Dear all,

Lately I have been working with some projects where timing is critical and hence have decided to learn how to write more efficient code.

Can you direct me some where or give some pointers regarding things like multiplication, division and ANY other ways of making my code run faster. I am using MikroC pr for PIC.

Best regards

 

Learn Assembly for PIC.

 

Learn Assembly for PIC.

+1
Max.

 

+1
Max.

Oh rubbish.

Mikro C is an beautiful language.

Post some of your code and people will advise.

 

Oh rubbish.

Mikro C is an beautiful language.

Didn't say it wasn't
And I didn't originate the comment, just endorsed the answer.
Max.

 

Dear all,

Lately I have been working with some projects where timing is critical and hence have decided to learn how to write more efficient code.

Can you direct me some where or give some pointers regarding things like multiplication, division and ANY other ways of making my code run faster. I am using MikroC pr for PIC.

Best regards

Use a PIC with hardware multiplier.
Optimize the source codes with the help of examining the disassembly, you'd wonder what some C lines produce compared to others.

Use precomputation by the means of tables, if the computation result is not depending on some varying input.

You can often consider that the actual result of a calculation or part of it will only be within a limited number of results. Even if it is some 1000s of ressults, it can be faster to calculate them once and pack into FLASH or external serial memory.

Learn optimization techniques such as loop unrolling.

 

And I didn't originate the comment, just endorsed the answer.
Max.

Guilt by association.

@ the OP, you can write good efficient code without having to use assembly. It just takes practice.

 

Why not to have an algorithm shown to see how it could be implemented?

Discussing in abstract brings you nowhere. Clock and required timing period would help to decide.

 

Guilt by association.

@ the OP, you can write good efficient code without having to use assembly. It just takes practice.

Assembly you need an IQ of 150+ to make an complex algorithm.

That's why they made C (to keep you from going insane)

Mikro C is an decade long project. So it is very mature.

I have done some pretty lengthy stuff in Mikro C.

It works.

 

Assembly you need an IQ of 150+ to make an complex algorithm.

I write entire applications in .asm, including complex algorithms and IEEE floating point math. My last piece of code was 12,527 lines long.

What is my IQ?

 

I write entire applications in .asm, including complex algorithms and IEEE floating point math. My last piece of code was 12,527 lines long.

What is my IQ?

30-years experience and you formally studied it?

 

Could be 166. I dunno mate.

 

I write entire applications in .asm, including complex algorithms and IEEE floating point math. My last piece of code was 12,527 lines long.

What is my IQ?

Well if it was just a BIT higher...

You would have used C for that.

 

I just love assembly, I dunno, maybe it's a masochistic thing? There is something about being in 'total control of every stinking bit' that makes me feel good.

 

Well if it was just a BIT higher...

You would have used C for that.

I used to know a bloke that did some assembly for the PC.

He wasn't all there.

 

IQ is flawed anyway.

Lots of people with high IQs are a few short of an six pack.

When I buy an six pack of beer I expect six bottles, not four and a half!

 

joeyd you are a very special case and someone I have nothing but respect for. I'd peg your IQ at "super genious" (the same as Wiley E. Coyote ) .

I just don't think assembler is the best choice for someone who may not do this stuff day to day. I once worked exclusively in assembler, back in the bad old days when compilers were very expensive. I was quite happy when first inexpensive then free compilers became available and have never looked back (with minor exceptions).

Using a better algorithm or arranging tasks properly is 90% of the battle. I am content with leaving that last 10% on the table for guys like you to optimize.

 

1. Microchip/AVR have library functions that perfor most of the maths (implemented in inline assembly);
2. Get uC with hardware arithemtic modules;
3. Optimize your code by using more bit operations instead of whole expressions;
4. Use more efficient algoriths;
5. Use comilers with higher optimization;

 

...I am content with leaving that last 10% on the table for guys like you to optimize

I appreciate the nice comments, Ernie. But 10% ain't enough:

Program Memory Bytes Used: 16274
Program Memory Bytes Free:   102

 

Here we go again...... the monthly slanging match. Assy Vs C
Max.