Re: struct lrecord_header size
On Sun, Jun 20, 1999 at 04:59:28PM -0700, Kyle Jones wrote:
Clever. But with this I think we'd be getting non-portable
again. malloc() isn't guaranteed to give us page-aligned blocks
and we'd need that to make the type check be fast.
In modern systems, unused memory has a close-to-zero cost, so we can
always do the "round up" trick, especially if we do the allocation in
chucks of, say, 1Mb.
The effects on the cache may be nasty, though. All type checks will
end up on the same cache line.
OG.