* Re: [PATCH v10 2/9] drm/mediatek: Add crtc path enum for all_drm_priv array
[not found] ` <20230927153833.23583-3-jason-jh.lin@mediatek.com>
@ 2023-09-28 9:21 ` AngeloGioacchino Del Regno
0 siblings, 0 replies; 10+ messages in thread
From: AngeloGioacchino Del Regno @ 2023-09-28 9:21 UTC (permalink / raw)
To: Jason-JH.Lin, Chun-Kuang Hu
Cc: Matthias Brugger, Jason-ch Chen, Johnson Wang, Singo Chang,
Nancy Lin, Shawn Sung, dri-devel, linux-mediatek,
linux-arm-kernel, linux-kernel,
Project_Global_Chrome_Upstream_Group
Il 27/09/23 17:38, Jason-JH.Lin ha scritto:
> Add mtk_drm_crtc_path enum for each display path.
>
> Instead of using array index of all_drm_priv in mtk_drm_kms_init(),
> mtk_drm_crtc_path enum can make code more readable.
>
> Signed-off-by: Jason-JH.Lin <jason-jh.lin@mediatek.com>
> Reviewed-by: Fei Shao <fshao@chromium.org>
> Reviewed-by: CK Hu <ck.hu@mediatek.com>
> Tested-by: Fei Shao <fshao@chromium.org>
Reviewed-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com>
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: [PATCH v10 7/9] drm/mediatek: Support dynamic selection of DP_INTF0 on MT8188 VDOSYS0
[not found] ` <20230927153833.23583-8-jason-jh.lin@mediatek.com>
@ 2023-09-28 9:24 ` AngeloGioacchino Del Regno
0 siblings, 0 replies; 10+ messages in thread
From: AngeloGioacchino Del Regno @ 2023-09-28 9:24 UTC (permalink / raw)
To: Jason-JH.Lin, Chun-Kuang Hu
Cc: Matthias Brugger, Jason-ch Chen, Johnson Wang, Singo Chang,
Nancy Lin, Shawn Sung, dri-devel, linux-mediatek,
linux-arm-kernel, linux-kernel,
Project_Global_Chrome_Upstream_Group
Il 27/09/23 17:38, Jason-JH.Lin ha scritto:
> Move DDP_COMPONENT_DP_INTF0 from mt8188_mtk_ddp_main array to a
> connector routes array called mt8188_mtk_ddp_main_routes to support
> dynamic selection capability for mt8188.
>
> Signed-off-by: Jason-JH.Lin <jason-jh.lin@mediatek.com>
Reviewed-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com>
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: [PATCH v10 4/9] drm/mediatek: Add encoder_index interface for mtk_ddp_comp_funcs
[not found] ` <20230927153833.23583-5-jason-jh.lin@mediatek.com>
@ 2023-09-28 9:24 ` AngeloGioacchino Del Regno
0 siblings, 0 replies; 10+ messages in thread
From: AngeloGioacchino Del Regno @ 2023-09-28 9:24 UTC (permalink / raw)
To: Jason-JH.Lin, Chun-Kuang Hu
Cc: Matthias Brugger, Jason-ch Chen, Johnson Wang, Singo Chang,
Nancy Lin, Shawn Sung, dri-devel, linux-mediatek,
linux-arm-kernel, linux-kernel,
Project_Global_Chrome_Upstream_Group
Il 27/09/23 17:38, Jason-JH.Lin ha scritto:
> To support dynamic connector selection function, each ddp_comp need to
> get their encoder_index to identify which connector should be selected.
>
> Add encoder_index interface for mtk_ddp_comp_funcs to get the encoder
> identifier by drm_encoder_index().
> Then drm driver will call mtk_ddp_comp_encoder_index_set() to store the
> encoder_index to each ddp_comp in connector routes.
>
> Signed-off-by: Jason-JH.Lin <jason-jh.lin@mediatek.com>
Reviewed-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com>
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: [PATCH v10 8/9] drm/mediatek: dsi: Support dynamic connector selection
[not found] ` <20230927153833.23583-9-jason-jh.lin@mediatek.com>
@ 2023-09-28 9:24 ` AngeloGioacchino Del Regno
0 siblings, 0 replies; 10+ messages in thread
From: AngeloGioacchino Del Regno @ 2023-09-28 9:24 UTC (permalink / raw)
To: Jason-JH.Lin, Chun-Kuang Hu
Cc: Matthias Brugger, Jason-ch Chen, Johnson Wang, Singo Chang,
Nancy Lin, Shawn Sung, dri-devel, linux-mediatek,
linux-arm-kernel, linux-kernel,
Project_Global_Chrome_Upstream_Group
Il 27/09/23 17:38, Jason-JH.Lin ha scritto:
> Add implementation of mtk_dsi_encoder_index to mtk_ddp_comp_func
> to make mtk_dsi support dynamic connector selection.
>
> Signed-off-by: Jason-JH.Lin <jason-jh.lin@mediatek.com>
> Reviewed-by: CK Hu <ck.hu@mediatek.com>
> Reviewed-by: Fei Shao <fshao@chromium.org>
> Tested-by: Fei Shao <fshao@chromium.org>
Reviewed-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com>
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: [PATCH v10 6/9] drm/mediatek: dpi: Support dynamic connector selection
[not found] ` <20230927153833.23583-7-jason-jh.lin@mediatek.com>
@ 2023-09-28 9:24 ` AngeloGioacchino Del Regno
0 siblings, 0 replies; 10+ messages in thread
From: AngeloGioacchino Del Regno @ 2023-09-28 9:24 UTC (permalink / raw)
To: Jason-JH.Lin, Chun-Kuang Hu
Cc: Matthias Brugger, Jason-ch Chen, Johnson Wang, Singo Chang,
Nancy Lin, Shawn Sung, dri-devel, linux-mediatek,
linux-arm-kernel, linux-kernel,
Project_Global_Chrome_Upstream_Group
Il 27/09/23 17:38, Jason-JH.Lin ha scritto:
> Add implementation of mtk_dpi_encoder_index to mtk_ddp_comp_func
> to make mtk_dpi support dynamic connector selection.
>
> Signed-off-by: Jason-JH.Lin <jason-jh.lin@mediatek.com>
Reviewed-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com>
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: [PATCH v10 9/9] drm/mediatek: Support dynamic selection of DSI0 on MT8188 VDOSYS0
[not found] ` <20230927153833.23583-10-jason-jh.lin@mediatek.com>
@ 2023-09-28 9:24 ` AngeloGioacchino Del Regno
0 siblings, 0 replies; 10+ messages in thread
From: AngeloGioacchino Del Regno @ 2023-09-28 9:24 UTC (permalink / raw)
To: Jason-JH.Lin, Chun-Kuang Hu
Cc: Matthias Brugger, Jason-ch Chen, Johnson Wang, Singo Chang,
Nancy Lin, Shawn Sung, dri-devel, linux-mediatek,
linux-arm-kernel, linux-kernel,
Project_Global_Chrome_Upstream_Group, Nathan Lu
Il 27/09/23 17:38, Jason-JH.Lin ha scritto:
> Add DDP_COMPONENT_DSI0 as a main display output selection on
> MT8188 VDOSYS0.
>
> Signed-off-by: Nathan Lu <nathan.lu@mediatek.com>
> Signed-off-by: Jason-JH.Lin <jason-jh.lin@mediatek.com>
> Reviewed-by: Matthias Brugger <matthias.bgg@gmail.com>
> Reviewed-by: Fei Shao <fshao@chromium.org>
> Tested-by: Fei Shao <fshao@chromium.org>
Reviewed-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com>
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: [PATCH v10 5/9] drm/mediatek: Add connector dynamic selection capability
[not found] ` <20230927153833.23583-6-jason-jh.lin@mediatek.com>
@ 2023-09-28 10:00 ` AngeloGioacchino Del Regno
2023-10-02 7:14 ` Jason-JH Lin (林睿祥)
0 siblings, 1 reply; 10+ messages in thread
From: AngeloGioacchino Del Regno @ 2023-09-28 10:00 UTC (permalink / raw)
To: Jason-JH.Lin, Chun-Kuang Hu
Cc: Matthias Brugger, Jason-ch Chen, Johnson Wang, Singo Chang,
Nancy Lin, Shawn Sung, dri-devel, linux-mediatek,
linux-arm-kernel, linux-kernel,
Project_Global_Chrome_Upstream_Group, Nathan Lu
Il 27/09/23 17:38, Jason-JH.Lin ha scritto:
> Add dynamic select available connector flow in mtk_drm_crtc_create()
> and mtk_drm_crtc_atomic_enable().
>
> In mtk_drm_crtc_create(), if there is a connector routes array in drm
> driver data, all components definded in the connector routes array will
> be checked and their encoder_index will be set.
>
> In mtk_drm_crtc_atomic_enable(), crtc will check its encoder_index to
> identify which componet in the connector routes array should append.
>
> Signed-off-by: Nancy Lin <nancy.lin@mediatek.com>
> Signed-off-by: Nathan Lu <nathan.lu@mediatek.com>
> Signed-off-by: Jason-JH.Lin <jason-jh.lin@mediatek.com>
> Tested-by: Fei Shao <fshao@chromium.org>
> ---
> drivers/gpu/drm/mediatek/mtk_drm_crtc.c | 78 ++++++++++++++++++++-
> drivers/gpu/drm/mediatek/mtk_drm_crtc.h | 5 +-
> drivers/gpu/drm/mediatek/mtk_drm_ddp_comp.c | 27 ++++++-
> drivers/gpu/drm/mediatek/mtk_drm_drv.c | 13 +++-
> drivers/gpu/drm/mediatek/mtk_drm_drv.h | 7 ++
> 5 files changed, 123 insertions(+), 7 deletions(-)
>
> diff --git a/drivers/gpu/drm/mediatek/mtk_drm_crtc.c b/drivers/gpu/drm/mediatek/mtk_drm_crtc.c
> index b6fa4ad2f94d..8eb4d2646a76 100644
> --- a/drivers/gpu/drm/mediatek/mtk_drm_crtc.c
> +++ b/drivers/gpu/drm/mediatek/mtk_drm_crtc.c
> @@ -63,6 +63,8 @@ struct mtk_drm_crtc {
> struct mtk_mutex *mutex;
> unsigned int ddp_comp_nr;
> struct mtk_ddp_comp **ddp_comp;
> + unsigned int num_conn_routes;
> + const struct mtk_drm_route *conn_routes;
>
> /* lock for display hardware access */
> struct mutex hw_lock;
> @@ -647,6 +649,45 @@ static void mtk_drm_crtc_disable_vblank(struct drm_crtc *crtc)
> mtk_ddp_comp_disable_vblank(comp);
> }
>
> +static void mtk_drm_crtc_update_output(struct drm_crtc *crtc,
> + struct drm_atomic_state *state)
> +{
> + int crtc_index = drm_crtc_index(crtc);
> + int i;
> + struct device *dev;
> + struct drm_crtc_state *crtc_state = state->crtcs[crtc_index].new_state;
> + struct mtk_drm_crtc *mtk_crtc = to_mtk_crtc(crtc);
> + struct mtk_drm_private *priv = crtc->dev->dev_private;
> + unsigned int comp_id;
You're not using comp_id globally in this function....
> + unsigned int encoder_mask = crtc_state->encoder_mask;
> +
> + if (!crtc_state->connectors_changed)
> + return;
> +
> + if (!mtk_crtc->num_conn_routes)
> + return;
> +
> + priv = priv->all_drm_private[crtc_index];
This is a bit confusing. You're reassigning priv: please avoid that.
I would prefer directly seeing the final assignment, or otherwise two pointers.
struct mtk_drm_private *priv;
if ....
...........
priv = crtc->dev->dev_private->all_drm_private[crtc_index];
> + dev = priv->dev;
> +
> + dev_dbg(dev, "connector change:%d, encoder mask:0x%x for crtc:%d\n",
> + crtc_state->connectors_changed, encoder_mask, crtc_index);
> +
> + for (i = 0; i < mtk_crtc->num_conn_routes; i++) {
> + struct mtk_ddp_comp *comp;
...so you can move it here...
unsigned int comp_id = mtk_crtc->conn_routes[i].route_ddp;
struct mtk_ddp_comp *comp = &priv->ddp_comp[comp_id];
> +
> + comp_id = mtk_crtc->conn_routes[i].route_ddp;
> + comp = &priv->ddp_comp[comp_id];
> + if (comp->encoder_index >= 0 &&
> + encoder_mask & BIT(comp->encoder_index)) {
For readability, I would prefer to see
if (comp->encoder_index >= 0 &&
(encoder_mask & BIT(comp->encoder_index))) {
...but I don't have strong opinions on that.
> + mtk_crtc->ddp_comp[mtk_crtc->ddp_comp_nr - 1] = comp;
> + dev_dbg(dev, "Add comp_id: %d at path index %d\n",
> + comp->id, mtk_crtc->ddp_comp_nr - 1);
> + break;
> + }
> + }
> +}
> +
> int mtk_drm_crtc_plane_check(struct drm_crtc *crtc, struct drm_plane *plane,
> struct mtk_plane_state *state)
> {
> @@ -679,6 +720,8 @@ static void mtk_drm_crtc_atomic_enable(struct drm_crtc *crtc,
>
> DRM_DEBUG_DRIVER("%s %d\n", __func__, crtc->base.id);
>
> + mtk_drm_crtc_update_output(crtc, state);
> +
What's the point of updating the output before pm_runtime_resume_and_get()?
If PM resume fails we're not enabling the CRTC so the update is actually useless.
Please move this after PM resume: that will also possibly come handy in the future.
> ret = pm_runtime_resume_and_get(comp->dev);
> if (ret < 0) {
> DRM_DEV_ERROR(comp->dev, "Failed to enable power domain: %d\n", ret);
> @@ -884,7 +927,8 @@ struct device *mtk_drm_crtc_dma_dev_get(struct drm_crtc *crtc)
>
> int mtk_drm_crtc_create(struct drm_device *drm_dev,
> const unsigned int *path, unsigned int path_len,
> - int priv_data_index)
> + int priv_data_index, const struct mtk_drm_route *conn_routes,
> + unsigned int num_conn_routes)
> {
> struct mtk_drm_private *priv = drm_dev->dev_private;
> struct device *dev = drm_dev->dev;
> @@ -935,7 +979,8 @@ int mtk_drm_crtc_create(struct drm_device *drm_dev,
>
> mtk_crtc->mmsys_dev = priv->mmsys_dev;
> mtk_crtc->ddp_comp_nr = path_len;
> - mtk_crtc->ddp_comp = devm_kmalloc_array(dev, mtk_crtc->ddp_comp_nr,
> + mtk_crtc->ddp_comp = devm_kmalloc_array(dev,
> + mtk_crtc->ddp_comp_nr + (conn_routes ? 1 : 0),
> sizeof(*mtk_crtc->ddp_comp),
> GFP_KERNEL);
> if (!mtk_crtc->ddp_comp)
> @@ -1038,5 +1083,34 @@ int mtk_drm_crtc_create(struct drm_device *drm_dev,
> init_waitqueue_head(&mtk_crtc->cb_blocking_queue);
> }
> #endif
> +
> + if (conn_routes) {
> + struct device_node *node;
> + struct mtk_ddp_comp *comp;
> + unsigned int comp_id;
> +
> + for (i = 0; i < num_conn_routes; i++) {
Same here, you're locally using comp_id, node and comp *only* in the for loop....
unsigned int comp_id = conn_routes[i].route_ddp;
struct device_node *node = priv->comp_node[comp_id];
struct mtk_ddp_comp *comp = &priv->ddp_comp[comp_id];
> + comp_id = conn_routes[i].route_ddp;
> + node = priv->comp_node[comp_id];
> + comp = &priv->ddp_comp[comp_id];
> +
> + if (!comp->dev) {
> + dev_dbg(dev, "comp_id:%d, Component %pOF not initialized\n",
> + comp_id, node);
> + /* mark encoder_index to -1, if route comp device is not enabled */
> + comp->encoder_index = -1;
> + continue;
> + }
> +
> + mtk_ddp_comp_encoder_index_set(&priv->ddp_comp[comp_id]);
> + }
> +
> + mtk_crtc->num_conn_routes = num_conn_routes;
> + mtk_crtc->conn_routes = conn_routes;
> +
> + /* increase ddp_comp_nr at the end of mtk_drm_crtc_create */
> + mtk_crtc->ddp_comp_nr++;
> + }
> +
> return 0;
> }
> diff --git a/drivers/gpu/drm/mediatek/mtk_drm_crtc.h b/drivers/gpu/drm/mediatek/mtk_drm_crtc.h
> index 3e9046993d09..3c224595fa71 100644
> --- a/drivers/gpu/drm/mediatek/mtk_drm_crtc.h
> +++ b/drivers/gpu/drm/mediatek/mtk_drm_crtc.h
> @@ -8,6 +8,7 @@
>
> #include <drm/drm_crtc.h>
> #include "mtk_drm_ddp_comp.h"
> +#include "mtk_drm_drv.h"
> #include "mtk_drm_plane.h"
>
> #define MTK_LUT_SIZE 512
> @@ -18,7 +19,9 @@ void mtk_drm_crtc_commit(struct drm_crtc *crtc);
> int mtk_drm_crtc_create(struct drm_device *drm_dev,
> const unsigned int *path,
> unsigned int path_len,
> - int priv_data_index);
> + int priv_data_index,
> + const struct mtk_drm_route *conn_routes,
> + unsigned int num_conn_routes);
> int mtk_drm_crtc_plane_check(struct drm_crtc *crtc, struct drm_plane *plane,
> struct mtk_plane_state *state);
> void mtk_drm_crtc_async_update(struct drm_crtc *crtc, struct drm_plane *plane,
> diff --git a/drivers/gpu/drm/mediatek/mtk_drm_ddp_comp.c b/drivers/gpu/drm/mediatek/mtk_drm_ddp_comp.c
> index 771f4e173353..4ddb5e561116 100644
> --- a/drivers/gpu/drm/mediatek/mtk_drm_ddp_comp.c
> +++ b/drivers/gpu/drm/mediatek/mtk_drm_ddp_comp.c
> @@ -507,6 +507,23 @@ static bool mtk_drm_find_comp_in_ddp(struct device *dev,
> return false;
> }
>
> +static int mtk_drm_find_comp_in_ddp_conn_path(struct device *dev,
> + const struct mtk_drm_route *routes,
> + unsigned int num_routes,
> + struct mtk_ddp_comp *ddp_comp)
> +{
> + unsigned int i;
> +
> + if (!routes)
> + return -EINVAL;
> +
> + for (i = 0; i < num_routes; i++)
> + if (dev == ddp_comp[routes[i].route_ddp].dev)
> + return BIT(routes[i].crtc_id);
> +
> + return -ENODEV;
> +}
> +
> int mtk_ddp_comp_get_id(struct device_node *node,
> enum mtk_ddp_comp_type comp_type)
> {
> @@ -538,7 +555,15 @@ unsigned int mtk_drm_find_possible_crtc_by_comp(struct drm_device *drm,
> private->data->third_len, private->ddp_comp))
> ret = BIT(2);
> else
> - DRM_INFO("Failed to find comp in ddp table\n");
> + ret = mtk_drm_find_comp_in_ddp_conn_path(dev,
> + private->data->conn_routes,
> + private->data->num_conn_routes,
> + private->ddp_comp);
> +
> + if (ret <= 0) {
> + DRM_INFO("Failed to find comp in ddp table, ret =%d\n", ret);
> + ret = 0;
Why are you returning 0 for error here?!
> + }
>
> return ret;
> }
Regards,
Angelo
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: [PATCH v10 5/9] drm/mediatek: Add connector dynamic selection capability
2023-09-28 10:00 ` [PATCH v10 5/9] drm/mediatek: Add connector dynamic selection capability AngeloGioacchino Del Regno
@ 2023-10-02 7:14 ` Jason-JH Lin (林睿祥)
2023-10-02 10:23 ` AngeloGioacchino Del Regno
0 siblings, 1 reply; 10+ messages in thread
From: Jason-JH Lin (林睿祥) @ 2023-10-02 7:14 UTC (permalink / raw)
To: angelogioacchino.delregno, chunkuang.hu
Cc: linux-mediatek, Singo Chang (張興國),
Johnson Wang (王聖鑫),
Jason-ch Chen (陳建豪),
Shawn Sung (宋孝謙),
linux-kernel, Nancy Lin (林欣螢),
dri-devel, Project_Global_Chrome_Upstream_Group,
Nathan Lu (呂東霖),
linux-arm-kernel, matthias.bgg
Hi Angelo,
Thanks for the reviews.
On Thu, 2023-09-28 at 12:00 +0200, AngeloGioacchino Del Regno wrote:
> Il 27/09/23 17:38, Jason-JH.Lin ha scritto:
> > Add dynamic select available connector flow in
> > mtk_drm_crtc_create()
> > and mtk_drm_crtc_atomic_enable().
> >
> > In mtk_drm_crtc_create(), if there is a connector routes array in
> > drm
> > driver data, all components definded in the connector routes array
> > will
> > be checked and their encoder_index will be set.
> >
> > In mtk_drm_crtc_atomic_enable(), crtc will check its encoder_index
> > to
> > identify which componet in the connector routes array should
> > append.
> >
> > Signed-off-by: Nancy Lin <nancy.lin@mediatek.com>
> > Signed-off-by: Nathan Lu <nathan.lu@mediatek.com>
> > Signed-off-by: Jason-JH.Lin <jason-jh.lin@mediatek.com>
> > Tested-by: Fei Shao <fshao@chromium.org>
> > ---
> > drivers/gpu/drm/mediatek/mtk_drm_crtc.c | 78
> > ++++++++++++++++++++-
> > drivers/gpu/drm/mediatek/mtk_drm_crtc.h | 5 +-
> > drivers/gpu/drm/mediatek/mtk_drm_ddp_comp.c | 27 ++++++-
> > drivers/gpu/drm/mediatek/mtk_drm_drv.c | 13 +++-
> > drivers/gpu/drm/mediatek/mtk_drm_drv.h | 7 ++
> > 5 files changed, 123 insertions(+), 7 deletions(-)
> >
> > diff --git a/drivers/gpu/drm/mediatek/mtk_drm_crtc.c
> > b/drivers/gpu/drm/mediatek/mtk_drm_crtc.c
> > index b6fa4ad2f94d..8eb4d2646a76 100644
> > --- a/drivers/gpu/drm/mediatek/mtk_drm_crtc.c
> > +++ b/drivers/gpu/drm/mediatek/mtk_drm_crtc.c
> > @@ -63,6 +63,8 @@ struct mtk_drm_crtc {
> > struct mtk_mutex *mutex;
> > unsigned int ddp_comp_nr;
> > struct mtk_ddp_comp **ddp_comp;
> > + unsigned int num_conn_routes;
> > + const struct mtk_drm_route *conn_routes;
> >
> > /* lock for display hardware access */
> > struct mutex hw_lock;
> > @@ -647,6 +649,45 @@ static void mtk_drm_crtc_disable_vblank(struct
> > drm_crtc *crtc)
> > mtk_ddp_comp_disable_vblank(comp);
> > }
> >
> > +static void mtk_drm_crtc_update_output(struct drm_crtc *crtc,
> > + struct drm_atomic_state *state)
> > +{
> > + int crtc_index = drm_crtc_index(crtc);
> > + int i;
> > + struct device *dev;
> > + struct drm_crtc_state *crtc_state = state-
> > >crtcs[crtc_index].new_state;
> > + struct mtk_drm_crtc *mtk_crtc = to_mtk_crtc(crtc);
> > + struct mtk_drm_private *priv = crtc->dev->dev_private;
> > + unsigned int comp_id;
>
> You're not using comp_id globally in this function....
OK, I'll move it to the for loop below.
>
> > + unsigned int encoder_mask = crtc_state->encoder_mask;
> > +
> > + if (!crtc_state->connectors_changed)
> > + return;
> > +
> > + if (!mtk_crtc->num_conn_routes)
> > + return;
> > +
> > + priv = priv->all_drm_private[crtc_index];
>
> This is a bit confusing. You're reassigning priv: please avoid that.
> I would prefer directly seeing the final assignment, or otherwise two
> pointers.
>
> struct mtk_drm_private *priv;
>
> if ....
> ...........
>
> priv = crtc->dev->dev_private->all_drm_private[crtc_index];
>
OK, because the type of dev_private is void*. I'll change the
assignment to:
priv = ((struct mtk_drm_private *)crtc->dev->dev_private)-
>all_drm_private[crtc_index];
> > + dev = priv->dev;
> > +
> > + dev_dbg(dev, "connector change:%d, encoder mask:0x%x for
> > crtc:%d\n",
> > + crtc_state->connectors_changed, encoder_mask,
> > crtc_index);
> > +
> > + for (i = 0; i < mtk_crtc->num_conn_routes; i++) {
> > + struct mtk_ddp_comp *comp;
>
> ...so you can move it here...
OK, I'll move it here.
> unsigned int comp_id = mtk_crtc-
> >conn_routes[i].route_ddp;
> struct mtk_ddp_comp *comp = &priv->ddp_comp[comp_id];
>
> > +
> > + comp_id = mtk_crtc->conn_routes[i].route_ddp;
> > + comp = &priv->ddp_comp[comp_id];
> > + if (comp->encoder_index >= 0 &&
> > + encoder_mask & BIT(comp->encoder_index)) {
>
> For readability, I would prefer to see
>
> if (comp->encoder_index >= 0 &&
> (encoder_mask & BIT(comp->encoder_index))) {
>
> ...but I don't have strong opinions on that.
>
OK, I'll modify it.
> > + mtk_crtc->ddp_comp[mtk_crtc->ddp_comp_nr - 1] =
> > comp;
> > + dev_dbg(dev, "Add comp_id: %d at path index
> > %d\n",
> > + comp->id, mtk_crtc->ddp_comp_nr - 1);
> > + break;
> > + }
> > + }
> > +}
> > +
> > int mtk_drm_crtc_plane_check(struct drm_crtc *crtc, struct
> > drm_plane *plane,
> > struct mtk_plane_state *state)
> > {
> > @@ -679,6 +720,8 @@ static void mtk_drm_crtc_atomic_enable(struct
> > drm_crtc *crtc,
> >
> > DRM_DEBUG_DRIVER("%s %d\n", __func__, crtc->base.id);
> >
> > + mtk_drm_crtc_update_output(crtc, state);
> > +
>
> What's the point of updating the output before
> pm_runtime_resume_and_get()?
> If PM resume fails we're not enabling the CRTC so the update is
> actually useless.
>
> Please move this after PM resume: that will also possibly come handy
> in the future.
OK, I'll move this after pm_runtime_resume_and_get().
>
>
> > ret = pm_runtime_resume_and_get(comp->dev);
> > if (ret < 0) {
> > DRM_DEV_ERROR(comp->dev, "Failed to enable power
> > domain: %d\n", ret);
> > @@ -884,7 +927,8 @@ struct device *mtk_drm_crtc_dma_dev_get(struct
> > drm_crtc *crtc)
> >
> > int mtk_drm_crtc_create(struct drm_device *drm_dev,
> > const unsigned int *path, unsigned int
> > path_len,
> > - int priv_data_index)
> > + int priv_data_index, const struct mtk_drm_route
> > *conn_routes,
> > + unsigned int num_conn_routes)
> > {
> > struct mtk_drm_private *priv = drm_dev->dev_private;
> > struct device *dev = drm_dev->dev;
> > @@ -935,7 +979,8 @@ int mtk_drm_crtc_create(struct drm_device
> > *drm_dev,
> >
> > mtk_crtc->mmsys_dev = priv->mmsys_dev;
> > mtk_crtc->ddp_comp_nr = path_len;
> > - mtk_crtc->ddp_comp = devm_kmalloc_array(dev, mtk_crtc-
> > >ddp_comp_nr,
> > + mtk_crtc->ddp_comp = devm_kmalloc_array(dev,
> > + mtk_crtc->ddp_comp_nr +
> > (conn_routes ? 1 : 0),
> > sizeof(*mtk_crtc-
> > >ddp_comp),
> > GFP_KERNEL);
> > if (!mtk_crtc->ddp_comp)
> > @@ -1038,5 +1083,34 @@ int mtk_drm_crtc_create(struct drm_device
> > *drm_dev,
> > init_waitqueue_head(&mtk_crtc->cb_blocking_queue);
> > }
> > #endif
> > +
> > + if (conn_routes) {
> > + struct device_node *node;
> > + struct mtk_ddp_comp *comp;
> > + unsigned int comp_id;
> > +
> > + for (i = 0; i < num_conn_routes; i++) {
>
> Same here, you're locally using comp_id, node and comp *only* in the
> for loop....
>
OK I'll move it here.
> unsigned int comp_id =
> conn_routes[i].route_ddp;
> struct device_node *node = priv-
> >comp_node[comp_id];
> struct mtk_ddp_comp *comp = &priv-
> >ddp_comp[comp_id];
>
> > + comp_id = conn_routes[i].route_ddp;
> > + node = priv->comp_node[comp_id];
> > + comp = &priv->ddp_comp[comp_id];
> > +
> > + if (!comp->dev) {
> > + dev_dbg(dev, "comp_id:%d, Component
> > %pOF not initialized\n",
> > + comp_id, node);
> > + /* mark encoder_index to -1, if route
> > comp device is not enabled */
> > + comp->encoder_index = -1;
> > + continue;
> > + }
> > +
> > + mtk_ddp_comp_encoder_index_set(&priv-
> > >ddp_comp[comp_id]);
> > + }
> > +
> > + mtk_crtc->num_conn_routes = num_conn_routes;
> > + mtk_crtc->conn_routes = conn_routes;
> > +
> > + /* increase ddp_comp_nr at the end of
> > mtk_drm_crtc_create */
> > + mtk_crtc->ddp_comp_nr++;
> > + }
> > +
> > return 0;
> > }
> > diff --git a/drivers/gpu/drm/mediatek/mtk_drm_crtc.h
> > b/drivers/gpu/drm/mediatek/mtk_drm_crtc.h
> > index 3e9046993d09..3c224595fa71 100644
> > --- a/drivers/gpu/drm/mediatek/mtk_drm_crtc.h
> > +++ b/drivers/gpu/drm/mediatek/mtk_drm_crtc.h
> > @@ -8,6 +8,7 @@
> >
> > #include <drm/drm_crtc.h>
> > #include "mtk_drm_ddp_comp.h"
> > +#include "mtk_drm_drv.h"
> > #include "mtk_drm_plane.h"
> >
> > #define MTK_LUT_SIZE 512
> > @@ -18,7 +19,9 @@ void mtk_drm_crtc_commit(struct drm_crtc *crtc);
> > int mtk_drm_crtc_create(struct drm_device *drm_dev,
> > const unsigned int *path,
> > unsigned int path_len,
> > - int priv_data_index);
> > + int priv_data_index,
> > + const struct mtk_drm_route *conn_routes,
> > + unsigned int num_conn_routes);
> > int mtk_drm_crtc_plane_check(struct drm_crtc *crtc, struct
> > drm_plane *plane,
> > struct mtk_plane_state *state);
> > void mtk_drm_crtc_async_update(struct drm_crtc *crtc, struct
> > drm_plane *plane,
> > diff --git a/drivers/gpu/drm/mediatek/mtk_drm_ddp_comp.c
> > b/drivers/gpu/drm/mediatek/mtk_drm_ddp_comp.c
> > index 771f4e173353..4ddb5e561116 100644
> > --- a/drivers/gpu/drm/mediatek/mtk_drm_ddp_comp.c
> > +++ b/drivers/gpu/drm/mediatek/mtk_drm_ddp_comp.c
> > @@ -507,6 +507,23 @@ static bool mtk_drm_find_comp_in_ddp(struct
> > device *dev,
> > return false;
> > }
> >
> > +static int mtk_drm_find_comp_in_ddp_conn_path(struct device *dev,
> > + const struct
> > mtk_drm_route *routes,
> > + unsigned int num_routes,
> > + struct mtk_ddp_comp
> > *ddp_comp)
> > +{
> > + unsigned int i;
> > +
> > + if (!routes)
> > + return -EINVAL;
> > +
> > + for (i = 0; i < num_routes; i++)
> > + if (dev == ddp_comp[routes[i].route_ddp].dev)
> > + return BIT(routes[i].crtc_id);
> > +
> > + return -ENODEV;
> > +}
> > +
> > int mtk_ddp_comp_get_id(struct device_node *node,
> > enum mtk_ddp_comp_type comp_type)
> > {
> > @@ -538,7 +555,15 @@ unsigned int
> > mtk_drm_find_possible_crtc_by_comp(struct drm_device *drm,
> > private->data->third_len,
> > private->ddp_comp))
> > ret = BIT(2);
> > else
> > - DRM_INFO("Failed to find comp in ddp table\n");
> > + ret = mtk_drm_find_comp_in_ddp_conn_path(dev,
> > + private->data-
> > >conn_routes,
> > + private->data-
> > >num_conn_routes,
> > + private-
> > >ddp_comp);
> > +
> > + if (ret <= 0) {
> > + DRM_INFO("Failed to find comp in ddp table, ret =%d\n",
> > ret);
> > + ret = 0;
>
> Why are you returning 0 for error here?!
Because this function return 0 means not found any valid crtc bits.
Should I change this?
Regards,
Jason-JH.Lin
>
> > + }
> >
> > return ret;
> > }
>
> Regards,
> Angelo
>
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: [PATCH v10 5/9] drm/mediatek: Add connector dynamic selection capability
2023-10-02 7:14 ` Jason-JH Lin (林睿祥)
@ 2023-10-02 10:23 ` AngeloGioacchino Del Regno
2023-10-03 3:34 ` Jason-JH Lin (林睿祥)
0 siblings, 1 reply; 10+ messages in thread
From: AngeloGioacchino Del Regno @ 2023-10-02 10:23 UTC (permalink / raw)
To: Jason-JH Lin (林睿祥), chunkuang.hu
Cc: linux-mediatek, Singo Chang (張興國),
Johnson Wang (王聖鑫),
Jason-ch Chen (陳建豪),
Shawn Sung (宋孝謙),
linux-kernel, Nancy Lin (林欣螢),
dri-devel, Project_Global_Chrome_Upstream_Group,
Nathan Lu (呂東霖),
linux-arm-kernel, matthias.bgg
Il 02/10/23 09:14, Jason-JH Lin (林睿祥) ha scritto:
> Hi Angelo,
>
> Thanks for the reviews.
>
> On Thu, 2023-09-28 at 12:00 +0200, AngeloGioacchino Del Regno wrote:
>> Il 27/09/23 17:38, Jason-JH.Lin ha scritto:
>>> Add dynamic select available connector flow in
>>> mtk_drm_crtc_create()
>>> and mtk_drm_crtc_atomic_enable().
>>>
>>> In mtk_drm_crtc_create(), if there is a connector routes array in
>>> drm
>>> driver data, all components definded in the connector routes array
>>> will
>>> be checked and their encoder_index will be set.
>>>
>>> In mtk_drm_crtc_atomic_enable(), crtc will check its encoder_index
>>> to
>>> identify which componet in the connector routes array should
>>> append.
>>>
>>> Signed-off-by: Nancy Lin <nancy.lin@mediatek.com>
>>> Signed-off-by: Nathan Lu <nathan.lu@mediatek.com>
>>> Signed-off-by: Jason-JH.Lin <jason-jh.lin@mediatek.com>
>>> Tested-by: Fei Shao <fshao@chromium.org>
>>> ---
>>> drivers/gpu/drm/mediatek/mtk_drm_crtc.c | 78
>>> ++++++++++++++++++++-
>>> drivers/gpu/drm/mediatek/mtk_drm_crtc.h | 5 +-
>>> drivers/gpu/drm/mediatek/mtk_drm_ddp_comp.c | 27 ++++++-
>>> drivers/gpu/drm/mediatek/mtk_drm_drv.c | 13 +++-
>>> drivers/gpu/drm/mediatek/mtk_drm_drv.h | 7 ++
>>> 5 files changed, 123 insertions(+), 7 deletions(-)
>>>
>>> diff --git a/drivers/gpu/drm/mediatek/mtk_drm_crtc.c
>>> b/drivers/gpu/drm/mediatek/mtk_drm_crtc.c
>>> index b6fa4ad2f94d..8eb4d2646a76 100644
>>> --- a/drivers/gpu/drm/mediatek/mtk_drm_crtc.c
>>> +++ b/drivers/gpu/drm/mediatek/mtk_drm_crtc.c
>>> @@ -63,6 +63,8 @@ struct mtk_drm_crtc {
>>> struct mtk_mutex *mutex;
>>> unsigned int ddp_comp_nr;
>>> struct mtk_ddp_comp **ddp_comp;
>>> + unsigned int num_conn_routes;
>>> + const struct mtk_drm_route *conn_routes;
>>>
>>> /* lock for display hardware access */
>>> struct mutex hw_lock;
>>> @@ -647,6 +649,45 @@ static void mtk_drm_crtc_disable_vblank(struct
>>> drm_crtc *crtc)
>>> mtk_ddp_comp_disable_vblank(comp);
>>> }
>>>
>>> +static void mtk_drm_crtc_update_output(struct drm_crtc *crtc,
>>> + struct drm_atomic_state *state)
>>> +{
>>> + int crtc_index = drm_crtc_index(crtc);
>>> + int i;
>>> + struct device *dev;
>>> + struct drm_crtc_state *crtc_state = state-
>>>> crtcs[crtc_index].new_state;
>>> + struct mtk_drm_crtc *mtk_crtc = to_mtk_crtc(crtc);
>>> + struct mtk_drm_private *priv = crtc->dev->dev_private;
>>> + unsigned int comp_id;
>>
>> You're not using comp_id globally in this function....
>
> OK, I'll move it to the for loop below.
>
>>
>>> + unsigned int encoder_mask = crtc_state->encoder_mask;
>>> +
>>> + if (!crtc_state->connectors_changed)
>>> + return;
>>> +
>>> + if (!mtk_crtc->num_conn_routes)
>>> + return;
>>> +
>>> + priv = priv->all_drm_private[crtc_index];
>>
>> This is a bit confusing. You're reassigning priv: please avoid that.
>> I would prefer directly seeing the final assignment, or otherwise two
>> pointers.
>>
>> struct mtk_drm_private *priv;
>>
>> if ....
>> ...........
>>
>> priv = crtc->dev->dev_private->all_drm_private[crtc_index];
>>
>
> OK, because the type of dev_private is void*. I'll change the
> assignment to:
>
> priv = ((struct mtk_drm_private *)crtc->dev->dev_private)-
>> all_drm_private[crtc_index];
>
>>> + dev = priv->dev;
>>> +
>>> + dev_dbg(dev, "connector change:%d, encoder mask:0x%x for
>>> crtc:%d\n",
>>> + crtc_state->connectors_changed, encoder_mask,
>>> crtc_index);
>>> +
>>> + for (i = 0; i < mtk_crtc->num_conn_routes; i++) {
>>> + struct mtk_ddp_comp *comp;
>>
>> ...so you can move it here...
>
> OK, I'll move it here.
>
>> unsigned int comp_id = mtk_crtc-
>>> conn_routes[i].route_ddp;
>> struct mtk_ddp_comp *comp = &priv->ddp_comp[comp_id];
>>
>>> +
>>> + comp_id = mtk_crtc->conn_routes[i].route_ddp;
>>> + comp = &priv->ddp_comp[comp_id];
>>> + if (comp->encoder_index >= 0 &&
>>> + encoder_mask & BIT(comp->encoder_index)) {
>>
>> For readability, I would prefer to see
>>
>> if (comp->encoder_index >= 0 &&
>> (encoder_mask & BIT(comp->encoder_index))) {
>>
>> ...but I don't have strong opinions on that.
>>
>
> OK, I'll modify it.
>
>>> + mtk_crtc->ddp_comp[mtk_crtc->ddp_comp_nr - 1] =
>>> comp;
>>> + dev_dbg(dev, "Add comp_id: %d at path index
>>> %d\n",
>>> + comp->id, mtk_crtc->ddp_comp_nr - 1);
>>> + break;
>>> + }
>>> + }
>>> +}
>>> +
>>> int mtk_drm_crtc_plane_check(struct drm_crtc *crtc, struct
>>> drm_plane *plane,
>>> struct mtk_plane_state *state)
>>> {
>>> @@ -679,6 +720,8 @@ static void mtk_drm_crtc_atomic_enable(struct
>>> drm_crtc *crtc,
>>>
>>> DRM_DEBUG_DRIVER("%s %d\n", __func__, crtc->base.id);
>>>
>>> + mtk_drm_crtc_update_output(crtc, state);
>>> +
>>
>> What's the point of updating the output before
>> pm_runtime_resume_and_get()?
>> If PM resume fails we're not enabling the CRTC so the update is
>> actually useless.
>>
>> Please move this after PM resume: that will also possibly come handy
>> in the future.
>
> OK, I'll move this after pm_runtime_resume_and_get().
>
>>
>>
>>> ret = pm_runtime_resume_and_get(comp->dev);
>>> if (ret < 0) {
>>> DRM_DEV_ERROR(comp->dev, "Failed to enable power
>>> domain: %d\n", ret);
>>> @@ -884,7 +927,8 @@ struct device *mtk_drm_crtc_dma_dev_get(struct
>>> drm_crtc *crtc)
>>>
>>> int mtk_drm_crtc_create(struct drm_device *drm_dev,
>>> const unsigned int *path, unsigned int
>>> path_len,
>>> - int priv_data_index)
>>> + int priv_data_index, const struct mtk_drm_route
>>> *conn_routes,
>>> + unsigned int num_conn_routes)
>>> {
>>> struct mtk_drm_private *priv = drm_dev->dev_private;
>>> struct device *dev = drm_dev->dev;
>>> @@ -935,7 +979,8 @@ int mtk_drm_crtc_create(struct drm_device
>>> *drm_dev,
>>>
>>> mtk_crtc->mmsys_dev = priv->mmsys_dev;
>>> mtk_crtc->ddp_comp_nr = path_len;
>>> - mtk_crtc->ddp_comp = devm_kmalloc_array(dev, mtk_crtc-
>>>> ddp_comp_nr,
>>> + mtk_crtc->ddp_comp = devm_kmalloc_array(dev,
>>> + mtk_crtc->ddp_comp_nr +
>>> (conn_routes ? 1 : 0),
>>> sizeof(*mtk_crtc-
>>>> ddp_comp),
>>> GFP_KERNEL);
>>> if (!mtk_crtc->ddp_comp)
>>> @@ -1038,5 +1083,34 @@ int mtk_drm_crtc_create(struct drm_device
>>> *drm_dev,
>>> init_waitqueue_head(&mtk_crtc->cb_blocking_queue);
>>> }
>>> #endif
>>> +
>>> + if (conn_routes) {
>>> + struct device_node *node;
>>> + struct mtk_ddp_comp *comp;
>>> + unsigned int comp_id;
>>> +
>>> + for (i = 0; i < num_conn_routes; i++) {
>>
>> Same here, you're locally using comp_id, node and comp *only* in the
>> for loop....
>>
>
> OK I'll move it here.
>
>> unsigned int comp_id =
>> conn_routes[i].route_ddp;
>> struct device_node *node = priv-
>>> comp_node[comp_id];
>> struct mtk_ddp_comp *comp = &priv-
>>> ddp_comp[comp_id];
>>
>>> + comp_id = conn_routes[i].route_ddp;
>>> + node = priv->comp_node[comp_id];
>>> + comp = &priv->ddp_comp[comp_id];
>>> +
>>> + if (!comp->dev) {
>>> + dev_dbg(dev, "comp_id:%d, Component
>>> %pOF not initialized\n",
>>> + comp_id, node);
>>> + /* mark encoder_index to -1, if route
>>> comp device is not enabled */
>>> + comp->encoder_index = -1;
>>> + continue;
>>> + }
>>> +
>>> + mtk_ddp_comp_encoder_index_set(&priv-
>>>> ddp_comp[comp_id]);
>>> + }
>>> +
>>> + mtk_crtc->num_conn_routes = num_conn_routes;
>>> + mtk_crtc->conn_routes = conn_routes;
>>> +
>>> + /* increase ddp_comp_nr at the end of
>>> mtk_drm_crtc_create */
>>> + mtk_crtc->ddp_comp_nr++;
>>> + }
>>> +
>>> return 0;
>>> }
>>> diff --git a/drivers/gpu/drm/mediatek/mtk_drm_crtc.h
>>> b/drivers/gpu/drm/mediatek/mtk_drm_crtc.h
>>> index 3e9046993d09..3c224595fa71 100644
>>> --- a/drivers/gpu/drm/mediatek/mtk_drm_crtc.h
>>> +++ b/drivers/gpu/drm/mediatek/mtk_drm_crtc.h
>>> @@ -8,6 +8,7 @@
>>>
>>> #include <drm/drm_crtc.h>
>>> #include "mtk_drm_ddp_comp.h"
>>> +#include "mtk_drm_drv.h"
>>> #include "mtk_drm_plane.h"
>>>
>>> #define MTK_LUT_SIZE 512
>>> @@ -18,7 +19,9 @@ void mtk_drm_crtc_commit(struct drm_crtc *crtc);
>>> int mtk_drm_crtc_create(struct drm_device *drm_dev,
>>> const unsigned int *path,
>>> unsigned int path_len,
>>> - int priv_data_index);
>>> + int priv_data_index,
>>> + const struct mtk_drm_route *conn_routes,
>>> + unsigned int num_conn_routes);
>>> int mtk_drm_crtc_plane_check(struct drm_crtc *crtc, struct
>>> drm_plane *plane,
>>> struct mtk_plane_state *state);
>>> void mtk_drm_crtc_async_update(struct drm_crtc *crtc, struct
>>> drm_plane *plane,
>>> diff --git a/drivers/gpu/drm/mediatek/mtk_drm_ddp_comp.c
>>> b/drivers/gpu/drm/mediatek/mtk_drm_ddp_comp.c
>>> index 771f4e173353..4ddb5e561116 100644
>>> --- a/drivers/gpu/drm/mediatek/mtk_drm_ddp_comp.c
>>> +++ b/drivers/gpu/drm/mediatek/mtk_drm_ddp_comp.c
>>> @@ -507,6 +507,23 @@ static bool mtk_drm_find_comp_in_ddp(struct
>>> device *dev,
>>> return false;
>>> }
>>>
>>> +static int mtk_drm_find_comp_in_ddp_conn_path(struct device *dev,
>>> + const struct
>>> mtk_drm_route *routes,
>>> + unsigned int num_routes,
>>> + struct mtk_ddp_comp
>>> *ddp_comp)
>>> +{
>>> + unsigned int i;
>>> +
>>> + if (!routes)
>>> + return -EINVAL;
>>> +
>>> + for (i = 0; i < num_routes; i++)
>>> + if (dev == ddp_comp[routes[i].route_ddp].dev)
>>> + return BIT(routes[i].crtc_id);
>>> +
>>> + return -ENODEV;
>>> +}
>>> +
>>> int mtk_ddp_comp_get_id(struct device_node *node,
>>> enum mtk_ddp_comp_type comp_type)
>>> {
>>> @@ -538,7 +555,15 @@ unsigned int
>>> mtk_drm_find_possible_crtc_by_comp(struct drm_device *drm,
>>> private->data->third_len,
>>> private->ddp_comp))
>>> ret = BIT(2);
>>> else
>>> - DRM_INFO("Failed to find comp in ddp table\n");
>>> + ret = mtk_drm_find_comp_in_ddp_conn_path(dev,
>>> + private->data-
>>>> conn_routes,
>>> + private->data-
>>>> num_conn_routes,
>>> + private-
>>>> ddp_comp);
>>> +
>>> + if (ret <= 0) {
>>> + DRM_INFO("Failed to find comp in ddp table, ret =%d\n",
>>> ret);
>>> + ret = 0;
>>
>> Why are you returning 0 for error here?!
>
> Because this function return 0 means not found any valid crtc bits.
> Should I change this?
>
Yes, please, return a relevant error code instead.
Thanks,
Angelo
> Regards,
> Jason-JH.Lin
>
>>
>>> + }
>>>
>>> return ret;
>>> }
>>
>> Regards,
>> Angelo
>>
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: [PATCH v10 5/9] drm/mediatek: Add connector dynamic selection capability
2023-10-02 10:23 ` AngeloGioacchino Del Regno
@ 2023-10-03 3:34 ` Jason-JH Lin (林睿祥)
0 siblings, 0 replies; 10+ messages in thread
From: Jason-JH Lin (林睿祥) @ 2023-10-03 3:34 UTC (permalink / raw)
To: angelogioacchino.delregno, chunkuang.hu
Cc: linux-mediatek, linux-kernel,
Singo Chang (張興國),
Johnson Wang (王聖鑫),
Jason-ch Chen (陳建豪),
Shawn Sung (宋孝謙),
Nancy Lin (林欣螢),
dri-devel, Project_Global_Chrome_Upstream_Group,
Nathan Lu (呂東霖),
linux-arm-kernel, matthias.bgg
Hi Angelo,
Thanks for the reviews.
[snip]
> > > >
> > > > +
> > > > + if (ret <= 0) {
> > > > + DRM_INFO("Failed to find comp in ddp table, ret
> > > > =%d\n",
> > > > ret);
> > > > + ret = 0;
> > >
> > > Why are you returning 0 for error here?!
> >
> > Because this function return 0 means not found any valid crtc bits.
> > Should I change this?
> >
>
> Yes, please, return a relevant error code instead.
>
I found that the return type of this function is 'unsigned int', I
think returning error code instead will change the original behavior of
returning 0 as not finding any possible crtc.
e.g. drm_encoder_ok() in drm_encoder.h will check:
!!(encoder->possible_crtcs & drm_crtc_mask(crtc))
The negative error code convert to unsigned int will be a huge number
and that may cause the unexpected result here.
Regards,
Jason-JH.Lin
> Thanks,
> Angelo
>
> > Regards,
> > Jason-JH.Lin
> >
> > >
> > > > + }
> > > >
> > > > return ret;
> > > > }
> > >
> > > Regards,
> > > Angelo
> > >
>
>
>
^ permalink raw reply [flat|nested] 10+ messages in thread
end of thread, other threads:[~2023-10-03 3:34 UTC | newest]
Thread overview: 10+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
[not found] <20230927153833.23583-1-jason-jh.lin@mediatek.com>
[not found] ` <20230927153833.23583-3-jason-jh.lin@mediatek.com>
2023-09-28 9:21 ` [PATCH v10 2/9] drm/mediatek: Add crtc path enum for all_drm_priv array AngeloGioacchino Del Regno
[not found] ` <20230927153833.23583-8-jason-jh.lin@mediatek.com>
2023-09-28 9:24 ` [PATCH v10 7/9] drm/mediatek: Support dynamic selection of DP_INTF0 on MT8188 VDOSYS0 AngeloGioacchino Del Regno
[not found] ` <20230927153833.23583-5-jason-jh.lin@mediatek.com>
2023-09-28 9:24 ` [PATCH v10 4/9] drm/mediatek: Add encoder_index interface for mtk_ddp_comp_funcs AngeloGioacchino Del Regno
[not found] ` <20230927153833.23583-9-jason-jh.lin@mediatek.com>
2023-09-28 9:24 ` [PATCH v10 8/9] drm/mediatek: dsi: Support dynamic connector selection AngeloGioacchino Del Regno
[not found] ` <20230927153833.23583-7-jason-jh.lin@mediatek.com>
2023-09-28 9:24 ` [PATCH v10 6/9] drm/mediatek: dpi: " AngeloGioacchino Del Regno
[not found] ` <20230927153833.23583-10-jason-jh.lin@mediatek.com>
2023-09-28 9:24 ` [PATCH v10 9/9] drm/mediatek: Support dynamic selection of DSI0 on MT8188 VDOSYS0 AngeloGioacchino Del Regno
[not found] ` <20230927153833.23583-6-jason-jh.lin@mediatek.com>
2023-09-28 10:00 ` [PATCH v10 5/9] drm/mediatek: Add connector dynamic selection capability AngeloGioacchino Del Regno
2023-10-02 7:14 ` Jason-JH Lin (林睿祥)
2023-10-02 10:23 ` AngeloGioacchino Del Regno
2023-10-03 3:34 ` Jason-JH Lin (林睿祥)
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox
all inboxes | Powered by JetHome®