From: Wolfgang Denk <wd@denx.de>
To: Marc Singer <elf@buici.com>
Cc: Joakim Tjernlund <Joakim.Tjernlund@lumentis.se>,
linux-mtd@lists.infradead.org
Subject: Re: crc32() optimization
Date: Mon, 11 Nov 2002 02:37:33 +0100 [thread overview]
Message-ID: <20021111013738.4FEB410162@denx.denx.de> (raw)
In-Reply-To: Your message of "Sun, 10 Nov 2002 17:31:14 PST." <20021111013114.GB27214@buici.com>
In message <20021111013114.GB27214@buici.com> you wrote:
>
> > > What's "Duff's Device"?
> >
> > It's a tricky way to implement general loop unrolling directly in C.
> > Applied to your problem, code that looks like this (instead of 8 any
> > other loop count may be used, but you need to adjust the "case"
> > statements then):
> >
> > register int n = (len + (8-1)) / 8;
> >
> > switch (len % 8) {
> > case 0: do { val = crc32_table ... ;
> > case 7: val = crc32_table ... ;
> > case 6: val = crc32_table ... ;
> > case 5: val = crc32_table ... ;
> > case 4: val = crc32_table ... ;
> > case 3: val = crc32_table ... ;
> > case 2: val = crc32_table ... ;
> > case 1: val = crc32_table ... ;
> > } while (--n > 0);
> > }
>
> This doesn't look right to me. You are decrementing n but using the
> modulus of len in the switch. The len modulus is correct when n == 1,
> but not when n > 1. The idea makes sense, but the implementation
> appears to be missing a detail.
You don't understand. The switch is only needed for the first,
partial loop where we want less than N statements; then we're nunning
the remaining fully unrolled loos in the do{}while loop.
> As for performance problems, I believe that the trouble is evident
> from the assembler output. The reason that the unrolled loop is more
> efficient than the simple loop is mainly because you don't jump as
> often. We all know that jumps tend to perturb the instruction fetch
> queue and cache.
Did you enable optimization?
> Here is an example. I know it is wrong, but it shows how the compiler
> implements the switch. I've marked the problem jump. It is executed
Irrelevant here.
> So, the version I posted may not be significantly better that the
> original one that unrolled 6 times. It is just that unrolling to a
You save the extra while{} loop.
Best regards,
Wolfgang Denk
--
Software Engineering: Embedded and Realtime Systems, Embedded Linux
Phone: (+49)-8142-4596-87 Fax: (+49)-8142-4596-88 Email: wd@denx.de
See us @ electronica 2002 in Munich, Nov 12-15, Hall A3, Booth A3.325
next prev parent reply other threads:[~2002-11-11 1:07 UTC|newest]
Thread overview: 17+ messages / expand[flat|nested] mbox.gz Atom feed top
[not found] <F122OgRGkQ6mBySsxVY00000854@hotmail.com>
[not found] ` <24987.1036797874@passion.cambridge.redhat.com>
2002-11-10 15:28 ` crc32() optimization Joakim Tjernlund
2002-11-10 18:43 ` Marc Singer
2002-11-10 19:25 ` Wolfgang Denk
2002-11-10 20:05 ` Joakim Tjernlund
2002-11-10 21:00 ` Wolfgang Denk
2002-11-10 21:22 ` Joakim Tjernlund
2002-11-10 22:35 ` Joakim Tjernlund
2002-11-10 22:41 ` Wolfgang Denk
2002-11-10 23:00 ` Joakim Tjernlund
2002-11-10 23:56 ` Eric W. Biederman
2002-11-11 1:31 ` Marc Singer
2002-11-11 1:37 ` Wolfgang Denk [this message]
2002-11-11 4:42 ` Marc Singer
2002-11-25 15:55 ` Herman Oosthuysen
2002-11-25 16:12 ` Joakim Tjernlund
2002-11-11 0:50 ` Marc Singer
2002-11-10 20:04 ` Joakim Tjernlund
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=20021111013738.4FEB410162@denx.denx.de \
--to=wd@denx.de \
--cc=Joakim.Tjernlund@lumentis.se \
--cc=elf@buici.com \
--cc=linux-mtd@lists.infradead.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