[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <Z9raotkELnHnK19a@gourry-fedora-PF4VCD3F>
Date: Wed, 19 Mar 2025 10:54:26 -0400
From: Gregory Price <gourry@...rry.net>
To: Yunjeong Mun <yunjeong.mun@...com>
Cc: kernel_team@...ynix.com, Joshua Hahn <joshua.hahnjy@...il.com>,
harry.yoo@...cle.com, ying.huang@...ux.alibaba.com,
gregkh@...uxfoundation.org, rakie.kim@...com,
akpm@...ux-foundation.org, rafael@...nel.org, lenb@...nel.org,
dan.j.williams@...el.com, Jonathan.Cameron@...wei.com,
dave.jiang@...el.com, horen.chuang@...ux.dev, hannes@...xchg.org,
linux-kernel@...r.kernel.org, linux-acpi@...r.kernel.org,
linux-mm@...ck.org, kernel-team@...a.com,
Honggyu Kim <honggyu.kim@...com>
Subject: Re: [PATCH 2/2 v6] mm/mempolicy: Don't create weight sysfs for
memoryless nodes
On Wed, Mar 19, 2025 at 06:56:21PM +0900, Yunjeong Mun wrote:
> > These are the root decoders that should map up directly with each CEDT
> > CFMWS entry.
> >
> > 2 of them should have interleave settings.
> >
> > If you were to then program the endpoint and hostbridge decoders with
> > the matching non-interleave address values from the other CEDT entries,
> > you could bring each individual device online in its own NUMA node.
> >
>
> I think this means that I can program the endpoint(=cxl_decoder_endpoint)
> to map to the 8 CFMWS, and the hostbridge decoder (=cxl_decoder switch) to map
> to another 2 CFMWS(cross-host bridge).
>
you can program any given endpoint decoders to map to either per-device
CFMWS, or to the interleave CFMWS - decoders are flexible in this way.
The existince of both sets gives you the option of how to program the
topology - it's up to you.
~Gregory
Powered by blists - more mailing lists