Menu
▾
▴
squeak-vmdev
|
From: Tim R. <ti...@su...> - 2002-09-13 23:59:55
|
I'd be ever so happy to see Ian's latest stable sources uploaded to SF and available that way along with all the other platforms (not to mention ditto for windows sources). So far as I can tell everyone who had any hand in altering the previous generation of sources has waved their hand and said someting to the effect of 'feel free to delete my old crap and replace it, Ian', so there should be no problem there. If Ian is too busy right now to deal with it, perhaps one of the more CVS competent of us could whack it in? And I know Andreas is painfully busy right now, so he may 'be behind' a little too. Of course, if I'm simply looking at things the wrong way and it is all already up to date, I'm somebody will slap me around and let me know I'm getting still more senile. Having got the stuff in there at all it would be ever so nice to have it up to date. tim -- Tim Rowledge, ti...@su..., http://sumeru.stanford.edu/tim Any nitwit can understand computers. Many do. - Ted Nelson |
|
From: Andreas R. <And...@gm...> - 2002-09-14 00:53:44
|
Tim, > I'd be ever so happy to see Ian's latest stable sources uploaded to SF > and available that way along with all the other platforms (not to > mention ditto for windows sources). What exactly do you mean?! The latest stable Windows source _are_ on SF. Cheers, - Andreas |
|
From: Ian P. <ian...@in...> - 2002-09-14 01:17:37
|
Hi Tim, > effect of 'feel free to delete my old crap and replace it, Ian', so > there should be no problem there. If Ian is too busy right now to deal > with it He's not too busy to (interrupt his regular hacking to) deal with it -- he's just currently wading through a bunch of changes in the sources. One he reaches shore again he'll drop the anchor in the relevant CVS repository. Real Soon Now. Promise. (For anyone who cares the changes include things like Andreas' recent mods to gl, some nifty stuff Ned send me for the teletype morph, giving the RePlugin a chance to settle down [which seems to have happened now, with corrected capitalisation to boot], checking some Solaris configuration weirdness, etc. On a day keeps the doctor at play.) > Of course, if I'm simply looking at things the wrong way and it is all > already up to date Everything is relative. (In this case, relative to where you look for your sources. ;) > somebody will slap me around I'll pass on that one: don't seem to have a wet fish handy... Ciao, Ian |
|
From: Tim R. <ti...@su...> - 2002-09-14 02:52:13
|
> > > somebody will slap me around > > I'll pass on that one: don't seem to have a wet fish handy... Hmm, I think I have some nice King salmon in the freezer. I'll get Bridget to slap me with it before we bake it with the fresh lemons & oranges out of the garden :-) The original reason I asked was because I'm finally getting to check some of the proposed vm changes stacked up on sqfixes; the crosby cache stuff etc. Obviously it's simplest if I can try with all latest stable code. I've built & tried on Acorn and to be honest it makesno difference forth old St-80 Green Book benchmarks. I know Scott claimed improvements in some macro benchmarks so I'll have to dig them out to try. I suspect that the puny cache on the strongarm is simply overwhelmed and hiding any effects of the changes. I really wish TI had carried through with their threat to make wafers with hundreds of SAs and many gigs of ram all integrated into a single unit.... tim -- Tim Rowledge, ti...@su..., http://sumeru.stanford.edu/tim "Bother," said Pooh, as he flunked the the sobriety test. |
|
From: Tim R. <ti...@su...> - 2002-09-14 19:28:06
|
Just to keep y'all up to date, the Acorn has no performance change at all on the 'Smalltalk macroBenchmarks' when using the crosby cache stuff. It does however seem to require a bunch more memory to achieve that same performance.... tim -- Tim Rowledge, ti...@su..., http://sumeru.stanford.edu/tim Hard work pays off in the future. Laziness pays off now. |
|
From: Ian P. <ian...@in...> - 2002-09-14 19:42:40
|
Tim, > Just to keep y'all up to date, the Acorn has no performance change at > all on the 'Smalltalk macroBenchmarks' when using the crosby cache > stuff. It does however seem to require a bunch more memory to achieve > that same performance.... Want to send me the change set? I'll file it in, rebuild and run the benchmarks. From what I remember of his posts he's talking about subtle effects (cache line invalidation cost) that are rather specific to his processor (some flavour of pentium clone) -- unless I'm reading _completely_ the wrong thread? This is pretty inspirational stuff! Unfortunately, what it inspires is a resounding "who the f**k cares?". His cpu might have a broken cache architecture, mine doesn't. (And given your results it would seem the ARM doesn't either.) Cheers, Ian |
|
From: Tim R. <ti...@su...> - 2002-09-15 00:21:27
|
In message <Pin...@ti...>
Ian Piumarta <ian...@in...> wrote:
> Want to send me the change set? I'll file it in, rebuild and run the
> benchmarks.
It's supposed to be http://swiki.gsug.org:8080/sqfixes/2201.html but I
can't get to it right now. I've dropped a copy on
http://sumeru.stanford.edu/tim/pooters/SqFiles/deltas/VMDevBcommonSendFi
xup.2.cs (and yes I realize that is a very confusing name but blame
scott not me!).
You'll note that there is a very odd and unused Interpreter class var by
the name of Must in there that I think is a stupid typo from a bodged
compile some where in a method with an improperly done comment saying
"Must be green " or somesuch. Now the really odd thing is that the first
time I filed this is the classvar CacheProbeShift went missing along
with Must. And it somehow corrupted the image in the process so I
couldn't restart it. Very odd.
>
> >From what I remember of his posts he's talking about subtle effects (cache
> line invalidation cost) that are rather specific to his processor (some
> flavour of pentium clone) -- unless I'm reading _completely_ the wrong
> thread?
That seems to have been the thrust of his claims. Of course, it
completely ignores any other kind of cpu or device. Amongst other things
that irritated me considerably was the repeated claim that you _must_
add more memory and _of course_ everyone can afford (or even fit)
hundreds of MB ram. I'm all for trying things out and finding ways to
benefit common cases, but really, claims like that just piss me off.
PDA's anyone? Older machines? Bare hardware for embedded stuff? Bah.
Anyway, I'll be interested to see what results you get. I've run the
modified vm with both the Smalltalk macroBenchmarks and the old Green
book benchmarks
(http://sumeru.stanford.edu/tim/pooters/SqFiles/deltas/sqBenchmarks.cs)
with essentially identical results on the Acorn.
>
> This is pretty inspirational stuff!
>
> Unfortunately, what it inspires is a resounding "who the f**k cares?".
> His cpu might have a broken cache architecture, mine doesn't. (And given
> your results it would seem the ARM doesn't either.)
Not broken, but teeny-tiny. And it's embarrassing to have to admit to
having a tiny cache. Even more embarrassing than it would be to admit to
a miniscule love sausage. Which I'm not. I'll refer you to my wife and
mistresses.
tim
--
Tim Rowledge, ti...@su..., http://sumeru.stanford.edu/tim
Useful random insult:- When a thought crosses her mind, it's a long and lonely journey.
|
|
From: Ian P. <ian...@in...> - 2002-09-15 01:26:23
|
On Sat, 14 Sep 2002, Tim Rowledge wrote: > That seems to have been the thrust of his claims. Of course, it > completely ignores any other kind of cpu or device. My particular device isn't impressed in the slightest. stock #(40788 315749 84660 44396 0 25997 14762) Total 526352 crosby #(42534 370191 100559 49918 0 28470 16948) Total 608620 Forget it. Go spend your time on something worthwile. Ian |
|
From: Tim R. <ti...@su...> - 2002-09-27 00:03:35
|
In message <Pin...@ti...>
Ian Piumarta <ian...@in...> wrote:
> On Sat, 14 Sep 2002, Tim Rowledge wrote:
>
> > That seems to have been the thrust of his claims. Of course, it
> > completely ignores any other kind of cpu or device.
>
> My particular device isn't impressed in the slightest.
>
> stock #(40788 315749 84660 44396 0 25997 14762) Total 526352
> crosby #(42534 370191 100559 49918 0 28470 16948) Total 608620
>
> Forget it. Go spend your time on something worthwile.
John did a quick test run on OS-X (since I can't make a non-debug vm
that actually runs - very wierd) and concluded that there is no
noticable benefit on OS-X either. SO, scratch that particular
'enhancement'. I have to admit to being a little surprised, since
normally one might expect some benefit from a bigger cache. Ah, well,
details, details.
tim
--
Tim Rowledge, ti...@su..., http://sumeru.stanford.edu/tim
"How many Kdatlyno does it take to change a lightbulb?" "None. It sounds perfectly OK to them."
|
|
From: Ian P. <ian...@in...> - 2002-09-29 06:53:08
|
Folks, On Fri, 13 Sep 2002, Tim Rowledge wrote: > I'd be ever so happy to see Ian's latest stable sources uploaded to SF Does anybody know if there is a good reason why RePlugin is sitting all lonesome on its own branch, refusing to participate in checkouts? Maybe somebody (Andrew?) would be so kind as to pull it onto the main branch, so that my sources will actually compile? (If nobody can be bothered then I can probably figure out how to do it.) I will then import unix-3_2-5 into SF (and/or destroy the entire repository in the attempt -- we'll find out shortly). Cheers, Ian |
|
From: <gor...@bl...> - 2002-10-02 10:07:32
|
Ian Piumarta <ian...@in...> wrote: > Folks, > > On Fri, 13 Sep 2002, Tim Rowledge wrote: > > I'd be ever so happy to see Ian's latest stable sources uploaded to SF > > Does anybody know if there is a good reason why RePlugin is sitting all > lonesome on its own branch, refusing to participate in checkouts? Yes, I know the reason. :-) It's because Andrew followed the instructions I gave him to play by the "rules" we agreed on. The idea was (if you recall) that noone but the port maintainers should commit on the trunk. Read more on: ..minnow down, can't remember the page... To get something into the trunk the person responsible for the branch (Andrew) needs to communicate with the port maintainer (Ian etc) and tell him something like "Hey, Ian, my super-nifty-plugin can be merged nto the trunk by typing xxxxx - will you do it?". Perhaps I am dumb but why <insert strong word of choice here> do you think I wrote down that darn guide? > Maybe somebody (Andrew?) would be so kind as to pull it onto the main > branch, so that my sources will actually compile? (If nobody can be I could of course do it, but I do not intend to turn into a CVS nanny just because I wrote the guide about principles we all agreed on. IMHO Andrew should have informed you about his stuff being ready for trunk inclusion - and you, being the port maintainer, should do the deed. If you want help though - of course, just say so. > bothered then I can probably figure out how to do it.) I will then import > unix-3_2-5 into SF (and/or destroy the entire repository in the attempt -- > we'll find out shortly). > > Cheers, > > Ian I may sound irritated but it's not that bad. Nothing a beer at OOPSLA won't fix. ;-) Cheers, Göran |
|
From: Ian P. <ian...@in...> - 2002-10-02 12:39:55
|
Hi Goran, > > Does anybody know if there is a good reason why RePlugin is sitting all > > lonesome on its own branch, refusing to participate in checkouts? > > Yes, I know the reason. :-) It's because Andrew followed the > instructions I gave him to play by the "rules" we agreed on. Good reason. :) I did read your mail to this effect (and understood both it and the "rules"). I think I just worded my question a little imprecisely... > Andrew should have informed you about his stuff being ready for trunk > inclusion - and you, being the port maintainer, should do the deed. It's in Cross not unix so I wouldn't want to act unilaterally. My question should really have been worded thus: Is there a good reason (according to Andrew, Andreas, John or Tim) not to do this? Since nobody spoke up I'll assume it's okay. > If you want help though - of course, just say so. Too late. ;) > I may sound irritated but it's not that bad. Nothing a beer at OOPSLA > won't fix. ;-) That's probably several I owe you by now. Regards, Ian |
|
From: <gor...@bl...> - 2002-10-02 12:52:17
|
Ian Piumarta <ian...@in...> wrote: [SNIP] > It's in Cross not unix so I wouldn't want to act unilaterally. My > question should really have been worded thus: Is there a good reason > (according to Andrew, Andreas, John or Tim) not to do this? > > Since nobody spoke up I'll assume it's okay. Sounds reasonable. > > If you want help though - of course, just say so. > > Too late. ;) Sorry! :-) I was in Greece wrecking a 56 foot sailing boat at the time. And that story is beyond belief... like something from a bad movie. Just consider this question: Did the 56.1 foot Ocean Star (worth about $500000) sailing boat: [ ] Run aground HARD. [ ] Almost pull a catamaran in two pieces during an "anchor duel". [ ] Crash into another sailing boat full of angry Germans :-) [ ] Crash into the dock [ ] Crash into a lamppost folding it to the ground [ ] Crash into a roadsign folding it to the ground [ ] All of the above. > > I may sound irritated but it's not that bad. Nothing a beer at OOPSLA > > won't fix. ;-) > > That's probably several I owe you by now. Nope, after your Socket overhaul I owe YOU. You will be at OOPSLA right? regards, Göran |
Thanks for helping keep SourceForge clean.
X