From: Michael Neuling <mikey@neuling.org>
To: Peter Zijlstra <peterz@infradead.org>,
Matt Fleming <matt@codeblueprint.co.uk>
Cc: mingo@kernel.org, linux-kernel@vger.kernel.org, clm@fb.com,
mgalbraith@suse.de, tglx@linutronix.de, fweisbec@gmail.com,
srikar@linux.vnet.ibm.com, anton@samba.org,
oliver <oohall@gmail.com>,
"Shreyas B. Prabhu" <shreyas@linux.vnet.ibm.com>
Subject: Re: [RFC][PATCH 4/7] sched: Replace sd_busy/nr_busy_cpus with sched_domain_shared
Date: Thu, 12 May 2016 12:05:37 +1000 [thread overview]
Message-ID: <1463018737.28449.38.camel@neuling.org> (raw)
In-Reply-To: <20160511182402.GD3205@twins.programming.kicks-ass.net>
On Wed, 2016-05-11 at 20:24 +0200, Peter Zijlstra wrote:
> On Wed, May 11, 2016 at 02:33:45PM +0200, Peter Zijlstra wrote:
> >
> > Hmm, PPC folks; what does your topology look like?
> >
> > Currently your sched_domain_topology, as per arch/powerpc/kernel/smp.c
> > seems to suggest your cores do not share cache at all.
> >
> > https://en.wikipedia.org/wiki/POWER7 seems to agree and states
> >
> > "4 MB L3 cache per C1 core"
> >
> > And http://www-03.ibm.com/systems/resources/systems_power_software_i_pe
> > rfmgmt_underthehood.pdf
> > also explicitly draws pictures with the L3 per core.
> >
> > _however_, that same document describes L3 inter-core fill and lateral
> > cast-out, which sounds like the L3s work together to form a node wide
> > caching system.
> >
> > Do we want to model this co-operative L3 slices thing as a sort of
> > node-wide LLC for the purpose of the scheduler ?
> Going back a generation; Power6 seems to have a shared L3 (off package)
> between the two cores on the package. The current topology does not
> reflect that at all.
>
> And going forward a generation; Power8 seems to share the per-core
> (chiplet) L3 amonst all cores (chiplets) + is has the centaur (memory
> controller) 16M L4.
Yep, L1/L2/L3 is per core on POWER8 and POWER7. POWER6 and POWER5 (both
dual core chips) had a shared off chip cache
The POWER8 L4 is really a bit different as it's out in the memory
controller. It's more of a memory DIMM buffer as it can only cache data
associated with the physical addresses on those DIMMS.
> So it seems the current topology setup is not describing these chips
> very well. Also note that the arch topology code can runtime select a
> topology, so you could make that topo setup micro-arch specific.
We are planning on making some topology changes for the upcoming P9 which
will share L2/L3 amongst pairs of cores (24 cores per chip).
FWIW our P9 upstreaming is still in it's infancy since P9 is not released
yet.
Mike
next prev parent reply other threads:[~2016-05-12 2:05 UTC|newest]
Thread overview: 39+ messages / expand[flat|nested] mbox.gz Atom feed top
2016-05-09 10:48 [RFC][PATCH 0/7] sched: select_idle_siblings rewrite Peter Zijlstra
2016-05-09 10:48 ` [RFC][PATCH 1/7] sched: Remove unused @cpu argument from destroy_sched_domain*() Peter Zijlstra
2016-05-09 10:48 ` [RFC][PATCH 2/7] sched: Restructure destroy_sched_domain() Peter Zijlstra
2016-05-09 14:46 ` Peter Zijlstra
2016-05-09 10:48 ` [RFC][PATCH 3/7] sched: Introduce struct sched_domain_shared Peter Zijlstra
2016-05-09 10:48 ` [RFC][PATCH 4/7] sched: Replace sd_busy/nr_busy_cpus with sched_domain_shared Peter Zijlstra
2016-05-11 11:55 ` Matt Fleming
2016-05-11 12:33 ` Peter Zijlstra
2016-05-11 18:11 ` Peter Zijlstra
2016-05-11 18:24 ` Peter Zijlstra
2016-05-12 2:05 ` Michael Neuling [this message]
2016-05-12 5:07 ` Peter Zijlstra
2016-05-12 11:07 ` Michael Neuling
2016-05-12 11:33 ` Peter Zijlstra
2016-05-13 0:12 ` Michael Neuling
2016-05-16 14:00 ` Peter Zijlstra
2016-05-17 10:20 ` Peter Zijlstra
2016-05-17 10:52 ` Srikar Dronamraju
2016-05-17 11:15 ` Peter Zijlstra
2016-05-11 17:37 ` Peter Zijlstra
2016-05-11 18:04 ` Matt Fleming
2016-05-16 15:31 ` Dietmar Eggemann
2016-05-16 17:02 ` Peter Zijlstra
2016-05-16 17:26 ` Dietmar Eggemann
2016-05-09 10:48 ` [RFC][PATCH 5/7] sched: Rewrite select_idle_siblings() Peter Zijlstra
2016-05-10 21:05 ` Yuyang Du
2016-05-11 7:00 ` Peter Zijlstra
2016-05-10 23:42 ` Yuyang Du
2016-05-11 7:43 ` Mike Galbraith
2016-05-09 10:48 ` [RFC][PATCH 6/7] sched: Optimize SCHED_SMT Peter Zijlstra
2016-05-09 10:48 ` [RFC][PATCH 7/7] sched: debug muck -- not for merging Peter Zijlstra
2016-05-10 0:50 ` [RFC][PATCH 0/7] sched: select_idle_siblings rewrite Chris Mason
2016-05-11 14:19 ` Chris Mason
2016-05-18 5:51 ` [RFC][PATCH 8/7] sched/fair: Use utilization distance to filter affine sync wakeups Mike Galbraith
2016-05-19 21:43 ` Rik van Riel
2016-05-20 2:52 ` Mike Galbraith
2016-05-25 14:51 ` [RFC][PATCH 0/7] sched: select_idle_siblings rewrite Chris Mason
2016-05-25 16:24 ` Peter Zijlstra
2016-05-25 17:11 ` Chris Mason
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=1463018737.28449.38.camel@neuling.org \
--to=mikey@neuling.org \
--cc=anton@samba.org \
--cc=clm@fb.com \
--cc=fweisbec@gmail.com \
--cc=linux-kernel@vger.kernel.org \
--cc=matt@codeblueprint.co.uk \
--cc=mgalbraith@suse.de \
--cc=mingo@kernel.org \
--cc=oohall@gmail.com \
--cc=peterz@infradead.org \
--cc=shreyas@linux.vnet.ibm.com \
--cc=srikar@linux.vnet.ibm.com \
--cc=tglx@linutronix.de \
/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
Powered by JetHome