Re: adding support for posix_fadvise() - Mailing list pgsql-hackers

From Neil Conway
Subject Re: adding support for posix_fadvise()
Date
Msg-id 1067874283.3089.241.camel@tokyo
Whole thread Raw
In response to Re: adding support for posix_fadvise()  (Tom Lane <tgl@sss.pgh.pa.us>)
Responses Re: adding support for posix_fadvise()  (Tom Lane <tgl@sss.pgh.pa.us>)
Re: adding support for posix_fadvise()  (Manfred Spraul <manfred@colorfullife.com>)
List pgsql-hackers
On Mon, 2003-11-03 at 10:01, Tom Lane wrote:
> Neil Conway <neilc@samurai.com> writes:
> > POSIX_FADV_RANDOM doesn't effect the page cache, it just determines how
> > aggressive the kernel is when doing readahead (at least on Linux, but
> > I'd expect to see other kernels implement similar behavior).
> 
> I would expect POSIX_FADV_SEQUENTIAL to reduce the chance that a page
> will be kept in buffer cache after it's been used.

I don't think that can be reasonably implied from the POSIX text, which
is merely:

POSIX_FADV_SEQUENTIAL       Specifies that the application expects to access the specified       data sequentially from
loweroffsets to higher offsets.
 

The present Linux implementation doesn't do this, AFAICS -- all it does
it increase the readahead for this file:
http://lxr.linux.no/source/mm/fadvise.c?v=2.6.0-test7

-Neil




pgsql-hackers by date:

Previous
From: Tom Lane
Date:
Subject: Re: adding support for posix_fadvise()
Next
From: Tom Lane
Date:
Subject: Re: Experimental patch for inter-page delay in VACUUM