media: m2m-deinterlace: add device-tree support
Subsystems (2)
v2 cover media: m2m-deinterlace: add device-tree support
Hello all, this patch series adds device-tree support for the generic m2m-deinterlace driver.
A specific dma controller channel can now be provided to the driver via the dt, but the driver still supports the original dma query if an of_node is not present. This should preserve backwards compatibility with platform drivers that directly instantiate the device.
Probing the m2m-deinterlace device via device-tree was tested on the BeagleBone Black with an m2m-appropriate edma channel.
Changes From v1 [1]:
- [PATCH 1/2] media: dt-bindings: Add dt bindings for m2m-deinterlace
- clarify device description and commit message for dt bindings
- correct format error reported by yamllint for dt bindings
- [PATCH 2/2] media: m2m-deinterlace: add device-tree support
- no changes
[1] https://lore.kernel.org/linux-media/[email protected]/
Best,
Matthew Majewski
Matthew Majewski (2):
media: dt-bindings: Add dt bindings for m2m-deinterlace device
media: m2m-deinterlace: add device-tree support
.../bindings/media/m2m-deinterlace.yaml | 44 +++++++++++++++++++ drivers/media/platform/m2m-deinterlace.c | 25 ++++++++--- 2 files changed, 63 insertions(+), 6 deletions(-) create mode 100644 Documentation/devicetree/bindings/media/m2m-deinterlace.yaml
Signature · 1 line
2.25.1
PATCH v2 1/2 media: dt-bindings: Add dt bindings for m2m-deinterlace device
Create a new yaml schema file to describe the device tree bindings for generic m2m-deinterlace device.
This device is supported on any hardware that provides a MEM_TO_MEM capable dma channel with interleaved trasfer support. Device tree bindings are for providing appropriate dma channel to device.
--- .../bindings/media/m2m-deinterlace.yaml | 44 +++++++++++++++++++ 1 file changed, 44 insertions(+) create mode 100644 Documentation/devicetree/bindings/media/m2m-deinterlace.yaml
Documentation/devicetree/bindings/media/m2m-deinterlace.yaml new file +44 -0
@@ -0,0 +1,44 @@ +# SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause) +%YAML 1.2 +--- +$id: http://devicetree.org/schemas/media/m2m-deinterlace.yaml# +$schema: http://devicetree.org/meta-schemas/core.yaml# + +title: M2M Deinterlacer + +maintainers: + - Mauro Carvalho Chehab <[email protected]> + +description: |- + A generic memory2memory device for deinterlacing video using dmaengine. It can + convert between interlaced buffer formats and can convert interlaced to + progressive using a simple line-doubling algorithm. This device can be used on + any hardware that provides a MEM_TO_MEM capable dma controller that supports + interleaved transfers. + +properties: + compatible: + const: m2m-deinterlace + + dma-names: + items: + - const: rxtx + + dmas: + items: + - description: MEM_TO_MEM capable DMA channel + +required: + - compatible + - dma-names + - dmas + +additionalProperties: false + +examples: + - | + m2m-deinterlace { + compatible = "m2m-deinterlace"; + dma-names = "rxtx"; + dmas = <&edma 20 0>; + };
Signature · 1 line
2.25.1
Re: [PATCH v2 1/2] media: dt-bindings: Add dt bindings for m2m-deinterlace device
Quoted 4 lines · Matthew Majewski
Create a new yaml schema file to describe the device tree bindings for generic m2m-deinterlace device.
This device is supported on any hardware that provides a
MEM_TO_MEM
Which device? I don't see here any device name/model.
I asked to provide here some examples of devices.
Quoted 2 lines
capable dma channel with interleaved trasfer support. Device tree bindings are for providing appropriate dma channel to device.
Don't describe what DT is, but the hardware.
Quoted 20 lines
Signed-off-by Matthew Majewski <[email protected]>--- .../bindings/media/m2m-deinterlace.yaml | 44 +++++++++++++++++++ 1 file changed, 44 insertions(+) create mode 100644 Documentation/devicetree/bindings/media/m2m-deinterlace.yamlDocumentation/devicetree/bindings/media/m2m-deinterlace.yaml new file +7
-0@@ -0,0 +1,44 @@ +# SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause) +%YAML 1.2 +--- +$id: http://devicetree.org/schemas/media/m2m-deinterlace.yaml# +$schema: http://devicetree.org/meta-schemas/core.yaml# + +title: M2M Deinterlacer
Documentation/devicetree/bindings/media/m2m-deinterlace.yaml line 17
Quoted 10 lines
Documentation/devicetree/bindings/media/m2m-deinterlace.yaml new file +10
-0+ +maintainers: + - Mauro Carvalho Chehab <[email protected]> + +description: |- + A generic memory2memory device for deinterlacing video using dmaengine. It can + convert between interlaced buffer formats and can convert interlaced to + progressive using a simple line-doubling algorithm. This device can be used on + any hardware that provides a MEM_TO_MEM capable dma controller that supports + interleaved transfers.
And how do you program that device to deinterlace? How do you signal end of frame/data when writing to the memory?
It still looks all this is for driver :/
Best regards,
Krzysztof
Re: [PATCH v2 1/2] media: dt-bindings: Add dt bindings for m2m-deinterlace device
Hi Krzysztof,
Quoted 8 lines · Krzysztof Kozlowski
Quoted 5 lines · Matthew Majewski
Create a new yaml schema file to describe the device tree bindings for
generic m2m-deinterlace device.This device is supported on any hardware that provides a
MEM_TO_MEMWhich device? I don't see here any device name/model.
By "device" I am referring to the m2m-deinterlace device, which I explained is a quasi-virtual device. If this is confusing wording I can change.
Quoted 1 line
I asked to provide here some examples of devices.
As I wrote, supported devices/hardware is anything that provides a MEM_TO_MEM capable dma-controller with interleaved transfer support. I did not list specific devices because the bindings are supposed to be generic, as they are not describing actual silicon. But if you want me to list some devices which provide a compatible dma-controller, here are devices I found in the current mainline kernel:
- TI OMAP Soc Family
- TI Davinci Soc Family
- TI Keystone Processor Family
- IMX27 Processor and variants
- Several Microchip Processors (sama5, sam9x7, sam9x60)
As I mentioned in my original email, I have personally tested on a BeagleBone Black with an AM335X OMAP processor. There are likely many more devices with compatible dma-controllers that could be supported with additional dmaengine driver support.
Quoted 5 lines
Quoted 2 lines
capable dma channel with interleaved trasfer support. Device tree bindings are for providing appropriate dma channel to device.
Don't describe what DT is, but the hardware.
Ok, will remove reference to DT.
Quoted 11 lines
Quoted 10 lines
+description: |-
+ A generic memory2memory device for deinterlacing video using dmaengine. It can
+ convert between interlaced buffer formats and can convert interlaced to
+ progressive using a simple line-doubling algorithm. This device can be used on
+ any hardware that provides aMEM_TO_MEMcapable dma controller that supports
+ interleaved transfers.
Quoted 6 lines
And how do you program that device to deinterlace? How do you signal end
of frame/data when writing to the memory?It still looks all this is for driver :/
All of the deinterlacing is handled by the dma channel. To simplify a bit, m2m-deinterlace basically just translates video format information into appropriate interleaved dma transfers. Everything else (and everything hardware specific) is handled by the dma engine, such as initiation and signaling completion of transfers.
I think an appropriate analogy for m2m-deinterlace would be spi-gpio. Since spi-gpio leverages gpio for bitbanging the spi protocol, the bindings do not need to describe any clocks, spi-controller registers, etc. All of the hardware specific components are abstracted away by the gpio controller. But the spi-gpio bindings still exist to specify which gpios are used.
Best,
Matthew
Re: [PATCH v2 1/2] media: dt-bindings: Add dt bindings for m2m-deinterlace device
Quoted 16 lines · Matthew Majewski
Hi Krzysztof,
Quoted 8 lines · Krzysztof Kozlowski
Quoted 5 lines · Matthew Majewski
Create a new yaml schema file to describe the device tree bindings for
generic m2m-deinterlace device.This device is supported on any hardware that provides a
MEM_TO_MEMWhich device? I don't see here any device name/model.
By "device" I am referring to the m2m-deinterlace device, which I explained is a quasi-virtual device. If this is confusing wording I can change.
Quoted 6 lines
Quoted 1 line
I asked to provide here some examples of devices.
As I wrote, supported devices/hardware is anything that provides a
MEM_TO_MEMcapable dma-controller with interleaved transfer support. I did not list specific devices because the bindings are supposed to be generic, as they are not describing actual silicon. But if you want me
I already told you that no. Bindings are not supposed to be generic.
From where did you get such information?
Quoted 8 lines
to list some devices which provide a compatible dma-controller, here are devices I found in the current mainline kernel:
- TI OMAP Soc Family
- TI Davinci Soc Family
- TI Keystone Processor Family
- IMX27 Processor and variants
- Several Microchip Processors (sama5, sam9x7, sam9x60)
That's too generic - you just listed SoCs, which consist of dozen or hundred of devices. Which hardware piece is here?
Maybe this is not for a real device, but then this should be marked clearly.
Quoted 30 lines
As I mentioned in my original email, I have personally tested on a BeagleBone Black with an AM335X OMAP processor. There are likely many more devices with compatible dma-controllers that could be supported with additional dmaengine driver support.
Quoted 5 lines
Quoted 2 lines
capable dma channel with interleaved trasfer support. Device tree bindings are for providing appropriate dma channel to device.
Don't describe what DT is, but the hardware.
Ok, will remove reference to DT.
Quoted 15 lines
Quoted 10 lines
+description: |-
+ A generic memory2memory device for deinterlacing video using dmaengine. It can
+ convert between interlaced buffer formats and can convert interlaced to
+ progressive using a simple line-doubling algorithm. This device can be used on
+ any hardware that provides aMEM_TO_MEMcapable dma controller that supports
+ interleaved transfers.And how do you program that device to deinterlace? How do you signal end
of frame/data when writing to the memory?
Quoted 8 lines
Quoted 2 lines
It still looks all this is for driver :/
All of the deinterlacing is handled by the dma channel. To simplify a bit, m2m-deinterlace basically just translates video format information into appropriate interleaved dma transfers. Everything else (and everything hardware specific) is handled by the dma engine, such as initiation and signaling completion of transfers.
So the device is the dma controller and maybe all this should be folded into that controller bindings.
Quoted 4 lines
I think an appropriate analogy for m2m-deinterlace would be spi-gpio. Since spi-gpio leverages gpio for bitbanging the spi protocol, the bindings do not need to describe any clocks, spi-controller registers,
Sure, SPI GPIO is Linux driver, not a device and I am asking about it all the time.
Quoted 3 lines
etc. All of the hardware specific components are abstracted away by the gpio controller. But the spi-gpio bindings still exist to specify which gpios are used.
Best regards,
Krzysztof
Re: [PATCH v2 1/2] media: dt-bindings: Add dt bindings for m2m-deinterlace device
Quoted 82 lines · Krzysztof Kozlowski
Quoted 16 lines · Matthew Majewski
Hi Krzysztof,
Quoted 8 lines · Krzysztof Kozlowski
Quoted 5 lines · Matthew Majewski
Create a new yaml schema file to describe the device tree bindings for
generic m2m-deinterlace device.This device is supported on any hardware that provides a
MEM_TO_MEMWhich device? I don't see here any device name/model.
By "device" I am referring to the m2m-deinterlace device, which I explained is a quasi-virtual device. If this is confusing wording I can change.
Quoted 6 lines
Quoted 1 line
I asked to provide here some examples of devices.
As I wrote, supported devices/hardware is anything that provides a
MEM_TO_MEMcapable dma-controller with interleaved transfer support. I did not list specific devices because the bindings are supposed to be generic, as they are not describing actual silicon. But if you want meI already told you that no. Bindings are not supposed to be generic.
From where did you get such information?
Quoted 8 lines
to list some devices which provide a compatible dma-controller, here are devices I found in the current mainline kernel:
- TI OMAP Soc Family
- TI Davinci Soc Family
- TI Keystone Processor Family
- IMX27 Processor and variants
- Several Microchip Processors (sama5, sam9x7, sam9x60)That's too generic - you just listed SoCs, which consist of dozen or hundred of devices. Which hardware piece is here?
Maybe this is not for a real device, but then this should be marked clearly.
Quoted 30 lines
As I mentioned in my original email, I have personally tested on a BeagleBone Black with an AM335X OMAP processor. There are likely many more devices with compatible dma-controllers that could be supported with additional dmaengine driver support.
Quoted 5 lines
Quoted 2 lines
capable dma channel with interleaved trasfer support. Device tree bindings are for providing appropriate dma channel to device.
Don't describe what DT is, but the hardware.
Ok, will remove reference to DT.
Quoted 15 lines
Quoted 10 lines
+description: |-
+ A generic memory2memory device for deinterlacing video using dmaengine. It can
+ convert between interlaced buffer formats and can convert interlaced to
+ progressive using a simple line-doubling algorithm. This device can be used on
+ any hardware that provides aMEM_TO_MEMcapable dma controller that supports
+ interleaved transfers.And how do you program that device to deinterlace? How do you signal end
of frame/data when writing to the memory?Quoted 8 lines
Quoted 2 lines
It still looks all this is for driver :/
All of the deinterlacing is handled by the dma channel. To simplify a bit, m2m-deinterlace basically just translates video format information into appropriate interleaved dma transfers. Everything else (and everything hardware specific) is handled by the dma engine, such as initiation and signaling completion of transfers.
Quoted 2 lines
So the device is the dma controller and maybe all this should be folded into that controller bindings.
Answering myself: obviously no, because interleaved DMA is not relevant to this device, so again: there is no device in SoC doing that. You just add bindings for specific Linux driver without saying that this is that driver and calling it "generic device". There is no device here.
Best regards,
Krzysztof
Re: [PATCH v2 1/2] media: dt-bindings: Add dt bindings for m2m-deinterlace device
In reply to Krzysztof KozlowskiQuoted 10 lines · Krzysztof Kozlowski
Quoted 8 lines · Matthew Majewski
As I wrote, supported devices/hardware is anything that provides a
MEM_TO_MEMcapable dma-controller with interleaved transfer support. I
did not list specific devices because the bindings are supposed to be
generic, as they are not describing actual silicon. But if you want me
Quoted 3 lines
I already told you that no. Bindings are not supposed to be generic.
From where did you get such information?
There are generic bindings in the kernel and I based my bindings off of them. spi-gpio.yaml, i2c-gpio.yaml, video-mux.yaml, etc are all generic bindings, no?
Quoted 11 lines
Quoted 10 lines
to list some devices which provide a compatible dma-controller, here
are devices I found in the current mainline kernel:- TI OMAP Soc Family
- TI Davinci Soc Family
- TI Keystone Processor Family
- IMX27 Processor and variants
- Several Microchip Processors (sama5, sam9x7, sam9x60)
Quoted 6 lines
That's too generic - you just listed SoCs, which consist of dozen or hundred of devices. Which hardware piece is here?
Maybe this is not for a real device, but then this should be marked clearly.
I listed devices that have a compatible dma-controller, so the list is a bit big, sorry. I also specifically mentioned the BeagleBone black board which I have been testing on.
"m2m-deinterlace" used to be a part of the mach-imx27_visstrim_m10.c board file, but was removed with commit 879c0e5e0ac711 (ARM: imx:
Remove i.MX27 board files). So at least the Vistrim M10 device was explicitly using the m2m-deinterlace device.
When the move away from board files was made towards device-tree, m2m- deinterlace support was never ported over to device-tree. This is what I am doing now.
And yes, m2m-deinterlace is not a "real device" if by "real device" you mean an actual piece of silicon on a specific piece of hardware. I think there is just some semantic confusion here. I will no longer refer to it as a "device" then, please let me know what the more appropriate term is and I will modify the description accordingly.
Quoted 10 lines
Quoted 6 lines
I think an appropriate analogy for m2m-deinterlace would be spi- gpio.
Since spi-gpio leverages gpio for bitbanging the spi protocol, the bindings do not need to describe any clocks, spi-controller registers,Sure, SPI GPIO is Linux driver, not a device and I am asking about it all the time.
My point was that spi-gpio has dt-bindings even though these bindings do not describe a specific hardware device, hence it is "generic".
Best,
Matthew
Re: [PATCH v2 1/2] media: dt-bindings: Add dt bindings for m2m-deinterlace device
Quoted 11 lines · Matthew Majewski
Quoted 10 lines · Krzysztof Kozlowski
Quoted 8 lines · Matthew Majewski
As I wrote, supported devices/hardware is anything that provides a
MEM_TO_MEMcapable dma-controller with interleaved transfer support. I
did not list specific devices because the bindings are supposed to be
generic, as they are not describing actual silicon. But if you want me
Quoted 7 lines
Quoted 3 lines
I already told you that no. Bindings are not supposed to be generic.
From where did you get such information?
There are generic bindings in the kernel and I based my bindings off of them.
spi-gpio.yaml,i2c-gpio.yaml,video-mux.yaml, etc are all generic bindings, no?
They are generic, but you said "supposed to be generic" and I am asking about this.
BTW, any "generic" or "simple" bindings often became non-generic and non-simple leading to mess.
Quoted 45 lines
Quoted 12 lines
Quoted 11 lines
to list some devices which provide a compatible dma-controller, here
are devices I found in the current mainline kernel:- TI OMAP Soc Family
- TI Davinci Soc Family
- TI Keystone Processor Family
- IMX27 Processor and variants
- Several Microchip Processors (sama5, sam9x7, sam9x60)Quoted 6 lines
That's too generic - you just listed SoCs, which consist of dozen or hundred of devices. Which hardware piece is here?
Maybe this is not for a real device, but then this should be marked clearly.
I listed devices that have a compatible dma-controller, so the list is a bit big, sorry. I also specifically mentioned the BeagleBone black board which I have been testing on.
"m2m-deinterlace" used to be a part of the
mach-imx27_visstrim_m10.cboard file, but was removed with commit879c0e5e0ac711(ARM: imx:
Remove i.MX27 board files). So at least the Vistrim M10 device was explicitly using the m2m-deinterlace device.When the move away from board files was made towards device-tree, m2m- deinterlace support was never ported over to device-tree. This is what I am doing now.
And yes, m2m-deinterlace is not a "real device" if by "real device" you mean an actual piece of silicon on a specific piece of hardware. I think there is just some semantic confusion here. I will no longer refer to it as a "device" then, please let me know what the more appropriate term is and I will modify the description accordingly.
Quoted 7 lines
Quoted 6 lines
I think an appropriate analogy for m2m-deinterlace would be spi- gpio.
Since spi-gpio leverages gpio for bitbanging the spi protocol, the bindings do not need to describe any clocks, spi-controller registers,
Quoted 6 lines
Quoted 3 lines
Sure, SPI GPIO is Linux driver, not a device and I am asking about it all the time.
My point was that spi-gpio has dt-bindings even though these bindings do not describe a specific hardware device, hence it is "generic".
SPI GPIO is physical GPIO lines bit-banged in software. I don't think it is possible to achieve such SPI controller as part of some other Linux structure.
Now your case - describe what you have here - SW for interlacing? But that is just memory operation. I really do not see why this is supposed to be separate device node in DTS. This is part of media pipeline, so easily can be part of some other device.
Best regards,
Krzysztof
PATCH v2 2/2 media: m2m-deinterlace: add device-tree support
In reply to Matthew MajewskiAdd support for loading the m2m-deinterlace driver through the device-tree.
Prefer to query the dma device through dma_request_chan(), which will get the dma device information from the device-tree node if present. Otherwise, fall back to the original dma_request_channel() to preserve backwards compatibility.
--- drivers/media/platform/m2m-deinterlace.c | 25 ++++++++++++++++++------ 1 file changed, 19 insertions(+), 6 deletions(-)
drivers/media/platform/m2m-deinterlace.c +19 -6
@@ -11,6 +11,7 @@#include <linux/interrupt.h> #include <linux/dmaengine.h> #include <linux/platform_device.h> +#include <linux/module.h> #include <media/v4l2-mem2mem.h> #include <media/v4l2-device.h>
@@ -914,7 +915,6 @@ static int deinterlace_probe(struct platform_device *pdev) {struct deinterlace_dev *pcdev;struct video_device *vfd; -dma_cap_mask_t mask;int ret =0 ; pcdev =devm_kzalloc (&pdev->dev,sizeof (*pcdev),GFP_KERNEL );
@@ -923,9 +923,16 @@ static int deinterlace_probe(struct platform_device *pdev)spin_lock_init (&pcdev->irqlock); -dma_cap_zero (mask); -dma_cap_set (DMA_INTERLEAVE , mask); - pcdev->dma_chan =dma_request_channel (mask,NULL , pcdev); +if (pdev->dev.of_node) { + pcdev->dma_chan =dma_request_chan (&pdev->dev,"rxtx" ); + }else { +dma_cap_mask_t mask; + +dma_cap_zero (mask); +dma_cap_set (DMA_INTERLEAVE , mask); + pcdev->dma_chan =dma_request_channel (mask,NULL , pcdev); + } +if (!pcdev->dma_chan)return -ENODEV ;
@@ -989,12 +996,18 @@ static void deinterlace_remove(struct platform_device *pdev)dma_release_channel (pcdev->dma_chan); } +static const struct of_device_id deinterlace_dt_match[] = { + { .compatible ="m2m-deinterlace" }, + {/* sentinel */ } +}; +MODULE_DEVICE_TABLE (of, deinterlace_dt_match); +static struct platform_driver deinterlace_pdrv = { .probe = deinterlace_probe, .remove = deinterlace_remove, .driver = { - .name =MEM2MEM_NAME , + .name =MEM2MEM_NAME , + .of_match_table = deinterlace_dt_match, }, };module_platform_driver (deinterlace_pdrv); -
Signature · 1 line
2.25.1