- Pharo-users - lists.pharo.org

Re: [Pharo-users] UFFI cleanup of externally allocated memory (was Re: UFFI const, unsigned, opaque-ish types)

by Esteban Lorenzano

Hi, sorry for arriving so late to this, but I was on holidays :) this is how autoRelease works: 1) #autoRelease of an object registers object for finalisation with a particular executor. Then behaviour is divided: 2.1.1) for ExternalAddresses, it just registers in regular way, who will call #finalize on GC 2.1.2) finalize will just call a free assuming ExternalAddress was allocated (which is a malloc) 2.2.1) for all FFIExternalReference, it will register for finalisation what #resourceData answers (normally, the handle of the object) 2.2.2) finalisation process will call the object class>>#finalizeResourceData: method, with the #resourceData result as parameter 2.2.3) each kind of external reference can decide how to free that data (by default is also just freeing). An example of this is how CairoFontFace works (or AthensCairoSurface). In you case, you will have something like: FFIExternalObject subclass: #CXString CXString class>>#finalizeResourceData: version self ffiCall: #(void clang_disposeString(void *version)) notice that here I casted to void*â¦ this is because actually resourceData answers a pointer, not the object. cheers, Esteban > On 06 Sep 2016, at 13:29, Ben Coman <btc(a)openinworld.com> wrote: > > Hi Pierce, > > I can't determine from DemoFFI how the autorelease works. > > I see memory is allocated by the C library here.... > DemoFFIAutoThing>>allocByReturnedPointer > super allocByReturnedPointer. > handle autoRelease > > where super is... > DemoFFIThing>>allocByReturnedPointer > ((handle := DemoFFILibrary uniqueInstance apiAllocByReturnedPointer) isNull) > ifTrue: [ DemoFFIError signal: 'bummer' ] > > and... > DemoFFILibrary>>apiAllocByReturnedPointer > "demo_thing* alloc_by_returned_pointer(void)" > ^ self ffiCall: #(DemoFFIExternalObject alloc_by_returned_pointer ()) > > > which I guess is to be released by... > DemoFFILibrary>>apiFree: handle > "int free_thing(demo_thing*)" > ^ self ffiCall: #(int free_thing (DemoFFIExternalObject handle)) > > but this has only one sender... > DemoFFIThing>>letGo > handle isNull ifFalse: [ > DemoFFILibrary uniqueInstance apiFree: handle ] > > and the only sender of #letGo is... > DEMOFFIThingTest >> tearDown > > From this I can't follow how #apiFree: ultimately gets automatically > invoked by the object finalization. Are you able to fill me in on the > details? > > cheers -ben > > > On Sun, Sep 4, 2016 at 10:40 PM, Ben Coman <btc(a)openinworld.com> wrote: >> On Fri, Sep 2, 2016 at 10:04 PM, Pierce Ng <pierce(a)samadhiweb.com> wrote: >>> On Fri, Sep 02, 2016 at 12:40:06PM +0800, Ben Coman wrote: >>>> Referring to the example show_clang_version() below, cleanup is done >>>> with clang_disposeString(). How does UFFI hook into the object >>>> finalisation mechanism to call this when the object is garage >>>> collected. >>> >>> Ben, see if this addresses your query: >>> >>> http://www.samadhiweb.com/blog/2016.03.12.demoffi.html >>> >>> Pierce >>> >> >> Thanks Pierce. I've loaded DemoFFI and browsing it now. >> autorelease looks like part of what I need. >> I guess it is #finalize where I should call the libraryAPI_free() function? >> >> cheers -ben >

Sept. 6, 2016

Re: [Pharo-users] UFFI cleanup of externally allocated memory (was Re: UFFI const, unsigned, opaque-ish types)

by Ben Coman

Hi Pierce, I can't determine from DemoFFI how the autorelease works. I see memory is allocated by the C library here.... DemoFFIAutoThing>>allocByReturnedPointer super allocByReturnedPointer. handle autoRelease where super is... DemoFFIThing>>allocByReturnedPointer ((handle := DemoFFILibrary uniqueInstance apiAllocByReturnedPointer) isNull) ifTrue: [ DemoFFIError signal: 'bummer' ] and... DemoFFILibrary>>apiAllocByReturnedPointer "demo_thing* alloc_by_returned_pointer(void)" ^ self ffiCall: #(DemoFFIExternalObject alloc_by_returned_pointer ()) which I guess is to be released by... DemoFFILibrary>>apiFree: handle "int free_thing(demo_thing*)" ^ self ffiCall: #(int free_thing (DemoFFIExternalObject handle)) but this has only one sender... DemoFFIThing>>letGo handle isNull ifFalse: [ DemoFFILibrary uniqueInstance apiFree: handle ] and the only sender of #letGo is... DEMOFFIThingTest >> tearDown >From this I can't follow how #apiFree: ultimately gets automatically invoked by the object finalization. Are you able to fill me in on the details? cheers -ben On Sun, Sep 4, 2016 at 10:40 PM, Ben Coman <btc(a)openinworld.com> wrote: > On Fri, Sep 2, 2016 at 10:04 PM, Pierce Ng <pierce(a)samadhiweb.com> wrote: >> On Fri, Sep 02, 2016 at 12:40:06PM +0800, Ben Coman wrote: >>> Referring to the example show_clang_version() below, cleanup is done >>> with clang_disposeString(). How does UFFI hook into the object >>> finalisation mechanism to call this when the object is garage >>> collected. >> >> Ben, see if this addresses your query: >> >> http://www.samadhiweb.com/blog/2016.03.12.demoffi.html >> >> Pierce >> > > Thanks Pierce. I've loaded DemoFFI and browsing it now. > autorelease looks like part of what I need. > I guess it is #finalize where I should call the libraryAPI_free() function? > > cheers -ben

Sept. 6, 2016

Re: [Pharo-users] PunQlite problems

by Esteban Lorenzano

Hi, Voyage-UnQLite uses PunQLite driver. I will check what happens, because it should be working. Esteban ps: numbering is a mess, it doesnât necessarily means anything. > On 06 Sep 2016, at 12:42, PBKResearch <peter(a)pbkresearch.co.uk> wrote: > > Hi Esteban > > Well, I tried re-installing from your repo, but now I cannot even run the tests in the mumez repo. Whatever I try, I get an error message > in trying to open the database, saying I cannot get a pointer to a byte array. > > Previously, my Monticello browser said I was using PunQLite-Core(TorstenBergmann.28); now it shows the version as Estebanlorenzano.17. I donât understand the naming and numbering, but surely the version number shouldnât go backwards? > > Should I give upon PunQLite, and maybe try Voyage-UnqLite instead? Basically all I need is a database that looks like a large dictionary, with rather lengthy entries (mainly in XML). > > Thanks. > > Peter Kenny > > > From: Pharo-users [mailto:pharo-users-bounces@lists.pharo.org] On Behalf Of Esteban Lorenzano > Sent: 06 September 2016 10:21 > To: Any question about pharo is welcome <pharo-users(a)lists.pharo.org> > Subject: Re: [Pharo-users] PunQlite problems > > Hi, > > most probably not related to the problem, but the currently maintained (for us) PunQLite is now here: https://github.com/pharo-nosql/PunQLite <https://github.com/pharo-nosql/PunQLite> > > and correct way to install it is: > > Metacello new > repository: 'github://pharo-nosql/PunQLite/repository' <github://pharo-nosql/PunQLite/repository'>; > baseline: 'PunQLite'; > load. > > â¦ even if it is basically the same (and we submit changes to mumez repository too), this is the correct place for us :) > > cheers, > Esteban > >> On 06 Sep 2016, at 00:37, PBKResearch <peter(a)pbkresearch.co.uk <mailto:peter@pbkresearch.co.uk>> wrote: >> >> Hi â just tried a further experiment, with more mysterious results. I created a database in memory with a subset of my data. This time the program will retrieve the values as well as the keys. However, the value presented is just the first 1024 bytes, i.e. the first buffer full. In addition, in the majority of cases the correct bytes are each followed by three zero bytes, e.g. the data begins: >> 3C 00 00 00 68 00 00 00 >> 74 00 00 00 6D 00 00 00 >> 6C 00 00 00 20 00 00 00 >> 63 00 00 00 6C 00 00 00 >> Rather than the expected: >> 3C 68 74 6D 6C 20 63 6C. >> This affects 75 of the 88 cases in my trial database. >> >> I presume the insertion of the zero bytes took place in the writing stage, but I have no idea what is going on. Are there any known problems with PunQLite on pharo 5? >> >> >> >> Hello >> >> I am using Moose 6.0 (Pharo 5.0 Latest update: #50761) on Windows 10. I have installed PunQLite following the instructions in https://github.com/mumez/PunQLite <https://github.com/mumez/PunQLite>. I have run all the examples that involve using files, and everything seems OK. I have tried creating a database using my own data, which seemed to run OK. The database will display its keys successfully, but whenever I try to retrieve a value I get PqFetchError. The debug information shows that an ffi fetch operation has given a return value of -10. I have searched for any information on unQLite return codes, but so far I have nothing. >> >> I tried the PqDatabaseBrowser in the Tools section. This recognised the database, opened it and displayed all the keys correctly. However, clicking on any key gave the same PqFetchError. >> >> Does anyone have any idea of what I have done wrong? >> >> Thanks for any help. >> >> Peter Kenny

Sept. 6, 2016

Re: [Pharo-users] PunQlite problems

by PBKResearch

Hi Esteban Well, I tried re-installing from your repo, but now I cannot even run the tests in the mumez repo. Whatever I try, I get an error message in trying to open the database, saying I cannot get a pointer to a byte array. Previously, my Monticello browser said I was using PunQLite-Core(TorstenBergmann.28); now it shows the version as Estebanlorenzano.17. I donât understand the naming and numbering, but surely the version number shouldnât go backwards? Should I give upon PunQLite, and maybe try Voyage-UnqLite instead? Basically all I need is a database that looks like a large dictionary, with rather lengthy entries (mainly in XML). Thanks. Peter Kenny From: Pharo-users [mailto:pharo-users-bounces@lists.pharo.org] On Behalf Of Esteban Lorenzano Sent: 06 September 2016 10:21 To: Any question about pharo is welcome <pharo-users(a)lists.pharo.org> Subject: Re: [Pharo-users] PunQlite problems Hi, most probably not related to the problem, but the currently maintained (for us) PunQLite is now here: https://github.com/pharo-nosql/PunQLite and correct way to install it is: Metacello new repository: 'github://pharo-nosql/PunQLite/repository'; baseline: 'PunQLite'; load. â¦ even if it is basically the same (and we submit changes to mumez repository too), this is the correct place for us :) cheers, Esteban On 06 Sep 2016, at 00:37, PBKResearch <peter(a)pbkresearch.co.uk <mailto:peter@pbkresearch.co.uk> > wrote: Hi â just tried a further experiment, with more mysterious results. I created a database in memory with a subset of my data. This time the program will retrieve the values as well as the keys. However, the value presented is just the first 1024 bytes, i.e. the first buffer full. In addition, in the majority of cases the correct bytes are each followed by three zero bytes, e.g. the data begins: 3C 00 00 00 68 00 00 00 74 00 00 00 6D 00 00 00 6C 00 00 00 20 00 00 00 63 00 00 00 6C 00 00 00 Rather than the expected: 3C 68 74 6D 6C 20 63 6C. This affects 75 of the 88 cases in my trial database. I presume the insertion of the zero bytes took place in the writing stage, but I have no idea what is going on. Are there any known problems with PunQLite on pharo 5? Hello I am using Moose 6.0 (Pharo 5.0 Latest update: #50761) on Windows 10. I have installed PunQLite following the instructions in <https://github.com/mumez/PunQLite> https://github.com/mumez/PunQLite. I have run all the examples that involve using files, and everything seems OK. I have tried creating a database using my own data, which seemed to run OK. The database will display its keys successfully, but whenever I try to retrieve a value I get PqFetchError. The debug information shows that an ffi fetch operation has given a return value of -10. I have searched for any information on unQLite return codes, but so far I have nothing. I tried the PqDatabaseBrowser in the Tools section. This recognised the database, opened it and displayed all the keys correctly. However, clicking on any key gave the same PqFetchError. Does anyone have any idea of what I have done wrong? Thanks for any help. Peter Kenny

Sept. 6, 2016

Re: [Pharo-users] [ANN] Territorial

by HernÃ¡n Morales Durand

2016-09-06 2:53 GMT-03:00 Serge Stinckwich <serge.stinckwich(a)gmail.com>: > On Tue, Sep 6, 2016 at 6:24 AM, HernÃ¡n Morales Durand > <hernan.morales(a)gmail.com> wrote: > > > > I am happy to announce the release of Territorial, a new Smalltalk > library > > for Geographical Information Retrieval in geopolitical objects. > > Great ! At the moment, I don't have immediate needs of this but this > might change in the future. > > > Blog post with details: > > http://80738163270632.blogspot.com/2016/08/territorial.html > > The link does not work, apparently the correct link is : > http://80738163270632.blogspot.fr/2016/09/territorial.html > > Thanks! I forgot to update, yes the correct link is http://80738163270632.blogspot.fr/2016/09/territorial.html > > User Manual: http://bit.ly/2c4RrCJ > > Nice documentation ! > > I see you are using github. Try to setup a CI job and maybe put your > code also on github. > Ask me if you need help. > > Regards, > -- > Serge Stinckwich > UCBN & UMI UMMISCO 209 (IRD/UPMC) > Every DSL ends up being Smalltalk > http://www.doesnotunderstand.org/ > >

Sept. 6, 2016

Re: [Pharo-users] [ANN] Territorial

by HernÃ¡n Morales Durand

Hi Stephan, 2016-09-06 2:52 GMT-03:00 Stephan Eggermont <stephan(a)stack.nl>: > On 06/09/16 06:24, HernÃ¡n Morales Durand wrote: > >> >> I am happy to announce the release of Territorial, a new Smalltalk >> library for Geographical Information Retrieval in geopolitical objects. >> > > Nice. Please tell us about your license choice > > License of the library is AGPL v3 (it is in the Notes and disclaimers of the manual) License of the documentation is CC BY-SA 3.0 Cheers, HernÃ¡n > Stephan > > > >

Sept. 6, 2016

Re: [Pharo-users] priority for startup actions

by Peter UhnÃ¡k

> single startup file that's too messy > they are executed in reading order, so if you name them something10.st, something20.st, etc. they will be in order. ah, I'll try that Thanks, Peter On Tue, Sep 6, 2016 at 9:44 AM, Esteban Lorenzano <estebanlm(a)gmail.com> wrote: > another solution is to put all your dependent startup actions in a single > startup fileâ¦ then you can load them in the order you want. > > Esteban > > > On 06 Sep 2016, at 09:43, Esteban Lorenzano <estebanlm(a)gmail.com> wrote: > > they are executed in reading order, so if you name them something10.st, > something20.st, etc. they will be in order. > > Esteban > > On 05 Sep 2016, at 10:10, Guille Polito <guillermopolito(a)gmail.com> wrote: > > Do you really mean startup actions or startup scripts (the .st files you > put in your settings directory)? > > -------- Original Message -------- > > Hi, > > Is there a way to specify a priority for startup scripts? > > I would like to setup some environment for all the startup scripts > (related to cross-platform compatibility), but I donât really understand in > what order are the startup scripts executed. > > Thanks, > Peter > > > >

Sept. 6, 2016

Re: [Pharo-users] PunQlite problems

by Esteban Lorenzano

Hi, most probably not related to the problem, but the currently maintained (for us) PunQLite is now here: https://github.com/pharo-nosql/PunQLite <https://github.com/pharo-nosql/PunQLite> and correct way to install it is: Metacello new repository: 'github://pharo-nosql/PunQLite/repository'; baseline: 'PunQLite'; load. â¦ even if it is basically the same (and we submit changes to mumez repository too), this is the correct place for us :) cheers, Esteban > On 06 Sep 2016, at 00:37, PBKResearch <peter(a)pbkresearch.co.uk> wrote: > > Hi â just tried a further experiment, with more mysterious results. I created a database in memory with a subset of my data. This time the program will retrieve the values as well as the keys. However, the value presented is just the first 1024 bytes, i.e. the first buffer full. In addition, in the majority of cases the correct bytes are each followed by three zero bytes, e.g. the data begins: > 3C 00 00 00 68 00 00 00 > 74 00 00 00 6D 00 00 00 > 6C 00 00 00 20 00 00 00 > 63 00 00 00 6C 00 00 00 > Rather than the expected: > 3C 68 74 6D 6C 20 63 6C. > This affects 75 of the 88 cases in my trial database. > > I presume the insertion of the zero bytes took place in the writing stage, but I have no idea what is going on. Are there any known problems with PunQLite on pharo 5? > > > > Hello > > I am using Moose 6.0 (Pharo 5.0 Latest update: #50761) on Windows 10. I have installed PunQLite following the instructions in https://github.com/mumez/PunQLite <https://github.com/mumez/PunQLite>. I have run all the examples that involve using files, and everything seems OK. I have tried creating a database using my own data, which seemed to run OK. The database will display its keys successfully, but whenever I try to retrieve a value I get PqFetchError. The debug information shows that an ffi fetch operation has given a return value of -10. I have searched for any information on unQLite return codes, but so far I have nothing. > > I tried the PqDatabaseBrowser in the Tools section. This recognised the database, opened it and displayed all the keys correctly. However, clicking on any key gave the same PqFetchError. > > Does anyone have any idea of what I have done wrong? > > Thanks for any help. > > Peter Kenny

Sept. 6, 2016

Re: [Pharo-users] priority for startup actions

by Esteban Lorenzano

another solution is to put all your dependent startup actions in a single startup fileâ¦ then you can load them in the order you want. Esteban > On 06 Sep 2016, at 09:43, Esteban Lorenzano <estebanlm(a)gmail.com> wrote: > > they are executed in reading order, so if you name them something10.st <http://something10.st/>, something20.st <http://something20.st/>, etc. they will be in order. > > Esteban > >> On 05 Sep 2016, at 10:10, Guille Polito <guillermopolito(a)gmail.com <mailto:guillermopolito@gmail.com>> wrote: >> >> Do you really mean startup actions or startup scripts (the .st files you put in your settings directory)? >> >> -------- Original Message -------- >>> Hi, >>> Is there a way to specify a priority for startup scripts? >>> >>> I would like to setup some environment for all the startup scripts (related to cross-platform compatibility), but I donât really understand in what order are the startup scripts executed. >>> >>> Thanks, >>> Peter >

Sept. 6, 2016

Re: [Pharo-users] priority for startup actions

by Esteban Lorenzano

they are executed in reading order, so if you name them something10.st <http://something10.st/>, something20.st <http://something20.st/>, etc. they will be in order. Esteban > On 05 Sep 2016, at 10:10, Guille Polito <guillermopolito(a)gmail.com> wrote: > > Do you really mean startup actions or startup scripts (the .st files you put in your settings directory)? > > -------- Original Message -------- >> Hi, >> Is there a way to specify a priority for startup scripts? >> >> I would like to setup some environment for all the startup scripts (related to cross-platform compatibility), but I donât really understand in what order are the startup scripts executed. >> >> Thanks, >> Peter

Sept. 6, 2016