Messages in this thread | | | From | Benno Senoner <> | Subject | Re: can't mlockall() more than 128MB, is this a kernel limitiation ? | Date | Sat, 5 Aug 2000 18:46:23 +0200 |
| |
On Sat, 05 Aug 2000, Benno Senoner wrote:
hmmm .. I looked at the source and seems that I can answer this question by myself:
> > Now my question: is this more than 128MB mlock() problem a limitation of the > 2.2.x kernel or can it be lifted by some sysctl ?
Yes, see mm/mlock.c ----- /* we may lock at most half of physical memory... */ /* (this check is pretty bogus, but doesn't hurt) */ if (locked > num_physpages/2) goto out; -----
Argh ! this is bogus ! , on my 250MB RAM BOX , I want to be able to mlock at least something like 200-220MB ,
What is the purpose of this artificial limitation ? I agree that running out of phys pages is bad, but with those big mem sizes like 256-512MB, a margin of 30-50MB (configurable) should be enough.
Can we make this configurable via sysctl in 2.4 ? (without this realtime multimedia apps will be unable to take advantage of the full amount of RAM, because swappable mem is useless in certain cases)
> > Does kernel 2.4 have the same limitiation ?
Yes :-(
same code present in 2.4-test6-pre2.
But I was really wondering : am I the only one that needs more than ramsize/2 mlocked() mem in a single process ? hopefully not ...
let me know
thanks, Benno.
- 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/
| |