`just slow it`....
Stefan Matthias Aust
sma at 3plus4.de
Sat Feb 10 10:35:20 UTC 2001
At 21:27 08.02.01 -0800, you wrote:
>Thanks for the information!
You're welcome.
>> [OS color depth should be Squeak color depth]
> > Good point. Unfortunately, the Unix VM has no 24 bit mode
> > which would be my OS mode. This could explain some part of the delay.
>
>It most certainly does. I am not sure (Unix people might be able to answer
>that question) but I also think that using the X shared memory extensions
>makes a huge difference. Is there any way of finding out if it's used or not?!
Well, I meant "Squeak" not "Unix VM". My X has a color depth of 24 bits
which isn't matched by Squeak. This means the VM has to translate Squeak's
16 bit or 32 bit into 24 bits. I use the "-xshm" extension so if shared
memory is available, I should get it. However, I never really noticed any
significant difference which might be because of the Sunray client stuff.
>Quite a difference I'd say. Must be Bowie ;-)
Okay, I rechecked while playing "Rosenstolz" and it's still the same
performance... :-)
Ahem, but here's more serious observation: I recompiled the Solaris VM
using Sun's CC and got a 13% performance improvement. I used CC=cc,
CFLAGS=-xO5 vs. CC=gcc, CFLAGS=-O3. Then I also checked the generated
SPARC assembler code (weird stuff) for interp.c and noticed also SUN's
compiler generates the superfluous "if byte >= 255" test. So it could be
even faster if you'd help it with some assembler, similar as the gnuify
script rewrites the switch to allow gcc better optimization.
>Most of these could be display related. Here is an interesting benchmark:
>[...]
I'll try it out Monday. Thanks.
From the windows benchmark, it seems that only ~8% are used to force the
display to the screen. 6% of processing time are used (wasted) to round the
corners. 18% are spent for layouting the browser. It takes 27% to build
the system categories list and of that, 13% to do no highlight.
> >[message tally]
>This looks like you might try to disable #smartUpdating of browsers (it's
>a preference). But everything else looks quite reasonable to me.
It's nice that you and Bob and others find places where you can save some
time by disabling features. However, I want all the features and still a
fast system. For the user's point of view, it's a poor choice just to
disable features to get a reasonable performance.
I think, the problem is not smart updating but how it's
implemented. Browsers should notify each other (as for example done in
Dolphin Smalltalk) and not poll for changes. Iehk.
bye
--
Stefan Matthias Aust \\ Truth Until Paradox
More information about the Squeak-dev
mailing list
|