[PATCH net-next 00/11] net: dsa: add generic debugfs interface

[PATCH net-next 00/11] net: dsa: add generic debugfs interface

[Vivien Didelot]

This patch series adds a generic debugfs interface for the DSA
framework, so that all switch devices benefit from it, e.g. Marvell,
Broadcom, Microchip or any other DSA driver.

This is really convenient for debugging, especially CPU ports and DSA
links which are not exposed to userspace as net device. This interface
is currently the only way to easily inspect the hardware for such ports.

With the patch series, any switch device user is able to query the
hardware for the supported tagging protocol, the ports stats and
registers, as well as their FDB, MDB and VLAN entries.

This support is only compiled if CONFIG_DEBUG_FS is enabled. Below is
and example of usage of this interface on a multi-chip switch fabric:

    # mount -t debugfs none /sys/kernel/debug
    # cd /sys/kernel/debug/dsa/
    # ls
    switch0  switch1 switch2
    # ls -l switch0/
    drwxr-xr-x 2 root root 0 Jan  1 00:00 port0
    drwxr-xr-x 2 root root 0 Jan  1 00:00 port1
    drwxr-xr-x 2 root root 0 Jan  1 00:00 port2
    drwxr-xr-x 2 root root 0 Jan  1 00:00 port5
    drwxr-xr-x 2 root root 0 Jan  1 00:00 port6
    -r--r--r-- 1 root root 0 Jan  1 00:00 tag_protocol
    -r--r--r-- 1 root root 0 Jan  1 00:00 tree
    # ls -l switch0/port6
    -r--r--r-- 1 root root 0 Jan  1 00:00 fdb
    -r--r--r-- 1 root root 0 Jan  1 00:00 mdb
    -r--r--r-- 1 root root 0 Jan  1 00:00 regs
    -r--r--r-- 1 root root 0 Jan  1 00:00 stats
    -r--r--r-- 1 root root 0 Jan  1 00:00 vlan
    # cat switch0/port2/vlan
    vid 42  pvid  untagged
    # cat switch0/port1/fdb
    vid 0    12:34:56:78:90:ab    static    unicast
    # pr -mt switch0/port{5,6}/stats
    in_good_octets      : 0             in_good_octets      : 13824
    in_bad_octets       : 0             in_bad_octets       : 0
    in_unicast          : 0             in_unicast          : 0
    in_broadcasts       : 0             in_broadcasts       : 216
    in_multicasts       : 0             in_multicasts       : 0
    in_pause            : 0             in_pause            : 0
    in_undersize        : 0             in_undersize        : 0
    ...
    # pr -mt switch0/port{5,6}/regs
     0: 4e07			     0: 4d04
     1: 403e			     1: 003d
     2: 0000			     2: 0000
     3: 3521			     3: 3521
     4: 0533			     4: 373f
     5: 8000			     5: 0000
     6: 005f			     6: 003f
     7: 002a			     7: 002a
    ...

where switch0 port5 and port6 are CPU and DSA ports of a ZII Rev B.

Vivien Didelot (11):
  net: dsa: legacy: assign dst->applied
  net: dsa: add debugfs interface
  net: dsa: debugfs: add tree
  net: dsa: debugfs: add tag_protocol
  net: dsa: debugfs: add port stats
  net: dsa: debugfs: add port registers
  net: dsa: debugfs: add port fdb
  net: dsa: restore mdb dump
  net: dsa: debugfs: add port mdb
  net: dsa: restore VLAN dump
  net: dsa: debugfs: add port vlan

 drivers/net/dsa/b53/b53_common.c       |  41 +++
 drivers/net/dsa/b53/b53_priv.h         |   2 +
 drivers/net/dsa/bcm_sf2.c              |   1 +
 drivers/net/dsa/dsa_loop.c             |  38 +++
 drivers/net/dsa/microchip/ksz_common.c |  41 +++
 drivers/net/dsa/mv88e6xxx/chip.c       |  82 +++++-
 include/net/dsa.h                      |  15 ++
 net/dsa/Kconfig                        |  14 +
 net/dsa/Makefile                       |   1 +
 net/dsa/debugfs.c                      | 453 +++++++++++++++++++++++++++++++++
 net/dsa/dsa.c                          |   3 +
 net/dsa/dsa2.c                         |   4 +
 net/dsa/dsa_priv.h                     |  13 +
 net/dsa/legacy.c                       |   7 +
 14 files changed, 707 insertions(+), 8 deletions(-)
 create mode 100644 net/dsa/debugfs.c

-- 
2.14.0

[PATCH net-next 01/11] net: dsa: legacy: assign dst->applied

[Vivien Didelot]

The "applied" boolean of the dsa_switch_tree is only set in the new
bindings. This patch sets it in the legacy code as well.

Signed-off-by: Vivien Didelot <vivien.didelot@savoirfairelinux.com>
---
 net/dsa/legacy.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/net/dsa/legacy.c b/net/dsa/legacy.c
index 91e6f7981d39..a6a0849483d1 100644
--- a/net/dsa/legacy.c
+++ b/net/dsa/legacy.c
@@ -605,6 +605,7 @@ static int dsa_setup_dst(struct dsa_switch_tree *dst, struct net_device *dev,
 	 */
 	wmb();
 	dev->dsa_ptr = dst;
+	dst->applied = true;
 
 	return 0;
 }
@@ -689,6 +690,8 @@ static void dsa_remove_dst(struct dsa_switch_tree *dst)
 	dsa_cpu_port_ethtool_restore(dst->cpu_dp);
 
 	dev_put(dst->cpu_dp->netdev);
+
+	dst->applied = false;
 }
 
 static int dsa_remove(struct platform_device *pdev)
-- 
2.14.0

Re: [PATCH net-next 01/11] net: dsa: legacy: assign dst->applied

[Andrew Lunn]

On Mon, Aug 14, 2017 at 06:22:32PM -0400, Vivien Didelot wrote:
> The "applied" boolean of the dsa_switch_tree is only set in the new
> bindings. This patch sets it in the legacy code as well.

What is missing here is: Why? 

I see the next patch does a WARN_ON(!applied). Why have a WARN_ON?

  Andrew

[PATCH net-next 02/11] net: dsa: add debugfs interface

[Vivien Didelot]

This commit adds a DEBUG_FS dependent DSA core file creating a generic
debug filesystem interface for the DSA switch devices.

The interface can be mounted with:

    # mount -t debugfs none /sys/kernel/debug

The dsa directory contains one directory per switch chip:

    # cd /sys/kernel/debug/dsa/
    # ls
    switch0  switch1 switch2

Each chip directory contains one directory per port:

    # ls -l switch0/
    drwxr-xr-x 2 root root 0 Jan  1 00:00 port0
    drwxr-xr-x 2 root root 0 Jan  1 00:00 port1
    drwxr-xr-x 2 root root 0 Jan  1 00:00 port2
    drwxr-xr-x 2 root root 0 Jan  1 00:00 port5
    drwxr-xr-x 2 root root 0 Jan  1 00:00 port6

Future patches will add entry files to these directories.

Signed-off-by: Vivien Didelot <vivien.didelot@savoirfairelinux.com>
---
 include/net/dsa.h  |   7 ++++
 net/dsa/Kconfig    |  14 +++++++
 net/dsa/Makefile   |   1 +
 net/dsa/debugfs.c  | 121 +++++++++++++++++++++++++++++++++++++++++++++++++++++
 net/dsa/dsa.c      |   3 ++
 net/dsa/dsa2.c     |   4 ++
 net/dsa/dsa_priv.h |  13 ++++++
 net/dsa/legacy.c   |   4 ++
 8 files changed, 167 insertions(+)
 create mode 100644 net/dsa/debugfs.c

diff --git a/include/net/dsa.h b/include/net/dsa.h
index 7f46b521313e..4ef5d38755d9 100644
--- a/include/net/dsa.h
+++ b/include/net/dsa.h
@@ -212,6 +212,13 @@ struct dsa_switch {
 	 */
 	void *priv;
 
+#ifdef CONFIG_NET_DSA_DEBUGFS
+	/*
+	 * Debugfs interface.
+	 */
+	struct dentry *debugfs_dir;
+#endif
+
 	/*
 	 * Configuration data for this switch.
 	 */
diff --git a/net/dsa/Kconfig b/net/dsa/Kconfig
index cc5f8f971689..0f05a1e59dd2 100644
--- a/net/dsa/Kconfig
+++ b/net/dsa/Kconfig
@@ -15,6 +15,20 @@ config NET_DSA
 
 if NET_DSA
 
+config NET_DSA_DEBUGFS
+	bool "Distributed Switch Architecture debugfs interface"
+	depends on DEBUG_FS
+	---help---
+	  Enable creation of debugfs files for the DSA core.
+
+	  These debugfs files provide per-switch information, such as the tag
+	  protocol in use and ports connectivity. They also allow querying the
+	  hardware directly through the switch operations for debugging instead
+	  of going through the bridge, switchdev and DSA layers.
+
+	  This is also a way to inspect the stats and FDB, MDB or VLAN entries
+	  of CPU and DSA links, since they are not exposed to userspace.
+
 # tagging formats
 config NET_DSA_TAG_BRCM
 	bool
diff --git a/net/dsa/Makefile b/net/dsa/Makefile
index fcce25da937c..7f60c6dfaffb 100644
--- a/net/dsa/Makefile
+++ b/net/dsa/Makefile
@@ -1,6 +1,7 @@
 # the core
 obj-$(CONFIG_NET_DSA) += dsa_core.o
 dsa_core-y += dsa.o dsa2.o legacy.o port.o slave.o switch.o
+dsa_core-$(CONFIG_NET_DSA_DEBUGFS) += debugfs.o
 
 # tagging formats
 dsa_core-$(CONFIG_NET_DSA_TAG_BRCM) += tag_brcm.o
diff --git a/net/dsa/debugfs.c b/net/dsa/debugfs.c
new file mode 100644
index 000000000000..68caf5a2c0c3
--- /dev/null
+++ b/net/dsa/debugfs.c
@@ -0,0 +1,121 @@
+/*
+ * net/dsa/debugfs.c - DSA debugfs interface
+ * Copyright (c) 2017 Savoir-faire Linux, Inc.
+ *	Vivien Didelot <vivien.didelot@savoirfairelinux.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ */
+
+#include <linux/debugfs.h>
+
+#include "dsa_priv.h"
+
+#define DSA_SWITCH_FMT	"switch%d"
+#define DSA_PORT_FMT	"port%d"
+
+/* DSA module debugfs directory */
+static struct dentry *dsa_debugfs_dir;
+
+static int dsa_debugfs_create_port(struct dsa_switch *ds, int port)
+{
+	struct dentry *dir;
+	char name[32];
+
+	snprintf(name, sizeof(name), DSA_PORT_FMT, port);
+
+	dir = debugfs_create_dir(name, ds->debugfs_dir);
+	if (IS_ERR_OR_NULL(dir))
+		return -EFAULT;
+
+	return 0;
+}
+
+static int dsa_debugfs_create_switch(struct dsa_switch *ds)
+{
+	char name[32];
+	int i;
+
+	/* skip if there is no debugfs support */
+	if (!dsa_debugfs_dir)
+		return 0;
+
+	snprintf(name, sizeof(name), DSA_SWITCH_FMT, ds->index);
+
+	ds->debugfs_dir = debugfs_create_dir(name, dsa_debugfs_dir);
+	if (IS_ERR_OR_NULL(ds->debugfs_dir))
+		return -EFAULT;
+
+	for (i = 0; i < ds->num_ports; i++) {
+		if (!ds->ports[i].dn)
+			continue;
+
+		err = dsa_debugfs_create_port(ds, i);
+		if (err)
+			return err;
+	}
+
+	return 0;
+}
+
+static void dsa_debugfs_destroy_switch(struct dsa_switch *ds)
+{
+	/* handles NULL */
+	debugfs_remove_recursive(ds->debugfs_dir);
+}
+
+void dsa_debugfs_create_tree(struct dsa_switch_tree *dst)
+{
+	struct dsa_switch *ds;
+	int i, err;
+
+	WARN_ON(!dst->applied);
+
+	for (i = 0; i < DSA_MAX_SWITCHES; i++) {
+		ds = dst->ds[i];
+		if (!ds)
+			continue;
+
+		err = dsa_debugfs_create_switch(ds);
+		if (err) {
+			pr_warn("DSA: failed to create debugfs interface for switch %d (%d)\n",
+				ds->index, err);
+			dsa_debugfs_destroy_tree(dst);
+			break;
+		}
+	}
+}
+
+void dsa_debugfs_destroy_tree(struct dsa_switch_tree *dst)
+{
+	struct dsa_switch *ds;
+	int i;
+
+	for (i = 0; i < DSA_MAX_SWITCHES; i++) {
+		ds = dst->ds[i];
+		if (!ds)
+			continue;
+
+		dsa_debugfs_destroy_switch(ds);
+	}
+}
+
+void dsa_debugfs_create_module(void)
+{
+	dsa_debugfs_dir = debugfs_create_dir("dsa", NULL);
+	if (IS_ERR(dsa_debugfs_dir)) {
+		pr_warn("DSA: failed to create debugfs interface\n");
+		dsa_debugfs_dir = NULL;
+	}
+
+	if (dsa_debugfs_dir)
+		pr_info("DSA: debugfs interface created\n");
+}
+
+void dsa_debugfs_destroy_module(void)
+{
+	/* handles NULL */
+	debugfs_remove_recursive(dsa_debugfs_dir);
+}
diff --git a/net/dsa/dsa.c b/net/dsa/dsa.c
index 99e38af85fc5..62e49ff6d737 100644
--- a/net/dsa/dsa.c
+++ b/net/dsa/dsa.c
@@ -308,12 +308,15 @@ static int __init dsa_init_module(void)
 
 	dev_add_pack(&dsa_pack_type);
 
+	dsa_debugfs_create_module();
+
 	return 0;
 }
 module_init(dsa_init_module);
 
 static void __exit dsa_cleanup_module(void)
 {
+	dsa_debugfs_destroy_module();
 	dsa_slave_unregister_notifier();
 	dev_remove_pack(&dsa_pack_type);
 	dsa_legacy_unregister();
diff --git a/net/dsa/dsa2.c b/net/dsa/dsa2.c
index cceaa4dd9f53..5912618ad63d 100644
--- a/net/dsa/dsa2.c
+++ b/net/dsa/dsa2.c
@@ -447,6 +447,8 @@ static int dsa_dst_apply(struct dsa_switch_tree *dst)
 	dst->cpu_dp->netdev->dsa_ptr = dst;
 	dst->applied = true;
 
+	dsa_debugfs_create_tree(dst);
+
 	return 0;
 }
 
@@ -458,6 +460,8 @@ static void dsa_dst_unapply(struct dsa_switch_tree *dst)
 	if (!dst->applied)
 		return;
 
+	dsa_debugfs_destroy_tree(dst);
+
 	dst->cpu_dp->netdev->dsa_ptr = NULL;
 
 	/* If we used a tagging format that doesn't have an ethertype
diff --git a/net/dsa/dsa_priv.h b/net/dsa/dsa_priv.h
index 9c3eeb72462d..84ca3a50a58b 100644
--- a/net/dsa/dsa_priv.h
+++ b/net/dsa/dsa_priv.h
@@ -93,6 +93,19 @@ struct dsa_slave_priv {
 	struct list_head	mall_tc_list;
 };
 
+/* debugfs.c */
+#ifdef CONFIG_NET_DSA_DEBUGFS
+void dsa_debugfs_create_module(void);
+void dsa_debugfs_destroy_module(void);
+void dsa_debugfs_create_tree(struct dsa_switch_tree *dst);
+void dsa_debugfs_destroy_tree(struct dsa_switch_tree *dst);
+#else
+static inline void dsa_debugfs_create_module(void) { }
+static inline void dsa_debugfs_destroy_module(void) { }
+static inline void dsa_debugfs_create_tree(struct dsa_switch_tree *dst) { }
+static inline void dsa_debugfs_destroy_tree(struct dsa_switch_tree *dst) { }
+#endif
+
 /* dsa.c */
 int dsa_cpu_dsa_setup(struct dsa_port *port);
 void dsa_cpu_dsa_destroy(struct dsa_port *dport);
diff --git a/net/dsa/legacy.c b/net/dsa/legacy.c
index a6a0849483d1..007034ebd218 100644
--- a/net/dsa/legacy.c
+++ b/net/dsa/legacy.c
@@ -607,6 +607,8 @@ static int dsa_setup_dst(struct dsa_switch_tree *dst, struct net_device *dev,
 	dev->dsa_ptr = dst;
 	dst->applied = true;
 
+	dsa_debugfs_create_tree(dst);
+
 	return 0;
 }
 
@@ -672,6 +674,8 @@ static void dsa_remove_dst(struct dsa_switch_tree *dst)
 {
 	int i;
 
+	dsa_debugfs_destroy_tree(dst);
+
 	dst->cpu_dp->netdev->dsa_ptr = NULL;
 
 	/* If we used a tagging format that doesn't have an ethertype
-- 
2.14.0

Re: [PATCH net-next 02/11] net: dsa: add debugfs interface

[Florian Fainelli]

On 08/14/2017 03:22 PM, Vivien Didelot wrote:
> This commit adds a DEBUG_FS dependent DSA core file creating a generic
> debug filesystem interface for the DSA switch devices.
> 
> The interface can be mounted with:
> 
>     # mount -t debugfs none /sys/kernel/debug
> 
> The dsa directory contains one directory per switch chip:
> 
>     # cd /sys/kernel/debug/dsa/
>     # ls
>     switch0  switch1 switch2
> 
> Each chip directory contains one directory per port:
> 
>     # ls -l switch0/
>     drwxr-xr-x 2 root root 0 Jan  1 00:00 port0
>     drwxr-xr-x 2 root root 0 Jan  1 00:00 port1
>     drwxr-xr-x 2 root root 0 Jan  1 00:00 port2
>     drwxr-xr-x 2 root root 0 Jan  1 00:00 port5
>     drwxr-xr-x 2 root root 0 Jan  1 00:00 port6
> 
> Future patches will add entry files to these directories.
> 
> Signed-off-by: Vivien Didelot <vivien.didelot@savoirfairelinux.com>
> ---

> +static int dsa_debugfs_create_switch(struct dsa_switch *ds)
> +{
> +	char name[32];
> +	int i;
> +
> +	/* skip if there is no debugfs support */
> +	if (!dsa_debugfs_dir)
> +		return 0;
> +
> +	snprintf(name, sizeof(name), DSA_SWITCH_FMT, ds->index);
> +
> +	ds->debugfs_dir = debugfs_create_dir(name, dsa_debugfs_dir);
> +	if (IS_ERR_OR_NULL(ds->debugfs_dir))
> +		return -EFAULT;
> +
> +	for (i = 0; i < ds->num_ports; i++) {
> +		if (!ds->ports[i].dn)
> +			continue;

This won't create port directories when using platform data, can you
check for BIT(i) & ds->enabled_port_mask instead?
-- 
Florian

[PATCH net-next 03/11] net: dsa: debugfs: add tree

[Vivien Didelot]

This commit adds the boiler plate to create a DSA related debug
filesystem entry as well as a "tree" file, containing the tree index.

    # cat switch1/tree
    0

Signed-off-by: Vivien Didelot <vivien.didelot@savoirfairelinux.com>
---
 net/dsa/debugfs.c | 108 ++++++++++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 108 insertions(+)

diff --git a/net/dsa/debugfs.c b/net/dsa/debugfs.c
index 68caf5a2c0c3..5607efdb924d 100644
--- a/net/dsa/debugfs.c
+++ b/net/dsa/debugfs.c
@@ -10,6 +10,7 @@
  */
 
 #include <linux/debugfs.h>
+#include <linux/seq_file.h>
 
 #include "dsa_priv.h"
 
@@ -19,6 +20,107 @@
 /* DSA module debugfs directory */
 static struct dentry *dsa_debugfs_dir;
 
+struct dsa_debugfs_ops {
+	int (*read)(struct dsa_switch *ds, int id, struct seq_file *seq);
+	int (*write)(struct dsa_switch *ds, int id, char *buf);
+};
+
+struct dsa_debugfs_priv {
+	const struct dsa_debugfs_ops *ops;
+	struct dsa_switch *ds;
+	int id;
+};
+
+static int dsa_debugfs_show(struct seq_file *seq, void *p)
+{
+	struct dsa_debugfs_priv *priv = seq->private;
+	struct dsa_switch *ds = priv->ds;
+
+	/* Somehow file mode is bypassed... Double check here */
+	if (!priv->ops->read)
+		return -EOPNOTSUPP;
+
+	return priv->ops->read(ds, priv->id, seq);
+}
+
+static ssize_t dsa_debugfs_write(struct file *file, const char __user *user_buf,
+				 size_t count, loff_t *ppos)
+{
+	struct seq_file *seq = file->private_data;
+	struct dsa_debugfs_priv *priv = seq->private;
+	struct dsa_switch *ds = priv->ds;
+	char buf[count + 1];
+	int err;
+
+	/* Somehow file mode is bypassed... Double check here */
+	if (!priv->ops->write)
+		return -EOPNOTSUPP;
+
+	if (copy_from_user(buf, user_buf, count))
+		return -EFAULT;
+
+	buf[count] = '\0';
+
+	err = priv->ops->write(ds, priv->id, buf);
+
+	return err ? err : count;
+}
+
+static int dsa_debugfs_open(struct inode *inode, struct file *file)
+{
+	return single_open(file, dsa_debugfs_show, inode->i_private);
+}
+
+static const struct file_operations dsa_debugfs_fops = {
+	.open = dsa_debugfs_open,
+	.read = seq_read,
+	.write = dsa_debugfs_write,
+	.llseek = no_llseek,
+	.release = single_release,
+	.owner = THIS_MODULE,
+};
+
+static int dsa_debugfs_create_file(struct dsa_switch *ds, struct dentry *dir,
+				   char *name, int id,
+				   const struct dsa_debugfs_ops *ops)
+{
+	struct dsa_debugfs_priv *priv;
+	struct dentry *entry;
+	umode_t mode;
+
+	priv = devm_kzalloc(ds->dev, sizeof(*priv), GFP_KERNEL);
+	if (!priv)
+		return -ENOMEM;
+
+	priv->ops = ops;
+	priv->ds = ds;
+	priv->id = id;
+
+	mode = 0;
+	if (ops->read)
+		mode |= 0444;
+	if (ops->write)
+		mode |= 0200;
+
+	entry = debugfs_create_file(name, mode, dir, priv, &dsa_debugfs_fops);
+	if (IS_ERR_OR_NULL(entry))
+		return -EFAULT;
+
+	return 0;
+}
+
+static int dsa_debugfs_tree_read(struct dsa_switch *ds, int id,
+				 struct seq_file *seq)
+{
+	seq_printf(seq, "%d\n", ds->dst->tree);
+
+	return 0;
+}
+
+static const struct dsa_debugfs_ops dsa_debugfs_tree_ops = {
+	.read = dsa_debugfs_tree_read,
+};
+
 static int dsa_debugfs_create_port(struct dsa_switch *ds, int port)
 {
 	struct dentry *dir;
@@ -36,6 +138,7 @@ static int dsa_debugfs_create_port(struct dsa_switch *ds, int port)
 static int dsa_debugfs_create_switch(struct dsa_switch *ds)
 {
 	char name[32];
+	int err;
 	int i;
 
 	/* skip if there is no debugfs support */
@@ -48,6 +151,11 @@ static int dsa_debugfs_create_switch(struct dsa_switch *ds)
 	if (IS_ERR_OR_NULL(ds->debugfs_dir))
 		return -EFAULT;
 
+	err = dsa_debugfs_create_file(ds, ds->debugfs_dir, "tree", -1,
+				      &dsa_debugfs_tree_ops);
+	if (err)
+		return err;
+
 	for (i = 0; i < ds->num_ports; i++) {
 		if (!ds->ports[i].dn)
 			continue;
-- 
2.14.0

Re: [PATCH net-next 03/11] net: dsa: debugfs: add tree

[Andrew Lunn]

On Mon, Aug 14, 2017 at 06:22:34PM -0400, Vivien Didelot wrote:
> This commit adds the boiler plate to create a DSA related debug
> filesystem entry as well as a "tree" file, containing the tree index.
> 
>     # cat switch1/tree
>     0
> 
> Signed-off-by: Vivien Didelot <vivien.didelot@savoirfairelinux.com>

Reviewed-by: Andrew Lunn <andrew@lunn.ch>

    Andrew

Re: [PATCH net-next 03/11] net: dsa: debugfs: add tree

[Florian Fainelli]

On 08/14/2017 03:22 PM, Vivien Didelot wrote:
> This commit adds the boiler plate to create a DSA related debug
> filesystem entry as well as a "tree" file, containing the tree index.
> 
>     # cat switch1/tree
>     0
> 
> Signed-off-by: Vivien Didelot <vivien.didelot@savoirfairelinux.com>

Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
-- 
Florian

[PATCH net-next 04/11] net: dsa: debugfs: add tag_protocol

[Vivien Didelot]

Add a debug filesystem "tag_protocol" entry to query the switch tagging
protocol through the .get_tag_protocol operation.

    # cat switch1/tag_protocol
    EDSA

Signed-off-by: Vivien Didelot <vivien.didelot@savoirfairelinux.com>
---
 net/dsa/debugfs.c | 54 ++++++++++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 54 insertions(+)

diff --git a/net/dsa/debugfs.c b/net/dsa/debugfs.c
index 5607efdb924d..30a732e86161 100644
--- a/net/dsa/debugfs.c
+++ b/net/dsa/debugfs.c
@@ -109,6 +109,55 @@ static int dsa_debugfs_create_file(struct dsa_switch *ds, struct dentry *dir,
 	return 0;
 }
 
+static int dsa_debugfs_tag_protocol_read(struct dsa_switch *ds, int id,
+					 struct seq_file *seq)
+{
+	enum dsa_tag_protocol proto;
+
+	if (!ds->ops->get_tag_protocol)
+		return -EOPNOTSUPP;
+
+	proto = ds->ops->get_tag_protocol(ds);
+
+	switch (proto) {
+	case DSA_TAG_PROTO_NONE:
+		seq_puts(seq, "NONE\n");
+		break;
+	case DSA_TAG_PROTO_BRCM:
+		seq_puts(seq, "BRCM\n");
+		break;
+	case DSA_TAG_PROTO_DSA:
+		seq_puts(seq, "DSA\n");
+		break;
+	case DSA_TAG_PROTO_EDSA:
+		seq_puts(seq, "EDSA\n");
+		break;
+	case DSA_TAG_PROTO_KSZ:
+		seq_puts(seq, "KSZ\n");
+		break;
+	case DSA_TAG_PROTO_LAN9303:
+		seq_puts(seq, "LAN9303\n");
+		break;
+	case DSA_TAG_PROTO_MTK:
+		seq_puts(seq, "MTK\n");
+		break;
+	case DSA_TAG_PROTO_QCA:
+		seq_puts(seq, "QCA\n");
+		break;
+	case DSA_TAG_PROTO_TRAILER:
+		seq_puts(seq, "TRAILER\n");
+		break;
+	default:
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+static const struct dsa_debugfs_ops dsa_debugfs_tag_protocol_ops = {
+	.read = dsa_debugfs_tag_protocol_read,
+};
+
 static int dsa_debugfs_tree_read(struct dsa_switch *ds, int id,
 				 struct seq_file *seq)
 {
@@ -151,6 +200,11 @@ static int dsa_debugfs_create_switch(struct dsa_switch *ds)
 	if (IS_ERR_OR_NULL(ds->debugfs_dir))
 		return -EFAULT;
 
+	err = dsa_debugfs_create_file(ds, ds->debugfs_dir, "tag_protocol", -1,
+				      &dsa_debugfs_tag_protocol_ops);
+	if (err)
+		return err;
+
 	err = dsa_debugfs_create_file(ds, ds->debugfs_dir, "tree", -1,
 				      &dsa_debugfs_tree_ops);
 	if (err)
-- 
2.14.0

Re: [PATCH net-next 04/11] net: dsa: debugfs: add tag_protocol

[Andrew Lunn]

On Mon, Aug 14, 2017 at 06:22:35PM -0400, Vivien Didelot wrote:
> Add a debug filesystem "tag_protocol" entry to query the switch tagging
> protocol through the .get_tag_protocol operation.
> 
>     # cat switch1/tag_protocol
>     EDSA
> 
> Signed-off-by: Vivien Didelot <vivien.didelot@savoirfairelinux.com>
> ---
>  net/dsa/debugfs.c | 54 ++++++++++++++++++++++++++++++++++++++++++++++++++++++
>  1 file changed, 54 insertions(+)
> 
> diff --git a/net/dsa/debugfs.c b/net/dsa/debugfs.c
> index 5607efdb924d..30a732e86161 100644
> --- a/net/dsa/debugfs.c
> +++ b/net/dsa/debugfs.c
> @@ -109,6 +109,55 @@ static int dsa_debugfs_create_file(struct dsa_switch *ds, struct dentry *dir,
>  	return 0;
>  }
>  
> +static int dsa_debugfs_tag_protocol_read(struct dsa_switch *ds, int id,
> +					 struct seq_file *seq)
> +{
> +	enum dsa_tag_protocol proto;
> +
> +	if (!ds->ops->get_tag_protocol)
> +		return -EOPNOTSUPP;
> +
> +	proto = ds->ops->get_tag_protocol(ds);
> +
> +	switch (proto) {
> +	case DSA_TAG_PROTO_NONE:
> +		seq_puts(seq, "NONE\n");
> +		break;
> +	case DSA_TAG_PROTO_BRCM:
> +		seq_puts(seq, "BRCM\n");
> +		break;
> +	case DSA_TAG_PROTO_DSA:
> +		seq_puts(seq, "DSA\n");
> +		break;
> +	case DSA_TAG_PROTO_EDSA:
> +		seq_puts(seq, "EDSA\n");
> +		break;
> +	case DSA_TAG_PROTO_KSZ:
> +		seq_puts(seq, "KSZ\n");
> +		break;
> +	case DSA_TAG_PROTO_LAN9303:
> +		seq_puts(seq, "LAN9303\n");
> +		break;
> +	case DSA_TAG_PROTO_MTK:
> +		seq_puts(seq, "MTK\n");
> +		break;
> +	case DSA_TAG_PROTO_QCA:
> +		seq_puts(seq, "QCA\n");
> +		break;
> +	case DSA_TAG_PROTO_TRAILER:
> +		seq_puts(seq, "TRAILER\n");
> +		break;
> +	default:
> +		return -EINVAL;
> +	}

Hi Vivien

Minor nitpick. Rather than -EINVAL, how about seq_puts(seq, "Unknown - Please fix %s\n", __func__);

Reviewed-by: Andrew Lunn <andrew@lunn.ch>

    Andrew

Re: [PATCH net-next 04/11] net: dsa: debugfs: add tag_protocol

[Florian Fainelli]

On 08/14/2017 03:22 PM, Vivien Didelot wrote:
> Add a debug filesystem "tag_protocol" entry to query the switch tagging
> protocol through the .get_tag_protocol operation.
> 
>     # cat switch1/tag_protocol
>     EDSA
> 
> Signed-off-by: Vivien Didelot <vivien.didelot@savoirfairelinux.com>
> ---
>  net/dsa/debugfs.c | 54 ++++++++++++++++++++++++++++++++++++++++++++++++++++++
>  1 file changed, 54 insertions(+)
> 
> diff --git a/net/dsa/debugfs.c b/net/dsa/debugfs.c
> index 5607efdb924d..30a732e86161 100644
> --- a/net/dsa/debugfs.c
> +++ b/net/dsa/debugfs.c
> @@ -109,6 +109,55 @@ static int dsa_debugfs_create_file(struct dsa_switch *ds, struct dentry *dir,
>  	return 0;
>  }
>  
> +static int dsa_debugfs_tag_protocol_read(struct dsa_switch *ds, int id,
> +					 struct seq_file *seq)
> +{
> +	enum dsa_tag_protocol proto;
> +
> +	if (!ds->ops->get_tag_protocol)
> +		return -EOPNOTSUPP;
> +
> +	proto = ds->ops->get_tag_protocol(ds);
> +
> +	switch (proto) {
> +	case DSA_TAG_PROTO_NONE:
> +		seq_puts(seq, "NONE\n");

Might be worth adding a helper function: does dsa_tag_protocol_to_str()
which is in include/net/dsa.h so it's easy to keep in sync when new
taggers are added, then you can just do:

	seq_puts(seq, dsa_tag_protocol_to_str(proto));

and/or use a temporary buffer for adding the trailing newline.
-- 
Florian

[PATCH net-next 05/11] net: dsa: debugfs: add port stats

[Vivien Didelot]

Add a debug filesystem "stats" entry to query a port's hardware
statistics through the DSA switch .get_sset_count, .get_strings and
.get_ethtool_stats operations.

This allows one to get statistics about DSA links interconnecting
switches, which is very convenient because this kind of port is not
exposed to userspace.

Here are the stats of a zii-rev-b DSA and CPU ports:

    # pr -mt switch0/port{5,6}/stats
    in_good_octets      : 0             in_good_octets      : 13824
    in_bad_octets       : 0             in_bad_octets       : 0
    in_unicast          : 0             in_unicast          : 0
    in_broadcasts       : 0             in_broadcasts       : 216
    in_multicasts       : 0             in_multicasts       : 0
    in_pause            : 0             in_pause            : 0
    in_undersize        : 0             in_undersize        : 0
    in_fragments        : 0             in_fragments        : 0
    in_oversize         : 0             in_oversize         : 0
    in_jabber           : 0             in_jabber           : 0
    in_rx_error         : 0             in_rx_error         : 0
    in_fcs_error        : 0             in_fcs_error        : 0
    out_octets          : 9216          out_octets          : 0
    out_unicast         : 0             out_unicast         : 0
    out_broadcasts      : 144           out_broadcasts      : 0
    out_multicasts      : 0             out_multicasts      : 0
    out_pause           : 0             out_pause           : 0
    excessive           : 0             excessive           : 0
    collisions          : 0             collisions          : 0
    deferred            : 0             deferred            : 0
    single              : 0             single              : 0
    multiple            : 0             multiple            : 0
    out_fcs_error       : 0             out_fcs_error       : 0
    late                : 0             late                : 0
    hist_64bytes        : 0             hist_64bytes        : 0
    hist_65_127bytes    : 0             hist_65_127bytes    : 0
    hist_128_255bytes   : 0             hist_128_255bytes   : 0
    hist_256_511bytes   : 0             hist_256_511bytes   : 0
    hist_512_1023bytes  : 0             hist_512_1023bytes  : 0
    hist_1024_max_bytes : 0             hist_1024_max_bytes : 0
    sw_in_discards      : 0             sw_in_discards      : 0
    sw_in_filtered      : 0             sw_in_filtered      : 0
    sw_out_filtered     : 0             sw_out_filtered     : 216

Signed-off-by: Vivien Didelot <vivien.didelot@savoirfairelinux.com>
---
 net/dsa/debugfs.c | 43 +++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 43 insertions(+)

diff --git a/net/dsa/debugfs.c b/net/dsa/debugfs.c
index 30a732e86161..5f91b4423404 100644
--- a/net/dsa/debugfs.c
+++ b/net/dsa/debugfs.c
@@ -109,6 +109,43 @@ static int dsa_debugfs_create_file(struct dsa_switch *ds, struct dentry *dir,
 	return 0;
 }
 
+static void dsa_debugfs_stats_read_count(struct dsa_switch *ds, int id,
+					 struct seq_file *seq, int count)
+{
+	u8 strings[count * ETH_GSTRING_LEN];
+	u64 stats[count];
+	int i;
+
+	ds->ops->get_strings(ds, id, strings);
+	ds->ops->get_ethtool_stats(ds, id, stats);
+
+	for (i = 0; i < count; i++)
+		seq_printf(seq, "%-20s: %lld\n", strings + i * ETH_GSTRING_LEN,
+			   stats[i]);
+}
+
+static int dsa_debugfs_stats_read(struct dsa_switch *ds, int id,
+				  struct seq_file *seq)
+{
+	int count;
+
+	if (!ds->ops->get_sset_count || !ds->ops->get_strings ||
+	    !ds->ops->get_ethtool_stats)
+		return -EOPNOTSUPP;
+
+	count = ds->ops->get_sset_count(ds);
+	if (count < 0)
+		return count;
+
+	dsa_debugfs_stats_read_count(ds, id, seq, count);
+
+	return 0;
+}
+
+static const struct dsa_debugfs_ops dsa_debugfs_stats_ops = {
+	.read = dsa_debugfs_stats_read,
+};
+
 static int dsa_debugfs_tag_protocol_read(struct dsa_switch *ds, int id,
 					 struct seq_file *seq)
 {
@@ -174,6 +211,7 @@ static int dsa_debugfs_create_port(struct dsa_switch *ds, int port)
 {
 	struct dentry *dir;
 	char name[32];
+	int err;
 
 	snprintf(name, sizeof(name), DSA_PORT_FMT, port);
 
@@ -181,6 +219,11 @@ static int dsa_debugfs_create_port(struct dsa_switch *ds, int port)
 	if (IS_ERR_OR_NULL(dir))
 		return -EFAULT;
 
+	err = dsa_debugfs_create_file(ds, dir, "stats", port,
+				      &dsa_debugfs_stats_ops);
+	if (err)
+		return err;
+
 	return 0;
 }
 
-- 
2.14.0

Re: [PATCH net-next 05/11] net: dsa: debugfs: add port stats

[Andrew Lunn]

On Mon, Aug 14, 2017 at 06:22:36PM -0400, Vivien Didelot wrote:
> Add a debug filesystem "stats" entry to query a port's hardware
> statistics through the DSA switch .get_sset_count, .get_strings and
> .get_ethtool_stats operations.
> 
> This allows one to get statistics about DSA links interconnecting
> switches, which is very convenient because this kind of port is not
> exposed to userspace.
> 
> Here are the stats of a zii-rev-b DSA and CPU ports:
> 
>     # pr -mt switch0/port{5,6}/stats
>     in_good_octets      : 0             in_good_octets      : 13824
>     in_bad_octets       : 0             in_bad_octets       : 0
>     in_unicast          : 0             in_unicast          : 0
>     in_broadcasts       : 0             in_broadcasts       : 216
>     in_multicasts       : 0             in_multicasts       : 0
>     in_pause            : 0             in_pause            : 0
>     in_undersize        : 0             in_undersize        : 0
>     in_fragments        : 0             in_fragments        : 0
>     in_oversize         : 0             in_oversize         : 0
>     in_jabber           : 0             in_jabber           : 0
>     in_rx_error         : 0             in_rx_error         : 0
>     in_fcs_error        : 0             in_fcs_error        : 0
>     out_octets          : 9216          out_octets          : 0
>     out_unicast         : 0             out_unicast         : 0
>     out_broadcasts      : 144           out_broadcasts      : 0
>     out_multicasts      : 0             out_multicasts      : 0
>     out_pause           : 0             out_pause           : 0
>     excessive           : 0             excessive           : 0
>     collisions          : 0             collisions          : 0
>     deferred            : 0             deferred            : 0
>     single              : 0             single              : 0
>     multiple            : 0             multiple            : 0
>     out_fcs_error       : 0             out_fcs_error       : 0
>     late                : 0             late                : 0
>     hist_64bytes        : 0             hist_64bytes        : 0
>     hist_65_127bytes    : 0             hist_65_127bytes    : 0
>     hist_128_255bytes   : 0             hist_128_255bytes   : 0
>     hist_256_511bytes   : 0             hist_256_511bytes   : 0
>     hist_512_1023bytes  : 0             hist_512_1023bytes  : 0
>     hist_1024_max_bytes : 0             hist_1024_max_bytes : 0
>     sw_in_discards      : 0             sw_in_discards      : 0
>     sw_in_filtered      : 0             sw_in_filtered      : 0
>     sw_out_filtered     : 0             sw_out_filtered     : 216
> 
> Signed-off-by: Vivien Didelot <vivien.didelot@savoirfairelinux.com>

Reviewed-by: Andrew Lunn <andrew@lunn.ch>

    Andrew

Re: [PATCH net-next 05/11] net: dsa: debugfs: add port stats

[Florian Fainelli]

On 08/14/2017 03:22 PM, Vivien Didelot wrote:
> Add a debug filesystem "stats" entry to query a port's hardware
> statistics through the DSA switch .get_sset_count, .get_strings and
> .get_ethtool_stats operations.
> 
> This allows one to get statistics about DSA links interconnecting
> switches, which is very convenient because this kind of port is not
> exposed to userspace.
> 
> Here are the stats of a zii-rev-b DSA and CPU ports:
> 
>     # pr -mt switch0/port{5,6}/stats
>     in_good_octets      : 0             in_good_octets      : 13824
>     in_bad_octets       : 0             in_bad_octets       : 0
>     in_unicast          : 0             in_unicast          : 0
>     in_broadcasts       : 0             in_broadcasts       : 216
>     in_multicasts       : 0             in_multicasts       : 0
>     in_pause            : 0             in_pause            : 0
>     in_undersize        : 0             in_undersize        : 0
>     in_fragments        : 0             in_fragments        : 0
>     in_oversize         : 0             in_oversize         : 0
>     in_jabber           : 0             in_jabber           : 0
>     in_rx_error         : 0             in_rx_error         : 0
>     in_fcs_error        : 0             in_fcs_error        : 0
>     out_octets          : 9216          out_octets          : 0
>     out_unicast         : 0             out_unicast         : 0
>     out_broadcasts      : 144           out_broadcasts      : 0
>     out_multicasts      : 0             out_multicasts      : 0
>     out_pause           : 0             out_pause           : 0
>     excessive           : 0             excessive           : 0
>     collisions          : 0             collisions          : 0
>     deferred            : 0             deferred            : 0
>     single              : 0             single              : 0
>     multiple            : 0             multiple            : 0
>     out_fcs_error       : 0             out_fcs_error       : 0
>     late                : 0             late                : 0
>     hist_64bytes        : 0             hist_64bytes        : 0
>     hist_65_127bytes    : 0             hist_65_127bytes    : 0
>     hist_128_255bytes   : 0             hist_128_255bytes   : 0
>     hist_256_511bytes   : 0             hist_256_511bytes   : 0
>     hist_512_1023bytes  : 0             hist_512_1023bytes  : 0
>     hist_1024_max_bytes : 0             hist_1024_max_bytes : 0
>     sw_in_discards      : 0             sw_in_discards      : 0
>     sw_in_filtered      : 0             sw_in_filtered      : 0
>     sw_out_filtered     : 0             sw_out_filtered     : 216
> 
> Signed-off-by: Vivien Didelot <vivien.didelot@savoirfairelinux.com>

Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
-- 
Florian

[PATCH net-next 06/11] net: dsa: debugfs: add port registers

[Vivien Didelot]

Add a debug filesystem "regs" entry to query a port's hardware registers
through the .get_regs_len and .get_regs_len switch operations.

This is very convenient because it allows one to dump the registers of
DSA links, which are not exposed to userspace.

Here are the registers of a zii-rev-b CPU and DSA ports:

    # pr -mt switch0/port{5,6}/regs
     0: 4e07			     0: 4d04
     1: 403e			     1: 003d
     2: 0000			     2: 0000
     3: 3521			     3: 3521
     4: 0533			     4: 373f
     5: 8000			     5: 0000
     6: 005f			     6: 003f
     7: 002a			     7: 002a
     8: 2080			     8: 2080
     9: 0001			     9: 0001
    10: 0000			    10: 0000
    11: 0020			    11: 0000
    12: 0000			    12: 0000
    13: 0000			    13: 0000
    14: 0000			    14: 0000
    15: 9100			    15: dada
    16: 0000			    16: 0000
    17: 0000			    17: 0000
    18: 0000			    18: 0000
    19: 0000			    19: 00d8
    20: 0000			    20: 0000
    21: 0000			    21: 0000
    22: 0022			    22: 0000
    23: 0000			    23: 0000
    24: 3210			    24: 3210
    25: 7654			    25: 7654
    26: 0000			    26: 0000
    27: 8000			    27: 8000
    28: 0000			    28: 0000
    29: 0000			    29: 0000
    30: 0000			    30: 0000
    31: 0000			    31: 0000

Signed-off-by: Vivien Didelot <vivien.didelot@savoirfairelinux.com>
---
 net/dsa/debugfs.c | 39 +++++++++++++++++++++++++++++++++++++++
 1 file changed, 39 insertions(+)

diff --git a/net/dsa/debugfs.c b/net/dsa/debugfs.c
index 5f91b4423404..012fcf466cc1 100644
--- a/net/dsa/debugfs.c
+++ b/net/dsa/debugfs.c
@@ -109,6 +109,40 @@ static int dsa_debugfs_create_file(struct dsa_switch *ds, struct dentry *dir,
 	return 0;
 }
 
+static void dsa_debugfs_regs_read_count(struct dsa_switch *ds, int id,
+					struct seq_file *seq, int count)
+{
+	u16 data[count * ETH_GSTRING_LEN];
+	struct ethtool_regs regs;
+	int i;
+
+	ds->ops->get_regs(ds, id, &regs, data);
+
+	for (i = 0; i < count / 2; i++)
+		seq_printf(seq, "%2d: %04x\n", i, data[i]);
+}
+
+static int dsa_debugfs_regs_read(struct dsa_switch *ds, int id,
+				 struct seq_file *seq)
+{
+	int count;
+
+	if (!ds->ops->get_regs_len || !ds->ops->get_regs)
+		return -EOPNOTSUPP;
+
+	count = ds->ops->get_regs_len(ds, id);
+	if (count < 0)
+		return count;
+
+	dsa_debugfs_regs_read_count(ds, id, seq, count);
+
+	return 0;
+}
+
+static const struct dsa_debugfs_ops dsa_debugfs_regs_ops = {
+	.read = dsa_debugfs_regs_read,
+};
+
 static void dsa_debugfs_stats_read_count(struct dsa_switch *ds, int id,
 					 struct seq_file *seq, int count)
 {
@@ -219,6 +253,11 @@ static int dsa_debugfs_create_port(struct dsa_switch *ds, int port)
 	if (IS_ERR_OR_NULL(dir))
 		return -EFAULT;
 
+	err = dsa_debugfs_create_file(ds, dir, "regs", port,
+				      &dsa_debugfs_regs_ops);
+	if (err)
+		return err;
+
 	err = dsa_debugfs_create_file(ds, dir, "stats", port,
 				      &dsa_debugfs_stats_ops);
 	if (err)
-- 
2.14.0

Re: [PATCH net-next 06/11] net: dsa: debugfs: add port registers

[Andrew Lunn]

On Mon, Aug 14, 2017 at 06:22:37PM -0400, Vivien Didelot wrote:
> Add a debug filesystem "regs" entry to query a port's hardware registers
> through the .get_regs_len and .get_regs_len switch operations.
> 
> This is very convenient because it allows one to dump the registers of
> DSA links, which are not exposed to userspace.
> 
> Here are the registers of a zii-rev-b CPU and DSA ports:
> 
>     # pr -mt switch0/port{5,6}/regs
>      0: 4e07			     0: 4d04
>      1: 403e			     1: 003d
>      2: 0000			     2: 0000
>      3: 3521			     3: 3521
>      4: 0533			     4: 373f
>      5: 8000			     5: 0000
>      6: 005f			     6: 003f
>      7: 002a			     7: 002a
>      8: 2080			     8: 2080
>      9: 0001			     9: 0001
>     10: 0000			    10: 0000
>     11: 0020			    11: 0000
>     12: 0000			    12: 0000
>     13: 0000			    13: 0000
>     14: 0000			    14: 0000
>     15: 9100			    15: dada
>     16: 0000			    16: 0000
>     17: 0000			    17: 0000
>     18: 0000			    18: 0000
>     19: 0000			    19: 00d8
>     20: 0000			    20: 0000
>     21: 0000			    21: 0000
>     22: 0022			    22: 0000
>     23: 0000			    23: 0000
>     24: 3210			    24: 3210
>     25: 7654			    25: 7654
>     26: 0000			    26: 0000
>     27: 8000			    27: 8000
>     28: 0000			    28: 0000
>     29: 0000			    29: 0000
>     30: 0000			    30: 0000
>     31: 0000			    31: 0000
> 
> Signed-off-by: Vivien Didelot <vivien.didelot@savoirfairelinux.com>

Reviewed-by: Andrew Lunn <andrew@lunn.ch>

    Andrew

RE: [PATCH net-next 06/11] net: dsa: debugfs: add port registers

[Woojung.Huh]

Vivien,

> Subject: [PATCH net-next 06/11] net: dsa: debugfs: add port registers
> 
> Add a debug filesystem "regs" entry to query a port's hardware registers
> through the .get_regs_len and .get_regs_len switch operations.
> 
> This is very convenient because it allows one to dump the registers of
> DSA links, which are not exposed to userspace.
This series will be very useful to get various debug information.
Do you have any plan to expand 32bit register access and switch-level registers
In addition to port-level registers?

> +static void dsa_debugfs_regs_read_count(struct dsa_switch *ds, int id,
> +					struct seq_file *seq, int count)
> +{
> +	u16 data[count * ETH_GSTRING_LEN];
I think this should be u16 data[count] because it is value of registers.

> +	struct ethtool_regs regs;
> +	int i;
> +
> +	ds->ops->get_regs(ds, id, &regs, data);
> +
> +	for (i = 0; i < count / 2; i++)
> +		seq_printf(seq, "%2d: %04x\n", i, data[i]);
> +}
> +
> +static int dsa_debugfs_regs_read(struct dsa_switch *ds, int id,
> +				 struct seq_file *seq)
> +{
> +	int count;
> +
> +	if (!ds->ops->get_regs_len || !ds->ops->get_regs)
> +		return -EOPNOTSUPP;
> +
> +	count = ds->ops->get_regs_len(ds, id);
> +	if (count < 0)
> +		return count;
Because get_regs_len returns length than count per mv88e6xxx/chip.c,
it requires some math before passing to dsa_debugfs_regs_read_count().

It does "count/2" in dsa-debugfs_regs_read_count(), however,
As commented above, "count" is used at "u16 data[...]".
So, it would be nice to match with name. Ie, count or size/length.

Thanks.
Woojung

Re: [PATCH net-next 06/11] net: dsa: debugfs: add port registers

[Florian Fainelli]

On 08/14/2017 03:22 PM, Vivien Didelot wrote:
> Add a debug filesystem "regs" entry to query a port's hardware registers
> through the .get_regs_len and .get_regs_len switch operations.
> 
> This is very convenient because it allows one to dump the registers of
> DSA links, which are not exposed to userspace.
> 
> Here are the registers of a zii-rev-b CPU and DSA ports:
> 
>     # pr -mt switch0/port{5,6}/regs
>      0: 4e07			     0: 4d04
>      1: 403e			     1: 003d
>      2: 0000			     2: 0000
>      3: 3521			     3: 3521
>      4: 0533			     4: 373f
>      5: 8000			     5: 0000
>      6: 005f			     6: 003f
>      7: 002a			     7: 002a
>      8: 2080			     8: 2080
>      9: 0001			     9: 0001
>     10: 0000			    10: 0000
>     11: 0020			    11: 0000
>     12: 0000			    12: 0000
>     13: 0000			    13: 0000
>     14: 0000			    14: 0000
>     15: 9100			    15: dada
>     16: 0000			    16: 0000
>     17: 0000			    17: 0000
>     18: 0000			    18: 0000
>     19: 0000			    19: 00d8
>     20: 0000			    20: 0000
>     21: 0000			    21: 0000
>     22: 0022			    22: 0000
>     23: 0000			    23: 0000
>     24: 3210			    24: 3210
>     25: 7654			    25: 7654
>     26: 0000			    26: 0000
>     27: 8000			    27: 8000
>     28: 0000			    28: 0000
>     29: 0000			    29: 0000
>     30: 0000			    30: 0000
>     31: 0000			    31: 0000
> 
> Signed-off-by: Vivien Didelot <vivien.didelot@savoirfairelinux.com>


Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
-- 
Florian

[PATCH net-next 07/11] net: dsa: debugfs: add port fdb

[Vivien Didelot]

Add a debug filesystem "fdb" entry to query a port's hardware FDB
entries through the .port_fdb_dump switch operation.

This is really convenient to query directly the hardware or inspect DSA
or CPU links, since these ports are not exposed to userspace.

    # cat port1/fdb
    vid 0    12:34:56:78:90:ab    static    unicast

Signed-off-by: Vivien Didelot <vivien.didelot@savoirfairelinux.com>
---
 net/dsa/debugfs.c | 36 ++++++++++++++++++++++++++++++++++++
 1 file changed, 36 insertions(+)

diff --git a/net/dsa/debugfs.c b/net/dsa/debugfs.c
index 012fcf466cc1..8204c62dc9c1 100644
--- a/net/dsa/debugfs.c
+++ b/net/dsa/debugfs.c
@@ -10,6 +10,7 @@
  */
 
 #include <linux/debugfs.h>
+#include <linux/etherdevice.h>
 #include <linux/seq_file.h>
 
 #include "dsa_priv.h"
@@ -109,6 +110,36 @@ static int dsa_debugfs_create_file(struct dsa_switch *ds, struct dentry *dir,
 	return 0;
 }
 
+static int dsa_debugfs_fdb_dump_cb(const unsigned char *addr, u16 vid,
+				   bool is_static, void *data)
+{
+	struct seq_file *seq = data;
+	int i;
+
+	seq_printf(seq, "vid %d", vid);
+	for (i = 0; i < ETH_ALEN; i++)
+		seq_printf(seq, "%s%02x", i ? ":" : "    ", addr[i]);
+	seq_printf(seq, "    %s", is_static ? "static" : "dynamic");
+	seq_printf(seq, "    %s", is_unicast_ether_addr(addr) ?
+		   "unicast" : "multicast");
+	seq_puts(seq, "\n");
+
+	return 0;
+}
+
+static int dsa_debugfs_fdb_read(struct dsa_switch *ds, int id,
+				struct seq_file *seq)
+{
+	if (!ds->ops->port_fdb_dump)
+		return -EOPNOTSUPP;
+
+	return ds->ops->port_fdb_dump(ds, id, dsa_debugfs_fdb_dump_cb, seq);
+}
+
+static const struct dsa_debugfs_ops dsa_debugfs_fdb_ops = {
+	.read = dsa_debugfs_fdb_read,
+};
+
 static void dsa_debugfs_regs_read_count(struct dsa_switch *ds, int id,
 					struct seq_file *seq, int count)
 {
@@ -253,6 +284,11 @@ static int dsa_debugfs_create_port(struct dsa_switch *ds, int port)
 	if (IS_ERR_OR_NULL(dir))
 		return -EFAULT;
 
+	err = dsa_debugfs_create_file(ds, dir, "fdb", port,
+				      &dsa_debugfs_fdb_ops);
+	if (err)
+		return err;
+
 	err = dsa_debugfs_create_file(ds, dir, "regs", port,
 				      &dsa_debugfs_regs_ops);
 	if (err)
-- 
2.14.0

Re: [PATCH net-next 07/11] net: dsa: debugfs: add port fdb

[Andrew Lunn]

On Mon, Aug 14, 2017 at 06:22:38PM -0400, Vivien Didelot wrote:
> Add a debug filesystem "fdb" entry to query a port's hardware FDB
> entries through the .port_fdb_dump switch operation.
> 
> This is really convenient to query directly the hardware or inspect DSA
> or CPU links, since these ports are not exposed to userspace.
> 
>     # cat port1/fdb
>     vid 0    12:34:56:78:90:ab    static    unicast
> 
> Signed-off-by: Vivien Didelot <vivien.didelot@savoirfairelinux.com>

Reviewed-by: Andrew Lunn <andrew@lunn.ch>

    Andrew

Re: [PATCH net-next 07/11] net: dsa: debugfs: add port fdb

[Florian Fainelli]

On 08/14/2017 03:22 PM, Vivien Didelot wrote:
> Add a debug filesystem "fdb" entry to query a port's hardware FDB
> entries through the .port_fdb_dump switch operation.
> 
> This is really convenient to query directly the hardware or inspect DSA
> or CPU links, since these ports are not exposed to userspace.
> 
>     # cat port1/fdb
>     vid 0    12:34:56:78:90:ab    static    unicast
> 
> Signed-off-by: Vivien Didelot <vivien.didelot@savoirfairelinux.com>
> ---
>  net/dsa/debugfs.c | 36 ++++++++++++++++++++++++++++++++++++
>  1 file changed, 36 insertions(+)
> 
> diff --git a/net/dsa/debugfs.c b/net/dsa/debugfs.c
> index 012fcf466cc1..8204c62dc9c1 100644
> --- a/net/dsa/debugfs.c
> +++ b/net/dsa/debugfs.c
> @@ -10,6 +10,7 @@
>   */
>  
>  #include <linux/debugfs.h>
> +#include <linux/etherdevice.h>
>  #include <linux/seq_file.h>
>  
>  #include "dsa_priv.h"
> @@ -109,6 +110,36 @@ static int dsa_debugfs_create_file(struct dsa_switch *ds, struct dentry *dir,
>  	return 0;
>  }
>  
> +static int dsa_debugfs_fdb_dump_cb(const unsigned char *addr, u16 vid,
> +				   bool is_static, void *data)
> +{
> +	struct seq_file *seq = data;
> +	int i;

unsigned int i?

> +
> +	seq_printf(seq, "vid %d", vid);
> +	for (i = 0; i < ETH_ALEN; i++)
> +		seq_printf(seq, "%s%02x", i ? ":" : "    ", addr[i]);

Too bad we can use %pM here, or can we?

Other than that:

Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
-- 
Florian

[PATCH net-next 08/11] net: dsa: restore mdb dump

[Vivien Didelot]

The same dsa_fdb_dump_cb_t callback is used since there is no
distinction to do between FDB and MDB entries at this layer.

Implement mv88e6xxx_port_mdb_dump so that multicast addresses associated
to a switch port can be dumped.

Signed-off-by: Vivien Didelot <vivien.didelot@savoirfairelinux.com>
---
 drivers/net/dsa/mv88e6xxx/chip.c | 33 +++++++++++++++++++++++++--------
 include/net/dsa.h                |  3 +++
 2 files changed, 28 insertions(+), 8 deletions(-)

diff --git a/drivers/net/dsa/mv88e6xxx/chip.c b/drivers/net/dsa/mv88e6xxx/chip.c
index 918d8f0fe091..2ad32734b6f6 100644
--- a/drivers/net/dsa/mv88e6xxx/chip.c
+++ b/drivers/net/dsa/mv88e6xxx/chip.c
@@ -1380,7 +1380,7 @@ static int mv88e6xxx_port_fdb_del(struct dsa_switch *ds, int port,
 }
 
 static int mv88e6xxx_port_db_dump_fid(struct mv88e6xxx_chip *chip,
-				      u16 fid, u16 vid, int port,
+				      u16 fid, u16 vid, int port, bool mc,
 				      dsa_fdb_dump_cb_t *cb, void *data)
 {
 	struct mv88e6xxx_atu_entry addr;
@@ -1401,11 +1401,14 @@ static int mv88e6xxx_port_db_dump_fid(struct mv88e6xxx_chip *chip,
 		if (addr.trunk || (addr.portvec & BIT(port)) == 0)
 			continue;
 
-		if (!is_unicast_ether_addr(addr.mac))
+		if ((is_unicast_ether_addr(addr.mac) && mc) ||
+		    (is_multicast_ether_addr(addr.mac) && !mc))
 			continue;
 
-		is_static = (addr.state ==
-			     MV88E6XXX_G1_ATU_DATA_STATE_UC_STATIC);
+		is_static = addr.state == mc ?
+			MV88E6XXX_G1_ATU_DATA_STATE_MC_STATIC :
+			MV88E6XXX_G1_ATU_DATA_STATE_UC_STATIC;
+
 		err = cb(addr.mac, vid, is_static, data);
 		if (err)
 			return err;
@@ -1415,7 +1418,7 @@ static int mv88e6xxx_port_db_dump_fid(struct mv88e6xxx_chip *chip,
 }
 
 static int mv88e6xxx_port_db_dump(struct mv88e6xxx_chip *chip, int port,
-				  dsa_fdb_dump_cb_t *cb, void *data)
+				  bool mc, dsa_fdb_dump_cb_t *cb, void *data)
 {
 	struct mv88e6xxx_vtu_entry vlan = {
 		.vid = chip->info->max_vid,
@@ -1428,7 +1431,7 @@ static int mv88e6xxx_port_db_dump(struct mv88e6xxx_chip *chip, int port,
 	if (err)
 		return err;
 
-	err = mv88e6xxx_port_db_dump_fid(chip, fid, 0, port, cb, data);
+	err = mv88e6xxx_port_db_dump_fid(chip, fid, 0, port, mc, cb, data);
 	if (err)
 		return err;
 
@@ -1442,7 +1445,7 @@ static int mv88e6xxx_port_db_dump(struct mv88e6xxx_chip *chip, int port,
 			break;
 
 		err = mv88e6xxx_port_db_dump_fid(chip, vlan.fid, vlan.vid, port,
-						 cb, data);
+						 mc, cb, data);
 		if (err)
 			return err;
 	} while (vlan.vid < chip->info->max_vid);
@@ -1457,7 +1460,7 @@ static int mv88e6xxx_port_fdb_dump(struct dsa_switch *ds, int port,
 	int err;
 
 	mutex_lock(&chip->reg_lock);
-	err = mv88e6xxx_port_db_dump(chip, port, cb, data);
+	err = mv88e6xxx_port_db_dump(chip, port, false, cb, data);
 	mutex_unlock(&chip->reg_lock);
 
 	return err;
@@ -3777,6 +3780,19 @@ static int mv88e6xxx_port_mdb_del(struct dsa_switch *ds, int port,
 	return err;
 }
 
+static int mv88e6xxx_port_mdb_dump(struct dsa_switch *ds, int port,
+				   dsa_fdb_dump_cb_t *cb, void *data)
+{
+	struct mv88e6xxx_chip *chip = ds->priv;
+	int err;
+
+	mutex_lock(&chip->reg_lock);
+	err = mv88e6xxx_port_db_dump(chip, port, true, cb, data);
+	mutex_unlock(&chip->reg_lock);
+
+	return err;
+}
+
 static const struct dsa_switch_ops mv88e6xxx_switch_ops = {
 	.probe			= mv88e6xxx_drv_probe,
 	.get_tag_protocol	= mv88e6xxx_get_tag_protocol,
@@ -3810,6 +3826,7 @@ static const struct dsa_switch_ops mv88e6xxx_switch_ops = {
 	.port_mdb_prepare       = mv88e6xxx_port_mdb_prepare,
 	.port_mdb_add           = mv88e6xxx_port_mdb_add,
 	.port_mdb_del           = mv88e6xxx_port_mdb_del,
+	.port_mdb_dump          = mv88e6xxx_port_mdb_dump,
 	.crosschip_bridge_join	= mv88e6xxx_crosschip_bridge_join,
 	.crosschip_bridge_leave	= mv88e6xxx_crosschip_bridge_leave,
 };
diff --git a/include/net/dsa.h b/include/net/dsa.h
index 4ef5d38755d9..1fd419031948 100644
--- a/include/net/dsa.h
+++ b/include/net/dsa.h
@@ -288,6 +288,7 @@ static inline u8 dsa_upstream_port(struct dsa_switch *ds)
 		return ds->rtable[dst->cpu_dp->ds->index];
 }
 
+/* FDB (and MDB) dump callback */
 typedef int dsa_fdb_dump_cb_t(const unsigned char *addr, u16 vid,
 			      bool is_static, void *data);
 struct dsa_switch_ops {
@@ -417,6 +418,8 @@ struct dsa_switch_ops {
 				struct switchdev_trans *trans);
 	int	(*port_mdb_del)(struct dsa_switch *ds, int port,
 				const struct switchdev_obj_port_mdb *mdb);
+	int	(*port_mdb_dump)(struct dsa_switch *ds, int port,
+				 dsa_fdb_dump_cb_t *cb, void *data);
 	/*
 	 * RXNFC
 	 */
-- 
2.14.0

Re: [PATCH net-next 08/11] net: dsa: restore mdb dump

[Andrew Lunn]

On Mon, Aug 14, 2017 at 06:22:39PM -0400, Vivien Didelot wrote:
> The same dsa_fdb_dump_cb_t callback is used since there is no
> distinction to do between FDB and MDB entries at this layer.
> 
> Implement mv88e6xxx_port_mdb_dump so that multicast addresses associated
> to a switch port can be dumped.
> 
> Signed-off-by: Vivien Didelot <vivien.didelot@savoirfairelinux.com>

Reviewed-by: Andrew Lunn <andrew@lunn.ch>

    Andrew

Re: [PATCH net-next 08/11] net: dsa: restore mdb dump

[Florian Fainelli]

On 08/14/2017 03:22 PM, Vivien Didelot wrote:
> The same dsa_fdb_dump_cb_t callback is used since there is no
> distinction to do between FDB and MDB entries at this layer.
> 
> Implement mv88e6xxx_port_mdb_dump so that multicast addresses associated
> to a switch port can be dumped.
> 
> Signed-off-by: Vivien Didelot <vivien.didelot@savoirfairelinux.com>
> ---
>  drivers/net/dsa/mv88e6xxx/chip.c | 33 +++++++++++++++++++++++++--------
>  include/net/dsa.h                |  3 +++
>  2 files changed, 28 insertions(+), 8 deletions(-)
> 
> diff --git a/drivers/net/dsa/mv88e6xxx/chip.c b/drivers/net/dsa/mv88e6xxx/chip.c
> index 918d8f0fe091..2ad32734b6f6 100644
> --- a/drivers/net/dsa/mv88e6xxx/chip.c
> +++ b/drivers/net/dsa/mv88e6xxx/chip.c
> @@ -1380,7 +1380,7 @@ static int mv88e6xxx_port_fdb_del(struct dsa_switch *ds, int port,
>  }
>  
>  static int mv88e6xxx_port_db_dump_fid(struct mv88e6xxx_chip *chip,
> -				      u16 fid, u16 vid, int port,
> +				      u16 fid, u16 vid, int port, bool mc,
>  				      dsa_fdb_dump_cb_t *cb, void *data)
>  {
>  	struct mv88e6xxx_atu_entry addr;
> @@ -1401,11 +1401,14 @@ static int mv88e6xxx_port_db_dump_fid(struct mv88e6xxx_chip *chip,
>  		if (addr.trunk || (addr.portvec & BIT(port)) == 0)
>  			continue;
>  
> -		if (!is_unicast_ether_addr(addr.mac))
> +		if ((is_unicast_ether_addr(addr.mac) && mc) ||

Should not this remain:

		if (!is_unicast_ether_addr(addr.mac) && mc) ||
		    (is_multicast_ether_addr(addr.mac) && !mc)

here, being:

- if this is an *not* an unicast address and MC dump is requested or
- if this is a multicast address and MC dump is not requested


-- 
Florian

[PATCH net-next 09/11] net: dsa: debugfs: add port mdb

[Vivien Didelot]

Add a debug filesystem "mdb" entry to query a port's hardware MDB
entries through the .port_mdb_dump switch operation.

This is really convenient to query directly the hardware or inspect DSA
or CPU links, since these ports are not exposed to userspace.

Signed-off-by: Vivien Didelot <vivien.didelot@savoirfairelinux.com>
---
 net/dsa/debugfs.c | 19 +++++++++++++++++++
 1 file changed, 19 insertions(+)

diff --git a/net/dsa/debugfs.c b/net/dsa/debugfs.c
index 8204c62dc9c1..98c5068d20da 100644
--- a/net/dsa/debugfs.c
+++ b/net/dsa/debugfs.c
@@ -140,6 +140,20 @@ static const struct dsa_debugfs_ops dsa_debugfs_fdb_ops = {
 	.read = dsa_debugfs_fdb_read,
 };
 
+static int dsa_debugfs_mdb_read(struct dsa_switch *ds, int id,
+				struct seq_file *seq)
+{
+	if (!ds->ops->port_mdb_dump)
+		return -EOPNOTSUPP;
+
+	/* same callback as for FDB dump */
+	return ds->ops->port_mdb_dump(ds, id, dsa_debugfs_fdb_dump_cb, seq);
+}
+
+static const struct dsa_debugfs_ops dsa_debugfs_mdb_ops = {
+	.read = dsa_debugfs_mdb_read,
+};
+
 static void dsa_debugfs_regs_read_count(struct dsa_switch *ds, int id,
 					struct seq_file *seq, int count)
 {
@@ -289,6 +303,11 @@ static int dsa_debugfs_create_port(struct dsa_switch *ds, int port)
 	if (err)
 		return err;
 
+	err = dsa_debugfs_create_file(ds, dir, "mdb", port,
+				      &dsa_debugfs_mdb_ops);
+	if (err)
+		return err;
+
 	err = dsa_debugfs_create_file(ds, dir, "regs", port,
 				      &dsa_debugfs_regs_ops);
 	if (err)
-- 
2.14.0

Re: [PATCH net-next 09/11] net: dsa: debugfs: add port mdb

[Andrew Lunn]

On Mon, Aug 14, 2017 at 06:22:40PM -0400, Vivien Didelot wrote:
> Add a debug filesystem "mdb" entry to query a port's hardware MDB
> entries through the .port_mdb_dump switch operation.
> 
> This is really convenient to query directly the hardware or inspect DSA
> or CPU links, since these ports are not exposed to userspace.
> 
> Signed-off-by: Vivien Didelot <vivien.didelot@savoirfairelinux.com>

Reviewed-by: Andrew Lunn <andrew@lunn.ch>

    Andrew

Re: [PATCH net-next 09/11] net: dsa: debugfs: add port mdb

[Florian Fainelli]

On 08/14/2017 03:22 PM, Vivien Didelot wrote:
> Add a debug filesystem "mdb" entry to query a port's hardware MDB
> entries through the .port_mdb_dump switch operation.
> 
> This is really convenient to query directly the hardware or inspect DSA
> or CPU links, since these ports are not exposed to userspace.
> 
> Signed-off-by: Vivien Didelot <vivien.didelot@savoirfairelinux.com>

Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
-- 
Florian

[PATCH net-next 10/11] net: dsa: restore VLAN dump

[Vivien Didelot]

This commit defines a dsa_vlan_dump_cb_t callback, similar to the FDB
dump callback and partly reverts commit a0b6b8c9fa3c ("net: dsa: Remove
support for vlan dump from DSA's drivers") to restore the DSA drivers
VLAN dump operations.

Signed-off-by: Vivien Didelot <vivien.didelot@savoirfairelinux.com>
---
 drivers/net/dsa/b53/b53_common.c       | 41 ++++++++++++++++++++++++++++
 drivers/net/dsa/b53/b53_priv.h         |  2 ++
 drivers/net/dsa/bcm_sf2.c              |  1 +
 drivers/net/dsa/dsa_loop.c             | 38 ++++++++++++++++++++++++++
 drivers/net/dsa/microchip/ksz_common.c | 41 ++++++++++++++++++++++++++++
 drivers/net/dsa/mv88e6xxx/chip.c       | 49 ++++++++++++++++++++++++++++++++++
 include/net/dsa.h                      |  5 ++++
 7 files changed, 177 insertions(+)

diff --git a/drivers/net/dsa/b53/b53_common.c b/drivers/net/dsa/b53/b53_common.c
index 274f3679f33d..be0c5fa8bd9b 100644
--- a/drivers/net/dsa/b53/b53_common.c
+++ b/drivers/net/dsa/b53/b53_common.c
@@ -1053,6 +1053,46 @@ int b53_vlan_del(struct dsa_switch *ds, int port,
 }
 EXPORT_SYMBOL(b53_vlan_del);
 
+int b53_vlan_dump(struct dsa_switch *ds, int port, dsa_vlan_dump_cb_t *cb,
+		  void *data)
+{
+	struct b53_device *dev = ds->priv;
+	u16 vid, vid_start = 0, pvid;
+	struct b53_vlan *vl;
+	bool untagged;
+	int err = 0;
+
+	if (is5325(dev) || is5365(dev))
+		vid_start = 1;
+
+	b53_read16(dev, B53_VLAN_PAGE, B53_VLAN_PORT_DEF_TAG(port), &pvid);
+
+	/* Use our software cache for dumps, since we do not have any HW
+	 * operation returning only the used/valid VLANs
+	 */
+	for (vid = vid_start; vid < dev->num_vlans; vid++) {
+		vl = &dev->vlans[vid];
+
+		if (!vl->valid)
+			continue;
+
+		if (!(vl->members & BIT(port)))
+			continue;
+
+		untagged = false;
+
+		if (vl->untag & BIT(port))
+			untagged = true;
+
+		err = cb(vid, pvid == vid, untagged, data);
+		if (err)
+			break;
+	}
+
+	return err;
+}
+EXPORT_SYMBOL(b53_vlan_dump);
+
 /* Address Resolution Logic routines */
 static int b53_arl_op_wait(struct b53_device *dev)
 {
@@ -1503,6 +1543,7 @@ static const struct dsa_switch_ops b53_switch_ops = {
 	.port_vlan_prepare	= b53_vlan_prepare,
 	.port_vlan_add		= b53_vlan_add,
 	.port_vlan_del		= b53_vlan_del,
+	.port_vlan_dump		= b53_vlan_dump,
 	.port_fdb_dump		= b53_fdb_dump,
 	.port_fdb_add		= b53_fdb_add,
 	.port_fdb_del		= b53_fdb_del,
diff --git a/drivers/net/dsa/b53/b53_priv.h b/drivers/net/dsa/b53/b53_priv.h
index 01bd8cbe9a3f..2b3e59d80fdb 100644
--- a/drivers/net/dsa/b53/b53_priv.h
+++ b/drivers/net/dsa/b53/b53_priv.h
@@ -393,6 +393,8 @@ void b53_vlan_add(struct dsa_switch *ds, int port,
 		  struct switchdev_trans *trans);
 int b53_vlan_del(struct dsa_switch *ds, int port,
 		 const struct switchdev_obj_port_vlan *vlan);
+int b53_vlan_dump(struct dsa_switch *ds, int port, dsa_vlan_dump_cb_t *cb,
+		  void *data);
 int b53_fdb_add(struct dsa_switch *ds, int port,
 		const unsigned char *addr, u16 vid);
 int b53_fdb_del(struct dsa_switch *ds, int port,
diff --git a/drivers/net/dsa/bcm_sf2.c b/drivers/net/dsa/bcm_sf2.c
index bbcb4053e04e..1907b27297c3 100644
--- a/drivers/net/dsa/bcm_sf2.c
+++ b/drivers/net/dsa/bcm_sf2.c
@@ -1021,6 +1021,7 @@ static const struct dsa_switch_ops bcm_sf2_ops = {
 	.port_vlan_prepare	= b53_vlan_prepare,
 	.port_vlan_add		= b53_vlan_add,
 	.port_vlan_del		= b53_vlan_del,
+	.port_vlan_dump		= b53_vlan_dump,
 	.port_fdb_dump		= b53_fdb_dump,
 	.port_fdb_add		= b53_fdb_add,
 	.port_fdb_del		= b53_fdb_del,
diff --git a/drivers/net/dsa/dsa_loop.c b/drivers/net/dsa/dsa_loop.c
index 7819a9fe8321..0407533f725f 100644
--- a/drivers/net/dsa/dsa_loop.c
+++ b/drivers/net/dsa/dsa_loop.c
@@ -257,6 +257,43 @@ static int dsa_loop_port_vlan_del(struct dsa_switch *ds, int port,
 	return 0;
 }
 
+static int dsa_loop_port_vlan_dump(struct dsa_switch *ds, int port,
+				   dsa_vlan_dump_cb_t *cb, void *data)
+{
+	struct dsa_loop_priv *ps = ds->priv;
+	struct mii_bus *bus = ps->bus;
+	struct dsa_loop_vlan *vl;
+	u16 vid, vid_start = 0;
+	bool pvid, untagged;
+	int err = 0;
+
+	dev_dbg(ds->dev, "%s\n", __func__);
+
+	/* Just do a sleeping operation to make lockdep checks effective */
+	mdiobus_read(bus, ps->port_base + port, MII_BMSR);
+
+	for (vid = vid_start; vid < DSA_LOOP_VLANS; vid++) {
+		vl = &ps->vlans[vid];
+
+		if (!(vl->members & BIT(port)))
+			continue;
+
+		untagged = false;
+		pvid = false;
+
+		if (vl->untagged & BIT(port))
+			untagged = true;
+		if (ps->pvid == vid)
+			pvid = true;
+
+		err = cb(vid, pvid, untagged, data);
+		if (err)
+			break;
+	}
+
+	return err;
+}
+
 static const struct dsa_switch_ops dsa_loop_driver = {
 	.get_tag_protocol	= dsa_loop_get_protocol,
 	.setup			= dsa_loop_setup,
@@ -273,6 +310,7 @@ static const struct dsa_switch_ops dsa_loop_driver = {
 	.port_vlan_prepare	= dsa_loop_port_vlan_prepare,
 	.port_vlan_add		= dsa_loop_port_vlan_add,
 	.port_vlan_del		= dsa_loop_port_vlan_del,
+	.port_vlan_dump		= dsa_loop_port_vlan_dump,
 };
 
 static int dsa_loop_drv_probe(struct mdio_device *mdiodev)
diff --git a/drivers/net/dsa/microchip/ksz_common.c b/drivers/net/dsa/microchip/ksz_common.c
index 56cd6d365352..52c7849acc3c 100644
--- a/drivers/net/dsa/microchip/ksz_common.c
+++ b/drivers/net/dsa/microchip/ksz_common.c
@@ -638,6 +638,46 @@ static int ksz_port_vlan_del(struct dsa_switch *ds, int port,
 	return 0;
 }
 
+static int ksz_port_vlan_dump(struct dsa_switch *ds, int port,
+			      dsa_vlan_dump_cb_t *cb, void *data)
+{
+	struct ksz_device *dev = ds->priv;
+	struct vlan_table *vlan_cache;
+	bool pvid, untagged;
+	u16 val;
+	int vid;
+	int err = 0;
+
+	mutex_lock(&dev->vlan_mutex);
+
+	/* use dev->vlan_cache due to lack of searching valid vlan entry */
+	for (vid = 0; vid < dev->num_vlans; vid++) {
+		vlan_cache = &dev->vlan_cache[vid];
+
+		if (!(vlan_cache->table[0] & VLAN_VALID))
+			continue;
+
+		untagged = false;
+		pvid = false;
+
+		if (vlan_cache->table[2] & BIT(port)) {
+			if (vlan_cache->table[1] & BIT(port))
+				untagged = true;
+			ksz_pread16(dev, port, REG_PORT_DEFAULT_VID, &val);
+			if (vid == (val & 0xFFFFF))
+				pvid = true;
+
+			err = cb(vid, pvid, untagged, data);
+			if (err)
+				break;
+		}
+	}
+
+	mutex_unlock(&dev->vlan_mutex);
+
+	return err;
+}
+
 struct alu_struct {
 	/* entry 1 */
 	u8	is_static:1;
@@ -1068,6 +1108,7 @@ static const struct dsa_switch_ops ksz_switch_ops = {
 	.port_vlan_prepare	= ksz_port_vlan_prepare,
 	.port_vlan_add		= ksz_port_vlan_add,
 	.port_vlan_del		= ksz_port_vlan_del,
+	.port_vlan_dump		= ksz_port_vlan_dump,
 	.port_fdb_dump		= ksz_port_fdb_dump,
 	.port_fdb_add		= ksz_port_fdb_add,
 	.port_fdb_del		= ksz_port_fdb_del,
diff --git a/drivers/net/dsa/mv88e6xxx/chip.c b/drivers/net/dsa/mv88e6xxx/chip.c
index 2ad32734b6f6..dc762e85c556 100644
--- a/drivers/net/dsa/mv88e6xxx/chip.c
+++ b/drivers/net/dsa/mv88e6xxx/chip.c
@@ -1011,6 +1011,54 @@ static int mv88e6xxx_vtu_loadpurge(struct mv88e6xxx_chip *chip,
 	return chip->info->ops->vtu_loadpurge(chip, entry);
 }
 
+static int mv88e6xxx_port_vlan_dump(struct dsa_switch *ds, int port,
+				    dsa_vlan_dump_cb_t *cb, void *data)
+{
+	struct mv88e6xxx_chip *chip = ds->priv;
+	struct mv88e6xxx_vtu_entry next = {
+		.vid = chip->info->max_vid,
+	};
+	bool untagged;
+	u16 pvid;
+	int err;
+
+	if (!chip->info->max_vid)
+		return -EOPNOTSUPP;
+
+	mutex_lock(&chip->reg_lock);
+
+	err = mv88e6xxx_port_get_pvid(chip, port, &pvid);
+	if (err)
+		goto unlock;
+
+	do {
+		err = mv88e6xxx_vtu_getnext(chip, &next);
+		if (err)
+			break;
+
+		if (!next.valid)
+			break;
+
+		if (next.member[port] ==
+		    MV88E6XXX_G1_VTU_DATA_MEMBER_TAG_NON_MEMBER)
+			continue;
+
+		untagged = false;
+		if (next.member[port] ==
+		    MV88E6XXX_G1_VTU_DATA_MEMBER_TAG_UNTAGGED)
+			untagged = true;
+
+		err = cb(next.vid, next.vid == pvid, untagged, data);
+		if (err)
+			break;
+	} while (next.vid < chip->info->max_vid);
+
+unlock:
+	mutex_unlock(&chip->reg_lock);
+
+	return err;
+}
+
 static int mv88e6xxx_atu_new(struct mv88e6xxx_chip *chip, u16 *fid)
 {
 	DECLARE_BITMAP(fid_bitmap, MV88E6XXX_N_FID);
@@ -3820,6 +3868,7 @@ static const struct dsa_switch_ops mv88e6xxx_switch_ops = {
 	.port_vlan_prepare	= mv88e6xxx_port_vlan_prepare,
 	.port_vlan_add		= mv88e6xxx_port_vlan_add,
 	.port_vlan_del		= mv88e6xxx_port_vlan_del,
+	.port_vlan_dump		= mv88e6xxx_port_vlan_dump,
 	.port_fdb_add           = mv88e6xxx_port_fdb_add,
 	.port_fdb_del           = mv88e6xxx_port_fdb_del,
 	.port_fdb_dump          = mv88e6xxx_port_fdb_dump,
diff --git a/include/net/dsa.h b/include/net/dsa.h
index 1fd419031948..317e0cd4bf70 100644
--- a/include/net/dsa.h
+++ b/include/net/dsa.h
@@ -291,6 +291,8 @@ static inline u8 dsa_upstream_port(struct dsa_switch *ds)
 /* FDB (and MDB) dump callback */
 typedef int dsa_fdb_dump_cb_t(const unsigned char *addr, u16 vid,
 			      bool is_static, void *data);
+typedef int dsa_vlan_dump_cb_t(u16 vid, bool pvid, bool untagged, void *data);
+
 struct dsa_switch_ops {
 	/*
 	 * Legacy probing.
@@ -397,6 +399,9 @@ struct dsa_switch_ops {
 				 struct switchdev_trans *trans);
 	int	(*port_vlan_del)(struct dsa_switch *ds, int port,
 				 const struct switchdev_obj_port_vlan *vlan);
+	int	(*port_vlan_dump)(struct dsa_switch *ds, int port,
+				  dsa_vlan_dump_cb_t *cb, void *data);
+
 	/*
 	 * Forwarding database
 	 */
-- 
2.14.0

Re: [PATCH net-next 10/11] net: dsa: restore VLAN dump

[Andrew Lunn]

On Mon, Aug 14, 2017 at 06:22:41PM -0400, Vivien Didelot wrote:
> This commit defines a dsa_vlan_dump_cb_t callback, similar to the FDB
> dump callback and partly reverts commit a0b6b8c9fa3c ("net: dsa: Remove
> support for vlan dump from DSA's drivers") to restore the DSA drivers
> VLAN dump operations.
> 
> Signed-off-by: Vivien Didelot <vivien.didelot@savoirfairelinux.com>

Reviewed-by: Andrew Lunn <andrew@lunn.ch>

    Andrew

Re: [PATCH net-next 10/11] net: dsa: restore VLAN dump

[Florian Fainelli]

On 08/14/2017 03:22 PM, Vivien Didelot wrote:
> This commit defines a dsa_vlan_dump_cb_t callback, similar to the FDB
> dump callback and partly reverts commit a0b6b8c9fa3c ("net: dsa: Remove
> support for vlan dump from DSA's drivers") to restore the DSA drivers
> VLAN dump operations.
> 
> Signed-off-by: Vivien Didelot <vivien.didelot@savoirfairelinux.com>

Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
-- 
Florian

[PATCH net-next 11/11] net: dsa: debugfs: add port vlan

[Vivien Didelot]

Add a debug filesystem "vlan" entry to query a port's hardware VLAN
entries through the .port_vlan_dump switch operation.

This is really convenient to query directly the hardware or inspect DSA
or CPU links, since these ports are not exposed to userspace.

Here are the VLAN entries for a CPU port:

    # cat port5/vlan
    vid 1
    vid 42  pvid

Signed-off-by: Vivien Didelot <vivien.didelot@savoirfairelinux.com>
---
 net/dsa/debugfs.c | 33 +++++++++++++++++++++++++++++++++
 1 file changed, 33 insertions(+)

diff --git a/net/dsa/debugfs.c b/net/dsa/debugfs.c
index 98c5068d20da..b00942368d29 100644
--- a/net/dsa/debugfs.c
+++ b/net/dsa/debugfs.c
@@ -286,6 +286,34 @@ static const struct dsa_debugfs_ops dsa_debugfs_tree_ops = {
 	.read = dsa_debugfs_tree_read,
 };
 
+static int dsa_debugfs_vlan_dump_cb(u16 vid, bool pvid, bool untagged,
+				    void *data)
+{
+	struct seq_file *seq = data;
+
+	seq_printf(seq, "vid %d", vid);
+	if (pvid)
+		seq_puts(seq, "  pvid");
+	if (untagged)
+		seq_puts(seq, "  untagged");
+	seq_puts(seq, "\n");
+
+	return 0;
+}
+
+static int dsa_debugfs_vlan_read(struct dsa_switch *ds, int id,
+				 struct seq_file *seq)
+{
+	if (!ds->ops->port_vlan_dump)
+		return -EOPNOTSUPP;
+
+	return ds->ops->port_vlan_dump(ds, id, dsa_debugfs_vlan_dump_cb, seq);
+}
+
+static const struct dsa_debugfs_ops dsa_debugfs_vlan_ops = {
+	.read = dsa_debugfs_vlan_read,
+};
+
 static int dsa_debugfs_create_port(struct dsa_switch *ds, int port)
 {
 	struct dentry *dir;
@@ -318,6 +346,11 @@ static int dsa_debugfs_create_port(struct dsa_switch *ds, int port)
 	if (err)
 		return err;
 
+	err = dsa_debugfs_create_file(ds, dir, "vlan", port,
+				      &dsa_debugfs_vlan_ops);
+	if (err)
+		return err;
+
 	return 0;
 }
 
-- 
2.14.0

Re: [PATCH net-next 11/11] net: dsa: debugfs: add port vlan

[Andrew Lunn]

On Mon, Aug 14, 2017 at 06:22:42PM -0400, Vivien Didelot wrote:
> Add a debug filesystem "vlan" entry to query a port's hardware VLAN
> entries through the .port_vlan_dump switch operation.
> 
> This is really convenient to query directly the hardware or inspect DSA
> or CPU links, since these ports are not exposed to userspace.
> 
> Here are the VLAN entries for a CPU port:
> 
>     # cat port5/vlan
>     vid 1
>     vid 42  pvid
> 
> Signed-off-by: Vivien Didelot <vivien.didelot@savoirfairelinux.com>

Reviewed-by: Andrew Lunn <andrew@lunn.ch>

    Andrew

Re: [PATCH net-next 11/11] net: dsa: debugfs: add port vlan

[Florian Fainelli]

On 08/14/2017 03:22 PM, Vivien Didelot wrote:
> Add a debug filesystem "vlan" entry to query a port's hardware VLAN
> entries through the .port_vlan_dump switch operation.
> 
> This is really convenient to query directly the hardware or inspect DSA
> or CPU links, since these ports are not exposed to userspace.
> 
> Here are the VLAN entries for a CPU port:
> 
>     # cat port5/vlan
>     vid 1
>     vid 42  pvid
> 
> Signed-off-by: Vivien Didelot <vivien.didelot@savoirfairelinux.com>
> ---
>  net/dsa/debugfs.c | 33 +++++++++++++++++++++++++++++++++
>  1 file changed, 33 insertions(+)
> 
> diff --git a/net/dsa/debugfs.c b/net/dsa/debugfs.c
> index 98c5068d20da..b00942368d29 100644
> --- a/net/dsa/debugfs.c
> +++ b/net/dsa/debugfs.c
> @@ -286,6 +286,34 @@ static const struct dsa_debugfs_ops dsa_debugfs_tree_ops = {
>  	.read = dsa_debugfs_tree_read,
>  };
>  
> +static int dsa_debugfs_vlan_dump_cb(u16 vid, bool pvid, bool untagged,
> +				    void *data)
> +{
> +	struct seq_file *seq = data;
> +
> +	seq_printf(seq, "vid %d", vid);
> +	if (pvid)
> +		seq_puts(seq, "  pvid");
> +	if (untagged)
> +		seq_puts(seq, "  untagged");

Personal preference: could we just specify something like:

vid 1 (t)
vid 42 (u) pvid

to clearly show which VLAN is tagged/untagged?

Other than that:

Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
-- 
Florian

Re: [PATCH net-next 00/11] net: dsa: add generic debugfs interface

[Florian Fainelli]

On 08/14/2017 03:22 PM, Vivien Didelot wrote:
> This patch series adds a generic debugfs interface for the DSA
> framework, so that all switch devices benefit from it, e.g. Marvell,
> Broadcom, Microchip or any other DSA driver.
> 
> This is really convenient for debugging, especially CPU ports and DSA
> links which are not exposed to userspace as net device. This interface
> is currently the only way to easily inspect the hardware for such ports.
> 
> With the patch series, any switch device user is able to query the
> hardware for the supported tagging protocol, the ports stats and
> registers, as well as their FDB, MDB and VLAN entries.
> 
> This support is only compiled if CONFIG_DEBUG_FS is enabled. Below is
> and example of usage of this interface on a multi-chip switch fabric:
> 
>     # mount -t debugfs none /sys/kernel/debug
>     # cd /sys/kernel/debug/dsa/
>     # ls
>     switch0  switch1 switch2
>     # ls -l switch0/
>     drwxr-xr-x 2 root root 0 Jan  1 00:00 port0
>     drwxr-xr-x 2 root root 0 Jan  1 00:00 port1
>     drwxr-xr-x 2 root root 0 Jan  1 00:00 port2
>     drwxr-xr-x 2 root root 0 Jan  1 00:00 port5
>     drwxr-xr-x 2 root root 0 Jan  1 00:00 port6
>     -r--r--r-- 1 root root 0 Jan  1 00:00 tag_protocol
>     -r--r--r-- 1 root root 0 Jan  1 00:00 tree
>     # ls -l switch0/port6
>     -r--r--r-- 1 root root 0 Jan  1 00:00 fdb
>     -r--r--r-- 1 root root 0 Jan  1 00:00 mdb
>     -r--r--r-- 1 root root 0 Jan  1 00:00 regs
>     -r--r--r-- 1 root root 0 Jan  1 00:00 stats
>     -r--r--r-- 1 root root 0 Jan  1 00:00 vlan
>     # cat switch0/port2/vlan
>     vid 42  pvid  untagged
>     # cat switch0/port1/fdb
>     vid 0    12:34:56:78:90:ab    static    unicast
>     # pr -mt switch0/port{5,6}/stats
>     in_good_octets      : 0             in_good_octets      : 13824
>     in_bad_octets       : 0             in_bad_octets       : 0
>     in_unicast          : 0             in_unicast          : 0
>     in_broadcasts       : 0             in_broadcasts       : 216
>     in_multicasts       : 0             in_multicasts       : 0
>     in_pause            : 0             in_pause            : 0
>     in_undersize        : 0             in_undersize        : 0
>     ...
>     # pr -mt switch0/port{5,6}/regs
>      0: 4e07			     0: 4d04
>      1: 403e			     1: 003d
>      2: 0000			     2: 0000
>      3: 3521			     3: 3521
>      4: 0533			     4: 373f
>      5: 8000			     5: 0000
>      6: 005f			     6: 003f
>      7: 002a			     7: 002a
>     ...
> 
> where switch0 port5 and port6 are CPU and DSA ports of a ZII Rev B.

For this whole series:

Tested-by: Florian Fainelli <f.fainelli@gmail.com>

On bcm_sf2

> 
> Vivien Didelot (11):
>   net: dsa: legacy: assign dst->applied
>   net: dsa: add debugfs interface
>   net: dsa: debugfs: add tree
>   net: dsa: debugfs: add tag_protocol
>   net: dsa: debugfs: add port stats
>   net: dsa: debugfs: add port registers
>   net: dsa: debugfs: add port fdb
>   net: dsa: restore mdb dump
>   net: dsa: debugfs: add port mdb
>   net: dsa: restore VLAN dump
>   net: dsa: debugfs: add port vlan
> 
>  drivers/net/dsa/b53/b53_common.c       |  41 +++
>  drivers/net/dsa/b53/b53_priv.h         |   2 +
>  drivers/net/dsa/bcm_sf2.c              |   1 +
>  drivers/net/dsa/dsa_loop.c             |  38 +++
>  drivers/net/dsa/microchip/ksz_common.c |  41 +++
>  drivers/net/dsa/mv88e6xxx/chip.c       |  82 +++++-
>  include/net/dsa.h                      |  15 ++
>  net/dsa/Kconfig                        |  14 +
>  net/dsa/Makefile                       |   1 +
>  net/dsa/debugfs.c                      | 453 +++++++++++++++++++++++++++++++++
>  net/dsa/dsa.c                          |   3 +
>  net/dsa/dsa2.c                         |   4 +
>  net/dsa/dsa_priv.h                     |  13 +
>  net/dsa/legacy.c                       |   7 +
>  14 files changed, 707 insertions(+), 8 deletions(-)
>  create mode 100644 net/dsa/debugfs.c
> 


-- 
Florian