From: Po-Kai Chi <pk.chi@mediatek.com>
To: Rob Herring <robh@kernel.org>
Cc: "Matthias Brugger" <matthias.bgg@gmail.com>,
"linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>,
"linux-arm-kernel@lists.infradead.org"
<linux-arm-kernel@lists.infradead.org>,
"linux-mediatek@lists.infradead.org"
<linux-mediatek@lists.infradead.org>,
"devicetree@vger.kernel.org" <devicetree@vger.kernel.org>,
wsd_upstream <wsd_upstream@mediatek.com>,
"CC Hwang (黃致銓)" <cc.hwang@mediatek.com>,
"Loda Chou (周宏霖)" <Loda.Chou@mediatek.com>
Subject: Re: [PATCH v1 1/4] dt-bindings: memory: Add binding for MediaTek Common DRAM Controller
Date: Thu, 1 Apr 2021 16:52:42 +0800 [thread overview]
Message-ID: <1617267162.24434.31.camel@mtkswgap22> (raw)
In-Reply-To: <20210330135814.GA224111@robh.at.kernel.org>
Hello Rob,
Thanks for the remind about dt_binding_check fail and the comments, my
reply is as follows and will fix it in the next version (v2).
Po-Kai
On Tue, 2021-03-30 at 21:58 +0800, Rob Herring wrote:
> On Tue, Mar 30, 2021 at 01:22:08PM +0800, Po-Kai Chi wrote:
> > This patch adds the documentation of the device-tree binding for
> > MediaTek Common DRAM Controller.
> >
> > Signed-off-by: Po-Kai Chi <pk.chi@mediatek.com>
> > ---
> > .../memory-controllers/mediatek,dramc.yaml | 155 ++++++++++++++++++++
> > 1 file changed, 155 insertions(+)
> > create mode 100644 Documentation/devicetree/bindings/memory-controllers/mediatek,dramc.yaml
> >
> > diff --git a/Documentation/devicetree/bindings/memory-controllers/mediatek,dramc.yaml b/Documentation/devicetree/bindings/memory-controllers/mediatek,dramc.yaml
> > new file mode 100644
> > index 0000000..0217ce0
> > --- /dev/null
> > +++ b/Documentation/devicetree/bindings/memory-controllers/mediatek,dramc.yaml
> > @@ -0,0 +1,155 @@
> > +# SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause)
> > +# Copyright (c) 2021 MediaTek Inc.
> > +%YAML 1.2
> > +---
> > +$id: https://urldefense.com/v3/__http://devicetree.org/schemas/memory-controllers/mediatek,dramc.yaml*__;Iw!!CTRNKA9wMg0ARbw!y9zM5d-aNLK99Y_ag2yvqq3TI1Xvm6TV_Vu03VVD3Qbe69N1qZXFFk2DUFb6CG0$
> > +$schema: https://urldefense.com/v3/__http://devicetree.org/meta-schemas/core.yaml*__;Iw!!CTRNKA9wMg0ARbw!y9zM5d-aNLK99Y_ag2yvqq3TI1Xvm6TV_Vu03VVD3Qbe69N1qZXFFk2DiMad89A$
> > +
> > +title: MediaTek DRAM Controller
> > +
> > +maintainers:
> > + - Po-Kai Chi <pk.chi@mediatek.com>
> > +
> > +description: |
> > + MediaTek DRAM controller (DRAMC) provides an interface to query information
> > + about DRAM which collected from bootloader and device tree.
> > + This is mainly used by MediaTek Extended Memory Interface (EMI) and DVFS Resource
> > + Control (DVFSRC).
> > +
> > +properties:
> > + compatible:
> > + items:
> > + - enum:
> > + - mediatek,mt6779-dramc
> > +
> > + reg:
> > + description:
> > + Base address of MediaTek DRAM related hardware modules, each channel has
> > + its own base address in order of
> > + DRAMC_AO_{CH}, DRAMC_NAO_{CH}, DDRPHY_AO_{CH}.
> > + minItems: 3 # 3 * N channels
> > + maxItems: 6
> > +
> > + dram_type:
>
> These need to be either common or have a vendor prefix.
>
> Also, s/_/-/
Okay, I have revised the naming rule according to writing-schema.rst.
> > + description:
> > + The DRAM type of current DRAM chip.
> > + This property is filled in by bootloader according to the board hardware
> > + configuration.
> > + $ref: /schemas/types.yaml#/definitions/uint32
> > + minimum: 0
> > + maximum: 7
> > +
> > + support_channel_cnt:
> > + description:
> > + The maximum DRAM channel count supported by SoC.
> > + $ref: /schemas/types.yaml#/definitions/uint32
> > + minimum: 1
> > + maximum: 4
> > +
> > + channel_cnt:
> > + description:
> > + The DRAM channel count of current DRAM chip.
> > + This property is filled in by bootloader according to the board hardware
> > + configuration.
> > + $ref: /schemas/types.yaml#/definitions/uint32
> > + minimum: 1
> > + maximum: 4
> > +
> > + rank_cnt:
> > + description:
> > + The DRAM rank count of current DRAM chip.
> > + This property is filled in by bootloader according to the board hardware
> > + configuration.
> > + $ref: /schemas/types.yaml#/definitions/uint32
> > + minimum: 1
> > + maximum: 2
> > +
> > + rank_size:
> > + description:
> > + The size of each DRAM rank.
> > + This property is filled in by bootloader according to the board hardware
> > + configuration.
> > + $ref: /schemas/types.yaml#/definitions/uint64
There may be some misunderstanding.
rank_size uses the full 64 bits to describe the size of each DRAM rank.
So the type of rank_size should be uint64-array, instead of uint64.
> > + minItems: 1
> > + maxItems: 2
> > + items:
> > + minimum: 0x0
> > + maximum: 0x100000000 # support up to 4GB in single rank
> > +
> > + mr_cnt:
> > + description:
> > + Specifies how many sets of DRAM mode register information to provide.
> > + This property is filled in by bootloader according to the board hardware
> > + configuration.
> > + $ref: /schemas/types.yaml#/definitions/uint32
> > + maximum: 40 # total 40 MRs for JEDEC LPDDR4X
> > +
> > + mr:
> > + description:
> > + Pair of DRAM mode register information.
> > + This property is filled in by bootloader according to the board hardware
> > + configuration.
> > + $ref: /schemas/types.yaml#/definitions/uint32-matrix
> > + maxItems: 40 # align with mr_cnt
> > + items:
> > + items:
> > + - description:
> > + Mode register index
> > + - description:
> > + Mode register value
> > +
> > + freq_cnt:
> > + description:
> > + Specifies how many sets of DRAM data clock rate supported by SoC.
> > + $ref: /schemas/types.yaml#/definitions/uint32
> > +
> > + freq_step:
> > + description:
> > + The DRAM data clock rate may be slightly different from those defined
> > + by the specification due to errors in multiples of the base frequency.
> > + This describe the mapping from real data clock rate measured by
> > + frequency meter to JEDEC data clock rate.
> > + $ref: /schemas/types.yaml#/definitions/uint32-matrix
> > + items:
> > + items:
> > + - description:
> > + Real data rate
> > + - description:
> > + Spec data rate
>
> Looks like an OPP table.
Yes, It's essentially an OPP table, but also records the relationship
between real data rate and spec data rate.
> > +
> > +required:
> > + - compatible
> > + - reg
> > + - dram_type
> > + - support_channel_cnt
> > + - channel_cnt
> > + - rank_cnt
> > + - mr_cnt
> > + - freq_cnt
> > +
> > +additionalProperties: false
> > +
> > +examples:
> > + - |
> > + dramc@10230000 {
> > + compatible = "mediatek,mt6779-dramc";
> > + reg = <0 0x10230000 0 0x2000>, /* DRAMC AO CHA */
> > + <0 0x10240000 0 0x2000>, /* DRAMC AO CHB */
> > + <0 0x10234000 0 0x1000>, /* DRAMC NAO CHA */
> > + <0 0x10244000 0 0x1000>, /* DRAMC NAO CHB */
> > + <0 0x10238000 0 0x2000>, /* DDRPHY AO CHA */
> > + <0 0x10248000 0 0x2000>; /* DDRPHY AO CHB */
> > + dram_type = <0>;
> > + support_channel_cnt = <2>;
> > + channel_cnt = <2>;
> > + rank_cnt = <2>;
> > + rank_size = <0x40000000 0x40000000>;
>
> You defined this as 64-bit, so this is a single value?
No, these are independent values.I have updated the description of this
property in the previous paragraph.
And the sample will be updated to
mediatek,rank-size = <0x40000000>, <0x40000000>;
>
> > + mr_cnt = <1>;
> > + mr = <0x5 0xff>;
> > + freq_cnt = <6>;
> > + freq_step = <3718 3733>,
> > + <3094 3200>,
> > + <2392 2400>,
> > + <1534 1600>,
> > + <1196 1200>,
> > + <754 800>;
> > + };
> > --
> > 1.7.9.5
> >
next prev parent reply other threads:[~2021-04-01 8:53 UTC|newest]
Thread overview: 8+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-03-30 5:22 [PATCH v1] memory: mediatek: add DRAM controller driver Po-Kai Chi
2021-03-30 5:22 ` [PATCH v1 1/4] dt-bindings: memory: Add binding for MediaTek Common DRAM Controller Po-Kai Chi
2021-03-30 13:08 ` Rob Herring
2021-03-30 13:58 ` Rob Herring
2021-04-01 8:52 ` Po-Kai Chi [this message]
2021-03-30 5:22 ` [PATCH v1 2/4] memory: mediatek: add DRAM controller driver Po-Kai Chi
2021-03-30 5:22 ` [PATCH v1 3/4] arm64: dts: add DRAMC node for MT6779 Po-Kai Chi
2021-03-30 5:22 ` [PATCH v1 4/4] arm64: defconfig: Enable MediaTek DRAMC common driver Po-Kai Chi
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=1617267162.24434.31.camel@mtkswgap22 \
--to=pk.chi@mediatek.com \
--cc=Loda.Chou@mediatek.com \
--cc=cc.hwang@mediatek.com \
--cc=devicetree@vger.kernel.org \
--cc=linux-arm-kernel@lists.infradead.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mediatek@lists.infradead.org \
--cc=matthias.bgg@gmail.com \
--cc=robh@kernel.org \
--cc=wsd_upstream@mediatek.com \
/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).