Messages in this thread | | | Date | Mon, 27 Apr 1998 17:34:53 +0200 (MET DST) | From | Rik van Riel <> | Subject | Re: Swap vs. ext2 fragmentation avoidance? |
| |
On Sat, 25 Apr 1998, Anthony Barbachan wrote: > From: James Mastros <root@jennifer-unix.dyn.ml.org> > > >Somthing has occured to me reading the discussions on > >memory-fragmentation: why we can't effectivily use ext2fs's > >fragmentation-avoidiance with our memory? I'm shure I'm missing somthing > >/really/ simple here, but I don't know what it is. > > Paging make memory fragmentation irrelevent as a page may be located on any > 4kb block. And there is no speed difference when accessing memory at > different locations. Unless that page has been swapped out the the disk.
- We do need to avoid memory fragmentation, since we need large memory blocks for DMA etc... - We _are_ thinking about using a zone allocator for memory usage, so that we can avoid the fragmentation (and do some easy page-coloring tricks on >16M machines).
We'll probably push this into the kernel in the early 2.3.x series...
Rik. +-------------------------------------------+--------------------------+ | Linux: - LinuxHQ MM-patches page | Scouting webmaster | | - kswapd ask-him & complain-to guy | Vries cubscout leader | | http://www.phys.uu.nl/~riel/ | <H.H.vanRiel@phys.uu.nl> | +-------------------------------------------+--------------------------+
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu
| |