- Pharo-dev - lists.pharo.org

Test - please ignore

by Torsten Bergmann

Oct. 9, 2013

Re: [Pharo-dev] Pharo for the Raspberry Pi

by Nick Ager

Hi Sven and others, I've attached a screen shot of Pharo 20623.image running on the Raspberry Pi - I'd connected through a VNC session into my Mac. The image is almost unusably slow - I don't know if the problem lays in the graphical performance + VNC rather than the bytecode execution. tinyBenchmarks give: 0 tinyBenchmarks '10471204 bytecodes/sec; 307552 sends/sec' I came across a "PrimitiveFailed" walk-back from NativeBoost class>>#isEnabled and just commented out the primitive call and instead returned false. I guess this is probably as the VM sources I have date back to April/May so probably aren't completely compatible with 2.0. I've shared my Raspberry Pi build folder "result" on dropbox, which contains the StackVM executable built on the Pi: https://www.dropbox.com/sh/cnta1hjoo7ptv1j/CELvuMjIRl I haven't tried installing the built StackVM onto a clean Pi distribution so don't know if any of the packages I installed during the build process will also be required for executing the StackVM - don't complain too loudly if just copying the contents from dropbox doesn't work directly on your Pi. I found some more notes I made while trying to compile the VM - though reading them back they seem more like the ramblings of a mad man and I'm sure the problems I ran into compiling the StackVM will be sorted by now: https://docs.google.com/document/d/1ZIzvTmV4zkYrFgOdAzfZzC35iMAqI2jc-44XQO9… Also note that Pharo1.2.1 works with the VM used for Scratch in the Raspberry Pi distribution - without the need to build your own VM. Sorry I don't have many free cycles to support, but hope this might help someone to create a "blessed" Raspberry Pi Pharo VM Cheers Nick On 2 October 2013 12:47, Sven Van Caekenberghe <sven(a)stfx.eu> wrote: > Hi Nick, > > On 02 Oct 2013, at 11:51, Nick Ager <nick.ager(a)gmail.com> wrote: > > > Hi, > > > > I managed to compile a StackVM on the raspberry Pi - a couple of months > ago - though I haven't tried with the latest StackVM source. I meant to > write it up, but life got in the way I'm afraid. > > What you did (and I can't) is indeed what should be done: recompile > locally ;-) > > So the conclusion is that it worked and that you can now run Pharo 2.0 > (and presumable 3.0) as well ? > > That would be excellent and important news ! And the best short term > solution. > > Either you (but you lack time) or somebody else who feels at home with > this stuff should redo your steps and produce the actual VM, so that us > weak high-level programmers can have our Pharo and play ;-) > > Thanks, > > Sven > > > The procedure I took was: > > * From a PC running Pharo install VMMaker packages (I found it was > easiest to run under a Linux (eg Ubuntu) to match the PIs OS. > > * Export the StackVM source > > * Install the build tools on the raspberry Pi > > * Copy exported StackVM source (need to ensure it is copied to exactly > the same place, eg /home/nick/build) > > * Build and test with Pharo2.0.image > > > > I choose to build on the Raspberry Pi as in my experience setting up a > reliable cross-compiling environment can take days. > > > > Here are some notes I made at the time: > > > > On the raspberry PI: > > # install build tools > > sudo apt-get install gcc g++ cmake > > > > # dependencies for vm plugins > > sudo apt-get install libasound2-dev libssl-dev libfreetype6-dev > libgl1-mesa-dev > > > > sudo apt-get install build-essential > > > > # to fix: > > # /usr/bin/ld: cannot find -lSM > > #/usr/bin/ld: cannot find -lICE > > # create the following links in: /usr/lib/arm-linux-gnueabihf/ > > > > sudo ln -s libSM.so.6 libSM.so > > sudo ln -s libICE.so.6 libICE.so > > > > Once the source is installed: > > > > chmod +x platforms/unix/config/version > > chmod +x platforms/unix/config/verstamp > > > > At the time I had to patch a couple of files, but that might have > changed with later versions of the StackVM source. My notes say I modified > the following, though don't indicate what I changed: > > > > platforms/Cross/vm/sqAtomicOps.h > > platforms/unix/vm/sqUnixITimerHeartbeat.c > > platforms/unix/vm/vmVersionInfo.h > > src/vm/gcc3x-interp.c > > > > > > Hope this helps > > > > Nick > > > > > > On 2 October 2013 10:31, Goubier Thierry <thierry.goubier(a)cea.fr> wrote: > > Ok, thanks for the info, Sven. > > > > Thierry > > > > Le 02/10/2013 11:06, Sven Van Caekenberghe a Ã©crit : > > > > Thierry, > > > > On 02 Oct 2013, at 10:14, Goubier Thierry <thierry.goubier(a)cea.fr> > wrote: > > > > Hi all, > > > > I'm trying to know if it is possible to run a version of pharo, recent > if possible, on the Raspberry Pi? > > > > I know that there is a squeak vm on Raspian (Debian for the RaspberryPi) > but it doesn't load recent pharo images. > > > > Using the VM that comes with Raspbian, you can run Pharo 1.4. > > Pharo 2.0 requires additional VM support. > > > > Sven > > > > Thanks, > > > > Thierry > > -- > > Thierry Goubier > > CEA list > > Laboratoire des Fondations des SystÃ¨mes Temps RÃ©el EmbarquÃ©s > > 91191 Gif sur Yvette Cedex > > France > > Phone/Fax: +33 (0) 1 69 08 32 92 / 83 95 > > > > > > > > > > > > > > -- > > Thierry Goubier > > CEA list > > Laboratoire des Fondations des SystÃ¨mes Temps RÃ©el EmbarquÃ©s > > 91191 Gif sur Yvette Cedex > > France > > Phone/Fax: +33 (0) 1 69 08 32 92 / 83 95 > > > > > > >

Oct. 9, 2013

Re: [Pharo-dev] Weird errors while loading code... reason why

by StÃ©phane Ducasse

On Oct 9, 2013, at 7:28 PM, phil(a)highoctane.be wrote: > There is stuff like this in the fileouts and in the MCZ source.st > > !XXXrView methodsFor: 'as yet unclassified' stamp: 'PhilippeBack 9/28/2013 00:25'! > ! ! > > So, nothing in the source of the method. > > Don't know how this occured, maybe refactorings normally not. This is the first time I hear that problem. May be an interaction with an external package doing strange thing with internals? > > > Removing these things allow me to filein the source.st > > Phil

Oct. 9, 2013

Re: [Pharo-dev] Strane error while loading code > Syntax Error : Message pattern expected

by phil＠highoctane.be

Source looked like ...! ! ! basically an empty method. --- Philippe Back Dramatic Performance Improvements Mob: +32(0) 478 650 140 | Fax: +32 (0) 70 408 027 Mail:phil@highoctane.be | Web: http://philippeback.eu Blog: http://philippeback.be | Twitter: @philippeback Youtube: http://www.youtube.com/user/philippeback/videos High Octane SPRL rue cour Boisacq 101 | 1301 Bierges | Belgium Pharo Consortium Member - http://consortium.pharo.org/ Featured on the Software Process and Measurement Cast - http://spamcast.libsyn.com Sparx Systems Enterprise Architect and Ability Engineering EADocX Value Added Reseller On Wed, Oct 9, 2013 at 9:23 PM, Camillo Bruni <camillobruni(a)gmail.com>wrote: > can you get the method source? > try inspecting the MCMethodAddition further down the stack... > > On 2013-10-09, at 19:19, "phil(a)highoctane.be" <phil(a)highoctane.be> wrote: > > > > > I am encountering a weird error (in 2.0) while loading code from a > monticello package. > > > > <09-10-13 19-10-36.png> > > > > This comes after a "compiling" progress box, then I get "Reloading..." > and that one. > > > > A right click on the top entry gives me the following: > > > > <09-10-13 19-14-36.png> > > > > The method is empty and its protocol is #'as yet unclassified' > > > > Really annoying > > > > The package was saved from a working image w/ the app working well. > > > > Ideas on how to fix this? > > Phil > > > > > >

Oct. 9, 2013

Re: [Pharo-dev] SmalltalkHub red

by Tudor Girba

Yes, running the Grease/Seaside tests works fine. But, you will only see the "packages" problems when you do something with packages. So you will have to trigger RPackage tests to notice the issue. After all, you spotted the issue by running the Pharo regression tests after loading Moose/Grease/Seaside. Doru On Wed, Oct 9, 2013 at 9:28 PM, Camillo Bruni <camillobruni(a)gmail.com>wrote: > so is seaside: https://ci.inria.fr/pharo-contribution/job/Seaside/ > > On 2013-10-09, at 17:48, Damien Cassou <damien.cassou(a)gmail.com> wrote: > > On Wed, Oct 9, 2013 at 4:47 PM, Yuriy Tymchuk <yuriy.tymchuk(a)me.com> > wrote: > >> Do you know who is working with grease? It doesn't sound too hard to > solve this. > > > > > > and Grease seems to be ok: > https://ci.inria.fr/pharo-contribution/job/Grease/ > > -- www.tudorgirba.com "Every thing has its own flow"

Oct. 9, 2013

Re: [Pharo-dev] SmalltalkHub red

by Camillo Bruni

so is seaside: https://ci.inria.fr/pharo-contribution/job/Seaside/ On 2013-10-09, at 17:48, Damien Cassou <damien.cassou(a)gmail.com> wrote: > On Wed, Oct 9, 2013 at 4:47 PM, Yuriy Tymchuk <yuriy.tymchuk(a)me.com> wrote: >> Do you know who is working with grease? It doesn't sound too hard to solve this. > > > and Grease seems to be ok: https://ci.inria.fr/pharo-contribution/job/Grease/

Oct. 9, 2013

Re: [Pharo-dev] Strane error while loading code > Syntax Error : Message pattern expected

by Camillo Bruni

can you get the method source? try inspecting the MCMethodAddition further down the stack... On 2013-10-09, at 19:19, "phil(a)highoctane.be" <phil(a)highoctane.be> wrote: > > I am encountering a weird error (in 2.0) while loading code from a monticello package. > > <09-10-13 19-10-36.png> > > This comes after a "compiling" progress box, then I get "Reloading..." and that one. > > A right click on the top entry gives me the following: > > <09-10-13 19-14-36.png> > > The method is empty and its protocol is #'as yet unclassified' > > Really annoying > > The package was saved from a working image w/ the app working well. > > Ideas on how to fix this? > Phil > >

Oct. 9, 2013

Re: [Pharo-dev] Strane error while loading code > Syntax Error : Message pattern expected

by ClÃ©ment Bera

This kind of errors are triggered by the parser. Something is probably miswritten in the method. Are you loading something with the old syntax ? (I mean <- instead of :=). Try in 3.0 to see if Opal can handle this case. 2013/10/9 phil(a)highoctane.be <phil(a)highoctane.be> > > I am encountering a weird error (in 2.0) while loading code from a > monticello package. > > [image: Inline image 1] > > This comes after a "compiling" progress box, then I get "Reloading..." and > that one. > > A right click on the top entry gives me the following: > > [image: Inline image 2] > > The method is empty and its protocol is #'as yet unclassified' > > Really annoying > > The package was saved from a working image w/ the app working well. > > Ideas on how to fix this? > Phil > > >

Oct. 9, 2013

Re: [Pharo-dev] ShortRunArray not in Pharo 3.0 anymore, leading to problems with FFI-Kernel-tbn.25

by phil＠highoctane.be

Well, no issues. It is too complex for me to get things loaded in 3.0 and be delivering what we need to, so I am back to 2.0 now. Spent the day on dealing with configs and packages. I am more educated about all that now anyway. Phil On Wed, Oct 9, 2013 at 7:43 PM, StÃ©phane Ducasse <stephane.ducasse(a)inria.fr>wrote: > Phil > > we are ready to help but then give the expression that shows the problem. > May be this is just that snapshotcello was simply never loaded in 30 > because moose is in 20. > and that there is just a platform definition missing. > > Stef > > > Snashotcello not loading in 3.0 as well and the 2.x trick. > > Also, Gofer gives me some headaches when loading from directories. > > Phil > > - > On Wed, Oct 9, 2013 at 4:05 PM, Esteban Lorenzano <estebanlm(a)gmail.com>wrote: > >> >> On Oct 9, 2013, at 3:30 PM, Marcus Denker <marcus.denker(a)inria.fr> wrote: >> >> > >> > On Oct 9, 2013, at 3:22 PM, phil(a)highoctane.be wrote: >> > >> >> There is an "ExternalObject class >>initialize" that attemps to place >> itself after ShortRunArray, which isn't there, leasgin to "Error: nil >> cannot be found in Smalltalk dictionary". >> >> >> >> Just to let you know, >> >> >> > >> > Do we support non-nativeBoost FFI in Pharo3? >> >> not yet >> >> > >> > >> >> (Was testing the loading of my confs in 3.0 with the 2.0 monticello >> trick". >> >> >> >> Phil >> > >> >> >> > >

Oct. 9, 2013

Re: [Pharo-dev] ShortRunArray not in Pharo 3.0 anymore, leading to problems with FFI-Kernel-tbn.25

by StÃ©phane Ducasse

Phil we are ready to help but then give the expression that shows the problem. May be this is just that snapshotcello was simply never loaded in 30 because moose is in 20. and that there is just a platform definition missing. Stef > Snashotcello not loading in 3.0 as well and the 2.x trick. > > Also, Gofer gives me some headaches when loading from directories. > > Phil > > - > On Wed, Oct 9, 2013 at 4:05 PM, Esteban Lorenzano <estebanlm(a)gmail.com> wrote: > > On Oct 9, 2013, at 3:30 PM, Marcus Denker <marcus.denker(a)inria.fr> wrote: > > > > > On Oct 9, 2013, at 3:22 PM, phil(a)highoctane.be wrote: > > > >> There is an "ExternalObject class >>initialize" that attemps to place itself after ShortRunArray, which isn't there, leasgin to "Error: nil cannot be found in Smalltalk dictionary". > >> > >> Just to let you know, > >> > > > > Do we support non-nativeBoost FFI in Pharo3? > > not yet > > > > > > >> (Was testing the loading of my confs in 3.0 with the 2.0 monticello trick". > >> > >> Phil > > > > >

Oct. 9, 2013