Skip to contentSkip to conversation

arm64: dts: imx8mp: Add easrc node

PATCH Applied Adam Ford 6 participants 16 messages Aug 31, 2023 – Oct 4, 2023

Subsystems (4)
lore
Share & tools
lore

Adam Ford

author
#0 lore raw

PATCH v3 1/3 arm64: dts: imx8mp: Add easrc node

The i.MX8MP has an asynchronous sample rate converter which seems to be the same as what is available on the i.MX8M Nano.

---
V3: No Change. The dt-binding update was already accepted into the sound tree, so that patch
was dropped from the series

V2: No Change.

arch/arm64/boot/dts/freescale/imx8mp.dtsi +20 -0
index 83d907294fbc..3167706d81e1 100644
@@ -1459,6 +1459,26 @@ sai7: sai@30c80000 {
 					interrupts = <GIC_SPI 111 IRQ_TYPE_LEVEL_HIGH>;
 					status = "disabled";
 				};
+
+				easrc: easrc@30c90000 {
+					compatible = "fsl,imx8mp-easrc", "fsl,imx8mn-easrc";
+					reg = <0x30c90000 0x10000>;
+					interrupts = <GIC_SPI 122 IRQ_TYPE_LEVEL_HIGH>;
+					clocks = <&audio_blk_ctrl IMX8MP_CLK_AUDIOMIX_ASRC_IPG>;
+					clock-names = "mem";
+					dmas = <&sdma2 16 23 0> , <&sdma2 17 23 0>,
+					       <&sdma2 18 23 0> , <&sdma2 19 23 0>,
+					       <&sdma2 20 23 0> , <&sdma2 21 23 0>,
+					       <&sdma2 22 23 0> , <&sdma2 23 23 0>;
+					dma-names = "ctx0_rx", "ctx0_tx",
+						    "ctx1_rx", "ctx1_tx",
+						    "ctx2_rx", "ctx2_tx",
+						    "ctx3_rx", "ctx3_tx";
+					firmware-name = "imx/easrc/easrc-imx8mn.bin";
+					fsl,asrc-rate  = <8000>;
+					fsl,asrc-format = <2>;
+					status = "disabled";
+				};
 			};
 
 			sdma3: dma-controller@30e00000 {
Signature · 1 line
2.39.2
author
#1 lore raw

PATCH v3 2/3 arm64: dts: imx8mp: Add micfil node

The i.MX8MP has a micfil controller which is used for interfacing with a pulse density microphone. Add the node and mark it as disabled by default.

---
V3: The AUDIOMIX_PDM_ROOT doesn't exist and the real clock is
called IMX8MP_CLK_AUDIOMIX_PDM_SEL, so swap it out.

V2: No change

arch/arm64/boot/dts/freescale/imx8mp.dtsi +21 -0
index 3167706d81e1..341fd0369ce9 100644
@@ -1479,6 +1479,27 @@ easrc: easrc@30c90000 {
 					fsl,asrc-format = <2>;
 					status = "disabled";
 				};
+
+				micfil: audio-controller@30ca0000 {
+					compatible = "fsl,imx8mp-micfil";
+					reg = <0x30ca0000 0x10000>;
+					#sound-dai-cells = <0>;
+					interrupts = <GIC_SPI 109 IRQ_TYPE_LEVEL_HIGH>,
+						     <GIC_SPI 110 IRQ_TYPE_LEVEL_HIGH>,
+						     <GIC_SPI 44 IRQ_TYPE_LEVEL_HIGH>,
+						     <GIC_SPI 45 IRQ_TYPE_LEVEL_HIGH>;
+					clocks = <&audio_blk_ctrl IMX8MP_CLK_AUDIOMIX_PDM_IPG>,
+						 <&audio_blk_ctrl IMX8MP_CLK_AUDIOMIX_PDM_SEL>,
+						 <&clk IMX8MP_AUDIO_PLL1_OUT>,
+						 <&clk IMX8MP_AUDIO_PLL2_OUT>,
+						 <&clk IMX8MP_CLK_EXT3>;
+					clock-names = "ipg_clk", "ipg_clk_app",
+						      "pll8k", "pll11k", "clkext3";
+					dmas = <&sdma2 24 25 0x80000000>;
+					dma-names = "rx";
+					status = "disabled";
+				};
+
 			};
 
 			sdma3: dma-controller@30e00000 {
Signature · 1 line
2.39.2
#2 lore raw

Re: [PATCH V3 2/3] arm64: dts: imx8mp: Add micfil node

Quoted 27 lines · Adam Ford

The i.MX8MP has a micfil controller which is used for interfacing with a pulse density microphone. Add the node and mark it as disabled by default.

---
V3: The AUDIOMIX_PDM_ROOT doesn't exist and the real clock is
called IMX8MP_CLK_AUDIOMIX_PDM_SEL, so swap it out.

V2: No change

arch/arm64/boot/dts/freescale/imx8mp.dtsi +8 -0
index 3167706d81e1..341fd0369ce9 100644
@@ -1479,6 +1479,27 @@ easrc: easrc@30c90000 {
  					fsl,asrc-format = <2>;
  					status = "disabled";
  				};
+
+				micfil: audio-controller@30ca0000 {
+					compatible = "fsl,imx8mp-micfil";
+					reg = <0x30ca0000 0x10000>;
+					#sound-dai-cells = <0>;
+					interrupts = <GIC_SPI 109 IRQ_TYPE_LEVEL_HIGH>,
+						     <GIC_SPI 110 IRQ_TYPE_LEVEL_HIGH>,
+						     <GIC_SPI 44 IRQ_TYPE_LEVEL_HIGH>,

arch/arm64/boot/dts/freescale/imx8mp.dtsi easrc: easrc@30c90000 { · line 1499

Quoted 10 lines
arch/arm64/boot/dts/freescale/imx8mp.dtsi +10 -0
+						     <GIC_SPI 45 IRQ_TYPE_LEVEL_HIGH>;
+					clocks = <&audio_blk_ctrl IMX8MP_CLK_AUDIOMIX_PDM_IPG>,
+						 <&audio_blk_ctrl IMX8MP_CLK_AUDIOMIX_PDM_SEL>,
+						 <&clk IMX8MP_AUDIO_PLL1_OUT>,
+						 <&clk IMX8MP_AUDIO_PLL2_OUT>,
+						 <&clk IMX8MP_CLK_EXT3>;
+					clock-names = "ipg_clk", "ipg_clk_app",
+						      "pll8k", "pll11k", "clkext3";
+					dmas = <&sdma2 24 25 0x80000000>;
+					dma-names = "rx";

Is dma-names really required if there is only a single DMA channel in DT ?

author
#3 lore raw

Re: [PATCH V3 2/3] arm64: dts: imx8mp: Add micfil node

Quoted 33 lines · Marek Vasut
Quoted 31 lines · Adam Ford

The i.MX8MP has a micfil controller which is used for interfacing with a pulse density microphone. Add the node and mark it as disabled by default.

---
V3: The AUDIOMIX_PDM_ROOT doesn't exist and the real clock is
called IMX8MP_CLK_AUDIOMIX_PDM_SEL, so swap it out.

V2: No change

arch/arm64/boot/dts/freescale/imx8mp.dtsi +12 -0
index 3167706d81e1..341fd0369ce9 100644
@@ -1479,6 +1479,27 @@ easrc: easrc@30c90000 {
                                      fsl,asrc-format = <2>;
                                      status = "disabled";
                              };
+
+                             micfil: audio-controller@30ca0000 {
+                                     compatible = "fsl,imx8mp-micfil";
+                                     reg = <0x30ca0000 0x10000>;
+                                     #sound-dai-cells = <0>;
+                                     interrupts = <GIC_SPI 109 IRQ_TYPE_LEVEL_HIGH>,
+                                                  <GIC_SPI 110 IRQ_TYPE_LEVEL_HIGH>,
+                                                  <GIC_SPI 44 IRQ_TYPE_LEVEL_HIGH>,
+                                                  <GIC_SPI 45 IRQ_TYPE_LEVEL_HIGH>;
+                                     clocks = <&audio_blk_ctrl IMX8MP_CLK_AUDIOMIX_PDM_IPG>,
+                                              <&audio_blk_ctrl IMX8MP_CLK_AUDIOMIX_PDM_SEL>,
+                                              <&clk IMX8MP_AUDIO_PLL1_OUT>,
Quoted 8 lines

arch/arm64/boot/dts/freescale/imx8mp.dtsi easrc: easrc@30c90000 { · line 1499

Quoted 6 lines
arch/arm64/boot/dts/freescale/imx8mp.dtsi +6 -0
+                                              <&clk IMX8MP_AUDIO_PLL2_OUT>,
+                                              <&clk IMX8MP_CLK_EXT3>;
+                                     clock-names = "ipg_clk", "ipg_clk_app",
+                                                   "pll8k", "pll11k", "clkext3";
+                                     dmas = <&sdma2 24 25 0x80000000>;
+                                     dma-names = "rx";

Is dma-names really required if there is only a single DMA channel in DT ?

I would normally agree with you, the DT binding file shows it's required, and the driver looks like it's searching for a channel name called 'rx'

adam

#4 lore raw

Re: [PATCH V3 2/3] arm64: dts: imx8mp: Add micfil node

Quoted 43 lines · Adam Ford
Quoted 33 lines · Marek Vasut
Quoted 31 lines · Adam Ford

The i.MX8MP has a micfil controller which is used for interfacing with a pulse density microphone. Add the node and mark it as disabled by default.

---
V3: The AUDIOMIX_PDM_ROOT doesn't exist and the real clock is
called IMX8MP_CLK_AUDIOMIX_PDM_SEL, so swap it out.

V2: No change

arch/arm64/boot/dts/freescale/imx8mp.dtsi +12 -0
index 3167706d81e1..341fd0369ce9 100644
@@ -1479,6 +1479,27 @@ easrc: easrc@30c90000 {
                                       fsl,asrc-format = <2>;
                                       status = "disabled";
                               };
+
+                             micfil: audio-controller@30ca0000 {
+                                     compatible = "fsl,imx8mp-micfil";
+                                     reg = <0x30ca0000 0x10000>;
+                                     #sound-dai-cells = <0>;
+                                     interrupts = <GIC_SPI 109 IRQ_TYPE_LEVEL_HIGH>,
+                                                  <GIC_SPI 110 IRQ_TYPE_LEVEL_HIGH>,
+                                                  <GIC_SPI 44 IRQ_TYPE_LEVEL_HIGH>,
+                                                  <GIC_SPI 45 IRQ_TYPE_LEVEL_HIGH>;
+                                     clocks = <&audio_blk_ctrl IMX8MP_CLK_AUDIOMIX_PDM_IPG>,
+                                              <&audio_blk_ctrl IMX8MP_CLK_AUDIOMIX_PDM_SEL>,
+                                              <&clk IMX8MP_AUDIO_PLL1_OUT>,
Quoted 8 lines

arch/arm64/boot/dts/freescale/imx8mp.dtsi easrc: easrc@30c90000 { · line 1499

Quoted 6 lines
arch/arm64/boot/dts/freescale/imx8mp.dtsi +6 -0
+                                              <&clk IMX8MP_AUDIO_PLL2_OUT>,
+                                              <&clk IMX8MP_CLK_EXT3>;
+                                     clock-names = "ipg_clk", "ipg_clk_app",
+                                                   "pll8k", "pll11k", "clkext3";
+                                     dmas = <&sdma2 24 25 0x80000000>;
+                                     dma-names = "rx";

Is dma-names really required if there is only a single DMA channel in DT ?

Quoted 3 lines

I would normally agree with you, the DT binding file shows it's required, and the driver looks like it's searching for a channel name called 'rx'

Maybe something that can be improved both in the driver and bindings ?

_______________________________________________
linux-arm-kernel mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

Applied
#5 lore raw

Re: [PATCH V3 2/3] arm64: dts: imx8mp: Add micfil node

In reply to Adam Ford
Quoted 5 lines · Adam Ford

The i.MX8MP has a micfil controller which is used for interfacing with a pulse density microphone. Add the node and mark it as disabled by default.

Applied, thanks!

reviewer
#6 lore raw

Re: [PATCH V3 2/3] arm64: dts: imx8mp: Add micfil node

In reply to Adam Ford

Hi Adam and devicetree folks,

Quoted 17 lines · Adam Ford

The i.MX8MP has a micfil controller which is used for interfacing with a pulse density microphone. Add the node and mark it as disabled by default.

---
V3: The AUDIOMIX_PDM_ROOT doesn't exist and the real clock is
called IMX8MP_CLK_AUDIOMIX_PDM_SEL, so swap it out.

V2: No change

arch/arm64/boot/dts/freescale/imx8mp.dtsi +0 -0
index 3167706d81e1..341fd0369ce9 100644

arch/arm64/boot/dts/freescale/imx8mp.dtsi easrc: easrc@30c90000 { · line 1486

Quoted 8 lines
arch/arm64/boot/dts/freescale/imx8mp.dtsi +5 -0
@@ -1479,6 +1479,27 @@ easrc: easrc@30c90000 {
                                        fsl,asrc-format = <2>;
                                        status = "disabled";
                                };
+
+                               micfil: audio-controller@30ca0000 {
+                                       compatible = "fsl,imx8mp-micfil";
+                                       reg = <0x30ca0000 0x10000>;
+                                       #sound-dai-cells = <0>;

After this patch, the following schema warning is seen:

imx8mp-beacon-kit.dtb: audio-controller@30ca0000: '#sound-dai-cells' does not match any of the regexes: 'pinctrl-[0-9]+'
from schema $id: http://devicetree.org/schemas/sound/fsl,micfil.yaml#

What is the correct way to solve this?

- Document #sound-dai-cells in fsl,micfil.yaml as an optional property
- Remove #sound-dai-cells from imx8mp.dtsi
- Document #sound-dai-cells in fsl,micfil.yaml as a required property and pass #sound-dai-cells to imx8mm/imx8mn/imx83.dtsi?

author
#7 lore raw

Re: [PATCH V3 2/3] arm64: dts: imx8mp: Add micfil node

Quoted 36 lines · Fabio Estevam

Hi Adam and devicetree folks,

Quoted 17 lines · Adam Ford

The i.MX8MP has a micfil controller which is used for interfacing with a pulse density microphone. Add the node and mark it as disabled by default.

---
V3: The AUDIOMIX_PDM_ROOT doesn't exist and the real clock is
called IMX8MP_CLK_AUDIOMIX_PDM_SEL, so swap it out.

V2: No change

arch/arm64/boot/dts/freescale/imx8mp.dtsi +0 -0
index 3167706d81e1..341fd0369ce9 100644

arch/arm64/boot/dts/freescale/imx8mp.dtsi easrc: easrc@30c90000 { · line 1486

Quoted 8 lines
arch/arm64/boot/dts/freescale/imx8mp.dtsi +5 -0
@@ -1479,6 +1479,27 @@ easrc: easrc@30c90000 {
                                        fsl,asrc-format = <2>;
                                        status = "disabled";
                                };
+
+                               micfil: audio-controller@30ca0000 {
+                                       compatible = "fsl,imx8mp-micfil";
+                                       reg = <0x30ca0000 0x10000>;
+                                       #sound-dai-cells = <0>;

After this patch, the following schema warning is seen:

imx8mp-beacon-kit.dtb: audio-controller@30ca0000: '#sound-dai-cells' does not match any of the regexes: 'pinctrl-[0-9]+'
from schema $id: http://devicetree.org/schemas/sound/fsl,micfil.yaml#

Quoted 6 lines

What is the correct way to solve this?

- Document #sound-dai-cells in fsl,micfil.yaml as an optional property
- Remove #sound-dai-cells from imx8mp.dtsi
- Document #sound-dai-cells in fsl,micfil.yaml as a required property and pass #sound-dai-cells to imx8mm/imx8mn/imx83.dtsi?

I am not an expert on yaml, in fact they quite confuse me, but I believe the proper solution is likely to document them in the yaml. When I built the device tree without the sound-dai-cells, it threw a message indicating that it needed to be there when it is configured to operate. I didn't see the schema warning that you noted, but my schema could have been out of date. Looking at the sai nodes and the yaml file, it appears that the sound-dai-cells is listed in the yaml.

adam

author
#8 lore raw

PATCH v3 3/3 clk: imx8mp: Remove non-existent IMX8MP_CLK_AUDIOMIX_PDM_ROOT

In reply to Adam Ford

The TRM shows there is only one AUDIOMIX PDM Root Clock Select register, and it's called IMX8MP_CLK_AUDIOMIX_PDM_SEL. That selector doesn't appear to have any more children and the MICFIL driver can reference IMX8MP_CLK_AUDIOMIX_PDM_SEL directly without the need for any other. Remove this errant clock, since it doesn't really exist.

---

V3: New to series

include/dt-bindings/clock/imx8mp-clock.h +0 -1
index 11cb0a4fe999..7da4243984b2 100644
@@ -376,7 +376,6 @@
 #define IMX8MP_CLK_AUDIOMIX_MU2_ROOT		36
 #define IMX8MP_CLK_AUDIOMIX_MU3_ROOT		37
 #define IMX8MP_CLK_AUDIOMIX_EARC_PHY		38
-#define IMX8MP_CLK_AUDIOMIX_PDM_ROOT		39
 #define IMX8MP_CLK_AUDIOMIX_SAI1_MCLK1_SEL	40
 #define IMX8MP_CLK_AUDIOMIX_SAI1_MCLK2_SEL	41
 #define IMX8MP_CLK_AUDIOMIX_SAI2_MCLK1_SEL	42
Signature · 1 line
2.39.2
maintainer Acked-by on v3 3/3
#9 lore raw

Re: [PATCH V3 3/3] clk: imx8mp: Remove non-existent IMX8MP_CLK_AUDIOMIX_PDM_ROOT

Quoted 11 lines · Adam Ford

The TRM shows there is only one AUDIOMIX PDM Root Clock Select register, and it's called IMX8MP_CLK_AUDIOMIX_PDM_SEL. That selector doesn't appear to have any more children and the MICFIL driver can reference IMX8MP_CLK_AUDIOMIX_PDM_SEL directly without the need for any other. Remove this errant clock, since it doesn't really exist.

---

V3: New to series

#10 lore raw
In reply to Adam Ford
Quoted 23 lines · Adam Ford

The i.MX8MP has an asynchronous sample rate converter which seems to be the same as what is available on the i.MX8M Nano.

---
V3: No Change. The dt-binding update was already accepted into the sound tree, so that patch
was dropped from the series

V2: No Change.

arch/arm64/boot/dts/freescale/imx8mp.dtsi +6 -0
index 83d907294fbc..3167706d81e1 100644
@@ -1459,6 +1459,26 @@ sai7: sai@30c80000 {
  					interrupts = <GIC_SPI 111 IRQ_TYPE_LEVEL_HIGH>;
  					status = "disabled";
  				};
+
+				easrc: easrc@30c90000 {
+					compatible = "fsl,imx8mp-easrc", "fsl,imx8mn-easrc";
+					reg = <0x30c90000 0x10000>;
+					interrupts = <GIC_SPI 122 IRQ_TYPE_LEVEL_HIGH>;
+					clocks = <&audio_blk_ctrl IMX8MP_CLK_AUDIOMIX_ASRC_IPG>;

arch/arm64/boot/dts/freescale/imx8mp.dtsi sai7: sai@30c80000 { · line 1477

Quoted 10 lines
arch/arm64/boot/dts/freescale/imx8mp.dtsi +10 -0
+					clock-names = "mem";
+					dmas = <&sdma2 16 23 0> , <&sdma2 17 23 0>,
+					       <&sdma2 18 23 0> , <&sdma2 19 23 0>,
+					       <&sdma2 20 23 0> , <&sdma2 21 23 0>,
+					       <&sdma2 22 23 0> , <&sdma2 23 23 0>;
+					dma-names = "ctx0_rx", "ctx0_tx",
+						    "ctx1_rx", "ctx1_tx",
+						    "ctx2_rx", "ctx2_tx",
+						    "ctx3_rx", "ctx3_tx";
+					firmware-name = "imx/easrc/easrc-imx8mn.bin";

Should the firmware name really be imx8mn or should it be imx8mp ?

I think the later is better, you can always add a symlink for the firmware name, and you can discern mx8mn/mp firmwares on the same rootfs in case that was ever needed in the future.

author
#11 lore raw
Quoted 29 lines · Marek Vasut
Quoted 27 lines · Adam Ford

The i.MX8MP has an asynchronous sample rate converter which seems to be the same as what is available on the i.MX8M Nano.

---
V3: No Change. The dt-binding update was already accepted into the sound tree, so that patch
was dropped from the series

V2: No Change.

arch/arm64/boot/dts/freescale/imx8mp.dtsi +10 -0
index 83d907294fbc..3167706d81e1 100644
@@ -1459,6 +1459,26 @@ sai7: sai@30c80000 {
                                      interrupts = <GIC_SPI 111 IRQ_TYPE_LEVEL_HIGH>;
                                      status = "disabled";
                              };
+
+                             easrc: easrc@30c90000 {
+                                     compatible = "fsl,imx8mp-easrc", "fsl,imx8mn-easrc";
+                                     reg = <0x30c90000 0x10000>;
+                                     interrupts = <GIC_SPI 122 IRQ_TYPE_LEVEL_HIGH>;
+                                     clocks = <&audio_blk_ctrl IMX8MP_CLK_AUDIOMIX_ASRC_IPG>;
+                                     clock-names = "mem";
+                                     dmas = <&sdma2 16 23 0> , <&sdma2 17 23 0>,
+                                            <&sdma2 18 23 0> , <&sdma2 19 23 0>,
+                                            <&sdma2 20 23 0> , <&sdma2 21 23 0>,
Quoted 8 lines

arch/arm64/boot/dts/freescale/imx8mp.dtsi sai7: sai@30c80000 { · line 1477

Quoted 6 lines
arch/arm64/boot/dts/freescale/imx8mp.dtsi +6 -0
+                                            <&sdma2 22 23 0> , <&sdma2 23 23 0>;
+                                     dma-names = "ctx0_rx", "ctx0_tx",
+                                                 "ctx1_rx", "ctx1_tx",
+                                                 "ctx2_rx", "ctx2_tx",
+                                                 "ctx3_rx", "ctx3_tx";
+                                     firmware-name = "imx/easrc/easrc-imx8mn.bin";

Should the firmware name really be imx8mn or should it be imx8mp ?

The firmware blobs provided by NXP only has easrc-imx8mn.bin and would be concerned that having a different name might confuse people if they try to go look for it and they only find easrc-imx8mn.bin. To me, this is no different than the SDMA driver using firmware called sdma-imx7d.bin

Quoted 4 lines

I think the later is better, you can always add a symlink for the firmware name, and you can discern mx8mn/mp firmwares on the same rootfs in case that was ever needed in the future.

#12 lore raw
Quoted 39 lines · Adam Ford
Quoted 29 lines · Marek Vasut
Quoted 27 lines · Adam Ford

The i.MX8MP has an asynchronous sample rate converter which seems to be the same as what is available on the i.MX8M Nano.

---
V3: No Change. The dt-binding update was already accepted into the sound tree, so that patch
was dropped from the series

V2: No Change.

arch/arm64/boot/dts/freescale/imx8mp.dtsi +10 -0
index 83d907294fbc..3167706d81e1 100644
@@ -1459,6 +1459,26 @@ sai7: sai@30c80000 {
                                       interrupts = <GIC_SPI 111 IRQ_TYPE_LEVEL_HIGH>;
                                       status = "disabled";
                               };
+
+                             easrc: easrc@30c90000 {
+                                     compatible = "fsl,imx8mp-easrc", "fsl,imx8mn-easrc";
+                                     reg = <0x30c90000 0x10000>;
+                                     interrupts = <GIC_SPI 122 IRQ_TYPE_LEVEL_HIGH>;
+                                     clocks = <&audio_blk_ctrl IMX8MP_CLK_AUDIOMIX_ASRC_IPG>;
+                                     clock-names = "mem";
+                                     dmas = <&sdma2 16 23 0> , <&sdma2 17 23 0>,
+                                            <&sdma2 18 23 0> , <&sdma2 19 23 0>,
+                                            <&sdma2 20 23 0> , <&sdma2 21 23 0>,
Quoted 8 lines

arch/arm64/boot/dts/freescale/imx8mp.dtsi sai7: sai@30c80000 { · line 1477

Quoted 6 lines
arch/arm64/boot/dts/freescale/imx8mp.dtsi +6 -0
+                                            <&sdma2 22 23 0> , <&sdma2 23 23 0>;
+                                     dma-names = "ctx0_rx", "ctx0_tx",
+                                                 "ctx1_rx", "ctx1_tx",
+                                                 "ctx2_rx", "ctx2_tx",
+                                                 "ctx3_rx", "ctx3_tx";
+                                     firmware-name = "imx/easrc/easrc-imx8mn.bin";

Should the firmware name really be imx8mn or should it be imx8mp ?

Quoted 5 lines

The firmware blobs provided by NXP only has easrc-imx8mn.bin and would be concerned that having a different name might confuse people if they try to go look for it and they only find easrc-imx8mn.bin. To me, this is no different than the SDMA driver using firmware called sdma-imx7d.bin

On the other hand, there is "fsl,imx8mp-easrc" compatible string -- to differentiate between IPs in case some difference between the IPs that cannot be predicted so far is found later on.

Quoted 3 lines
Quoted 3 lines

I think the later is better, you can always add a symlink for the firmware name, and you can discern mx8mn/mp firmwares on the same rootfs in case that was ever needed in the future.

Which is what this part of my comment ^ is all about.

_______________________________________________
linux-arm-kernel mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

#13 lore raw
In reply to Adam Ford
Quoted 24 lines · Adam Ford

The i.MX8MP has an asynchronous sample rate converter which seems to be the same as what is available on the i.MX8M Nano.

---
V3: No Change. The dt-binding update was already accepted into the sound tree, so that patch
was dropped from the series

V2: No Change.

arch/arm64/boot/dts/freescale/imx8mp.dtsi +7 -0
index 83d907294fbc..3167706d81e1 100644
@@ -1459,6 +1459,26 @@ sai7: sai@30c80000 {
 					interrupts = <GIC_SPI 111 IRQ_TYPE_LEVEL_HIGH>;
 					status = "disabled";
 				};
+
+				easrc: easrc@30c90000 {
+					compatible = "fsl,imx8mp-easrc", "fsl,imx8mn-easrc";
+					reg = <0x30c90000 0x10000>;
+					interrupts = <GIC_SPI 122 IRQ_TYPE_LEVEL_HIGH>;
+					clocks = <&audio_blk_ctrl IMX8MP_CLK_AUDIOMIX_ASRC_IPG>;
+					clock-names = "mem";

arch/arm64/boot/dts/freescale/imx8mp.dtsi sai7: sai@30c80000 { · line 1478

Quoted 10 lines
arch/arm64/boot/dts/freescale/imx8mp.dtsi +10 -0
+					dmas = <&sdma2 16 23 0> , <&sdma2 17 23 0>,
+					       <&sdma2 18 23 0> , <&sdma2 19 23 0>,
+					       <&sdma2 20 23 0> , <&sdma2 21 23 0>,
+					       <&sdma2 22 23 0> , <&sdma2 23 23 0>;
+					dma-names = "ctx0_rx", "ctx0_tx",
+						    "ctx1_rx", "ctx1_tx",
+						    "ctx2_rx", "ctx2_tx",
+						    "ctx3_rx", "ctx3_tx";
+					firmware-name = "imx/easrc/easrc-imx8mn.bin";
+					fsl,asrc-rate  = <8000>;

One space before =

Shawn

Quoted 9 lines
arch/arm64/boot/dts/freescale/imx8mp.dtsi +3 -0
+					fsl,asrc-format = <2>;
+					status = "disabled";
+				};
 			};
 
 			sdma3: dma-controller@30e00000 {
Signature · 1 line
2.39.2
#14 lore raw
Quoted 30 lines · Shawn Guo
Quoted 28 lines · Adam Ford

The i.MX8MP has an asynchronous sample rate converter which seems to be the same as what is available on the i.MX8M Nano.

---
V3: No Change. The dt-binding update was already accepted into the sound tree, so that patch
was dropped from the series

V2: No Change.

arch/arm64/boot/dts/freescale/imx8mp.dtsi +11 -0
index 83d907294fbc..3167706d81e1 100644
@@ -1459,6 +1459,26 @@ sai7: sai@30c80000 {
                                      interrupts = <GIC_SPI 111 IRQ_TYPE_LEVEL_HIGH>;
                                      status = "disabled";
                              };
+
+                             easrc: easrc@30c90000 {
+                                     compatible = "fsl,imx8mp-easrc", "fsl,imx8mn-easrc";
+                                     reg = <0x30c90000 0x10000>;
+                                     interrupts = <GIC_SPI 122 IRQ_TYPE_LEVEL_HIGH>;
+                                     clocks = <&audio_blk_ctrl IMX8MP_CLK_AUDIOMIX_ASRC_IPG>;
+                                     clock-names = "mem";
+                                     dmas = <&sdma2 16 23 0> , <&sdma2 17 23 0>,
+                                            <&sdma2 18 23 0> , <&sdma2 19 23 0>,
+                                            <&sdma2 20 23 0> , <&sdma2 21 23 0>,
+                                            <&sdma2 22 23 0> , <&sdma2 23 23 0>;
Quoted 8 lines

arch/arm64/boot/dts/freescale/imx8mp.dtsi sai7: sai@30c80000 { · line 1478

Quoted 6 lines
arch/arm64/boot/dts/freescale/imx8mp.dtsi +6 -0
+                                     dma-names = "ctx0_rx", "ctx0_tx",
+                                                 "ctx1_rx", "ctx1_tx",
+                                                 "ctx2_rx", "ctx2_tx",
+                                                 "ctx3_rx", "ctx3_tx";
+                                     firmware-name = "imx/easrc/easrc-imx8mn.bin";
+                                     fsl,asrc-rate  = <8000>;

One space before =

I fixed it up and applied, thanks!

Applied
#15 lore raw

Re: (subset) [PATCH V3 1/3] arm64: dts: imx8mp: Add easrc node

In reply to Adam Ford
Quoted 4 lines · Adam Ford

The i.MX8MP has an asynchronous sample rate converter which seems to be the same as what is available on the i.MX8M Nano.

Applied, thanks!

[3/3] clk: imx8mp: Remove non-existent IMX8MP_CLK_AUDIOMIX_PDM_ROOT

      commit: 05eeeff22b216e871be76dcf216a2fdda48880e2

Best regards,

Signature · 1 line
Abel Vesa <[email protected]>