Am 07.06.2018 um 12:42 schrieb Martok:
What actually happens is that the memory is released back to the heap
(but *not* to the OS, at least on Linux), with the result that
concatenating elements will introduce a substantial hit particularly if
space for a new element allocated from the heap isn't contiguous.
Writing a preallocating wrapper *where needed* for heavily grown arrays is
fairly simple.
The thing is: it's rarely worth it ;-)
I recently did that for a wavefront mesh loader that appends vertices one-by-one
to an array, and for a scene with some 400k vertices, the difference was just
some milliseconds out of several seconds overall.
Turns out the allocator usually finds a spot where the array doesn't need to
actually be copied around for a while, and the pure bookkeeping of realloc is
very cheap.
It would be worth though to think about extending the memory manager to make use of the realloc syscall (not sure about
its actual name) on some OSes as the OS can "copy" memory to a different virtual address by just re-mapping pages. I
think at least linux really does this.
_______________________________________________
fpc-pascal maillist - fpc-pascal@lists.freepascal.org
http://lists.freepascal.org/cgi-bin/mailman/listinfo/fpc-pascal
