Email: Password: Remember Me | Create Account (Free)

Back to Subject List

Old thread has been locked -- no new posts accepted in this thread

??? 03/07/09 00:23 Read: times	#163186 - Absolutely true! Responding to: ???'s previous message

List of 76 messages in thread

Topic	Author	Date
NXP P89LPC936 Auxilary RAM		01/01/70 00:00
Compiler Cannot Save the Day At Runtime		01/01/70 00:00
I found it.....		01/01/70 00:00
Incorrect!		01/01/70 00:00
To be fair,...		01/01/70 00:00
Its good info to know...		01/01/70 00:00
how is it done in C?		01/01/70 00:00
Maybe you should try doing it ASM!		01/01/70 00:00
You should start a new thread on that!		01/01/70 00:00
Not in this case!		01/01/70 00:00
A compiler should support ALL MCU features ...		01/01/70 00:00
A compiler should translate the language		01/01/70 00:00
Still C - just not strict.		01/01/70 00:00
I didn't say that!		01/01/70 00:00
It is not that clean and clear		01/01/70 00:00
Deviants are deviant		01/01/70 00:00
Keil and SDCC need no macros ...		01/01/70 00:00
Are you sure?		01/01/70 00:00
That's probably correct, but ...		01/01/70 00:00
A C compiler can map the hardware quite well		01/01/70 00:00
compiler vendors are looking at the new processors		01/01/70 00:00
I once mentioned that ...		01/01/70 00:00
It's not what I'd choose, but it is a matter of perefernce		01/01/70 00:00
As flash gets bigger, the code steps do too.		01/01/70 00:00
Still not needed for other architectures		01/01/70 00:00
We will have to agree to disagree ... I guess		01/01/70 00:00
Which single-clocker is cheaper than an ARM?		01/01/70 00:00
RE: Which single-clocker is cheaper than an ARM?		01/01/70 00:00
differs considerably from the classic microcontroller?		01/01/70 00:00
Horses for courses		01/01/70 00:00
Always start each project by scanning the market		01/01/70 00:00
On that we can agree		01/01/70 00:00
Maybe, but what are they comparing?		01/01/70 00:00
Did you actually look?		01/01/70 00:00
Yes, I did.		01/01/70 00:00
it's a tradeoff		01/01/70 00:00
Is it 2006 already?		01/01/70 00:00
Really?		01/01/70 00:00
that's outright silly		01/01/70 00:00
Maybe ... which is why it is not yet the case		01/01/70 00:00
the eyes of the beholder		01/01/70 00:00
Look at it from another viewpoint for a moment		01/01/70 00:00
RE: I'm not the one to ask about IC prices		01/01/70 00:00
doesn't mean I'm totally in the dark		01/01/70 00:00
Richard Erlacher has left the planet		01/01/70 00:00
Maybe Andy's the one who's lost		01/01/70 00:00
I cannot believe you even looked at ARM		01/01/70 00:00
It's all relative		01/01/70 00:00
Price doesn't directly follow processor size		01/01/70 00:00
What about performance?		01/01/70 00:00
Be specific please.		01/01/70 00:00
You can use your own supplier		01/01/70 00:00
Your test simulates as 41.04us		01/01/70 00:00
RE: ARM compiles do not like byte-addressing		01/01/70 00:00
A typo on my part		01/01/70 00:00
Case of full unroll		01/01/70 00:00
... and what does THAT ARM MCU cost?		01/01/70 00:00
WHO CARES		01/01/70 00:00
Absolutely true!		01/01/70 00:00
$5 or lower		01/01/70 00:00
My Simulation times were wrong.		01/01/70 00:00
67% of your loop was your loop		01/01/70 00:00
re-written the loop in C		01/01/70 00:00
Try without loop counter		01/01/70 00:00
You're right!		01/01/70 00:00
Similar trick with ARM7 would require 66.67MHz		01/01/70 00:00
60ns with no store instruction.		01/01/70 00:00
25% speedup		01/01/70 00:00
Actually, I found the for loop faster.		01/01/70 00:00
Compiler setting?		01/01/70 00:00
I made NO optimisation		01/01/70 00:00
Avoid variable aliasing if you like high optimization levels		01/01/70 00:00
The RealView compiler is very competent		01/01/70 00:00
Yes, caching		01/01/70 00:00
Yes, caching		01/01/70 00:00
It's much easier with the 8-bit single-clocker		01/01/70 00:00

Back to Subject List