2022-04-09 00:36:13

by Bjorn Andersson

[permalink] [raw]
Subject: [PATCH v3 1/2] dt-bindings: display: msm: Add optional resets

Add an optional reference to the MDSS_CORE reset, which when specified
can be used by the implementation to reset the hardware blocks.

Signed-off-by: Bjorn Andersson <[email protected]>
---

Resending these two patches again as I put "v2" in the subject, even though I
meant v3. Sorry about that.

Changes since v2:
- None

.../devicetree/bindings/display/msm/dpu-qcm2290.yaml | 4 ++++
Documentation/devicetree/bindings/display/msm/dpu-sc7180.yaml | 4 ++++
Documentation/devicetree/bindings/display/msm/dpu-sc7280.yaml | 4 ++++
Documentation/devicetree/bindings/display/msm/dpu-sdm845.yaml | 4 ++++
4 files changed, 16 insertions(+)

diff --git a/Documentation/devicetree/bindings/display/msm/dpu-qcm2290.yaml b/Documentation/devicetree/bindings/display/msm/dpu-qcm2290.yaml
index 6fb7e321f011..734d14de966d 100644
--- a/Documentation/devicetree/bindings/display/msm/dpu-qcm2290.yaml
+++ b/Documentation/devicetree/bindings/display/msm/dpu-qcm2290.yaml
@@ -66,6 +66,10 @@ properties:
interconnect-names:
const: mdp0-mem

+ resets:
+ items:
+ - description: MDSS_CORE reset
+
patternProperties:
"^display-controller@[0-9a-f]+$":
type: object
diff --git a/Documentation/devicetree/bindings/display/msm/dpu-sc7180.yaml b/Documentation/devicetree/bindings/display/msm/dpu-sc7180.yaml
index 12a86b1ec1bc..b41991eaa454 100644
--- a/Documentation/devicetree/bindings/display/msm/dpu-sc7180.yaml
+++ b/Documentation/devicetree/bindings/display/msm/dpu-sc7180.yaml
@@ -65,6 +65,10 @@ properties:
interconnect-names:
const: mdp0-mem

+ resets:
+ items:
+ - description: MDSS_CORE reset
+
patternProperties:
"^display-controller@[0-9a-f]+$":
type: object
diff --git a/Documentation/devicetree/bindings/display/msm/dpu-sc7280.yaml b/Documentation/devicetree/bindings/display/msm/dpu-sc7280.yaml
index fbeb931a026e..6e417d06fc79 100644
--- a/Documentation/devicetree/bindings/display/msm/dpu-sc7280.yaml
+++ b/Documentation/devicetree/bindings/display/msm/dpu-sc7280.yaml
@@ -64,6 +64,10 @@ properties:
interconnect-names:
const: mdp0-mem

+ resets:
+ items:
+ - description: MDSS_CORE reset
+
patternProperties:
"^display-controller@[0-9a-f]+$":
type: object
diff --git a/Documentation/devicetree/bindings/display/msm/dpu-sdm845.yaml b/Documentation/devicetree/bindings/display/msm/dpu-sdm845.yaml
index 0dca4b3d66e4..1a42491efdbc 100644
--- a/Documentation/devicetree/bindings/display/msm/dpu-sdm845.yaml
+++ b/Documentation/devicetree/bindings/display/msm/dpu-sdm845.yaml
@@ -57,6 +57,10 @@ properties:

ranges: true

+ resets:
+ items:
+ - description: MDSS_CORE reset
+
patternProperties:
"^display-controller@[0-9a-f]+$":
type: object
--
2.35.1


2022-04-09 04:55:49

by Bjorn Andersson

[permalink] [raw]
Subject: [PATCH v3 2/2] drm/msm/dpu: Issue MDSS reset during initialization

It's typical for the bootloader to bring up the display for showing a
boot splash or efi framebuffer. But in some cases the kernel driver ends
up only partially configuring (in particular) the DPU, which might
result in e.g. that two different data paths attempts to push data to
the interface - with resulting graphical artifacts.

Naturally the end goal would be to inherit the bootloader's
configuration and provide the user with a glitch free handover from the
boot configuration to a running DPU.

But as implementing seamless transition from the bootloader
configuration to the running OS will be a considerable effort, start by
simply resetting the entire MDSS to its power-on state, to avoid the
partial configuration.

Signed-off-by: Bjorn Andersson <[email protected]>
---

Changes since v2:
- Move the reset before the probing of the children, to avoid issues if child
drivers touched hardware at probe time.
- Extend reset assert time as the previous 1us showed reliability issues on
Lenovo Flex 5G.

drivers/gpu/drm/msm/disp/dpu1/dpu_mdss.c | 32 ++++++++++++++++++++++++
1 file changed, 32 insertions(+)

diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_mdss.c b/drivers/gpu/drm/msm/disp/dpu1/dpu_mdss.c
index b10ca505f9ac..6f921910d0e0 100644
--- a/drivers/gpu/drm/msm/disp/dpu1/dpu_mdss.c
+++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_mdss.c
@@ -3,10 +3,12 @@
* Copyright (c) 2018, The Linux Foundation
*/

+#include <linux/delay.h>
#include <linux/irq.h>
#include <linux/irqchip.h>
#include <linux/irqdesc.h>
#include <linux/irqchip/chained_irq.h>
+#include <linux/reset.h>
#include "dpu_kms.h"

#define to_dpu_mdss(x) container_of(x, struct dpu_mdss, base)
@@ -197,6 +199,32 @@ static void dpu_mdss_destroy(struct msm_mdss *mdss)
dpu_mdss->mmio = NULL;
}

+static int dpu_mdss_reset(struct device *dev)
+{
+ struct reset_control *reset;
+
+ reset = reset_control_get_optional_exclusive(dev, NULL);
+ if (!reset) {
+ /* Optional reset not specified */
+ return 0;
+ } else if (IS_ERR(reset)) {
+ DPU_ERROR("failed to acquire mdss reset, ret=%ld", PTR_ERR(reset));
+ return PTR_ERR(reset);
+ }
+
+ reset_control_assert(reset);
+ /*
+ * Tests indicate that reset has to be held for some period of time,
+ * make it one frame in a typical system
+ */
+ msleep(20);
+ reset_control_deassert(reset);
+
+ reset_control_put(reset);
+
+ return 0;
+}
+
static const struct msm_mdss_funcs mdss_funcs = {
.enable = dpu_mdss_enable,
.disable = dpu_mdss_disable,
@@ -210,6 +238,10 @@ int dpu_mdss_init(struct platform_device *pdev)
int ret;
int irq;

+ ret = dpu_mdss_reset(&pdev->dev);
+ if (ret)
+ return ret;
+
dpu_mdss = devm_kzalloc(&pdev->dev, sizeof(*dpu_mdss), GFP_KERNEL);
if (!dpu_mdss)
return -ENOMEM;
--
2.35.1

2022-04-10 11:05:55

by Dmitry Baryshkov

[permalink] [raw]
Subject: Re: [PATCH v3 1/2] dt-bindings: display: msm: Add optional resets

On Sat, 9 Apr 2022 at 00:05, Bjorn Andersson <[email protected]> wrote:
>
> Add an optional reference to the MDSS_CORE reset, which when specified
> can be used by the implementation to reset the hardware blocks.
>
> Signed-off-by: Bjorn Andersson <[email protected]>

Reviewed-by: Dmitry Baryshkov <[email protected]>


> ---
>
> Resending these two patches again as I put "v2" in the subject, even though I
> meant v3. Sorry about that.
>
> Changes since v2:
> - None
>
> .../devicetree/bindings/display/msm/dpu-qcm2290.yaml | 4 ++++
> Documentation/devicetree/bindings/display/msm/dpu-sc7180.yaml | 4 ++++
> Documentation/devicetree/bindings/display/msm/dpu-sc7280.yaml | 4 ++++
> Documentation/devicetree/bindings/display/msm/dpu-sdm845.yaml | 4 ++++
> 4 files changed, 16 insertions(+)
>
> diff --git a/Documentation/devicetree/bindings/display/msm/dpu-qcm2290.yaml b/Documentation/devicetree/bindings/display/msm/dpu-qcm2290.yaml
> index 6fb7e321f011..734d14de966d 100644
> --- a/Documentation/devicetree/bindings/display/msm/dpu-qcm2290.yaml
> +++ b/Documentation/devicetree/bindings/display/msm/dpu-qcm2290.yaml
> @@ -66,6 +66,10 @@ properties:
> interconnect-names:
> const: mdp0-mem
>
> + resets:
> + items:
> + - description: MDSS_CORE reset
> +
> patternProperties:
> "^display-controller@[0-9a-f]+$":
> type: object
> diff --git a/Documentation/devicetree/bindings/display/msm/dpu-sc7180.yaml b/Documentation/devicetree/bindings/display/msm/dpu-sc7180.yaml
> index 12a86b1ec1bc..b41991eaa454 100644
> --- a/Documentation/devicetree/bindings/display/msm/dpu-sc7180.yaml
> +++ b/Documentation/devicetree/bindings/display/msm/dpu-sc7180.yaml
> @@ -65,6 +65,10 @@ properties:
> interconnect-names:
> const: mdp0-mem
>
> + resets:
> + items:
> + - description: MDSS_CORE reset
> +
> patternProperties:
> "^display-controller@[0-9a-f]+$":
> type: object
> diff --git a/Documentation/devicetree/bindings/display/msm/dpu-sc7280.yaml b/Documentation/devicetree/bindings/display/msm/dpu-sc7280.yaml
> index fbeb931a026e..6e417d06fc79 100644
> --- a/Documentation/devicetree/bindings/display/msm/dpu-sc7280.yaml
> +++ b/Documentation/devicetree/bindings/display/msm/dpu-sc7280.yaml
> @@ -64,6 +64,10 @@ properties:
> interconnect-names:
> const: mdp0-mem
>
> + resets:
> + items:
> + - description: MDSS_CORE reset
> +
> patternProperties:
> "^display-controller@[0-9a-f]+$":
> type: object
> diff --git a/Documentation/devicetree/bindings/display/msm/dpu-sdm845.yaml b/Documentation/devicetree/bindings/display/msm/dpu-sdm845.yaml
> index 0dca4b3d66e4..1a42491efdbc 100644
> --- a/Documentation/devicetree/bindings/display/msm/dpu-sdm845.yaml
> +++ b/Documentation/devicetree/bindings/display/msm/dpu-sdm845.yaml
> @@ -57,6 +57,10 @@ properties:
>
> ranges: true
>
> + resets:
> + items:
> + - description: MDSS_CORE reset
> +
> patternProperties:
> "^display-controller@[0-9a-f]+$":
> type: object
> --
> 2.35.1
>


--
With best wishes
Dmitry

2022-04-11 06:40:38

by Krzysztof Kozlowski

[permalink] [raw]
Subject: Re: [PATCH v3 1/2] dt-bindings: display: msm: Add optional resets

On 08/04/2022 23:08, Bjorn Andersson wrote:
> Add an optional reference to the MDSS_CORE reset, which when specified
> can be used by the implementation to reset the hardware blocks.
>
> Signed-off-by: Bjorn Andersson <[email protected]>
> ---
>
> Resending these two patches again as I put "v2" in the subject, even though I
> meant v3. Sorry about that.
>

I reviewed v2, so one here as well:

Acked-by: Krzysztof Kozlowski <[email protected]>


Best regards,
Krzysztof

2022-04-12 11:08:56

by Dmitry Baryshkov

[permalink] [raw]
Subject: Re: [PATCH v3 2/2] drm/msm/dpu: Issue MDSS reset during initialization

On Sat, 9 Apr 2022 at 00:05, Bjorn Andersson <[email protected]> wrote:
>
> It's typical for the bootloader to bring up the display for showing a
> boot splash or efi framebuffer. But in some cases the kernel driver ends
> up only partially configuring (in particular) the DPU, which might
> result in e.g. that two different data paths attempts to push data to
> the interface - with resulting graphical artifacts.
>
> Naturally the end goal would be to inherit the bootloader's
> configuration and provide the user with a glitch free handover from the
> boot configuration to a running DPU.
>
> But as implementing seamless transition from the bootloader
> configuration to the running OS will be a considerable effort, start by
> simply resetting the entire MDSS to its power-on state, to avoid the
> partial configuration.
>
> Signed-off-by: Bjorn Andersson <[email protected]>

Reviewed-by: Dmitry Baryshkov <[email protected]>


> ---
>
> Changes since v2:
> - Move the reset before the probing of the children, to avoid issues if child
> drivers touched hardware at probe time.
> - Extend reset assert time as the previous 1us showed reliability issues on
> Lenovo Flex 5G.
>
> drivers/gpu/drm/msm/disp/dpu1/dpu_mdss.c | 32 ++++++++++++++++++++++++
> 1 file changed, 32 insertions(+)
>
> diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_mdss.c b/drivers/gpu/drm/msm/disp/dpu1/dpu_mdss.c
> index b10ca505f9ac..6f921910d0e0 100644
> --- a/drivers/gpu/drm/msm/disp/dpu1/dpu_mdss.c
> +++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_mdss.c
> @@ -3,10 +3,12 @@
> * Copyright (c) 2018, The Linux Foundation
> */
>
> +#include <linux/delay.h>
> #include <linux/irq.h>
> #include <linux/irqchip.h>
> #include <linux/irqdesc.h>
> #include <linux/irqchip/chained_irq.h>
> +#include <linux/reset.h>
> #include "dpu_kms.h"
>
> #define to_dpu_mdss(x) container_of(x, struct dpu_mdss, base)
> @@ -197,6 +199,32 @@ static void dpu_mdss_destroy(struct msm_mdss *mdss)
> dpu_mdss->mmio = NULL;
> }
>
> +static int dpu_mdss_reset(struct device *dev)
> +{
> + struct reset_control *reset;
> +
> + reset = reset_control_get_optional_exclusive(dev, NULL);
> + if (!reset) {
> + /* Optional reset not specified */
> + return 0;
> + } else if (IS_ERR(reset)) {
> + DPU_ERROR("failed to acquire mdss reset, ret=%ld", PTR_ERR(reset));
> + return PTR_ERR(reset);
> + }
> +
> + reset_control_assert(reset);
> + /*
> + * Tests indicate that reset has to be held for some period of time,
> + * make it one frame in a typical system
> + */
> + msleep(20);
> + reset_control_deassert(reset);
> +
> + reset_control_put(reset);
> +
> + return 0;
> +}
> +
> static const struct msm_mdss_funcs mdss_funcs = {
> .enable = dpu_mdss_enable,
> .disable = dpu_mdss_disable,
> @@ -210,6 +238,10 @@ int dpu_mdss_init(struct platform_device *pdev)
> int ret;
> int irq;
>
> + ret = dpu_mdss_reset(&pdev->dev);
> + if (ret)
> + return ret;
> +
> dpu_mdss = devm_kzalloc(&pdev->dev, sizeof(*dpu_mdss), GFP_KERNEL);
> if (!dpu_mdss)
> return -ENOMEM;
> --
> 2.35.1
>


--
With best wishes
Dmitry