Subject:
|
Re: Memory question
|
Newsgroups:
|
lugnet.robotics.rcx.nqc
|
Date:
|
Tue, 13 Dec 2005 22:17:41 GMT
|
Viewed:
|
6272 times
|
| |
|
|
In lugnet.robotics.rcx.nqc, Brian Davis wrote:
> Most functions in NQC are in-line functions, so in a sense they probably
> *are* in main(). One thing I do is place all my initialization code (as much as
> possible) into in-line functions. That way, when the compiler runs through, it
> can allocate variable space as best it can, reusing variables in different
> functions as they fall in and out of scope. Saves me from running out of
> variables in some cases.
So what your saying is I should split my program into more functions not less,
so that the local ints are destroyed and available for re-use. That's a great
idea, it's the complete opposite of my question but fulfills my goal perfectly
:-).
>
> Another thing you can do is have a separate program do all your
> initialization, and then have the main program run "preinitialized" as it were
> (thanks SMART). The standard firmware doesn't zero-out global variables between
> runs, so *as long as the variable lists correspond*, the global variables
> initialized by one program will be accessible to another.
I'd never thought of that as a way of re-using variables before. I might try it
in future projects if need be.
Thanks
Steve
|
|
Message is in Reply To:
 | | Re: Memory question
|
| (...) Most functions in NQC are in-line functions, so in a sense they probably *are* in main(). One thing I do is place all my initalization code (as much as possible) into inline functions. That way, when the compilier runs through, it can allocate (...) (19 years ago, 13-Dec-05, to lugnet.robotics.rcx.nqc)
|
3 Messages in This Thread:
 
- Entire Thread on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
|
|
|
|
