Skip to contentSkip to conversation

kvaser_{pciefd,usb} fixes

PATCH Superseded Jimmy Assarsson 3 participants 13 messages Nov 19, 2021 – Dec 8, 2021 CAN network drivers

v1 of 5 Superseded Latest is v5 →

lore
Share & tools
lore
#0 lore raw

cover kvaser_{pciefd,usb} fixes

The first patches fix net_device_stats when receiving error frames. The last two patches fix the CAN clock frequency for Kvaser Leaf devices.

Note that this series contains patches for both kvaser_usb and kvaser_pciefd driver.

Jimmy Assarsson (5):
can: kvaser_pciefd: Do not increase stats->rx_{packets,bytes} for
error frames
can: kvaser_pciefd: Increase correct stats->{rx,tx}_errors counter
can: kvaser_usb: Do not increase stats->rx_{packets,bytes} for error
frames
can: kvaser_usb: Use CAN_MHZ define in assignment of frequency
can: kvaser_usb: Get CAN clock frequency from device

 drivers/net/can/kvaser_pciefd.c               | 11 +--
 .../net/can/usb/kvaser_usb/kvaser_usb_hydra.c | 13 ++--
 .../net/can/usb/kvaser_usb/kvaser_usb_leaf.c  | 70 ++++++++++++++++---
 3 files changed, 72 insertions(+), 22 deletions(-)
Signature · 1 line
2.31.1
#1 lore raw

PATCH 1/5 can: kvaser_pciefd: Do not increase stats->rx_{packets,bytes} for error frames

Do not increase net_device_stats rx_{packets,bytes} when receiving error frames.

---
 drivers/net/can/kvaser_pciefd.c | 3 ---
 1 file changed, 3 deletions(-)
drivers/net/can/kvaser_pciefd.c +0 -3
index 74d9899fc904..2c98befcf2a0 100644
@@ -1304,9 +1304,6 @@ static int kvaser_pciefd_rx_error_frame(struct kvaser_pciefd_can *can,
 	cf->data[6] = bec.txerr;
 	cf->data[7] = bec.rxerr;
 
-	stats->rx_packets++;
-	stats->rx_bytes += cf->len;
-
 	netif_rx(skb);
 	return 0;
 }
Signature · 1 line
2.31.1
#2 lore raw

Re: [PATCH v1 1/5] can: kvaser_pciefd: Do not increase stats->rx_{packets,bytes} for error frames

Quoted 13 lines · Jimmy Assarsson

Do not increase net_device_stats rx_{packets,bytes} when receiving error frames.

---
 drivers/net/can/kvaser_pciefd.c | 3 ---
 1 file changed, 3 deletions(-)
drivers/net/can/kvaser_pciefd.c +0 -0
index 74d9899fc904..2c98befcf2a0 100644

drivers/net/can/kvaser_pciefd.c static int kvaser_pciefd_rx_error_frame(struct kvaser_pciefd_can *can, · line 1309

Quoted 9 lines
drivers/net/can/kvaser_pciefd.c +0 -3
@@ -1304,9 +1304,6 @@ static int kvaser_pciefd_rx_error_frame(struct kvaser_pciefd_can *can,
 	cf->data[6] = bec.txerr;
 	cf->data[7] = bec.rxerr;
 
-	stats->rx_packets++;
-	stats->rx_bytes += cf->len;
-
 	netif_rx(skb);
 	return 0;
 }

I think that this patch makes sense because the CAN error frames do not exists on the wire: only an error flag is transmitted.

However, the current consensus is that the rx_packets and rx_bytes statistics should be incremented for CAN error frames. And I think that consistency between the drivers is the first priority.

I inquired here in the past to ask if it made sense to stop increasing the rx stats for CAN error frames:

https://lore.kernel.org/linux-can/CAMZ6Rq+8YSRqXU7CPrT9FKnWZ1G9xkSr3wt185r2CswmxhXPVg@mail.gmail.com/t/#u

But the discussion did not raise interest. I am fine to send a tree wide cleaning patch, but first, I would like to have people agree on this.

#3 lore raw

Re: [PATCH v1 1/5] can: kvaser_pciefd: Do not increase stats->rx_{packets,bytes} for error frames

Quoted 31 lines · Vincent Mailhol
Quoted 13 lines · Jimmy Assarsson

Do not increase net_device_stats rx_{packets,bytes} when receiving error frames.

---
  drivers/net/can/kvaser_pciefd.c | 3 ---
  1 file changed, 3 deletions(-)
drivers/net/can/kvaser_pciefd.c +0 -0
index 74d9899fc904..2c98befcf2a0 100644

drivers/net/can/kvaser_pciefd.c static int kvaser_pciefd_rx_error_frame(struct kvaser_pciefd_can *can, · line 1309

Quoted 9 lines
drivers/net/can/kvaser_pciefd.c +0 -3
@@ -1304,9 +1304,6 @@ static int kvaser_pciefd_rx_error_frame(struct kvaser_pciefd_can *can,
  	cf->data[6] = bec.txerr;
  	cf->data[7] = bec.rxerr;
 
-	stats->rx_packets++;
-	stats->rx_bytes += cf->len;
-
  	netif_rx(skb);
  	return 0;
  }

I think that this patch makes sense because the CAN error frames do not exists on the wire: only an error flag is transmitted.

However, the current consensus is that the rx_packets and rx_bytes statistics should be incremented for CAN error frames. And I think that consistency between the drivers is the first priority.

Quoted 8 lines

I inquired here in the past to ask if it made sense to stop increasing the rx stats for CAN error frames:

https://lore.kernel.org/linux-can/CAMZ6Rq+8YSRqXU7CPrT9FKnWZ1G9xkSr3wt185r2CswmxhXPVg@mail.gmail.com/t/#u

But the discussion did not raise interest. I am fine to send a tree wide cleaning patch, but first, I would like to have people agree on this.

Thanks, I've not seen that discussion.
I agree, it doesn't make sense for a user to see the rx_packets and rx_bytes increase, when in reality they didn't receive any valid frames.

This change started as a request from one of our customers, that got confused by the increase of rx_packets, when connected to a bus with wrong bitrate.

I'm in favour of a tree wide patch.

Best regards,
jimmy

#4 lore raw

PATCH 2/5 can: kvaser_pciefd: Increase correct stats->{rx,tx}_errors counter

In reply to Jimmy Assarsson

Check the direction bit in the error frame packet (EPACK) to determine which net_device_stats {rx,tx}_errors counter to increase.

---
 drivers/net/can/kvaser_pciefd.c | 8 +++++++-
 1 file changed, 7 insertions(+), 1 deletion(-)
drivers/net/can/kvaser_pciefd.c +7 -1
index 2c98befcf2a0..6fb18ced1806 100644
@@ -248,6 +248,9 @@ MODULE_DESCRIPTION("CAN driver for Kvaser CAN/PCIe devices");
 #define KVASER_PCIEFD_SPACK_EWLR BIT(23)
 #define KVASER_PCIEFD_SPACK_EPLR BIT(24)
 
+/* Kvaser KCAN_EPACK second word */
+#define KVASER_PCIEFD_EPACK_DIR_TX BIT(0)
+
 struct kvaser_pciefd;
 
 struct kvaser_pciefd_can {
@@ -1285,7 +1288,10 @@ static int kvaser_pciefd_rx_error_frame(struct kvaser_pciefd_can *can,
 
 	can->err_rep_cnt++;
 	can->can.can_stats.bus_error++;
-	stats->rx_errors++;
+	if (p->header[1] & KVASER_PCIEFD_EPACK_DIR_TX)
+		stats->tx_errors++;
+	else
+		stats->rx_errors++;
 
 	can->bec.txerr = bec.txerr;
 	can->bec.rxerr = bec.rxerr;
Signature · 1 line
2.31.1
#5 lore raw

PATCH 3/5 can: kvaser_usb: Do not increase stats->rx_{packets,bytes} for error frames

In reply to Jimmy Assarsson

Do not increase net_device_stats rx_{packets,bytes} when receiving error frames.

---
 drivers/net/can/usb/kvaser_usb/kvaser_usb_hydra.c | 6 ------
 1 file changed, 6 deletions(-)
drivers/net/can/usb/kvaser_usb/kvaser_usb_hydra.c +0 -6
index dcee8dc828ec..aa81b92237ca 100644
@@ -920,8 +920,6 @@ static void kvaser_usb_hydra_update_state(struct kvaser_usb_net_priv *priv,
 	cf->data[7] = bec->rxerr;
 
 	stats = &netdev->stats;
-	stats->rx_packets++;
-	stats->rx_bytes += cf->len;
 	netif_rx(skb);
 }
 
@@ -1074,8 +1072,6 @@ kvaser_usb_hydra_error_frame(struct kvaser_usb_net_priv *priv,
 	cf->data[6] = bec.txerr;
 	cf->data[7] = bec.rxerr;
 
-	stats->rx_packets++;
-	stats->rx_bytes += cf->len;
 	netif_rx(skb);
 
 	priv->bec.txerr = bec.txerr;
@@ -1109,8 +1105,6 @@ static void kvaser_usb_hydra_one_shot_fail(struct kvaser_usb_net_priv *priv,
 	}
 
 	stats->tx_errors++;
-	stats->rx_packets++;
-	stats->rx_bytes += cf->len;
 	netif_rx(skb);
 }
 
Signature · 1 line
2.31.1
#6 lore raw

PATCH 4/5 can: kvaser_usb: Use CAN_MHZ define in assignment of frequency

In reply to Jimmy Assarsson

Use the CAN_MHZ define when assigning frequencies.

---
 drivers/net/can/usb/kvaser_usb/kvaser_usb_hydra.c | 7 ++++---
 drivers/net/can/usb/kvaser_usb/kvaser_usb_leaf.c  | 3 ++-
 2 files changed, 6 insertions(+), 4 deletions(-)
drivers/net/can/usb/kvaser_usb/kvaser_usb_hydra.c +4 -3
index aa81b92237ca..221339338414 100644
@@ -25,6 +25,7 @@
 #include <linux/usb.h>
 
 #include <linux/can.h>
+#include <linux/can/bittiming.h>
 #include <linux/can/dev.h>
 #include <linux/can/error.h>
 #include <linux/can/netlink.h>
@@ -2034,7 +2035,7 @@ const struct kvaser_usb_dev_ops kvaser_usb_hydra_dev_ops = {
 
 static const struct kvaser_usb_dev_cfg kvaser_usb_hydra_dev_cfg_kcan = {
 	.clock = {
-		.freq = 80000000,
+		.freq = 80 * CAN_MHZ,
 	},
 	.timestamp_freq = 80,
 	.bittiming_const = &kvaser_usb_hydra_kcan_bittiming_c,
@@ -2043,7 +2044,7 @@ static const struct kvaser_usb_dev_cfg kvaser_usb_hydra_dev_cfg_kcan = {
 
 static const struct kvaser_usb_dev_cfg kvaser_usb_hydra_dev_cfg_flexc = {
 	.clock = {
-		.freq = 24000000,
+		.freq = 24 * CAN_MHZ,
 	},
 	.timestamp_freq = 1,
 	.bittiming_const = &kvaser_usb_hydra_flexc_bittiming_c,
@@ -2051,7 +2052,7 @@ static const struct kvaser_usb_dev_cfg kvaser_usb_hydra_dev_cfg_flexc = {
 
 static const struct kvaser_usb_dev_cfg kvaser_usb_hydra_dev_cfg_rt = {
 	.clock = {
-		.freq = 80000000,
+		.freq = 80 * CAN_MHZ,
 	},
 	.timestamp_freq = 24,
 	.bittiming_const = &kvaser_usb_hydra_rt_bittiming_c,
drivers/net/can/usb/kvaser_usb/kvaser_usb_leaf.c +2 -1
index 59ba7c7beec0..506e4b76d119 100644
@@ -22,6 +22,7 @@
 #include <linux/usb.h>
 
 #include <linux/can.h>
+#include <linux/can/bittiming.h>
 #include <linux/can/dev.h>
 #include <linux/can/error.h>
 #include <linux/can/netlink.h>
@@ -31,7 +32,7 @@
 /* Forward declaration */
 static const struct kvaser_usb_dev_cfg kvaser_usb_leaf_dev_cfg;
 
-#define CAN_USB_CLOCK			8000000
+#define CAN_USB_CLOCK			(8 * CAN_MHZ)
 #define MAX_USBCAN_NET_DEVICES		2
 
 /* Command header size */
Signature · 1 line
2.31.1
#7 lore raw

Re: [PATCH v1 4/5] can: kvaser_usb: Use CAN_MHZ define in assignment of frequency

Quoted 1 line · Jimmy Assarsson

Use the CAN_MHZ define when assigning frequencies.

Maybe we should use the HZ_PER_MHZ define introduced in kernel 5.15 in linux/units.h. It also got defines for KILO and MEGA, which we also got in linux/can/bittiming.h.

What do you think?

...

Quoted 5 lines
  static const struct kvaser_usb_dev_cfg kvaser_usb_hydra_dev_cfg_kcan = {
  	.clock = {
drivers/net/can/usb/kvaser_usb/kvaser_usb_hydra.c +1 -1
-		.freq = 80000000,
+		.freq = 80 * CAN_MHZ,
  	},
#8 lore raw

Re: [PATCH v1 4/5] can: kvaser_usb: Use CAN_MHZ define in assignment of frequency

Quoted 9 lines · Jimmy Assarsson
Quoted 1 line · Jimmy Assarsson

Use the CAN_MHZ define when assigning frequencies.

Maybe we should use the HZ_PER_MHZ define introduced in kernel 5.15 in linux/units.h. It also got defines for KILO and MEGA, which we also got in linux/can/bittiming.h.

What do you think?

With the recent changes in linux/units.h, I am perfectly fine to remove the CAN_MHZ, CAN_KBPS and CAN_MBPS macros from linux/can/bittiming.h.

I am just bothered by the name of the macro HZ_PER_MHZ. Hertz per Mega Hertz is a ratio without a unit. The clock speed is expressed in hertz. So:

.frep = 80 * MEGA,

#9 lore raw

Re: [PATCH v1 4/5] can: kvaser_usb: Use CAN_MHZ define in assignment of frequency

(I inadvertently sent my previous message while I was still writing it, resending with full comments)

Quoted 9 lines · Jimmy Assarsson
Quoted 1 line · Jimmy Assarsson

Use the CAN_MHZ define when assigning frequencies.

Maybe we should use the HZ_PER_MHZ define introduced in kernel 5.15 in linux/units.h. It also got defines for KILO and MEGA, which we also got in linux/can/bittiming.h.

What do you think?

With the recent changes in linux/units.h, I am perfectly fine to remove the CAN_MHZ, CAN_KBPS and CAN_MBPS macros from linux/can/bittiming.h.

I am just bothered by the name of the macro HZ_PER_MHZ. Hertz per Mega Hertz is a ratio without a unit (physically speaking). The clock speed is expressed in hertz. So:

| .frep = 80 * MEGA,

would make more sense.

I see HZ_PER_MHZ as something used for conversion, e.g.:
| freq_in_hz = HZ_PER_MHZ * frep_in_mhz;

Yours sincerely,
Vincent Mailhol

#10 lore raw

PATCH 5/5 can: kvaser_usb: Get CAN clock frequency from device

In reply to Jimmy Assarsson

Get the CAN clock frequency from device, since the various Kvaser Leaf products use different CAN clocks.

---
 .../net/can/usb/kvaser_usb/kvaser_usb_leaf.c  | 69 ++++++++++++++++---
 1 file changed, 60 insertions(+), 9 deletions(-)
drivers/net/can/usb/kvaser_usb/kvaser_usb_leaf.c +60 -9
index 506e4b76d119..5153af71c47a 100644
@@ -29,10 +29,12 @@
 
 #include "kvaser_usb.h"
 
-/* Forward declaration */
-static const struct kvaser_usb_dev_cfg kvaser_usb_leaf_dev_cfg;
+/* Forward declarations */
+static const struct kvaser_usb_dev_cfg kvaser_usb_leaf_dev_cfg_8mhz;
+static const struct kvaser_usb_dev_cfg kvaser_usb_leaf_dev_cfg_16mhz;
+static const struct kvaser_usb_dev_cfg kvaser_usb_leaf_dev_cfg_24mhz;
+static const struct kvaser_usb_dev_cfg kvaser_usb_leaf_dev_cfg_32mhz;
 
-#define CAN_USB_CLOCK			(8 * CAN_MHZ)
 #define MAX_USBCAN_NET_DEVICES		2
 
 /* Command header size */
@@ -81,6 +83,12 @@ static const struct kvaser_usb_dev_cfg kvaser_usb_leaf_dev_cfg;
 
 #define CMD_LEAF_LOG_MESSAGE		106
 
+/* Leaf frequency options */
+#define KVASER_USB_LEAF_SWOPTION_FREQ_MASK 0x60
+#define KVASER_USB_LEAF_SWOPTION_FREQ_16_MHZ_CLK 0
+#define KVASER_USB_LEAF_SWOPTION_FREQ_32_MHZ_CLK BIT(5)
+#define KVASER_USB_LEAF_SWOPTION_FREQ_24_MHZ_CLK BIT(6)
+
 /* error factors */
 #define M16C_EF_ACKE			BIT(0)
 #define M16C_EF_CRCE			BIT(1)
@@ -472,6 +480,27 @@ static int kvaser_usb_leaf_send_simple_cmd(const struct kvaser_usb *dev,
 	return rc;
 }
 
+static void kvaser_usb_leaf_get_software_info_leaf(struct kvaser_usb *dev,
+						   const struct leaf_cmd_softinfo *softinfo)
+{
+	u32 sw_options = le32_to_cpu(softinfo->sw_options);
+
+	dev->fw_version = le32_to_cpu(softinfo->fw_version);
+	dev->max_tx_urbs = le16_to_cpu(softinfo->max_outstanding_tx);
+
+	switch (sw_options & KVASER_USB_LEAF_SWOPTION_FREQ_MASK) {
+	case KVASER_USB_LEAF_SWOPTION_FREQ_16_MHZ_CLK:
+		dev->cfg = &kvaser_usb_leaf_dev_cfg_16mhz;
+		break;
+	case KVASER_USB_LEAF_SWOPTION_FREQ_24_MHZ_CLK:
+		dev->cfg = &kvaser_usb_leaf_dev_cfg_24mhz;
+		break;
+	case KVASER_USB_LEAF_SWOPTION_FREQ_32_MHZ_CLK:
+		dev->cfg = &kvaser_usb_leaf_dev_cfg_32mhz;
+		break;
+	}
+}
+
 static int kvaser_usb_leaf_get_software_info_inner(struct kvaser_usb *dev)
 {
 	struct kvaser_cmd cmd;
@@ -487,14 +516,13 @@ static int kvaser_usb_leaf_get_software_info_inner(struct kvaser_usb *dev)
 
 	switch (dev->card_data.leaf.family) {
 	case KVASER_LEAF:
-		dev->fw_version = le32_to_cpu(cmd.u.leaf.softinfo.fw_version);
-		dev->max_tx_urbs =
-			le16_to_cpu(cmd.u.leaf.softinfo.max_outstanding_tx);
+		kvaser_usb_leaf_get_software_info_leaf(dev, &cmd.u.leaf.softinfo);
 		break;
 	case KVASER_USBCAN:
 		dev->fw_version = le32_to_cpu(cmd.u.usbcan.softinfo.fw_version);
 		dev->max_tx_urbs =
 			le16_to_cpu(cmd.u.usbcan.softinfo.max_outstanding_tx);
+		dev->cfg = &kvaser_usb_leaf_dev_cfg_8mhz;
 		break;
 	}
 
@@ -1226,7 +1254,6 @@ static int kvaser_usb_leaf_init_card(struct kvaser_usb *dev)
 {
 	struct kvaser_usb_dev_card_data *card_data = &dev->card_data;
 
-	dev->cfg = &kvaser_usb_leaf_dev_cfg;
 	card_data->ctrlmode_supported |= CAN_CTRLMODE_3_SAMPLES;
 
 	return 0;
@@ -1350,9 +1377,33 @@ const struct kvaser_usb_dev_ops kvaser_usb_leaf_dev_ops = {
 	.dev_frame_to_cmd = kvaser_usb_leaf_frame_to_cmd,
 };
 
-static const struct kvaser_usb_dev_cfg kvaser_usb_leaf_dev_cfg = {
+static const struct kvaser_usb_dev_cfg kvaser_usb_leaf_dev_cfg_8mhz = {
+	.clock = {
+		.freq = 8 * CAN_MHZ,
+	},
+	.timestamp_freq = 1,
+	.bittiming_const = &kvaser_usb_leaf_bittiming_const,
+};
+
+static const struct kvaser_usb_dev_cfg kvaser_usb_leaf_dev_cfg_16mhz = {
+	.clock = {
+		.freq = 16 * CAN_MHZ,
+	},
+	.timestamp_freq = 1,
+	.bittiming_const = &kvaser_usb_leaf_bittiming_const,
+};
+
+static const struct kvaser_usb_dev_cfg kvaser_usb_leaf_dev_cfg_24mhz = {
+	.clock = {
+		.freq = 24 * CAN_MHZ,
+	},
+	.timestamp_freq = 1,
+	.bittiming_const = &kvaser_usb_leaf_bittiming_const,
+};
+
+static const struct kvaser_usb_dev_cfg kvaser_usb_leaf_dev_cfg_32mhz = {
 	.clock = {
-		.freq = CAN_USB_CLOCK,
+		.freq = 32 * CAN_MHZ,
 	},
 	.timestamp_freq = 1,
 	.bittiming_const = &kvaser_usb_leaf_bittiming_const,
Signature · 1 line
2.31.1
#11 lore raw
In reply to Jimmy Assarsson
Quoted 5 lines · Jimmy Assarsson

The first patches fix net_device_stats when receiving error frames. The last two patches fix the CAN clock frequency for Kvaser Leaf devices.

Note that this series contains patches for both kvaser_usb and kvaser_pciefd driver.

What's left of this series, after applying Vincent's stats cleanup [1] series?

[1]
https://lore.kernel.org/all/[email protected]/

Can you send a series only containing the remaining patches?

regards,
Marc

Signature · 4 lines
Pengutronix e.K.                 | Marc Kleine-Budde           |
Embedded Linux                   | https://www.pengutronix.de  |
Vertretung West/Dortmund         | Phone: +49-231-2826-924     |
Amtsgericht Hildesheim, HRA 2686 | Fax:   +49-5121-206917-5555 |
#12 lore raw
Quoted 7 lines · Marc Kleine-Budde
Quoted 5 lines · Jimmy Assarsson

The first patches fix net_device_stats when receiving error frames. The last two patches fix the CAN clock frequency for Kvaser Leaf devices.

Note that this series contains patches for both kvaser_usb and kvaser_pciefd driver.

Quoted 7 lines

What's left of this series, after applying Vincent's stats cleanup [1] series?

[1]
https://lore.kernel.org/all/[email protected]/

Can you send a series only containing the remaining patches?

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

Best regards,
jimmy

Quoted 3 lines

regards,
Marc