Linux NFS development
 help / color / mirror / Atom feed
From: Chuck Lever <chuck.lever@oracle.com>
To: "J. Bruce Fields" <bfields@fieldses.org>
Cc: nfs@lists.sourceforge.net, trond.myklebust@fys.uio.no
Subject: Re: [PATCH 13/15] SUNRPC: RPC buffer size estimates are too large
Date: Wed, 24 Jan 2007 15:50:37 -0500	[thread overview]
Message-ID: <45B7C69D.1090709@oracle.com> (raw)
In-Reply-To: <20070124204724.GB6587@fieldses.org>

[-- Attachment #1: Type: text/plain, Size: 1545 bytes --]

J. Bruce Fields wrote:
> On Wed, Jan 24, 2007 at 02:20:20PM -0500, Chuck Lever wrote:
>> The RPC buffer size estimation logic in net/sunrpc/clnt.c always
>> significantly overestimates the requirements for the buffer size.
>> A little instrumentation demonstrated that in fact rpc_malloc was never
>> allocating the buffer from the mempool, but almost always called kmalloc.
>>
>> To compute the size of the RPC buffer more precisely, split p_bufsiz into
>> two fields; one for the argument size, and one for the result size.
>>
>> So now we will compute the sum of the exact call and reply header sizes,
>> and split the RPC buffer precisely between the two.  That should keep
>> almost all RPC buffers within the 2KiB buffer mempool limit.
> 
> One other worry: after this patch, as I understand it, a small mistake
> in any of the definitions of the *_sz macros could allow a buffer
> overflow that RPC_SLACK_SPACE previously prevented.
> 
> Those definitions are all handwritten, and I'd be suprised if there
> aren't at least a few minor errors.
> 
> I'd certainly rather have these calculations be exact than depend on
> this big RPC_SLACK_SPACE fudge factor, but is there some way we could
> verify those macros first?

I've been running Connectathon and fsx-odirect with hundreds of 
thousands of ops on a kernel that has all the slab debugging options 
enabled.  I haven't hit a problem yet.

And besides, the RPC buffers are all a fixed size (2kiB), and much 
larger than required already.  The slack is "built in" to the buffer size.

[-- Attachment #2: chuck.lever.vcf --]
[-- Type: text/x-vcard, Size: 265 bytes --]

begin:vcard
fn:Chuck Lever
n:Lever;Chuck
org:Oracle Corporation;Corporate Architecture Linux Projects Group
email;internet:chuck dot lever at nospam oracle dot com
title:Principle Member of Staff
tel;work:+1 248 614 5091
x-mozilla-html:FALSE
version:2.1
end:vcard


[-- Attachment #3: Type: text/plain, Size: 347 bytes --]

-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys - and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV

[-- Attachment #4: Type: text/plain, Size: 140 bytes --]

_______________________________________________
NFS maillist  -  NFS@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/nfs

  reply	other threads:[~2007-01-24 20:51 UTC|newest]

Thread overview: 26+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2007-01-24 19:17 [PATCH 00/15] patches for 2.6.21 Chuck Lever
2007-01-24 19:19 ` [PATCH 01/15] NFS: fix print format for tk_pid Chuck Lever
2007-01-24 19:19 ` [PATCH 02/15] SUNRPC: fix print format for tk_pid in auth_gss support Chuck Lever
2007-01-24 19:19 ` [PATCH 03/15] SUNRPC: fix print format for tk_pid Chuck Lever
2007-01-24 19:19 ` [PATCH 04/15] SUNRPC: Eliminate side effects from rpc_malloc Chuck Lever
2007-01-24 19:19 ` [PATCH 05/15] SUNRPC: Make rpc_free API more generic Chuck Lever
2007-01-24 19:19 ` [PATCH 06/15] NFS: disconnect before retrying NFSv4 requests over TCP Chuck Lever
2007-01-24 19:20 ` [PATCH 07/15] SUNRPC: introduce rpcbind: replacement for in-kernel portmapper Chuck Lever
2007-01-24 19:20 ` [PATCH 08/15] SUNRPC: switch socket-based RPC transports to use rpcbind Chuck Lever
2007-01-24 19:20 ` [PATCH 09/15] SUNRPC: switch the RPC server to use the new rpcbind registration API Chuck Lever
2007-01-24 19:20 ` [PATCH 10/15] NFS: switch NFSROOT to use new rpcbind client Chuck Lever
2007-01-24 19:20 ` [PATCH 11/15] SUNRPC: remove old portmapper Chuck Lever
2007-01-24 19:20 ` [PATCH 12/15] SUNRPC: RPC client should retry with different versions of rpcbind Chuck Lever
2007-01-24 19:20 ` [PATCH 13/15] SUNRPC: RPC buffer size estimates are too large Chuck Lever
2007-01-24 20:37   ` J. Bruce Fields
2007-01-24 20:41     ` Chuck Lever
2007-01-24 20:51       ` J. Bruce Fields
2007-01-24 21:09         ` Chuck Lever
2007-01-24 21:25           ` J. Bruce Fields
2007-01-24 21:28             ` Chuck Lever
2007-01-24 20:47   ` J. Bruce Fields
2007-01-24 20:50     ` Chuck Lever [this message]
2007-01-24 21:01       ` J. Bruce Fields
2007-01-24 21:11         ` Chuck Lever
2007-01-24 19:20 ` [PATCH 14/15] NLM: Shrink the maximum request size of NLM4 requests Chuck Lever
2007-01-24 19:20 ` [PATCH 15/15] SUNRPC: Debugging aid Chuck Lever

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=45B7C69D.1090709@oracle.com \
    --to=chuck.lever@oracle.com \
    --cc=bfields@fieldses.org \
    --cc=nfs@lists.sourceforge.net \
    --cc=trond.myklebust@fys.uio.no \
    /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