Subject:
|
Re: Proposal: int program_running()
|
Newsgroups:
|
lugnet.robotics.rcx.legos
|
Date:
|
Thu, 21 Mar 2002 15:07:43 GMT
|
Viewed:
|
1929 times
|
| |
|
|
ROSCO,
Well, I was wrong about the inline function. I sure am learning a lot these
days 8-)
I tried it both ways (inline func and macro)... turns out the macro produces
less code by about 50 bytes. (that will be the last time I post without
trying it out first! sorry about that!)
Also, last night I discovered that in mm_try_join (mm.c) The while loop is
only terminated when a non-free memory block is found. I am wondering what
will happen if you mm_try_join on a block and all the remaining blocks are
free... the next pointer would overflow (wrap-around?). Is there something
that I don't know... like a in-use block at the end that would prevent this?
Joe
"Joseph Woolley" <jwoolley@spamblock.insight.rr.com> wrote in message
news:GtAt5u.MGJ@lugnet.com...
<snip>
>
> I like the idea of the inline function best, but will leave that up for
> debate. I was wrong about the placement though. I think tm.h is a better
> place for it. When it comes to the #ifdef... it has to be there anyway;
> nb_tasks is inside the #ifdef CONF_TM, so this macro/function would
> logically be as well.
>
> Joe
<snip>
>
> "Ross Crawford" <rcrawford@csi.com> wrote in message
> news:GtAM9t.4uo@lugnet.com...
> > In lugnet.robotics.rcx.legos, Joseph Woolley writes:
> > >
> > > I would like to propose that the following inline function (or if • necessary,
> > > macro) be used instead. This would likely be placed in program.h • within
> > > #ifdef CONF_PROGRAM ... #endif
> > >
> > > inline int program_running() {
> > > return (nb_tasks > 3);
> > > }
> > >
> > > Any comments or suggestions? I would make the patch myself, but am not
> > > familiar with the tools to do so.
> >
> > I've thought about this before but never did anything 8?( I think I'd • prefer a
> > macro, thus removing the need for #ifdef. Although you wouldn't want the • #ifdef
> > anyway, as you need to call it in the kernel, and #ifdefing it out would • be
> > rather deterimental 8?)
> >
> > ROSCO
>
>
|
|
Message is in Reply To:
 | | Re: Proposal: int program_running()
|
| I know it is fairly trivial, but it is a start. I am working on some other things... hopefully to be included either in patch form or in 0.2.6 (if/when it comes about). I like the idea of the inline function best, but will leave that up for debate. (...) (23 years ago, 21-Mar-02, to lugnet.robotics.rcx.legos)
|
4 Messages in This Thread:
 
- Entire Thread on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
|
|
|
|
