mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: Badari Pulavarty <pbadari@us.ibm.com>
To: akpm@zip.com.au (Andrew Morton)
Cc: pbadari@us.ibm.com (Badari Pulavarty),
	linux-kernel@vger.kernel.org, gerrit@us.ibm.com (Gerrit Huizenga),
	hjt@us.ibm.com (Hans-J Tannenberger),
	janetmor@us.ibm.com (Janet Morgan),
	andmike@us.ibm.com (Mike Anderson),
	mjbligh@us.ibm.com (Martin Bligh)
Subject: Re: 2.5.32 IO performance issues
Date: Thu, 29 Aug 2002 13:55:40 -0700 (PDT)	[thread overview]
Message-ID: <200208292055.g7TKte224951@eng2.beaverton.ibm.com> (raw)
In-Reply-To: <3D6E6B64.66203783@zip.com.au> from "Andrew Morton" at Aug 29, 2002 10:43:48 AM PST

> 
> Badari Pulavarty wrote:
> > 
> > Hi,
> > 
> > I am having severe IO performance problems with 2.5.32 (2.5.31 works fine).
> > I was wondering what caused this.
> > 
> > As you can see, IO rate went from
> > 
> >                 384MB/sec with 6% CPU utilization on 2.5.31
> >                         to
> >                 120MB/sec with 19% CPU utilization on 2.5.32
> > 
> > ...
> > 151712 default_idle                             2370.5000
> > 21622 __scsi_end_request                       122.8523
> 
> block-highmem is bust for scsi. (aic7xxx at least).  Does
> http://www.zip.com.au/~akpm/linux/patches/2.5/2.5.32/2.5.32-mm2/broken-out/scsi_hack.patch
> fix it?

Hmm !! This patch fixed it. I remember you gave me this patch for 2.5.31. But 2.5.31 
was doing fine without it. But 2.5.32 seem to need it.

   procs                      memory    swap          io     system         cpu
 r  b  w   swpd   free   buff  cache  si  so    bi    bo   in    cs  us  sy  id
 1 38  0      0 3811160      0  23532   0   0 384461     0 3981  6346   0   5  95
 0 39  0      0 3811152      0  23532   0   0 384499     0 3967  6354   0   5  95
 0 39  0      0 3811148      0  23532   0   0 384154     0 3974  6341   0   5  95
 0 39  0      0 3811144      0  23532   0   0 384013     0 3964  6344   0   5  95
 0 39  0      0 3811204      0  23532   0   0 384154     0 3966  6357   0   5  95
 0 39  0      0 3811176      0  23532   0   0 384256     0 3973  6341   0   5  95
 0 39  0      0 3811168      0  23532   0   0 384320     0 3969  6329   0   5  95
 0 39  0      0 3811152      0  23532   0   0 384294     0 3980  6335   0   5  95

Here is the profile output with this patch:

# readprofile  | sort -nr +2 
221145 default_idle                             3455.3906
  3163 do_softirq                                16.4740
  1489 __scsi_end_request                         8.4602
  1696 scsi_dispatch_cmd                          2.7895
   371 dio_bio_complete                           2.1080
   509 blk_rq_map_sg                              1.3255
   198 dio_await_one                              1.2375
    57 dio_prep_bio                               1.1875
    56 bio_put                                    1.1667
    71 __generic_copy_to_user                     1.1094
   630 do_direct_IO                               0.9157
    42 __wake_up                                  0.8750
   297 __set_page_dirty_buffers                   0.8438
   622 schedule                                   0.6942
   629 __make_request                             0.5460
   402 get_user_pages                             0.5346
    17 scsi_sense_valid                           0.5312
    42 bio_destructor                             0.5250
   198 bio_alloc                                  0.4760
    32 dio_get_page                               0.4000


  reply	other threads:[~2002-08-29 20:51 UTC|newest]

Thread overview: 8+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2002-08-29 18:20 Badari Pulavarty
2002-08-29 18:39 ` Rik van Riel
2002-08-29 19:02   ` Dave Hansen
2002-08-29 18:43 ` Andrew Morton
2002-08-29 20:55   ` Badari Pulavarty [this message]
2002-08-29 21:53     ` Patrick Mansfield
2002-08-29 22:23       ` Andrew Morton
2002-08-31  9:14         ` Jens Axboe

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=200208292055.g7TKte224951@eng2.beaverton.ibm.com \
    --to=pbadari@us.ibm.com \
    --cc=akpm@zip.com.au \
    --cc=andmike@us.ibm.com \
    --cc=gerrit@us.ibm.com \
    --cc=hjt@us.ibm.com \
    --cc=janetmor@us.ibm.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mjbligh@us.ibm.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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox

all inboxes | Powered by JetHome®