All of lore.kernel.org
 help / color / mirror / Atom feed
From: Maneesh Soni <maneesh@in.ibm.com>
To: "Martin J. Bligh" <mbligh@aracnet.com>
Cc: Jesse Barnes <jbarnes@engr.sgi.com>,
	hawkes@sgi.com, linux-kernel@vger.kernel.org, wli@holomorphy.com
Subject: Re: kernbench on 512p
Date: Fri, 20 Aug 2004 14:36:35 -0500	[thread overview]
Message-ID: <20040820193635.GA4161@in.ibm.com> (raw)
In-Reply-To: <270470000.1092952599@flay>

On Thu, Aug 19, 2004 at 02:56:39PM -0700, Martin J. Bligh wrote:
> --On Thursday, August 19, 2004 17:24:04 -0400 Jesse Barnes <jbarnes@engr.sgi.com> wrote:
> 
> > On Thursday, August 19, 2004 5:11 pm, Jesse Barnes wrote:
> >> The output is attached (my mailer insists on wrapping it if I inline it). 
> >> I used 'lockstat -w'.
> > 
> > The highlights:
> > 
> >  nw   spin   rjct  lock & function
> > 19.0% 81.0%    0%  dcache_lock
> >  3.3% 96.7%    0%    d_alloc+0x270
> >  2.7% 97.3%    0%    d_delete+0x40
> > 18.3% 81.7%    0%    d_instantiate+0x90
> >  4.7% 95.3%    0%    d_move+0x60
> > 34.6% 65.4%    0%    d_rehash+0xe0
> > 19.1% 80.9%    0%    dput+0x40
> > 10.5% 89.5%    0%    link_path_walk+0xef0
> >    0%  100%    0%    sys_getcwd+0x210
> > 
> > 41.4% 58.6%    0%  rcu_state
> > 61.3% 38.7%    0%    __rcu_process_callbacks+0x260
> > 41.4% 58.6%    0%    rcu_check_quiescent_state+0xf0
> > 
> > So it looks like the dcache lock is the biggest problem on this system with 
> > this load.  And although the rcu stuff has improved tremendously for this 
> > system, it's still highly contended.
> 
> Hmmm. dcache_lock is known-fucked, though I'm suprised at d_rehash
> (file deletion)?
> 

The thing which is hurting most is dput(), almost 90% of the dcache_lock 
aquistion is there.

(deleted few columns)

 23.8% 67.3%     9955785 32.7% 67.3%    0%  dcache_lock
 0.06% 70.5%       17068 29.5% 70.5%    0%    d_alloc+0x270
 0.02% 25.4%       15340 74.6% 25.4%    0%    d_delete+0x40
 0.06% 65.9%       30485 34.1% 65.9%    0%    d_instantiate+0x90
 0.06% 78.0%        4461 22.0% 78.0%    0%    d_move+0x60
 0.00%    0%           2  100%    0%    0%    d_path+0x120
 0.04% 52.5%       17068 47.5% 52.5%    0%    d_rehash+0xe0
 0.00% 26.7%          15 73.3% 26.7%    0%    d_splice_alias+0xc0
 0.00% 20.0%           5 80.0% 20.0%    0%    dentry_unhash+0x70
 0.00%    0%           4  100%    0%    0%    dentry_unhash+0xc0
 23.5% 67.5%     9827270 32.5% 67.5%    0%    dput+0x40
 0.09% 62.3%       36875 37.7% 62.3%    0%    link_path_walk+0xef0
 0.00% 0.24%        5068 99.8% 0.24%    0%    link_path_walk+0x1cc0
 0.00% 33.3%           3 66.7% 33.3%    0%    proc_pid_unhash+0x50
 0.00%  7.1%          14 92.9%  7.1%    0%    prune_dcache+0x50
 0.00%    0%         167  100%    0%    0%    prune_dcache+0x3d0
 0.00%    0%          21  100%    0%    0%    select_parent+0x40
 0.00% 24.2%        1919 75.8% 24.2%    0%    sys_getcwd+0x210



-- 
Maneesh Soni
Linux Technology Center, 
IBM Austin
email: maneesh@in.ibm.com
Phone: 1-512-838-1896 Fax: 
T/L : 6781896

  reply	other threads:[~2004-08-20 19:38 UTC|newest]

Thread overview: 35+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2004-08-19 16:16 kernbench on 512p Jesse Barnes
2004-08-19 16:22 ` William Lee Irwin III
2004-08-19 16:29 ` David Mosberger
2004-08-19 16:37   ` Jesse Barnes
2004-08-19 16:40     ` David Mosberger
2004-08-19 17:55       ` Alan Cox
2004-08-20  7:53         ` David Mosberger
2004-08-19 18:25     ` Martin J. Bligh
2004-08-19 18:42       ` Jesse Barnes
2004-08-19 21:11       ` Jesse Barnes
2004-08-19 21:24         ` Jesse Barnes
2004-08-19 21:56           ` Martin J. Bligh
2004-08-20 19:36             ` Maneesh Soni [this message]
2004-08-19 23:38           ` Paul E. McKenney
2004-08-20  0:16             ` Jesse Barnes
2004-08-20 15:57               ` Paul E. McKenney
2004-08-20 17:24                 ` Jesse Barnes
2004-08-20 19:22                   ` Manfred Spraul
2004-08-20 20:19                   ` Manfred Spraul
2004-08-23 21:23                     ` Jesse Barnes
2004-08-23 23:13                       ` Jesse Barnes
2004-09-10 19:01                     ` Greg Edwards
2004-09-13 18:21                       ` Manfred Spraul
2004-09-14 17:52                         ` Greg Edwards
2004-09-14 18:16                           ` Manfred Spraul
2004-09-14 18:43                             ` Greg Edwards
2004-09-14 19:16                               ` Manfred Spraul
2004-08-19 21:50         ` Ray Bryant
2004-08-19 22:00           ` Jesse Barnes
2004-08-19 23:03 ` William Lee Irwin III
2004-08-20 17:35   ` William Lee Irwin III
2004-08-20  0:56 ` remove dentry_open::file_ra_init_state() duplicated memset was " Marcelo Tosatti
2004-08-20  6:21   ` Andrew Morton
2004-08-20  7:28     ` Marcelo Tosatti
2004-08-20  8:34       ` Andrew Morton

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=20040820193635.GA4161@in.ibm.com \
    --to=maneesh@in.ibm.com \
    --cc=hawkes@sgi.com \
    --cc=jbarnes@engr.sgi.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mbligh@aracnet.com \
    --cc=wli@holomorphy.com \
    /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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.