent: domingo, 29 de junio de 2014 22:52 > > > > > >> To: Eric Dobson > > > > > >> Cc: Jos Koot; Matthew Flatt; Jens Axel Søgaard; Racket > > > Users List > > > > > >> Subject: Re: [racket] FW: q about code for partitions > > > > > >> > > > > > >> Ah, that explains it. Thanks! > > > > > >> > > > > > >> Vincent > > > > > >> > > > > > >> > > > > > >> At Sun, 29 Jun 2014 12:32:48 -0700, > > > > > >> Eric Dobson wrote: > > > > > >> > > > > > > >> > Vincent: exact-zero? is defined through 'make-predicate' > > > > > which uses > > > > > >> > contract machinery to generate the function. I filed a > > > > > >> couple of bugs > > > > > >> > tracking some of the slowness issues. There is no > > > `contract` form > > > > > >> > though so I doubt that the coach will find it. > > > > > >> > > > > > > >> > > > > > > >> > http://bugs.racket-lang.org/query/?cmd=view%20audit-trail&data > > > > > > base=default&pr=14610 > > > > > >> > > > > > > >> > http://bugs.racket-lang.org/query/?cmd=view%20audit-trail&data > > > > > > base=default&pr=14611 > > > > > >> > > > > > > >> > On Sun, Jun 29, 2014 at 11:53 AM, Jos Koot > > > > > >> <jos.k...@gmail.com> wrote: > > > > > >> > > Great work, Jens. I am glad my approach as been adopted > > > > > >> (and much improved > > > > > >> > > without deviating from the original idea of simpler > > > > > >> recurrence). When can we > > > > > >> > > expect it in the next nightly build? > > > > > >> > > > > > > > >> > > Thanks, Jos > > > > > >> > > > > > > > >> > >> -----Original Message----- > > > > > >> > >> From: jensaxelsoega...@gmail.com > > > > > >> > >> [mailto:jensaxelsoega...@gmail.com] On Behalf Of Jens > > > > > >> Axel Søgaard > > > > > >> > >> Sent: domingo, 29 de junio de 2014 12:48 > > > > > >> > >> To: Matthew Flatt > > > > > >> > >> Cc: Jos Koot; Racket Users List > > > > > >> > >> Subject: Re: [racket] FW: q about code for partitions > > > > > >> > >> > > > > > >> > >> I have made a new vector version using zero? > instead of > > > > > >> exact-zero?. > > > > > >> > >> > > > > > >> > >> To give users a chance to remove the cache after doing > > > > > >> > >> partitions calculations, > > > > > >> > >> I have added set-partitions-cache. > > > > > >> > >> > > > > > >> > >> Code: > > > > > >> > >> > > > > > >> > >> > > > https://github.com/soegaard/racket/blob/patch-14/pkgs/math-pkg > > > > > >> > > s/math-lib/math/private/number-theory/partitions.rkt > > > > > >> > >> > > > > > >> > >> Discussion: > > > > > >> > >> https://github.com/plt/racket/pull/697 > > > > > >> > >> > > > > > >> > >> /Jens Axel > > > > > >> > >> > > > > > >> > >> > > > > > >> > >> 2014-06-29 12:44 GMT+02:00 Jens Axel Søgaard > > > > > >> <jensa...@soegaard.net>: > > > > > >> > >> > 2014-06-29 8:47 GMT+02:00 Matthew Flatt > > > > > <mfl...@cs.utah.edu>: > > > > > >> > >> >> It looks like "partitions2.rkt" ends up calling a > > > > > >> contract-wrapped > > > > > >> > >> >> variant of `exact-zero?`. > > > > > >> > >> > > > > > > >> > >> > That explains why Eric saw an improvement, when > > > the used #f > > > > > >> > >> instead of > > > > > >> > >> > 0 as the not-cached-yet value. > > > > > >> > >> > > > > > > >> > >> > /Jens Axel > > > > > >> > >> > > > > > >> > >> > > > > > >> > >> > > > > > >> > >> -- > > > > > >> > >> -- > > > > > >> > >> Jens Axel Søgaard > > > > > >> > > > > > > > >> > > > > > > > >> > > ____________________ > > > > > >> > > Racket Users list: > > > > > >> > > http://lists.racket-lang.org/users > > > > > >> > > > > > > >> > ____________________ > > > > > >> > Racket Users list: > > > > > >> > http://lists.racket-lang.org/users > > > > > > > > > > > > > > > > > > ____________________ > > > > > > Racket Users list: > > > > > > http://lists.racket-lang.org/users > > > > > > > > > > > > > > > > > > > > -- > > > > > -- > > > > > Jens Axel Søgaard ____________________ Racket Users list: http://lists.racket-lang.org/users
