Squeak-dev October 2002

squeak-dev@lists.squeakfoundation.org

183 participants
406 discussions

RE: Progrmaming in Bytecode?
by Swan, Dean 24 Nov '06

24 Nov '06

>> It's not uncommon to find x86 code padded with a lot of NOPs >> that weren't in the original source. > >Cowardly assembler! ;-) > >> [NOPs...] Sometimes they're there to pad the execution pipeline, and >> sometimes it's just because the code generator allocates a whole large >> enough for the largest form of an instruction to simplify the branch >> target calculations. > >We must be talking about M$oft assemblers here. ;^p) Yes. This has been pretty characteristic of MS assemblers as far back as I can remember (i.e. about MASM 2.0 - I remember being quite excited when MASM 4.0 came out - '186 and '286 code!). >The real problem is that the solution in general might require N >iterations for a piece of code with N branches -- if promoting one of them >from short to long causes the next (or, much more problematically, a >preceding) one to become long too. You're such an optimist. O(n) is close to best case for this problem. >This is a weak form of what I suppose one might call an "oscillating >constraints" problem. (Hands up anyone who has ever seen [La]TeX get Hence, stochastic optimization. Even MS uses stochastic optimizers now, even in vb. (didn't want to say that too loud here. ;) ) >The usual cause for "padding" in (capable) assemblers (other than for >prefetch alignment, as you point out) is due to dynamic linking where the >size of call sites isn't known until link time. (On modern systems the Not only prefetch alignment. I have heard, although I don't know the details that sometimes it's desirable to force a dead cycle in the execution pipeline. I can only guess that this would be to influence the scheduling of the execution units for complex instructions, but I consider myself to be only a "casual observer" with respect to this kind of thing. >In Smalltalk you've also got a bigger problem: for very long methods the >range of even unconditional branches might not be enough, at which point >I know this has nothing whatsoever to do with Smalltalk...) Well, to kind of tie this back into something relevant to Squeak - I know a few people have talked about "new" VM instruction sets. It seems that it might help to think about a VM where all instructions are the same size. This could simplify code generation (i.e. compiling methods), and possibly improve performance and *reduce* code size (at least in the case of long methods). The anticipated performance improvement would come from forcing aligned fetches more often and potentially simpler instruction decoding. It's really nice to work with an architecture where all instructions take up one word of memory, take one cyle to execute, and sometimes can cause five or six different useful things to occur during that one cycle. (ok, so I write a lot of DSP code...) The single cycle thing isn't necessarily relevant for Smalltalk, but those "extended" bytecodes are really distasteful to me. -Dean

10 15

[ENH] Undeclared cleanup
by Michael Rueger 23 Aug '04

23 Aug '04

Hi all, this tiny goodie removes all entries without references from Undeclared. Maybe this should be called after every file in. Michael 'From Squeak2.8alpha of 19 January 2000 [latest update: #2052] on 5 May 2000 at 1:27:13 pm'! !Utilities class methodsFor: 'miscellaneous' stamp: 'mir 5/5/2000 13:26'! cleanupUndeclared | toBeRemoved refs | "Utilities cleanupUndeclared" toBeRemoved _ Set new. Undeclared associationsDo: [:assoc | refs _ Smalltalk allCallsOn: assoc. refs isEmpty ifTrue: [toBeRemoved add: assoc key]]. toBeRemoved do: [:each | Undeclared removeKey: each]! !

4 19

[ENH] new String modifier: String>>asPlural
by Norton, Chris 24 May '04

24 May '04

Hi Folks. Here's the preamble from my changeset: ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ "OK. You could say that this is a silly crusade, but I like my Squeak messages to be in proper English (at least when the language you're using is English). I was messing around with the HangMan file-in and I got a walkback that had the following title: 'Error: HangManPhrases classs are not indexable'. I just couldn't stand the extra 's' on 'class', so I created a nifty new String method called asPlural and I modified the sender of the odd string to use the new String modifier. Take it or leave it, but I think this makes auto-generated text just a little easier to read." ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ I hope you find this new String modifier useful too. Cheers! ---==> Chris <<PluralStrings-ccn.cs>>

4 4

[FIX] CollectionFixes
by black＠cse.ogi.edu 06 Mar '04

06 Mar '04

from preamble: "Change Set: CollectionFixes Date: 29 October 2002 Author: Andrew P. Black Change 'Set new' to 'self species new' so that subclasses behave correctly."!

2 1

[BUG][FIX] A Morph that rendered by a TransformationMorph is not saved correctly
by Kazuhiro ABE 14 Oct '03

14 Oct '03

Hi all, A Morph that rendered by a TransformationMorph is not saved correctly. In this case, it needs to save a renderer instead of itself. 1. make an Ellipse morph 2. write a following script: Ellipse turn: 5 3. tick the script 4. save this morph by 'save morph in file' menu of debug halo 5. load the morph by 'load as morph' menu from filelist. 6. tick the script again 7. OOPS!! File in the attached changes, now saving and loading will work. Regards, Kazuhiro ABE NISHIHARA Satoshi

2 1

RE: destructors, how?
by Peter Crowther 18 Jul '03

18 Jul '03

> From: Ragnar Hojland Espinosa [mailto:ragnar@linalco.com] > I want to automatically call self terminate on an object on > destruction (or when it goes out of scope..) Smalltalk objects are not 'destroyed'. They are garbage-collected at some time after no other object references them. The 'some time after' could be microseconds or days. > I can't find out how to write a destructor > or something functionally equivalent. Check out implementors of 'finalize' (IIRC). In essence, you hold onto the object using a weak array; are notified when it is garbage-collected; and clear appropriate fields based on copies of the data in the object. - Peter

17 22

Re: [Modules] C and Smalltalk modules
by Anthony Hannan 10 Mar '03

10 Mar '03

Hello again, here's a simpler description of the modules I presented in my previous email. A module is an object that introduces new objects and specifies changes to existing objects in imported modules. When a module is linked its changes are applied to the existing objects. When a module is unlinked its changes are undone. A module is automatically linked when a process that has the module in its environment becomes active. Process has a new instance variable called environment that points to a module. A module is IN an environment iff the environment directly or indirectly imports the module. A module is automatically unlinked when a process that does NOT have the module in its environment becomes active. Pointers that cross module boundaries are converted to identifiers when unlinked and converted back to pointers when linked. For example, a class Foo introduced in ModuleB may have its superclass Bar introduced in ModuleA, an import of ModuleB. When ModuleB is unlinked, Foo's superclass will contain the identifier #Bar. When it is linked back it will contain the class Bar. If there is no global pointing directly to Bar, its identifier will specify a path to Bar from a global, for example #Classes.Bar. Modules only deal with objects and identifiers and is independent of any class (or classless) structure. Thanks, Anthony

7 7

[ALPHA] Merchandise
by Marcus Denker 13 Dec '02

13 Dec '02

Hi! I've set up an account at cafepress.com http://www.cafeshops.com/squeakev Cafepress is a nice service that allows to set up a small shop for typical mechandising stuff (Mugs, T-shirts, Mouse-pads, stickers...). (see http://www.cafepress.com/cp/info/products/index.aspx) The nice thing is that it a) costs nothing to set up and b) they do all the work. The prices in the squeak-ev shop are a little higher than base price (.50 - $2), all profits go to Squeak Germany e.V. There are only 4 products set up because this needs to be testet first. Anyone in the US care to do alpha-testing? They have a 30-day-money-back garantee. Marcus -- Marcus Denker marcus(a)ira.uka.de -- Squeak! http://squeakland.org

3 5

[BUG] SeesColor Tile gone?
by Bert Freudenberg 11 Dec '02

11 Dec '02

I can't find the seesColor:/isOverColor: tile anymore. But the code is still there. Where did it go? Funny enough, the viewer for the car in the "Building with Squeak" project in the prebuilt Squeak3.3a-4981 image has a "test" category open still showing that tile. But in a newly opened viewer it's gone. -- Bert

3 2

ESUG Free Smalltalk on line now
by Stephane Ducasse 09 Dec '02

09 Dec '02

Hi all The October issue of the ESUG Smalltalk Cd is now onlline at on my webpages. - http://www.iam.unibe.ch/~ducasse/WebPages/ESUGCD/ Stef Dr. Stéphane DUCASSE (ducasse(a)iam.unibe.ch) http://www.iam.unibe.ch/~ducasse/ "if you knew today was your last day on earth, what would you do different? ... especially if, by doing something different, today might not be your last day on earth" Calvin&Hobbes

4 7

Jump to page: