Phylink PCS updates
Subsystems (4)
Russell King - ARM Linux admin
maintainercover Phylink PCS updates
Hi,
This series updates the rudimentary phylink PCS support with the results of the last four months of development of that. Phylink PCS support was initially added back at the end of March, when it became clear that the current approach of treating everything at the MAC end as being part of the MAC was inadequate.
However, this rudimentary implementation was fine initially for mvneta and similar, but in practice had a fair number of issues, particularly when ethtool interfaces were used to change various link properties.
It became apparent that relying on the phylink_config structure for the PCS was also bad when it became clear that the same PCS was used in DSA drivers as well as in NXPs other offerings, and there was a desire to re-use that code.
It also became apparent that splitting the "configuration" step on an interface mode configuration between the MAC and PCS using just mac_config() and pcs_config() methods was not sufficient for some setups, as the MAC needed to be "taken down" prior to making changes, and once all settings were complete, the MAC could only then be resumed.
This series addresses these points, progressing PCS support, and has been developed with mvneta and DPAA2 setups, with work on both those drivers to prove this approach. It has been rigorously tested with mvneta, as that provides the most flexibility for testing the various code paths.
To solve the phylink_config reuse problem, we introduce a struct phylink_pcs, which contains the minimal information necessary, and it is intended that this is embedded in the PCS private data structure.
To solve the interface mode configuration problem, we introduce two new MAC methods, mac_prepare() and mac_finish() which wrap the entire interface mode configuration only. This has the additional benefit of relieving MAC drivers from working out whether an interface change has occurred, and whether they need to do some major work.
I have not yet updated all the interface documentation for these changes yet, that work remains, but this patch set is provided in the hope that those working on PCS support in NXP will find this useful.
Since there is a lot of change here, this is the reason why I strongly advise that everyone has converted to the mac_link_up() way of configuring the link parameters when the link comes up, rather than the old way of using mac_config() - especially as splitting the PCS changes how and when phylink calls mac_config(). Although no change for existing users is intended, that is something I no longer am able to test.
Changes since RFC:
- fix bisect build failure
- add patch to use config.an_enabled
- rename phylink_config_interface to phylink_major_reconfig
- add expanded documentation for phylink_set_pcs()
drivers/net/phy/phylink.c | 372 ++++++++++++++++++++++++++++++++-------------- include/linux/phylink.h | 103 ++++++++++--- 2 files changed, 344 insertions(+), 131 deletions(-)
Signature · 2 lines
RMK's Patch system: https://www.armlinux.org.uk/developer/patches/ FTTP is here! 40Mbps down 10Mbps up. Decent connectivity at last!
PATCH 1/14 net: phylink: update ethtool reporting for fixed-link modes
Comparing the ethtool output from phylink and non-phylink fixed-link setups shows that we have some differences:
- The "auto-negotiation" fields are different; phylink reports these as "No", non-phylink reports these as "Yes" for the supported and advertising masks.
- The link partner advertisement is set to the link speed with non- phylink, but phylink leaves this unset, causing all link partner fields to be omitted.
The phylink ethtool output also disagrees with the software emulated PHY dump via the MII registers.
Update the phylink fixed-link parsing code so that we better reflect the behaviour of the non-phylink code that this facility replaces, and bring the ethtool interface more into line with the report from via the MII interface.
--- drivers/net/phy/phylink.c | 2 ++ 1 file changed, 2 insertions(+)
drivers/net/phy/phylink.c +2 -0
@@ -241,8 +241,10 @@ static int phylink_parse_fixedlink(struct phylink *pl,phylink_set (pl->supported,MII );phylink_set (pl->supported, Pause);phylink_set (pl->supported, Asym_Pause); +phylink_set (pl->supported, Autoneg);if (s) {__set_bit (s->bit, pl->supported); +__set_bit (s->bit, pl->link_config.lp_advertising); }else { phylink_warn(pl, "fixed link %s duplex %dMbps not recognised\n", pl->link_config.duplex == DUPLEX_FULL ? "full" : "half",
Signature · 1 line
2.20.1
PATCH 2/14 net: phylink: rejig link state tracking
In reply to Russell King - ARM Linux adminRejig the link state tracking, so that we can use the current state in a future patch.
--- drivers/net/phy/phylink.c | 14 +++++++------- 1 file changed, 7 insertions(+), 7 deletions(-)
drivers/net/phy/phylink.c +7 -7
@@ -578,9 +578,14 @@ static void phylink_resolve(struct work_struct *w) struct phylink *pl = container_of(w, struct phylink, resolve);struct phylink_link_state link_state;struct net_device *ndev = pl->netdev; -int link_changed; +bool cur_link_state;mutex_lock (&pl->state_mutex); +if (pl->netdev) + cur_link_state =netif_carrier_ok (ndev); +else + cur_link_state = pl->old_link_state; +if (pl->phylink_disable_state) { pl->mac_link_dropped = false; link_state.link = false;
@@ -623,12 +628,7 @@ static void phylink_resolve(struct work_struct *w) } } -if (pl->netdev) - link_changed = (link_state.link !=netif_carrier_ok (ndev)); -else - link_changed = (link_state.link != pl->old_link_state); - -if (link_changed) { +if (link_state.link != cur_link_state) { pl->old_link_state = link_state.link;if (!link_state.link)phylink_link_down (pl);
Signature · 1 line
2.20.1
PATCH 3/14 net: phylink: rearrange resolve mac_config() call
In reply to Russell King - ARM Linux adminUse a boolean to indicate whether mac_config() should be called during a resolution. This allows resolution to have a single location where mac_config() will be called, which will allow us to make decisions about how and what we do.
--- drivers/net/phy/phylink.c | 21 ++++++++------------- 1 file changed, 8 insertions(+), 13 deletions(-)
drivers/net/phy/phylink.c +8 -13
@@ -421,13 +421,6 @@ static void phylink_mac_config(struct phylink *pl, pl->mac_ops->mac_config (pl->config, pl->cur_link_an_mode, state); } -static void phylink_mac_config_up (struct phylink *pl, -const struct phylink_link_state *state) -{ -if (state->link) -phylink_mac_config (pl, state); -} -static void phylink_mac_pcs_an_restart (struct phylink *pl) { if (pl->link_config.an_enabled &&
@@ -578,6 +571,7 @@ static void phylink_resolve(struct work_struct *w) struct phylink *pl = container_of(w, struct phylink, resolve);struct phylink_link_state link_state;struct net_device *ndev = pl->netdev; +bool mac_config = false;bool cur_link_state;mutex_lock (&pl->state_mutex);
@@ -596,12 +590,12 @@ static void phylink_resolve(struct work_struct *w)case MLO_AN_PHY : link_state = pl->phy_state;phylink_apply_manual_flow (pl, &link_state); -phylink_mac_config_up (pl, &link_state); + mac_config = link_state.link;break ;case MLO_AN_FIXED :phylink_get_fixed_state (pl, &link_state); -phylink_mac_config_up (pl, &link_state); + mac_config = link_state.link;break ;case MLO_AN_INBAND :
@@ -619,15 +613,16 @@ static void phylink_resolve(struct work_struct *w)/* If we have a PHY, we need to update with * the PHY flow control bits. */ link_state.pause = pl->phy_state.pause; -phylink_apply_manual_flow (pl, &link_state); -phylink_mac_config (pl, &link_state); - }else { -phylink_apply_manual_flow (pl, &link_state); + mac_config = true; } +phylink_apply_manual_flow (pl, &link_state);break ; } } +if (mac_config) +phylink_mac_config (pl, &link_state); +if (link_state.link != cur_link_state) { pl->old_link_state = link_state.link;if (!link_state.link)
Signature · 1 line
2.20.1
PATCH 4/14 net: phylink: ensure link is down when changing interface
In reply to Russell King - ARM Linux adminThe only PHYs that are used with phylink which change their interface are the BCM84881 and MV88X3310 family, both of which only change their interface modes on link-up events. However, rather than relying upon this behaviour by the PHY, we should give a stronger guarantee when resolving that the link will be down whenever we change the interface mode. This patch implements that stronger guarantee for resolve.
--- drivers/net/phy/phylink.c | 12 +++++++++++- 1 file changed, 11 insertions(+), 1 deletion(-)
drivers/net/phy/phylink.c +11 -1
@@ -620,8 +620,18 @@ static void phylink_resolve(struct work_struct *w) } } -if (mac_config) +if (mac_config) { +if (link_state.interface != pl->link_config.interface) { +/* The interface has changed, force the link down and +* then reconfigure. +*/ +if (cur_link_state) { +phylink_link_down (pl); + cur_link_state = false; + } + }phylink_mac_config (pl, &link_state); + }if (link_state.link != cur_link_state) { pl->old_link_state = link_state.link;
Signature · 1 line
2.20.1
PATCH 5/14 net: phylink: update PCS when changing interface during resolution
In reply to Russell King - ARM Linux adminThe only PHYs that are used with phylink which change their interface are the BCM84881 and MV88X3310 family, both of which only change their interface modes on link-up events. This will break when drivers are converted to split-PCS. Fix this.
--- drivers/net/phy/phylink.c | 9 ++++++++- 1 file changed, 8 insertions(+), 1 deletion(-)
drivers/net/phy/phylink.c +8 -1
@@ -629,8 +629,15 @@ static void phylink_resolve(struct work_struct *w)phylink_link_down (pl); cur_link_state = false; } +phylink_pcs_config (pl, false, &link_state); + pl->link_config.interface = link_state.interface; + }else { +/* The interface remains unchanged, only the speed, +* duplex or pause settings have changed. Call the +* old mac_config() method to configure the MAC/PCS. +*/ +phylink_mac_config (pl, &link_state); } -phylink_mac_config (pl, &link_state); }if (link_state.link != cur_link_state) {
Signature · 1 line
2.20.1
PATCH 6/14 net: phylink: avoid mac_config calls
In reply to Russell King - ARM Linux adminAvoid calling mac_config() when using split PCS, and the interface remains the same.
--- drivers/net/phy/phylink.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-)
drivers/net/phy/phylink.c +4 -2
@@ -631,10 +631,12 @@ static void phylink_resolve(struct work_struct *w) }phylink_pcs_config (pl, false, &link_state); pl->link_config.interface = link_state.interface; - }else { + }else if (!pl->pcs_ops) {/* The interface remains unchanged, only the speed, * duplex or pause settings have changed. Call the -* old mac_config() method to configure the MAC/PCS. +* old mac_config() method to configure the MAC/PCS +* only if we do not have a PCS installed (an +* unconverted user.) */ phylink_mac_config (pl, &link_state); }
Signature · 1 line
2.20.1
PATCH 7/14 net: phylink: simplify ksettings_set() implementation
In reply to Russell King - ARM Linux adminSimplify the ksettings_set() implementation to look more like phylib's implementation; use a switch() for validating the autoneg setting, and use the linkmode_modify() helper to set the autoneg bit in the advertisement mask.
--- drivers/net/phy/phylink.c | 25 ++++++++++++------------- 1 file changed, 12 insertions(+), 13 deletions(-)
drivers/net/phy/phylink.c +12 -13
@@ -1314,25 +1314,24 @@ int phylink_ethtool_ksettings_set(struct phylink *pl,__ETHTOOL_DECLARE_LINK_MODE_MASK (support);struct ethtool_link_ksettings our_kset;struct phylink_link_state config; +const struct phy_setting *s;int ret;ASSERT_RTNL (); -if (kset->base.autoneg !=AUTONEG_DISABLE && - kset->base.autoneg !=AUTONEG_ENABLE ) -return -EINVAL ; -linkmode_copy (support, pl->supported); config = pl->link_config; + config.an_enabled = kset->base.autoneg ==AUTONEG_ENABLE ; -/* Mask out unsupported advertisements */ +/* Mask out unsupported advertisements, and force the autoneg bit */ linkmode_and (config.advertising, kset->link_modes.advertising, support); +linkmode_mod_bit (ETHTOOL_LINK_MODE_Autoneg_BIT, config.advertising, + config.an_enabled);/* FIXME: should we reject autoneg if phy/mac does not support it? */ -if (kset->base.autoneg ==AUTONEG_DISABLE ) { -const struct phy_setting *s; - +switch (kset->base.autoneg) { +case AUTONEG_DISABLE :/* Autonegotiation disabled, select a suitable speed and * duplex. */
@@ -1351,19 +1350,19 @@ int phylink_ethtool_ksettings_set(struct phylink *pl, config.speed = s->speed; config.duplex = s->duplex; - config.an_enabled = false; +break ; -__clear_bit (ETHTOOL_LINK_MODE_Autoneg_BIT, config.advertising); - }else { +case AUTONEG_ENABLE :/* If we have a fixed link, refuse to enable autonegotiation */ if (pl->cur_link_an_mode ==MLO_AN_FIXED )return -EINVAL ; config.speed =SPEED_UNKNOWN ; config.duplex =DUPLEX_UNKNOWN ; - config.an_enabled = true; +break ; -__set_bit (ETHTOOL_LINK_MODE_Autoneg_BIT, config.advertising); +default : +return -EINVAL ; }if (pl->phydev) {
Signature · 1 line
2.20.1
PATCH 8/14 net: phylink: simplify phy case for ksettings_set method
In reply to Russell King - ARM Linux adminWhen we have a PHY attached, an ethtool ksettings_set() call only really needs to call through to the phylib equivalent; phylib will call back to us when the link changes so we can update our state. Therefore, we can bypass most of our ksettings_set() call for this case.
--- drivers/net/phy/phylink.c | 104 +++++++++++++++++--------------------- 1 file changed, 47 insertions(+), 57 deletions(-)
drivers/net/phy/phylink.c +47 -57
@@ -1312,13 +1312,33 @@ int phylink_ethtool_ksettings_set(struct phylink *pl, const struct ethtool_link_ksettings *kset) {__ETHTOOL_DECLARE_LINK_MODE_MASK (support); -struct ethtool_link_ksettings our_kset;struct phylink_link_state config;const struct phy_setting *s; -int ret;ASSERT_RTNL (); +if (pl->phydev) { +/* We can rely on phylib for this update; we also do not need +* to update the pl->link_config settings: +* - the configuration returned via ksettings_get() will come +* from phylib whenever a PHY is present. +* - link_config.interface will be updated by the PHY calling +* back via phylink_phy_change() and a subsequent resolve. +* - initial link configuration for PHY mode comes from the +* last phy state updated via phylink_phy_change(). +* - other configuration changes (e.g. pause modes) are +* performed directly via phylib. +* - if in in-band mode with a PHY, the link configuration +* is passed on the link from the PHY, and all of +* link_config.{speed,duplex,an_enabled,pause} are not used. +* - the only possible use would be link_config.advertising +* pause modes when in 1000base-X mode with a PHY, but in +* the presence of a PHY, this should not be changed as that +* should be determined from the media side advertisement. +*/ +return phy_ethtool_ksettings_set (pl->phydev, kset); + } +linkmode_copy (support, pl->supported); config = pl->link_config; config.an_enabled = kset->base.autoneg ==AUTONEG_ENABLE ;
@@ -1365,65 +1385,35 @@ int phylink_ethtool_ksettings_set(struct phylink *pl,return -EINVAL ; } -if (pl->phydev) { -/* If we have a PHY, we process the kset change via phylib. -* phylib will call our link state function if the PHY -* parameters have changed, which will trigger a resolve -* and update the MAC configuration. -*/ - our_kset = *kset; -linkmode_copy (our_kset.link_modes.advertising, - config.advertising); - our_kset.base.speed = config.speed; - our_kset.base.duplex = config.duplex; +/* For a fixed link, this isn't able to change any parameters, +* which just leaves inband mode. +*/ +if (phylink_validate (pl, support, &config)) +return -EINVAL ; - ret =phy_ethtool_ksettings_set (pl->phydev, &our_kset); -if (ret) -return ret; +/* If autonegotiation is enabled, we must have an advertisement */ +if (config.an_enabled &&phylink_is_empty_linkmode (config.advertising)) +return -EINVAL ; -mutex_lock (&pl->state_mutex); -/* Save the new configuration */ -linkmode_copy (pl->link_config.advertising, - our_kset.link_modes.advertising); - pl->link_config.interface = config.interface; - pl->link_config.speed = our_kset.base.speed; - pl->link_config.duplex = our_kset.base.duplex; - pl->link_config.an_enabled = our_kset.base.autoneg != -AUTONEG_DISABLE ; -mutex_unlock (&pl->state_mutex); - }else { -/* For a fixed link, this isn't able to change any parameters, -* which just leaves inband mode. +mutex_lock (&pl->state_mutex); +linkmode_copy (pl->link_config.advertising, config.advertising); + pl->link_config.interface = config.interface; + pl->link_config.speed = config.speed; + pl->link_config.duplex = config.duplex; + pl->link_config.an_enabled = kset->base.autoneg != +AUTONEG_DISABLE ; + +if (pl->cur_link_an_mode ==MLO_AN_INBAND && + !test_bit (PHYLINK_DISABLE_STOPPED , &pl->phylink_disable_state)) { +/* If in 802.3z mode, this updates the advertisement. +* +* If we are in SGMII mode without a PHY, there is no +* advertisement; the only thing we have is the pause +* modes which can only come from a PHY. */ -if (phylink_validate (pl, support, &config)) -return -EINVAL ; - -/* If autonegotiation is enabled, we must have an advertisement */ -if (config.an_enabled && -phylink_is_empty_linkmode (config.advertising)) -return -EINVAL ; - -mutex_lock (&pl->state_mutex); -linkmode_copy (pl->link_config.advertising, config.advertising); - pl->link_config.interface = config.interface; - pl->link_config.speed = config.speed; - pl->link_config.duplex = config.duplex; - pl->link_config.an_enabled = kset->base.autoneg != -AUTONEG_DISABLE ; - -if (pl->cur_link_an_mode ==MLO_AN_INBAND && - !test_bit (PHYLINK_DISABLE_STOPPED , - &pl->phylink_disable_state)) { -/* If in 802.3z mode, this updates the advertisement. -* -* If we are in SGMII mode without a PHY, there is no -* advertisement; the only thing we have is the pause -* modes which can only come from a PHY. -*/ -phylink_pcs_config (pl, true, &pl->link_config); - } -mutex_unlock (&pl->state_mutex); +phylink_pcs_config (pl, true, &pl->link_config); } +mutex_unlock (&pl->state_mutex);return 0 ; }
Signature · 1 line
2.20.1
PATCH 9/14 net: phylink: use config.an_enabled in ksettings_set method
In reply to Russell King - ARM Linux adminRather than recomputing whether AN is enabled, use config.an_enabled.
--- drivers/net/phy/phylink.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-)
drivers/net/phy/phylink.c +1 -2
@@ -1400,8 +1400,7 @@ int phylink_ethtool_ksettings_set(struct phylink *pl, pl->link_config.interface = config.interface; pl->link_config.speed = config.speed; pl->link_config.duplex = config.duplex; - pl->link_config.an_enabled = kset->base.autoneg != -AUTONEG_DISABLE ; + pl->link_config.an_enabled = config.an_enabled;if (pl->cur_link_an_mode ==MLO_AN_INBAND && !test_bit (PHYLINK_DISABLE_STOPPED , &pl->phylink_disable_state)) {
Signature · 1 line
2.20.1
PATCH 10/14 net: phylink: simplify fixed-link case for ksettings_set method
In reply to Russell King - ARM Linux adminFor fixed links, we only allow the current settings, so this should be a matter of merely rejecting an attempt to change the settings. If the settings agree, then there is nothing more we need to do.
--- drivers/net/phy/phylink.c | 31 ++++++++++++++++++++----------- 1 file changed, 20 insertions(+), 11 deletions(-)
drivers/net/phy/phylink.c +20 -11
@@ -1360,22 +1360,31 @@ int phylink_ethtool_ksettings_set(struct phylink *pl,if (!s)return -EINVAL ; -/* If we have a fixed link (as specified by firmware), refuse -* to change link parameters. +/* If we have a fixed link, refuse to change link parameters. +* If the link parameters match, accept them but do nothing. */ -if (pl->cur_link_an_mode ==MLO_AN_FIXED && - (s->speed != pl->link_config.speed || - s->duplex != pl->link_config.duplex)) -return -EINVAL ; +if (pl->cur_link_an_mode ==MLO_AN_FIXED ) { +if (s->speed != pl->link_config.speed || + s->duplex != pl->link_config.duplex) +return -EINVAL ; +return 0 ; + } config.speed = s->speed; config.duplex = s->duplex;break ;case AUTONEG_ENABLE : -/* If we have a fixed link, refuse to enable autonegotiation */ -if (pl->cur_link_an_mode ==MLO_AN_FIXED ) -return -EINVAL ; +/* If we have a fixed link, allow autonegotiation (since that +* is our default case) but do not allow the advertisement to +* be changed. If the advertisement matches, simply return. +*/ +if (pl->cur_link_an_mode ==MLO_AN_FIXED ) { +if (!linkmode_equal (config.advertising, + pl->link_config.advertising)) +return -EINVAL ; +return 0 ; + } config.speed =SPEED_UNKNOWN ; config.duplex =DUPLEX_UNKNOWN ;
@@ -1385,8 +1394,8 @@ int phylink_ethtool_ksettings_set(struct phylink *pl,return -EINVAL ; } -/* For a fixed link, this isn't able to change any parameters, -* which just leaves inband mode. +/* We have ruled out the case with a PHY attached, and the +* fixed-link cases. All that is left are in-band links. */ if (phylink_validate (pl, support, &config))return -EINVAL ;
Signature · 1 line
2.20.1
PATCH 11/14 net: phylink: in-band pause mode advertisement update for PCS
In reply to Russell King - ARM Linux adminRe-code the pause in-band advertisement update in light of the addition of PCS support, so that we perform the minimum required; only the PCS configuration function needs to be called in this case, followed by the request to trigger a restart of negotiation if the programmed advertisement changed.
We need to change the pcs_config() signature to pass whether resolved pause should be passed to the MAC for setups such as mvneta and mvpp2 where doing so overrides the MAC manual flow controls.
--- drivers/net/phy/phylink.c | 55 ++++++++++++++++++++++++++++++++++++--- include/linux/phylink.h | 7 +++-- 2 files changed, 56 insertions(+), 6 deletions(-)
drivers/net/phy/phylink.c +51 -4
@@ -441,7 +441,9 @@ static void phylink_pcs_config(struct phylink *pl, bool force_restart,if (pl->pcs_ops && pl->pcs_ops->pcs_config (pl->config, pl->cur_link_an_mode, state->interface, - state->advertising)) + state->advertising, + !!(pl->link_config.pause & +MLO_PAUSE_AN ))) restart = true;phylink_mac_config (pl, state);
@@ -450,6 +452,49 @@ static void phylink_pcs_config(struct phylink *pl, bool force_restart,phylink_mac_pcs_an_restart (pl); } +/* +* Reconfigure for a change of inband advertisement. +* If we have a separate PCS, we only need to call its pcs_config() method, +* and then restart AN if it indicates something changed. Otherwise, we do +* the full MAC reconfiguration. +*/ +static int phylink_change_inband_advert (struct phylink *pl) +{ +int ret; + +if (test_bit (PHYLINK_DISABLE_STOPPED , &pl->phylink_disable_state)) +return 0 ; + +if (!pl->pcs_ops) { +/* Legacy method */ +phylink_mac_config (pl, &pl->link_config); +phylink_mac_pcs_an_restart (pl); +return 0 ; + } + +phylink_dbg (pl,"%s: mode=%s/%s adv=%*pb pause=%02x\n" , __func__, +phylink_an_mode_str (pl->cur_link_an_mode), +phy_modes (pl->link_config.interface), + __ETHTOOL_LINK_MODE_MASK_NBITS, pl->link_config.advertising, + pl->link_config.pause); + +/* Modern PCS-based method; update the advert at the PCS, and +* restart negotiation if the pcs_config() helper indicates that +* the programmed advertisement has changed. +*/ + ret = pl->pcs_ops->pcs_config (pl->config, pl->cur_link_an_mode, + pl->link_config.interface, + pl->link_config.advertising, + !!(pl->link_config.pause &MLO_PAUSE_AN )); +if (ret <0 ) +return ret; + +if (ret >0 ) +phylink_mac_pcs_an_restart (pl); + +return 0 ; +} +static void phylink_mac_pcs_get_state (struct phylink *pl,struct phylink_link_state *state) {
@@ -1524,9 +1569,11 @@ int phylink_ethtool_set_pauseparam(struct phylink *pl, config->pause = pause_state; -if (!pl->phydev && !test_bit (PHYLINK_DISABLE_STOPPED , - &pl->phylink_disable_state)) -phylink_pcs_config (pl, true, &pl->link_config); +/* Update our in-band advertisement, triggering a renegotiation if +* the advertisement changed. +*/ +if (!pl->phydev) +phylink_change_inband_advert (pl);mutex_unlock (&pl->state_mutex);
include/linux/phylink.h +5 -2
@@ -286,7 +286,8 @@ struct phylink_pcs_ops { struct phylink_link_state *state);int (*pcs_config)(struct phylink_config *config,unsigned int mode,phy_interface_t interface, -const unsigned long *advertising); +const unsigned long *advertising, +bool permit_pause_to_mac);void (*pcs_an_restart)(struct phylink_config *config);void (*pcs_link_up)(struct phylink_config *config,unsigned int mode,phy_interface_t interface,int speed,int duplex);
@@ -317,9 +318,11 @@ void pcs_get_state(struct phylink_config *config,* @mode: one of %MLO_AN_FIXED, %MLO_AN_PHY, %MLO_AN_INBAND. * @interface: interface mode to be used * @advertising: adertisement ethtool link mode mask +* @permit_pause_to_mac: permit forwarding pause resolution to MAC * * Configure the PCS for the operating mode, the interface mode, and set -* the advertisement mask. +* the advertisement mask. @permit_pause_to_mac indicates whether the +* hardware may forward the pause mode resolution to the MAC. * * When operating in %MLO_AN_INBAND, inband should always be enabled, * otherwise inband should be disabled.
Signature · 1 line
2.20.1
PATCH 12/14 net: phylink: re-implement interface configuration with PCS
In reply to Russell King - ARM Linux adminWith PCS support, how we implement interface reconfiguration (or other major reconfiguration) is not up to the job; we end up reconfiguring the PCS for an interface change while the link could potentially be up. In order to solve this, add two additional MAC methods for major configuration, one to prepare for the change, and one to finish the change.
This allows mvneta and mvpp2 to shutdown what they require prior to the MAC and PCS configuration calls, and then restart as appropriate.
This impacts ksettings_set(), which now needs to identify whether the change is a minor tweak to the advertisement masks or whether the interface mode has changed, and call the appropriate function for that update.
--- drivers/net/phy/phylink.c | 75 +++++++++++++++++++++++++++------------ include/linux/phylink.h | 48 +++++++++++++++++++++++++ 2 files changed, 100 insertions(+), 23 deletions(-)
drivers/net/phy/phylink.c +52 -23
@@ -433,23 +433,47 @@ static void phylink_mac_pcs_an_restart(struct phylink *pl) } } -static void phylink_pcs_config (struct phylink *pl,bool force_restart, -const struct phylink_link_state *state) +static void phylink_major_config (struct phylink *pl,bool restart, +const struct phylink_link_state *state) { -bool restart = force_restart; +int err; + +phylink_dbg (pl,"major config %s\n" ,phy_modes (state->interface)); -if (pl->pcs_ops && pl->pcs_ops->pcs_config (pl->config, - pl->cur_link_an_mode, - state->interface, - state->advertising, - !!(pl->link_config.pause & -MLO_PAUSE_AN ))) - restart = true; +if (pl->mac_ops->mac_prepare) { + err = pl->mac_ops->mac_prepare (pl->config, pl->cur_link_an_mode, + state->interface); +if (err <0 ) { +phylink_err (pl,"mac_prepare failed: %pe\n" , +ERR_PTR (err)); +return ; + } + }phylink_mac_config (pl, state); +if (pl->pcs_ops) { + err = pl->pcs_ops->pcs_config (pl->config, pl->cur_link_an_mode, + state->interface, + state->advertising, + !!(pl->link_config.pause & +MLO_PAUSE_AN )); +if (err <0 ) +phylink_err (pl,"pcs_config failed: %pe\n" , +ERR_PTR (err)); +if (err >0 ) + restart = true; + }if (restart)phylink_mac_pcs_an_restart (pl); + +if (pl->mac_ops->mac_finish) { + err = pl->mac_ops->mac_finish (pl->config, pl->cur_link_an_mode, + state->interface); +if (err <0 ) +phylink_err (pl,"mac_prepare failed: %pe\n" , +ERR_PTR (err)); + } }/*
@@ -555,7 +579,7 @@ static void phylink_mac_initial_config(struct phylink *pl, bool force_restart) link_state.link = false;phylink_apply_manual_flow (pl, &link_state); -phylink_pcs_config (pl, force_restart, &link_state); +phylink_major_config (pl, force_restart, &link_state); }static const char *phylink_pause_to_str (int pause)
@@ -674,7 +698,7 @@ static void phylink_resolve(struct work_struct *w)phylink_link_down (pl); cur_link_state = false; } -phylink_pcs_config (pl, false, &link_state); +phylink_major_config (pl, false, &link_state); pl->link_config.interface = link_state.interface; }else if (!pl->pcs_ops) {/* The interface remains unchanged, only the speed,
@@ -1450,21 +1474,26 @@ int phylink_ethtool_ksettings_set(struct phylink *pl,return -EINVAL ;mutex_lock (&pl->state_mutex); -linkmode_copy (pl->link_config.advertising, config.advertising); - pl->link_config.interface = config.interface; pl->link_config.speed = config.speed; pl->link_config.duplex = config.duplex; pl->link_config.an_enabled = config.an_enabled; -if (pl->cur_link_an_mode ==MLO_AN_INBAND && - !test_bit (PHYLINK_DISABLE_STOPPED , &pl->phylink_disable_state)) { -/* If in 802.3z mode, this updates the advertisement. -* -* If we are in SGMII mode without a PHY, there is no -* advertisement; the only thing we have is the pause -* modes which can only come from a PHY. -*/ -phylink_pcs_config (pl, true, &pl->link_config); +if (pl->link_config.interface != config.interface) { +/* The interface changed, e.g. 1000base-X <-> 2500base-X */ +/* We need to force the link down, then change the interface */ +if (pl->old_link_state) { +phylink_link_down (pl); + pl->old_link_state = false; + } +if (!test_bit (PHYLINK_DISABLE_STOPPED , + &pl->phylink_disable_state)) +phylink_major_config (pl, false, &config); + pl->link_config.interface = config.interface; +linkmode_copy (pl->link_config.advertising, config.advertising); + }else if (!linkmode_equal (pl->link_config.advertising, + config.advertising)) { +linkmode_copy (pl->link_config.advertising, config.advertising); +phylink_change_inband_advert (pl); }mutex_unlock (&pl->state_mutex);
include/linux/phylink.h +48 -0
75 lines not on this page · raw message
Signature · 1 line
2.20.1
PATCH 13/14 net: phylink: add struct phylink_pcs
In reply to Russell King - ARM Linux adminAdd a way for MAC PCS to have private data while keeping independence from struct phylink_config, which is used for the MAC itself. We need this independence as we will have stand-alone code for PCS that is independent of the MAC. Introduce struct phylink_pcs, which is designed to be embedded in a driver private data structure.
This structure does not include a mdio_device as there are PCS implementations such as the Marvell DSA and network drivers where this is not necessary.
--- drivers/net/phy/phylink.c | 34 ++++++++++++++++++++++------- include/linux/phylink.h | 45 ++++++++++++++++++++++++++------------- 2 files changed, 56 insertions(+), 23 deletions(-)
drivers/net/phy/phylink.c +26 -8
@@ -43,6 +43,7 @@ struct phylink {const struct phylink_mac_ops *mac_ops;const struct phylink_pcs_ops *pcs_ops;struct phylink_config *config; +struct phylink_pcs *pcs;struct device *dev; unsigned int old_link_state:1;
@@ -427,7 +428,7 @@ static void phylink_mac_pcs_an_restart(struct phylink *pl) phy_interface_mode_is_8023z(pl->link_config.interface) && phylink_autoneg_inband(pl->cur_link_an_mode)) {if (pl->pcs_ops) - pl->pcs_ops->pcs_an_restart (pl->config); + pl->pcs_ops->pcs_an_restart (pl->pcs);else pl->mac_ops->mac_an_restart (pl->config); }
@@ -453,7 +454,7 @@ static void phylink_major_config(struct phylink *pl, bool restart,phylink_mac_config (pl, state);if (pl->pcs_ops) { - err = pl->pcs_ops->pcs_config(pl->config, pl->cur_link_an_mode, + err = pl->pcs_ops->pcs_config(pl->pcs, pl->cur_link_an_mode, state->interface, state->advertising, !!(pl->link_config.pause &
@@ -506,7 +507,7 @@ static int phylink_change_inband_advert(struct phylink *pl)* restart negotiation if the pcs_config() helper indicates that * the programmed advertisement has changed. */ - ret = pl->pcs_ops->pcs_config (pl->config, pl->cur_link_an_mode, + ret = pl->pcs_ops->pcs_config (pl->pcs, pl->cur_link_an_mode, pl->link_config.interface, pl->link_config.advertising, !!(pl->link_config.pause &MLO_PAUSE_AN ));
@@ -533,7 +534,7 @@ static void phylink_mac_pcs_get_state(struct phylink *pl, state->link =1 ;if (pl->pcs_ops) - pl->pcs_ops->pcs_get_state (pl->config, state); + pl->pcs_ops->pcs_get_state (pl->pcs, state);else pl->mac_ops->mac_pcs_get_state (pl->config, state); }
@@ -604,7 +605,7 @@ static void phylink_link_up(struct phylink *pl, pl->cur_interface = link_state.interface;if (pl->pcs_ops && pl->pcs_ops->pcs_link_up) - pl->pcs_ops->pcs_link_up (pl->config, pl->cur_link_an_mode, + pl->pcs_ops->pcs_link_up (pl->pcs, pl->cur_link_an_mode, pl->cur_interface, link_state.speed, link_state.duplex);
@@ -863,11 +864,26 @@ struct phylink *phylink_create(struct phylink_config *config, }EXPORT_SYMBOL_GPL (phylink_create); -void phylink_add_pcs (struct phylink *pl,const struct phylink_pcs_ops *ops) +/** +* phylink_set_pcs() - set the current PCS for phylink to use +* @pl: a pointer to a &struct phylink returned from phylink_create() +* @pcs: a pointer to the &struct phylink_pcs +* +* Bind the MAC PCS to phylink. This may be called after phylink_create(), +* in mac_prepare() or mac_config() methods if it is desired to dynamically +* change the PCS. +* +* Please note that there are behavioural changes with the mac_config() +* callback if a PCS is present (denoting a newer setup) so removing a PCS +* is not supported, and if a PCS is going to be used, it must be registered +* by calling phylink_set_pcs() at the latest in the first mac_config() call. +*/ +void phylink_set_pcs (struct phylink *pl,struct phylink_pcs *pcs) { - pl->pcs_ops = ops; + pl->pcs = pcs; + pl->pcs_ops = pcs->ops; } -EXPORT_SYMBOL_GPL (phylink_add_pcs); +EXPORT_SYMBOL_GPL (phylink_set_pcs);/** * phylink_destroy() - cleanup and destroy the phylink instance
@@ -1212,6 +1228,8 @@ void phylink_start(struct phylink *pl)break ;case MLO_AN_INBAND : poll |= pl->config->pcs_poll; +if (pl->pcs) + poll |= pl->pcs->poll;break ; }if (poll)
include/linux/phylink.h +30 -15
104 lines not on this page · raw message
Signature · 1 line
2.20.1
PATCH 14/14 net: phylink: add interface to configure clause 22 PCS PHY
In reply to Russell King - ARM Linux adminAdd an interface to configure the advertisement for a clause 22 PCS PHY, and set the AN enable flag in the BMCR appropriately.
--- drivers/net/phy/phylink.c | 37 +++++++++++++++++++++++++++++++++++++ include/linux/phylink.h | 3 +++ 2 files changed, 40 insertions(+)
drivers/net/phy/phylink.c +37 -0
@@ -2442,6 +2442,43 @@ int phylink_mii_c22_pcs_set_advertisement(struct mdio_device *pcs, }EXPORT_SYMBOL_GPL (phylink_mii_c22_pcs_set_advertisement); +/** +* phylink_mii_c22_pcs_config() - configure clause 22 PCS +* @pcs: a pointer to a &struct mdio_device. +* @mode: link autonegotiation mode +* @interface: the PHY interface mode being configured +* @advertising: the ethtool advertisement mask +* +* Configure a Clause 22 PCS PHY with the appropriate negotiation +* parameters for the @mode, @interface and @advertising parameters. +* Returns negative error number on failure, zero if the advertisement +* has not changed, or positive if there is a change. +*/ +int phylink_mii_c22_pcs_config (struct mdio_device *pcs,unsigned int mode, +phy_interface_t interface, +const unsigned long *advertising) +{ +bool changed; +u16 bmcr; +int ret; + + ret =phylink_mii_c22_pcs_set_advertisement (pcs, interface, + advertising); +if (ret <0 ) +return ret; + + changed = ret >0 ; + + bmcr = mode ==MLO_AN_INBAND ?BMCR_ANENABLE :0 ; + ret =mdiobus_modify (pcs->bus, pcs->addr,MII_BMCR , +BMCR_ANENABLE , bmcr); +if (ret <0 ) +return ret; + +return changed ?1 :0 ; +} +EXPORT_SYMBOL_GPL (phylink_mii_c22_pcs_config); +/** * phylink_mii_c22_pcs_an_restart() - restart 802.3z autonegotiation * @pcs: a pointer to a &struct mdio_device.
include/linux/phylink.h +3 -0
@@ -478,6 +478,9 @@ void phylink_mii_c22_pcs_get_state(struct mdio_device *pcs,int phylink_mii_c22_pcs_set_advertisement (struct mdio_device *pcs,phy_interface_t interface,const unsigned long *advertising); +int phylink_mii_c22_pcs_config (struct mdio_device *pcs,unsigned int mode, +phy_interface_t interface, +const unsigned long *advertising);void phylink_mii_c22_pcs_an_restart (struct mdio_device *pcs); void phylink_mii_c45_pcs_get_state(struct mdio_device *pcs,
Signature · 1 line
2.20.1
David Miller
maintainerFrom: Russell King - ARM Linux admin <[email protected]>
Date: Tue, 21 Jul 2020 12:01:52 +0100
Quoted 23 lines
This series updates the rudimentary phylink PCS support with the results of the last four months of development of that. Phylink PCS support was initially added back at the end of March, when it became clear that the current approach of treating everything at the MAC end as being part of the MAC was inadequate.
However, this rudimentary implementation was fine initially for mvneta and similar, but in practice had a fair number of issues, particularly when ethtool interfaces were used to change various link properties.
It became apparent that relying on the
phylink_configstructure for the PCS was also bad when it became clear that the same PCS was used in DSA drivers as well as in NXPs other offerings, and there was a desire to re-use that code.It also became apparent that splitting the "configuration" step on an interface mode configuration between the MAC and PCS using just
mac_config()andpcs_config()methods was not sufficient for some setups, as the MAC needed to be "taken down" prior to making changes, and once all settings were complete, the MAC could only then be resumed.
Quoted 5 lines
This series addresses these points, progressing PCS support, and has been developed with mvneta and DPAA2 setups, with work on both those drivers to prove this approach. It has been rigorously tested with mvneta, as that provides the most flexibility for testing the various code paths.
...
Series applied, thank you.