Hi Jean-Daniel,
On Jan 16, 2009, at 09:32, Jean-Daniel Dupas wrote:
You don't want to understand.
I'm sorry, if it appears to be like I do not want to understand. I've
been working all night on this, and I'm quite sleepy right now.
My results after changing to plain C are much better. My conclusion is
based upon what I can see as results, they may still be wrong.
You CANNOT mesure elapsed time for ONE call. You have to bench
thousand of call in one loop and take an average time.
Yes and no. It depends on what the purpose is.
To see an average value, you'll have to do just that, however, in my
case, to discover when/why the chops happen, I did this - initially
because I thought that my code sometimes was to blame.
You first conclusion IS WRONG and it does not depend if you have 20
threads or only one. Your PEAK VALUE is WRONG too. It's not possible
to mesure a peak value, especially in a preemptive environment.
You're right about that, and I agree to a certain point.
My rendering is pretty quick (it's using highly optimized code and
already prepared data).
When the code I have is using a lot of ObjC invocations, it appears to
me that it's more likely to be hit, than when it doesn't.
-Both in theory, and when I look at the result.
The only way I can explain that it works better with clean C (apart
from superstitious explanations), is that something that happens in
the message dispatcher, is not happening anymore.
I've tried debugging into the message-dispatcher, however I failed to
coninue at some point, which is expected (I've written debuggers in
the past as well).
Love,
Jens
_______________________________________________
Cocoa-dev mailing list (Cocoa-dev@lists.apple.com)
Please do not post admin requests or moderator comments to the list.
Contact the moderators at cocoa-dev-admins(at)lists.apple.com
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/cocoa-dev/archive%40mail-archive.com
This email sent to arch...@mail-archive.com
