Re: Weird behavior with dynamically sized arrays

[Dan <dan27649@xxxxxxxxx>]

O.k., o.k., o.k.  I've seen the light.  Scott, thank you for reminding me
of the context of 80k on today's Power i systems.  Of course, this was just
an example, but I'm sure most arrays in our production code aren't too far
off that figure.

Thanks everyone!
- Dan

On Wed, Mar 8, 2017 at 3:20 PM, Scott Klement <rpg400-l@xxxxxxxxxxxxxxxx>
wrote:

> Dan,
>
> The code that makes the dump doesn't know how much memory you have/haven't
> allocated.  That information is only known to YOUR code.
>
> So when you say DIM(5000), the dump program tries to list 5000 elements.
> Again, it doesn't know that you've only reserved memory for 32 elements!
> You said 5000!   So when it tries to read the full 5000 elements, it runs
> into an error -- and therefore prints that the data wasn't addressable.
>
> What frustrates me is to see someone using dynamic memory at all for
> this!  It's only an 80k total size array...   believe me, you'll never in a
> zillion years notice that 80k.  You are wasting time and creating headaches
> by keeping tracking of allocation sizes (and from your description, there
> are bugs in that) and giving yourself headaches keeping track of it for
> WHAT, exactly?
>
> Even if this were megabytes, it might be hard to argue that it's
> worthwhile to use dynamic allocation.  But for 80k?!  Frankly, you're nuts
> for using dynamic memory for that.
>
> And this is coming from someone who LIKES working with pointers!
>
> -SK