Messages in this thread | | | From | "Stephen C. Tweedie" <> | Date | Mon, 8 Nov 1999 11:44:16 +0000 (GMT) | Subject | RE: Buffer and page cache |
| |
Hi,
On Sun, 7 Nov 1999 21:49:46 -0700, "braam" <braam@carissimi.coda.cs.cmu.edu> said:
-> > What appears to be needed is the following - probably it's mostly -> > lacking in my understanding, but I'd appreciate to be advised how to -> > attack the following points: -> -> > - a bit to keep shrink_mmap away from the page. -> -> Yes, bumping the page count is the perfect way to do this.
> If we do that, how can the VM still exert pressure to get rid of dirty > pages?
We were talking about pages with IO in progress IIRC. There is no need to worry about VM pressure on those.
More concerning is the general issue of memory pressure on filesystems which, due to complex mechanisms such as transaction write-ordering requirements or defered allocation of disk blocks to data, cannot arbitrarily flush individual pages back to disk on demand. For such long-lived structures in memory, we do need a decent place for the VM to exert pressure --- this is still being worked out for 2.3.
> Great I agree: a file system generic pointer would help a lot. Can we > have it?
Once we've persuaded Linus, yes. :)
--STephen
- 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/
| |