From: Segher Boessenkool <segher@kernel.crashing.org>
To: Matthew Wilcox <willy@infradead.org>
Cc: Christophe LEROY <christophe.leroy@c-s.fr>,
paulus@samba.org, linuxppc-dev@lists.ozlabs.org
Subject: Re: Optimised memset64/memset32 for powerpc
Date: Tue, 21 Mar 2017 11:45:29 -0500 [thread overview]
Message-ID: <20170321164527.GJ4402@gate.crashing.org> (raw)
In-Reply-To: <20170321132910.GA4482@bombadil.infradead.org>
On Tue, Mar 21, 2017 at 06:29:10AM -0700, Matthew Wilcox wrote:
> > Unrolling the loop could help a bit on old powerpc32s that don't have branch
> > units, but on those processors the main driver is the time spent to do the
> > effective write to memory, and the operations necessary to unroll the loop
> > are not worth the cycle added by the branch.
> >
> > On more modern powerpc32s, the branch unit implies that branches have a zero
> > cost.
>
> Fair enough. I'm just surprised it was worth unrolling the loop on
> powerpc64 and not on powerpc32 -- see mem_64.S.
We can do at most one loop iteration per cycle, but we can do multiple
stores per cycle, on modern, bigger CPUs. Many old or small CPUs have
only one load/store unit on the other hand. There are other issues,
but that is the biggest difference.
Segher
next prev parent reply other threads:[~2017-03-21 17:44 UTC|newest]
Thread overview: 20+ messages / expand[flat|nested] mbox.gz Atom feed top
2017-03-20 21:14 Optimised memset64/memset32 for powerpc Matthew Wilcox
2017-03-20 21:23 ` Benjamin Herrenschmidt
2017-03-21 12:23 ` Christophe LEROY
2017-03-21 13:29 ` Matthew Wilcox
2017-03-21 16:45 ` Segher Boessenkool [this message]
2017-03-21 21:26 ` Benjamin Herrenschmidt
2017-03-22 13:18 ` Matthew Wilcox
2017-03-22 19:30 ` Matthew Wilcox
2017-03-27 19:37 ` Naveen N. Rao
2017-03-27 19:37 ` [PATCH 1/2] powerpc: string: implement optimized memset variants Naveen N. Rao
2017-03-28 0:44 ` Michael Ellerman
2017-03-28 10:21 ` Naveen N. Rao
2017-03-29 11:36 ` Michael Ellerman
2017-03-30 7:16 ` Naveen N. Rao
2017-04-04 12:00 ` Michael Ellerman
2017-04-18 6:45 ` Michael Ellerman
2017-04-05 5:51 ` PrasannaKumar Muralidharan
2017-04-12 15:05 ` Naveen N. Rao
2017-08-18 12:50 ` [1/2] " Michael Ellerman
2017-03-27 19:37 ` [PATCH 2/2] powerpc: bpf: use memset32() to pre-fill traps in BPF page(s) Naveen N. Rao
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=20170321164527.GJ4402@gate.crashing.org \
--to=segher@kernel.crashing.org \
--cc=christophe.leroy@c-s.fr \
--cc=linuxppc-dev@lists.ozlabs.org \
--cc=paulus@samba.org \
--cc=willy@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;
as well as URLs for NNTP newsgroup(s).