- Pharo-dev - lists.pharo.org

Re: [Pharo-dev] The Second Coming of Java article

by Alexandre Bergel

Actually, do I want to have Pharo running on Java? My answer was âyesâ a couple of years before. Today, I want to have Pharo (or at least most of it) running in a web browser. Distributing applications to clients is then so easy... Alexandre -- _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: Alexandre Bergel http://www.bergel.eu ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;. On Nov 24, 2013, at 6:19 PM, askoh <askoh(a)askoh.com> wrote: > http://www.wired.com/wiredenterprise/2013/09/the-second-coming-of-java/ > > The article talks about Java Virtual Machine running dynamic languages like > Scala and Clojure in addition to Java. What prevents the Smalltalk Virtual > Machine from running Java, Scala, Clojure and other JVM languages? Is that > something we can ride on to greater awareness and usage of Smalltalk Virtual > Machine and Smalltalk? > > All the best, > Aik-Siong Koh > > > > -- > View this message in context: http://forum.world.st/The-Second-Coming-of-Java-article-tp4724932.html > Sent from the Pharo Smalltalk Developers mailing list archive at Nabble.com. >

Nov. 25, 2013

Re: [Pharo-dev] about miniimage

by Eliot Miranda

Hi Pavel, Hi All, On Sun, Nov 24, 2013 at 10:11 PM, Pavel Krivanek <pavel.krivanek(a)gmail.com>wrote: > Hi Stef, > > our starting point looks like this: > - we have a method how to produce small image without network etc. > - we are able to load network, Monticello and Gofer in it (this job is > currently broken) > - we are able to load Metacello too - this should be the basic stage for > normal users > - than we are a le to load rest od the system at once > - we have several configurations that we are able to load and unload. > > Our biggest problem is the huge nonmodular step between Metacello image > and full Pharo. I think we shoud move forward using division. To define how > an image without development tools should look like and create two big > configurations for them. Then continue with the next splitting. > > From the practical point of view, it's always faster to remove. something > than to load something. And it's much faster to unload it without > Monticello. So I would use ugly removeAllButPackages: because it's fast, > then fix the problems like obsolete classes an Undeclared, continue with > the pretty unloding part of the configuration and finally loading part will > be easy. > Tangentially can anyone point me to a headless image with a REPL for use in testing Spur? > > -- Pavel > > 24. 11. 2013 v 22:36, StÃ©phane Ducasse <stephane.ducasse(a)inria.fr>: > > > Hi pavel > > > > may be I should start from your miniimage and start to making sure that > the configurations can load then in a second step > > I can make sure that they can unload. > > > > What do you think? > > > > Stef > > -- best, Eliot

Nov. 25, 2013

Re: [Pharo-dev] The Second Coming of Java article

by kilon alios

Its also way simpler and faster to just make java apps communicate with pharo via sockets. You dont need pharo to run JVM , all you need is a way to communicate with it. So why do it the hard way ? Î£ÏÎ¹Ï 24 ÎÎ¿Îµ 2013 11:20 Î¼.Î¼., Î¿ ÏÏÎ®ÏÏÎ·Ï "askoh" <askoh(a)askoh.com> ÎÎ³ÏÎ±ÏÎµ: > http://www.wired.com/wiredenterprise/2013/09/the-second-coming-of-java/ > > The article talks about Java Virtual Machine running dynamic languages like > Scala and Clojure in addition to Java. What prevents the Smalltalk Virtual > Machine from running Java, Scala, Clojure and other JVM languages? Is that > something we can ride on to greater awareness and usage of Smalltalk > Virtual > Machine and Smalltalk? > > All the best, > Aik-Siong Koh > > > > -- > View this message in context: > http://forum.world.st/The-Second-Coming-of-Java-article-tp4724932.html > Sent from the Pharo Smalltalk Developers mailing list archive at > Nabble.com. > >

Nov. 25, 2013

Re: [Pharo-dev] show stpper

by Eliot Miranda

On Mon, Nov 25, 2013 at 3:49 AM, Camillo Bruni <camillobruni(a)gmail.com>wrote: > we have now 5 cases defined as show stopper... > > https://pharo.fogbugz.com/f/filters/62/All-priority > > if that were really the case the image would no longer run ;) > My understanding of the term "show stopper" is a deficiency that is judged unacceptable in an up-coming release, i.e. an issue that must be fixed before the system can be released. It does not refer, necessarily, to something that stops the system functioning. It could be something as non-functional as a splash screen graphic or a copyright issue. -- best, Eliot

Nov. 25, 2013

Re: [Pharo-dev] The Second Coming of Java article

by Eliot Miranda

On Sun, Nov 24, 2013 at 1:19 PM, askoh <askoh(a)askoh.com> wrote: > http://www.wired.com/wiredenterprise/2013/09/the-second-coming-of-java/ > > The article talks about Java Virtual Machine running dynamic languages like > Scala and Clojure in addition to Java. What prevents the Smalltalk Virtual > Machine from running Java, Scala, Clojure and other JVM languages? Is that > something we can ride on to greater awareness and usage of Smalltalk > Virtual > Machine and Smalltalk? > Time and money. > > All the best, > Aik-Siong Koh > > > > -- > View this message in context: > http://forum.world.st/The-Second-Coming-of-Java-article-tp4724932.html > Sent from the Pharo Smalltalk Developers mailing list archive at > Nabble.com. > > -- best, Eliot

Nov. 25, 2013

Re: [Pharo-dev] VM with fragmented memory

by Alexandre Bergel

> If I add, letâs say, 30 000 000 elements into c. After the addition the array will be very very big. Which means that it may be in a different segment memory than the one that contains âcâ. Do you feel this may be a problem? I am thinking about a long jump between c and array. Long jumps cost more than short jumps right? > > I don't think it'll be an issue because statistically it'll affect a minority of collections. No, its not necessarily the case that there may be a higher cost. This depends on how the processor's memory hierarchy works. Further, performance may be more affected by how clustered the elements of the collection are than the collection itself. Ah okay. What do you mean by how the elements are clustered? > Do you feel there are some suboptimal situation with this situation? > > I think it's way too early to worry about this kind of thing. The system needs to work and be tuned. There's plenty of time to address clustering, etc. I am not sure to understand what clustering means here. Thanks again for taking the time to answer questions. Alexandre > > > Alexandre > > NB: I cc the mailing list to see whether other may help in understand better whatâs going on with segment and generations > -- > _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: > Alexandre Bergel http://www.bergel.eu > ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;. > > > > On Nov 19, 2013, at 1:37 PM, Eliot Miranda <eliot.miranda(a)gmail.com> wrote: > > > Hi Alexandre, > > > > > > On Mon, Nov 18, 2013 at 7:40 AM, Alexandre Bergel <alexandre.bergel(a)me.com> wrote: > > Hi Eliot, > > > > Stef told me you are working hard on improving the Pharo VM. > > In particular making the VM support fragmented memory. If I understand correctly, this will relieve the VM from having a continuous block of memory to hold the whole image. Really cool! > > > > A better term is segmented memory. fragmentation in memory management means free space getting divided into too small pieces to be useful. > > > > I have a general question regarding VM. Are you aware of other Virtual machines supporting this? Does the Java VM support fragmented memory? .Net? V8? > > > > Any VM that uses the train algorithm supports segmented memory. Any VM that uses "pools" or "regions" uses segmented memory. VisualWorks' VM supports segments. I'm pretty sure that V8 .Net & HotSpot provide segmented memory. > > > > > > I would like to mention this in a research paper I am working on. > > > > Cheers, > > Alexandre > > -- > > _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: > > Alexandre Bergel http://www.bergel.eu > > ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;. > > > > > > > > > > > > > > -- > > best, > > Eliot > > > > > -- > best, > Eliot

Nov. 25, 2013

Re: [Pharo-dev] VM with fragmented memory

by Alexandre Bergel

Ok, thanks ClÃ©ment for the explanation. what do you mean with "now if you a 3 million of elements to this OrderedCollection, you need to handle specificallyâ. What the VM does exactly in that case? Alexandre -- _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: Alexandre Bergel http://www.bergel.eu ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;. On Nov 25, 2013, at 1:46 PM, ClÃ©ment Bera <bera.clement(a)gmail.com> wrote: > Hello, > > To answer your question, V8 definitely supports memory segments. I don't know about papers but in the code each segment is an instance of Heap (see heap.cc and heap.h in the source). The 2 young spaces are in another contiguous segment of 8Mb. > > Each segment of V8 are one of this kind (all instances of Heap): > > * NewSpace: New object allocations go here. Capacity is 8MB. It's divided into two contiguous semispaces, and periodically collected using Cheney's algorithm. Uses bump-allocation. > * OldSpaces: These allocate memory in 64-128KB chunks of 8KB pages. They come in the following flavors: > * MapSpace: Exclusively for Maps (similar to js::Shape), and Maps *always* go here. > * CodeSpace: Exclusively for Code. > * DataSpace: For flat strings and numbers; can't reference objects, except for maps. > * PointerSpace: For anything not a Map or Code that can reference other objects. > * CellSpace: Seems to store global properties or something. > * LOSpace: For large objects, where "large" means it needs more than ~8KB of storage. > > > I have no proof for hotspot (their code is too complex) but I'm sure they have it too. > > For the other question (adding elements to OrderedCollection), Eliot will answer it in more details, but the main idea is that you have the heap chopped into segments of a fixed size (1Mb in Cog, and 64-128KB chunks in V8), therefore for object bigger than a segment, you need to handle specifically the object, i.e. having a segment of the size of the object. If you add 30 000 elements to an OrderedCollection, the object will still be under 1MB so it is ok, now if you a 3 million of elements to this OrderedCollection, you need to handle specifically, Therefore you have performance loss for object of more than 1Mb but there are not common so you will not notice it (you can hardly have thousands of objects bigger than 1Mb allocated and deallocated regularly, and you need this amount to notice the performance loss). > > An interesting point in V8 is that the data and pointers are not in the same spaces. It would mean in Cog, if I understand correctly, that byte and word objects are not in the same space as other objects. I guess this simplifies the GC logic. I don't know if they did that in VW. > > > 2013/11/25 Alexandre Bergel <alexandre.bergel(a)me.com> > Hi Eliot, > > I have a question still on segmented memory. > > Consider an object âc := OrderedCollection newâ. It is likely that the internal "arrayâ and the object c will be in the same generation and in the same memory segment. > If I add, letâs say, 30 000 000 elements into c. After the addition the array will be very very big. Which means that it may be in a different segment memory than the one that contains âcâ. Do you feel this may be a problem? I am thinking about a long jump between c and array. Long jumps cost more than short jumps right? > > Do you feel there are some suboptimal situation with this situation? > > Alexandre > > NB: I cc the mailing list to see whether other may help in understand better whatâs going on with segment and generations > -- > _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: > Alexandre Bergel http://www.bergel.eu > ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;. > > > > On Nov 19, 2013, at 1:37 PM, Eliot Miranda <eliot.miranda(a)gmail.com> wrote: > > > Hi Alexandre, > > > > > > On Mon, Nov 18, 2013 at 7:40 AM, Alexandre Bergel <alexandre.bergel(a)me.com> wrote: > > Hi Eliot, > > > > Stef told me you are working hard on improving the Pharo VM. > > In particular making the VM support fragmented memory. If I understand correctly, this will relieve the VM from having a continuous block of memory to hold the whole image. Really cool! > > > > A better term is segmented memory. fragmentation in memory management means free space getting divided into too small pieces to be useful. > > > > I have a general question regarding VM. Are you aware of other Virtual machines supporting this? Does the Java VM support fragmented memory? .Net? V8? > > > > Any VM that uses the train algorithm supports segmented memory. Any VM that uses "pools" or "regions" uses segmented memory. VisualWorks' VM supports segments. I'm pretty sure that V8 .Net & HotSpot provide segmented memory. > > > > > > I would like to mention this in a research paper I am working on. > > > > Cheers, > > Alexandre > > -- > > _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: > > Alexandre Bergel http://www.bergel.eu > > ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;. > > > > > > > > > > > > > > -- > > best, > > Eliot > > >

Nov. 25, 2013

Re: [Pharo-dev] PharoLauncher

by Sean P. DeNigris

Damien Cassou wrote > done. Please open the launcher, then its preference settings, and > change the directory. Cool, thanks! ----- Cheers, Sean -- View this message in context: http://forum.world.st/PharoLauncher-tp4723991p4725135.html Sent from the Pharo Smalltalk Developers mailing list archive at Nabble.com.

Nov. 25, 2013

Re: [Pharo-dev] VM with fragmented memory

by Eliot Miranda

Hi Alexandre, On Mon, Nov 25, 2013 at 7:42 AM, Alexandre Bergel <alexandre.bergel(a)me.com>wrote: > Hi Eliot, > > I have a question still on segmented memory. > > Consider an object âc := OrderedCollection newâ. It is likely that the > internal "arrayâ and the object c will be in the same generation and in the > same memory segment. > There is no guarantee but it is very likely they will end up in the same segment, yes. > If I add, letâs say, 30 000 000 elements into c. After the addition the > array will be very very big. Which means that it may be in a different > segment memory than the one that contains âcâ. Do you feel this may be a > problem? I am thinking about a long jump between c and array. Long jumps > cost more than short jumps right? > I don't think it'll be an issue because statistically it'll affect a minority of collections. No, its not necessarily the case that there may be a higher cost. This depends on how the processor's memory hierarchy works. Further, performance may be more affected by how clustered the elements of the collection are than the collection itself. Do you feel there are some suboptimal situation with this situation? > I think it's way too early to worry about this kind of thing. The system needs to work and be tuned. There's plenty of time to address clustering, etc. > Alexandre > > NB: I cc the mailing list to see whether other may help in understand > better whatâs going on with segment and generations > -- > _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: > Alexandre Bergel http://www.bergel.eu > ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;. > > > > On Nov 19, 2013, at 1:37 PM, Eliot Miranda <eliot.miranda(a)gmail.com> > wrote: > > > Hi Alexandre, > > > > > > On Mon, Nov 18, 2013 at 7:40 AM, Alexandre Bergel < > alexandre.bergel(a)me.com> wrote: > > Hi Eliot, > > > > Stef told me you are working hard on improving the Pharo VM. > > In particular making the VM support fragmented memory. If I understand > correctly, this will relieve the VM from having a continuous block of > memory to hold the whole image. Really cool! > > > > A better term is segmented memory. fragmentation in memory management > means free space getting divided into too small pieces to be useful. > > > > I have a general question regarding VM. Are you aware of other Virtual > machines supporting this? Does the Java VM support fragmented memory? .Net? > V8? > > > > Any VM that uses the train algorithm supports segmented memory. Any VM > that uses "pools" or "regions" uses segmented memory. VisualWorks' VM > supports segments. I'm pretty sure that V8 .Net & HotSpot provide > segmented memory. > > > > > > I would like to mention this in a research paper I am working on. > > > > Cheers, > > Alexandre > > -- > > _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: > > Alexandre Bergel http://www.bergel.eu > > ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;. > > > > > > > > > > > > > > -- > > best, > > Eliot > > -- best, Eliot

Nov. 25, 2013

Re: [Pharo-dev] [Vm-dev] Frequent SegFaults in PharoVM with Pharo 3.0

by Eliot Miranda

On Mon, Nov 25, 2013 at 9:37 AM, Sven Van Caekenberghe <sven(a)stfx.eu> wrote: > > On 25 Nov 2013, at 17:56, ClÃ©ment Bera <bera.clement(a)gmail.com> wrote: > > > Yeah you cannot compile without inlining specific messages. You can > disable inlining of #timesRepeat: at image level because we added it for > fun to see if it was easy to do in Opal. Now all other optimizations were > in the old compiler and are mandatories. However, you can now in Opal > disable these optimized messages in a restricted area (such as a class and > its subclasses). > > > > To me, it is kind of a bug that we cannot disable those optimizations, > we should be able to do it (so 1 day in the very very far future we could > have the JIT inlining these control structure allowing us to implement all > these inlined messages in all classes). But one thing is that some methods > such as #whileTrue are not implemented in a way they really work (they have > not stop condition, so if the compiler does not inline it, it gives you an > infinite loop). Another thing is the interrupt point problem as Marcus said. > > > > I don't think your other bug is related Opal optimization, only > #timesRepeat: is new and therefore can be faulty (and in this case, the > bytecode is correct, so I guess it is faulty because it removes an > interrupt point). > > It is a bit scary that there is no explicit, crystal clear list of those > interrupt critical points, especially if skipping them can lead to VM > crashes. > Before we raise the alarm the first thing is to identify the cause of the VM crash. It could still be a VM bug. First track down the bug. Don't try and find work-arounds. Try and find the cause. Only then can one develop a correct fix. Anything else buries the true cause under hacks, a process that is truly scary ;-) > > > 2013/11/25 Marcus Denker <marcus.denker(a)inria.fr> > > > > On 25 Nov 2013, at 16:26, Max Leske <maxleske(a)gmail.com> wrote: > > > > > Thanks ClÃ©ment, that seems to be it. Disabling the timesRepeat > inlining makes most of the builds run through but not all of them. It seems > like thereâs another (Opal related?) problem (with the exact same > symptomsâ¦). > > > > > > Iâve tried disabling all options, just to see what happens: all builds > fail with SegFaultsâ¦ > > > > You can not compile without optimisations. There are places in the image > that would not work with e.g. whileTrue: optimisation disabled as it would > e.g. add an interrupt point where none is now > > And other things, like performance, or e.g. the loop that goes over all > objects, if that creates objects while running you have a problemâ¦ > > > > Marcus > > > > > -- best, Eliot

Nov. 25, 2013