Messages in this thread | | | From | (Ramakrishna K) | Subject | Re: swap cache | Date | Sat, 19 Dec 1998 18:54:22 -0500 (GMT) |
| |
> > On Sat, 19 Dec 1998, Prasun Kapoor wrote: > > > > Linux VM, but we don't need that functionality: in Linux we have always > > > swapped out ptes on a per-page-table basis, not on a per-physical-page > > > basis. We *DO* swap processes out independently. > > > > I dont think I understand. The whole idea of doing it globally is > > to do some kind of LRU on a global scale and swap out pages based > > on that. > > This terribly fragments swap space -- you seem to be forgetting > that we also need to swap things back _in_ and that is most > efficient when the process' pages are contiguous on disk. > > LRU is just a good guess, the method Linux uses might be a bit > worse on swap-out, but since we do less disk _seek_ I/O we come > out slightly ahead of the normal way... > > For Linux 2.3 we are planning several major enhancements to > this scheme though (much better I/O clustering and swap read- > ahead/behind and several other goodies).
read ahead/behind i assume refers to clustering for I/O.
Also i was thinking about one of the other suggestions that was made regarding the compression of a page(s) to a portion of the RAM, and uncompress it as part of the swap back in. Although the compression/un-compression has to be measured against the I/O cost. This will also involve looking at the available RAM. On a small system with less RAM, this may not be a wise-thing to do.
It would be intersting to know if any1 has implemented such a thing and what's the cost involved vis-a-vis available RAM, i/o etc.
thanks, Rama.
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/
| |