All of lore.kernel.org
 help / color / mirror / Atom feed
From: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
To: Jacopo Mondi <jacopo@jmondi.org>
Cc: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>,
	dri-devel@lists.freedesktop.org,
	linux-renesas-soc@vger.kernel.org,
	Kieran Bingham <kieran.bingham@ideasonboard.com>,
	Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Subject: Re: [PATCH 5/9] drm: rcar-du: Use DRM-managed allocation for encoders
Date: Mon, 14 Dec 2020 15:28:56 +0200	[thread overview]
Message-ID: <X9domDvwBxTw5npR@pendragon.ideasonboard.com> (raw)
In-Reply-To: <20201214103750.bpwmxe4icjtika4v@uno.localdomain>

Hi Jacopo,

On Mon, Dec 14, 2020 at 11:37:50AM +0100, Jacopo Mondi wrote:
> On Sat, Dec 05, 2020 at 12:01:35AM +0200, Laurent Pinchart wrote:
> > devm_kzalloc() is the wrong API to allocate encoders, as the lifetime of
> > the encoders is tied to the DRM device, not the device to driver
> > binding. drmm_kzalloc() isn't a good option either, as it would result
> > in the encoder being freed before being unregistered during the managed
> > cleanup of the DRM objects. Use a plain kzalloc(), and register a drmm
> > action to cleanup the encoder.
> >
> > Signed-off-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
> > ---
> >  drivers/gpu/drm/rcar-du/rcar_du_encoder.c | 47 ++++++++++++++---------
> >  1 file changed, 29 insertions(+), 18 deletions(-)
> >
> > diff --git a/drivers/gpu/drm/rcar-du/rcar_du_encoder.c b/drivers/gpu/drm/rcar-du/rcar_du_encoder.c
> > index 0edce24f2053..7c491ff72cd2 100644
> > --- a/drivers/gpu/drm/rcar-du/rcar_du_encoder.c
> > +++ b/drivers/gpu/drm/rcar-du/rcar_du_encoder.c
> > @@ -8,12 +8,13 @@
> >   */
> >
> >  #include <linux/export.h>
> > +#include <linux/slab.h>
> >
> >  #include <drm/drm_bridge.h>
> >  #include <drm/drm_crtc.h>
> > +#include <drm/drm_managed.h>
> >  #include <drm/drm_modeset_helper_vtables.h>
> >  #include <drm/drm_panel.h>
> > -#include <drm/drm_simple_kms_helper.h>
> >
> >  #include "rcar_du_drv.h"
> >  #include "rcar_du_encoder.h"
> > @@ -44,6 +45,17 @@ static unsigned int rcar_du_encoder_count_ports(struct device_node *node)
> >  	return num_ports;
> >  }
> >
> > +static const struct drm_encoder_funcs rcar_du_encoder_funcs = {
> > +};
> > +
> > +static void rcar_du_encoder_release(struct drm_device *dev, void *res)
> > +{
> > +	struct rcar_du_encoder *renc = res;
> > +
> > +	drm_encoder_cleanup(&renc->base);
> > +	kfree(renc);
> > +}
> > +
> >  int rcar_du_encoder_init(struct rcar_du_device *rcdu,
> >  			 enum rcar_du_output output,
> >  			 struct device_node *enc_node)
> > @@ -53,7 +65,7 @@ int rcar_du_encoder_init(struct rcar_du_device *rcdu,
> >  	struct drm_bridge *bridge;
> >  	int ret;
> >
> > -	renc = devm_kzalloc(rcdu->dev, sizeof(*renc), GFP_KERNEL);
> > +	renc = kzalloc(sizeof(*renc), GFP_KERNEL);
> >  	if (renc == NULL)
> >  		return -ENOMEM;
> >
> > @@ -76,20 +88,20 @@ int rcar_du_encoder_init(struct rcar_du_device *rcdu,
> >
> >  		if (IS_ERR(panel)) {
> >  			ret = PTR_ERR(panel);
> > -			goto done;
> > +			goto error;
> >  		}
> >
> >  		bridge = devm_drm_panel_bridge_add_typed(rcdu->dev, panel,
> >  							 DRM_MODE_CONNECTOR_DPI);
> >  		if (IS_ERR(bridge)) {
> >  			ret = PTR_ERR(bridge);
> > -			goto done;
> > +			goto error;
> >  		}
> >  	} else {
> >  		bridge = of_drm_find_bridge(enc_node);
> >  		if (!bridge) {
> >  			ret = -EPROBE_DEFER;
> > -			goto done;
> > +			goto error;
> >  		}
> >
> >  		if (output == RCAR_DU_OUTPUT_LVDS0 ||
> > @@ -104,28 +116,27 @@ int rcar_du_encoder_init(struct rcar_du_device *rcdu,
> >  	if (rcdu->info->gen >= 3 && output == RCAR_DU_OUTPUT_LVDS1) {
> >  		if (rcar_lvds_dual_link(bridge)) {
> >  			ret = -ENOLINK;
> > -			goto done;
> > +			goto error;
> >  		}
> >  	}
> >
> > -	ret = drm_simple_encoder_init(rcdu->ddev, encoder,
> > -				      DRM_MODE_ENCODER_NONE);
> > +	ret = drm_encoder_init(rcdu->ddev, encoder, &rcar_du_encoder_funcs,
> > +			       DRM_MODE_ENCODER_NONE, NULL);
> 
> I might have missed the reason for this.
> Why add an action later instead of making rcar_du_encoder_release part
> of rcar_du_encoder_funcs ?

Because the subsystem is moving towards drmm_* instead of manual release
callbacks (I don't mind the manual release callbacks personally). Note
that there's "[PATCH v4 02/19] drm: add drmm_encoder_alloc()" that
should be merged soon, which will allow simplifying this.

> >  	if (ret < 0)
> > -		goto done;
> > +		goto error;
> > +
> > +	ret = drmm_add_action_or_reset(rcdu->ddev, rcar_du_encoder_release,
> > +				       renc);
> > +	if (ret)
> > +		return ret;
> >
> >  	/*
> >  	 * Attach the bridge to the encoder. The bridge will create the
> >  	 * connector.
> >  	 */
> > -	ret = drm_bridge_attach(encoder, bridge, NULL, 0);
> > -	if (ret) {
> > -		drm_encoder_cleanup(encoder);
> > -		return ret;
> > -	}
> > -
> > -done:
> > -	if (ret < 0)
> > -		devm_kfree(rcdu->dev, renc);
> > +	return drm_bridge_attach(encoder, bridge, NULL, 0);
> >
> > +error:
> > +	kfree(renc);
> >  	return ret;
> >  }

-- 
Regards,

Laurent Pinchart

WARNING: multiple messages have this Message-ID (diff)
From: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
To: Jacopo Mondi <jacopo@jmondi.org>
Cc: linux-renesas-soc@vger.kernel.org,
	Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>,
	Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>,
	Kieran Bingham <kieran.bingham@ideasonboard.com>,
	dri-devel@lists.freedesktop.org
Subject: Re: [PATCH 5/9] drm: rcar-du: Use DRM-managed allocation for encoders
Date: Mon, 14 Dec 2020 15:28:56 +0200	[thread overview]
Message-ID: <X9domDvwBxTw5npR@pendragon.ideasonboard.com> (raw)
In-Reply-To: <20201214103750.bpwmxe4icjtika4v@uno.localdomain>

Hi Jacopo,

On Mon, Dec 14, 2020 at 11:37:50AM +0100, Jacopo Mondi wrote:
> On Sat, Dec 05, 2020 at 12:01:35AM +0200, Laurent Pinchart wrote:
> > devm_kzalloc() is the wrong API to allocate encoders, as the lifetime of
> > the encoders is tied to the DRM device, not the device to driver
> > binding. drmm_kzalloc() isn't a good option either, as it would result
> > in the encoder being freed before being unregistered during the managed
> > cleanup of the DRM objects. Use a plain kzalloc(), and register a drmm
> > action to cleanup the encoder.
> >
> > Signed-off-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
> > ---
> >  drivers/gpu/drm/rcar-du/rcar_du_encoder.c | 47 ++++++++++++++---------
> >  1 file changed, 29 insertions(+), 18 deletions(-)
> >
> > diff --git a/drivers/gpu/drm/rcar-du/rcar_du_encoder.c b/drivers/gpu/drm/rcar-du/rcar_du_encoder.c
> > index 0edce24f2053..7c491ff72cd2 100644
> > --- a/drivers/gpu/drm/rcar-du/rcar_du_encoder.c
> > +++ b/drivers/gpu/drm/rcar-du/rcar_du_encoder.c
> > @@ -8,12 +8,13 @@
> >   */
> >
> >  #include <linux/export.h>
> > +#include <linux/slab.h>
> >
> >  #include <drm/drm_bridge.h>
> >  #include <drm/drm_crtc.h>
> > +#include <drm/drm_managed.h>
> >  #include <drm/drm_modeset_helper_vtables.h>
> >  #include <drm/drm_panel.h>
> > -#include <drm/drm_simple_kms_helper.h>
> >
> >  #include "rcar_du_drv.h"
> >  #include "rcar_du_encoder.h"
> > @@ -44,6 +45,17 @@ static unsigned int rcar_du_encoder_count_ports(struct device_node *node)
> >  	return num_ports;
> >  }
> >
> > +static const struct drm_encoder_funcs rcar_du_encoder_funcs = {
> > +};
> > +
> > +static void rcar_du_encoder_release(struct drm_device *dev, void *res)
> > +{
> > +	struct rcar_du_encoder *renc = res;
> > +
> > +	drm_encoder_cleanup(&renc->base);
> > +	kfree(renc);
> > +}
> > +
> >  int rcar_du_encoder_init(struct rcar_du_device *rcdu,
> >  			 enum rcar_du_output output,
> >  			 struct device_node *enc_node)
> > @@ -53,7 +65,7 @@ int rcar_du_encoder_init(struct rcar_du_device *rcdu,
> >  	struct drm_bridge *bridge;
> >  	int ret;
> >
> > -	renc = devm_kzalloc(rcdu->dev, sizeof(*renc), GFP_KERNEL);
> > +	renc = kzalloc(sizeof(*renc), GFP_KERNEL);
> >  	if (renc == NULL)
> >  		return -ENOMEM;
> >
> > @@ -76,20 +88,20 @@ int rcar_du_encoder_init(struct rcar_du_device *rcdu,
> >
> >  		if (IS_ERR(panel)) {
> >  			ret = PTR_ERR(panel);
> > -			goto done;
> > +			goto error;
> >  		}
> >
> >  		bridge = devm_drm_panel_bridge_add_typed(rcdu->dev, panel,
> >  							 DRM_MODE_CONNECTOR_DPI);
> >  		if (IS_ERR(bridge)) {
> >  			ret = PTR_ERR(bridge);
> > -			goto done;
> > +			goto error;
> >  		}
> >  	} else {
> >  		bridge = of_drm_find_bridge(enc_node);
> >  		if (!bridge) {
> >  			ret = -EPROBE_DEFER;
> > -			goto done;
> > +			goto error;
> >  		}
> >
> >  		if (output == RCAR_DU_OUTPUT_LVDS0 ||
> > @@ -104,28 +116,27 @@ int rcar_du_encoder_init(struct rcar_du_device *rcdu,
> >  	if (rcdu->info->gen >= 3 && output == RCAR_DU_OUTPUT_LVDS1) {
> >  		if (rcar_lvds_dual_link(bridge)) {
> >  			ret = -ENOLINK;
> > -			goto done;
> > +			goto error;
> >  		}
> >  	}
> >
> > -	ret = drm_simple_encoder_init(rcdu->ddev, encoder,
> > -				      DRM_MODE_ENCODER_NONE);
> > +	ret = drm_encoder_init(rcdu->ddev, encoder, &rcar_du_encoder_funcs,
> > +			       DRM_MODE_ENCODER_NONE, NULL);
> 
> I might have missed the reason for this.
> Why add an action later instead of making rcar_du_encoder_release part
> of rcar_du_encoder_funcs ?

Because the subsystem is moving towards drmm_* instead of manual release
callbacks (I don't mind the manual release callbacks personally). Note
that there's "[PATCH v4 02/19] drm: add drmm_encoder_alloc()" that
should be merged soon, which will allow simplifying this.

> >  	if (ret < 0)
> > -		goto done;
> > +		goto error;
> > +
> > +	ret = drmm_add_action_or_reset(rcdu->ddev, rcar_du_encoder_release,
> > +				       renc);
> > +	if (ret)
> > +		return ret;
> >
> >  	/*
> >  	 * Attach the bridge to the encoder. The bridge will create the
> >  	 * connector.
> >  	 */
> > -	ret = drm_bridge_attach(encoder, bridge, NULL, 0);
> > -	if (ret) {
> > -		drm_encoder_cleanup(encoder);
> > -		return ret;
> > -	}
> > -
> > -done:
> > -	if (ret < 0)
> > -		devm_kfree(rcdu->dev, renc);
> > +	return drm_bridge_attach(encoder, bridge, NULL, 0);
> >
> > +error:
> > +	kfree(renc);
> >  	return ret;
> >  }

-- 
Regards,

Laurent Pinchart
_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel

  reply	other threads:[~2020-12-14 13:30 UTC|newest]

Thread overview: 66+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-12-04 22:01 [PATCH 0/9] drm: rcar-du: Fix LVDS-related crash Laurent Pinchart
2020-12-04 22:01 ` Laurent Pinchart
2020-12-04 22:01 ` [PATCH 1/9] drm: rcar-du: Fix crash when using LVDS1 clock for CRTC Laurent Pinchart
2020-12-04 22:01   ` Laurent Pinchart
2020-12-07  8:15   ` Geert Uytterhoeven
2020-12-07  8:15     ` Geert Uytterhoeven
2020-12-08  0:42     ` Laurent Pinchart
2020-12-08  0:42       ` Laurent Pinchart
2020-12-14  9:43   ` Jacopo Mondi
2020-12-14  9:43     ` Jacopo Mondi
2020-12-14 15:36   ` Kieran Bingham
2020-12-14 15:36     ` Kieran Bingham
2020-12-04 22:01 ` [PATCH 2/9] drm: rcar-du: Release vsp device reference in all error paths Laurent Pinchart
2020-12-04 22:01   ` Laurent Pinchart
2020-12-14 10:03   ` Jacopo Mondi
2020-12-14 10:03     ` Jacopo Mondi
2020-12-14 15:42   ` Kieran Bingham
2020-12-14 15:42     ` Kieran Bingham
2020-12-04 22:01 ` [PATCH 3/9] drm: rcar-du: Drop unneeded encoder cleanup in error path Laurent Pinchart
2020-12-04 22:01   ` Laurent Pinchart
2020-12-14 10:11   ` Jacopo Mondi
2020-12-14 10:11     ` Jacopo Mondi
2020-12-14 13:19     ` Laurent Pinchart
2020-12-14 13:19       ` Laurent Pinchart
2020-12-14 15:59   ` Kieran Bingham
2020-12-14 15:59     ` Kieran Bingham
2020-12-04 22:01 ` [PATCH 4/9] drm: rcar-du: Use DRM-managed allocation for VSP planes Laurent Pinchart
2020-12-04 22:01   ` Laurent Pinchart
2020-12-14 16:20   ` Kieran Bingham
2020-12-14 16:20     ` Kieran Bingham
2020-12-14 16:26     ` Laurent Pinchart
2020-12-14 16:26       ` Laurent Pinchart
2020-12-15 15:42       ` Geert Uytterhoeven
2020-12-15 15:42         ` Geert Uytterhoeven
2020-12-04 22:01 ` [PATCH 5/9] drm: rcar-du: Use DRM-managed allocation for encoders Laurent Pinchart
2020-12-04 22:01   ` Laurent Pinchart
2020-12-14 10:37   ` Jacopo Mondi
2020-12-14 10:37     ` Jacopo Mondi
2020-12-14 13:28     ` Laurent Pinchart [this message]
2020-12-14 13:28       ` Laurent Pinchart
2020-12-04 22:01 ` [PATCH 6/9] drm: rcar-du: Embed drm_device in rcar_du_device Laurent Pinchart
2020-12-04 22:01   ` Laurent Pinchart
2020-12-14 10:55   ` Jacopo Mondi
2020-12-14 10:55     ` Jacopo Mondi
2020-12-14 17:02   ` Kieran Bingham
2020-12-14 17:02     ` Kieran Bingham
2020-12-04 22:01 ` [PATCH 7/9] drm: rcar-du: Replace dev_private with container_of Laurent Pinchart
2020-12-04 22:01   ` Laurent Pinchart
2020-12-14 10:58   ` Jacopo Mondi
2020-12-14 10:58     ` Jacopo Mondi
2020-12-14 17:11     ` Kieran Bingham
2020-12-14 17:11       ` Kieran Bingham
2020-12-04 22:01 ` [PATCH 8/9] drm: rcar-du: Skip encoder allocation for LVDS1 in dual-link mode Laurent Pinchart
2020-12-04 22:01   ` Laurent Pinchart
2020-12-14 11:04   ` Jacopo Mondi
2020-12-14 11:04     ` Jacopo Mondi
2020-12-14 13:32     ` Laurent Pinchart
2020-12-14 13:32       ` Laurent Pinchart
2020-12-14 17:21   ` Kieran Bingham
2020-12-14 17:21     ` Kieran Bingham
2020-12-04 22:01 ` [PATCH 9/9] drm: rcar-du: Drop local encoder variable Laurent Pinchart
2020-12-04 22:01   ` Laurent Pinchart
2020-12-14 11:05   ` Jacopo Mondi
2020-12-14 11:05     ` Jacopo Mondi
2020-12-14 17:16   ` Kieran Bingham
2020-12-14 17:16     ` Kieran Bingham

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=X9domDvwBxTw5npR@pendragon.ideasonboard.com \
    --to=laurent.pinchart@ideasonboard.com \
    --cc=dri-devel@lists.freedesktop.org \
    --cc=jacopo@jmondi.org \
    --cc=kieran.bingham@ideasonboard.com \
    --cc=kuninori.morimoto.gx@renesas.com \
    --cc=laurent.pinchart+renesas@ideasonboard.com \
    --cc=linux-renesas-soc@vger.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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.