[PATCH 1/4] can: sja1000: Add support for CAN_CTRLMODE_LOOPBACK

[PATCH 1/4] can: sja1000: Add support for CAN_CTRLMODE_LOOPBACK

[Nikita Edward Baruzdin]

This adds support for hardware loopback in SJA1000 by utilising its self
reception request (SRR) feature. Upon SRR the message is transmitted and
received simultaneously, meaning you can't have hardware loopback
without actually sending a message to the CAN bus in case of SJA1000.

Signed-off-by: Nikita Edward Baruzdin <nebaruzdin@gmail.com>
---
 drivers/net/can/sja1000/sja1000.c | 18 +++++++++++++-----
 1 file changed, 13 insertions(+), 5 deletions(-)

diff --git a/drivers/net/can/sja1000/sja1000.c b/drivers/net/can/sja1000/sja1000.c
index f31499a..c480bce 100644
--- a/drivers/net/can/sja1000/sja1000.c
+++ b/drivers/net/can/sja1000/sja1000.c
@@ -278,6 +278,7 @@ static netdev_tx_t sja1000_start_xmit(struct sk_buff *skb,
 	uint8_t dlc;
 	canid_t id;
 	uint8_t dreg;
+	u8 cmd_reg_val = 0x00;
 	int i;
 
 	if (can_dropped_invalid_skb(dev, skb))
@@ -312,9 +313,14 @@ static netdev_tx_t sja1000_start_xmit(struct sk_buff *skb,
 	can_put_echo_skb(skb, dev, 0);
 
 	if (priv->can.ctrlmode & CAN_CTRLMODE_ONE_SHOT)
-		sja1000_write_cmdreg(priv, CMD_TR | CMD_AT);
+		cmd_reg_val |= CMD_AT;
+
+	if (priv->can.ctrlmode & CAN_CTRLMODE_LOOPBACK)
+		cmd_reg_val |= CMD_SRR;
 	else
-		sja1000_write_cmdreg(priv, CMD_TR);
+		cmd_reg_val |= CMD_TR;
+
+	sja1000_write_cmdreg(priv, cmd_reg_val);
 
 	return NETDEV_TX_OK;
 }
@@ -622,9 +628,11 @@ struct net_device *alloc_sja1000dev(int sizeof_priv)
 	priv->can.do_set_bittiming = sja1000_set_bittiming;
 	priv->can.do_set_mode = sja1000_set_mode;
 	priv->can.do_get_berr_counter = sja1000_get_berr_counter;
-	priv->can.ctrlmode_supported = CAN_CTRLMODE_3_SAMPLES |
-		CAN_CTRLMODE_BERR_REPORTING | CAN_CTRLMODE_LISTENONLY |
-		CAN_CTRLMODE_ONE_SHOT;
+	priv->can.ctrlmode_supported = CAN_CTRLMODE_LOOPBACK
+				     | CAN_CTRLMODE_LISTENONLY
+				     | CAN_CTRLMODE_3_SAMPLES
+				     | CAN_CTRLMODE_ONE_SHOT
+				     | CAN_CTRLMODE_BERR_REPORTING;
 
 	spin_lock_init(&priv->cmdreg_lock);
 
-- 
2.0.1

[PATCH 2/4] can: netlink: Remove space before tab

[Nikita Edward Baruzdin]

Fixes the corresponing checkpatch.pl warning.

Signed-off-by: Nikita Edward Baruzdin <nebaruzdin@gmail.com>
---
 include/uapi/linux/can/netlink.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/include/uapi/linux/can/netlink.h b/include/uapi/linux/can/netlink.h
index 813d11f..3bbf5c7 100644
--- a/include/uapi/linux/can/netlink.h
+++ b/include/uapi/linux/can/netlink.h
@@ -92,7 +92,7 @@ struct can_ctrlmode {
 };
 
 #define CAN_CTRLMODE_LOOPBACK		0x01	/* Loopback mode */
-#define CAN_CTRLMODE_LISTENONLY		0x02 	/* Listen-only mode */
+#define CAN_CTRLMODE_LISTENONLY		0x02	/* Listen-only mode */
 #define CAN_CTRLMODE_3_SAMPLES		0x04	/* Triple sampling mode */
 #define CAN_CTRLMODE_ONE_SHOT		0x08	/* One-Shot mode */
 #define CAN_CTRLMODE_BERR_REPORTING	0x10	/* Bus-error reporting */
-- 
2.0.1

[PATCH 3/4] can: netlink: Add CAN_CTRLMODE_PRESUME_ACK flag

[Nikita Edward Baruzdin]

Most CAN controllers have a support for ignoring ACK absence. Some of
them refer to this feature as a self test mode (e. g. SJA1000) and some
include it as a part of a loopback mode (e. g. MCP2510).

Setting the introduced flag via netlink should make CAN controller
perform a successful transmission, even if there is no acknowledgement
(dominant ACK bit) received.

Signed-off-by: Nikita Edward Baruzdin <nebaruzdin@gmail.com>
---
 include/uapi/linux/can/netlink.h | 1 +
 1 file changed, 1 insertion(+)

diff --git a/include/uapi/linux/can/netlink.h b/include/uapi/linux/can/netlink.h
index 3bbf5c7..3e4323a 100644
--- a/include/uapi/linux/can/netlink.h
+++ b/include/uapi/linux/can/netlink.h
@@ -97,6 +97,7 @@ struct can_ctrlmode {
 #define CAN_CTRLMODE_ONE_SHOT		0x08	/* One-Shot mode */
 #define CAN_CTRLMODE_BERR_REPORTING	0x10	/* Bus-error reporting */
 #define CAN_CTRLMODE_FD			0x20	/* CAN FD mode */
+#define CAN_CTRLMODE_PRESUME_ACK	0x40	/* Ignore missing CAN ACKs */
 
 /*
  * CAN device statistics
-- 
2.0.1

Re: [PATCH 3/4] can: netlink: Add CAN_CTRLMODE_PRESUME_ACK flag

[Oliver Hartkopp]

Hello Marc and Wolfgang,

what do you think about this extension.

 From my side the patchset makes sense.

Regards,
Oliver

On 09.07.2014 15:31, Nikita Edward Baruzdin wrote:
> Most CAN controllers have a support for ignoring ACK absence. Some of
> them refer to this feature as a self test mode (e. g. SJA1000) and some
> include it as a part of a loopback mode (e. g. MCP2510).
>
> Setting the introduced flag via netlink should make CAN controller
> perform a successful transmission, even if there is no acknowledgement
> (dominant ACK bit) received.
>
> Signed-off-by: Nikita Edward Baruzdin <nebaruzdin@gmail.com>
> ---
>   include/uapi/linux/can/netlink.h | 1 +
>   1 file changed, 1 insertion(+)
>
> diff --git a/include/uapi/linux/can/netlink.h b/include/uapi/linux/can/netlink.h
> index 3bbf5c7..3e4323a 100644
> --- a/include/uapi/linux/can/netlink.h
> +++ b/include/uapi/linux/can/netlink.h
> @@ -97,6 +97,7 @@ struct can_ctrlmode {
>   #define CAN_CTRLMODE_ONE_SHOT		0x08	/* One-Shot mode */
>   #define CAN_CTRLMODE_BERR_REPORTING	0x10	/* Bus-error reporting */
>   #define CAN_CTRLMODE_FD			0x20	/* CAN FD mode */
> +#define CAN_CTRLMODE_PRESUME_ACK	0x40	/* Ignore missing CAN ACKs */
>
>   /*
>    * CAN device statistics
>

Re: [PATCH 3/4] can: netlink: Add CAN_CTRLMODE_PRESUME_ACK flag

[Marc Kleine-Budde]

[-- Attachment #1: Type: text/plain, Size: 581 bytes --]

On 07/10/2014 04:06 PM, Oliver Hartkopp wrote:
> what do you think about this extension.
> 
> From my side the patchset makes sense.
>> +#define CAN_CTRLMODE_PRESUME_ACK    0x40    /* Ignore missing CAN
>> ACKs */

Is this a good name? Does anyone come up with an alternative/better one?

Marc

-- 
Pengutronix e.K.                  | Marc Kleine-Budde           |
Industrial Linux Solutions        | Phone: +49-231-2826-924     |
Vertretung West/Dortmund          | Fax:   +49-5121-206917-5555 |
Amtsgericht Hildesheim, HRA 2686  | http://www.pengutronix.de   |


[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 242 bytes --]

Re: [PATCH 3/4] can: netlink: Add CAN_CTRLMODE_PRESUME_ACK flag

[Oliver Hartkopp]

I tried to find a short name for:

- do not wait for ack
- ignore missing ack
- broadcast that doesn't require an ack
- ...

I'm really open for an alternative ;-)

On 10.07.2014 10:12, Marc Kleine-Budde wrote:
> On 07/10/2014 04:06 PM, Oliver Hartkopp wrote:
>> what do you think about this extension.
>>
>>  From my side the patchset makes sense.
>>> +#define CAN_CTRLMODE_PRESUME_ACK    0x40    /* Ignore missing CAN
>>> ACKs */
>
> Is this a good name? Does anyone come up with an alternative/better one?
>
> Marc
>

Re: [PATCH 3/4] can: netlink: Add CAN_CTRLMODE_PRESUME_ACK flag

[Nikita Edward Baruzdin]

I actually think the word "PRESUME" is good enough. We could also use
"NOT_REQUIRE" instead: it's a bit more explicit, but a bit longer.
At least I couldn't think of anything better.

"NOT_WAIT" doesn't reflect the behaviour precisely in my view.
I also believe that the word "IGNORE" is better when it comes to some
"active" message (that supposes some reaction), like "IGNORE_ERRORS" or
"IGNORE_SERVICE_REQUEST". And ACK is more of a reply than a request for
action.

On Fri, 2014-07-11 at 09:18 -0400, Oliver Hartkopp wrote:
> I tried to find a short name for:
> 
> - do not wait for ack
> - ignore missing ack
> - broadcast that doesn't require an ack
> - ...
> 
> I'm really open for an alternative ;-)
> 
> On 10.07.2014 10:12, Marc Kleine-Budde wrote:
> > On 07/10/2014 04:06 PM, Oliver Hartkopp wrote:
> >> what do you think about this extension.
> >>
> >>  From my side the patchset makes sense.
> >>> +#define CAN_CTRLMODE_PRESUME_ACK    0x40    /* Ignore missing CAN
> >>> ACKs */
> >
> > Is this a good name? Does anyone come up with an alternative/better one?
> >
> > Marc
> >

Re: [PATCH 3/4] can: netlink: Add CAN_CTRLMODE_PRESUME_ACK flag

[Marc Kleine-Budde]

[-- Attachment #1: Type: text/plain, Size: 478 bytes --]

On 07/10/2014 04:06 PM, Oliver Hartkopp wrote:
> Hello Marc and Wolfgang,
> 
> what do you think about this extension.
> 
> From my side the patchset makes sense.

Is this an Acked-by?

Marc

-- 
Pengutronix e.K.                  | Marc Kleine-Budde           |
Industrial Linux Solutions        | Phone: +49-231-2826-924     |
Vertretung West/Dortmund          | Fax:   +49-5121-206917-5555 |
Amtsgericht Hildesheim, HRA 2686  | http://www.pengutronix.de   |


[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 242 bytes --]

Re: [PATCH 3/4] can: netlink: Add CAN_CTRLMODE_PRESUME_ACK flag

[Oliver Hartkopp]

Yes.

Acked-by: Oliver Hartkopp <socketcan@hartkopp.net>

Just wanted to know if it's ok for you too.

Tnx,
Oliver

On 14.07.2014 08:21, Marc Kleine-Budde wrote:
> On 07/10/2014 04:06 PM, Oliver Hartkopp wrote:
>> Hello Marc and Wolfgang,
>>
>> what do you think about this extension.
>>
>>  From my side the patchset makes sense.
>
> Is this an Acked-by?
>
> Marc
>

[PATCH 4/4] can: sja1000: Add support for CAN_CTRLMODE_PRESUME_ACK

[Nikita Edward Baruzdin]

SJA1000 has a self test mode (STM) which does not require
acknowledgement for the successful message transmission. In this mode a
node test is possible without any other active node on the bus.

This patch adds a possibility to set STM for SJA1000 controller through
specifying the corresponding CAN_CTRLMODE_PRESUME_ACK netlink flag.

Signed-off-by: Nikita Edward Baruzdin <nebaruzdin@gmail.com>
---
 drivers/net/can/sja1000/sja1000.c | 11 +++++++----
 1 file changed, 7 insertions(+), 4 deletions(-)

diff --git a/drivers/net/can/sja1000/sja1000.c b/drivers/net/can/sja1000/sja1000.c
index c480bce..fd6ee05 100644
--- a/drivers/net/can/sja1000/sja1000.c
+++ b/drivers/net/can/sja1000/sja1000.c
@@ -141,6 +141,7 @@ static void set_normal_mode(struct net_device *dev)
 {
 	struct sja1000_priv *priv = netdev_priv(dev);
 	unsigned char status = priv->read_reg(priv, SJA1000_MOD);
+	u8 mod_reg_val = 0x00;
 	int i;
 
 	for (i = 0; i < 100; i++) {
@@ -158,9 +159,10 @@ static void set_normal_mode(struct net_device *dev)
 
 		/* set chip to normal mode */
 		if (priv->can.ctrlmode & CAN_CTRLMODE_LISTENONLY)
-			priv->write_reg(priv, SJA1000_MOD, MOD_LOM);
-		else
-			priv->write_reg(priv, SJA1000_MOD, 0x00);
+			mod_reg_val |= MOD_LOM;
+		if (priv->can.ctrlmode & CAN_CTRLMODE_PRESUME_ACK)
+			mod_reg_val |= MOD_STM;
+		priv->write_reg(priv, SJA1000_MOD, mod_reg_val);
 
 		udelay(10);
 
@@ -632,7 +634,8 @@ struct net_device *alloc_sja1000dev(int sizeof_priv)
 				     | CAN_CTRLMODE_LISTENONLY
 				     | CAN_CTRLMODE_3_SAMPLES
 				     | CAN_CTRLMODE_ONE_SHOT
-				     | CAN_CTRLMODE_BERR_REPORTING;
+				     | CAN_CTRLMODE_BERR_REPORTING
+				     | CAN_CTRLMODE_PRESUME_ACK;
 
 	spin_lock_init(&priv->cmdreg_lock);
 
-- 
2.0.1

Re: [PATCH 1/4] can: sja1000: Add support for CAN_CTRLMODE_LOOPBACK

[Marc Kleine-Budde]

[-- Attachment #1: Type: text/plain, Size: 2537 bytes --]

On 07/09/2014 09:31 PM, Nikita Edward Baruzdin wrote:
> This adds support for hardware loopback in SJA1000 by utilising its self
> reception request (SRR) feature. Upon SRR the message is transmitted and
> received simultaneously, meaning you can't have hardware loopback
> without actually sending a message to the CAN bus in case of SJA1000.
> 
> Signed-off-by: Nikita Edward Baruzdin <nebaruzdin@gmail.com>

Looks good, coding style nitpick inline...

> ---
>  drivers/net/can/sja1000/sja1000.c | 18 +++++++++++++-----
>  1 file changed, 13 insertions(+), 5 deletions(-)
> 
> diff --git a/drivers/net/can/sja1000/sja1000.c b/drivers/net/can/sja1000/sja1000.c
> index f31499a..c480bce 100644
> --- a/drivers/net/can/sja1000/sja1000.c
> +++ b/drivers/net/can/sja1000/sja1000.c
> @@ -278,6 +278,7 @@ static netdev_tx_t sja1000_start_xmit(struct sk_buff *skb,
>  	uint8_t dlc;
>  	canid_t id;
>  	uint8_t dreg;
> +	u8 cmd_reg_val = 0x00;
>  	int i;
>  
>  	if (can_dropped_invalid_skb(dev, skb))
> @@ -312,9 +313,14 @@ static netdev_tx_t sja1000_start_xmit(struct sk_buff *skb,
>  	can_put_echo_skb(skb, dev, 0);
>  
>  	if (priv->can.ctrlmode & CAN_CTRLMODE_ONE_SHOT)
> -		sja1000_write_cmdreg(priv, CMD_TR | CMD_AT);
> +		cmd_reg_val |= CMD_AT;
> +
> +	if (priv->can.ctrlmode & CAN_CTRLMODE_LOOPBACK)
> +		cmd_reg_val |= CMD_SRR;
>  	else
> -		sja1000_write_cmdreg(priv, CMD_TR);
> +		cmd_reg_val |= CMD_TR;
> +
> +	sja1000_write_cmdreg(priv, cmd_reg_val);
>  
>  	return NETDEV_TX_OK;
>  }
> @@ -622,9 +628,11 @@ struct net_device *alloc_sja1000dev(int sizeof_priv)
>  	priv->can.do_set_bittiming = sja1000_set_bittiming;
>  	priv->can.do_set_mode = sja1000_set_mode;
>  	priv->can.do_get_berr_counter = sja1000_get_berr_counter;
> -	priv->can.ctrlmode_supported = CAN_CTRLMODE_3_SAMPLES |
> -		CAN_CTRLMODE_BERR_REPORTING | CAN_CTRLMODE_LISTENONLY |
> -		CAN_CTRLMODE_ONE_SHOT;
> +	priv->can.ctrlmode_supported = CAN_CTRLMODE_LOOPBACK
> +				     | CAN_CTRLMODE_LISTENONLY
> +				     | CAN_CTRLMODE_3_SAMPLES
> +				     | CAN_CTRLMODE_ONE_SHOT
> +				     | CAN_CTRLMODE_BERR_REPORTING;

Please keep the '|' at the end of the lines.

>  
>  	spin_lock_init(&priv->cmdreg_lock);
>  
> 

Marc

-- 
Pengutronix e.K.                  | Marc Kleine-Budde           |
Industrial Linux Solutions        | Phone: +49-231-2826-924     |
Vertretung West/Dortmund          | Fax:   +49-5121-206917-5555 |
Amtsgericht Hildesheim, HRA 2686  | http://www.pengutronix.de   |


[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 242 bytes --]

[PATCH v2 1/4] can: sja1000: Add support for CAN_CTRLMODE_LOOPBACK

[Nikita Edward Baruzdin]

This adds support for hardware loopback in SJA1000 by utilising its self
reception request (SRR) feature. Upon SRR the message is transmitted and
received simultaneously, meaning you can't have hardware loopback
without actually sending a message to the CAN bus in case of SJA1000.

Signed-off-by: Nikita Edward Baruzdin <nebaruzdin@gmail.com>
---
 drivers/net/can/sja1000/sja1000.c | 18 +++++++++++++-----
 1 file changed, 13 insertions(+), 5 deletions(-)

diff --git a/drivers/net/can/sja1000/sja1000.c b/drivers/net/can/sja1000/sja1000.c
index f31499a..45400d9 100644
--- a/drivers/net/can/sja1000/sja1000.c
+++ b/drivers/net/can/sja1000/sja1000.c
@@ -278,6 +278,7 @@ static netdev_tx_t sja1000_start_xmit(struct sk_buff *skb,
 	uint8_t dlc;
 	canid_t id;
 	uint8_t dreg;
+	u8 cmd_reg_val = 0x00;
 	int i;
 
 	if (can_dropped_invalid_skb(dev, skb))
@@ -312,9 +313,14 @@ static netdev_tx_t sja1000_start_xmit(struct sk_buff *skb,
 	can_put_echo_skb(skb, dev, 0);
 
 	if (priv->can.ctrlmode & CAN_CTRLMODE_ONE_SHOT)
-		sja1000_write_cmdreg(priv, CMD_TR | CMD_AT);
+		cmd_reg_val |= CMD_AT;
+
+	if (priv->can.ctrlmode & CAN_CTRLMODE_LOOPBACK)
+		cmd_reg_val |= CMD_SRR;
 	else
-		sja1000_write_cmdreg(priv, CMD_TR);
+		cmd_reg_val |= CMD_TR;
+
+	sja1000_write_cmdreg(priv, cmd_reg_val);
 
 	return NETDEV_TX_OK;
 }
@@ -622,9 +628,11 @@ struct net_device *alloc_sja1000dev(int sizeof_priv)
 	priv->can.do_set_bittiming = sja1000_set_bittiming;
 	priv->can.do_set_mode = sja1000_set_mode;
 	priv->can.do_get_berr_counter = sja1000_get_berr_counter;
-	priv->can.ctrlmode_supported = CAN_CTRLMODE_3_SAMPLES |
-		CAN_CTRLMODE_BERR_REPORTING | CAN_CTRLMODE_LISTENONLY |
-		CAN_CTRLMODE_ONE_SHOT;
+	priv->can.ctrlmode_supported = CAN_CTRLMODE_LOOPBACK |
+				       CAN_CTRLMODE_LISTENONLY |
+				       CAN_CTRLMODE_3_SAMPLES |
+				       CAN_CTRLMODE_ONE_SHOT |
+				       CAN_CTRLMODE_BERR_REPORTING;
 
 	spin_lock_init(&priv->cmdreg_lock);
 
-- 
2.0.1

[PATCH v2 2/4] can: netlink: Remove space before tab

[Nikita Edward Baruzdin]

Fixes the corresponing checkpatch.pl warning.

Signed-off-by: Nikita Edward Baruzdin <nebaruzdin@gmail.com>
---
 include/uapi/linux/can/netlink.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/include/uapi/linux/can/netlink.h b/include/uapi/linux/can/netlink.h
index 813d11f..3bbf5c7 100644
--- a/include/uapi/linux/can/netlink.h
+++ b/include/uapi/linux/can/netlink.h
@@ -92,7 +92,7 @@ struct can_ctrlmode {
 };
 
 #define CAN_CTRLMODE_LOOPBACK		0x01	/* Loopback mode */
-#define CAN_CTRLMODE_LISTENONLY		0x02 	/* Listen-only mode */
+#define CAN_CTRLMODE_LISTENONLY		0x02	/* Listen-only mode */
 #define CAN_CTRLMODE_3_SAMPLES		0x04	/* Triple sampling mode */
 #define CAN_CTRLMODE_ONE_SHOT		0x08	/* One-Shot mode */
 #define CAN_CTRLMODE_BERR_REPORTING	0x10	/* Bus-error reporting */
-- 
2.0.1

[PATCH v2 3/4] can: netlink: Add CAN_CTRLMODE_PRESUME_ACK flag

[Nikita Edward Baruzdin]

Most CAN controllers have a support for ignoring ACK absence. Some of
them refer to this feature as a self test mode (e. g. SJA1000) and some
include it as a part of a loopback mode (e. g. MCP2510).

Setting the introduced flag via netlink should make CAN controller
perform a successful transmission, even if there is no acknowledgement
(dominant ACK bit) received.

Signed-off-by: Nikita Edward Baruzdin <nebaruzdin@gmail.com>
---
 include/uapi/linux/can/netlink.h | 1 +
 1 file changed, 1 insertion(+)

diff --git a/include/uapi/linux/can/netlink.h b/include/uapi/linux/can/netlink.h
index 3bbf5c7..3e4323a 100644
--- a/include/uapi/linux/can/netlink.h
+++ b/include/uapi/linux/can/netlink.h
@@ -97,6 +97,7 @@ struct can_ctrlmode {
 #define CAN_CTRLMODE_ONE_SHOT		0x08	/* One-Shot mode */
 #define CAN_CTRLMODE_BERR_REPORTING	0x10	/* Bus-error reporting */
 #define CAN_CTRLMODE_FD			0x20	/* CAN FD mode */
+#define CAN_CTRLMODE_PRESUME_ACK	0x40	/* Ignore missing CAN ACKs */
 
 /*
  * CAN device statistics
-- 
2.0.1

[PATCH v2 4/4] can: sja1000: Add support for CAN_CTRLMODE_PRESUME_ACK

[Nikita Edward Baruzdin]

SJA1000 has a self test mode (STM) which does not require
acknowledgement for the successful message transmission. In this mode a
node test is possible without any other active node on the bus.

This patch adds a possibility to set STM for SJA1000 controller through
specifying the corresponding CAN_CTRLMODE_PRESUME_ACK netlink flag.

Signed-off-by: Nikita Edward Baruzdin <nebaruzdin@gmail.com>
---
 drivers/net/can/sja1000/sja1000.c | 11 +++++++----
 1 file changed, 7 insertions(+), 4 deletions(-)

diff --git a/drivers/net/can/sja1000/sja1000.c b/drivers/net/can/sja1000/sja1000.c
index 45400d9..d169215 100644
--- a/drivers/net/can/sja1000/sja1000.c
+++ b/drivers/net/can/sja1000/sja1000.c
@@ -141,6 +141,7 @@ static void set_normal_mode(struct net_device *dev)
 {
 	struct sja1000_priv *priv = netdev_priv(dev);
 	unsigned char status = priv->read_reg(priv, SJA1000_MOD);
+	u8 mod_reg_val = 0x00;
 	int i;
 
 	for (i = 0; i < 100; i++) {
@@ -158,9 +159,10 @@ static void set_normal_mode(struct net_device *dev)
 
 		/* set chip to normal mode */
 		if (priv->can.ctrlmode & CAN_CTRLMODE_LISTENONLY)
-			priv->write_reg(priv, SJA1000_MOD, MOD_LOM);
-		else
-			priv->write_reg(priv, SJA1000_MOD, 0x00);
+			mod_reg_val |= MOD_LOM;
+		if (priv->can.ctrlmode & CAN_CTRLMODE_PRESUME_ACK)
+			mod_reg_val |= MOD_STM;
+		priv->write_reg(priv, SJA1000_MOD, mod_reg_val);
 
 		udelay(10);
 
@@ -632,7 +634,8 @@ struct net_device *alloc_sja1000dev(int sizeof_priv)
 				       CAN_CTRLMODE_LISTENONLY |
 				       CAN_CTRLMODE_3_SAMPLES |
 				       CAN_CTRLMODE_ONE_SHOT |
-				       CAN_CTRLMODE_BERR_REPORTING;
+				       CAN_CTRLMODE_BERR_REPORTING |
+				       CAN_CTRLMODE_PRESUME_ACK;
 
 	spin_lock_init(&priv->cmdreg_lock);
 
-- 
2.0.1

Re: [PATCH v2 1/4] can: sja1000: Add support for CAN_CTRLMODE_LOOPBACK

[Marc Kleine-Budde]

[-- Attachment #1: Type: text/plain, Size: 796 bytes --]

On 07/11/2014 02:13 PM, Nikita Edward Baruzdin wrote:
> This adds support for hardware loopback in SJA1000 by utilising its self
> reception request (SRR) feature. Upon SRR the message is transmitted and
> received simultaneously, meaning you can't have hardware loopback
> without actually sending a message to the CAN bus in case of SJA1000.
> 
> Signed-off-by: Nikita Edward Baruzdin <nebaruzdin@gmail.com>

Thanks, applied to can-next. Do you care to provide patches for ip and
libsocketcan, too?

Marc

-- 
Pengutronix e.K.                  | Marc Kleine-Budde           |
Industrial Linux Solutions        | Phone: +49-231-2826-924     |
Vertretung West/Dortmund          | Fax:   +49-5121-206917-5555 |
Amtsgericht Hildesheim, HRA 2686  | http://www.pengutronix.de   |


[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 242 bytes --]

Re: [PATCH v2 1/4] can: sja1000: Add support for CAN_CTRLMODE_LOOPBACK

[Nikita Edward Baruzdin]

On Tue, 2014-07-15 at 09:35 +0200, Marc Kleine-Budde wrote:
> Thanks, applied to can-next. Do you care to provide patches for ip and
> libsocketcan, too?

Thank you. Yes, I'm going to a bit later.