Internal Register File (Custom Instructions)

Hi.

When I was writing the original question, I was assuming that I can

read/write the memory easily if I know the memory's address.

I was wrong.

Recently someone kindly gave me advice on this. I should turn my

VHDL code into a component which connects to the Avalon Bus.

So, I'd like to make this new component for SOPC Builder, but

I've never done this before.

Could anyone tell me where I should look into?

If there are VHDL sample files, it would be great,

but I'll be deeply grateful for any suggestion.

cheers,

shino

Well, what you're asking to do isn't exactly easy, so make sure you really need this to run that fast. If not, I'd recommend sticking with a custom instruction and using a loop.

If so... then you want to make a bus-mastering Avalon bus peripheral. There's a PDF file on Altera's site called "Avalon Bus Specification" you should check out first. This should tell you enough that you could write a VHDL file which has all the peripheral's logic. You'll probably need at least an Avalon Slave (registers) interface to let the CPU control this thing (and let it interrupt the CPU when it's done), and an Avalon Master port so it can read and write memory. Test it out in Altera's simulator by itself to make sure it does what you expect.

Once you have the component's VHDL file, use it to create a new component in SOPC Builder, and add it to your system.

Once again, this is the hard way, but it has the potential to do up to one word every two clocks, so make sure you really need the speed.

Dear Mike,

Thank you very much for your comment.

I didn't realize this could become such a difficult task.

Do I need this to run that fast? Well..., I'm not sure.

Probably I could pick up several other functions for speeding up (= by turning

them into VHDL) to achieve the total speed, but these other functions may

have other difficult problems such as "recursive part", "structures", ... etc.

So, maybe I should stick to this approach a little longer, and see how far I

can go. Anyway, in this stage, any experience (know-how) should be valued.

> There's a PDF file on Altera's site called "Avalon Bus Specification"

> you should check out first. This should tell you enough that you could

> write a VHDL file which has all the peripheral's logic.

Thank you.

I've downloaded the file. I'll tackle it. (At least, I think I can try...)

By the way, when I have questions while I'm trying to make a bus-mastering

Avalon bus peripheral, do you think I can ask here? Or "General Discussion

Forum" would be more suitable?

Cheers,

shino

--- Quote Start ---

originally posted by shino+oct 6 2005, 04:34 am--><div class='quotetop'>quote (shino @ oct 6 2005, 04:34 am)</div>

--- quote start ---

do i need this to run that fast? well..., i&#39;m not sure.[/b]

--- quote end ---

usually you can figure it out from project requirements. or you can build the non-state-machine version and see if it&#39;s fast enough.

<!--quotebegin-shino@Oct 6 2005, 04:34 AM

by the way, when i have questions while i&#39;m trying to make a bus-mastering avalon bus peripheral, do you think i can ask here? or "general discussion forum" would be more suitable?

--- Quote End ---

I&#39;d put VHDL/Verilog/Avalon questions in the General Discussion Forum, and IDE/HAL/C/C++ questions here.

Dear Mike,

Thank you again for your kind remarks.

> >Do I need this to run that fast? Well..., I&#39;m not sure.

> Usually you can figure it out from project requirements.

> Or you can build the non-state-machine version and see if it&#39;s fast enough.

What I meant by "not sure" was the following.

(If I had caused any confusion, I&#39;m sorry.)

Suppose I want to speed up (10-20%) one project, which consists

of hundreds c-coded functions. Profiling the project told me that

function1 occupies 10% of the project. When I&#39;ve turned function1

into "VHDL+loop", the new function1 achieved a fourfold speedup,

but as a total, it&#39;s only 7.5% (0.1 x 0.75 = 0.075) speedup.

So, I have to either

a) turn function1 into VHDL completely

and/or

b) turn other functions into VHDL

as a next step.

Because I knew there were many difficulties in b), I just thought

I should try a), but after reading your comments, I became unsure.

And, the good thing is that this is my first exercise on converting c-

to-VHDL, and I&#39;m allowed to take a little "detour".

> I&#39;d put VHDL/Verilog/Avalon questions in the General Discussion Forum, and IDE/HAL/C/C++ questions here.

Thanks. I&#39;ll follow your advice.

Cheers,

shino

--- Quote Start ---

originally posted by shino@Oct 7 2005, 01:06 AM

suppose i want to speed up (10-20%) one project, which consists

of hundreds c-coded functions. profiling the project told me that

function1 occupies 10% of the project. when i&#39;ve turned function1

into "vhdl+loop", the new function1 achieved a fourfold speedup,

but as a total, it&#39;s only 7.5% (0.1 x 0.75 = 0.075) speedup.

--- Quote End ---

And, if you somehow optimized function1 to take zero execution time(!), you&#39;d still only get a 10% speedup. Is this function1 at 10% the heaviest-use function in the system?

Honestly, and I say this as a guess since I don&#39;t know much about what you&#39;re doing, it looks like you might want to figure out a way to speed up the system clock, since it doesn&#39;t sound like your processing is concentrated in any one place. That or increase caches.

Alternatively, if you think you can keep track of things and have room on the chip, you might want to add another Nios2 processor, with just on-chip RAM or something, and use it as a coprocessor. There&#39;s a Mutex peripheral for synchronizing multiple CPUs like that.

Just some alternatives to think about.

Dear Mike and BadOmen,

Thank you very much for teaching me many possibilities.

Quote(Mike DeSimone @ Posted Oct 7 2005, 11:02 AM)

> And, if you somehow optimized function1 to take zero execution time(!),

> you&#39;d still only get a 10% speedup.

Yes, it&#39;s true...very true... but still...

I thought I could get much speedup, let&#39;s say, 95% speedup by

turning whole function1 into VHDL. When I can get 9.5% (0.1 x 0.95)

speedup by function1, it would make the rest considerably easier.

Well..., was I too optimistic? Do you think 95% speedup unusual?

It wouldn&#39;t be worth even trying?

I&#39;m now reading "Avalon Bus Specification". Thank you for mentioning

"Avalon Slave (registers)", "Avalon Master port", and so on.

This mentioning is really helping me. So far, "Streaming Transfer"

looks promising, although obviously I need (much) more reading.

Quote(Mike DeSimone @ Posted Oct 7 2005, 11:02 AM)

> Is this function1 at 10% the heaviest-use function in the system?

Yes. This function1 is most frequently used. In addition to that,

it&#39;s far simpler than other functions. Because these functions were

all written by other people, simplicity is very important factor to me

when rewriting.

As for speeding up the system clock or multi-processor are not

the options at this moment. But the fact your recommending these

rather than "Avalon Bus component" is interesting. It means

that they are more preferable for the experts like you, right?

Quote(BadOmen @ Posted Oct 7 2005, 12:38 PM)

> I would look through the code to see if you have any calculations

> (not just functions) that are used often that do not run fast on your

> system (maybe they lend themselves well to a custom instruction)

Ah, right. "calculations(not just functions)" yes.

I overlooked the basic fact that even when a function has some

difficulties such as "recursive part", "structures", other parts of the

function can be turned into VHDL. In fact, the status of function1

is exactly this... but still, I overlooked it. What a tunnel vision!

It must be worth checking. Thank you for pointing it out.

Cheers,

shino

--- Quote Start ---

originally posted by shino+oct 11 2005, 03:23 am--><div class='quotetop'>quote (shino @ oct 11 2005, 03:23 am)</div>

--- quote start ---

well..., was i too optimistic? do you think 95% speedup unusual?

it wouldn&#39;t be worth even trying?[/b]

--- quote end ---

i simply have no idea. there isn&#39;t enough info here on what you&#39;re really[/i] trying to do to answer that.

<!--QuoteBegin-shino[/i]@Oct 11 2005, 03:23 AM

so far, "streaming transfer"

looks promising, although obviously i need (much) more reading.

--- Quote End ---

You might not need the streaming stuff. I haven&#39;t needed it yet, myself.

"Streaming transfers" are for when the master doesn&#39;t know how much data it needs to move, but the slave does. I&#39;ve always transferred things in fixed amounts, so I&#39;ve never used streaming.

--- Quote Start ---

originally posted by shino+oct 11 2005, 03:23 am--><div class='quotetop'>quote (shino @ oct 11 2005, 03:23 am)</div>

--- quote start ---

because these functions were

all written by other people, simplicity is very important factor to me

when rewriting.[/b]

--- quote end ---

agreed.

<!--quotebegin-shino@Oct 11 2005, 03:23 AM

as for speeding up the system clock or multi-processor are not

the options at this moment. but the fact your recommending these

rather than "avalon bus component" is interesting. it means

that they are more preferable for the experts like you, right?

--- Quote End ---

It means that they are simpler options. Speeding up the clock is a matter of making sure the chip can go that fast, and either changing some PLL settings or a clock oscillator. Interfacing with other components can still be done at the slower speeds now that SOPC Builder supports multiple clocks. Writing your own Avalon component is just more work. Tedious more than difficult. Also, you can&#39;t use the debugger to debug Avalon components&#39; internals; you have use something like SignalTap.

I have the same problem and need to have some example of doing that. Can we include a new macro function in C header file for setting register indexes or is there any other way of doing this using C language?

Thanks in advance.