This patch series focuses on making vsc73xx usable.
The first patch was added in v2; it switches from a poll loop to
read_poll_timeout.
The second patch is a simple conversion to phylink because adjust_link
won't work anymore.
The third patch is preparation for future use. Using the
"phy_interface_mode_is_rgmii" macro allows for the proper recognition
of all RGMII modes.
Patches 4-5 involve some cleanup: The fourth patch introduces
a definition with the maximum number of ports to avoid using
magic numbers. The next one fills in documentation.
The sixth patch implements port state configuration, which is required
for bridge functionality. STP frames are not forwarded at this moment.
BPDU frames are only forwarded from/to the PI/SI interface.
For more information, see chapter 2.7.1 (CPU Forwarding) in the
datasheet.
Patches 7, 12-15 provide a basic implementation of tag_8021q
functionality with QinQ support, without VLAN filtering in
the bridge and simple VLAN awareness in VLAN filtering mode.
Patches 8-11 came from Vladimir Oltean. They prepare for making
tag8021q more common. VSC73XX uses very similar tag recognition,
and some code from tag_sja1105 could be moved to tag_8021q for
common use.
Patch 16 is required to avoid problem with learning on standalone ports.
Pawel Dembicki (12):
net: dsa: vsc73xx: use read_poll_timeout instead delay loop
net: dsa: vsc73xx: convert to PHYLINK
net: dsa: vsc73xx: use macros for rgmii recognition
net: dsa: vsc73xx: Add define for max num of ports
net: dsa: vsc73xx: add structure descriptions
net: dsa: vsc73xx: add port_stp_state_set function
net: dsa: vsc73xx: Add vlan filtering
net: dsa: vsc73xx: introduce tag 8021q for vsc73xx
net: dsa: vsc73xx: Implement the tag_8021q VLAN operations
net: dsa: Define max num of bridges in tag8021q implementation
net: dsa: vsc73xx: Add bridge support
net: dsa: vsc73xx: start treating the BR_LEARNING flag
Vladimir Oltean (4):
net: dsa: tag_sja1105: absorb logic for not overwriting precise info
into dsa_8021q_rcv()
net: dsa: tag_sja1105: absorb entire sja1105_vlan_rcv() into
dsa_8021q_rcv()
net: dsa: tag_sja1105: prefer precise source port info on SJA1110 too
net: dsa: tag_sja1105: refactor skb->dev assignment to
dsa_tag_8021q_find_user()
drivers/net/dsa/Kconfig | 2 +-
drivers/net/dsa/sja1105/sja1105_main.c | 3 +-
drivers/net/dsa/vitesse-vsc73xx-core.c | 979 +++++++++++++++++++++----
drivers/net/dsa/vitesse-vsc73xx.h | 69 +-
include/linux/dsa/8021q.h | 5 +
include/net/dsa.h | 2 +
net/dsa/Kconfig | 6 +
net/dsa/Makefile | 1 +
net/dsa/tag_8021q.c | 81 +-
net/dsa/tag_8021q.h | 7 +-
net/dsa/tag_ocelot_8021q.c | 2 +-
net/dsa/tag_sja1105.c | 72 +-
net/dsa/tag_vsc73xx_8021q.c | 68 ++
13 files changed, 1074 insertions(+), 223 deletions(-)
create mode 100644 net/dsa/tag_vsc73xx_8021q.c
--
2.34.1
Switch the delay loop during the Arbiter empty check from
vsc73xx_adjust_link() to use read_poll_timeout(). Functionally,
one msleep() call is eliminated at the end of the loop in the timeout
case.
As Russell King suggested:
"This [change] avoids the issue that on the last iteration, the code reads
the register, tests it, finds the condition that's being waiting for is
false, _then_ waits and end up printing the error message - that last
wait is rather useless, and as the arbiter state isn't checked after
waiting, it could be that we had success during the last wait."
Suggested-by: Russell King <[email protected]>
Reviewed-by: Andrew Lunn <[email protected]>
Reviewed-by: Linus Walleij <[email protected]>
Reviewed-by: Florian Fainelli <[email protected]>
Signed-off-by: Pawel Dembicki <[email protected]>
---
v8,v7,v6:
- resend only
v5:
- improved commit description
- added two defines instead magic numbers
- added include with iopoll.h
v4:
- resend patch
v3:
- added "Reviewed-by" to commit message only
v2:
- introduced patch
drivers/net/dsa/vitesse-vsc73xx-core.c | 30 ++++++++++++++------------
1 file changed, 16 insertions(+), 14 deletions(-)
diff --git a/drivers/net/dsa/vitesse-vsc73xx-core.c b/drivers/net/dsa/vitesse-vsc73xx-core.c
index ae70eac3be28..ab5771d4d828 100644
--- a/drivers/net/dsa/vitesse-vsc73xx-core.c
+++ b/drivers/net/dsa/vitesse-vsc73xx-core.c
@@ -17,6 +17,7 @@
#include <linux/kernel.h>
#include <linux/module.h>
#include <linux/device.h>
+#include <linux/iopoll.h>
#include <linux/of.h>
#include <linux/of_mdio.h>
#include <linux/bitops.h>
@@ -268,6 +269,9 @@
#define IS_7398(a) ((a)->chipid == VSC73XX_CHIPID_ID_7398)
#define IS_739X(a) (IS_7395(a) || IS_7398(a))
+#define VSC73XX_POLL_SLEEP_US 1000
+#define VSC73XX_POLL_TIMEOUT_US 10000
+
struct vsc73xx_counter {
u8 counter;
const char *name;
@@ -779,7 +783,7 @@ static void vsc73xx_adjust_link(struct dsa_switch *ds, int port,
* after a PHY or the CPU port comes up or down.
*/
if (!phydev->link) {
- int maxloop = 10;
+ int ret, err;
dev_dbg(vsc->dev, "port %d: went down\n",
port);
@@ -794,19 +798,17 @@ static void vsc73xx_adjust_link(struct dsa_switch *ds, int port,
VSC73XX_ARBDISC, BIT(port), BIT(port));
/* Wait until queue is empty */
- vsc73xx_read(vsc, VSC73XX_BLOCK_ARBITER, 0,
- VSC73XX_ARBEMPTY, &val);
- while (!(val & BIT(port))) {
- msleep(1);
- vsc73xx_read(vsc, VSC73XX_BLOCK_ARBITER, 0,
- VSC73XX_ARBEMPTY, &val);
- if (--maxloop == 0) {
- dev_err(vsc->dev,
- "timeout waiting for block arbiter\n");
- /* Continue anyway */
- break;
- }
- }
+ ret = read_poll_timeout(vsc73xx_read, err,
+ err < 0 || (val & BIT(port)),
+ VSC73XX_POLL_SLEEP_US,
+ VSC73XX_POLL_TIMEOUT_US, false,
+ vsc, VSC73XX_BLOCK_ARBITER, 0,
+ VSC73XX_ARBEMPTY, &val);
+ if (ret)
+ dev_err(vsc->dev,
+ "timeout waiting for block arbiter\n");
+ else if (err < 0)
+ dev_err(vsc->dev, "error reading arbiter\n");
/* Put this port into reset */
vsc73xx_write(vsc, VSC73XX_BLOCK_MAC, port, VSC73XX_MAC_CFG,
--
2.34.1
This patch replaces the adjust_link api with the phylink apis that provide
equivalent functionality.
The remaining functionality from the adjust_link is now covered in the
phylink_mac_link_* and phylink_mac_config.
Removes:
adjust_link
Adds:
phylink_mac_config
phylink_mac_link_up
phylink_mac_link_down
Signed-off-by: Pawel Dembicki <[email protected]>
Reviewed-by: Florian Fainelli <[email protected]>
Reviewed-by: Linus Walleij <[email protected]>
Reviewed-by: Russell King (Oracle) <[email protected]>
---
v8:
- resend only
v7:
- added 'Reviewed-by' to commit message only
v6:
- resend only
v5:
- introduce MAC reset function with procedure described in datasheet
- reorganise routines
- move 'Accept packets again' routine to 'phylink_mac_link_up'
- fix one FIXME
v4:
- update commit descripion
- remove phylink_get_caps after rebase to current net-next/main
v3:
- remove legacy_pre_march2020 after rebase
v2:
- replace switch to if and get rid of macros in
vsc73xx_phylink_mac_link_up function
drivers/net/dsa/vitesse-vsc73xx-core.c | 235 ++++++++++++-------------
1 file changed, 109 insertions(+), 126 deletions(-)
diff --git a/drivers/net/dsa/vitesse-vsc73xx-core.c b/drivers/net/dsa/vitesse-vsc73xx-core.c
index ab5771d4d828..cb2e7e256279 100644
--- a/drivers/net/dsa/vitesse-vsc73xx-core.c
+++ b/drivers/net/dsa/vitesse-vsc73xx-core.c
@@ -717,52 +717,43 @@ static void vsc73xx_init_port(struct vsc73xx *vsc, int port)
port, VSC73XX_C_RX0, 0);
}
-static void vsc73xx_adjust_enable_port(struct vsc73xx *vsc,
- int port, struct phy_device *phydev,
- u32 initval)
+static void vsc73xx_reset_port(struct vsc73xx *vsc, int port, u32 initval)
{
- u32 val = initval;
- u8 seed;
-
- /* Reset this port FIXME: break out subroutine */
- val |= VSC73XX_MAC_CFG_RESET;
- vsc73xx_write(vsc, VSC73XX_BLOCK_MAC, port, VSC73XX_MAC_CFG, val);
-
- /* Seed the port randomness with randomness */
- get_random_bytes(&seed, 1);
- val |= seed << VSC73XX_MAC_CFG_SEED_OFFSET;
- val |= VSC73XX_MAC_CFG_SEED_LOAD;
- val |= VSC73XX_MAC_CFG_WEXC_DIS;
- vsc73xx_write(vsc, VSC73XX_BLOCK_MAC, port, VSC73XX_MAC_CFG, val);
+ int ret, err;
+ u32 val;
- /* Flow control for the PHY facing ports:
- * Use a zero delay pause frame when pause condition is left
- * Obey pause control frames
- * When generating pause frames, use 0xff as pause value
- */
- vsc73xx_write(vsc, VSC73XX_BLOCK_MAC, port, VSC73XX_FCCONF,
- VSC73XX_FCCONF_ZERO_PAUSE_EN |
- VSC73XX_FCCONF_FLOW_CTRL_OBEY |
- 0xff);
+ /* Disable RX on this port */
+ vsc73xx_update_bits(vsc, VSC73XX_BLOCK_MAC, port,
+ VSC73XX_MAC_CFG,
+ VSC73XX_MAC_CFG_RX_EN, 0);
- /* Disallow backward dropping of frames from this port */
+ /* Discard packets */
vsc73xx_update_bits(vsc, VSC73XX_BLOCK_ARBITER, 0,
- VSC73XX_SBACKWDROP, BIT(port), 0);
+ VSC73XX_ARBDISC, BIT(port), BIT(port));
+
+ /* Wait until queue is empty */
+ ret = read_poll_timeout(vsc73xx_read, err,
+ err < 0 || (val & BIT(port)),
+ VSC73XX_POLL_SLEEP_US,
+ VSC73XX_POLL_TIMEOUT_US, false,
+ vsc, VSC73XX_BLOCK_ARBITER, 0,
+ VSC73XX_ARBEMPTY, &val);
+ if (ret)
+ dev_err(vsc->dev,
+ "timeout waiting for block arbiter\n");
+ else if (err < 0)
+ dev_err(vsc->dev, "error reading arbiter\n");
- /* Enable TX, RX, deassert reset, stop loading seed */
- vsc73xx_update_bits(vsc, VSC73XX_BLOCK_MAC, port,
- VSC73XX_MAC_CFG,
- VSC73XX_MAC_CFG_RESET | VSC73XX_MAC_CFG_SEED_LOAD |
- VSC73XX_MAC_CFG_TX_EN | VSC73XX_MAC_CFG_RX_EN,
- VSC73XX_MAC_CFG_TX_EN | VSC73XX_MAC_CFG_RX_EN);
+ /* Put this port into reset */
+ vsc73xx_write(vsc, VSC73XX_BLOCK_MAC, port, VSC73XX_MAC_CFG,
+ VSC73XX_MAC_CFG_RESET | initval);
}
-static void vsc73xx_adjust_link(struct dsa_switch *ds, int port,
- struct phy_device *phydev)
+static void vsc73xx_phylink_mac_config(struct dsa_switch *ds, int port,
+ unsigned int mode,
+ const struct phylink_link_state *state)
{
struct vsc73xx *vsc = ds->priv;
- u32 val;
-
/* Special handling of the CPU-facing port */
if (port == CPU_PORT) {
/* Other ports are already initialized but not this one */
@@ -778,102 +769,92 @@ static void vsc73xx_adjust_link(struct dsa_switch *ds, int port,
VSC73XX_ADVPORTM_ENA_GTX |
VSC73XX_ADVPORTM_DDR_MODE);
}
+}
+
+static void vsc73xx_phylink_mac_link_down(struct dsa_switch *ds, int port,
+ unsigned int mode,
+ phy_interface_t interface)
+{
+ struct vsc73xx *vsc = ds->priv;
- /* This is the MAC confiuration that always need to happen
- * after a PHY or the CPU port comes up or down.
+ /* This routine is described in the datasheet (below ARBDISC register
+ * description)
*/
- if (!phydev->link) {
- int ret, err;
-
- dev_dbg(vsc->dev, "port %d: went down\n",
- port);
-
- /* Disable RX on this port */
- vsc73xx_update_bits(vsc, VSC73XX_BLOCK_MAC, port,
- VSC73XX_MAC_CFG,
- VSC73XX_MAC_CFG_RX_EN, 0);
-
- /* Discard packets */
- vsc73xx_update_bits(vsc, VSC73XX_BLOCK_ARBITER, 0,
- VSC73XX_ARBDISC, BIT(port), BIT(port));
-
- /* Wait until queue is empty */
- ret = read_poll_timeout(vsc73xx_read, err,
- err < 0 || (val & BIT(port)),
- VSC73XX_POLL_SLEEP_US,
- VSC73XX_POLL_TIMEOUT_US, false,
- vsc, VSC73XX_BLOCK_ARBITER, 0,
- VSC73XX_ARBEMPTY, &val);
- if (ret)
- dev_err(vsc->dev,
- "timeout waiting for block arbiter\n");
- else if (err < 0)
- dev_err(vsc->dev, "error reading arbiter\n");
-
- /* Put this port into reset */
- vsc73xx_write(vsc, VSC73XX_BLOCK_MAC, port, VSC73XX_MAC_CFG,
- VSC73XX_MAC_CFG_RESET);
-
- /* Accept packets again */
- vsc73xx_update_bits(vsc, VSC73XX_BLOCK_ARBITER, 0,
- VSC73XX_ARBDISC, BIT(port), 0);
-
- /* Allow backward dropping of frames from this port */
- vsc73xx_update_bits(vsc, VSC73XX_BLOCK_ARBITER, 0,
- VSC73XX_SBACKWDROP, BIT(port), BIT(port));
-
- /* Receive mask (disable forwarding) */
- vsc73xx_update_bits(vsc, VSC73XX_BLOCK_ANALYZER, 0,
- VSC73XX_RECVMASK, BIT(port), 0);
+ vsc73xx_reset_port(vsc, port, 0);
- return;
- }
+ /* Allow backward dropping of frames from this port */
+ vsc73xx_update_bits(vsc, VSC73XX_BLOCK_ARBITER, 0,
+ VSC73XX_SBACKWDROP, BIT(port), BIT(port));
- /* Figure out what speed was negotiated */
- if (phydev->speed == SPEED_1000) {
- dev_dbg(vsc->dev, "port %d: 1000 Mbit mode full duplex\n",
- port);
-
- /* Set up default for internal port or external RGMII */
- if (phydev->interface == PHY_INTERFACE_MODE_RGMII)
- val = VSC73XX_MAC_CFG_1000M_F_RGMII;
- else
- val = VSC73XX_MAC_CFG_1000M_F_PHY;
- vsc73xx_adjust_enable_port(vsc, port, phydev, val);
- } else if (phydev->speed == SPEED_100) {
- if (phydev->duplex == DUPLEX_FULL) {
- val = VSC73XX_MAC_CFG_100_10M_F_PHY;
- dev_dbg(vsc->dev,
- "port %d: 100 Mbit full duplex mode\n",
- port);
- } else {
- val = VSC73XX_MAC_CFG_100_10M_H_PHY;
- dev_dbg(vsc->dev,
- "port %d: 100 Mbit half duplex mode\n",
- port);
- }
- vsc73xx_adjust_enable_port(vsc, port, phydev, val);
- } else if (phydev->speed == SPEED_10) {
- if (phydev->duplex == DUPLEX_FULL) {
- val = VSC73XX_MAC_CFG_100_10M_F_PHY;
- dev_dbg(vsc->dev,
- "port %d: 10 Mbit full duplex mode\n",
- port);
- } else {
- val = VSC73XX_MAC_CFG_100_10M_H_PHY;
- dev_dbg(vsc->dev,
- "port %d: 10 Mbit half duplex mode\n",
- port);
- }
- vsc73xx_adjust_enable_port(vsc, port, phydev, val);
- } else {
- dev_err(vsc->dev,
- "could not adjust link: unknown speed\n");
- }
+ /* Receive mask (disable forwarding) */
+ vsc73xx_update_bits(vsc, VSC73XX_BLOCK_ANALYZER, 0,
+ VSC73XX_RECVMASK, BIT(port), 0);
+}
+
+static void vsc73xx_phylink_mac_link_up(struct dsa_switch *ds, int port,
+ unsigned int mode,
+ phy_interface_t interface,
+ struct phy_device *phydev,
+ int speed, int duplex,
+ bool tx_pause, bool rx_pause)
+{
+ struct vsc73xx *vsc = ds->priv;
+ u32 val;
+ u8 seed;
+
+ if (speed == SPEED_1000)
+ val = VSC73XX_MAC_CFG_GIGA_MODE | VSC73XX_MAC_CFG_TX_IPG_1000M;
+ else
+ val = VSC73XX_MAC_CFG_TX_IPG_100_10M;
+
+ if (interface == PHY_INTERFACE_MODE_RGMII)
+ val |= VSC73XX_MAC_CFG_CLK_SEL_1000M;
+ else
+ val |= VSC73XX_MAC_CFG_CLK_SEL_EXT;
+
+ if (duplex == DUPLEX_FULL)
+ val |= VSC73XX_MAC_CFG_FDX;
+
+ /* This routine is described in the datasheet (below ARBDISC register
+ * description)
+ */
+ vsc73xx_reset_port(vsc, port, val);
+
+ /* Seed the port randomness with randomness */
+ get_random_bytes(&seed, 1);
+ val |= seed << VSC73XX_MAC_CFG_SEED_OFFSET;
+ val |= VSC73XX_MAC_CFG_SEED_LOAD;
+ val |= VSC73XX_MAC_CFG_WEXC_DIS;
+ vsc73xx_write(vsc, VSC73XX_BLOCK_MAC, port, VSC73XX_MAC_CFG, val);
+
+ /* Flow control for the PHY facing ports:
+ * Use a zero delay pause frame when pause condition is left
+ * Obey pause control frames
+ * When generating pause frames, use 0xff as pause value
+ */
+ vsc73xx_write(vsc, VSC73XX_BLOCK_MAC, port, VSC73XX_FCCONF,
+ VSC73XX_FCCONF_ZERO_PAUSE_EN |
+ VSC73XX_FCCONF_FLOW_CTRL_OBEY |
+ 0xff);
+
+ /* Accept packets again */
+ vsc73xx_update_bits(vsc, VSC73XX_BLOCK_ARBITER, 0,
+ VSC73XX_ARBDISC, BIT(port), 0);
/* Enable port (forwarding) in the receieve mask */
vsc73xx_update_bits(vsc, VSC73XX_BLOCK_ANALYZER, 0,
VSC73XX_RECVMASK, BIT(port), BIT(port));
+
+ /* Disallow backward dropping of frames from this port */
+ vsc73xx_update_bits(vsc, VSC73XX_BLOCK_ARBITER, 0,
+ VSC73XX_SBACKWDROP, BIT(port), 0);
+
+ /* Enable TX, RX, deassert reset, stop loading seed */
+ vsc73xx_update_bits(vsc, VSC73XX_BLOCK_MAC, port,
+ VSC73XX_MAC_CFG,
+ VSC73XX_MAC_CFG_RESET | VSC73XX_MAC_CFG_SEED_LOAD |
+ VSC73XX_MAC_CFG_TX_EN | VSC73XX_MAC_CFG_RX_EN,
+ VSC73XX_MAC_CFG_TX_EN | VSC73XX_MAC_CFG_RX_EN);
}
static int vsc73xx_port_enable(struct dsa_switch *ds, int port,
@@ -1060,7 +1041,9 @@ static const struct dsa_switch_ops vsc73xx_ds_ops = {
.setup = vsc73xx_setup,
.phy_read = vsc73xx_phy_read,
.phy_write = vsc73xx_phy_write,
- .adjust_link = vsc73xx_adjust_link,
+ .phylink_mac_config = vsc73xx_phylink_mac_config,
+ .phylink_mac_link_down = vsc73xx_phylink_mac_link_down,
+ .phylink_mac_link_up = vsc73xx_phylink_mac_link_up,
.get_strings = vsc73xx_get_strings,
.get_ethtool_stats = vsc73xx_get_ethtool_stats,
.get_sset_count = vsc73xx_get_sset_count,
--
2.34.1
It's preparation for future use. At this moment, the RGMII port is used
only for a connection to the MAC interface, but in the future, someone
could connect a PHY to it. Using the "phy_interface_mode_is_rgmii" macro
allows for the proper recognition of all RGMII modes.
Suggested-by: Russell King (Oracle) <[email protected]>
Signed-off-by: Pawel Dembicki <[email protected]>
Reviewed-by: Linus Walleij <[email protected]>
Reviewed-by: Florian Fainelli <[email protected]>
Reviewed-by: Vladimir Oltean <[email protected]>
Reviewed-by: Russell King (Oracle) <[email protected]>
---
v8:
- resend only
v7:
- added 'Reviewed-by' and fix 'Suggested-by' in commit message
v6:
- resend only
v5:
- added 'Reviewed-by' only
v4:
- introduced patch
drivers/net/dsa/vitesse-vsc73xx-core.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/net/dsa/vitesse-vsc73xx-core.c b/drivers/net/dsa/vitesse-vsc73xx-core.c
index cb2e7e256279..8ed8fc997d6d 100644
--- a/drivers/net/dsa/vitesse-vsc73xx-core.c
+++ b/drivers/net/dsa/vitesse-vsc73xx-core.c
@@ -807,7 +807,7 @@ static void vsc73xx_phylink_mac_link_up(struct dsa_switch *ds, int port,
else
val = VSC73XX_MAC_CFG_TX_IPG_100_10M;
- if (interface == PHY_INTERFACE_MODE_RGMII)
+ if (phy_interface_mode_is_rgmii(interface))
val |= VSC73XX_MAC_CFG_CLK_SEL_1000M;
else
val |= VSC73XX_MAC_CFG_CLK_SEL_EXT;
--
2.34.1
This patch introduces a new define: VSC73XX_MAX_NUM_PORTS, which can be
used in the future instead of a hardcoded value.
Currently, the only hardcoded value is vsc->ds->num_ports. It is being
replaced with the new define.
Suggested-by: Vladimir Oltean <[email protected]>
Signed-off-by: Pawel Dembicki <[email protected]>
Reviewed-by: Vladimir Oltean <[email protected]>
Reviewed-by: Linus Walleij <[email protected]>
Reviewed-by: Florian Fainelli <[email protected]>
---
v8:
- resend only
v7:
- added 'Reviewed-by' only
v6:
- resend only
v5:
- added 'Reviewed-by' only
v4:
- Resend patch
v3:
- Introduce patch
drivers/net/dsa/vitesse-vsc73xx-core.c | 13 +------------
drivers/net/dsa/vitesse-vsc73xx.h | 11 +++++++++++
2 files changed, 12 insertions(+), 12 deletions(-)
diff --git a/drivers/net/dsa/vitesse-vsc73xx-core.c b/drivers/net/dsa/vitesse-vsc73xx-core.c
index 8ed8fc997d6d..425999d7bf41 100644
--- a/drivers/net/dsa/vitesse-vsc73xx-core.c
+++ b/drivers/net/dsa/vitesse-vsc73xx-core.c
@@ -1180,23 +1180,12 @@ int vsc73xx_probe(struct vsc73xx *vsc)
vsc->addr[0], vsc->addr[1], vsc->addr[2],
vsc->addr[3], vsc->addr[4], vsc->addr[5]);
- /* The VSC7395 switch chips have 5+1 ports which means 5
- * ordinary ports and a sixth CPU port facing the processor
- * with an RGMII interface. These ports are numbered 0..4
- * and 6, so they leave a "hole" in the port map for port 5,
- * which is invalid.
- *
- * The VSC7398 has 8 ports, port 7 is again the CPU port.
- *
- * We allocate 8 ports and avoid access to the nonexistant
- * ports.
- */
vsc->ds = devm_kzalloc(dev, sizeof(*vsc->ds), GFP_KERNEL);
if (!vsc->ds)
return -ENOMEM;
vsc->ds->dev = dev;
- vsc->ds->num_ports = 8;
+ vsc->ds->num_ports = VSC73XX_MAX_NUM_PORTS;
vsc->ds->priv = vsc;
vsc->ds->ops = &vsc73xx_ds_ops;
diff --git a/drivers/net/dsa/vitesse-vsc73xx.h b/drivers/net/dsa/vitesse-vsc73xx.h
index 30b1f0a36566..fee1378508b5 100644
--- a/drivers/net/dsa/vitesse-vsc73xx.h
+++ b/drivers/net/dsa/vitesse-vsc73xx.h
@@ -3,6 +3,17 @@
#include <linux/etherdevice.h>
#include <linux/gpio/driver.h>
+/* The VSC7395 switch chips have 5+1 ports which means 5 ordinary ports and
+ * a sixth CPU port facing the processor with an RGMII interface. These ports
+ * are numbered 0..4 and 6, so they leave a "hole" in the port map for port 5,
+ * which is invalid.
+ *
+ * The VSC7398 has 8 ports, port 7 is again the CPU port.
+ *
+ * We allocate 8 ports and avoid access to the nonexistent ports.
+ */
+#define VSC73XX_MAX_NUM_PORTS 8
+
/**
* struct vsc73xx - VSC73xx state container
*/
--
2.34.1
This isn't a fully functional implementation of 802.1D, but
port_stp_state_set is required for a future tag8021q operations.
This implementation handles properly all states, but vsc73xx doesn't
forward STP packets.
Reviewed-by: Linus Walleij <[email protected]>
Reviewed-by: Florian Fainelli <[email protected]>
Reviewed-by: Vladimir Oltean <[email protected]>
Signed-off-by: Pawel Dembicki <[email protected]>
---
v8:
- resend only
v7:
- implement 'vsc73xx_refresh_fwd_map' simplification
v6:
- fix inconsistent indenting
v5:
- remove unneeded 'RECVMASK' operations
- reorganise vsc73xx_refresh_fwd_map function
v4:
- fully reworked port_stp_state_set
v3:
- use 'VSC73XX_MAX_NUM_PORTS' define
- add 'state == BR_STATE_DISABLED' condition
- fix style issues
v2:
- fix kdoc
drivers/net/dsa/vitesse-vsc73xx-core.c | 96 +++++++++++++++++++++++---
1 file changed, 85 insertions(+), 11 deletions(-)
diff --git a/drivers/net/dsa/vitesse-vsc73xx-core.c b/drivers/net/dsa/vitesse-vsc73xx-core.c
index 425999d7bf41..07f982cf62dd 100644
--- a/drivers/net/dsa/vitesse-vsc73xx-core.c
+++ b/drivers/net/dsa/vitesse-vsc73xx-core.c
@@ -164,6 +164,10 @@
#define VSC73XX_AGENCTRL 0xf0
#define VSC73XX_CAPRST 0xff
+#define VSC73XX_SRCMASKS_CPU_COPY BIT(27)
+#define VSC73XX_SRCMASKS_MIRROR BIT(26)
+#define VSC73XX_SRCMASKS_PORTS_MASK GENMASK(7, 0)
+
#define VSC73XX_MACACCESS_CPU_COPY BIT(14)
#define VSC73XX_MACACCESS_FWD_KILL BIT(13)
#define VSC73XX_MACACCESS_IGNORE_VLAN BIT(12)
@@ -623,9 +627,6 @@ static int vsc73xx_setup(struct dsa_switch *ds)
vsc73xx_write(vsc, VSC73XX_BLOCK_SYSTEM, 0, VSC73XX_GMIIDELAY,
VSC73XX_GMIIDELAY_GMII0_GTXDELAY_2_0_NS |
VSC73XX_GMIIDELAY_GMII0_RXDELAY_2_0_NS);
- /* Enable reception of frames on all ports */
- vsc73xx_write(vsc, VSC73XX_BLOCK_ANALYZER, 0, VSC73XX_RECVMASK,
- 0x5f);
/* IP multicast flood mask (table 144) */
vsc73xx_write(vsc, VSC73XX_BLOCK_ANALYZER, 0, VSC73XX_IFLODMSK,
0xff);
@@ -785,10 +786,6 @@ static void vsc73xx_phylink_mac_link_down(struct dsa_switch *ds, int port,
/* Allow backward dropping of frames from this port */
vsc73xx_update_bits(vsc, VSC73XX_BLOCK_ARBITER, 0,
VSC73XX_SBACKWDROP, BIT(port), BIT(port));
-
- /* Receive mask (disable forwarding) */
- vsc73xx_update_bits(vsc, VSC73XX_BLOCK_ANALYZER, 0,
- VSC73XX_RECVMASK, BIT(port), 0);
}
static void vsc73xx_phylink_mac_link_up(struct dsa_switch *ds, int port,
@@ -841,10 +838,6 @@ static void vsc73xx_phylink_mac_link_up(struct dsa_switch *ds, int port,
vsc73xx_update_bits(vsc, VSC73XX_BLOCK_ARBITER, 0,
VSC73XX_ARBDISC, BIT(port), 0);
- /* Enable port (forwarding) in the receieve mask */
- vsc73xx_update_bits(vsc, VSC73XX_BLOCK_ANALYZER, 0,
- VSC73XX_RECVMASK, BIT(port), BIT(port));
-
/* Disallow backward dropping of frames from this port */
vsc73xx_update_bits(vsc, VSC73XX_BLOCK_ARBITER, 0,
VSC73XX_SBACKWDROP, BIT(port), 0);
@@ -1036,6 +1029,86 @@ static void vsc73xx_phylink_get_caps(struct dsa_switch *dsa, int port,
config->mac_capabilities = MAC_SYM_PAUSE | MAC_10 | MAC_100 | MAC_1000;
}
+static void vsc73xx_refresh_fwd_map(struct dsa_switch *ds, int port, u8 state)
+{
+ struct dsa_port *other_dp, *dp = dsa_to_port(ds, port);
+ struct vsc73xx *vsc = ds->priv;
+ u16 mask;
+
+ if (state != BR_STATE_FORWARDING) {
+ /* Ports that aren't in the forwarding state must not
+ * forward packets anywhere.
+ */
+ vsc73xx_update_bits(vsc, VSC73XX_BLOCK_ANALYZER, 0,
+ VSC73XX_SRCMASKS + port,
+ VSC73XX_SRCMASKS_PORTS_MASK, 0);
+
+ dsa_switch_for_each_available_port(other_dp, ds) {
+ if (other_dp == dp)
+ continue;
+ vsc73xx_update_bits(vsc, VSC73XX_BLOCK_ANALYZER, 0,
+ VSC73XX_SRCMASKS + other_dp->index,
+ BIT(port), 0);
+ }
+
+ return;
+ }
+
+ /* Forwarding ports must forward to the CPU and to other ports
+ * in the same bridge
+ */
+ vsc73xx_update_bits(vsc, VSC73XX_BLOCK_ANALYZER, 0,
+ VSC73XX_SRCMASKS + CPU_PORT, BIT(port), BIT(port));
+
+ mask = BIT(CPU_PORT);
+
+ dsa_switch_for_each_user_port(other_dp, ds) {
+ int other_port = other_dp->index;
+
+ if (port == other_port || !dsa_port_bridge_same(dp, other_dp) ||
+ other_dp->stp_state != BR_STATE_FORWARDING)
+ continue;
+
+ mask |= BIT(other_port);
+
+ vsc73xx_update_bits(vsc, VSC73XX_BLOCK_ANALYZER, 0,
+ VSC73XX_SRCMASKS + other_port,
+ BIT(port), BIT(port));
+ }
+
+ vsc73xx_update_bits(vsc, VSC73XX_BLOCK_ANALYZER, 0,
+ VSC73XX_SRCMASKS + port,
+ VSC73XX_SRCMASKS_PORTS_MASK, mask);
+}
+
+/* FIXME: STP frames aren't forwarded at this moment. BPDU frames are
+ * forwarded only from and to PI/SI interface. For more info see chapter
+ * 2.7.1 (CPU Forwarding) in datasheet.
+ * This function is required for tag_8021q operations.
+ */
+static void vsc73xx_port_stp_state_set(struct dsa_switch *ds, int port,
+ u8 state)
+{
+ struct vsc73xx *vsc = ds->priv;
+ u32 val;
+
+ val = (state == BR_STATE_BLOCKING || state == BR_STATE_DISABLED) ?
+ 0 : BIT(port);
+ vsc73xx_update_bits(vsc, VSC73XX_BLOCK_ANALYZER, 0,
+ VSC73XX_RECVMASK, BIT(port), val);
+
+ val = (state == BR_STATE_LEARNING || state == BR_STATE_FORWARDING) ?
+ BIT(port) : 0;
+ vsc73xx_update_bits(vsc, VSC73XX_BLOCK_ANALYZER, 0,
+ VSC73XX_LEARNMASK, BIT(port), val);
+
+ /* CPU Port should always forward packets when user ports are forwarding
+ * so let's configure it from other ports only.
+ */
+ if (port != CPU_PORT)
+ vsc73xx_refresh_fwd_map(ds, port, state);
+}
+
static const struct dsa_switch_ops vsc73xx_ds_ops = {
.get_tag_protocol = vsc73xx_get_tag_protocol,
.setup = vsc73xx_setup,
@@ -1051,6 +1124,7 @@ static const struct dsa_switch_ops vsc73xx_ds_ops = {
.port_disable = vsc73xx_port_disable,
.port_change_mtu = vsc73xx_change_mtu,
.port_max_mtu = vsc73xx_get_max_mtu,
+ .port_stp_state_set = vsc73xx_port_stp_state_set,
.phylink_get_caps = vsc73xx_phylink_get_caps,
};
--
2.34.1
This commit adds updates to the documentation describing the structures
used in vsc73xx. This will help prevent kdoc-related issues in the future.
Reviewed-by: Linus Walleij <[email protected]>
Reviewed-by: Florian Fainelli <[email protected]>
Signed-off-by: Pawel Dembicki <[email protected]>
---
v8,v7,v6:
- resend only
v5:
- improved description and stick to 80 characters per line
v4:
- introduced patch
drivers/net/dsa/vitesse-vsc73xx.h | 16 +++++++++++++++-
1 file changed, 15 insertions(+), 1 deletion(-)
diff --git a/drivers/net/dsa/vitesse-vsc73xx.h b/drivers/net/dsa/vitesse-vsc73xx.h
index fee1378508b5..e7b08599a625 100644
--- a/drivers/net/dsa/vitesse-vsc73xx.h
+++ b/drivers/net/dsa/vitesse-vsc73xx.h
@@ -15,7 +15,16 @@
#define VSC73XX_MAX_NUM_PORTS 8
/**
- * struct vsc73xx - VSC73xx state container
+ * struct vsc73xx - VSC73xx state container: main data structure
+ * @dev: The device pointer
+ * @reset: The descriptor for the GPIO line tied to the reset pin
+ * @ds: Pointer to the DSA core structure
+ * @gc: Main structure of the GPIO controller
+ * @chipid: Storage for the Chip ID value read from the CHIPID register of the
+ * sswitch
+ * @addr: MAC address used in flow control frames
+ * @ops: Structure with hardware-dependent operations
+ * @priv: Pointer to the configuration interface structure
*/
struct vsc73xx {
struct device *dev;
@@ -28,6 +37,11 @@ struct vsc73xx {
void *priv;
};
+/**
+ * struct vsc73xx_ops - VSC73xx methods container
+ * @read: Method for register reading over the hardware-dependent interface
+ * @write: Method for register writing over the hardware-dependent interface
+ */
struct vsc73xx_ops {
int (*read)(struct vsc73xx *vsc, u8 block, u8 subblock, u8 reg,
u32 *val);
--
2.34.1
This patch implements VLAN filtering for the vsc73xx driver.
After starting VLAN filtering, the switch is reconfigured from QinQ to
a simple VLAN aware mode. This is required because VSC73XX chips do not
support inner VLAN tag filtering.
Signed-off-by: Pawel Dembicki <[email protected]>
---
v8:
- set init value of 'ret' in 'vsc73xx_port_vlan_add'
v7:
- rework pvid and untagged configuration routines
- introduce portinfo structure which should make pvid/untagged
procedures simpler
- introduce 'vsc73xx_vlan_summary' structure
- replace tagged/untagged count functions with 'vsc73xx_bridge_vlan_summary'
- fix VSC73XX_VLANMASK configuration. It was copy from existing code.
- stop configuring pvid/untagged registers whed pvid/untagged is
disabled
v6:
- resend only
v5:
- fix possible leak in 'vsc73xx_port_vlan_add'
- use proper variable in statement from 'vsc73xx_port_vlan_filtering'
- change 'vlan_no' name to 'vid'
- codding style improvements
- comment improvements
- handle return of 'vsc73xx_update_bits'
- reduce I/O operations
- use 'size_t' for counting variables
v4:
- reworked most of conditional register configs
- simplified port_vlan function
- move vlan table clearing from port_setup to setup
- pvid configuration simplified (now kernel take care about no of
pvids per port)
- port vlans are stored in list now
- introduce implementation of all untagged vlans state
- many minor changes
v3:
- reworked all vlan commits
- added storage variables for pvid and untagged vlans
- move length extender settings to port setup
- remove vlan table cleaning in wrong places
v2:
- no changes done
drivers/net/dsa/vitesse-vsc73xx-core.c | 545 ++++++++++++++++++++++++-
drivers/net/dsa/vitesse-vsc73xx.h | 42 ++
2 files changed, 585 insertions(+), 2 deletions(-)
diff --git a/drivers/net/dsa/vitesse-vsc73xx-core.c b/drivers/net/dsa/vitesse-vsc73xx-core.c
index 07f982cf62dd..433ffd4f44fd 100644
--- a/drivers/net/dsa/vitesse-vsc73xx-core.c
+++ b/drivers/net/dsa/vitesse-vsc73xx-core.c
@@ -22,9 +22,11 @@
#include <linux/of_mdio.h>
#include <linux/bitops.h>
#include <linux/if_bridge.h>
+#include <linux/if_vlan.h>
#include <linux/etherdevice.h>
#include <linux/gpio/consumer.h>
#include <linux/gpio/driver.h>
+#include <linux/dsa/8021q.h>
#include <linux/random.h>
#include <net/dsa.h>
@@ -62,6 +64,8 @@
#define VSC73XX_CAT_DROP 0x6e
#define VSC73XX_CAT_PR_MISC_L2 0x6f
#define VSC73XX_CAT_PR_USR_PRIO 0x75
+#define VSC73XX_CAT_VLAN_MISC 0x79
+#define VSC73XX_CAT_PORT_VLAN 0x7a
#define VSC73XX_Q_MISC_CONF 0xdf
/* MAC_CFG register bits */
@@ -122,6 +126,17 @@
#define VSC73XX_ADVPORTM_IO_LOOPBACK BIT(1)
#define VSC73XX_ADVPORTM_HOST_LOOPBACK BIT(0)
+/* TXUPDCFG transmit modify setup bits */
+#define VSC73XX_TXUPDCFG_DSCP_REWR_MODE GENMASK(20, 19)
+#define VSC73XX_TXUPDCFG_DSCP_REWR_ENA BIT(18)
+#define VSC73XX_TXUPDCFG_TX_INT_TO_USRPRIO_ENA BIT(17)
+#define VSC73XX_TXUPDCFG_TX_UNTAGGED_VID GENMASK(15, 4)
+#define VSC73XX_TXUPDCFG_TX_UNTAGGED_VID_ENA BIT(3)
+#define VSC73XX_TXUPDCFG_TX_UPDATE_CRC_CPU_ENA BIT(1)
+#define VSC73XX_TXUPDCFG_TX_INSERT_TAG BIT(0)
+
+#define VSC73XX_TXUPDCFG_TX_UNTAGGED_VID_SHIFT 4
+
/* CAT_DROP categorizer frame dropping register bits */
#define VSC73XX_CAT_DROP_DROP_MC_SMAC_ENA BIT(6)
#define VSC73XX_CAT_DROP_FWD_CTRL_ENA BIT(4)
@@ -135,6 +150,15 @@
#define VSC73XX_Q_MISC_CONF_EARLY_TX_512 (1 << 1)
#define VSC73XX_Q_MISC_CONF_MAC_PAUSE_MODE BIT(0)
+/* CAT_VLAN_MISC categorizer VLAN miscellaneous bits */
+#define VSC73XX_CAT_VLAN_MISC_VLAN_TCI_IGNORE_ENA BIT(8)
+#define VSC73XX_CAT_VLAN_MISC_VLAN_KEEP_TAG_ENA BIT(7)
+
+/* CAT_PORT_VLAN categorizer port VLAN */
+#define VSC73XX_CAT_PORT_VLAN_VLAN_CFI BIT(15)
+#define VSC73XX_CAT_PORT_VLAN_VLAN_USR_PRIO GENMASK(14, 12)
+#define VSC73XX_CAT_PORT_VLAN_VLAN_VID GENMASK(11, 0)
+
/* Frame analyzer block 2 registers */
#define VSC73XX_STORMLIMIT 0x02
#define VSC73XX_ADVLEARN 0x03
@@ -189,7 +213,8 @@
#define VSC73XX_VLANACCESS_VLAN_MIRROR BIT(29)
#define VSC73XX_VLANACCESS_VLAN_SRC_CHECK BIT(28)
#define VSC73XX_VLANACCESS_VLAN_PORT_MASK GENMASK(9, 2)
-#define VSC73XX_VLANACCESS_VLAN_TBL_CMD_MASK GENMASK(2, 0)
+#define VSC73XX_VLANACCESS_VLAN_PORT_MASK_SHIFT 2
+#define VSC73XX_VLANACCESS_VLAN_TBL_CMD_MASK GENMASK(1, 0)
#define VSC73XX_VLANACCESS_VLAN_TBL_CMD_IDLE 0
#define VSC73XX_VLANACCESS_VLAN_TBL_CMD_READ_ENTRY 1
#define VSC73XX_VLANACCESS_VLAN_TBL_CMD_WRITE_ENTRY 2
@@ -347,6 +372,17 @@ static const struct vsc73xx_counter vsc73xx_tx_counters[] = {
{ 29, "TxQoSClass3" }, /* non-standard counter */
};
+struct vsc73xx_vlan_summary {
+ size_t num_tagged;
+ size_t num_untagged;
+};
+
+enum vsc73xx_port_vlan_conf {
+ VSC73XX_VLAN_FILTER,
+ VSC73XX_VLAN_FILTER_UNTAG_ALL,
+ VSC73XX_VLAN_IGNORE,
+};
+
int vsc73xx_is_addr_valid(u8 block, u8 subblock)
{
switch (block) {
@@ -564,6 +600,90 @@ static enum dsa_tag_protocol vsc73xx_get_tag_protocol(struct dsa_switch *ds,
return DSA_TAG_PROTO_NONE;
}
+static int vsc73xx_wait_for_vlan_table_cmd(struct vsc73xx *vsc)
+{
+ int ret, err;
+ u32 val;
+
+ ret = read_poll_timeout(vsc73xx_read, err,
+ err < 0 ||
+ ((val & VSC73XX_VLANACCESS_VLAN_TBL_CMD_MASK) ==
+ VSC73XX_VLANACCESS_VLAN_TBL_CMD_IDLE),
+ VSC73XX_POLL_SLEEP_US, VSC73XX_POLL_TIMEOUT_US,
+ false, vsc, VSC73XX_BLOCK_ANALYZER,
+ 0, VSC73XX_VLANACCESS, &val);
+ if (ret)
+ return ret;
+ return err;
+}
+
+static int
+vsc73xx_read_vlan_table_entry(struct vsc73xx *vsc, u16 vid, u8 *portmap)
+{
+ u32 val;
+ int ret;
+
+ vsc73xx_write(vsc, VSC73XX_BLOCK_ANALYZER, 0, VSC73XX_VLANTIDX, vid);
+
+ ret = vsc73xx_wait_for_vlan_table_cmd(vsc);
+ if (ret)
+ return ret;
+
+ vsc73xx_update_bits(vsc, VSC73XX_BLOCK_ANALYZER, 0, VSC73XX_VLANACCESS,
+ VSC73XX_VLANACCESS_VLAN_TBL_CMD_MASK,
+ VSC73XX_VLANACCESS_VLAN_TBL_CMD_READ_ENTRY);
+
+ ret = vsc73xx_wait_for_vlan_table_cmd(vsc);
+ if (ret)
+ return ret;
+
+ vsc73xx_read(vsc, VSC73XX_BLOCK_ANALYZER, 0, VSC73XX_VLANACCESS, &val);
+ *portmap = (val & VSC73XX_VLANACCESS_VLAN_PORT_MASK) >>
+ VSC73XX_VLANACCESS_VLAN_PORT_MASK_SHIFT;
+
+ return 0;
+}
+
+static int
+vsc73xx_write_vlan_table_entry(struct vsc73xx *vsc, u16 vid, u8 portmap)
+{
+ int ret;
+
+ vsc73xx_write(vsc, VSC73XX_BLOCK_ANALYZER, 0, VSC73XX_VLANTIDX, vid);
+
+ ret = vsc73xx_wait_for_vlan_table_cmd(vsc);
+ if (ret)
+ return ret;
+
+ vsc73xx_update_bits(vsc, VSC73XX_BLOCK_ANALYZER, 0, VSC73XX_VLANACCESS,
+ VSC73XX_VLANACCESS_VLAN_TBL_CMD_MASK |
+ VSC73XX_VLANACCESS_VLAN_SRC_CHECK |
+ VSC73XX_VLANACCESS_VLAN_PORT_MASK,
+ VSC73XX_VLANACCESS_VLAN_TBL_CMD_WRITE_ENTRY |
+ VSC73XX_VLANACCESS_VLAN_SRC_CHECK |
+ (portmap << VSC73XX_VLANACCESS_VLAN_PORT_MASK_SHIFT));
+
+ return vsc73xx_wait_for_vlan_table_cmd(vsc);
+}
+
+static int
+vsc73xx_update_vlan_table(struct vsc73xx *vsc, int port, u16 vid, bool set)
+{
+ u8 portmap;
+ int ret;
+
+ ret = vsc73xx_read_vlan_table_entry(vsc, vid, &portmap);
+ if (ret)
+ return ret;
+
+ if (set)
+ portmap |= BIT(port);
+ else
+ portmap &= ~BIT(port);
+
+ return vsc73xx_write_vlan_table_entry(vsc, vid, portmap);
+}
+
static int vsc73xx_setup(struct dsa_switch *ds)
{
struct vsc73xx *vsc = ds->priv;
@@ -598,7 +718,7 @@ static int vsc73xx_setup(struct dsa_switch *ds)
VSC73XX_MACACCESS,
VSC73XX_MACACCESS_CMD_CLEAR_TABLE);
- /* Clear VLAN table */
+ /* Set VLAN table to default values */
vsc73xx_write(vsc, VSC73XX_BLOCK_ANALYZER, 0,
VSC73XX_VLANACCESS,
VSC73XX_VLANACCESS_VLAN_TBL_CMD_CLEAR_TABLE);
@@ -627,6 +747,9 @@ static int vsc73xx_setup(struct dsa_switch *ds)
vsc73xx_write(vsc, VSC73XX_BLOCK_SYSTEM, 0, VSC73XX_GMIIDELAY,
VSC73XX_GMIIDELAY_GMII0_GTXDELAY_2_0_NS |
VSC73XX_GMIIDELAY_GMII0_RXDELAY_2_0_NS);
+ /* Ingess VLAN reception mask (table 145) */
+ vsc73xx_write(vsc, VSC73XX_BLOCK_ANALYZER, 0, VSC73XX_VLANMASK,
+ 0xff);
/* IP multicast flood mask (table 144) */
vsc73xx_write(vsc, VSC73XX_BLOCK_ANALYZER, 0, VSC73XX_IFLODMSK,
0xff);
@@ -639,6 +762,12 @@ static int vsc73xx_setup(struct dsa_switch *ds)
udelay(4);
+ /* Clear VLAN table */
+ for (i = 0; i < VLAN_N_VID; i++)
+ vsc73xx_write_vlan_table_entry(vsc, i, 0);
+
+ INIT_LIST_HEAD(&vsc->vlans);
+
return 0;
}
@@ -822,6 +951,12 @@ static void vsc73xx_phylink_mac_link_up(struct dsa_switch *ds, int port,
val |= seed << VSC73XX_MAC_CFG_SEED_OFFSET;
val |= VSC73XX_MAC_CFG_SEED_LOAD;
val |= VSC73XX_MAC_CFG_WEXC_DIS;
+
+ /* Those bits are responsible for MTU only. Kernel takes care about MTU,
+ * let's enable +8 bytes frame length unconditionally.
+ */
+ val |= VSC73XX_MAC_CFG_VLAN_AWR | VSC73XX_MAC_CFG_VLAN_DBLAWR;
+
vsc73xx_write(vsc, VSC73XX_BLOCK_MAC, port, VSC73XX_MAC_CFG, val);
/* Flow control for the PHY facing ports:
@@ -1029,6 +1164,408 @@ static void vsc73xx_phylink_get_caps(struct dsa_switch *dsa, int port,
config->mac_capabilities = MAC_SYM_PAUSE | MAC_10 | MAC_100 | MAC_1000;
}
+static bool vsc73xx_tag_8021q_active(struct dsa_port *dp)
+{
+ return !dsa_port_is_vlan_filtering(dp);
+}
+
+static void
+vsc73xx_set_vlan_conf(struct vsc73xx *vsc, int port,
+ enum vsc73xx_port_vlan_conf port_vlan_conf)
+{
+ u32 val = 0;
+
+ if (port_vlan_conf == VSC73XX_VLAN_IGNORE)
+ val = VSC73XX_CAT_VLAN_MISC_VLAN_TCI_IGNORE_ENA |
+ VSC73XX_CAT_VLAN_MISC_VLAN_KEEP_TAG_ENA;
+
+ vsc73xx_update_bits(vsc, VSC73XX_BLOCK_MAC, port, VSC73XX_CAT_VLAN_MISC,
+ VSC73XX_CAT_VLAN_MISC_VLAN_TCI_IGNORE_ENA |
+ VSC73XX_CAT_VLAN_MISC_VLAN_KEEP_TAG_ENA, val);
+
+ val = (port_vlan_conf == VSC73XX_VLAN_FILTER) ?
+ VSC73XX_TXUPDCFG_TX_INSERT_TAG : 0;
+ vsc73xx_update_bits(vsc, VSC73XX_BLOCK_MAC, port, VSC73XX_TXUPDCFG,
+ VSC73XX_TXUPDCFG_TX_INSERT_TAG, val);
+}
+
+static int vsc73xx_vlan_change_untagged_hw(struct vsc73xx *vsc, int port,
+ u16 vid, bool set)
+{
+ u32 val = 0;
+
+ if (set)
+ val = VSC73XX_TXUPDCFG_TX_UNTAGGED_VID_ENA |
+ ((vid << VSC73XX_TXUPDCFG_TX_UNTAGGED_VID_SHIFT) &
+ VSC73XX_TXUPDCFG_TX_UNTAGGED_VID);
+
+ return vsc73xx_update_bits(vsc, VSC73XX_BLOCK_MAC, port,
+ VSC73XX_TXUPDCFG,
+ VSC73XX_TXUPDCFG_TX_UNTAGGED_VID_ENA |
+ VSC73XX_TXUPDCFG_TX_UNTAGGED_VID, val);
+}
+
+static int vsc73xx_vlan_set_untagged_hw(struct vsc73xx *vsc, int port, u16 vid)
+{
+ return vsc73xx_vlan_change_untagged_hw(vsc, port, vid, true);
+}
+
+static int
+vsc73xx_vlan_change_pvid_hw(struct vsc73xx *vsc, int port, u16 vid, bool set)
+{
+ u32 val = 0;
+ int ret;
+
+ val = set ? 0 : VSC73XX_CAT_DROP_UNTAGGED_ENA;
+
+ ret = vsc73xx_update_bits(vsc, VSC73XX_BLOCK_MAC, port,
+ VSC73XX_CAT_DROP,
+ VSC73XX_CAT_DROP_UNTAGGED_ENA, val);
+ if (!set || ret)
+ return ret;
+
+ return vsc73xx_update_bits(vsc, VSC73XX_BLOCK_MAC, port,
+ VSC73XX_CAT_PORT_VLAN,
+ VSC73XX_CAT_PORT_VLAN_VLAN_VID,
+ vid & VSC73XX_CAT_PORT_VLAN_VLAN_VID);
+}
+
+static int vsc73xx_vlan_change_pvid(struct vsc73xx *vsc, int port, u16 vid,
+ bool valid, bool operate_on_storage,
+ bool tag_8021q_mode)
+{
+ struct vsc73xx_portinfo *portinfo = &vsc->portinfo[port];
+
+ if (tag_8021q_mode) {
+ portinfo->pvid_tag_8021q_configured = valid;
+ portinfo->pvid_tag_8021q = vid;
+ } else {
+ portinfo->pvid_vlan_filtering_configured = valid;
+ portinfo->pvid_vlan_filtering = vid;
+ }
+
+ if (operate_on_storage)
+ return 0;
+
+ return vsc73xx_vlan_change_pvid_hw(vsc, port, vid, valid);
+}
+
+static int vsc73xx_vlan_clear_pvid(struct vsc73xx *vsc, int port,
+ bool operate_on_storage, bool tag_8021q_mode)
+{
+ return vsc73xx_vlan_change_pvid(vsc, port, 0, false,
+ operate_on_storage, tag_8021q_mode);
+}
+
+static int vsc73xx_vlan_set_pvid(struct vsc73xx *vsc, int port, u16 vid,
+ bool operate_on_storage, bool tag_8021q_mode)
+{
+ return vsc73xx_vlan_change_pvid(vsc, port, vid, true,
+ operate_on_storage, tag_8021q_mode);
+}
+
+static bool vsc73xx_port_get_pvid(struct vsc73xx *vsc, int port, u16 *vid,
+ bool tag_8021q_mode)
+{
+ struct vsc73xx_portinfo *portinfo = &vsc->portinfo[port];
+
+ if (tag_8021q_mode) {
+ if (portinfo->pvid_tag_8021q_configured)
+ *vid = portinfo->pvid_tag_8021q;
+ return portinfo->pvid_tag_8021q_configured;
+ }
+
+ if (portinfo->pvid_vlan_filtering_configured)
+ *vid = portinfo->pvid_vlan_filtering;
+ return portinfo->pvid_vlan_filtering_configured;
+}
+
+static struct vsc73xx_bridge_vlan *
+vsc73xx_bridge_vlan_find(struct vsc73xx *vsc, u16 vid)
+{
+ struct vsc73xx_bridge_vlan *vlan;
+
+ list_for_each_entry(vlan, &vsc->vlans, list)
+ if (vlan->vid == vid)
+ return vlan;
+
+ return NULL;
+}
+
+static void vsc73xx_bridge_vlan_summary(struct vsc73xx *vsc, int port,
+ struct vsc73xx_vlan_summary *summary,
+ u16 ignored_vid)
+{
+ size_t num_tagged = 0, num_untagged = 0;
+ struct vsc73xx_bridge_vlan *vlan;
+
+ list_for_each_entry(vlan, &vsc->vlans, list) {
+ if (!(vlan->portmask & BIT(port)) || vlan->vid == ignored_vid)
+ continue;
+
+ if (vlan->untagged & BIT(port))
+ num_untagged++;
+ else
+ num_tagged++;
+ }
+
+ summary->num_untagged = num_untagged;
+ summary->num_tagged = num_tagged;
+}
+
+static u16 vsc73xx_find_first_vlan_untagged(struct vsc73xx *vsc, int port)
+{
+ struct vsc73xx_bridge_vlan *vlan;
+
+ list_for_each_entry(vlan, &vsc->vlans, list)
+ if ((vlan->portmask & BIT(port)) &&
+ (vlan->untagged & BIT(port)))
+ return vlan->vid;
+
+ return VLAN_N_VID;
+}
+
+static int
+vsc73xx_port_vlan_filtering(struct dsa_switch *ds, int port,
+ bool vlan_filtering, struct netlink_ext_ack *extack)
+{
+ enum vsc73xx_port_vlan_conf port_vlan_conf = VSC73XX_VLAN_IGNORE;
+ u16 vid_pvid = 0, vid_untagged = 0;
+ struct vsc73xx_portinfo *portinfo;
+ struct vsc73xx *vsc = ds->priv;
+ bool set_untagged = false;
+ bool set_pvid = false;
+
+ portinfo = &vsc->portinfo[port];
+
+ /* The swap processed below is required because vsc73xx is using
+ * tag_8021q. When vlan_filtering is disabled, tag_8021q uses
+ * pvid/untagged vlans for port recognition. The values configured for
+ * vlans and pvid/untagged states are stored in portinfo structure.
+ * When vlan_filtering is enabled, we need to restore pvid/untagged from
+ * portinfo structure. Analogic routine is processed when vlan_filtering
+ * is disabled, but values used for tag_8021q are restored.
+ */
+ if (vlan_filtering) {
+ struct vsc73xx_vlan_summary summary;
+
+ /* Use VLAN_N_VID to count all vlans */
+ vsc73xx_bridge_vlan_summary(vsc, port, &summary, VLAN_N_VID);
+
+ port_vlan_conf = (summary.num_untagged > 1) ?
+ VSC73XX_VLAN_FILTER_UNTAG_ALL :
+ VSC73XX_VLAN_FILTER;
+
+ if (summary.num_untagged == 1) {
+ vid_untagged = vsc73xx_find_first_vlan_untagged(vsc,
+ port);
+ set_untagged = true;
+ }
+
+ vid_pvid = portinfo->pvid_vlan_filtering;
+ set_pvid = portinfo->pvid_vlan_filtering_configured;
+ } else {
+ vid_untagged = portinfo->untagged_tag_8021q;
+ set_untagged = portinfo->untagged_tag_8021q_configured;
+ vid_pvid = portinfo->pvid_tag_8021q;
+ set_pvid = portinfo->pvid_tag_8021q_configured;
+ }
+
+ vsc73xx_vlan_change_untagged_hw(vsc, port, vid_untagged, set_untagged);
+ vsc73xx_vlan_change_pvid_hw(vsc, port, vid_pvid, set_pvid);
+ vsc73xx_set_vlan_conf(vsc, port, port_vlan_conf);
+
+ return 0;
+}
+
+static int vsc73xx_port_vlan_add(struct dsa_switch *ds, int port,
+ const struct switchdev_obj_port_vlan *vlan,
+ struct netlink_ext_ack *extack)
+{
+ bool untagged = vlan->flags & BRIDGE_VLAN_INFO_UNTAGGED;
+ bool pvid = vlan->flags & BRIDGE_VLAN_INFO_PVID;
+ struct dsa_port *dp = dsa_to_port(ds, port);
+ enum vsc73xx_port_vlan_conf port_vlan_conf;
+ struct vsc73xx_bridge_vlan *vsc73xx_vlan;
+ struct vsc73xx_vlan_summary summary;
+ struct vsc73xx *vsc = ds->priv;
+ bool operate_on_storage;
+ int ret = 0;
+ u16 vid;
+
+ /* Be sure to deny alterations to the configuration done by tag_8021q.
+ */
+ if (vid_is_dsa_8021q(vlan->vid)) {
+ NL_SET_ERR_MSG_MOD(extack,
+ "Range 3072-4095 reserved for dsa_8021q operation");
+ return -EBUSY;
+ }
+
+ /* The processed vlan->vid is excluded from the search because the VLAN
+ * can be re-added with a different set of flags, so it's easiest to
+ * ignore its old flags from the VLAN database software copy.
+ */
+ vsc73xx_bridge_vlan_summary(vsc, port, &summary, vlan->vid);
+
+ /* VSC73XX allow only three untagged states: none, one or all */
+ if ((untagged && summary.num_tagged > 0 && summary.num_untagged > 0) ||
+ (!untagged && summary.num_untagged > 1)) {
+ NL_SET_ERR_MSG_MOD(extack,
+ "Port can have only none, one or all untagged vlan");
+ return -EBUSY;
+ }
+
+ vsc73xx_vlan = vsc73xx_bridge_vlan_find(vsc, vlan->vid);
+
+ if (!vsc73xx_vlan) {
+ vsc73xx_vlan = kzalloc(sizeof(*vsc73xx_vlan), GFP_KERNEL);
+ if (!vsc73xx_vlan)
+ return -ENOMEM;
+
+ vsc73xx_vlan->vid = vlan->vid;
+ vsc73xx_vlan->portmask = 0;
+ vsc73xx_vlan->untagged = 0;
+
+ INIT_LIST_HEAD(&vsc73xx_vlan->list);
+ list_add_tail(&vsc73xx_vlan->list, &vsc->vlans);
+ }
+
+ vsc73xx_vlan->portmask |= BIT(port);
+
+ if (untagged)
+ vsc73xx_vlan->untagged |= BIT(port);
+ else
+ vsc73xx_vlan->untagged &= ~BIT(port);
+
+ /* CPU port must be always tagged because port separation is based on
+ * tag_8021q.
+ */
+ if (port == CPU_PORT)
+ goto update_vlan_table;
+
+ operate_on_storage = vsc73xx_tag_8021q_active(dp);
+
+ if (pvid)
+ ret = vsc73xx_vlan_set_pvid(vsc, port, vlan->vid,
+ operate_on_storage, false);
+ else if (vsc73xx_port_get_pvid(vsc, port, &vid, false) &&
+ vid == vlan->vid)
+ ret = vsc73xx_vlan_clear_pvid(vsc, port, operate_on_storage,
+ false);
+ if (ret)
+ goto err;
+
+ if (operate_on_storage)
+ goto update_vlan_table;
+
+ port_vlan_conf = VSC73XX_VLAN_FILTER;
+
+ if (summary.num_tagged == 0 && untagged)
+ port_vlan_conf = VSC73XX_VLAN_FILTER_UNTAG_ALL;
+ vsc73xx_set_vlan_conf(vsc, port, port_vlan_conf);
+
+ if (port_vlan_conf == VSC73XX_VLAN_FILTER_UNTAG_ALL)
+ goto update_vlan_table;
+
+ if (untagged) {
+ ret = vsc73xx_vlan_set_untagged_hw(vsc, port, vlan->vid);
+ } else if (summary.num_untagged == 1) {
+ vid = vsc73xx_find_first_vlan_untagged(vsc, port);
+ ret = vsc73xx_vlan_set_untagged_hw(vsc, port, vid);
+ }
+ if (ret)
+ goto err;
+
+update_vlan_table:
+ ret = vsc73xx_update_vlan_table(vsc, port, vlan->vid, true);
+ if (!ret)
+ return 0;
+err:
+ list_del(&vsc73xx_vlan->list);
+ kfree(vsc73xx_vlan);
+ return ret;
+}
+
+static int vsc73xx_port_vlan_del(struct dsa_switch *ds, int port,
+ const struct switchdev_obj_port_vlan *vlan)
+{
+ struct vsc73xx_bridge_vlan *vsc73xx_vlan;
+ struct vsc73xx_vlan_summary summary;
+ struct vsc73xx *vsc = ds->priv;
+ bool operate_on_storage;
+ int ret;
+ u16 vid;
+
+ vsc73xx_bridge_vlan_summary(vsc, port, &summary, vlan->vid);
+
+ ret = vsc73xx_update_vlan_table(vsc, port, vlan->vid, false);
+ if (ret)
+ return ret;
+
+ operate_on_storage = vsc73xx_tag_8021q_active(dsa_to_port(ds, port));
+
+ if (!operate_on_storage) {
+ enum vsc73xx_port_vlan_conf port_vlan_conf =
+ VSC73XX_VLAN_FILTER;
+
+ if (summary.num_tagged == 0)
+ port_vlan_conf = VSC73XX_VLAN_FILTER_UNTAG_ALL;
+ vsc73xx_set_vlan_conf(vsc, port, port_vlan_conf);
+
+ if (summary.num_untagged <= 1) {
+ vid = vsc73xx_find_first_vlan_untagged(vsc, port);
+ vsc73xx_vlan_change_untagged_hw(vsc, port, vid,
+ summary.num_untagged);
+ }
+ }
+
+ if (vsc73xx_port_get_pvid(vsc, port, &vid, false) && vid == vlan->vid)
+ vsc73xx_vlan_clear_pvid(vsc, port, operate_on_storage, false);
+
+ vsc73xx_vlan = vsc73xx_bridge_vlan_find(vsc, vlan->vid);
+
+ if (vsc73xx_vlan) {
+ vsc73xx_vlan->portmask &= ~BIT(port);
+
+ if (vsc73xx_vlan->portmask)
+ return 0;
+
+ list_del(&vsc73xx_vlan->list);
+ kfree(vsc73xx_vlan);
+ }
+
+ return 0;
+}
+
+static int vsc73xx_port_setup(struct dsa_switch *ds, int port)
+{
+ struct vsc73xx_portinfo *portinfo;
+ struct vsc73xx *vsc = ds->priv;
+
+ portinfo = &vsc->portinfo[port];
+
+ vsc73xx_update_bits(vsc, VSC73XX_BLOCK_MAC, port, VSC73XX_CAT_DROP,
+ VSC73XX_CAT_DROP_TAGGED_ENA |
+ VSC73XX_CAT_DROP_UNTAGGED_ENA,
+ VSC73XX_CAT_DROP_UNTAGGED_ENA);
+ vsc73xx_update_bits(vsc, VSC73XX_BLOCK_MAC, port, VSC73XX_TXUPDCFG,
+ VSC73XX_TXUPDCFG_TX_UNTAGGED_VID_ENA |
+ VSC73XX_TXUPDCFG_TX_UNTAGGED_VID, 0);
+ vsc73xx_update_bits(vsc, VSC73XX_BLOCK_MAC, port, VSC73XX_CAT_PORT_VLAN,
+ VSC73XX_CAT_PORT_VLAN_VLAN_VID, 0);
+
+ if (port == CPU_PORT)
+ vsc73xx_set_vlan_conf(vsc, port, VSC73XX_VLAN_FILTER);
+ else
+ vsc73xx_set_vlan_conf(vsc, port, VSC73XX_VLAN_IGNORE);
+
+ portinfo->pvid_vlan_filtering_configured = false;
+ portinfo->pvid_tag_8021q_configured = false;
+ portinfo->untagged_tag_8021q_configured = false;
+
+ return 0;
+}
+
static void vsc73xx_refresh_fwd_map(struct dsa_switch *ds, int port, u8 state)
{
struct dsa_port *other_dp, *dp = dsa_to_port(ds, port);
@@ -1120,11 +1657,15 @@ static const struct dsa_switch_ops vsc73xx_ds_ops = {
.get_strings = vsc73xx_get_strings,
.get_ethtool_stats = vsc73xx_get_ethtool_stats,
.get_sset_count = vsc73xx_get_sset_count,
+ .port_setup = vsc73xx_port_setup,
.port_enable = vsc73xx_port_enable,
.port_disable = vsc73xx_port_disable,
.port_change_mtu = vsc73xx_change_mtu,
.port_max_mtu = vsc73xx_get_max_mtu,
.port_stp_state_set = vsc73xx_port_stp_state_set,
+ .port_vlan_filtering = vsc73xx_port_vlan_filtering,
+ .port_vlan_add = vsc73xx_port_vlan_add,
+ .port_vlan_del = vsc73xx_port_vlan_del,
.phylink_get_caps = vsc73xx_phylink_get_caps,
};
diff --git a/drivers/net/dsa/vitesse-vsc73xx.h b/drivers/net/dsa/vitesse-vsc73xx.h
index e7b08599a625..ce007273eafe 100644
--- a/drivers/net/dsa/vitesse-vsc73xx.h
+++ b/drivers/net/dsa/vitesse-vsc73xx.h
@@ -14,6 +14,27 @@
*/
#define VSC73XX_MAX_NUM_PORTS 8
+/**
+ * struct vsc73xx_portinfo - port data structure: contains storage data
+ * @pvid_vlan_filtering_configured: imforms if port have configured pvid in vlan
+ * fitering mode
+ * @pvid_vlan_filtering: pvid vlan number used in vlan fitering mode
+ * @pvid_tag_8021q_configured: imforms if port have configured pvid in tag_8021q
+ * mode
+ * @pvid_tag_8021q: pvid vlan number used in tag_8021q mode
+ * @untagged_tag_8021q_configured: imforms if port have configured untagged vlan
+ * in tag_8021q mode
+ * @untagged_tag_8021q: untagged vlan number used in tag_8021q mode
+ */
+struct vsc73xx_portinfo {
+ bool pvid_vlan_filtering_configured;
+ u16 pvid_vlan_filtering;
+ bool pvid_tag_8021q_configured;
+ u16 pvid_tag_8021q;
+ bool untagged_tag_8021q_configured;
+ u16 untagged_tag_8021q;
+};
+
/**
* struct vsc73xx - VSC73xx state container: main data structure
* @dev: The device pointer
@@ -25,6 +46,10 @@
* @addr: MAC address used in flow control frames
* @ops: Structure with hardware-dependent operations
* @priv: Pointer to the configuration interface structure
+ * @portinfo: Storage table portinfo structructures
+ * @vlans: List of configured vlans. Contains port mask and untagged status of
+ * every vlan configured in port vlan operation. It doesn't cover tag_8021q
+ * vlans.
*/
struct vsc73xx {
struct device *dev;
@@ -35,6 +60,8 @@ struct vsc73xx {
u8 addr[ETH_ALEN];
const struct vsc73xx_ops *ops;
void *priv;
+ struct vsc73xx_portinfo portinfo[VSC73XX_MAX_NUM_PORTS];
+ struct list_head vlans;
};
/**
@@ -49,6 +76,21 @@ struct vsc73xx_ops {
u32 val);
};
+/**
+ * struct vsc73xx_bridge_vlan - VSC73xx driver structure which keeps vlan
+ * database copy
+ * @vid: VLAN number
+ * @portmask: each bit represents one port
+ * @untagged: each bit represents one port configured with @vid untagged
+ * @list: list structure
+ */
+struct vsc73xx_bridge_vlan {
+ u16 vid;
+ u8 portmask;
+ u8 untagged;
+ struct list_head list;
+};
+
int vsc73xx_is_addr_valid(u8 block, u8 subblock);
int vsc73xx_probe(struct vsc73xx *vsc);
void vsc73xx_remove(struct vsc73xx *vsc);
--
2.34.1
From: Vladimir Oltean <[email protected]>
tag_sja1105 has a wrapper over dsa_8021q_rcv(): sja1105_vlan_rcv(),
which determines whether the packet came from a bridge with
vlan_filtering=1 (the case resolved via
dsa_find_designated_bridge_port_by_vid()), or if it contains a tag_8021q
header.
Looking at a new tagger implementation for vsc73xx, based also on
tag_8021q, it is becoming clear that the logic is needed there as well.
So instead of forcing each tagger to wrap around dsa_8021q_rcv(), let's
merge the logic into the core.
Signed-off-by: Vladimir Oltean <[email protected]>
Signed-off-by: Pawel Dembicki <[email protected]>
---
v8, v7, v6:
- resend only
v5:
- add missing SoB
v4:
- introduced patch
net/dsa/tag_8021q.c | 34 ++++++++++++++++++++++++++++------
net/dsa/tag_8021q.h | 2 +-
net/dsa/tag_ocelot_8021q.c | 2 +-
net/dsa/tag_sja1105.c | 32 ++++----------------------------
4 files changed, 34 insertions(+), 36 deletions(-)
diff --git a/net/dsa/tag_8021q.c b/net/dsa/tag_8021q.c
index 3cb0293793a5..332b0ae02645 100644
--- a/net/dsa/tag_8021q.c
+++ b/net/dsa/tag_8021q.c
@@ -507,27 +507,39 @@ EXPORT_SYMBOL_GPL(dsa_tag_8021q_find_port_by_vbid);
* @vbid: pointer to storage for imprecise bridge ID. Must be pre-initialized
* with -1. If a positive value is returned, the source_port and switch_id
* are invalid.
+ * @vid: pointer to storage for original VID, in case tag_8021q decoding failed.
+ *
+ * If the packet has a tag_8021q header, decode it and set @source_port,
+ * @switch_id and @vbid, and strip the header. Otherwise set @vid and keep the
+ * header in the hwaccel area of the packet.
*/
void dsa_8021q_rcv(struct sk_buff *skb, int *source_port, int *switch_id,
- int *vbid)
+ int *vbid, int *vid)
{
int tmp_source_port, tmp_switch_id, tmp_vbid;
- u16 vid, tci;
+ __be16 vlan_proto;
+ u16 tmp_vid, tci;
if (skb_vlan_tag_present(skb)) {
+ vlan_proto = skb->vlan_proto;
tci = skb_vlan_tag_get(skb);
__vlan_hwaccel_clear_tag(skb);
} else {
+ struct vlan_ethhdr *hdr = vlan_eth_hdr(skb);
+
+ vlan_proto = hdr->h_vlan_proto;
skb_push_rcsum(skb, ETH_HLEN);
__skb_vlan_pop(skb, &tci);
skb_pull_rcsum(skb, ETH_HLEN);
}
- vid = tci & VLAN_VID_MASK;
+ tmp_vid = tci & VLAN_VID_MASK;
+ if (!vid_is_dsa_8021q(tmp_vid))
+ goto not_tag_8021q;
- tmp_source_port = dsa_8021q_rx_source_port(vid);
- tmp_switch_id = dsa_8021q_rx_switch_id(vid);
- tmp_vbid = dsa_tag_8021q_rx_vbid(vid);
+ tmp_source_port = dsa_8021q_rx_source_port(tmp_vid);
+ tmp_switch_id = dsa_8021q_rx_switch_id(tmp_vid);
+ tmp_vbid = dsa_tag_8021q_rx_vbid(tmp_vid);
/* Precise source port information is unknown when receiving from a
* VLAN-unaware bridging domain, and tmp_source_port and tmp_switch_id
@@ -546,5 +558,15 @@ void dsa_8021q_rcv(struct sk_buff *skb, int *source_port, int *switch_id,
*vbid = tmp_vbid;
skb->priority = (tci & VLAN_PRIO_MASK) >> VLAN_PRIO_SHIFT;
+ return;
+
+not_tag_8021q:
+ if (vid)
+ *vid = tmp_vid;
+ if (vbid)
+ *vbid = -1;
+
+ /* Put the tag back */
+ __vlan_hwaccel_put_tag(skb, vlan_proto, tci);
}
EXPORT_SYMBOL_GPL(dsa_8021q_rcv);
diff --git a/net/dsa/tag_8021q.h b/net/dsa/tag_8021q.h
index 41f7167ac520..0c6671d7c1c2 100644
--- a/net/dsa/tag_8021q.h
+++ b/net/dsa/tag_8021q.h
@@ -14,7 +14,7 @@ struct sk_buff *dsa_8021q_xmit(struct sk_buff *skb, struct net_device *netdev,
u16 tpid, u16 tci);
void dsa_8021q_rcv(struct sk_buff *skb, int *source_port, int *switch_id,
- int *vbid);
+ int *vbid, int *vid);
struct net_device *dsa_tag_8021q_find_port_by_vbid(struct net_device *conduit,
int vbid);
diff --git a/net/dsa/tag_ocelot_8021q.c b/net/dsa/tag_ocelot_8021q.c
index b059381310fe..8e8b1bef6af6 100644
--- a/net/dsa/tag_ocelot_8021q.c
+++ b/net/dsa/tag_ocelot_8021q.c
@@ -81,7 +81,7 @@ static struct sk_buff *ocelot_rcv(struct sk_buff *skb,
{
int src_port, switch_id;
- dsa_8021q_rcv(skb, &src_port, &switch_id, NULL);
+ dsa_8021q_rcv(skb, &src_port, &switch_id, NULL, NULL);
skb->dev = dsa_conduit_find_user(netdev, switch_id, src_port);
if (!skb->dev)
diff --git a/net/dsa/tag_sja1105.c b/net/dsa/tag_sja1105.c
index 48886d4b7e3e..7639ccb94d35 100644
--- a/net/dsa/tag_sja1105.c
+++ b/net/dsa/tag_sja1105.c
@@ -472,37 +472,14 @@ static bool sja1110_skb_has_inband_control_extension(const struct sk_buff *skb)
return ntohs(eth_hdr(skb)->h_proto) == ETH_P_SJA1110;
}
-/* If the VLAN in the packet is a tag_8021q one, set @source_port and
- * @switch_id and strip the header. Otherwise set @vid and keep it in the
- * packet.
- */
-static void sja1105_vlan_rcv(struct sk_buff *skb, int *source_port,
- int *switch_id, int *vbid, u16 *vid)
-{
- struct vlan_ethhdr *hdr = vlan_eth_hdr(skb);
- u16 vlan_tci;
-
- if (skb_vlan_tag_present(skb))
- vlan_tci = skb_vlan_tag_get(skb);
- else
- vlan_tci = ntohs(hdr->h_vlan_TCI);
-
- if (vid_is_dsa_8021q(vlan_tci & VLAN_VID_MASK))
- return dsa_8021q_rcv(skb, source_port, switch_id, vbid);
-
- /* Try our best with imprecise RX */
- *vid = vlan_tci & VLAN_VID_MASK;
-}
-
static struct sk_buff *sja1105_rcv(struct sk_buff *skb,
struct net_device *netdev)
{
- int source_port = -1, switch_id = -1, vbid = -1;
+ int source_port = -1, switch_id = -1, vbid = -1, vid = -1;
struct sja1105_meta meta = {0};
struct ethhdr *hdr;
bool is_link_local;
bool is_meta;
- u16 vid;
hdr = eth_hdr(skb);
is_link_local = sja1105_is_link_local(skb);
@@ -525,7 +502,7 @@ static struct sk_buff *sja1105_rcv(struct sk_buff *skb,
* a tag_8021q VLAN which we have to strip
*/
if (sja1105_skb_has_tag_8021q(skb))
- sja1105_vlan_rcv(skb, &source_port, &switch_id, &vbid, &vid);
+ dsa_8021q_rcv(skb, &source_port, &switch_id, &vbid, &vid);
else if (source_port == -1 && switch_id == -1)
/* Packets with no source information have no chance of
* getting accepted, drop them straight away.
@@ -660,9 +637,8 @@ static struct sk_buff *sja1110_rcv_inband_control_extension(struct sk_buff *skb,
static struct sk_buff *sja1110_rcv(struct sk_buff *skb,
struct net_device *netdev)
{
- int source_port = -1, switch_id = -1, vbid = -1;
+ int source_port = -1, switch_id = -1, vbid = -1, vid = -1;
bool host_only = false;
- u16 vid = 0;
if (sja1110_skb_has_inband_control_extension(skb)) {
skb = sja1110_rcv_inband_control_extension(skb, &source_port,
@@ -674,7 +650,7 @@ static struct sk_buff *sja1110_rcv(struct sk_buff *skb,
/* Packets with in-band control extensions might still have RX VLANs */
if (likely(sja1105_skb_has_tag_8021q(skb)))
- sja1105_vlan_rcv(skb, &source_port, &switch_id, &vbid, &vid);
+ dsa_8021q_rcv(skb, &source_port, &switch_id, &vbid, &vid);
if (vbid >= 1)
skb->dev = dsa_tag_8021q_find_port_by_vbid(netdev, vbid);
--
2.34.1
This commit introduces a new tagger based on 802.1q tagging.
It's designed for the vsc73xx driver. The VSC73xx family doesn't have
any tag support for the RGMII port, but it could be based on VLANs.
Reviewed-by: Florian Fainelli <[email protected]>
Reviewed-by: Vladimir Oltean <[email protected]>
Signed-off-by: Pawel Dembicki <[email protected]>
Reviewed-by: Linus Walleij <[email protected]>
---
v8:
- added 'Reviewed-by' only
v7:
- replace netdev_warn by dev_warn_ratelimited
v6:
- added missing MODULE_DESCRIPTION()
v5:
- removed skb_vlan_tag_present(skb) checking
- use 80 characters per line limit
v4:
- rebase to net-next/main
v3:
- Introduce a patch after the tagging patch split
include/net/dsa.h | 2 ++
net/dsa/Kconfig | 6 ++++
net/dsa/Makefile | 1 +
net/dsa/tag_vsc73xx_8021q.c | 68 +++++++++++++++++++++++++++++++++++++
4 files changed, 77 insertions(+)
create mode 100644 net/dsa/tag_vsc73xx_8021q.c
diff --git a/include/net/dsa.h b/include/net/dsa.h
index 7c0da9effe4e..b79e136e4c41 100644
--- a/include/net/dsa.h
+++ b/include/net/dsa.h
@@ -56,6 +56,7 @@ struct phylink_link_state;
#define DSA_TAG_PROTO_RTL8_4T_VALUE 25
#define DSA_TAG_PROTO_RZN1_A5PSW_VALUE 26
#define DSA_TAG_PROTO_LAN937X_VALUE 27
+#define DSA_TAG_PROTO_VSC73XX_8021Q_VALUE 28
enum dsa_tag_protocol {
DSA_TAG_PROTO_NONE = DSA_TAG_PROTO_NONE_VALUE,
@@ -86,6 +87,7 @@ enum dsa_tag_protocol {
DSA_TAG_PROTO_RTL8_4T = DSA_TAG_PROTO_RTL8_4T_VALUE,
DSA_TAG_PROTO_RZN1_A5PSW = DSA_TAG_PROTO_RZN1_A5PSW_VALUE,
DSA_TAG_PROTO_LAN937X = DSA_TAG_PROTO_LAN937X_VALUE,
+ DSA_TAG_PROTO_VSC73XX_8021Q = DSA_TAG_PROTO_VSC73XX_8021Q_VALUE,
};
struct dsa_switch;
diff --git a/net/dsa/Kconfig b/net/dsa/Kconfig
index 8e698bea99a3..e59360071c67 100644
--- a/net/dsa/Kconfig
+++ b/net/dsa/Kconfig
@@ -166,6 +166,12 @@ config NET_DSA_TAG_TRAILER
Say Y or M if you want to enable support for tagging frames at
with a trailed. e.g. Marvell 88E6060.
+config NET_DSA_TAG_VSC73XX_8021Q
+ tristate "Tag driver for Microchip/Vitesse VSC73xx family of switches, using VLAN"
+ help
+ Say Y or M if you want to enable support for tagging frames with a
+ custom VLAN-based header.
+
config NET_DSA_TAG_XRS700X
tristate "Tag driver for XRS700x switches"
help
diff --git a/net/dsa/Makefile b/net/dsa/Makefile
index 8a1894a42552..555c07cfeb71 100644
--- a/net/dsa/Makefile
+++ b/net/dsa/Makefile
@@ -37,6 +37,7 @@ obj-$(CONFIG_NET_DSA_TAG_RTL8_4) += tag_rtl8_4.o
obj-$(CONFIG_NET_DSA_TAG_RZN1_A5PSW) += tag_rzn1_a5psw.o
obj-$(CONFIG_NET_DSA_TAG_SJA1105) += tag_sja1105.o
obj-$(CONFIG_NET_DSA_TAG_TRAILER) += tag_trailer.o
+obj-$(CONFIG_NET_DSA_TAG_VSC73XX_8021Q) += tag_vsc73xx_8021q.o
obj-$(CONFIG_NET_DSA_TAG_XRS700X) += tag_xrs700x.o
# for tracing framework to find trace.h
diff --git a/net/dsa/tag_vsc73xx_8021q.c b/net/dsa/tag_vsc73xx_8021q.c
new file mode 100644
index 000000000000..af121a9aff7f
--- /dev/null
+++ b/net/dsa/tag_vsc73xx_8021q.c
@@ -0,0 +1,68 @@
+// SPDX-License-Identifier: GPL-2.0 OR MIT
+/* Copyright (C) 2024 Pawel Dembicki <[email protected]>
+ */
+#include <linux/dsa/8021q.h>
+
+#include "tag.h"
+#include "tag_8021q.h"
+
+#define VSC73XX_8021Q_NAME "vsc73xx-8021q"
+
+static struct sk_buff *
+vsc73xx_xmit(struct sk_buff *skb, struct net_device *netdev)
+{
+ struct dsa_port *dp = dsa_user_to_port(netdev);
+ u16 queue_mapping = skb_get_queue_mapping(skb);
+ u16 tx_vid = dsa_tag_8021q_standalone_vid(dp);
+ u8 pcp;
+
+ if (skb->offload_fwd_mark) {
+ unsigned int bridge_num = dsa_port_bridge_num_get(dp);
+ struct net_device *br = dsa_port_bridge_dev_get(dp);
+
+ if (br_vlan_enabled(br))
+ return skb;
+
+ tx_vid = dsa_tag_8021q_bridge_vid(bridge_num);
+ }
+
+ pcp = netdev_txq_to_tc(netdev, queue_mapping);
+
+ return dsa_8021q_xmit(skb, netdev, ETH_P_8021Q,
+ ((pcp << VLAN_PRIO_SHIFT) | tx_vid));
+}
+
+static struct sk_buff *
+vsc73xx_rcv(struct sk_buff *skb, struct net_device *netdev)
+{
+ int src_port = -1, switch_id = -1, vbid = -1, vid = -1;
+
+ dsa_8021q_rcv(skb, &src_port, &switch_id, &vbid, &vid);
+
+ skb->dev = dsa_tag_8021q_find_user(netdev, src_port, switch_id,
+ vid, vbid);
+ if (!skb->dev) {
+ dev_warn_ratelimited(&netdev->dev,
+ "Couldn't decode source port\n");
+ return NULL;
+ }
+
+ dsa_default_offload_fwd_mark(skb);
+
+ return skb;
+}
+
+static const struct dsa_device_ops vsc73xx_8021q_netdev_ops = {
+ .name = VSC73XX_8021Q_NAME,
+ .proto = DSA_TAG_PROTO_VSC73XX_8021Q,
+ .xmit = vsc73xx_xmit,
+ .rcv = vsc73xx_rcv,
+ .needed_headroom = VLAN_HLEN,
+ .promisc_on_conduit = true,
+};
+
+MODULE_LICENSE("GPL");
+MODULE_DESCRIPTION("DSA tag driver for VSC73XX family of switches, using VLAN");
+MODULE_ALIAS_DSA_TAG_DRIVER(DSA_TAG_PROTO_VSC73XX_8021Q, VSC73XX_8021Q_NAME);
+
+module_dsa_tag_driver(vsc73xx_8021q_netdev_ops);
--
2.34.1
This patch adds bridge support for vsc73xx driver.
It introduce two functions for port_bridge_join and
vsc73xx_port_bridge_leave handling.
Those functions implement forwarding adjust and use
dsa_tag_8021q_bridge_* api for adjust VLAN configuration.
Reviewed-by: Linus Walleij <[email protected]>
Signed-off-by: Pawel Dembicki <[email protected]>
Reviewed-by: Vladimir Oltean <[email protected]>
Reviewed-by: Florian Fainelli <[email protected]>
---
v8:
- resend only
v7:
- added 'Reviewed-by' only
v6:
- resend only
v5:
- added 'Reviewed-by' only
v4:
- remove forward configuration after stp patch refactoring
- implement new define with max num of bridges for tag8021q devices
v3:
- All vlan commits was reworked
- move VLAN_AWR and VLAN_DBLAWR to port setup in other commit
- drop vlan table upgrade
v2:
- no changes done
drivers/net/dsa/vitesse-vsc73xx-core.c | 21 +++++++++++++++++++++
1 file changed, 21 insertions(+)
diff --git a/drivers/net/dsa/vitesse-vsc73xx-core.c b/drivers/net/dsa/vitesse-vsc73xx-core.c
index d0442ea8a610..87922804d1c1 100644
--- a/drivers/net/dsa/vitesse-vsc73xx-core.c
+++ b/drivers/net/dsa/vitesse-vsc73xx-core.c
@@ -691,6 +691,9 @@ static int vsc73xx_setup(struct dsa_switch *ds)
dev_info(vsc->dev, "set up the switch\n");
+ ds->untag_bridge_pvid = true;
+ ds->max_num_bridges = DSA_TAG_8021Q_MAX_NUM_BRIDGES;
+
/* Issue RESET */
vsc73xx_write(vsc, VSC73XX_BLOCK_SYSTEM, 0, VSC73XX_GLORESET,
VSC73XX_GLORESET_MASTER_RESET);
@@ -1585,6 +1588,22 @@ static int vsc73xx_tag_8021q_vlan_del(struct dsa_switch *ds, int port, u16 vid)
return vsc73xx_update_vlan_table(vsc, port, vid, false);
}
+static int vsc73xx_port_bridge_join(struct dsa_switch *ds, int port,
+ struct dsa_bridge bridge,
+ bool *tx_fwd_offload,
+ struct netlink_ext_ack *extack)
+{
+ *tx_fwd_offload = true;
+
+ return dsa_tag_8021q_bridge_join(ds, port, bridge);
+}
+
+static void vsc73xx_port_bridge_leave(struct dsa_switch *ds, int port,
+ struct dsa_bridge bridge)
+{
+ dsa_tag_8021q_bridge_leave(ds, port, bridge);
+}
+
static int vsc73xx_port_setup(struct dsa_switch *ds, int port)
{
struct vsc73xx_portinfo *portinfo;
@@ -1708,6 +1727,8 @@ static const struct dsa_switch_ops vsc73xx_ds_ops = {
.port_setup = vsc73xx_port_setup,
.port_enable = vsc73xx_port_enable,
.port_disable = vsc73xx_port_disable,
+ .port_bridge_join = vsc73xx_port_bridge_join,
+ .port_bridge_leave = vsc73xx_port_bridge_leave,
.port_change_mtu = vsc73xx_change_mtu,
.port_max_mtu = vsc73xx_get_max_mtu,
.port_stp_state_set = vsc73xx_port_stp_state_set,
--
2.34.1
Max number of bridges in tag8021q implementation is strictly limited
by VBID size: 3 bits. But zero is reserved and only 7 values can be used.
This patch adds define which describe maximum possible value.
Suggested-by: Vladimir Oltean <[email protected]>
Signed-off-by: Pawel Dembicki <[email protected]>
Reviewed-by: Florian Fainelli <[email protected]>
Reviewed-by: Linus Walleij <[email protected]>
---
v8:
- resend only
v7:
- added 'Reviewed-by' only
v6:
- resend only
v5:
- added 'Reviewed-by' only
v4:
- introduce patch
drivers/net/dsa/sja1105/sja1105_main.c | 3 +--
include/linux/dsa/8021q.h | 5 +++++
2 files changed, 6 insertions(+), 2 deletions(-)
diff --git a/drivers/net/dsa/sja1105/sja1105_main.c b/drivers/net/dsa/sja1105/sja1105_main.c
index fc262348a134..155fd28419f9 100644
--- a/drivers/net/dsa/sja1105/sja1105_main.c
+++ b/drivers/net/dsa/sja1105/sja1105_main.c
@@ -3156,8 +3156,7 @@ static int sja1105_setup(struct dsa_switch *ds)
ds->vlan_filtering_is_global = true;
ds->untag_bridge_pvid = true;
ds->fdb_isolation = true;
- /* tag_8021q has 3 bits for the VBID, and the value 0 is reserved */
- ds->max_num_bridges = 7;
+ ds->max_num_bridges = DSA_TAG_8021Q_MAX_NUM_BRIDGES;
/* Advertise the 8 egress queues */
ds->num_tx_queues = SJA1105_NUM_TC;
diff --git a/include/linux/dsa/8021q.h b/include/linux/dsa/8021q.h
index f3664ee12170..1dda2a13b832 100644
--- a/include/linux/dsa/8021q.h
+++ b/include/linux/dsa/8021q.h
@@ -8,6 +8,11 @@
#include <net/dsa.h>
#include <linux/types.h>
+/* VBID is limited to three bits only and zero is reserved.
+ * Only 7 bridges can be enumerated.
+ */
+#define DSA_TAG_8021Q_MAX_NUM_BRIDGES 7
+
int dsa_tag_8021q_register(struct dsa_switch *ds, __be16 proto);
void dsa_tag_8021q_unregister(struct dsa_switch *ds);
--
2.34.1
This patch implements .port_pre_bridge_flags() and .port_bridge_flags(),
which are required for properly treating the BR_LEARNING flag. Also,
port_stp_state_set() is tweaked and now disables learning for standalone
ports.
Disabling learning for standalone ports is required to avoid situations
where one port sees traffic originating from another, which could cause
packet drops.
Signed-off-by: Pawel Dembicki <[email protected]>
Acked-by: Linus Walleij <[email protected]>
Reviewed-by: Florian Fainelli <[email protected]>
Reviewed-by: Vladimir Oltean <[email protected]>
---
v8:
- resend only
v7:
- added 'Acked-by' and 'Reviewed-by' and improve commit message
v6:
- fix arranging local variables in reverse xmas tree order
v5:
- introduce patch
drivers/net/dsa/vitesse-vsc73xx-core.c | 41 ++++++++++++++++++++++----
1 file changed, 35 insertions(+), 6 deletions(-)
diff --git a/drivers/net/dsa/vitesse-vsc73xx-core.c b/drivers/net/dsa/vitesse-vsc73xx-core.c
index 87922804d1c1..90ca791add06 100644
--- a/drivers/net/dsa/vitesse-vsc73xx-core.c
+++ b/drivers/net/dsa/vitesse-vsc73xx-core.c
@@ -1588,6 +1588,31 @@ static int vsc73xx_tag_8021q_vlan_del(struct dsa_switch *ds, int port, u16 vid)
return vsc73xx_update_vlan_table(vsc, port, vid, false);
}
+static int vsc73xx_port_pre_bridge_flags(struct dsa_switch *ds, int port,
+ struct switchdev_brport_flags flags,
+ struct netlink_ext_ack *extack)
+{
+ if (flags.mask & ~BR_LEARNING)
+ return -EINVAL;
+
+ return 0;
+}
+
+static int vsc73xx_port_bridge_flags(struct dsa_switch *ds, int port,
+ struct switchdev_brport_flags flags,
+ struct netlink_ext_ack *extack)
+{
+ if (flags.mask & BR_LEARNING) {
+ u32 val = flags.val & BR_LEARNING ? BIT(port) : 0;
+ struct vsc73xx *vsc = ds->priv;
+
+ return vsc73xx_update_bits(vsc, VSC73XX_BLOCK_ANALYZER, 0,
+ VSC73XX_LEARNMASK, BIT(port), val);
+ }
+
+ return 0;
+}
+
static int vsc73xx_port_bridge_join(struct dsa_switch *ds, int port,
struct dsa_bridge bridge,
bool *tx_fwd_offload,
@@ -1693,19 +1718,21 @@ static void vsc73xx_refresh_fwd_map(struct dsa_switch *ds, int port, u8 state)
static void vsc73xx_port_stp_state_set(struct dsa_switch *ds, int port,
u8 state)
{
+ struct dsa_port *dp = dsa_to_port(ds, port);
struct vsc73xx *vsc = ds->priv;
- u32 val;
+ u32 val = 0;
+
+ if (state == BR_STATE_LEARNING || state == BR_STATE_FORWARDING)
+ val = dp->learning ? BIT(port) : 0;
+
+ vsc73xx_update_bits(vsc, VSC73XX_BLOCK_ANALYZER, 0,
+ VSC73XX_LEARNMASK, BIT(port), val);
val = (state == BR_STATE_BLOCKING || state == BR_STATE_DISABLED) ?
0 : BIT(port);
vsc73xx_update_bits(vsc, VSC73XX_BLOCK_ANALYZER, 0,
VSC73XX_RECVMASK, BIT(port), val);
- val = (state == BR_STATE_LEARNING || state == BR_STATE_FORWARDING) ?
- BIT(port) : 0;
- vsc73xx_update_bits(vsc, VSC73XX_BLOCK_ANALYZER, 0,
- VSC73XX_LEARNMASK, BIT(port), val);
-
/* CPU Port should always forward packets when user ports are forwarding
* so let's configure it from other ports only.
*/
@@ -1727,6 +1754,8 @@ static const struct dsa_switch_ops vsc73xx_ds_ops = {
.port_setup = vsc73xx_port_setup,
.port_enable = vsc73xx_port_enable,
.port_disable = vsc73xx_port_disable,
+ .port_pre_bridge_flags = vsc73xx_port_pre_bridge_flags,
+ .port_bridge_flags = vsc73xx_port_bridge_flags,
.port_bridge_join = vsc73xx_port_bridge_join,
.port_bridge_leave = vsc73xx_port_bridge_leave,
.port_change_mtu = vsc73xx_change_mtu,
--
2.34.1
From: Vladimir Oltean <[email protected]>
In both sja1105_rcv() and sja1110_rcv(), we may have precise source port
information coming from parallel hardware mechanisms, in addition to the
tag_8021q header.
Only sja1105_rcv() has extra logic to not overwrite that precise info
with what's present in the VLAN tag. This is because sja1110_rcv() gets
by, by having a reversed set of checks when assigning skb->dev. When the
source port is imprecise (vbid >=1), source_port and switch_id will be
set to zeroes by dsa_8021q_rcv(), which might be problematic. But by
checking for vbid >= 1 first, sja1110_rcv() fends that off.
We would like to make more code common between sja1105_rcv() and
sja1110_rcv(), and for that, we need to make sure that sja1110_rcv()
also goes through the precise source port preservation logic.
Signed-off-by: Vladimir Oltean <[email protected]>
Signed-off-by: Pawel Dembicki <[email protected]>
---
v8, v7, v6:
- resend only
v5:
- add missing SoB
v4:
- introduced patch
net/dsa/tag_8021q.c | 32 +++++++++++++++++++++++++++++---
net/dsa/tag_sja1105.c | 23 +++--------------------
2 files changed, 32 insertions(+), 23 deletions(-)
diff --git a/net/dsa/tag_8021q.c b/net/dsa/tag_8021q.c
index 71b26ae6db39..3cb0293793a5 100644
--- a/net/dsa/tag_8021q.c
+++ b/net/dsa/tag_8021q.c
@@ -497,9 +497,21 @@ struct net_device *dsa_tag_8021q_find_port_by_vbid(struct net_device *conduit,
}
EXPORT_SYMBOL_GPL(dsa_tag_8021q_find_port_by_vbid);
+/**
+ * dsa_8021q_rcv - Decode source information from tag_8021q header
+ * @skb: RX socket buffer
+ * @source_port: pointer to storage for precise source port information.
+ * If this is known already from outside tag_8021q, the pre-initialized
+ * value is preserved. If not known, pass -1.
+ * @switch_id: similar to source_port.
+ * @vbid: pointer to storage for imprecise bridge ID. Must be pre-initialized
+ * with -1. If a positive value is returned, the source_port and switch_id
+ * are invalid.
+ */
void dsa_8021q_rcv(struct sk_buff *skb, int *source_port, int *switch_id,
int *vbid)
{
+ int tmp_source_port, tmp_switch_id, tmp_vbid;
u16 vid, tci;
if (skb_vlan_tag_present(skb)) {
@@ -513,11 +525,25 @@ void dsa_8021q_rcv(struct sk_buff *skb, int *source_port, int *switch_id,
vid = tci & VLAN_VID_MASK;
- *source_port = dsa_8021q_rx_source_port(vid);
- *switch_id = dsa_8021q_rx_switch_id(vid);
+ tmp_source_port = dsa_8021q_rx_source_port(vid);
+ tmp_switch_id = dsa_8021q_rx_switch_id(vid);
+ tmp_vbid = dsa_tag_8021q_rx_vbid(vid);
+
+ /* Precise source port information is unknown when receiving from a
+ * VLAN-unaware bridging domain, and tmp_source_port and tmp_switch_id
+ * are zeroes in this case.
+ *
+ * Preserve the source information from hardware-specific mechanisms,
+ * if available. This allows us to not overwrite a valid source port
+ * and switch ID with less precise values.
+ */
+ if (tmp_vbid == 0 && *source_port == -1)
+ *source_port = tmp_source_port;
+ if (tmp_vbid == 0 && *switch_id == -1)
+ *switch_id = tmp_switch_id;
if (vbid)
- *vbid = dsa_tag_8021q_rx_vbid(vid);
+ *vbid = tmp_vbid;
skb->priority = (tci & VLAN_PRIO_MASK) >> VLAN_PRIO_SHIFT;
}
diff --git a/net/dsa/tag_sja1105.c b/net/dsa/tag_sja1105.c
index 1aba1d05c27a..48886d4b7e3e 100644
--- a/net/dsa/tag_sja1105.c
+++ b/net/dsa/tag_sja1105.c
@@ -524,30 +524,13 @@ static struct sk_buff *sja1105_rcv(struct sk_buff *skb,
/* Normal data plane traffic and link-local frames are tagged with
* a tag_8021q VLAN which we have to strip
*/
- if (sja1105_skb_has_tag_8021q(skb)) {
- int tmp_source_port = -1, tmp_switch_id = -1;
-
- sja1105_vlan_rcv(skb, &tmp_source_port, &tmp_switch_id, &vbid,
- &vid);
- /* Preserve the source information from the INCL_SRCPT option,
- * if available. This allows us to not overwrite a valid source
- * port and switch ID with zeroes when receiving link-local
- * frames from a VLAN-unaware bridged port (non-zero vbid) or a
- * VLAN-aware bridged port (non-zero vid). Furthermore, the
- * tag_8021q source port information is only of trust when the
- * vbid is 0 (precise port). Otherwise, tmp_source_port and
- * tmp_switch_id will be zeroes.
- */
- if (vbid == 0 && source_port == -1)
- source_port = tmp_source_port;
- if (vbid == 0 && switch_id == -1)
- switch_id = tmp_switch_id;
- } else if (source_port == -1 && switch_id == -1) {
+ if (sja1105_skb_has_tag_8021q(skb))
+ sja1105_vlan_rcv(skb, &source_port, &switch_id, &vbid, &vid);
+ else if (source_port == -1 && switch_id == -1)
/* Packets with no source information have no chance of
* getting accepted, drop them straight away.
*/
return NULL;
- }
if (source_port != -1 && switch_id != -1)
skb->dev = dsa_conduit_find_user(netdev, switch_id, source_port);
--
2.34.1
From: Vladimir Oltean <[email protected]>
A new tagging protocol implementation based on tag_8021q is on the
horizon, and it appears that it also has to open-code the complicated
logic of finding a source port based on a VLAN header.
Create a single dsa_tag_8021q_find_user() and make sja1105 call it.
Signed-off-by: Vladimir Oltean <[email protected]>
Signed-off-by: Pawel Dembicki <[email protected]>
Reviewed-by: Florian Fainelli <[email protected]>
---
v8:
- resend only
v7:
- added 'Reviewed-by' only
v6, v5:
- resend only
v4:
- introduce patch and change from master to conduit and slave to user
net/dsa/tag_8021q.c | 19 ++++++++++++++++---
net/dsa/tag_8021q.h | 5 +++--
net/dsa/tag_sja1105.c | 17 +++++------------
3 files changed, 24 insertions(+), 17 deletions(-)
diff --git a/net/dsa/tag_8021q.c b/net/dsa/tag_8021q.c
index 332b0ae02645..454d36c84671 100644
--- a/net/dsa/tag_8021q.c
+++ b/net/dsa/tag_8021q.c
@@ -468,8 +468,8 @@ struct sk_buff *dsa_8021q_xmit(struct sk_buff *skb, struct net_device *netdev,
}
EXPORT_SYMBOL_GPL(dsa_8021q_xmit);
-struct net_device *dsa_tag_8021q_find_port_by_vbid(struct net_device *conduit,
- int vbid)
+static struct net_device *
+dsa_tag_8021q_find_port_by_vbid(struct net_device *conduit, int vbid)
{
struct dsa_port *cpu_dp = conduit->dsa_ptr;
struct dsa_switch_tree *dst = cpu_dp->dst;
@@ -495,7 +495,20 @@ struct net_device *dsa_tag_8021q_find_port_by_vbid(struct net_device *conduit,
return NULL;
}
-EXPORT_SYMBOL_GPL(dsa_tag_8021q_find_port_by_vbid);
+
+struct net_device *dsa_tag_8021q_find_user(struct net_device *conduit,
+ int source_port, int switch_id,
+ int vid, int vbid)
+{
+ /* Always prefer precise source port information, if available */
+ if (source_port != -1 && switch_id != -1)
+ return dsa_conduit_find_user(conduit, switch_id, source_port);
+ else if (vbid >= 1)
+ return dsa_tag_8021q_find_port_by_vbid(conduit, vbid);
+
+ return dsa_find_designated_bridge_port_by_vid(conduit, vid);
+}
+EXPORT_SYMBOL_GPL(dsa_tag_8021q_find_user);
/**
* dsa_8021q_rcv - Decode source information from tag_8021q header
diff --git a/net/dsa/tag_8021q.h b/net/dsa/tag_8021q.h
index 0c6671d7c1c2..27b8906f99ec 100644
--- a/net/dsa/tag_8021q.h
+++ b/net/dsa/tag_8021q.h
@@ -16,8 +16,9 @@ struct sk_buff *dsa_8021q_xmit(struct sk_buff *skb, struct net_device *netdev,
void dsa_8021q_rcv(struct sk_buff *skb, int *source_port, int *switch_id,
int *vbid, int *vid);
-struct net_device *dsa_tag_8021q_find_port_by_vbid(struct net_device *conduit,
- int vbid);
+struct net_device *dsa_tag_8021q_find_user(struct net_device *conduit,
+ int source_port, int switch_id,
+ int vid, int vbid);
int dsa_switch_tag_8021q_vlan_add(struct dsa_switch *ds,
struct dsa_notifier_tag_8021q_vlan_info *info);
diff --git a/net/dsa/tag_sja1105.c b/net/dsa/tag_sja1105.c
index 35a6346549f2..3e902af7eea6 100644
--- a/net/dsa/tag_sja1105.c
+++ b/net/dsa/tag_sja1105.c
@@ -509,12 +509,8 @@ static struct sk_buff *sja1105_rcv(struct sk_buff *skb,
*/
return NULL;
- if (source_port != -1 && switch_id != -1)
- skb->dev = dsa_conduit_find_user(netdev, switch_id, source_port);
- else if (vbid >= 1)
- skb->dev = dsa_tag_8021q_find_port_by_vbid(netdev, vbid);
- else
- skb->dev = dsa_find_designated_bridge_port_by_vid(netdev, vid);
+ skb->dev = dsa_tag_8021q_find_user(netdev, source_port, switch_id,
+ vid, vbid);
if (!skb->dev) {
netdev_warn(netdev, "Couldn't decode source port\n");
return NULL;
@@ -652,12 +648,9 @@ static struct sk_buff *sja1110_rcv(struct sk_buff *skb,
if (likely(sja1105_skb_has_tag_8021q(skb)))
dsa_8021q_rcv(skb, &source_port, &switch_id, &vbid, &vid);
- if (source_port != -1 && switch_id != -1)
- skb->dev = dsa_conduit_find_user(netdev, switch_id, source_port);
- else if (vbid >= 1)
- skb->dev = dsa_tag_8021q_find_port_by_vbid(netdev, vbid);
- else
- skb->dev = dsa_find_designated_bridge_port_by_vid(netdev, vid);
+ skb->dev = dsa_tag_8021q_find_user(netdev, source_port, switch_id,
+ vid, vbid);
+
if (!skb->dev) {
netdev_warn(netdev, "Couldn't decode source port\n");
return NULL;
--
2.34.1
This patch is a simple implementation of 802.1q tagging in the vsc73xx
driver. Currently, devices with DSA_TAG_PROTO_NONE are not functional.
The VSC73XX family doesn't provide any tag support for external Ethernet
ports.
The only option available is VLAN-based tagging, which requires constant
hardware VLAN filtering. While the VSC73XX family supports provider
bridging, it only supports QinQ without full implementation of 802.1AD.
This means it only allows the doubled 0x8100 TPID.
In the simple port mode, QinQ is enabled to preserve forwarding of
VLAN-tagged frames.
Signed-off-by: Pawel Dembicki <[email protected]>
---
v8:
- resend only
v7:
- adjust tag8021q implementation for vlan filtering implementation
changes
v6:
- resend only
v5:
- improve commit message
v4:
- adjust tag8021q implementation for changed untagged vlan storage
- minor fixes
v3:
- Split tagger and tag implementation into separate commits
drivers/net/dsa/Kconfig | 2 +-
drivers/net/dsa/vitesse-vsc73xx-core.c | 54 +++++++++++++++++++++++++-
2 files changed, 53 insertions(+), 3 deletions(-)
diff --git a/drivers/net/dsa/Kconfig b/drivers/net/dsa/Kconfig
index 3092b391031a..22a04636d09e 100644
--- a/drivers/net/dsa/Kconfig
+++ b/drivers/net/dsa/Kconfig
@@ -126,7 +126,7 @@ config NET_DSA_SMSC_LAN9303_MDIO
config NET_DSA_VITESSE_VSC73XX
tristate
- select NET_DSA_TAG_NONE
+ select NET_DSA_TAG_VSC73XX_8021Q
select FIXED_PHY
select VITESSE_PHY
select GPIOLIB
diff --git a/drivers/net/dsa/vitesse-vsc73xx-core.c b/drivers/net/dsa/vitesse-vsc73xx-core.c
index 433ffd4f44fd..d0442ea8a610 100644
--- a/drivers/net/dsa/vitesse-vsc73xx-core.c
+++ b/drivers/net/dsa/vitesse-vsc73xx-core.c
@@ -597,7 +597,7 @@ static enum dsa_tag_protocol vsc73xx_get_tag_protocol(struct dsa_switch *ds,
* cannot access the tag. (See "Internal frame header" section
* 3.9.1 in the manual.)
*/
- return DSA_TAG_PROTO_NONE;
+ return DSA_TAG_PROTO_VSC73XX_8021Q;
}
static int vsc73xx_wait_for_vlan_table_cmd(struct vsc73xx *vsc)
@@ -687,7 +687,7 @@ vsc73xx_update_vlan_table(struct vsc73xx *vsc, int port, u16 vid, bool set)
static int vsc73xx_setup(struct dsa_switch *ds)
{
struct vsc73xx *vsc = ds->priv;
- int i;
+ int i, ret;
dev_info(vsc->dev, "set up the switch\n");
@@ -756,6 +756,12 @@ static int vsc73xx_setup(struct dsa_switch *ds)
mdelay(50);
+ rtnl_lock();
+ ret = dsa_tag_8021q_register(ds, htons(ETH_P_8021Q));
+ rtnl_unlock();
+ if (ret)
+ return ret;
+
/* Release reset from the internal PHYs */
vsc73xx_write(vsc, VSC73XX_BLOCK_SYSTEM, 0, VSC73XX_GLORESET,
VSC73XX_GLORESET_PHY_RESET);
@@ -1210,6 +1216,21 @@ static int vsc73xx_vlan_set_untagged_hw(struct vsc73xx *vsc, int port, u16 vid)
return vsc73xx_vlan_change_untagged_hw(vsc, port, vid, true);
}
+static int vsc73xx_vlan_change_untagged_tag_8021q(struct vsc73xx *vsc, int port,
+ u16 vid, bool valid,
+ bool operate_on_storage)
+{
+ struct vsc73xx_portinfo *portinfo = &vsc->portinfo[port];
+
+ portinfo->untagged_tag_8021q_configured = valid;
+ portinfo->untagged_tag_8021q = vid;
+
+ if (operate_on_storage)
+ return 0;
+
+ return vsc73xx_vlan_change_untagged_hw(vsc, port, vid, valid);
+}
+
static int
vsc73xx_vlan_change_pvid_hw(struct vsc73xx *vsc, int port, u16 vid, bool set)
{
@@ -1537,6 +1558,33 @@ static int vsc73xx_port_vlan_del(struct dsa_switch *ds, int port,
return 0;
}
+static int vsc73xx_tag_8021q_vlan_add(struct dsa_switch *ds, int port, u16 vid,
+ u16 flags)
+{
+ bool untagged = flags & BRIDGE_VLAN_INFO_UNTAGGED;
+ bool pvid = flags & BRIDGE_VLAN_INFO_PVID;
+ struct vsc73xx *vsc = ds->priv;
+ bool operate_on_storage;
+
+ operate_on_storage = !vsc73xx_tag_8021q_active(dsa_to_port(ds, port));
+
+ if (untagged)
+ vsc73xx_vlan_change_untagged_tag_8021q(vsc, port, vid, true,
+ operate_on_storage);
+ if (pvid)
+ vsc73xx_vlan_change_pvid(vsc, port, vid, true,
+ operate_on_storage, true);
+
+ return vsc73xx_update_vlan_table(vsc, port, vid, true);
+}
+
+static int vsc73xx_tag_8021q_vlan_del(struct dsa_switch *ds, int port, u16 vid)
+{
+ struct vsc73xx *vsc = ds->priv;
+
+ return vsc73xx_update_vlan_table(vsc, port, vid, false);
+}
+
static int vsc73xx_port_setup(struct dsa_switch *ds, int port)
{
struct vsc73xx_portinfo *portinfo;
@@ -1667,6 +1715,8 @@ static const struct dsa_switch_ops vsc73xx_ds_ops = {
.port_vlan_add = vsc73xx_port_vlan_add,
.port_vlan_del = vsc73xx_port_vlan_del,
.phylink_get_caps = vsc73xx_phylink_get_caps,
+ .tag_8021q_vlan_add = vsc73xx_tag_8021q_vlan_add,
+ .tag_8021q_vlan_del = vsc73xx_tag_8021q_vlan_del,
};
static int vsc73xx_gpio_get(struct gpio_chip *chip, unsigned int offset)
--
2.34.1
From: Vladimir Oltean <[email protected]>
Now that dsa_8021q_rcv() handles better the case where we don't
overwrite the precise source information if it comes from an external
(non-tag_8021q) source, we can now unify the call sequence between
sja1105_rcv() and sja1110_rcv().
This is a preparatory change for creating a higher-level wrapper for the
entire sequence which will live in tag_8021q.
Signed-off-by: Vladimir Oltean <[email protected]>
Signed-off-by: Pawel Dembicki <[email protected]>
---
v8,v7,v6,v5:
- resend only
v4:
- introduce patch and replace 'slave' with 'conduit' after rebase
net/dsa/tag_sja1105.c | 8 ++++----
1 file changed, 4 insertions(+), 4 deletions(-)
diff --git a/net/dsa/tag_sja1105.c b/net/dsa/tag_sja1105.c
index 7639ccb94d35..35a6346549f2 100644
--- a/net/dsa/tag_sja1105.c
+++ b/net/dsa/tag_sja1105.c
@@ -652,12 +652,12 @@ static struct sk_buff *sja1110_rcv(struct sk_buff *skb,
if (likely(sja1105_skb_has_tag_8021q(skb)))
dsa_8021q_rcv(skb, &source_port, &switch_id, &vbid, &vid);
- if (vbid >= 1)
+ if (source_port != -1 && switch_id != -1)
+ skb->dev = dsa_conduit_find_user(netdev, switch_id, source_port);
+ else if (vbid >= 1)
skb->dev = dsa_tag_8021q_find_port_by_vbid(netdev, vbid);
- else if (source_port == -1 || switch_id == -1)
- skb->dev = dsa_find_designated_bridge_port_by_vid(netdev, vid);
else
- skb->dev = dsa_conduit_find_user(netdev, switch_id, source_port);
+ skb->dev = dsa_find_designated_bridge_port_by_vid(netdev, vid);
if (!skb->dev) {
netdev_warn(netdev, "Couldn't decode source port\n");
return NULL;
--
2.34.1
On Wed, Apr 03, 2024 at 12:37:17PM +0200, Pawel Dembicki wrote:
> Switch the delay loop during the Arbiter empty check from
> vsc73xx_adjust_link() to use read_poll_timeout(). Functionally,
> one msleep() call is eliminated at the end of the loop in the timeout
> case.
>
> As Russell King suggested:
>
> "This [change] avoids the issue that on the last iteration, the code reads
> the register, tests it, finds the condition that's being waiting for is
> false, _then_ waits and end up printing the error message - that last
> wait is rather useless, and as the arbiter state isn't checked after
> waiting, it could be that we had success during the last wait."
>
> Suggested-by: Russell King <[email protected]>
> Reviewed-by: Andrew Lunn <[email protected]>
> Reviewed-by: Linus Walleij <[email protected]>
> Reviewed-by: Florian Fainelli <[email protected]>
> Signed-off-by: Pawel Dembicki <[email protected]>
> ---
Reviewed-by: Vladimir Oltean <[email protected]>
On Wed, Apr 03, 2024 at 12:37:21PM +0200, Pawel Dembicki wrote:
> This commit adds updates to the documentation describing the structures
> used in vsc73xx. This will help prevent kdoc-related issues in the future.
>
> Reviewed-by: Linus Walleij <[email protected]>
> Reviewed-by: Florian Fainelli <[email protected]>
> Signed-off-by: Pawel Dembicki <[email protected]>
> ---
Reviewed-by: Vladimir Oltean <[email protected]>
> diff --git a/drivers/net/dsa/vitesse-vsc73xx.h b/drivers/net/dsa/vitesse-vsc73xx.h
> index fee1378508b5..e7b08599a625 100644
> --- a/drivers/net/dsa/vitesse-vsc73xx.h
> +++ b/drivers/net/dsa/vitesse-vsc73xx.h
> @@ -15,7 +15,16 @@
> #define VSC73XX_MAX_NUM_PORTS 8
>
> /**
> - * struct vsc73xx - VSC73xx state container
> + * struct vsc73xx - VSC73xx state container: main data structure
> + * @dev: The device pointer
> + * @reset: The descriptor for the GPIO line tied to the reset pin
> + * @ds: Pointer to the DSA core structure
> + * @gc: Main structure of the GPIO controller
> + * @chipid: Storage for the Chip ID value read from the CHIPID register of the
> + * sswitch
Nitpick (not a request to resend): "switch"
> + * @addr: MAC address used in flow control frames
> + * @ops: Structure with hardware-dependent operations
> + * @priv: Pointer to the configuration interface structure
> */
> struct vsc73xx {
> struct device *dev;
On Wed, Apr 03, 2024 at 12:37:29PM +0200, Pawel Dembicki wrote:
> @@ -756,6 +756,12 @@ static int vsc73xx_setup(struct dsa_switch *ds)
>
> mdelay(50);
>
> + rtnl_lock();
> + ret = dsa_tag_8021q_register(ds, htons(ETH_P_8021Q));
> + rtnl_unlock();
> + if (ret)
> + return ret;
> +
No dsa_tag_8021q_unregister() in vsc73xx_teardown()?
Also, a bit odd placement, while the resets haven't all finished (this
falls in between the IP multicast flood mask and the PHY resets. It
makes future refactoring of the other code harder.
Additionally (and much worse), even later in vsc73xx_setup(), you clear
the VLAN table with portmaps of 0. Doesn't this wipe out the
configuration made by
dsa_tag_8021q_register()
-> dsa_tag_8021q_setup()
-> dsa_tag_8021q_port_setup()
-> dsa_port_tag_8021q_vlan_add()
-> dsa_switch_tag_8021q_vlan_add()
-> dsa_port_do_tag_8021q_vlan_add()
-> vsc73xx_tag_8021q_vlan_add()
-> vsc73xx_tag_8021q_vlan_add()
?
On Wed, Apr 03, 2024 at 12:37:31PM +0200, Pawel Dembicki wrote:
> This patch adds bridge support for vsc73xx driver.
> It introduce two functions for port_bridge_join and
> vsc73xx_port_bridge_leave handling.
>
> Those functions implement forwarding adjust and use
> dsa_tag_8021q_bridge_* api for adjust VLAN configuration.
>
> Reviewed-by: Linus Walleij <[email protected]>
> Signed-off-by: Pawel Dembicki <[email protected]>
> Reviewed-by: Vladimir Oltean <[email protected]>
> Reviewed-by: Florian Fainelli <[email protected]>
> ---
> diff --git a/drivers/net/dsa/vitesse-vsc73xx-core.c b/drivers/net/dsa/vitesse-vsc73xx-core.c
> index d0442ea8a610..87922804d1c1 100644
> --- a/drivers/net/dsa/vitesse-vsc73xx-core.c
> +++ b/drivers/net/dsa/vitesse-vsc73xx-core.c
> +static int vsc73xx_port_bridge_join(struct dsa_switch *ds, int port,
> + struct dsa_bridge bridge,
> + bool *tx_fwd_offload,
> + struct netlink_ext_ack *extack)
> +{
> + *tx_fwd_offload = true;
> +
> + return dsa_tag_8021q_bridge_join(ds, port, bridge);
> +}
> +
> +static void vsc73xx_port_bridge_leave(struct dsa_switch *ds, int port,
> + struct dsa_bridge bridge)
> +{
> + dsa_tag_8021q_bridge_leave(ds, port, bridge);
> +}
> +
> static int vsc73xx_port_setup(struct dsa_switch *ds, int port)
> {
> struct vsc73xx_portinfo *portinfo;
> @@ -1708,6 +1727,8 @@ static const struct dsa_switch_ops vsc73xx_ds_ops = {
> .port_setup = vsc73xx_port_setup,
> .port_enable = vsc73xx_port_enable,
> .port_disable = vsc73xx_port_disable,
> + .port_bridge_join = vsc73xx_port_bridge_join,
> + .port_bridge_leave = vsc73xx_port_bridge_leave,
> .port_change_mtu = vsc73xx_change_mtu,
> .port_max_mtu = vsc73xx_get_max_mtu,
> .port_stp_state_set = vsc73xx_port_stp_state_set,
> --
> 2.34.1
>
Making a mental note here, out loud, that after this patch set is
merged, "*tx_fwd_offload = true" should be absorbed into
dsa_tag_8021q_bridge_join(). Then, you could simply point
ds->ops->port_bridge_join() and ds->ops->port_bridge_leave() to the
generic tag_8021q methods, without the need for intermediary shims.
Looking at my upcoming schedule, I'll probably forget. Could you please
try to remember to do it?
On Wed, Apr 03, 2024 at 12:37:23PM +0200, Pawel Dembicki wrote:
> +static int
> +vsc73xx_port_vlan_filtering(struct dsa_switch *ds, int port,
> + bool vlan_filtering, struct netlink_ext_ack *extack)
> +{
> + enum vsc73xx_port_vlan_conf port_vlan_conf = VSC73XX_VLAN_IGNORE;
> + u16 vid_pvid = 0, vid_untagged = 0;
> + struct vsc73xx_portinfo *portinfo;
> + struct vsc73xx *vsc = ds->priv;
> + bool set_untagged = false;
> + bool set_pvid = false;
> +
> + portinfo = &vsc->portinfo[port];
> +
> + /* The swap processed below is required because vsc73xx is using
> + * tag_8021q. When vlan_filtering is disabled, tag_8021q uses
> + * pvid/untagged vlans for port recognition. The values configured for
> + * vlans and pvid/untagged states are stored in portinfo structure.
> + * When vlan_filtering is enabled, we need to restore pvid/untagged from
> + * portinfo structure. Analogic routine is processed when vlan_filtering
Analogous
> + * is disabled, but values used for tag_8021q are restored.
> + */
> + if (vlan_filtering) {
> + struct vsc73xx_vlan_summary summary;
> +
> + /* Use VLAN_N_VID to count all vlans */
> + vsc73xx_bridge_vlan_summary(vsc, port, &summary, VLAN_N_VID);
> +
> + port_vlan_conf = (summary.num_untagged > 1) ?
> + VSC73XX_VLAN_FILTER_UNTAG_ALL :
> + VSC73XX_VLAN_FILTER;
> +
> + if (summary.num_untagged == 1) {
> + vid_untagged = vsc73xx_find_first_vlan_untagged(vsc,
> + port);
> + set_untagged = true;
> + }
Is there really any functional difference between the above, vs this in
port_vlan_add():
port_vlan_conf = VSC73XX_VLAN_FILTER;
if (summary.num_tagged == 0 && untagged)
port_vlan_conf = VSC73XX_VLAN_FILTER_UNTAG_ALL;
vsc73xx_set_vlan_conf(vsc, port, port_vlan_conf);
if (port_vlan_conf == VSC73XX_VLAN_FILTER_UNTAG_ALL)
goto update_vlan_table;
vs this in port_vlan_del():
enum vsc73xx_port_vlan_conf port_vlan_conf =
VSC73XX_VLAN_FILTER;
if (summary.num_tagged == 0)
port_vlan_conf = VSC73XX_VLAN_FILTER_UNTAG_ALL;
vsc73xx_set_vlan_conf(vsc, port, port_vlan_conf);
if (summary.num_untagged <= 1) {
vid = vsc73xx_find_first_vlan_untagged(vsc, port);
vsc73xx_vlan_change_untagged_hw(vsc, port, vid,
summary.num_untagged);
}
?
If not, I have to agree with Florian that it's dizzying to follow the
vsc73xx_set_vlan_conf() calls, all with slightly different input.
Even worse now than before, I think.
I see you've changed the storage variable names, so maybe you are open
to some more refactoring, to make the code more maintainable.
I would suggest calling a single vsc73xx_update_vlan_conf() from all
places, which figures out what to do by itself, based on the current
(updated) state. All ifs and buts are on the inside, not on the outside.
Also, the operate_on_storage variable is not a great name now. Because
you always operate on storage. It means "operate _just_ on storage", aka
"don't commit to hardware".
I wanted to avoid requesting you to make more changes to the
implementation, but you started it, and now it's a confusing mix of
concepts from old and new patch sets.
I would like to also see the "operate_on_storage" logic rewritten and
embedded into those functions which must update shared resources (port
PVID, port untagged VID, etc).
Basically, in the end I would like the driver to be written in the style
of every other similar function: ocelot_port_manage_port_tag(),
sja1105_commit_pvid(), mv88e6xxx_commit_pvid() etc. No arguments, except
the port on which to run. Figure out what to do from the current state,
and make sure it is called from all places that change that state.
It also serves very well as self-documenting code, because "how register
X should be set" is centralized into a single function.
Sorry for making this request so late in the development process. I hope
it's not discouraging.
On Wed, Apr 03, 2024 at 12:37:18PM +0200, Pawel Dembicki wrote:
> This patch replaces the adjust_link api with the phylink apis that provide
> equivalent functionality.
>
> The remaining functionality from the adjust_link is now covered in the
> phylink_mac_link_* and phylink_mac_config.
>
> Removes:
> .adjust_link
> Adds:
> .phylink_mac_config
> .phylink_mac_link_up
> .phylink_mac_link_down
>
> Signed-off-by: Pawel Dembicki <[email protected]>
> Reviewed-by: Florian Fainelli <[email protected]>
> Reviewed-by: Linus Walleij <[email protected]>
> Reviewed-by: Russell King (Oracle) <[email protected]>
> ---
I think you should split out the phylink patches from VLAN work, since
these seem good to go.