linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Chen-Yu Tsai <wenst@chromium.org>
To: AngeloGioacchino Del Regno  <angelogioacchino.delregno@collabora.com>
Cc: Allen-KH Cheng <allen-kh.cheng@mediatek.com>,
	Rob Herring <robh+dt@kernel.org>,
	Krzysztof Kozlowski <krzysztof.kozlowski+dt@linaro.org>,
	Matthias Brugger <matthias.bgg@gmail.com>,
	Project_Global_Chrome_Upstream_Group@mediatek.com,
	devicetree@vger.kernel.org, linux-arm-kernel@lists.infradead.org,
	linux-kernel@vger.kernel.org, linux-mediatek@lists.infradead.org
Subject: Re: [PATCH] arm64: dts: mt8192: Add adsp power domain controller
Date: Thu, 1 Dec 2022 18:05:39 +0800	[thread overview]
Message-ID: <CAGXv+5H+msQ1ct5CUzYnT_BmYPnS72HbivvyO39uOfMfXbo5qg@mail.gmail.com> (raw)
In-Reply-To: <8ae2a510-abf6-6e1b-9893-293db7d930e7@collabora.com>

On Thu, Dec 1, 2022 at 5:39 PM AngeloGioacchino Del Regno
<angelogioacchino.delregno@collabora.com> wrote:
>
> Il 01/12/22 10:10, Chen-Yu Tsai ha scritto:
> > On Thu, Dec 1, 2022 at 5:07 PM AngeloGioacchino Del Regno
> > <angelogioacchino.delregno@collabora.com> wrote:
> >>
> >> Il 01/12/22 08:33, Allen-KH Cheng ha scritto:
> >>> Add adsp power domain controller node for mt8192 SoC.
> >>>
> >>> Signed-off-by: Allen-KH Cheng <allen-kh.cheng@mediatek.com>
> >>> ---
> >>> Ref: https://lore.kernel.org/all/2ec80bd8-dfef-d2e6-eb41-6e6088043e33@collabora.com/
> >>>       [Allen-KH Cheng <allen-kh.cheng@mediatek.com>]
> >>> ---
> >>> ---
> >>>    arch/arm64/boot/dts/mediatek/mt8192.dtsi | 8 ++++++++
> >>>    include/dt-bindings/power/mt8192-power.h | 1 +
> >>>    2 files changed, 9 insertions(+)
> >>>
> >>
> >> Allen, thanks for this one, but it's incomplete...
> >>
> >> First of all, you must add the power domain on the driver itself, specifically,
> >> in drivers/soc/mediatek/mt8192-pm-domains.h - otherwise this change will have no
> >> effect!
> >
> > Actually it's worse. The driver doesn't know about the new power domain,
> > and so it will fail to probe. We might need to make the power domain driver
> > fail gracefully and skip unknown power domains.
> >
>
> Right. It's worse. I don't know, though, if gracefully skipping unknown power
> domains in the driver would be a good decision... as sometimes error messages
> go unnoticed.
>
> When the platform "explodes" instead, you're forced to read that log carefully
> and get it working again... Anyway, I'm only thinking out loud, nothing less and
> nothing more than that :-)

Me too. :)

> By the way, we can probably expand on that topic a bit later, as it's outside of
> the scope of this specific change.
>
> Back to topic, if we get one series containing both changes (devicetree, bindings
> and driver) with the right Fixes tags and/or Cc stable, we shouldn't have such
> issue on backports so we can probably ignore that potential issue, I think? :-)

Everything goes through the soc tree, so they should appear in Linus's tree
and get picked by stable at more or less the same time. I think we would
want the driver change to appear before the dts change, for bisectability's
sake (because of header dependencies and driver errors).

So we probably want:
1. driver + binding header changes
2. dtsi changes

And have these merged through fixes so that the history between them is linear.


ChenYu

> Cheers,
> Angelo
>
> > ChenYu
> >
> >> ...Then, as Chen-Yu said, you should also add the power domain to the scp_adsp
> >> clock node as that's solving the lockup issue...
> >>
> >> .......and last, but not least: we need a Fixes tag to backport this fix, here
> >> and on the commit that adds the missing power domain in the driver.
> >>
> >> Thanks,
> >> Angelo
> >>
> >>> diff --git a/arch/arm64/boot/dts/mediatek/mt8192.dtsi b/arch/arm64/boot/dts/mediatek/mt8192.dtsi
> >>> index 424fc89cc6f7..e71afba871fc 100644
> >>> --- a/arch/arm64/boot/dts/mediatek/mt8192.dtsi
> >>> +++ b/arch/arm64/boot/dts/mediatek/mt8192.dtsi
> >>> @@ -514,6 +514,14 @@
> >>>                                                };
> >>>                                        };
> >>>                                };
> >>> +
> >>> +                             power-domain@MT8192_POWER_DOMAIN_ADSP {
> >>> +                                     reg = <MT8192_POWER_DOMAIN_ADSP>;
> >>> +                                     clocks = <&topckgen CLK_TOP_ADSP_SEL>;
> >>> +                                     clock-names = "adsp";
> >>> +                                     mediatek,infracfg = <&infracfg>;
> >>> +                                     #power-domain-cells = <0>;
> >>> +                             };
> >>>                        };
> >>>                };
> >>>
> >>> diff --git a/include/dt-bindings/power/mt8192-power.h b/include/dt-bindings/power/mt8192-power.h
> >>> index 4eaa53d7270a..63e81cd0d06d 100644
> >>> --- a/include/dt-bindings/power/mt8192-power.h
> >>> +++ b/include/dt-bindings/power/mt8192-power.h
> >>> @@ -28,5 +28,6 @@
> >>>    #define MT8192_POWER_DOMAIN_CAM_RAWA        18
> >>>    #define MT8192_POWER_DOMAIN_CAM_RAWB        19
> >>>    #define MT8192_POWER_DOMAIN_CAM_RAWC        20
> >>> +#define MT8192_POWER_DOMAIN_ADSP     21
> >>>
> >>>    #endif /* _DT_BINDINGS_POWER_MT8192_POWER_H */
> >>>
> >>
>
>

  reply	other threads:[~2022-12-01 10:05 UTC|newest]

Thread overview: 10+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-12-01  7:33 [PATCH] arm64: dts: mt8192: Add adsp power domain controller Allen-KH Cheng
2022-12-01  8:58 ` Chen-Yu Tsai
2022-12-01  9:07 ` AngeloGioacchino Del Regno
2022-12-01  9:10   ` Chen-Yu Tsai
2022-12-01  9:39     ` AngeloGioacchino Del Regno
2022-12-01 10:05       ` Chen-Yu Tsai [this message]
2022-12-01 10:10         ` AngeloGioacchino Del Regno
2022-12-02  8:00           ` Allen-KH Cheng (程冠勳)
2022-12-01 10:22 ` Krzysztof Kozlowski
2022-12-02  8:13   ` Allen-KH Cheng (程冠勳)

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=CAGXv+5H+msQ1ct5CUzYnT_BmYPnS72HbivvyO39uOfMfXbo5qg@mail.gmail.com \
    --to=wenst@chromium.org \
    --cc=Project_Global_Chrome_Upstream_Group@mediatek.com \
    --cc=allen-kh.cheng@mediatek.com \
    --cc=angelogioacchino.delregno@collabora.com \
    --cc=devicetree@vger.kernel.org \
    --cc=krzysztof.kozlowski+dt@linaro.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+dt@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).