Messages in this thread | | | Date | Thu, 27 May 1999 19:54:22 -0400 (EDT) | From | Emil Briggs <> | Subject | Re: Q: PAGE_CACHE_SIZE? |
| |
> >> For large amounts of memory on fast boxes you want a higher page >> size. Some vendors even pick page size based on memory size at >> boot up. > >This sounds suspiciously like the 'larger-blocks-for-larger-FSes' >tactic other systems have been using to hide the bad scalability >of their algorithms. > >A larger page size is no compensation for the lack of a decent >read-{ahead,back,anywhere} I/O clustering algorithm in the OS. >I believe we should take the more appropriate path and build >a proper 'smart' algorithm. Once we're optimizing for I/O >minimization, CPU is relatively cheap anyway... >
Another reason for large page sizes has nothing to do with I/O. On an architecture where TLB misses are expensive (e.g the Origin2000) a larger page size can have a big effect on performance for applications that access a lot of memory. Running one numerical simulation that I use on the O2K with 16K pages is around 20% slower than running it using 64K pages.
Regards Emil
- 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/
| |