Messages in this thread | | | Date | Thu, 26 Nov 1998 23:10:47 -0800 | From | Benjamin Redelings I <> | Subject | Re: [2.1.130-3] Page cache DEFINATELY too persistant... feature? |
| |
Well, a further (unintentional) test shows that this isn't just a problem with the buffer cache or with writes. I did a simple "find . -name '*.c' | xargs grep shrink_mmap"
This command proceeded to fill my PAGE cache (and I assume it wasn't doing much writing) and swap a lot of stuff out to disk as soon as free memory was exhausted. I had at a small (~3Mb buffer cache), and at least 44 out of 63 Mb in the page cache.
So ONE problem, at least, is that the only strategy that seems to be used to free memory is to swap, whereas in this case the page cache should have been chopped. Is there a function that does something like "free X number of pages from the page cache?" I really don't know Linux's MM strategy, so perhaps this is expected behaviour, though it seems kind of "unfair" to me.
If that problem were solved, the swapping problem with the buffers wouldn't be nearly so bad, because I wouldn't have the page cache sitting there and hogging 33% of my RAM. Maybe only 20% :)
Oh, the glories of running a "low memory machine!" (64Mb RAM) I should really learn how the MM subsystem works...
Thanks, -benRI -- If E. Coli was a computer: "OPERON.DLL: Mutation in LACTOSE.DLL. Entire system halted."
Benjamin Redelings I <>< http://sdcc13.ucsd.edu/~bredelin
- 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/
| |