[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <3a2v2i5lximl6q7eidd5zhjoyckz2fvgv52rhjruiuyuc6v3a3@5kdqa2eypidt>
Date: Mon, 13 Oct 2025 17:34:06 +0300
From: Dmitry Baryshkov <dmitry.baryshkov@....qualcomm.com>
To: Junjie Cao <caojunjie650@...il.com>
Cc: Neil Armstrong <neil.armstrong@...aro.org>,
Jessica Zhang <jessica.zhang@....qualcomm.com>,
Maarten Lankhorst <maarten.lankhorst@...ux.intel.com>,
Maxime Ripard <mripard@...nel.org>,
Thomas Zimmermann <tzimmermann@...e.de>,
David Airlie <airlied@...il.com>, Simona Vetter <simona@...ll.ch>,
Rob Herring <robh@...nel.org>,
Krzysztof Kozlowski <krzk+dt@...nel.org>,
Conor Dooley <conor+dt@...nel.org>,
Rob Clark <robin.clark@....qualcomm.com>,
Dmitry Baryshkov <lumag@...nel.org>,
Abhinav Kumar <abhinav.kumar@...ux.dev>, Sean Paul <sean@...rly.run>,
Marijn Suijten <marijn.suijten@...ainline.org>,
Antonino Maniscalco <antomani103@...il.com>,
Jonathan Marek <jonathan@...ek.ca>, Eugene Lepshy <fekz115@...il.com>,
Jun Nie <jun.nie@...aro.org>, dri-devel@...ts.freedesktop.org,
devicetree@...r.kernel.org, linux-kernel@...r.kernel.org,
linux-arm-msm@...r.kernel.org, freedreno@...ts.freedesktop.org
Subject: Re: [PATCH 2/3] drm/msm/dsi: support DSC configurations with
slice_per_pkt > 1
On Mon, Oct 13, 2025 at 09:17:04PM +0800, Junjie Cao wrote:
> Dmitry Baryshkov <dmitry.baryshkov@....qualcomm.com> 于2025年10月13日周一 20:31写道:
> > On Mon, Oct 13, 2025 at 07:04:43PM +0800, Junjie Cao wrote:
> > > Dmitry Baryshkov <dmitry.baryshkov@....qualcomm.com> 于2025年10月13日周一 17:39写道:
> > > > On 13/10/2025 04:52, 曹俊杰 wrote:
> > > > > >Dmitry Baryshkov <dmitry.baryshkov@....qualcomm.com
> > > > > <mailto:dmitry.baryshkov@....qualcomm.com>> 于2025年10月2日周四 10:04写道:
> > > > > >On Wed, Oct 01, 2025 at 09:59:13PM +0800, Junjie Cao wrote:
> > > > > >> From: Jun Nie <jun.nie@...aro.org <mailto:jun.nie@...aro.org>>
> > > > > >>
> > > > > >> Some panels support multiple slice to be sent in a single DSC
> > > > > packet. And
> > > > > >> this feature is a must for specific panels, such as JDI LPM026M648C.
> > > > > Add a
> > > > > >> dsc_slice_per_pkt member into struct mipi_dsi_device and support the
> > > > > >> feature in msm mdss driver.
> > > > > >>
> > > > > >> Co-developed-by: Jonathan Marek <jonathan@...ek.ca
> > > > > <mailto:jonathan@...ek.ca>>
> > > > > >> Signed-off-by: Jonathan Marek <jonathan@...ek.ca
> > > > > <mailto:jonathan@...ek.ca>>
> > > > > >> Signed-off-by: Jun Nie <jun.nie@...aro.org <mailto:jun.nie@...aro.org>>
> > > > > >> Signed-off-by: Junjie Cao <caojunjie650@...il.com
> > > > > <mailto:caojunjie650@...il.com>>
> > > > > >> ---
> > > > > >> drivers/gpu/drm/msm/dsi/dsi_host.c | 25 ++++++++++---------------
> > > > > >> include/drm/drm_mipi_dsi.h | 2 ++
> > > > > >> 2 files changed, 12 insertions(+), 15 deletions(-)
> > > > > >
> > > > > >Please extract the generic part, so that it can be merged through a
> > > > > >generic tree.
> > > > > >
> > > > >
> > > > > Sorry, I don't get it. The generic part, generic tree? Do you mean
> > > > > the drm tree? `slice_per_pkt >= 2` is seen on the panels of these
> > > > > tablets that are equipped with qcom chips. I don't know if these
> > > > > panels are used on other platforms, and if it is necessary to do it
> > > > > in drm.
> > > >
> > > > There are two changes here:
> > > > - MIPI DSI header change
> > > > - msm DSI driver
> > > >
> > > > I've asked to split it to those two commits so that he change for
> > > > drm_mipi_dsi.h is more obvious for reviewers and so that it can be
> > > > merged through a drm-misc tree (or through drm-msm tree provided it gets
> > > > a necessary ack).
> > > >
> > >
> > > Thanks for your clear explanation.
> > >
> > > I don't mind to add the field separately. But should I submit it
> > > with the panel driver together? Otherwise, this field is unused
> > > for a while.
> > >
> > > However, as you mentioned, this is not a part of standard, neither
> > > mipi dsi nor VESA DSC. Recently, only Qualcomm devices require it
> > > to calculate parameters, then we use them to program registers. Why
> > > don't we parse the field from devicetree?
> >
> > Because the value is uniquelly identified by the panel's compat string.
> >
>
> Yes, it is panel specified.
> But can we set it for every panel like
>
> &mdss_dsi0 {
> qcom,mdss-dsc-slice-per-pkt = <2>;
>
> status = "okay";
>
> panel: panel@0 {
> compatible = "foo,bar";
> reg = <0>;
> };
> };
>
> or moving the property to panel node? We access it from child node.
Why do you need it in DT if the panel driver can provide this
information.
--
With best wishes
Dmitry
Powered by blists - more mailing lists