Distraction layer of IDE, inverted by the aplication boundries of C++

Hi All, I am on learning curve towards coming up with a design for creating OS abstraction Layer for one of our legacy product.Heres what I proposed my team:

Create a posix interface compliant API and inside the API have #defines to differentiate between multiple OS.

Is this approach right one?

I have a query regarding this: Every OS will have its own booting requirements in startup phase.How can OS abstraction layer handle this in a neat way?

What are the guidelines interms of design one should have in mind while creating an OS abstraction layer?

Can anyone provide me links ,tips or tutorials on creating OS Abstraction Layer?

Looking farward for your replies and Advanced thanks for the same,

Regards, s.subbarayan _________________________________________________

Modules loaded or compiled after the image is running , is poison . Make files and includes and etc etc are poison ..

They do not work . They cause flood and bloat and doc's of 100's of pages . Forth has no docs nor manuals .

You cant port s/w by flipping a s/w switch ... You must integrate everything , test it , and flash it for each CPU . No Exceptions ., no later additions ...

Linux has problems because they try to "build" a kernel using "make" files , after its compiled ! It does NOT work . Compiling does NOT work .

What does work , is to write a thin layer at the bottom of Host Forth , called Meta Forth . Thus 90% of the host Forth will be ported to target , unaltered Only the thin layer will be targets opcodes Use Forths Prims to create target assembler , and you are 90% done And it goes into the same dictionary ! When your ready , Forth builds the binary , ready to boot on target cpu .

There is no easier nor faster method to port OpSys . And its so simple to show a programmer !

Everything is simpler in Forth .

Reply to
werty
Loading thread data ...

Do you want to comment on the replies from the first time you asked the question? Thread "Developing OS Abstraction Layer" - 8 Jan.

Regards, Richard.

  • formatting link
  • formatting link
    for Cortex-M3, ARM7, ARM9, HCS12, H8S, MSP430 Microblaze, Coldfire, AVR, x86, 8051, PIC24 & dsPIC .... and soon AVR32
Reply to
Bill Blass

Yes, I can help. First of all using C is a dead-end because it's ASCII. That's not the way to do it. I suggest you use image based Forth tools and your right brain. Please stop being a Luddite.

A very clever individual will very soon produce a breakthrough product called Forthrite that will be free and even make your coffee. It's an image based Forth tool. I think if you wait for that. By the time you design, code, and test your distraction layer you could be done with your project using Forthrite.

Good luck and please tell us how it turns out.

JJS

Reply to
johnspeth

The reasoning is simple . If one "waits" for Linux/C/C++ or other

"popular" software , to develope an easier to use OpSys ...

Did you get the point ?

Its us hackers that simplify , we will give a free Forth ,on ARM9

that will be faster than a P4 and much easier to do low level

coding . Any buyer faced with the decision , will reject

Linux et al , and thus hand the power to Forthrite ....

Why fight us ! Join us ..

------------------------------------------------------------------------------- BTW , Interupts on NDS ( Mexican game box "Noentiendo DS ")

are not arbitrary as programers countered ( against me ).

My OpSys polls INTs , it never enables them .

This alllows the kernel to complete its proper tasks ,

and makes Forthrite Mult-Tasking run w/o errors .

NDS ARM9 may appear to enable INTs , but it enables

them when its completeed to top thread , anyway !

The old method , says , as you program , you enable

if there is not a compelling reason NOT to .

My Forthrite OpSys is "first things first" , so Forthrite

polls the next proper INT , thus it never errors as Linux does .

Try to argue that !

I am the fastest Systems Programmer on earth , and Forthrite

will nix Linux and C++ .

I will piggyback PSRAM on top of SDARM and wire the CS OE pins

to ARM 9 MMU so i can boot Linux , in PSRAM

in 0.1 seconds .

___________________________________________

snipped-for-privacy@yahoo.com wrote:

Reply to
werty

I swear one or more of the posters on these streams is just a Markov Model spewing randomness.

Reply to
tbroberg

"werty" is great!

You should do a search for him on comp.sys.hp48. He posts there as well.

I got this gut feeling that if could just decipher his code (like taking the first letters of each sentence, or using the random spacing between words and sentences as an index into a table, etc.), I'd learn where the Arc of the Covenant was hidden, or the landing site of Ancient Astronauts, or how to fill a mayonnaise jar with Dark Matter.

I just can't help from feeling there's something there I'm missing...

-Zonn

--
Zonn Moore            Remove the ".EXAMPLE" from the
Zektor, LLC           email address to reply.
www.zektor.com
Reply to
Zonn

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.