Is this a reasonable way to implement a stack ?

[Harovan]

Originally posted by Urban Ranger
For a general purpose CPU with a hardware stack pointer, the size of the stack is bound by the size of the stack pointer. Microcontrollers are more interesting since the available RAM is usually much more limited.

Are we meddling call stacks with storage stacks? Come on UR, what kind of stack he means is glaring obvious from the opening post.

[DaShi]

Be nice, he's trying to save face.

[Skanky Burns]

Originally posted by Sir Ralph
What's interesting is, that I used them (try quoting my post), but they didn't do squat other than changing the font to Courier.

Possibly because you used html code tags rather than vB-code code tags (try quoting my post).

[Harovan]

Originally posted by Skanky Burns
Possibly because you used html code tags rather than vB-code code tags (try quoting my post).

I have to steal loinburgers excellent smiley here:

[loinburger]

Originally posted by Urban Ranger
For a general purpose CPU with a hardware stack pointer, the size of the stack is bound by the size of the stack pointer.

It is true that you would have a bit of difficulty growing a stack beyond 4294967296 bytes on a 32-bit machine, but I don't think that necessarily justifies allocating 4294967296 bytes for a stack.

[Provost Harrison]

Thankyou for this thread. This week I have had difficulty sleeping, but now I just have to glance at this thread and I am unconscious for days...

[Harovan]

Originally posted by Provost Harrison
Thankyou for this thread. This week I have had difficulty sleeping, but now I just have to glance at this thread and I am unconscious for days...

Does this mean you won't come again for a few days with your insightful additions?

Oh noes! How will we survive this?

[Urban Ranger]

Originally posted by Sir Ralph
Are we meddling call stacks with storage stacks? Come on UR, what kind of stack he means is glaring obvious from the opening post.

It makes no diffence. The call stack is just a special stack due to hardware awareness. In fact, some CPUs have no stack pointers and you can use any register to create the call stack.

[loinburger]

Originally posted by Urban Ranger
It makes no diffence. The call stack is just a special stack due to hardware awareness. In fact, some CPUs have no stack pointers and you can use any register to create the call stack.

Originally posted by loinburger
It is true that you would have a bit of difficulty growing a stack beyond 4294967296 bytes on a 32-bit machine, but I don't think that necessarily justifies allocating 4294967296 bytes for a stack.

[DaShi]

[loinburger]

Originally posted by DaShi

If you've got a 32-bit machine, then it's tricky (but certainly not impossible) to address more than 2^32 bytes of memory. So you could allocate a 2^32 byte fixed size array to implement a stack, and then say "But I didn't unnecessarily bound the size of the stack!" to defend your design choice.

[Urban Ranger]

Originally posted by loinburger
It is true that you would have a bit of difficulty growing a stack beyond 4294967296 bytes on a 32-bit machine, but I don't think that necessarily justifies allocating 4294967296 bytes for a stack.

Unless you have a special machine that can exceed the hard 32-bit limit addressing space, I don't think it's physically possible to allocate that much to the stack.

Besides, you probably will run into physical limits before that, esp. with microcontrollers.

[loinburger]

If your stack implementation is external then you're limited only by the capacity of your persistent storage device(s).

[Fve Crathva]

This is a very silly discussion.

SP

[Asher]

Originally posted by Urban Ranger
The simplest way is to use a fixed sized array. This way you will never get a stack overrun or underrrun.

Honestly, what is your alma mater? This is fascinating.