| Message ID | 20230218230636.5528-1-luizluca@gmail.com (mailing list archive) |
|---|---|
| State | Deferred |
| Delegated to: | Netdev Maintainers |
| Headers | show |
| Series | [net-next,v3] net: dsa: realtek: rtl8365mb: add change_mtu | expand |
On Sat, 18 Feb 2023 20:06:37 -0300 Luiz Angelo Daros de Luca wrote: > The rtl8365mb was using a fixed MTU size of 1536, which was probably > inspired by the rtl8366rb's initial packet size. However, unlike that > family, the rtl8365mb family can specify the max packet size in bytes, > rather than in fixed steps. The max packet size now defaults to > VLAN_ETH_HLEN+ETH_DATA_LEN+ETH_FCS_LEN, which is 1522 bytes. > > DSA calls change_mtu for the CPU port once the max MTU value among the > ports changes. As the max packet size is defined globally, the switch > is configured only when the call affects the CPU port. > > The available specifications do not directly define the max supported > packet size, but it mentions a 16k limit. This driver will use the 0x3FFF > limit as it is used in the vendor API code. However, the switch sets the > max packet size to 16368 bytes (0x3FF0) after it resets. > > change_mtu uses MTU size, or ethernet payload size, while the switch > works with frame size. The frame size is calculated considering the > ethernet header (14 bytes), a possible 802.1Q tag (4 bytes), the payload > size (MTU), and the Ethernet FCS (4 bytes). The CPU tag (8 bytes) is > consumed before the switch enforces the limit. > > MTU was tested up to 2018 (with 802.1Q) as that is as far as mt7620 > (where rtl8367s is stacked) can go. The register was manually > manipulated byte-by-byte to ensure the MTU to frame size conversion was > correct. For frames without 802.1Q tag, the frame size limit will be 4 > bytes over the required size. > > There is a jumbo register, enabled by default at 6k packet size. > However, the jumbo settings do not seem to limit nor expand the maximum > tested MTU (2018), even when jumbo is disabled. More tests are needed > with a device that can handle larger frames. # Form letter - net-next is closed The merge window for v6.3 has begun and therefore net-next is closed for new drivers, features, code refactoring and optimizations. We are currently accepting bug fixes only. Please repost when net-next reopens after Mar 6th. RFC patches sent for review only are obviously welcome at any time.
On Sat, Feb 18, 2023 at 08:06:37PM -0300, Luiz Angelo Daros de Luca wrote: > The rtl8365mb was using a fixed MTU size of 1536, which was probably > inspired by the rtl8366rb's initial packet size. However, unlike that > family, the rtl8365mb family can specify the max packet size in bytes, > rather than in fixed steps. The max packet size now defaults to > VLAN_ETH_HLEN+ETH_DATA_LEN+ETH_FCS_LEN, which is 1522 bytes. > > DSA calls change_mtu for the CPU port once the max MTU value among the > ports changes. As the max packet size is defined globally, the switch > is configured only when the call affects the CPU port. > > The available specifications do not directly define the max supported > packet size, but it mentions a 16k limit. This driver will use the 0x3FFF > limit as it is used in the vendor API code. However, the switch sets the > max packet size to 16368 bytes (0x3FF0) after it resets. > > change_mtu uses MTU size, or ethernet payload size, while the switch > works with frame size. The frame size is calculated considering the > ethernet header (14 bytes), a possible 802.1Q tag (4 bytes), the payload > size (MTU), and the Ethernet FCS (4 bytes). The CPU tag (8 bytes) is > consumed before the switch enforces the limit. > > MTU was tested up to 2018 (with 802.1Q) as that is as far as mt7620 > (where rtl8367s is stacked) can go. The register was manually > manipulated byte-by-byte to ensure the MTU to frame size conversion was > correct. For frames without 802.1Q tag, the frame size limit will be 4 > bytes over the required size. > > There is a jumbo register, enabled by default at 6k packet size. > However, the jumbo settings do not seem to limit nor expand the maximum > tested MTU (2018), even when jumbo is disabled. More tests are needed > with a device that can handle larger frames. > > Signed-off-by: Luiz Angelo Daros de Luca <luizluca@gmail.com> > --- > > v2->v3: > - changed max frame size to 0x3FFF (used by vendor API) > - added info about how frame size is calculated, some more description > about the tests performed and the 4 extra bytes when untagged frame is > used. > > v1->v2: > - dropped jumbo code as it was not changing the behavior (up to 2k MTU) > - fixed typos > - fixed code alignment > - renamed rtl8365mb_(change|max)_mtu to rtl8365mb_port_(change|max)_mtu > > drivers/net/dsa/realtek/rtl8365mb.c | 41 ++++++++++++++++++++++++++--- > 1 file changed, 37 insertions(+), 4 deletions(-) > > diff --git a/drivers/net/dsa/realtek/rtl8365mb.c b/drivers/net/dsa/realtek/rtl8365mb.c > index da31d8b839ac..05642c617c3b 100644 > --- a/drivers/net/dsa/realtek/rtl8365mb.c > +++ b/drivers/net/dsa/realtek/rtl8365mb.c > @@ -98,6 +98,7 @@ > #include <linux/of_irq.h> > #include <linux/regmap.h> > #include <linux/if_bridge.h> > +#include <linux/if_vlan.h> > > #include "realtek.h" > > @@ -267,6 +268,7 @@ > /* Maximum packet length register */ > #define RTL8365MB_CFG0_MAX_LEN_REG 0x088C > #define RTL8365MB_CFG0_MAX_LEN_MASK 0x3FFF > +#define RTL8365MB_CFG0_MAX_LEN_MAX 0x3FFF > > /* Port learning limit registers */ > #define RTL8365MB_LUT_PORT_LEARN_LIMIT_BASE 0x0A20 > @@ -1135,6 +1137,36 @@ static void rtl8365mb_phylink_mac_link_up(struct dsa_switch *ds, int port, > } > } > > +static int rtl8365mb_port_change_mtu(struct dsa_switch *ds, int port, > + int new_mtu) > +{ > + struct realtek_priv *priv = ds->priv; > + int frame_size; > + > + /* When a new MTU is set, DSA always sets the CPU port's MTU to the > + * largest MTU of the slave ports. Because the switch only has a global > + * RX length register, only allowing CPU port here is enough. > + */ > + You forgot to remove this spurious newline. With that fixed you can add: Reviewed-by: Alvin Šipraga <alsi@bang-olufsen.dk> Thanks! > + if (!dsa_is_cpu_port(ds, port)) > + return 0; > + > + frame_size = new_mtu + VLAN_ETH_HLEN + ETH_FCS_LEN; > + > + dev_dbg(priv->dev, "changing mtu to %d (frame size: %d)\n", > + new_mtu, frame_size); > + > + return regmap_update_bits(priv->map, RTL8365MB_CFG0_MAX_LEN_REG, > + RTL8365MB_CFG0_MAX_LEN_MASK, > + FIELD_PREP(RTL8365MB_CFG0_MAX_LEN_MASK, > + frame_size)); > +} > + > +static int rtl8365mb_port_max_mtu(struct dsa_switch *ds, int port) > +{ > + return RTL8365MB_CFG0_MAX_LEN_MAX - VLAN_ETH_HLEN - ETH_FCS_LEN; > +} > + > static void rtl8365mb_port_stp_state_set(struct dsa_switch *ds, int port, > u8 state) > { > @@ -1980,10 +2012,7 @@ static int rtl8365mb_setup(struct dsa_switch *ds) > p->index = i; > } > > - /* Set maximum packet length to 1536 bytes */ > - ret = regmap_update_bits(priv->map, RTL8365MB_CFG0_MAX_LEN_REG, > - RTL8365MB_CFG0_MAX_LEN_MASK, > - FIELD_PREP(RTL8365MB_CFG0_MAX_LEN_MASK, 1536)); > + ret = rtl8365mb_port_change_mtu(ds, cpu->trap_port, ETH_DATA_LEN); > if (ret) > goto out_teardown_irq; > > @@ -2103,6 +2132,8 @@ static const struct dsa_switch_ops rtl8365mb_switch_ops_smi = { > .get_eth_mac_stats = rtl8365mb_get_mac_stats, > .get_eth_ctrl_stats = rtl8365mb_get_ctrl_stats, > .get_stats64 = rtl8365mb_get_stats64, > + .port_change_mtu = rtl8365mb_port_change_mtu, > + .port_max_mtu = rtl8365mb_port_max_mtu, > }; > > static const struct dsa_switch_ops rtl8365mb_switch_ops_mdio = { > @@ -2124,6 +2155,8 @@ static const struct dsa_switch_ops rtl8365mb_switch_ops_mdio = { > .get_eth_mac_stats = rtl8365mb_get_mac_stats, > .get_eth_ctrl_stats = rtl8365mb_get_ctrl_stats, > .get_stats64 = rtl8365mb_get_stats64, > + .port_change_mtu = rtl8365mb_port_change_mtu, > + .port_max_mtu = rtl8365mb_port_max_mtu, > }; > > static const struct realtek_ops rtl8365mb_ops = { > -- > 2.39.1 >
diff --git a/drivers/net/dsa/realtek/rtl8365mb.c b/drivers/net/dsa/realtek/rtl8365mb.c index da31d8b839ac..05642c617c3b 100644 --- a/drivers/net/dsa/realtek/rtl8365mb.c +++ b/drivers/net/dsa/realtek/rtl8365mb.c @@ -98,6 +98,7 @@ #include <linux/of_irq.h> #include <linux/regmap.h> #include <linux/if_bridge.h> +#include <linux/if_vlan.h> #include "realtek.h" @@ -267,6 +268,7 @@ /* Maximum packet length register */ #define RTL8365MB_CFG0_MAX_LEN_REG 0x088C #define RTL8365MB_CFG0_MAX_LEN_MASK 0x3FFF +#define RTL8365MB_CFG0_MAX_LEN_MAX 0x3FFF /* Port learning limit registers */ #define RTL8365MB_LUT_PORT_LEARN_LIMIT_BASE 0x0A20 @@ -1135,6 +1137,36 @@ static void rtl8365mb_phylink_mac_link_up(struct dsa_switch *ds, int port, } } +static int rtl8365mb_port_change_mtu(struct dsa_switch *ds, int port, + int new_mtu) +{ + struct realtek_priv *priv = ds->priv; + int frame_size; + + /* When a new MTU is set, DSA always sets the CPU port's MTU to the + * largest MTU of the slave ports. Because the switch only has a global + * RX length register, only allowing CPU port here is enough. + */ + + if (!dsa_is_cpu_port(ds, port)) + return 0; + + frame_size = new_mtu + VLAN_ETH_HLEN + ETH_FCS_LEN; + + dev_dbg(priv->dev, "changing mtu to %d (frame size: %d)\n", + new_mtu, frame_size); + + return regmap_update_bits(priv->map, RTL8365MB_CFG0_MAX_LEN_REG, + RTL8365MB_CFG0_MAX_LEN_MASK, + FIELD_PREP(RTL8365MB_CFG0_MAX_LEN_MASK, + frame_size)); +} + +static int rtl8365mb_port_max_mtu(struct dsa_switch *ds, int port) +{ + return RTL8365MB_CFG0_MAX_LEN_MAX - VLAN_ETH_HLEN - ETH_FCS_LEN; +} + static void rtl8365mb_port_stp_state_set(struct dsa_switch *ds, int port, u8 state) { @@ -1980,10 +2012,7 @@ static int rtl8365mb_setup(struct dsa_switch *ds) p->index = i; } - /* Set maximum packet length to 1536 bytes */ - ret = regmap_update_bits(priv->map, RTL8365MB_CFG0_MAX_LEN_REG, - RTL8365MB_CFG0_MAX_LEN_MASK, - FIELD_PREP(RTL8365MB_CFG0_MAX_LEN_MASK, 1536)); + ret = rtl8365mb_port_change_mtu(ds, cpu->trap_port, ETH_DATA_LEN); if (ret) goto out_teardown_irq; @@ -2103,6 +2132,8 @@ static const struct dsa_switch_ops rtl8365mb_switch_ops_smi = { .get_eth_mac_stats = rtl8365mb_get_mac_stats, .get_eth_ctrl_stats = rtl8365mb_get_ctrl_stats, .get_stats64 = rtl8365mb_get_stats64, + .port_change_mtu = rtl8365mb_port_change_mtu, + .port_max_mtu = rtl8365mb_port_max_mtu, }; static const struct dsa_switch_ops rtl8365mb_switch_ops_mdio = { @@ -2124,6 +2155,8 @@ static const struct dsa_switch_ops rtl8365mb_switch_ops_mdio = { .get_eth_mac_stats = rtl8365mb_get_mac_stats, .get_eth_ctrl_stats = rtl8365mb_get_ctrl_stats, .get_stats64 = rtl8365mb_get_stats64, + .port_change_mtu = rtl8365mb_port_change_mtu, + .port_max_mtu = rtl8365mb_port_max_mtu, }; static const struct realtek_ops rtl8365mb_ops = {
The rtl8365mb was using a fixed MTU size of 1536, which was probably inspired by the rtl8366rb's initial packet size. However, unlike that family, the rtl8365mb family can specify the max packet size in bytes, rather than in fixed steps. The max packet size now defaults to VLAN_ETH_HLEN+ETH_DATA_LEN+ETH_FCS_LEN, which is 1522 bytes. DSA calls change_mtu for the CPU port once the max MTU value among the ports changes. As the max packet size is defined globally, the switch is configured only when the call affects the CPU port. The available specifications do not directly define the max supported packet size, but it mentions a 16k limit. This driver will use the 0x3FFF limit as it is used in the vendor API code. However, the switch sets the max packet size to 16368 bytes (0x3FF0) after it resets. change_mtu uses MTU size, or ethernet payload size, while the switch works with frame size. The frame size is calculated considering the ethernet header (14 bytes), a possible 802.1Q tag (4 bytes), the payload size (MTU), and the Ethernet FCS (4 bytes). The CPU tag (8 bytes) is consumed before the switch enforces the limit. MTU was tested up to 2018 (with 802.1Q) as that is as far as mt7620 (where rtl8367s is stacked) can go. The register was manually manipulated byte-by-byte to ensure the MTU to frame size conversion was correct. For frames without 802.1Q tag, the frame size limit will be 4 bytes over the required size. There is a jumbo register, enabled by default at 6k packet size. However, the jumbo settings do not seem to limit nor expand the maximum tested MTU (2018), even when jumbo is disabled. More tests are needed with a device that can handle larger frames. Signed-off-by: Luiz Angelo Daros de Luca <luizluca@gmail.com> --- v2->v3: - changed max frame size to 0x3FFF (used by vendor API) - added info about how frame size is calculated, some more description about the tests performed and the 4 extra bytes when untagged frame is used. v1->v2: - dropped jumbo code as it was not changing the behavior (up to 2k MTU) - fixed typos - fixed code alignment - renamed rtl8365mb_(change|max)_mtu to rtl8365mb_port_(change|max)_mtu drivers/net/dsa/realtek/rtl8365mb.c | 41 ++++++++++++++++++++++++++--- 1 file changed, 37 insertions(+), 4 deletions(-)