Skip to contentSkip to conversation

Enable multiple MCAN on AM62x

PATCH Superseded Judith Mendez 5 participants 15 messages May 1, 2023 – May 10, 2023

Subsystems (6)

v4 of 10 Superseded Latest is v10 →

lore
Share & tools
lore
#0 lore raw

v4 cover Enable multiple MCAN on AM62x

On AM62x there is one MCAN in MAIN domain and two in MCU domain. The MCANs in MCU domain were not enabled since there is no hardware interrupt routed to A53 GIC interrupt controller. Therefore A53 Linux cannot be interrupted by MCU MCANs.

This solution instantiates a hrtimer with 1 ms polling interval for MCAN device when there is no hardware interrupt and there is poll-interval property in DTB MCAN node. The hrtimer generates a recurring software interrupt which allows to call the isr. The isr will check if there is pending transaction by reading a register and proceed normally if there is.

On AM62x, this series enables two MCU MCAN which will use the hrtimer implementation. MCANs with hardware interrupt routed to A53 Linux will continue to use the hardware interrupt as expected.

Timer polling method was tested on both classic CAN and CAN-FD at 125 KBPS, 250 KBPS, 1 MBPS and 2.5 MBPS with 4 MBPS bitrate switching.

Letency and CPU load benchmarks were tested on 3x MCAN on AM62x. 1 MBPS timer polling interval is the better timer polling interval since it has comparable latency to hardware interrupt with the worse case being 1ms + CAN frame propagation time and CPU load is not substantial. Latency can be improved further with less than 1 ms polling intervals, howerver it is at the cost of CPU usage since CPU load increases at 0.5 ms.

Note that in terms of power, enabling MCU MCANs with timer-polling implementation might have negative impact since we will have to wake up every 1 ms whether there are CAN packets pending in the RX FIFO or not. This might prevent the CPU from entering into deeper idle states for extended periods of time.

This patch series depends on 'Enable CAN PHY transceiver driver':

v2:

V1:

RFC:

Changes since v3:
- Wrong patch sent, resend correct patch series

Changes since v2:
- Change binding patch first
- Update binding poll-interval description
- Add oneOf to select either interrupts/interrupt-names or poll-interval
- Sort list of includes
- Create a define for 1 ms polling interval
- Change plarform_get_irq to optional to not print error msg
- Fix indentations, lengths of code lines, and added other style changes

Changes since v1:
- Add poll-interval property to bindings and MCAN DTB node
- Add functionality to check for 'poll-interval' property in MCAN node
- Bindings: add an example using poll-interval
- Add 'polling' flag in driver to check if device is using polling method
- Check for both timer polling and hardware interrupt case, default to hardware interrupt method
- Change ns_to_ktime() to ms_to_ktime()

Judith Mendez (4):
dt-bindings: net: can: Add poll-interval for MCAN
can: m_can: Add hrtimer to generate software interrupt
arm64: dts: ti: Add AM62x MCAN MAIN domain transceiver overlay
arm64: dts: ti: Enable MCU MCANs for AM62x

 .../bindings/net/can/bosch,m_can.yaml         | 36 +++++++++++-
 arch/arm64/boot/dts/ti/Makefile               |  2 +
 arch/arm64/boot/dts/ti/k3-am62-mcu.dtsi       | 24 ++++++++
 .../boot/dts/ti/k3-am625-sk-mcan-main.dtso    | 35 ++++++++++++
 .../boot/dts/ti/k3-am625-sk-mcan-mcu.dtso     | 57 +++++++++++++++++++
 drivers/net/can/m_can/m_can.c                 | 29 +++++++++-
 drivers/net/can/m_can/m_can.h                 |  4 ++
 drivers/net/can/m_can/m_can_platform.c        | 32 ++++++++++-
 8 files changed, 212 insertions(+), 7 deletions(-)
 create mode 100644 arch/arm64/boot/dts/ti/k3-am625-sk-mcan-main.dtso
 create mode 100644 arch/arm64/boot/dts/ti/k3-am625-sk-mcan-mcu.dtso

base-commit: 92e815cf07ed24ee1c51b122f24ffcf2964b4b13

Signature · 1 line
2.17.1
#1 lore raw

PATCH v4 1/4 dt-bindings: net: can: Add poll-interval for MCAN

On AM62x SoC, MCANs on MCU domain do not have hardware interrupt routed to A53 Linux, instead they will use software interrupt by hrtimer. To enable timer method, interrupts should be optional so remove interrupts property from required section and introduce poll-interval property.

---
Changelog:
v3:
1. Move binding patch to first in series
2. Update description for poll-interval
3. Add oneOf to specify using interrupts/interrupt-names or poll-interval
4. Fix example property: add comment below 'example'

v2:
1. Add poll-interval property to enable timer polling method
2. Add example using poll-interval property

 .../bindings/net/can/bosch,m_can.yaml         | 36 +++++++++++++++++--
 1 file changed, 34 insertions(+), 2 deletions(-)
Documentation/devicetree/bindings/net/can/bosch,m_can.yaml +34 -2
index 67879aab623b..c024ee49962c 100644
@@ -14,6 +14,13 @@ maintainers:
 allOf:
   - $ref: can-controller.yaml#
 
+oneOf:
+  - required:
+      - interrupts
+      - interrupt-names
+  - required:
+      - poll-interval
+
 properties:
   compatible:
     const: bosch,m_can
@@ -40,6 +47,14 @@ properties:
       - const: int1
     minItems: 1
 
+  poll-interval:
+    $ref: /schemas/types.yaml#/definitions/flag
+    description: Enable hrtimer polling method for an M_CAN device.
+      If this property is defined in MCAN node, it tells the driver to
+      enable polling method for an MCAN device. If for an MCAN device,
+      hardware interrupt is found and hrtimer polling method is enabled,
+      the driver will use hardware interrupt method.
+
   clocks:
     items:
       - description: peripheral clock
@@ -122,8 +137,6 @@ required:
   - compatible
   - reg
   - reg-names
-  - interrupts
-  - interrupt-names
   - clocks
   - clock-names
   - bosch,mram-cfg
@@ -132,6 +145,7 @@ additionalProperties: false
 
 examples:
   - |
+    // Example with interrupts
     #include <dt-bindings/clock/imx6sx-clock.h>
     can@20e8000 {
       compatible = "bosch,m_can";
@@ -149,4 +163,22 @@ examples:
       };
     };
 
+  - |
+    // Example with timer polling
+    #include <dt-bindings/clock/imx6sx-clock.h>
+    can@20e8000 {
+      compatible = "bosch,m_can";
+      reg = <0x020e8000 0x4000>, <0x02298000 0x4000>;
+      reg-names = "m_can", "message_ram";
+      poll-interval;
+      clocks = <&clks IMX6SX_CLK_CANFD>,
+               <&clks IMX6SX_CLK_CANFD>;
+      clock-names = "hclk", "cclk";
+      bosch,mram-cfg = <0x0 0 0 32 0 0 0 1>;
+
+      can-transceiver {
+        max-bitrate = <5000000>;
+      };
+    };
+
 ...
Signature · 1 line
2.17.1
maintainer
#2 lore raw

Re: [PATCH v4 1/4] dt-bindings: net: can: Add poll-interval for MCAN

Quoted 27 lines · Judith Mendez

On AM62x SoC, MCANs on MCU domain do not have hardware interrupt routed to A53 Linux, instead they will use software interrupt by hrtimer. To enable timer method, interrupts should be optional so remove interrupts property from required section and introduce poll-interval property.

---
Changelog:
v3:
1. Move binding patch to first in series
2. Update description for poll-interval
3. Add oneOf to specify using interrupts/interrupt-names or poll-interval
4. Fix example property: add comment below 'example'

v2:
1. Add poll-interval property to enable timer polling method
2. Add example using poll-interval property

 .../bindings/net/can/bosch,m_can.yaml         | 36 +++++++++++++++++--
 1 file changed, 34 insertions(+), 2 deletions(-)
Documentation/devicetree/bindings/net/can/bosch,m_can.yaml +0 -0
index 67879aab623b..c024ee49962c 100644

Documentation/devicetree/bindings/net/can/bosch,m_can.yaml maintainers: · line 22

Quoted 9 lines
Documentation/devicetree/bindings/net/can/bosch,m_can.yaml +6 -0
@@ -14,6 +14,13 @@ maintainers:
 allOf:
   - $ref: can-controller.yaml#
 
+oneOf:
+  - required:
+      - interrupts
+      - interrupt-names
+  - required:
+      - poll-interval

Move this next to 'required'.

Documentation/devicetree/bindings/net/can/bosch,m_can.yaml properties: · line 51

Quoted 10 lines
Documentation/devicetree/bindings/net/can/bosch,m_can.yaml +3 -0
+
 properties:
   compatible:
     const: bosch,m_can
@@ -40,6 +47,14 @@ properties:
       - const: int1
     minItems: 1
 
+  poll-interval:
+    $ref: /schemas/types.yaml#/definitions/flag

This is a common property already defined as a uint32. You shouldn't define a new type.

A flag doesn't even make sense. If that's all you need, then just enable polling if no interrupt is present.

Documentation/devicetree/bindings/net/can/bosch,m_can.yaml properties: · line 55

Quoted 4 lines
Documentation/devicetree/bindings/net/can/bosch,m_can.yaml +4 -0
+    description: Enable hrtimer polling method for an M_CAN device.
+      If this property is defined in MCAN node, it tells the driver to
+      enable polling method for an MCAN device. If for an MCAN device,
+      hardware interrupt is found and hrtimer polling method is enabled,

What's hrtimer? (Don't put Linuxisms in bindings)

Quoted 48 lines
Documentation/devicetree/bindings/net/can/bosch,m_can.yaml +21 -2
+      the driver will use hardware interrupt method.
+
   clocks:
     items:
       - description: peripheral clock
@@ -122,8 +137,6 @@ required:
   - compatible
   - reg
   - reg-names
-  - interrupts
-  - interrupt-names
   - clocks
   - clock-names
   - bosch,mram-cfg
@@ -132,6 +145,7 @@ additionalProperties: false
 
 examples:
   - |
+    // Example with interrupts
     #include <dt-bindings/clock/imx6sx-clock.h>
     can@20e8000 {
       compatible = "bosch,m_can";
@@ -149,4 +163,22 @@ examples:
       };
     };
 
+  - |
+    // Example with timer polling
+    #include <dt-bindings/clock/imx6sx-clock.h>
+    can@20e8000 {
+      compatible = "bosch,m_can";
+      reg = <0x020e8000 0x4000>, <0x02298000 0x4000>;
+      reg-names = "m_can", "message_ram";
+      poll-interval;
+      clocks = <&clks IMX6SX_CLK_CANFD>,
+               <&clks IMX6SX_CLK_CANFD>;
+      clock-names = "hclk", "cclk";
+      bosch,mram-cfg = <0x0 0 0 32 0 0 0 1>;
+
+      can-transceiver {
+        max-bitrate = <5000000>;
+      };
+    };
+
...
Signature · 1 line
2.17.1
#3 lore raw

Re: [PATCH v4 1/4] dt-bindings: net: can: Add poll-interval for MCAN

Quoted 48 lines · Rob Herring
Quoted 27 lines · Judith Mendez

On AM62x SoC, MCANs on MCU domain do not have hardware interrupt routed to A53 Linux, instead they will use software interrupt by hrtimer. To enable timer method, interrupts should be optional so remove interrupts property from required section and introduce poll-interval property.

---
Changelog:
v3:
1. Move binding patch to first in series
2. Update description for poll-interval
3. Add oneOf to specify using interrupts/interrupt-names or poll-interval
4. Fix example property: add comment below 'example'

v2:
1. Add poll-interval property to enable timer polling method
2. Add example using poll-interval property

 .../bindings/net/can/bosch,m_can.yaml         | 36 +++++++++++++++++--
 1 file changed, 34 insertions(+), 2 deletions(-)
Documentation/devicetree/bindings/net/can/bosch,m_can.yaml +0 -0
index 67879aab623b..c024ee49962c 100644

Documentation/devicetree/bindings/net/can/bosch,m_can.yaml maintainers: · line 22

Quoted 9 lines
Documentation/devicetree/bindings/net/can/bosch,m_can.yaml +6 -0
@@ -14,6 +14,13 @@ maintainers:
 allOf:
   - $ref: can-controller.yaml#
 
+oneOf:
+  - required:
+      - interrupts
+      - interrupt-names
+  - required:
+      - poll-interval

Move this next to 'required'.

Quoted 8 lines
Documentation/devicetree/bindings/net/can/bosch,m_can.yaml +1 -0
+
 properties:
   compatible:
     const: bosch,m_can
@@ -40,6 +47,14 @@ properties:
       - const: int1
     minItems: 1
 
Quoted 8 lines

Documentation/devicetree/bindings/net/can/bosch,m_can.yaml properties: · line 51

Quoted 2 lines
Documentation/devicetree/bindings/net/can/bosch,m_can.yaml +2 -0
+  poll-interval:
+    $ref: /schemas/types.yaml#/definitions/flag

This is a common property already defined as a uint32. You shouldn't define a new type.

A flag doesn't even make sense. If that's all you need, then just enable polling if no interrupt is present.

Ok, then it's implicit. No IRQs -> polling.

Marc

Signature · 4 lines
Pengutronix e.K.                 | Marc Kleine-Budde          |
Embedded Linux                   | https://www.pengutronix.de |
Vertretung Nürnberg              | Phone: +49-5121-206917-129 |
Amtsgericht Hildesheim, HRA 2686 | Fax:   +49-5121-206917-9   |
#4 lore raw

Re: [PATCH v4 1/4] dt-bindings: net: can: Add poll-interval for MCAN

Hello all,

Quoted 55 lines · Marc Kleine-Budde
Quoted 54 lines · Rob Herring
Quoted 27 lines · Judith Mendez

On AM62x SoC, MCANs on MCU domain do not have hardware interrupt routed to A53 Linux, instead they will use software interrupt by hrtimer. To enable timer method, interrupts should be optional so remove interrupts property from required section and introduce poll-interval property.

---
Changelog:
v3:
1. Move binding patch to first in series
2. Update description for poll-interval
3. Add oneOf to specify using interrupts/interrupt-names or poll-interval
4. Fix example property: add comment below 'example'

v2:
1. Add poll-interval property to enable timer polling method
2. Add example using poll-interval property

 .../bindings/net/can/bosch,m_can.yaml         | 36 +++++++++++++++++--
 1 file changed, 34 insertions(+), 2 deletions(-)
Documentation/devicetree/bindings/net/can/bosch,m_can.yaml +0 -0
index 67879aab623b..c024ee49962c 100644

Documentation/devicetree/bindings/net/can/bosch,m_can.yaml maintainers: · line 22

Quoted 9 lines
Documentation/devicetree/bindings/net/can/bosch,m_can.yaml +6 -0
@@ -14,6 +14,13 @@ maintainers:
 allOf:
   - $ref: can-controller.yaml#
 
+oneOf:
+  - required:
+      - interrupts
+      - interrupt-names
+  - required:
+      - poll-interval

Move this next to 'required'.

Documentation/devicetree/bindings/net/can/bosch,m_can.yaml properties: · line 51

Quoted 10 lines
Documentation/devicetree/bindings/net/can/bosch,m_can.yaml +3 -0
+
 properties:
   compatible:
     const: bosch,m_can
@@ -40,6 +47,14 @@ properties:
       - const: int1
     minItems: 1
 
+  poll-interval:
+    $ref: /schemas/types.yaml#/definitions/flag

This is a common property already defined as a uint32. You shouldn't define a new type.

Quoted 4 lines
Quoted 2 lines

A flag doesn't even make sense. If that's all you need, then just enable polling if no interrupt is present.

Ok, then it's implicit. No IRQs -> polling.

Ok, will send out a v5 with these updates. Thanks.

regards,
Judith

#5 lore raw

PATCH v4 2/4 can: m_can: Add hrtimer to generate software interrupt

In reply to Judith Mendez

Add an hrtimer to MCAN class device. Each MCAN will have its own hrtimer instantiated if there is no hardware interrupt found and poll-interval property is defined in device tree M_CAN node.

The hrtimer will generate a software interrupt every 1 ms. In hrtimer callback, we check if there is a transaction pending by reading a register, then process by calling the isr if there is.

---
Changelog:
v1:
1. Sort list of includes
2. Create a define for HR_TIMER_POLL_INTERVAL
3. Fix indentations and style issues/warnings
4. Change polling variable to type bool
5. Change platform_get_irq to optional so not to print error msg
6. Move error check for addr directly after assignment
7. Print appropriate error msg with dev_err_probe insead of dev_dbg

v2:
1. Add poll-interval to MCAN class device to check if poll-interval propery is present in MCAN node, this enables timer polling method
2. Add 'polling' flag to MCAN class device to check if a device is using timer polling method
3. Check if both timer polling and hardware interrupt are enabled for a MCAN device, default to hardware interrupt mode if both are enabled
4. Change ms_to_ktime() to ns_to_ktime()
5. Remove newlines, tabs, and restructure if/else section

 drivers/net/can/m_can/m_can.c          | 29 +++++++++++++++++++++--
 drivers/net/can/m_can/m_can.h          |  4 ++++
 drivers/net/can/m_can/m_can_platform.c | 32 +++++++++++++++++++++++---
 3 files changed, 60 insertions(+), 5 deletions(-)
drivers/net/can/m_can/m_can.c +27 -2
index a5003435802b..e1ac0c1d85a3 100644
@@ -11,6 +11,7 @@
 #include <linux/bitfield.h>
 #include <linux/can/dev.h>
 #include <linux/ethtool.h>
+#include <linux/hrtimer.h>
 #include <linux/interrupt.h>
 #include <linux/io.h>
 #include <linux/iopoll.h>
@@ -308,6 +309,9 @@ enum m_can_reg {
 #define TX_EVENT_MM_MASK	GENMASK(31, 24)
 #define TX_EVENT_TXTS_MASK	GENMASK(15, 0)
 
+/* Hrtimer polling interval */
+#define HRTIMER_POLL_INTERVAL		1
+
 /* The ID and DLC registers are adjacent in M_CAN FIFO memory,
  * and we can save a (potentially slow) bus round trip by combining
  * reads and writes to them.
@@ -1587,6 +1591,11 @@ static int m_can_close(struct net_device *dev)
 	if (!cdev->is_peripheral)
 		napi_disable(&cdev->napi);
 
+	if (cdev->polling) {
+		dev_dbg(cdev->dev, "Disabling the hrtimer\n");
+		hrtimer_cancel(&cdev->hrtimer);
+	}
+
 	m_can_stop(dev);
 	m_can_clk_stop(cdev);
 	free_irq(dev->irq, dev);
@@ -1793,6 +1802,18 @@ static netdev_tx_t m_can_start_xmit(struct sk_buff *skb,
 	return NETDEV_TX_OK;
 }
 
+static enum hrtimer_restart hrtimer_callback(struct hrtimer *timer)
+{
+	struct m_can_classdev *cdev = container_of(timer, struct
+						   m_can_classdev, hrtimer);
+
+	m_can_isr(0, cdev->net);
+
+	hrtimer_forward_now(timer, ms_to_ktime(HRTIMER_POLL_INTERVAL));
+
+	return HRTIMER_RESTART;
+}
+
 static int m_can_open(struct net_device *dev)
 {
 	struct m_can_classdev *cdev = netdev_priv(dev);
@@ -1827,13 +1848,17 @@ static int m_can_open(struct net_device *dev)
 		}
 
 		INIT_WORK(&cdev->tx_work, m_can_tx_work_queue);
-
 		err = request_threaded_irq(dev->irq, NULL, m_can_isr,
 					   IRQF_ONESHOT,
 					   dev->name, dev);
-	} else {
+	} else if (!cdev->polling) {
 		err = request_irq(dev->irq, m_can_isr, IRQF_SHARED, dev->name,
 				  dev);
+	} else {
+		dev_dbg(cdev->dev, "Start hrtimer\n");
+		cdev->hrtimer.function = &hrtimer_callback;
+		hrtimer_start(&cdev->hrtimer, ms_to_ktime(HRTIMER_POLL_INTERVAL),
+			      HRTIMER_MODE_REL_PINNED);
 	}
 
 	if (err < 0) {
drivers/net/can/m_can/m_can.h +4 -0
index a839dc71dc9b..e9db5cce4e68 100644
@@ -15,6 +15,7 @@
 #include <linux/device.h>
 #include <linux/dma-mapping.h>
 #include <linux/freezer.h>
+#include <linux/hrtimer.h>
 #include <linux/interrupt.h>
 #include <linux/io.h>
 #include <linux/iopoll.h>
@@ -93,6 +94,9 @@ struct m_can_classdev {
 	int is_peripheral;
 
 	struct mram_cfg mcfg[MRAM_CFG_NUM];
+
+	struct hrtimer hrtimer;
+	bool polling;
 };
 
 struct m_can_classdev *m_can_class_allocate_dev(struct device *dev, int sizeof_priv);
drivers/net/can/m_can/m_can_platform.c +29 -3
index 9c1dcf838006..0fcb436298f8 100644
@@ -5,6 +5,7 @@
 //
 // Copyright (C) 2018-19 Texas Instruments Incorporated - http://www.ti.com/
 
+#include <linux/hrtimer.h>
 #include <linux/phy/phy.h>
 #include <linux/platform_device.h>
 
@@ -96,12 +97,37 @@ static int m_can_plat_probe(struct platform_device *pdev)
 		goto probe_fail;
 
 	addr = devm_platform_ioremap_resource_byname(pdev, "m_can");
-	irq = platform_get_irq_byname(pdev, "int0");
-	if (IS_ERR(addr) || irq < 0) {
-		ret = -EINVAL;
+	if (IS_ERR(addr)) {
+		ret = PTR_ERR(addr);
 		goto probe_fail;
 	}
 
+	irq = platform_get_irq_byname_optional(pdev, "int0");
+	if (irq == -EPROBE_DEFER) {
+		ret = -EPROBE_DEFER;
+		goto probe_fail;
+	}
+
+	if (device_property_present(mcan_class->dev, "poll-interval"))
+		mcan_class->polling = 1;
+
+	if (!mcan_class->polling && irq < 0) {
+		ret = -ENXIO;
+		dev_err_probe(mcan_class->dev, ret, "IRQ int0 not found and polling not activated\n");
+		goto probe_fail;
+	}
+
+	if (mcan_class->polling) {
+		if (irq > 0) {
+			mcan_class->polling = 0;
+			dev_dbg(mcan_class->dev, "Polling enabled and hardware IRQ found, use hardware IRQ\n");
+		} else {
+			dev_dbg(mcan_class->dev, "Polling enabled, initialize hrtimer");
+			hrtimer_init(&mcan_class->hrtimer, CLOCK_MONOTONIC,
+				     HRTIMER_MODE_REL_PINNED);
+		}
+	}
+
 	/* message ram could be shared */
 	res = platform_get_resource_byname(pdev, IORESOURCE_MEM, "message_ram");
 	if (!res) {
Signature · 1 line
2.17.1
#6 lore raw

Re: [PATCH v4 2/4] can: m_can: Add hrtimer to generate software interrupt

Quoted 9 lines · Judith Mendez

Add an hrtimer to MCAN class device. Each MCAN will have its own hrtimer instantiated if there is no hardware interrupt found and poll-interval property is defined in device tree M_CAN node.

The hrtimer will generate a software interrupt every 1 ms. In hrtimer callback, we check if there is a transaction pending by reading a register, then process by calling the isr if there is.

I think this patch is as good as it gets, given the HW and SW limitations of the coprocessor.

Some minor nitpicks inline. No need to resend from my point of view, I'll fixup while applying the patch.

Marc

Quoted 80 lines

---
Changelog:
v1:
1. Sort list of includes
2. Create a define for HR_TIMER_POLL_INTERVAL
3. Fix indentations and style issues/warnings
4. Change polling variable to type bool
5. Change platform_get_irq to optional so not to print error msg
6. Move error check for addr directly after assignment
7. Print appropriate error msg with dev_err_probe insead of dev_dbg

v2:
1. Add poll-interval to MCAN class device to check if poll-interval propery is present in MCAN node, this enables timer polling method
2. Add 'polling' flag to MCAN class device to check if a device is using timer polling method
3. Check if both timer polling and hardware interrupt are enabled for a MCAN device, default to hardware interrupt mode if both are enabled
4. Change ms_to_ktime() to ns_to_ktime()
5. Remove newlines, tabs, and restructure if/else section

 drivers/net/can/m_can/m_can.c          | 29 +++++++++++++++++++++--
 drivers/net/can/m_can/m_can.h          |  4 ++++
 drivers/net/can/m_can/m_can_platform.c | 32 +++++++++++++++++++++++---
 3 files changed, 60 insertions(+), 5 deletions(-)
drivers/net/can/m_can/m_can.c +21 -0
index a5003435802b..e1ac0c1d85a3 100644
@@ -11,6 +11,7 @@
 #include <linux/bitfield.h>
 #include <linux/can/dev.h>
 #include <linux/ethtool.h>
+#include <linux/hrtimer.h>
 #include <linux/interrupt.h>
 #include <linux/io.h>
 #include <linux/iopoll.h>
@@ -308,6 +309,9 @@ enum m_can_reg {
 #define TX_EVENT_MM_MASK	GENMASK(31, 24)
 #define TX_EVENT_TXTS_MASK	GENMASK(15, 0)
 
+/* Hrtimer polling interval */
+#define HRTIMER_POLL_INTERVAL		1
+
 /* The ID and DLC registers are adjacent in M_CAN FIFO memory,
  * and we can save a (potentially slow) bus round trip by combining
  * reads and writes to them.
@@ -1587,6 +1591,11 @@ static int m_can_close(struct net_device *dev)
 	if (!cdev->is_peripheral)
 		napi_disable(&cdev->napi);
 
+	if (cdev->polling) {
+		dev_dbg(cdev->dev, "Disabling the hrtimer\n");
+		hrtimer_cancel(&cdev->hrtimer);
+	}
+
 	m_can_stop(dev);
 	m_can_clk_stop(cdev);
 	free_irq(dev->irq, dev);
@@ -1793,6 +1802,18 @@ static netdev_tx_t m_can_start_xmit(struct sk_buff *skb,
 	return NETDEV_TX_OK;
 }
 
+static enum hrtimer_restart hrtimer_callback(struct hrtimer *timer)
+{
+	struct m_can_classdev *cdev = container_of(timer, struct
+						   m_can_classdev, hrtimer);
+
+	m_can_isr(0, cdev->net);
+
+	hrtimer_forward_now(timer, ms_to_ktime(HRTIMER_POLL_INTERVAL));
+
+	return HRTIMER_RESTART;
+}
+
 static int m_can_open(struct net_device *dev)
 {
 	struct m_can_classdev *cdev = netdev_priv(dev);

drivers/net/can/m_can/m_can.c static int m_can_open(struct net_device *dev) · line 1830

Quoted 4 lines
drivers/net/can/m_can/m_can.c +0 -1
@@ -1827,13 +1848,17 @@ static int m_can_open(struct net_device *dev)
 		}
 
 		INIT_WORK(&cdev->tx_work, m_can_tx_work_queue);
-

unrelated change

Quoted 59 lines
drivers/net/can/m_can/m_can.c +6 -1
 		err = request_threaded_irq(dev->irq, NULL, m_can_isr,
 					   IRQF_ONESHOT,
 					   dev->name, dev);
-	} else {
+	} else if (!cdev->polling) {
 		err = request_irq(dev->irq, m_can_isr, IRQF_SHARED, dev->name,
 				  dev);
+	} else {
+		dev_dbg(cdev->dev, "Start hrtimer\n");
+		cdev->hrtimer.function = &hrtimer_callback;
+		hrtimer_start(&cdev->hrtimer, ms_to_ktime(HRTIMER_POLL_INTERVAL),
+			      HRTIMER_MODE_REL_PINNED);
 	}
 
 	if (err < 0) {
drivers/net/can/m_can/m_can.h +4 -0
index a839dc71dc9b..e9db5cce4e68 100644
@@ -15,6 +15,7 @@
 #include <linux/device.h>
 #include <linux/dma-mapping.h>
 #include <linux/freezer.h>
+#include <linux/hrtimer.h>
 #include <linux/interrupt.h>
 #include <linux/io.h>
 #include <linux/iopoll.h>
@@ -93,6 +94,9 @@ struct m_can_classdev {
 	int is_peripheral;
 
 	struct mram_cfg mcfg[MRAM_CFG_NUM];
+
+	struct hrtimer hrtimer;
+	bool polling;
 };
 
 struct m_can_classdev *m_can_class_allocate_dev(struct device *dev, int sizeof_priv);
drivers/net/can/m_can/m_can_platform.c +3 -3
index 9c1dcf838006..0fcb436298f8 100644
@@ -5,6 +5,7 @@
 //
 // Copyright (C) 2018-19 Texas Instruments Incorporated - http://www.ti.com/
 
+#include <linux/hrtimer.h>
 #include <linux/phy/phy.h>
 #include <linux/platform_device.h>
 
@@ -96,12 +97,37 @@ static int m_can_plat_probe(struct platform_device *pdev)
 		goto probe_fail;
 
 	addr = devm_platform_ioremap_resource_byname(pdev, "m_can");
-	irq = platform_get_irq_byname(pdev, "int0");
-	if (IS_ERR(addr) || irq < 0) {
-		ret = -EINVAL;
+	if (IS_ERR(addr)) {
+		ret = PTR_ERR(addr);
 		goto probe_fail;

drivers/net/can/m_can/m_can_platform.c static int m_can_plat_probe(struct platform_device *pdev) · line 112

Quoted 10 lines
drivers/net/can/m_can/m_can_platform.c +8 -0
 	}
 
+	irq = platform_get_irq_byname_optional(pdev, "int0");
+	if (irq == -EPROBE_DEFER) {
+		ret = -EPROBE_DEFER;
+		goto probe_fail;
+	}
+
+	if (device_property_present(mcan_class->dev, "poll-interval"))
+		mcan_class->polling = 1;

true

drivers/net/can/m_can/m_can_platform.c static int m_can_plat_probe(struct platform_device *pdev) · line 122

Quoted 10 lines
drivers/net/can/m_can/m_can_platform.c +10 -0
+
+	if (!mcan_class->polling && irq < 0) {
+		ret = -ENXIO;
+		dev_err_probe(mcan_class->dev, ret, "IRQ int0 not found and polling not activated\n");
+		goto probe_fail;
+	}
+
+	if (mcan_class->polling) {
+		if (irq > 0) {
+			mcan_class->polling = 0;

false

drivers/net/can/m_can/m_can_platform.c static int m_can_plat_probe(struct platform_device *pdev) · line 123

Quoted 1 line
drivers/net/can/m_can/m_can_platform.c +1 -0
+			dev_dbg(mcan_class->dev, "Polling enabled and hardware IRQ found, use hardware IRQ\n");

"...using hardware IRQ"

Use dev_info(), as there is something not 100% correct with the DT.

Quoted 14 lines
drivers/net/can/m_can/m_can_platform.c +7 -0
+		} else {
+			dev_dbg(mcan_class->dev, "Polling enabled, initialize hrtimer");
+			hrtimer_init(&mcan_class->hrtimer, CLOCK_MONOTONIC,
+				     HRTIMER_MODE_REL_PINNED);
+		}
+	}
+
 	/* message ram could be shared */
 	res = platform_get_resource_byname(pdev, IORESOURCE_MEM, "message_ram");
 	if (!res) {
Signature · 1 line
2.17.1
Signature · 4 lines
Pengutronix e.K.                 | Marc Kleine-Budde          |
Embedded Linux                   | https://www.pengutronix.de |
Vertretung Nürnberg              | Phone: +49-5121-206917-129 |
Amtsgericht Hildesheim, HRA 2686 | Fax:   +49-5121-206917-9   |
#7 lore raw

Re: [PATCH v4 2/4] can: m_can: Add hrtimer to generate software interrupt

Hello Marc

Quoted 9 lines · Marc Kleine-Budde
Quoted 8 lines · Judith Mendez

Add an hrtimer to MCAN class device. Each MCAN will have its own hrtimer instantiated if there is no hardware interrupt found and poll-interval property is defined in device tree M_CAN node.

The hrtimer will generate a software interrupt every 1 ms. In hrtimer callback, we check if there is a transaction pending by reading a register, then process by calling the isr if there is.

Quoted 7 lines
Quoted 1 line

I think this patch is as good as it gets, given the HW and SW limitations of the coprocessor.

Some minor nitpicks inline. No need to resend from my point of view, I'll fixup while applying the patch.

Thanks Marc, really appreciate your feedback and attention.

Same to everyone who helped make these patches better. (:

regards,
Judith

#8 lore raw

Re: [PATCH v4 2/4] can: m_can: Add hrtimer to generate software interrupt

In reply to Marc Kleine-Budde

Hello Marc,

Quoted 191 lines · Marc Kleine-Budde
Quoted 9 lines · Judith Mendez

Add an hrtimer to MCAN class device. Each MCAN will have its own hrtimer instantiated if there is no hardware interrupt found and poll-interval property is defined in device tree M_CAN node.

The hrtimer will generate a software interrupt every 1 ms. In hrtimer callback, we check if there is a transaction pending by reading a register, then process by calling the isr if there is.

I think this patch is as good as it gets, given the HW and SW limitations of the coprocessor.

Some minor nitpicks inline. No need to resend from my point of view, I'll fixup while applying the patch.

Marc

Quoted 80 lines

---
Changelog:
v1:
1. Sort list of includes
2. Create a define for HR_TIMER_POLL_INTERVAL
3. Fix indentations and style issues/warnings
4. Change polling variable to type bool
5. Change platform_get_irq to optional so not to print error msg
6. Move error check for addr directly after assignment
7. Print appropriate error msg with dev_err_probe insead of dev_dbg

v2:
1. Add poll-interval to MCAN class device to check if poll-interval propery is present in MCAN node, this enables timer polling method
2. Add 'polling' flag to MCAN class device to check if a device is using timer polling method
3. Check if both timer polling and hardware interrupt are enabled for a MCAN device, default to hardware interrupt mode if both are enabled
4. Change ms_to_ktime() to ns_to_ktime()
5. Remove newlines, tabs, and restructure if/else section

 drivers/net/can/m_can/m_can.c          | 29 +++++++++++++++++++++--
 drivers/net/can/m_can/m_can.h          |  4 ++++
 drivers/net/can/m_can/m_can_platform.c | 32 +++++++++++++++++++++++---
 3 files changed, 60 insertions(+), 5 deletions(-)
drivers/net/can/m_can/m_can.c +21 -0
index a5003435802b..e1ac0c1d85a3 100644
@@ -11,6 +11,7 @@
 #include <linux/bitfield.h>
 #include <linux/can/dev.h>
 #include <linux/ethtool.h>
+#include <linux/hrtimer.h>
 #include <linux/interrupt.h>
 #include <linux/io.h>
 #include <linux/iopoll.h>
@@ -308,6 +309,9 @@ enum m_can_reg {
 #define TX_EVENT_MM_MASK	GENMASK(31, 24)
 #define TX_EVENT_TXTS_MASK	GENMASK(15, 0)
 
+/* Hrtimer polling interval */
+#define HRTIMER_POLL_INTERVAL		1
+
 /* The ID and DLC registers are adjacent in M_CAN FIFO memory,
  * and we can save a (potentially slow) bus round trip by combining
  * reads and writes to them.
@@ -1587,6 +1591,11 @@ static int m_can_close(struct net_device *dev)
 	if (!cdev->is_peripheral)
 		napi_disable(&cdev->napi);
 
+	if (cdev->polling) {
+		dev_dbg(cdev->dev, "Disabling the hrtimer\n");
+		hrtimer_cancel(&cdev->hrtimer);
+	}
+
 	m_can_stop(dev);
 	m_can_clk_stop(cdev);
 	free_irq(dev->irq, dev);
@@ -1793,6 +1802,18 @@ static netdev_tx_t m_can_start_xmit(struct sk_buff *skb,
 	return NETDEV_TX_OK;
 }
 
+static enum hrtimer_restart hrtimer_callback(struct hrtimer *timer)
+{
+	struct m_can_classdev *cdev = container_of(timer, struct
+						   m_can_classdev, hrtimer);
+
+	m_can_isr(0, cdev->net);
+
+	hrtimer_forward_now(timer, ms_to_ktime(HRTIMER_POLL_INTERVAL));
+
+	return HRTIMER_RESTART;
+}
+
 static int m_can_open(struct net_device *dev)
 {
 	struct m_can_classdev *cdev = netdev_priv(dev);

drivers/net/can/m_can/m_can.c static int m_can_open(struct net_device *dev) · line 1830

Quoted 4 lines
drivers/net/can/m_can/m_can.c +0 -1
@@ -1827,13 +1848,17 @@ static int m_can_open(struct net_device *dev)
 		}
 
 		INIT_WORK(&cdev->tx_work, m_can_tx_work_queue);
-

unrelated change

Quoted 59 lines
drivers/net/can/m_can/m_can.c +6 -1
 		err = request_threaded_irq(dev->irq, NULL, m_can_isr,
 					   IRQF_ONESHOT,
 					   dev->name, dev);
-	} else {
+	} else if (!cdev->polling) {
 		err = request_irq(dev->irq, m_can_isr, IRQF_SHARED, dev->name,
 				  dev);
+	} else {
+		dev_dbg(cdev->dev, "Start hrtimer\n");
+		cdev->hrtimer.function = &hrtimer_callback;
+		hrtimer_start(&cdev->hrtimer, ms_to_ktime(HRTIMER_POLL_INTERVAL),
+			      HRTIMER_MODE_REL_PINNED);
 	}
 
 	if (err < 0) {
drivers/net/can/m_can/m_can.h +4 -0
index a839dc71dc9b..e9db5cce4e68 100644
@@ -15,6 +15,7 @@
 #include <linux/device.h>
 #include <linux/dma-mapping.h>
 #include <linux/freezer.h>
+#include <linux/hrtimer.h>
 #include <linux/interrupt.h>
 #include <linux/io.h>
 #include <linux/iopoll.h>
@@ -93,6 +94,9 @@ struct m_can_classdev {
 	int is_peripheral;
 
 	struct mram_cfg mcfg[MRAM_CFG_NUM];
+
+	struct hrtimer hrtimer;
+	bool polling;
 };
 
 struct m_can_classdev *m_can_class_allocate_dev(struct device *dev, int sizeof_priv);
drivers/net/can/m_can/m_can_platform.c +3 -3
index 9c1dcf838006..0fcb436298f8 100644
@@ -5,6 +5,7 @@
 //
 // Copyright (C) 2018-19 Texas Instruments Incorporated - http://www.ti.com/
 
+#include <linux/hrtimer.h>
 #include <linux/phy/phy.h>
 #include <linux/platform_device.h>
 
@@ -96,12 +97,37 @@ static int m_can_plat_probe(struct platform_device *pdev)
 		goto probe_fail;
 
 	addr = devm_platform_ioremap_resource_byname(pdev, "m_can");
-	irq = platform_get_irq_byname(pdev, "int0");
-	if (IS_ERR(addr) || irq < 0) {
-		ret = -EINVAL;
+	if (IS_ERR(addr)) {
+		ret = PTR_ERR(addr);
 		goto probe_fail;

drivers/net/can/m_can/m_can_platform.c static int m_can_plat_probe(struct platform_device *pdev) · line 112

Quoted 10 lines
drivers/net/can/m_can/m_can_platform.c +8 -0
 	}
 
+	irq = platform_get_irq_byname_optional(pdev, "int0");
+	if (irq == -EPROBE_DEFER) {
+		ret = -EPROBE_DEFER;
+		goto probe_fail;
+	}
+
+	if (device_property_present(mcan_class->dev, "poll-interval"))
+		mcan_class->polling = 1;

true

drivers/net/can/m_can/m_can_platform.c static int m_can_plat_probe(struct platform_device *pdev) · line 122

Quoted 10 lines
drivers/net/can/m_can/m_can_platform.c +10 -0
+
+	if (!mcan_class->polling && irq < 0) {
+		ret = -ENXIO;
+		dev_err_probe(mcan_class->dev, ret, "IRQ int0 not found and polling not activated\n");
+		goto probe_fail;
+	}
+
+	if (mcan_class->polling) {
+		if (irq > 0) {
+			mcan_class->polling = 0;

false

Quoted 5 lines

drivers/net/can/m_can/m_can_platform.c static int m_can_plat_probe(struct platform_device *pdev) · line 123

Quoted 1 line
drivers/net/can/m_can/m_can_platform.c +1 -0
+			dev_dbg(mcan_class->dev, "Polling enabled and hardware IRQ found, use hardware IRQ\n");

"...using hardware IRQ"

Use dev_info(), as there is something not 100% correct with the DT.

Is it dev_info or dev_dbg? I used to have dev_info since it was nice to see when polling was enabled. Also, I had seen this print and the next as informative prints, hence the dev_info(). However, I was told in this review process to change to dev_dbg. Which is correct?

Quoted 6 lines
Quoted 6 lines
drivers/net/can/m_can/m_can_platform.c +6 -0
+		} else {
+			dev_dbg(mcan_class->dev, "Polling enabled, initialize hrtimer");
+			hrtimer_init(&mcan_class->hrtimer, CLOCK_MONOTONIC,
+				     HRTIMER_MODE_REL_PINNED);
+		}
+	}
Quoted 9 lines
Quoted 8 lines
drivers/net/can/m_can/m_can_platform.c +1 -0
+
 	/* message ram could be shared */
 	res = platform_get_resource_byname(pdev, IORESOURCE_MEM, "message_ram");
 	if (!res) {
Signature · 1 line
2.17.1

regards,
Judith

#9 lore raw

Re: [PATCH v4 2/4] can: m_can: Add hrtimer to generate software interrupt

On 09.05.2023 17:18:06, Judith Mendez wrote:
[...]

Quoted 12 lines
Quoted 12 lines

drivers/net/can/m_can/m_can_platform.c static int m_can_plat_probe(struct platform_device *pdev) · line 122

Quoted 9 lines
drivers/net/can/m_can/m_can_platform.c +9 -0
+	if (!mcan_class->polling && irq < 0) {
+		ret = -ENXIO;
+		dev_err_probe(mcan_class->dev, ret, "IRQ int0 not found and polling not activated\n");
+		goto probe_fail;
+	}
+
+	if (mcan_class->polling) {
+		if (irq > 0) {
+			mcan_class->polling = 0;

false

Quoted 7 lines
Quoted 5 lines

drivers/net/can/m_can/m_can_platform.c static int m_can_plat_probe(struct platform_device *pdev) · line 123

Quoted 1 line
drivers/net/can/m_can/m_can_platform.c +1 -0
+			dev_dbg(mcan_class->dev, "Polling enabled and hardware IRQ found, use hardware IRQ\n");

"...using hardware IRQ"

Use dev_info(), as there is something not 100% correct with the DT.

Is it dev_info or dev_dbg?

dev_info() - But without an explicit "poll-interval' in the DT, this code path doesn't exist anymore.

Quoted 2 lines

I used to have dev_info since it was nice to see when polling was enabled.

Re-read your code, this is not about enabling polling. This message handles the case where an IRQ was given _and_ "poll-interval" was specified. So there is something not 100% correct with the DT (IRQ _and_ polling), but this is obsolete now.

Quoted 1 line

Also, I had seen this print and the next as informative prints, hence the dev_info().

We don't print messages when IRQs are enabled, so enabling polling should be a dev_dbg(), too.

Quoted 1 line

However, I was told in this review process to change to dev_dbg. Which is correct?

Driver works correct -> dev_dbg()
Something is strange -> dev_info()

Hope that helps,
Marc

Signature · 4 lines
Pengutronix e.K.                 | Marc Kleine-Budde          |
Embedded Linux                   | https://www.pengutronix.de |
Vertretung Nürnberg              | Phone: +49-5121-206917-129 |
Amtsgericht Hildesheim, HRA 2686 | Fax:   +49-5121-206917-9   |
#10 lore raw

Re: [PATCH v4 2/4] can: m_can: Add hrtimer to generate software interrupt

Hello Marc,

Quoted 25 lines · Marc Kleine-Budde

On 09.05.2023 17:18:06, Judith Mendez wrote:
[...]

Quoted 12 lines
Quoted 12 lines

drivers/net/can/m_can/m_can_platform.c static int m_can_plat_probe(struct platform_device *pdev) · line 122

Quoted 9 lines
drivers/net/can/m_can/m_can_platform.c +9 -0
+	if (!mcan_class->polling && irq < 0) {
+		ret = -ENXIO;
+		dev_err_probe(mcan_class->dev, ret, "IRQ int0 not found and polling not activated\n");
+		goto probe_fail;
+	}
+
+	if (mcan_class->polling) {
+		if (irq > 0) {
+			mcan_class->polling = 0;

false

Quoted 7 lines
Quoted 5 lines

drivers/net/can/m_can/m_can_platform.c static int m_can_plat_probe(struct platform_device *pdev) · line 123

Quoted 1 line
drivers/net/can/m_can/m_can_platform.c +1 -0
+			dev_dbg(mcan_class->dev, "Polling enabled and hardware IRQ found, use hardware IRQ\n");

"...using hardware IRQ"

Use dev_info(), as there is something not 100% correct with the DT.

Is it dev_info or dev_dbg?

dev_info() - But without an explicit "poll-interval' in the DT, this code path doesn't exist anymore.

Quoted 7 lines
Quoted 2 lines

I used to have dev_info since it was nice to see when polling was enabled.

Re-read your code, this is not about enabling polling. This message handles the case where an IRQ was given _and_ "poll-interval" was specified. So there is something not 100% correct with the DT (IRQ _and_ polling), but this is obsolete now.

Sorry, I meant it was nice to see when polling was enabled or not. In

this case, if polling was enabled but hardware IRQ exists, it is good

information to see this print. But I understand now how this is a case

where 'something is strange'.

Quoted 9 lines
Quoted 1 line

Also, I had seen this print and the next as informative prints, hence the dev_info().

We don't print messages when IRQs are enabled, so enabling polling should be a dev_dbg(), too.

Quoted 1 line

However, I was told in this review process to change to dev_dbg. Which is correct?

Driver works correct -> dev_dbg()
Something is strange -> dev_info()

Got it, this is very helpful. Thanks.

regards,

Judith

#11 lore raw

PATCH v4 3/4 DO_NOT_MERGE arm64: dts: ti: Add AM62x MCAN MAIN domain transceiver overlay

In reply to Judith Mendez

Add an overlay for main domain MCAN on AM62x SK. The AM62x SK board does not have on-board CAN transceiver so instead of changing the DTB permanently, add an overlay to enable MAIN domain MCAN and support for 1 CAN transceiver.

This DT overlay can be used with the following EVM:

---
Changelog:
v3:
1. Add link for specific board

 arch/arm64/boot/dts/ti/Makefile               |  2 ++
 .../boot/dts/ti/k3-am625-sk-mcan-main.dtso    | 35 +++++++++++++++++++
 2 files changed, 37 insertions(+)
 create mode 100644 arch/arm64/boot/dts/ti/k3-am625-sk-mcan-main.dtso
arch/arm64/boot/dts/ti/Makefile +2 -0
index c83c9d772b81..abe15e76b614 100644
@@ -9,8 +9,10 @@
 # alphabetically.
 
 # Boards with AM62x SoC
+k3-am625-sk-mcan-dtbs := k3-am625-sk.dtb k3-am625-sk-mcan-main.dtbo
 dtb-$(CONFIG_ARCH_K3) += k3-am625-beagleplay.dtb
 dtb-$(CONFIG_ARCH_K3) += k3-am625-sk.dtb
+dtb-$(CONFIG_ARCH_K3) += k3-am625-sk-mcan.dtb
 dtb-$(CONFIG_ARCH_K3) += k3-am62-lp-sk.dtb
 
 # Boards with AM62Ax SoC
arch/arm64/boot/dts/ti/k3-am625-sk-mcan-main.dtso new file +35 -0
new file mode 100644
index 000000000000..0a7b2f394f87
@@ -0,0 +1,35 @@
+// SPDX-License-Identifier: GPL-2.0
+/**
+ * DT overlay for MCAN transceiver in main domain on AM625 SK
+ *
+ * Copyright (C) 2022 Texas Instruments Incorporated - https://www.ti.com/
+ */
+
+/dts-v1/;
+/plugin/;
+
+#include "k3-pinctrl.h"
+
+&{/} {
+	transceiver1: can-phy0 {
+		compatible = "ti,tcan1042";
+		#phy-cells = <0>;
+		max-bitrate = <5000000>;
+	};
+};
+
+&main_pmx0 {
+	main_mcan0_pins_default: main-mcan0-pins-default {
+		pinctrl-single,pins = <
+			AM62X_IOPAD(0x1dc, PIN_INPUT, 0) /* (E15) MCAN0_RX */
+			AM62X_IOPAD(0x1d8, PIN_OUTPUT, 0) /* (C15) MCAN0_TX */
+		>;
+	};
+};
+
+&main_mcan0 {
+	status = "okay";
+	pinctrl-names = "default";
+	pinctrl-0 = <&main_mcan0_pins_default>;
+	phys = <&transceiver1>;
+};
Signature · 1 line
2.17.1
maintainer NAK
#12 lore raw

Re: [PATCH v4 3/4] DO_NOT_MERGE arm64: dts: ti: Add AM62x MCAN MAIN domain transceiver overlay

Quoted 24 lines · Judith Mendez

Add an overlay for main domain MCAN on AM62x SK. The AM62x SK board does not have on-board CAN transceiver so instead of changing the DTB permanently, add an overlay to enable MAIN domain MCAN and support for 1 CAN transceiver.

This DT overlay can be used with the following EVM:

---
Changelog:
v3:
1. Add link for specific board

 arch/arm64/boot/dts/ti/Makefile               |  2 ++
 .../boot/dts/ti/k3-am625-sk-mcan-main.dtso    | 35 +++++++++++++++++++
 2 files changed, 37 insertions(+)
 create mode 100644 arch/arm64/boot/dts/ti/k3-am625-sk-mcan-main.dtso
arch/arm64/boot/dts/ti/Makefile +0 -0
index c83c9d772b81..abe15e76b614 100644

arch/arm64/boot/dts/ti/Makefile line 12

Quoted 4 lines
arch/arm64/boot/dts/ti/Makefile +1 -0
@@ -9,8 +9,10 @@
 # alphabetically.
 
 # Boards with AM62x SoC
+k3-am625-sk-mcan-dtbs := k3-am625-sk.dtb k3-am625-sk-mcan-main.dtbo

NAK.
https://lore.kernel.org/all/[email protected]/

Same reasons - we don't want specific instance based overlays please - that would'nt make sense - maxbitrate will depend on transceiver so it has nothing to do with mcan-main or mcan-mcu and has everything to do with the board that is plugged in.

arch/arm64/boot/dts/ti/k3-am625-sk-mcan-main.dtso

Quoted 11 lines
arch/arm64/boot/dts/ti/Makefile +1 -0
 dtb-$(CONFIG_ARCH_K3) += k3-am625-beagleplay.dtb
 dtb-$(CONFIG_ARCH_K3) += k3-am625-sk.dtb
+dtb-$(CONFIG_ARCH_K3) += k3-am625-sk-mcan.dtb
 dtb-$(CONFIG_ARCH_K3) += k3-am62-lp-sk.dtb
 
 # Boards with AM62Ax SoC
arch/arm64/boot/dts/ti/k3-am625-sk-mcan-main.dtso new file +0 -0
new file mode 100644
index 000000000000..0a7b2f394f87

If you are going down this road: am625-sk-tcan10242d.dsto (enable main and mcu?) or something reasonable. Though looking at the pins, I fail to see how this physically plugs into AM625-SK (I am hoping the answer isn't breadboard or jumper wires..).

Quoted 39 lines
arch/arm64/boot/dts/ti/k3-am625-sk-mcan-main.dtso new file +35 -0
@@ -0,0 +1,35 @@
+// SPDX-License-Identifier: GPL-2.0
+/**
+ * DT overlay for MCAN transceiver in main domain on AM625 SK
+ *
+ * Copyright (C) 2022 Texas Instruments Incorporated - https://www.ti.com/
+ */
+
+/dts-v1/;
+/plugin/;
+
+#include "k3-pinctrl.h"
+
+&{/} {
+	transceiver1: can-phy0 {
+		compatible = "ti,tcan1042";
+		#phy-cells = <0>;
+		max-bitrate = <5000000>;
+	};
+};
+
+&main_pmx0 {
+	main_mcan0_pins_default: main-mcan0-pins-default {
+		pinctrl-single,pins = <
+			AM62X_IOPAD(0x1dc, PIN_INPUT, 0) /* (E15) MCAN0_RX */
+			AM62X_IOPAD(0x1d8, PIN_OUTPUT, 0) /* (C15) MCAN0_TX */
+		>;
+	};
+};
+
+&main_mcan0 {
+	status = "okay";
+	pinctrl-names = "default";
+	pinctrl-0 = <&main_mcan0_pins_default>;
+	phys = <&transceiver1>;
+};
Signature · 1 line
2.17.1
Signature · 3 lines
Regards,
Nishanth Menon
Key (0xDDB5849D1736249D) / Fingerprint: F8A2 8693 54EB 8232 17A3  1A34 DDB5 849D 1736 249D
#13 lore raw

PATCH v4 4/4 DO_NOT_MERGE arm64: dts: ti: Enable MCU MCANs for AM62x

In reply to Judith Mendez

On AM62x there are no hardware interrupts routed to A53 GIC interrupt controller for MCU MCAN IPs, so MCU MCANs were not added to the MCU dtsi. In this patch series an hrtimer is introduced to MCAN driver to generate software interrupts. Now add MCU MCAN nodes to the MCU dtsi but disable the MCAN devices by default.

AM62x does not carry on-board CAN transceivers, so instead of changing DTB permanently use an overlay to enable MCU MCANs and to add CAN transceiver nodes.

If there is no hardware interrupt and timer method is used, remove interrupt properties and add poll-interval to enable the hrtimer per MCAN node.

This DT overlay can be used with the following EVM:

---
Changelog:
v3:
1. Add link for specific board

 arch/arm64/boot/dts/ti/Makefile               |  2 +-
 arch/arm64/boot/dts/ti/k3-am62-mcu.dtsi       | 24 ++++++++
 .../boot/dts/ti/k3-am625-sk-mcan-mcu.dtso     | 57 +++++++++++++++++++
 3 files changed, 82 insertions(+), 1 deletion(-)
 create mode 100644 arch/arm64/boot/dts/ti/k3-am625-sk-mcan-mcu.dtso
arch/arm64/boot/dts/ti/Makefile +1 -1
index abe15e76b614..c76be3888e4d 100644
@@ -9,7 +9,7 @@
 # alphabetically.
 
 # Boards with AM62x SoC
-k3-am625-sk-mcan-dtbs := k3-am625-sk.dtb k3-am625-sk-mcan-main.dtbo
+k3-am625-sk-mcan-dtbs := k3-am625-sk.dtb k3-am625-sk-mcan-main.dtbo k3-am625-sk-mcan-mcu.dtbo
 dtb-$(CONFIG_ARCH_K3) += k3-am625-beagleplay.dtb
 dtb-$(CONFIG_ARCH_K3) += k3-am625-sk.dtb
 dtb-$(CONFIG_ARCH_K3) += k3-am625-sk-mcan.dtb
arch/arm64/boot/dts/ti/k3-am62-mcu.dtsi +24 -0
index 076601a41e84..20462f457643 100644
@@ -141,4 +141,28 @@
 		/* Tightly coupled to M4F */
 		status = "reserved";
 	};
+
+	mcu_mcan1: can@4e00000 {
+		compatible = "bosch,m_can";
+		reg = <0x00 0x4e00000 0x00 0x8000>,
+			  <0x00 0x4e08000 0x00 0x200>;
+		reg-names = "message_ram", "m_can";
+		power-domains = <&k3_pds 188 TI_SCI_PD_EXCLUSIVE>;
+		clocks = <&k3_clks 188 6>, <&k3_clks 188 1>;
+		clock-names = "hclk", "cclk";
+		bosch,mram-cfg = <0x0 128 64 64 64 64 32 32>;
+		status = "disabled";
+	};
+
+	mcu_mcan2: can@4e10000 {
+		compatible = "bosch,m_can";
+		reg = <0x00 0x4e10000 0x00 0x8000>,
+			  <0x00 0x4e18000 0x00 0x200>;
+		reg-names = "message_ram", "m_can";
+		power-domains = <&k3_pds 189 TI_SCI_PD_EXCLUSIVE>;
+		clocks = <&k3_clks 189 6>, <&k3_clks 189 1>;
+		clock-names = "hclk", "cclk";
+		bosch,mram-cfg = <0x0 128 64 64 64 64 32 32>;
+		status = "disabled";
+	};
 };
arch/arm64/boot/dts/ti/k3-am625-sk-mcan-mcu.dtso new file +57 -0

57 lines not on this page · raw message

Signature · 1 line
2.17.1
#14 lore raw
In reply to Judith Mendez
Quoted 40 lines · Judith Mendez

On AM62x there is one MCAN in MAIN domain and two in MCU domain. The MCANs in MCU domain were not enabled since there is no hardware interrupt routed to A53 GIC interrupt controller. Therefore A53 Linux cannot be interrupted by MCU MCANs.

This solution instantiates a hrtimer with 1 ms polling interval for MCAN device when there is no hardware interrupt and there is poll-interval property in DTB MCAN node. The hrtimer generates a recurring software interrupt which allows to call the isr. The isr will check if there is pending transaction by reading a register and proceed normally if there is.

On AM62x, this series enables two MCU MCAN which will use the hrtimer implementation. MCANs with hardware interrupt routed to A53 Linux will continue to use the hardware interrupt as expected.

Timer polling method was tested on both classic CAN and CAN-FD at 125 KBPS, 250 KBPS, 1 MBPS and 2.5 MBPS with 4 MBPS bitrate switching.

Letency and CPU load benchmarks were tested on 3x MCAN on AM62x. 1 MBPS timer polling interval is the better timer polling interval since it has comparable latency to hardware interrupt with the worse case being 1ms + CAN frame propagation time and CPU load is not substantial. Latency can be improved further with less than 1 ms polling intervals, howerver it is at the cost of CPU usage since CPU load increases at 0.5 ms.

Note that in terms of power, enabling MCU MCANs with timer-polling implementation might have negative impact since we will have to wake up every 1 ms whether there are CAN packets pending in the RX FIFO or not. This might prevent the CPU from entering into deeper idle states for extended periods of time.

This patch series depends on 'Enable CAN PHY transceiver driver':

v2:

Quoted 8 lines

V1:

RFC:

Changes since v3:
- Wrong patch sent, resend correct patch series

Sending patchsets every 10 minutes does not help us...

Best regards,
Krzysztof

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