linux-mediatek.lists.infradead.org archive mirror
 help / color / mirror / Atom feed
From: Enric Balletbo i Serra <enric.balletbo@collabora.com>
To: Eizan Miyamoto <eizan@chromium.org>, linux-kernel@vger.kernel.org
Cc: wenst@chromium.org, chunkuang.hu@kernel.org,
	yong.wu@mediatek.com, houlong.wei@mediatek.com,
	Andrew-CT Chen <andrew-ct.chen@mediatek.com>,
	Matthias Brugger <matthias.bgg@gmail.com>,
	Mauro Carvalho Chehab <mchehab@kernel.org>,
	Minghsiu Tsai <minghsiu.tsai@mediatek.com>,
	linux-arm-kernel@lists.infradead.org,
	linux-media@vger.kernel.org, linux-mediatek@lists.infradead.org
Subject: Re: [PATCH v5 3/8] media: mtk-mdp: don't pm_run_time_get/put for master comp in clock_on
Date: Mon, 12 Jul 2021 12:12:26 +0200	[thread overview]
Message-ID: <9ee56e91-88c1-c24d-c797-34d7defaef2e@collabora.com> (raw)
In-Reply-To: <20210709122040.v5.3.Ifd87681955689a08cce4c65e72ee5e8c6d09af95@changeid>

Hi Eizan,

Thank you for your patch.

On 9/7/21 4:23, Eizan Miyamoto wrote:
> The original intent of commit 86698b9505bbc ("media: mtk-mdp: convert
> mtk_mdp_dev.comp array to list") was to create a list to track all the
> MDP components that needed to have their clocks enabled/disabled when
> calling mtk_mdp_comp_clock_on/off. However, there was a bug inside
> mtk_mdp_register_component where the args to a call to list_add were
> swapped. The result is that only one component was added to
> mtk_mdp_dev.comp_list because comp_list was instead being
> repeatedly added to the single element lists headed by each
> mtk_mdp_comp.
> 
> The order of the args to list_add in mtk_mdp_register_component was
> fixed in https://patchwork.kernel.org/patch/11742895/ (Fix Null pointer
> dereference when calling list_add).
> 
> Then, as a result of https://patchwork.kernel.org/patch/11530769/
> (mtk-mdp: use pm_runtime in MDP component driver) if all the components
> are added to the component list, the mdp "master" / rdma0 component
> ends up having pm_runtime_get_sync() called on it twice recursively:
> 
>     rpm_resume+0x694/0x8f8
>     __pm_runtime_resume+0x7c/0xa0 ***NOTE***
>     mtk_mdp_comp_clock_on+0x48/0x104 [mtk_mdp]
>     mtk_mdp_pm_resume+0x2c/0x44 [mtk_mdp]
>     pm_generic_runtime_resume+0x34/0x48
>     __genpd_runtime_resume+0x6c/0x80
>     genpd_runtime_resume+0x104/0x1ac
>     __rpm_callback+0x120/0x238
>     rpm_callback+0x34/0x8c
>     rpm_resume+0x7a0/0x8f8
>     __pm_runtime_resume+0x7c/0xa0 ***NOTE***
>     mtk_mdp_m2m_start_streaming+0x2c/0x3c [mtk_mdp]
> 
> (The calls to pm_runtime_get_sync are inlined and correspond to the
> calls to __pm_runtime_resume). It is not correct to have
> pm_runtime_get_sync called recursively and the second call will block
> indefinitely.
> 
> As a result of all that, this change factors mtk_mdp_comp_clock_on/off
> into mtk_mdp_comp_power_on/off and moves the calls to
> pm_runtime_get/put into the power_on/off functions.
> 
> This change then special-cases the master/rdma0 MDP component and does
> these things:
> - the master/rdma0 component is not added to mtk_mdp_dev.comp_list
> - the master/rdma0 component has its clocks (*but not power*) toggled
>   by mtk_mpd_comp_clock_on/off inside mtk_mdp_clock_on/off.
> - the other components have their clocks *and* power toggled with
>   mtk_mdp_comp_power_on/off.
> 
> This change introduces the assumption that mtk_mdp_pm_resume will
> always be called though a callback from pm_runtime_get_sync made on the
> master / rdma0 component.
> 
> Signed-off-by: Eizan Miyamoto <eizan@chromium.org>
> ---
> 
> (no changes since v1)
> 
>  drivers/media/platform/mtk-mdp/mtk_mdp_comp.c | 28 ++++++++++++++-----
>  drivers/media/platform/mtk-mdp/mtk_mdp_comp.h |  3 ++
>  drivers/media/platform/mtk-mdp/mtk_mdp_core.c | 23 ++++++++++-----
>  3 files changed, 40 insertions(+), 14 deletions(-)
> 
> diff --git a/drivers/media/platform/mtk-mdp/mtk_mdp_comp.c b/drivers/media/platform/mtk-mdp/mtk_mdp_comp.c
> index 8b6158379f41..ce54863c24ce 100644
> --- a/drivers/media/platform/mtk-mdp/mtk_mdp_comp.c
> +++ b/drivers/media/platform/mtk-mdp/mtk_mdp_comp.c
> @@ -53,9 +53,9 @@ static const struct of_device_id mtk_mdp_comp_driver_dt_match[] = {
>  };
>  MODULE_DEVICE_TABLE(of, mtk_mdp_comp_driver_dt_match);
>  
> -void mtk_mdp_comp_clock_on(struct mtk_mdp_comp *comp)
> +void mtk_mdp_comp_power_on(struct mtk_mdp_comp *comp)
>  {
> -	int i, err;
> +	int err;
>  
>  	if (comp->larb_dev) {
>  		err = mtk_smi_larb_get(comp->larb_dev);
> @@ -66,9 +66,25 @@ void mtk_mdp_comp_clock_on(struct mtk_mdp_comp *comp)
>  
>  	err = pm_runtime_get_sync(comp->dev);
>  	if (err < 0)
> -		dev_err(comp->dev,
> -			"failed to runtime get, err %d.\n",
> -			err);
> +		dev_err(comp->dev, "failed to runtime get, err %d.\n", err);

You can do that change in the previous patch which is the one that introduces
this dev_err call.

> +
> +	mtk_mdp_comp_clock_on(comp);
> +}
> +
> +void mtk_mdp_comp_power_off(struct mtk_mdp_comp *comp)
> +{
> +	int err;
> +
> +	mtk_mdp_comp_clock_off(comp);
> +
> +	err = pm_runtime_put_sync(comp->dev);
> +	if (err < 0)
> +		dev_err(comp->dev, "failed to runtime put, err %d.\n", err);
> +}
> +
> +void mtk_mdp_comp_clock_on(struct mtk_mdp_comp *comp)
> +{
> +	int i, err;
>  
>  	for (i = 0; i < ARRAY_SIZE(comp->clk); i++) {
>  		if (IS_ERR(comp->clk[i]))
> @@ -91,8 +107,6 @@ void mtk_mdp_comp_clock_off(struct mtk_mdp_comp *comp)
>  
>  	if (comp->larb_dev)
>  		mtk_smi_larb_put(comp->larb_dev);
> -
> -	pm_runtime_put_sync(comp->dev);

Ok, now that you are changing these functions I think it will be good to rework
these clock_on and off properly. First returning an error when is needed and
handling it properly on the error path instead of ignoring.

>  }
>  
>  static int mtk_mdp_comp_bind(struct device *dev, struct device *master,
> diff --git a/drivers/media/platform/mtk-mdp/mtk_mdp_comp.h b/drivers/media/platform/mtk-mdp/mtk_mdp_comp.h
> index 5e2ee94a1b37..15ef9539a5a7 100644
> --- a/drivers/media/platform/mtk-mdp/mtk_mdp_comp.h
> +++ b/drivers/media/platform/mtk-mdp/mtk_mdp_comp.h
> @@ -23,6 +23,9 @@ struct mtk_mdp_comp {
>  
>  int mtk_mdp_comp_init(struct mtk_mdp_comp *comp, struct device *dev);
>  
> +void mtk_mdp_comp_power_on(struct mtk_mdp_comp *comp);
> +void mtk_mdp_comp_power_off(struct mtk_mdp_comp *comp);
> +
>  void mtk_mdp_comp_clock_on(struct mtk_mdp_comp *comp);
>  void mtk_mdp_comp_clock_off(struct mtk_mdp_comp *comp);
>  
> diff --git a/drivers/media/platform/mtk-mdp/mtk_mdp_core.c b/drivers/media/platform/mtk-mdp/mtk_mdp_core.c
> index 78c40a61df1d..3558a6587f51 100644
> --- a/drivers/media/platform/mtk-mdp/mtk_mdp_core.c
> +++ b/drivers/media/platform/mtk-mdp/mtk_mdp_core.c
> @@ -54,8 +54,15 @@ static void mtk_mdp_clock_on(struct mtk_mdp_dev *mdp)
>  {
>  	struct mtk_mdp_comp *comp_node;
>  
> +	/*
> +	 * The master / rdma0 component will have pm_runtime_get_sync called
> +	 * on it through mtk_mdp_m2m_start_streaming, making it unnecessary to
> +	 * have mtk_mdp_comp_power_on called on it.
> +	 */
> +	mtk_mdp_comp_clock_on(&mdp->comp_self);
> +
>  	list_for_each_entry(comp_node, &mdp->comp_list, node)
> -		mtk_mdp_comp_clock_on(comp_node);
> +		mtk_mdp_comp_power_on(comp_node);
>  }
>  
>  static void mtk_mdp_clock_off(struct mtk_mdp_dev *mdp)
> @@ -63,7 +70,14 @@ static void mtk_mdp_clock_off(struct mtk_mdp_dev *mdp)
>  	struct mtk_mdp_comp *comp_node;
>  
>  	list_for_each_entry(comp_node, &mdp->comp_list, node)
> -		mtk_mdp_comp_clock_off(comp_node);
> +		mtk_mdp_comp_power_off(comp_node);
> +
> +	/*
> +	 * The master / rdma0 component will have pm_runtime_put called
> +	 * on it through mtk_mdp_m2m_stop_streaming, making it unnecessary to
> +	 * have mtk_mdp_comp_power_off called on it.
> +	 */
> +	mtk_mdp_comp_clock_off(&mdp->comp_self);
>  }
>  
>  static void mtk_mdp_wdt_worker(struct work_struct *work)
> @@ -102,8 +116,6 @@ static int mtk_mdp_master_bind(struct device *dev)
>  	int status;
>  	struct mtk_mdp_dev *mdp = dev_get_drvdata(dev);
>  
> -	mtk_mdp_register_component(mdp, &mdp->comp_self);
> -
>  	status = component_bind_all(dev, mdp);
>  	if (status) {
>  		dev_err(dev, "Failed to bind all components: %d\n", status);
> @@ -125,8 +137,6 @@ static int mtk_mdp_master_bind(struct device *dev)
>  	component_unbind_all(dev, mdp);
>  
>  err_component_bind_all:
> -	mtk_mdp_unregister_component(mdp, &mdp->comp_self);
> -
>  	return status;
>  }
>  
> @@ -137,7 +147,6 @@ static void mtk_mdp_master_unbind(struct device *dev)
>  	pm_runtime_disable(dev);
>  	mtk_mdp_unregister_m2m_device(mdp);
>  	component_unbind_all(dev, mdp);
> -	mtk_mdp_unregister_component(mdp, &mdp->comp_self);
>  }
>  
>  static const struct component_master_ops mtk_mdp_com_ops = {
> 

_______________________________________________
Linux-mediatek mailing list
Linux-mediatek@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-mediatek

  reply	other threads:[~2021-07-12 10:14 UTC|newest]

Thread overview: 16+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-07-09  2:23 [PATCH v5 0/8] Refactor MTK MDP driver into core/components Eizan Miyamoto
2021-07-09  2:23 ` [PATCH v5 1/8] mtk-mdp: add driver to probe mdp components Eizan Miyamoto
2021-07-12 10:11   ` Enric Balletbo i Serra
2021-07-09  2:23 ` [PATCH v5 2/8] mtk-mdp: use pm_runtime in MDP component driver Eizan Miyamoto
2021-07-12 10:12   ` Enric Balletbo i Serra
2021-07-09  2:23 ` [PATCH v5 3/8] media: mtk-mdp: don't pm_run_time_get/put for master comp in clock_on Eizan Miyamoto
2021-07-12 10:12   ` Enric Balletbo i Serra [this message]
2021-07-09  2:23 ` [PATCH v5 4/8] mtk-mdp: soc: mediatek: register mdp from mmsys Eizan Miyamoto
2021-07-12 10:12   ` Enric Balletbo i Serra
2021-07-09  2:23 ` [PATCH v5 5/8] media: mtk-mdp: search for vpu node instead of linking it to a property Eizan Miyamoto
2021-07-09  2:23 ` [PATCH v5 6/8] media: mtk-mdp: propagate errors better in pm_suspend/resume Eizan Miyamoto
2021-07-12 10:12   ` Enric Balletbo i Serra
2021-07-09  2:23 ` [PATCH v5 7/8] media: mtk-mdp: use mdp-rdma0 alias to point to MDP master Eizan Miyamoto
2021-07-12 10:12   ` Enric Balletbo i Serra
2021-07-09  2:23 ` [PATCH v5 8/8] dts: mtk-mdp: remove mediatek, vpu property from primary MDP device Eizan Miyamoto
2021-07-12 10:13   ` [PATCH v5 8/8] dts: mtk-mdp: remove mediatek,vpu " Enric Balletbo i Serra

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=9ee56e91-88c1-c24d-c797-34d7defaef2e@collabora.com \
    --to=enric.balletbo@collabora.com \
    --cc=andrew-ct.chen@mediatek.com \
    --cc=chunkuang.hu@kernel.org \
    --cc=eizan@chromium.org \
    --cc=houlong.wei@mediatek.com \
    --cc=linux-arm-kernel@lists.infradead.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-media@vger.kernel.org \
    --cc=linux-mediatek@lists.infradead.org \
    --cc=matthias.bgg@gmail.com \
    --cc=mchehab@kernel.org \
    --cc=minghsiu.tsai@mediatek.com \
    --cc=wenst@chromium.org \
    --cc=yong.wu@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).