From: ebiederm+eric@npwt.net (Eric W. Biederman)
To: Jason Crawford <jasonc@cacr.caltech.edu>
Cc: Linux MM <linux-mm@kvack.org>
Subject: Re: Linux wppage patch (fwd)
Date: 30 Jun 1998 12:59:11 -0500 [thread overview]
Message-ID: <m1ww9y7ouo.fsf@flinx.npwt.net> (raw)
In-Reply-To: Rik van Riel's message of Fri, 26 Jun 1998 07:34:10 +0200 (CEST)
>>>>> "JC" == Rik van Riel <H.H.vanRiel@phys.uu.nl> writes:
JC> ---------- Forwarded message ----------
JC> Date: Thu, 25 Jun 1998 21:10:00 -0700 (PDT)
JC> From: Jason Crawford <jasonc@cacr.caltech.edu>
JC> To: h.h.vanriel@phys.uu.nl
JC> Subject: Linux wppage patch
JC> 2. Make a slight change to the way the custom nopage routine is called.
JC> The third argument to nopage is declared as "write_access" in the
JC> definition of the VM operations struct in mm.h. But when it's called, it
JC> is actually "no_share", computed as:
JC> (vma->vm_flags & VM_SHARED) ? 0 : write_access
JC> My code, however, needs to know whether the access was a write even
JC> though it is shared memory, so I would like to change this argument to
JC> just "write_access". Since the VMA is passed in to the routine anyway,
JC> the VM flags will be available, and any routine which wants to calculate
JC> "no_share" can do so. Again, I searched the Linux source tree, and only
JC> the generic filemap_nopage routine uses the no_share argument. It can
JC> easily be changed to accept "write_access" instead of "no_share" and
JC> calculate "no_share" before it does any work.
Your code basically looks reasonable but there is a potential gotcha
in the works.
Shared pages are never write protected by the nopage routine so you
will never discover if a shared page has been written too...
Which could cause all kinds of havoc for distrubuted shared memory.
Eric
prev parent reply other threads:[~1998-06-30 17:47 UTC|newest]
Thread overview: 2+ messages / expand[flat|nested] mbox.gz Atom feed top
1998-06-26 5:34 Rik van Riel
1998-06-30 17:59 ` Eric W. Biederman [this message]
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=m1ww9y7ouo.fsf@flinx.npwt.net \
--to=ebiederm+eric@npwt.net \
--cc=jasonc@cacr.caltech.edu \
--cc=linux-mm@kvack.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox