Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
MIME-Version: 1.0
References: <20211006193819.2654854-1-swboyd@chromium.org> <20211006193819.2654854-3-swboyd@chromium.org>
 <CAGETcx9T59dHXodt9MW=tTV_hYhtNOZzYFT=35D--VN7WJ0GqQ@mail.gmail.com>
 <CAE-0n50YqKr1nKy-4WaxsfuwPiJ5kZcf46t-U_4i-TpfXzOX1g@mail.gmail.com>
 <CAE-0n532XYgT=dTTCyLcwikvxgUyGi=TcybDh=v3wQTNb=wqyw@mail.gmail.com>
 <CAGETcx_0GRg7u3dAxP9u0qO-hfJ0N3V44CGLwFFX1kVxZ00g+w@mail.gmail.com> <CAE-0n52ytNtf5atojYrLviuAKOt3v+mLVoPeqtRUf-6uBpkiJA@mail.gmail.com>
In-Reply-To: <CAE-0n52ytNtf5atojYrLviuAKOt3v+mLVoPeqtRUf-6uBpkiJA@mail.gmail.com>
From:   Saravana Kannan <saravanak@google.com>
Date:   Thu, 7 Oct 2021 18:32:20 -0700
Message-ID: <CAGETcx9+pL1DxDyZOBGFFVsmb0haAzJTFtyQFEO9ab9u8N=FXQ@mail.gmail.com>
Subject: Re: [PATCH v2 02/34] component: Introduce the aggregate bus_type
To:     Stephen Boyd <swboyd@chromium.org>
Cc:     Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
        linux-kernel@vger.kernel.org, linux-arm-msm@vger.kernel.org,
        dri-devel@lists.freedesktop.org, freedreno@lists.freedesktop.org,
        Daniel Vetter <daniel.vetter@ffwll.ch>,
        "Rafael J. Wysocki" <rafael@kernel.org>,
        Rob Clark <robdclark@gmail.com>,
        Russell King <rmk+kernel@arm.linux.org.uk>
Content-Type: text/plain; charset="UTF-8"
Precedence: bulk

On Thu, Oct 7, 2021 at 6:24 PM Stephen Boyd <swboyd@chromium.org> wrote:
>
> Quoting Saravana Kannan (2021-10-07 18:10:24)
> > On Thu, Oct 7, 2021 at 1:11 PM Stephen Boyd <swboyd@chromium.org> wrote:
> > >
> > > Quoting Stephen Boyd (2021-10-07 11:40:07)
> > > > Quoting Saravana Kannan (2021-10-06 20:07:11)
> > > > > On Wed, Oct 6, 2021 at 12:38 PM Stephen Boyd <swboyd@chromium.org> wrote:
> > > > > > diff --git a/drivers/base/component.c b/drivers/base/component.c
> > > > > > index 0a41bbe14981..d99e99cabb99 100644
> > > > > > --- a/drivers/base/component.c
> > > > > > +++ b/drivers/base/component.c
> > > > [...]
> > > > > > +                       continue;
> > > > > > +
> > > > > > +               /* Matches put in component_del() */
> > > > > > +               get_device(&adev->dev);
> > > > > > +               c->link = device_link_add(&adev->dev, c->dev,
> > > > > > +                                         DL_FLAG_STATELESS | DL_FLAG_PM_RUNTIME);
> > > > >
> > > > > Remove the STATELESS flag and you'll get a bunch of other stuff done for free:
> > > >
> > > > I tried that and it didn't work for me. The aggregate device never
> > > > probed and I was left with no display. Let me see if I can reproduce it
> > > > with logging to provide more details.
> > >
> > > This patch fixes it (whitespace damaged sorry).
> >
> > Not sure why you have to trigger an explicit rescan, but instead of
> > this patch below, you could also try setting this flag instead?
> > DL_FLAG_AUTOPROBE_CONSUMER
> >
>
> I'd rather not conflate component driver probe with component_add()
> being called. It's simpler if that is the case, i.e. all component
> drivers are calling component_add() in their driver probe routine, but I
> don't know if that's always true. I did this poor audit of the kernel
>
> $ git grep -p \[^_]component_add | grep \.c=  | grep -v probe
> drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c=static int
> amdgpu_dm_audio_init(struct amdgpu_device *adev)
> drivers/gpu/drm/i915/display/intel_audio.c=static void
> i915_audio_component_init(struct drm_i915_private *dev_priv)
> drivers/gpu/drm/i915/display/intel_hdcp.c=void
> intel_hdcp_component_init(struct drm_i915_private *dev_priv)
> drivers/gpu/drm/nouveau/dispnv50/disp.c=nv50_audio_component_init(struct
> nouveau_drm *drm)
> drivers/gpu/drm/rockchip/dw-mipi-dsi-rockchip.c=static int
> dw_mipi_dsi_rockchip_host_attach(void *priv_data,
> drivers/gpu/drm/rockchip/dw-mipi-dsi-rockchip.c=static int
> dw_mipi_dsi_dphy_init(struct phy *phy)
> drivers/gpu/drm/vc4/vc4_dsi.c=static int vc4_dsi_host_attach(struct
> mipi_dsi_host *host,
>
> and then peeking at rockchip above I see that component_add() is called
> in the mipi dsi attach ops and then I got lost trying to figure out
> where it ends up. Maybe it is still in some probe call?
>
> Anyway, I think we still have to do a rescan so that we can try to bind
> the aggregate device. Is there a better API to use for that?

If you know the exact device, you could call device_attach()? That's
what bus_rescan_devices() eventually calls, but you skip the
unnecessary looping over all the other devices.

-Saravana