From: Yazen Ghannam <yazen.ghannam@amd.com>
To: Naveen Krishna Chatradhi <nchatrad@amd.com>
Cc: linux-edac@vger.kernel.org, x86@kernel.org,
linux-kernel@vger.kernel.org, bp@alien8.de, mingo@redhat.com,
mchehab@kernel.org, Muralidhara M K <muralimk@amd.com>
Subject: Re: [PATCH v7 01/12] EDAC/amd64: Document heterogeneous enumeration
Date: Wed, 9 Feb 2022 22:34:14 +0000 [thread overview]
Message-ID: <YgRBZjUN3AoHr7q+@yaz-ubuntu> (raw)
In-Reply-To: <20220203174942.31630-2-nchatrad@amd.com>
On Thu, Feb 03, 2022 at 11:49:31AM -0600, Naveen Krishna Chatradhi wrote:
> From: Muralidhara M K <muralimk@amd.com>
>
> The Documentation notes have been added in amd64_edac.h and will be
> referring to driver-api wherever needed.
I don't see the comment in amd64_edac.h referring to driver-api/edac.rst. So
I'm not sure what this sentence is saying.
>
> Explains how the physical topology is enumerated in the software and
> edac module populates the sysfs ABIs.
>
Also, please make sure the message is imperative, e.g "Add...", "Explain...",
etc.
> Signed-off-by: Muralidhara M K <muralimk@amd.com>
> Signed-off-by: Naveen Krishna Chatradhi <nchatrad@amd.com>
> ---
> v6->v7:
> * New in v7
>
> Documentation/driver-api/edac.rst | 9 +++
> drivers/edac/amd64_edac.h | 101 ++++++++++++++++++++++++++++++
> 2 files changed, 110 insertions(+)
>
> diff --git a/Documentation/driver-api/edac.rst b/Documentation/driver-api/edac.rst
> index b8c742aa0a71..0dd07d0d0e47 100644
> --- a/Documentation/driver-api/edac.rst
> +++ b/Documentation/driver-api/edac.rst
> @@ -106,6 +106,15 @@ will occupy those chip-select rows.
> This term is avoided because it is unclear when needing to distinguish
> between chip-select rows and socket sets.
>
> +* High Bandwidth Memory (HBM)
> +
> +HBM is a new type of memory chip with low power consumption and ultra-wide
> +communication lanes. It uses vertically stacked memory chips (DRAM dies)
> +interconnected by microscopic wires called "through-silicon vias," or TSVs.
> +
> +Several stacks of HBM chips connect to the CPU or GPU through an ultra-fast
> +interconnect called the “interposer". So that HBM’s characteristics are
> +nearly indistinguishable from on-chip integrated RAM.
>
I think this makes sense.
> Memory Controllers
> ------------------
> diff --git a/drivers/edac/amd64_edac.h b/drivers/edac/amd64_edac.h
> index 6f8147abfa71..6a112270a84b 100644
> --- a/drivers/edac/amd64_edac.h
> +++ b/drivers/edac/amd64_edac.h
> @@ -559,3 +559,104 @@ static inline u32 dct_sel_baseaddr(struct amd64_pvt *pvt)
> }
> return (pvt)->dct_sel_lo & 0xFFFFF800;
> }
> +
> +/*
> + * AMD Heterogeneous system support on EDAC subsystem
> + * --------------------------------------------------
> + *
> + * An AMD heterogeneous system built by connecting the data fabrics of both CPUs
> + * and GPUs via custom xGMI links. So, the Data Fabric on the GPU nodes can be
> + * accessed the same way as the Data Fabric on CPU nodes.
> + *
> + * An Aldebaran GPUs has 2 Data Fabrics, each GPU DF contains four Unified
> + * Memory Controllers (UMC). Each UMC contains eight Channels. Each UMC Channel
> + * controls one 128-bit HBM2e (2GB) channel (equivalent to 8 X 2GB ranks),
> + * this creates a total of 4096-bits of DRAM data bus.
> + *
> + * While UMC is interfacing a 16GB (8H X 2GB DRAM) HBM stack, each UMC channel is
What is "8H"? Is that 8 "high"?
> + * interfacing 2GB of DRAM (represented as rank).
> + *
> + * Memory controllers on AMD GPU nodes can be represented in EDAC is as below:
> + * GPU DF / GPU Node -> EDAC MC
> + * GPU UMC -> EDAC CSROW
> + * GPU UMC channel -> EDAC CHANNEL
> + *
> + * Eg: An heterogeneous system with 1 AMD CPU is connected to 4 Aldebaran GPUs using xGMI.
> + *
> + * AMD GPU Nodes are enumerated in sequential order based on the PCI hierarchy, and the
> + * first GPU node is assumed to have an "Node ID" value after CPU Nodes are fully
> + * populated.
> + *
> + * $ ls /sys/devices/system/edac/mc/
> + * mc0 - CPU MC node 0
> + * mc1 |
> + * mc2 |- GPU card[0] => node 0(mc1), node 1(mc2)
> + * mc3 |
> + * mc4 |- GPU card[1] => node 0(mc3), node 1(mc4)
> + * mc5 |
> + * mc6 |- GPU card[2] => node 0(mc5), node 1(mc6)
> + * mc7 |
> + * mc8 |- GPU card[3] => node 0(mc7), node 1(mc8)
> + *
> + * sysfs entries will be populated as below:
> + *
> + * CPU # CPU node
> + * ├── mc 0
> + *
> + * GPU Nodes are enumerated sequentially after CPU nodes are populated
> + * GPU card 1 # Each Aldebaran GPU has 2 nodes/mcs
> + * ├── mc 1 # GPU node 0 == mc1, Each MC node has 4 UMCs/CSROWs
> + * │ ├── csrow 0 # UMC 0
> + * │ │ ├── channel 0 # Each UMC has 8 channels
> + * │ │ ├── channel 1 # size of each channel is 2 GB, so each UMC has 16 GB
> + * │ │ ├── channel 2
> + * │ │ ├── channel 3
> + * │ │ ├── channel 4
> + * │ │ ├── channel 5
> + * │ │ ├── channel 6
> + * │ │ ├── channel 7
> + * │ ├── csrow 1 # UMC 1
> + * │ │ ├── channel 0
> + * │ │ ├── ..
> + * │ │ ├── channel 7
> + * │ ├── .. ..
> + * │ ├── csrow 3 # UMC 3
> + * │ │ ├── channel 0
> + * │ │ ├── ..
> + * │ │ ├── channel 7
> + * │ ├── rank 0
> + * │ ├── .. ..
> + * │ ├── rank 31 # total 32 ranks/dimms from 4 UMCs
> + * ├
> + * ├── mc 2 # GPU node 1 == mc2
> + * │ ├── .. # each GPU has total 64 GB
> + *
> + * GPU card 2
> + * ├── mc 3
> + * │ ├── ..
> + * ├── mc 4
> + * │ ├── ..
> + *
> + * GPU card 3
> + * ├── mc 5
> + * │ ├── ..
> + * ├── mc 6
> + * │ ├── ..
> + *
> + * GPU card 4
> + * ├── mc 7
> + * │ ├── ..
> + * ├── mc 8
> + * │ ├── ..
> + *
> + *
> + * Heterogeneous hardware details for above context as below:
> + * - The CPU UMC (Unified Memory Controller) is mostly the same as the GPU UMC.
> + * They have chip selects (csrows) and channels. However, the layouts are different
> + * for performance, physical layout, or other reasons.
> + * - CPU UMCs use 1 channel. So we say UMC = EDAC Channel. This follows the
> + * marketing speak, example. CPU has X memory channels, etc.
> + * - CPU UMCs use up to 4 chip selects. So we say UMC chip select = EDAC CSROW.
> + * - GPU UMCs use 1 chip select. So we say UMC = EDAC CSROW.
> + * - GPU UMCs use 8 channels. So we say UMC Channel = EDAC Channel.
> + */
> --
This makes sense to me. I'm interested to see if there's any feedback from
others though.
Please fix up the commit message. Otherwise, I think this looks good.
Reviewed-by: Yazen Ghannam <yazen.ghannam@amd.com>
Thanks,
Yazen
next prev parent reply other threads:[~2022-02-09 22:34 UTC|newest]
Thread overview: 22+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-02-03 17:49 [PATCH v7 00/12] x86/edac/amd64: Add support for GPU nodes Naveen Krishna Chatradhi
2022-02-03 17:49 ` [PATCH v7 01/12] EDAC/amd64: Document heterogeneous enumeration Naveen Krishna Chatradhi
2022-02-09 22:34 ` Yazen Ghannam [this message]
2022-02-03 17:49 ` [PATCH v7 02/12] x86/amd_nb: Add support for northbridges on Aldebaran Naveen Krishna Chatradhi
2022-02-09 23:23 ` Yazen Ghannam
2022-02-03 17:49 ` [PATCH v7 03/12] EDAC/mce_amd: Extract node id from MCA_IPID Naveen Krishna Chatradhi
2022-02-09 23:31 ` Yazen Ghannam
2022-02-14 17:54 ` Chatradhi, Naveen Krishna
2022-02-03 17:49 ` [PATCH v7 04/12] EDAC/amd64: Move struct fam_type variables into amd64_pvt structure Naveen Krishna Chatradhi
2022-02-15 15:39 ` Yazen Ghannam
2022-02-03 17:49 ` [PATCH v7 05/12] EDAC/amd64: Define dynamic family ops routines Naveen Krishna Chatradhi
2022-02-15 15:49 ` Yazen Ghannam
2022-02-03 17:49 ` [PATCH v7 06/12] EDAC/amd64: Add AMD heterogeneous family 19h Model 30h-3fh Naveen Krishna Chatradhi
2022-02-15 16:20 ` Yazen Ghannam
2022-02-03 17:49 ` [PATCH v7 07/12] EDAC/amd64: Enumerate Aldebaran GPU nodes by adding family ops Naveen Krishna Chatradhi
2022-02-15 16:34 ` Yazen Ghannam
2022-02-03 17:49 ` [PATCH v7 08/12] EDAC/amd64: Add Family ops to update GPU csrow and channel info Naveen Krishna Chatradhi
2022-02-15 16:43 ` Yazen Ghannam
2022-02-03 17:49 ` [PATCH v7 09/12] EDAC/amd64: Add check for when to add DRAM base and hole Naveen Krishna Chatradhi
2022-02-03 17:49 ` [PATCH v7 10/12] EDAC/amd64: Save the number of block instances Naveen Krishna Chatradhi
2022-02-03 17:49 ` [PATCH v7 11/12] EDAC/amd64: Add address translation support for DF3.5 Naveen Krishna Chatradhi
2022-02-03 17:49 ` [PATCH v7 12/12] EDAC/amd64: Add fixed UMC to CS mapping Naveen Krishna Chatradhi
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=YgRBZjUN3AoHr7q+@yaz-ubuntu \
--to=yazen.ghannam@amd.com \
--cc=bp@alien8.de \
--cc=linux-edac@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=mchehab@kernel.org \
--cc=mingo@redhat.com \
--cc=muralimk@amd.com \
--cc=nchatrad@amd.com \
--cc=x86@kernel.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).