lkml.org 
[lkml]   [1998]   [Mar]   [8]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectRe: mmap() versus read()
I thought madvise() only supplied clues, that it did not tell you how to
respond to those clues?
That would mean that for a ramdisk it could simply be ignored by the
FS. Thus my contention that it enables yet another optimization that
has its place, though not necessarily the most important of places.

Your point in regards to persistence is a good one. One should be able
to do a persistent madvise() saved into the FS, and autodetectors should
employ the same call as well. But I don't volunteer to write that
code.:-)

Best,

Hans

MOLNAR Ingo wrote:
>
> On Sun, 8 Mar 1998, Hans Reiser wrote:
>
> > I disagree. If you page fault more than once, so as to establish a
> > pattern of sequentiality, you are already losing compared to madvise().
>
> not if the (autodetected) access pattern is persistant, eg. you save the
> pattern into the filesystem itself ... this is IMO better than madvise(),
> since madvise() carries only very few information, and madvise() has no
> knowledge about the underlying block device, eg. you might be using a
> solid state disk with no seek cost (or a ramdisk).
>
> thus madvise() might even turn out to be an overhead, on a sufficiently
> smart kernel and/or on sufficiently smart hardware.
>
> -- mingo

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu

\
 
 \ /
  Last update: 2005-03-22 13:41    [from the cache]
©2003-2011 Jasper Spaans