What you can do is, take an existing implementation and preallocate a fixed number of objects in a linked list, like a stack. Then you pop off the first object whereever there is a malloc() and push it on again whereever there is a free()

On Sat, 2010-10-16 at 00:35 +0200, Lieven Moors wrote: -- eins, zwei, drei ... tekno tekno?? http://www.youtube.com/watch?v=ZEgbW1FxR78

On Sat, Oct 16, 2010 at 6:41 AM, lieven moors <lievenmoors(a)gmail.com> wrote: Hi, I'm not sure if this will help you, but I wrote a fixed block size memory allocator which is really simple for the TI89 calculator. I wrote it a while ago, and I think I'm the only one who's actually used it till now, so it's not well tested, but it's available on google code: http://code.google.com/p/lardalloc/ It should fit your requirements, as it just uses the free blocks to implement a stack, and thus is really fast. The only problem you might encounter is that because implemented for a graphing calculator, it is limited to 2^16 blocks maximum. I wouldn't expect it to cause any problems running on a non-calculator platform, because it is written in standard C. Again, you'd still have to write a tree implementation on top of it. Jeremy

On Sat, Oct 16, 2010 at 11:06:19AM +0200, Jens M Andreasen wrote: > What you can do is, take an existing implementation and preallocate a > fixed number of objects in a linked list, like a stack. Then you pop off > the first object whereever there is a malloc() and push it on again > whereever there is a free() > >