Subject: Re: copyin/out
To: Allen Briggs <briggs@wasabisystems.com>
From: Ben Harris <bjh21@netbsd.org>
List: port-arm
Date: 08/11/2002 13:37:52
  This message is in MIME format.  The first part should be readable text,
  while the remaining parts are likely unreadable without MIME-aware tools.
  Send mail to mime@docserver.cac.washington.edu for more info.

---559023410-851401618-1029069472=:26268
Content-Type: TEXT/PLAIN; charset=US-ASCII

More experimentation.  I've replaced runs of STR and LDR with STM and
LDM, tweaking the register usage in copyin to make the runs longer.  The
diff's attached.

Same test as ever: dd if=/dev/zero of=/dev/null count=1024 bs=1024k

ARM610, 30 MHz, 4KB unified cache:
brini:  1073741824 bytes transferred in 206.649 secs (5195969 bytes/sec)
briggs: 1073741824 bytes transferred in 92.540 secs (11603002 bytes/sec)
bjh21:  1073741824 bytes transferred in 75.677 secs (14188482 bytes/sec)

ARM710a, 40 MHz, 8KB unified cache:
brini:  1073741824 bytes transferred in 148.135 secs (7248400 bytes/sec)
briggs: 1073741824 bytes transferred in 48.377 secs (22195295 bytes/sec)
bjh21:  1073741824 bytes transferred in 48.248 secs (22254639 bytes/sec)

SA-110, 233 MHz, 16KB I-cache, 16KB D-cache:
brini:  1073741824 bytes transferred in 50.537 secs (21246647 bytes/sec)
briggs: 1073741824 bytes transferred in 43.931 secs (24441552 bytes/sec)
bjh21:  1073741824 bytes transferred in 43.889 secs (24464941 bytes/sec)

So significantly better performance on the ARM610, and almost no change on
ARM710a and SA-110.

-- 
Ben Harris                                                   <bjh21@netbsd.org>
Portmaster, NetBSD/acorn26           <URL:http://www.netbsd.org/Ports/acorn26/>


---559023410-851401618-1029069472=:26268
Content-Type: TEXT/PLAIN; charset=US-ASCII; name="bcopyinout.patch"
Content-Transfer-Encoding: BASE64
Content-ID: <Pine.SOL.4.44.0208111337520.26268@virgo.cus.cam.ac.uk>
Content-Description: 
Content-Disposition: attachment; filename="bcopyinout.patch"
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---559023410-851401618-1029069472=:26268--