public-mirrors/linux
1
0
Fork 0
mirror of git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git synced 2025-08-05 16:54:27 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions
linux/drivers/net/ethernet/mscc/ocelot_net.c

1333 lines
35 KiB
C
Raw Normal View History

net: mscc: ocelot: move net_device related functions to ocelot_net.c The ocelot hardware library shouldn't contain too much net_device specific code, since it is shared with DSA which abstracts that structure away. So much as much of this code as possible into the mscc_ocelot driver and outside of the common library. We're making an exception for MDB and LAG code. That is not yet exported to DSA, but when it will, most of the code that's already in ocelot.c will remain there. So, there's no point in moving code to ocelot_net.c just to move it back later. We could have moved all net_device code to ocelot_vsc7514.c directly, but let's operate under the assumption that if a new switchdev ocelot driver gets added, it'll define its SoC-specific stuff in a new ocelot_vsc*.c file and it'll reuse the rest of the code. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2020-06-20 18:43:44 +03:00
// SPDX-License-Identifier: (GPL-2.0 OR MIT)
/* Microsemi Ocelot Switch driver
net: mscc: ocelot: configure watermarks using devlink-sb Using devlink-sb, we can configure 12/16 (the important 75%) of the switch's controlling watermarks for congestion drops, and we can monitor 50% of the watermark occupancies (we can monitor the reservation watermarks, but not the sharing watermarks, which are exposed as pool sizes). The following definitions can be made: SB_BUF=0 # The devlink-sb for frame buffers SB_REF=1 # The devlink-sb for frame references POOL_ING=0 # The pool for ingress traffic. Both devlink-sb instances # have one of these. POOL_EGR=1 # The pool for egress traffic. Both devlink-sb instances # have one of these. Editing the hardware watermarks is done in the following way: BUF_xxxx_I is accessed when sb=$SB_BUF and pool=$POOL_ING REF_xxxx_I is accessed when sb=$SB_REF and pool=$POOL_ING BUF_xxxx_E is accessed when sb=$SB_BUF and pool=$POOL_EGR REF_xxxx_E is accessed when sb=$SB_REF and pool=$POOL_EGR Configuring the sharing watermarks for COL_SHR(dp=0) is done implicitly by modifying the corresponding pool size. By default, the pool size has maximum size, so this can be skipped. devlink sb pool set pci/0000:00:00.5 sb $SB_BUF pool $POOL_ING \ size 129840 thtype static Since by default there is no buffer reservation, the above command has maxed out BUF_COL_SHR_I(dp=0). Configuring the per-port reservation watermark (P_RSRV) is done in the following way: devlink sb port pool set pci/0000:00:00.5/0 sb $SB_BUF \ pool $POOL_ING th 1000 The above command sets BUF_P_RSRV_I(port 0) to 1000 bytes. After this command, the sharing watermarks are internally reconfigured with 1000 bytes less, i.e. from 129840 bytes to 128840 bytes. Configuring the per-port-tc reservation watermarks (Q_RSRV) is done in the following way: for tc in {0..7}; do devlink sb tc bind set pci/0000:00:00.5/0 sb 0 tc $tc \ type ingress pool $POOL_ING \ th 3000 done The above command sets BUF_Q_RSRV_I(port 0, tc 0..7) to 3000 bytes. The sharing watermarks are again reconfigured with 24000 bytes less. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2021-01-15 04:11:20 +02:00
*
* This contains glue logic between the switchdev driver operations and the
* mscc_ocelot_switch_lib.
net: mscc: ocelot: move net_device related functions to ocelot_net.c The ocelot hardware library shouldn't contain too much net_device specific code, since it is shared with DSA which abstracts that structure away. So much as much of this code as possible into the mscc_ocelot driver and outside of the common library. We're making an exception for MDB and LAG code. That is not yet exported to DSA, but when it will, most of the code that's already in ocelot.c will remain there. So, there's no point in moving code to ocelot_net.c just to move it back later. We could have moved all net_device code to ocelot_vsc7514.c directly, but let's operate under the assumption that if a new switchdev ocelot driver gets added, it'll define its SoC-specific stuff in a new ocelot_vsc*.c file and it'll reuse the rest of the code. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2020-06-20 18:43:44 +03:00
*
* Copyright (c) 2017, 2019 Microsemi Corporation
net: mscc: ocelot: configure watermarks using devlink-sb Using devlink-sb, we can configure 12/16 (the important 75%) of the switch's controlling watermarks for congestion drops, and we can monitor 50% of the watermark occupancies (we can monitor the reservation watermarks, but not the sharing watermarks, which are exposed as pool sizes). The following definitions can be made: SB_BUF=0 # The devlink-sb for frame buffers SB_REF=1 # The devlink-sb for frame references POOL_ING=0 # The pool for ingress traffic. Both devlink-sb instances # have one of these. POOL_EGR=1 # The pool for egress traffic. Both devlink-sb instances # have one of these. Editing the hardware watermarks is done in the following way: BUF_xxxx_I is accessed when sb=$SB_BUF and pool=$POOL_ING REF_xxxx_I is accessed when sb=$SB_REF and pool=$POOL_ING BUF_xxxx_E is accessed when sb=$SB_BUF and pool=$POOL_EGR REF_xxxx_E is accessed when sb=$SB_REF and pool=$POOL_EGR Configuring the sharing watermarks for COL_SHR(dp=0) is done implicitly by modifying the corresponding pool size. By default, the pool size has maximum size, so this can be skipped. devlink sb pool set pci/0000:00:00.5 sb $SB_BUF pool $POOL_ING \ size 129840 thtype static Since by default there is no buffer reservation, the above command has maxed out BUF_COL_SHR_I(dp=0). Configuring the per-port reservation watermark (P_RSRV) is done in the following way: devlink sb port pool set pci/0000:00:00.5/0 sb $SB_BUF \ pool $POOL_ING th 1000 The above command sets BUF_P_RSRV_I(port 0) to 1000 bytes. After this command, the sharing watermarks are internally reconfigured with 1000 bytes less, i.e. from 129840 bytes to 128840 bytes. Configuring the per-port-tc reservation watermarks (Q_RSRV) is done in the following way: for tc in {0..7}; do devlink sb tc bind set pci/0000:00:00.5/0 sb 0 tc $tc \ type ingress pool $POOL_ING \ th 3000 done The above command sets BUF_Q_RSRV_I(port 0, tc 0..7) to 3000 bytes. The sharing watermarks are again reconfigured with 24000 bytes less. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2021-01-15 04:11:20 +02:00
* Copyright 2020-2021 NXP Semiconductors
net: mscc: ocelot: move net_device related functions to ocelot_net.c The ocelot hardware library shouldn't contain too much net_device specific code, since it is shared with DSA which abstracts that structure away. So much as much of this code as possible into the mscc_ocelot driver and outside of the common library. We're making an exception for MDB and LAG code. That is not yet exported to DSA, but when it will, most of the code that's already in ocelot.c will remain there. So, there's no point in moving code to ocelot_net.c just to move it back later. We could have moved all net_device code to ocelot_vsc7514.c directly, but let's operate under the assumption that if a new switchdev ocelot driver gets added, it'll define its SoC-specific stuff in a new ocelot_vsc*.c file and it'll reuse the rest of the code. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2020-06-20 18:43:44 +03:00
*/
#include <linux/if_bridge.h>
net: mscc: ocelot: export VCAP structures to include/soc/mscc The Felix driver will need to preinstall some VCAP filters for its tag_8021q implementation (outside of the tc-flower offload logic), so these need to be exported to the common includes. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Reviewed-by: Florian Fainelli <f.fainelli@gmail.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2021-01-29 03:00:00 +02:00
#include <net/pkt_cls.h>
net: mscc: ocelot: move net_device related functions to ocelot_net.c The ocelot hardware library shouldn't contain too much net_device specific code, since it is shared with DSA which abstracts that structure away. So much as much of this code as possible into the mscc_ocelot driver and outside of the common library. We're making an exception for MDB and LAG code. That is not yet exported to DSA, but when it will, most of the code that's already in ocelot.c will remain there. So, there's no point in moving code to ocelot_net.c just to move it back later. We could have moved all net_device code to ocelot_vsc7514.c directly, but let's operate under the assumption that if a new switchdev ocelot driver gets added, it'll define its SoC-specific stuff in a new ocelot_vsc*.c file and it'll reuse the rest of the code. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2020-06-20 18:43:44 +03:00
#include "ocelot.h"
net: mscc: ocelot: rename ocelot_ace.{c, h} to ocelot_vcap.{c,h} Access Control Lists (and their respective Access Control Entries) are specifically entries in the VCAP IS2, the security enforcement block, according to the documentation. Let's rename the files that deal with generic operations on the VCAP TCAM, so that VCAP IS1 and ES0 can reuse the same code without confusion. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2020-06-20 18:43:45 +03:00
#include "ocelot_vcap.h"
net: mscc: ocelot: move net_device related functions to ocelot_net.c The ocelot hardware library shouldn't contain too much net_device specific code, since it is shared with DSA which abstracts that structure away. So much as much of this code as possible into the mscc_ocelot driver and outside of the common library. We're making an exception for MDB and LAG code. That is not yet exported to DSA, but when it will, most of the code that's already in ocelot.c will remain there. So, there's no point in moving code to ocelot_net.c just to move it back later. We could have moved all net_device code to ocelot_vsc7514.c directly, but let's operate under the assumption that if a new switchdev ocelot driver gets added, it'll define its SoC-specific stuff in a new ocelot_vsc*.c file and it'll reuse the rest of the code. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2020-06-20 18:43:44 +03:00
net: mscc: ocelot: configure watermarks using devlink-sb Using devlink-sb, we can configure 12/16 (the important 75%) of the switch's controlling watermarks for congestion drops, and we can monitor 50% of the watermark occupancies (we can monitor the reservation watermarks, but not the sharing watermarks, which are exposed as pool sizes). The following definitions can be made: SB_BUF=0 # The devlink-sb for frame buffers SB_REF=1 # The devlink-sb for frame references POOL_ING=0 # The pool for ingress traffic. Both devlink-sb instances # have one of these. POOL_EGR=1 # The pool for egress traffic. Both devlink-sb instances # have one of these. Editing the hardware watermarks is done in the following way: BUF_xxxx_I is accessed when sb=$SB_BUF and pool=$POOL_ING REF_xxxx_I is accessed when sb=$SB_REF and pool=$POOL_ING BUF_xxxx_E is accessed when sb=$SB_BUF and pool=$POOL_EGR REF_xxxx_E is accessed when sb=$SB_REF and pool=$POOL_EGR Configuring the sharing watermarks for COL_SHR(dp=0) is done implicitly by modifying the corresponding pool size. By default, the pool size has maximum size, so this can be skipped. devlink sb pool set pci/0000:00:00.5 sb $SB_BUF pool $POOL_ING \ size 129840 thtype static Since by default there is no buffer reservation, the above command has maxed out BUF_COL_SHR_I(dp=0). Configuring the per-port reservation watermark (P_RSRV) is done in the following way: devlink sb port pool set pci/0000:00:00.5/0 sb $SB_BUF \ pool $POOL_ING th 1000 The above command sets BUF_P_RSRV_I(port 0) to 1000 bytes. After this command, the sharing watermarks are internally reconfigured with 1000 bytes less, i.e. from 129840 bytes to 128840 bytes. Configuring the per-port-tc reservation watermarks (Q_RSRV) is done in the following way: for tc in {0..7}; do devlink sb tc bind set pci/0000:00:00.5/0 sb 0 tc $tc \ type ingress pool $POOL_ING \ th 3000 done The above command sets BUF_Q_RSRV_I(port 0, tc 0..7) to 3000 bytes. The sharing watermarks are again reconfigured with 24000 bytes less. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2021-01-15 04:11:20 +02:00
static struct ocelot *devlink_port_to_ocelot(struct devlink_port *dlp)
{
return devlink_priv(dlp->devlink);
}
static int devlink_port_to_port(struct devlink_port *dlp)
{
struct ocelot *ocelot = devlink_port_to_ocelot(dlp);
return dlp - ocelot->devlink_ports;
}
static int ocelot_devlink_sb_pool_get(struct devlink *dl,
unsigned int sb_index, u16 pool_index,
struct devlink_sb_pool_info *pool_info)
{
struct ocelot *ocelot = devlink_priv(dl);
return ocelot_sb_pool_get(ocelot, sb_index, pool_index, pool_info);
}
static int ocelot_devlink_sb_pool_set(struct devlink *dl, unsigned int sb_index,
u16 pool_index, u32 size,
enum devlink_sb_threshold_type threshold_type,
struct netlink_ext_ack *extack)
{
struct ocelot *ocelot = devlink_priv(dl);
return ocelot_sb_pool_set(ocelot, sb_index, pool_index, size,
threshold_type, extack);
}
static int ocelot_devlink_sb_port_pool_get(struct devlink_port *dlp,
unsigned int sb_index, u16 pool_index,
u32 *p_threshold)
{
struct ocelot *ocelot = devlink_port_to_ocelot(dlp);
int port = devlink_port_to_port(dlp);
return ocelot_sb_port_pool_get(ocelot, port, sb_index, pool_index,
p_threshold);
}
static int ocelot_devlink_sb_port_pool_set(struct devlink_port *dlp,
unsigned int sb_index, u16 pool_index,
u32 threshold,
struct netlink_ext_ack *extack)
{
struct ocelot *ocelot = devlink_port_to_ocelot(dlp);
int port = devlink_port_to_port(dlp);
return ocelot_sb_port_pool_set(ocelot, port, sb_index, pool_index,
threshold, extack);
}
static int
ocelot_devlink_sb_tc_pool_bind_get(struct devlink_port *dlp,
unsigned int sb_index, u16 tc_index,
enum devlink_sb_pool_type pool_type,
u16 *p_pool_index, u32 *p_threshold)
{
struct ocelot *ocelot = devlink_port_to_ocelot(dlp);
int port = devlink_port_to_port(dlp);
return ocelot_sb_tc_pool_bind_get(ocelot, port, sb_index, tc_index,
pool_type, p_pool_index,
p_threshold);
}
static int
ocelot_devlink_sb_tc_pool_bind_set(struct devlink_port *dlp,
unsigned int sb_index, u16 tc_index,
enum devlink_sb_pool_type pool_type,
u16 pool_index, u32 threshold,
struct netlink_ext_ack *extack)
{
struct ocelot *ocelot = devlink_port_to_ocelot(dlp);
int port = devlink_port_to_port(dlp);
return ocelot_sb_tc_pool_bind_set(ocelot, port, sb_index, tc_index,
pool_type, pool_index, threshold,
extack);
}
static int ocelot_devlink_sb_occ_snapshot(struct devlink *dl,
unsigned int sb_index)
{
struct ocelot *ocelot = devlink_priv(dl);
return ocelot_sb_occ_snapshot(ocelot, sb_index);
}
static int ocelot_devlink_sb_occ_max_clear(struct devlink *dl,
unsigned int sb_index)
{
struct ocelot *ocelot = devlink_priv(dl);
return ocelot_sb_occ_max_clear(ocelot, sb_index);
}
static int ocelot_devlink_sb_occ_port_pool_get(struct devlink_port *dlp,
unsigned int sb_index,
u16 pool_index, u32 *p_cur,
u32 *p_max)
{
struct ocelot *ocelot = devlink_port_to_ocelot(dlp);
int port = devlink_port_to_port(dlp);
return ocelot_sb_occ_port_pool_get(ocelot, port, sb_index, pool_index,
p_cur, p_max);
}
static int
ocelot_devlink_sb_occ_tc_port_bind_get(struct devlink_port *dlp,
unsigned int sb_index, u16 tc_index,
enum devlink_sb_pool_type pool_type,
u32 *p_cur, u32 *p_max)
{
struct ocelot *ocelot = devlink_port_to_ocelot(dlp);
int port = devlink_port_to_port(dlp);
return ocelot_sb_occ_tc_port_bind_get(ocelot, port, sb_index,
tc_index, pool_type,
p_cur, p_max);
}
net: mscc: ocelot: register devlink ports Add devlink integration into the mscc_ocelot switchdev driver. All physical ports (i.e. the unused ones as well) except the CPU port module at ocelot->num_phys_ports are registered with devlink, and that requires keeping the devlink_port structure outside struct ocelot_port_private, since the latter has a 1:1 mapping with a struct net_device (which does not exist for unused ports). Since we use devlink_port_type_eth_set to link the devlink port to the net_device, we can as well remove the .ndo_get_phys_port_name and .ndo_get_port_parent_id implementations, since devlink takes care of retrieving the port name and number automatically, once .ndo_get_devlink_port is implemented. Note that the felix DSA driver is already integrated with devlink by default, since that is a thing that the DSA core takes care of. This is the reason why these devlink stubs were put in ocelot_net.c and not in the common library. It is also the reason why ocelot::devlink is a pointer and not a full structure embedded inside struct ocelot: because the mscc_ocelot driver allocates that by itself (as the container of struct ocelot, in fact), but in the case of felix, it is DSA who allocates the devlink, and felix just propagates the pointer towards struct ocelot. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Reviewed-by: Florian Fainelli <f.fainelli@gmail.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2021-01-15 04:11:18 +02:00
const struct devlink_ops ocelot_devlink_ops = {
net: mscc: ocelot: configure watermarks using devlink-sb Using devlink-sb, we can configure 12/16 (the important 75%) of the switch's controlling watermarks for congestion drops, and we can monitor 50% of the watermark occupancies (we can monitor the reservation watermarks, but not the sharing watermarks, which are exposed as pool sizes). The following definitions can be made: SB_BUF=0 # The devlink-sb for frame buffers SB_REF=1 # The devlink-sb for frame references POOL_ING=0 # The pool for ingress traffic. Both devlink-sb instances # have one of these. POOL_EGR=1 # The pool for egress traffic. Both devlink-sb instances # have one of these. Editing the hardware watermarks is done in the following way: BUF_xxxx_I is accessed when sb=$SB_BUF and pool=$POOL_ING REF_xxxx_I is accessed when sb=$SB_REF and pool=$POOL_ING BUF_xxxx_E is accessed when sb=$SB_BUF and pool=$POOL_EGR REF_xxxx_E is accessed when sb=$SB_REF and pool=$POOL_EGR Configuring the sharing watermarks for COL_SHR(dp=0) is done implicitly by modifying the corresponding pool size. By default, the pool size has maximum size, so this can be skipped. devlink sb pool set pci/0000:00:00.5 sb $SB_BUF pool $POOL_ING \ size 129840 thtype static Since by default there is no buffer reservation, the above command has maxed out BUF_COL_SHR_I(dp=0). Configuring the per-port reservation watermark (P_RSRV) is done in the following way: devlink sb port pool set pci/0000:00:00.5/0 sb $SB_BUF \ pool $POOL_ING th 1000 The above command sets BUF_P_RSRV_I(port 0) to 1000 bytes. After this command, the sharing watermarks are internally reconfigured with 1000 bytes less, i.e. from 129840 bytes to 128840 bytes. Configuring the per-port-tc reservation watermarks (Q_RSRV) is done in the following way: for tc in {0..7}; do devlink sb tc bind set pci/0000:00:00.5/0 sb 0 tc $tc \ type ingress pool $POOL_ING \ th 3000 done The above command sets BUF_Q_RSRV_I(port 0, tc 0..7) to 3000 bytes. The sharing watermarks are again reconfigured with 24000 bytes less. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2021-01-15 04:11:20 +02:00
.sb_pool_get = ocelot_devlink_sb_pool_get,
.sb_pool_set = ocelot_devlink_sb_pool_set,
.sb_port_pool_get = ocelot_devlink_sb_port_pool_get,
.sb_port_pool_set = ocelot_devlink_sb_port_pool_set,
.sb_tc_pool_bind_get = ocelot_devlink_sb_tc_pool_bind_get,
.sb_tc_pool_bind_set = ocelot_devlink_sb_tc_pool_bind_set,
.sb_occ_snapshot = ocelot_devlink_sb_occ_snapshot,
.sb_occ_max_clear = ocelot_devlink_sb_occ_max_clear,
.sb_occ_port_pool_get = ocelot_devlink_sb_occ_port_pool_get,
.sb_occ_tc_port_bind_get = ocelot_devlink_sb_occ_tc_port_bind_get,
net: mscc: ocelot: register devlink ports Add devlink integration into the mscc_ocelot switchdev driver. All physical ports (i.e. the unused ones as well) except the CPU port module at ocelot->num_phys_ports are registered with devlink, and that requires keeping the devlink_port structure outside struct ocelot_port_private, since the latter has a 1:1 mapping with a struct net_device (which does not exist for unused ports). Since we use devlink_port_type_eth_set to link the devlink port to the net_device, we can as well remove the .ndo_get_phys_port_name and .ndo_get_port_parent_id implementations, since devlink takes care of retrieving the port name and number automatically, once .ndo_get_devlink_port is implemented. Note that the felix DSA driver is already integrated with devlink by default, since that is a thing that the DSA core takes care of. This is the reason why these devlink stubs were put in ocelot_net.c and not in the common library. It is also the reason why ocelot::devlink is a pointer and not a full structure embedded inside struct ocelot: because the mscc_ocelot driver allocates that by itself (as the container of struct ocelot, in fact), but in the case of felix, it is DSA who allocates the devlink, and felix just propagates the pointer towards struct ocelot. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Reviewed-by: Florian Fainelli <f.fainelli@gmail.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2021-01-15 04:11:18 +02:00
};
int ocelot_port_devlink_init(struct ocelot *ocelot, int port,
enum devlink_port_flavour flavour)
{
struct devlink_port *dlp = &ocelot->devlink_ports[port];
int id_len = sizeof(ocelot->base_mac);
struct devlink *dl = ocelot->devlink;
struct devlink_port_attrs attrs = {};
memcpy(attrs.switch_id.id, &ocelot->base_mac, id_len);
attrs.switch_id.id_len = id_len;
attrs.phys.port_number = port;
attrs.flavour = flavour;
devlink_port_attrs_set(dlp, &attrs);
return devlink_port_register(dl, dlp, port);
}
void ocelot_port_devlink_teardown(struct ocelot *ocelot, int port)
{
struct devlink_port *dlp = &ocelot->devlink_ports[port];
devlink_port_unregister(dlp);
}
static struct devlink_port *ocelot_get_devlink_port(struct net_device *dev)
{
struct ocelot_port_private *priv = netdev_priv(dev);
struct ocelot *ocelot = priv->port.ocelot;
int port = priv->chip_port;
return &ocelot->devlink_ports[port];
}
net: mscc: ocelot: move net_device related functions to ocelot_net.c The ocelot hardware library shouldn't contain too much net_device specific code, since it is shared with DSA which abstracts that structure away. So much as much of this code as possible into the mscc_ocelot driver and outside of the common library. We're making an exception for MDB and LAG code. That is not yet exported to DSA, but when it will, most of the code that's already in ocelot.c will remain there. So, there's no point in moving code to ocelot_net.c just to move it back later. We could have moved all net_device code to ocelot_vsc7514.c directly, but let's operate under the assumption that if a new switchdev ocelot driver gets added, it'll define its SoC-specific stuff in a new ocelot_vsc*.c file and it'll reuse the rest of the code. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2020-06-20 18:43:44 +03:00
int ocelot_setup_tc_cls_flower(struct ocelot_port_private *priv,
struct flow_cls_offload *f,
bool ingress)
{
struct ocelot *ocelot = priv->port.ocelot;
int port = priv->chip_port;
if (!ingress)
return -EOPNOTSUPP;
switch (f->command) {
case FLOW_CLS_REPLACE:
return ocelot_cls_flower_replace(ocelot, port, f, ingress);
case FLOW_CLS_DESTROY:
return ocelot_cls_flower_destroy(ocelot, port, f, ingress);
case FLOW_CLS_STATS:
return ocelot_cls_flower_stats(ocelot, port, f, ingress);
default:
return -EOPNOTSUPP;
}
}
static int ocelot_setup_tc_cls_matchall(struct ocelot_port_private *priv,
struct tc_cls_matchall_offload *f,
bool ingress)
{
struct netlink_ext_ack *extack = f->common.extack;
struct ocelot *ocelot = priv->port.ocelot;
struct ocelot_policer pol = { 0 };
struct flow_action_entry *action;
int port = priv->chip_port;
int err;
if (!ingress) {
NL_SET_ERR_MSG_MOD(extack, "Only ingress is supported");
return -EOPNOTSUPP;
}
switch (f->command) {
case TC_CLSMATCHALL_REPLACE:
if (!flow_offload_has_one_action(&f->rule->action)) {
NL_SET_ERR_MSG_MOD(extack,
"Only one action is supported");
return -EOPNOTSUPP;
}
if (priv->tc.block_shared) {
NL_SET_ERR_MSG_MOD(extack,
"Rate limit is not supported on shared blocks");
return -EOPNOTSUPP;
}
action = &f->rule->action.entries[0];
if (action->id != FLOW_ACTION_POLICE) {
NL_SET_ERR_MSG_MOD(extack, "Unsupported action");
return -EOPNOTSUPP;
}
if (priv->tc.police_id && priv->tc.police_id != f->cookie) {
NL_SET_ERR_MSG_MOD(extack,
"Only one policer per port is supported");
return -EEXIST;
}
pol.rate = (u32)div_u64(action->police.rate_bytes_ps, 1000) * 8;
net:qos: police action offloading parameter 'burst' change to the original value Since 'tcfp_burst' with TICK factor, driver side always need to recover it to the original value, this patch moves the generic calculation and recover to the 'burst' original value before offloading to device driver. Signed-off-by: Po Liu <po.liu@nxp.com> Acked-by: Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2020-06-29 14:54:16 +08:00
pol.burst = action->police.burst;
net: mscc: ocelot: move net_device related functions to ocelot_net.c The ocelot hardware library shouldn't contain too much net_device specific code, since it is shared with DSA which abstracts that structure away. So much as much of this code as possible into the mscc_ocelot driver and outside of the common library. We're making an exception for MDB and LAG code. That is not yet exported to DSA, but when it will, most of the code that's already in ocelot.c will remain there. So, there's no point in moving code to ocelot_net.c just to move it back later. We could have moved all net_device code to ocelot_vsc7514.c directly, but let's operate under the assumption that if a new switchdev ocelot driver gets added, it'll define its SoC-specific stuff in a new ocelot_vsc*.c file and it'll reuse the rest of the code. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2020-06-20 18:43:44 +03:00
err = ocelot_port_policer_add(ocelot, port, &pol);
if (err) {
NL_SET_ERR_MSG_MOD(extack, "Could not add policer");
return err;
}
priv->tc.police_id = f->cookie;
priv->tc.offload_cnt++;
return 0;
case TC_CLSMATCHALL_DESTROY:
if (priv->tc.police_id != f->cookie)
return -ENOENT;
err = ocelot_port_policer_del(ocelot, port);
if (err) {
NL_SET_ERR_MSG_MOD(extack,
"Could not delete policer");
return err;
}
priv->tc.police_id = 0;
priv->tc.offload_cnt--;
return 0;
case TC_CLSMATCHALL_STATS:
default:
return -EOPNOTSUPP;
}
}
static int ocelot_setup_tc_block_cb(enum tc_setup_type type,
void *type_data,
void *cb_priv, bool ingress)
{
struct ocelot_port_private *priv = cb_priv;
if (!tc_cls_can_offload_and_chain0(priv->dev, type_data))
return -EOPNOTSUPP;
switch (type) {
case TC_SETUP_CLSMATCHALL:
return ocelot_setup_tc_cls_matchall(priv, type_data, ingress);
case TC_SETUP_CLSFLOWER:
return ocelot_setup_tc_cls_flower(priv, type_data, ingress);
default:
return -EOPNOTSUPP;
}
}
static int ocelot_setup_tc_block_cb_ig(enum tc_setup_type type,
void *type_data,
void *cb_priv)
{
return ocelot_setup_tc_block_cb(type, type_data,
cb_priv, true);
}
static int ocelot_setup_tc_block_cb_eg(enum tc_setup_type type,
void *type_data,
void *cb_priv)
{
return ocelot_setup_tc_block_cb(type, type_data,
cb_priv, false);
}
static LIST_HEAD(ocelot_block_cb_list);
static int ocelot_setup_tc_block(struct ocelot_port_private *priv,
struct flow_block_offload *f)
{
struct flow_block_cb *block_cb;
flow_setup_cb_t *cb;
if (f->binder_type == FLOW_BLOCK_BINDER_TYPE_CLSACT_INGRESS) {
cb = ocelot_setup_tc_block_cb_ig;
priv->tc.block_shared = f->block_shared;
} else if (f->binder_type == FLOW_BLOCK_BINDER_TYPE_CLSACT_EGRESS) {
cb = ocelot_setup_tc_block_cb_eg;
} else {
return -EOPNOTSUPP;
}
f->driver_block_list = &ocelot_block_cb_list;
switch (f->command) {
case FLOW_BLOCK_BIND:
if (flow_block_cb_is_busy(cb, priv, &ocelot_block_cb_list))
return -EBUSY;
block_cb = flow_block_cb_alloc(cb, priv, priv, NULL);
if (IS_ERR(block_cb))
return PTR_ERR(block_cb);
flow_block_cb_add(block_cb, f);
list_add_tail(&block_cb->driver_list, f->driver_block_list);
return 0;
case FLOW_BLOCK_UNBIND:
block_cb = flow_block_cb_lookup(f->block, cb, priv);
if (!block_cb)
return -ENOENT;
flow_block_cb_remove(block_cb, f);
list_del(&block_cb->driver_list);
return 0;
default:
return -EOPNOTSUPP;
}
}
static int ocelot_setup_tc(struct net_device *dev, enum tc_setup_type type,
void *type_data)
{
struct ocelot_port_private *priv = netdev_priv(dev);
switch (type) {
case TC_SETUP_BLOCK:
return ocelot_setup_tc_block(priv, type_data);
default:
return -EOPNOTSUPP;
}
return 0;
}
static void ocelot_port_adjust_link(struct net_device *dev)
{
struct ocelot_port_private *priv = netdev_priv(dev);
struct ocelot *ocelot = priv->port.ocelot;
int port = priv->chip_port;
ocelot_adjust_link(ocelot, port, dev->phydev);
}
net: mscc: ocelot: deny changing the native VLAN from the prepare phase Put the preparation phase of switchdev VLAN objects to some good use, and move the check we already had, for preventing the existence of more than one egress-untagged VLAN per port, to the preparation phase of the addition. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2020-10-31 12:29:15 +02:00
static int ocelot_vlan_vid_prepare(struct net_device *dev, u16 vid, bool pvid,
bool untagged)
{
struct ocelot_port_private *priv = netdev_priv(dev);
struct ocelot_port *ocelot_port = &priv->port;
struct ocelot *ocelot = ocelot_port->ocelot;
int port = priv->chip_port;
return ocelot_vlan_prepare(ocelot, port, vid, pvid, untagged);
}
net: mscc: ocelot: move net_device related functions to ocelot_net.c The ocelot hardware library shouldn't contain too much net_device specific code, since it is shared with DSA which abstracts that structure away. So much as much of this code as possible into the mscc_ocelot driver and outside of the common library. We're making an exception for MDB and LAG code. That is not yet exported to DSA, but when it will, most of the code that's already in ocelot.c will remain there. So, there's no point in moving code to ocelot_net.c just to move it back later. We could have moved all net_device code to ocelot_vsc7514.c directly, but let's operate under the assumption that if a new switchdev ocelot driver gets added, it'll define its SoC-specific stuff in a new ocelot_vsc*.c file and it'll reuse the rest of the code. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2020-06-20 18:43:44 +03:00
static int ocelot_vlan_vid_add(struct net_device *dev, u16 vid, bool pvid,
bool untagged)
{
struct ocelot_port_private *priv = netdev_priv(dev);
struct ocelot_port *ocelot_port = &priv->port;
struct ocelot *ocelot = ocelot_port->ocelot;
int port = priv->chip_port;
int ret;
ret = ocelot_vlan_add(ocelot, port, vid, pvid, untagged);
if (ret)
return ret;
/* Add the port MAC address to with the right VLAN information */
ocelot_mact_learn(ocelot, PGID_CPU, dev->dev_addr, vid,
ENTRYTYPE_LOCKED);
return 0;
}
static int ocelot_vlan_vid_del(struct net_device *dev, u16 vid)
{
struct ocelot_port_private *priv = netdev_priv(dev);
struct ocelot *ocelot = priv->port.ocelot;
int port = priv->chip_port;
int ret;
/* 8021q removes VID 0 on module unload for all interfaces
* with VLAN filtering feature. We need to keep it to receive
* untagged traffic.
*/
if (vid == 0)
return 0;
ret = ocelot_vlan_del(ocelot, port, vid);
if (ret)
return ret;
/* Del the port MAC address to with the right VLAN information */
ocelot_mact_forget(ocelot, dev->dev_addr, vid);
return 0;
}
static int ocelot_port_open(struct net_device *dev)
{
struct ocelot_port_private *priv = netdev_priv(dev);
struct ocelot_port *ocelot_port = &priv->port;
struct ocelot *ocelot = ocelot_port->ocelot;
int port = priv->chip_port;
int err;
if (priv->serdes) {
err = phy_set_mode_ext(priv->serdes, PHY_MODE_ETHERNET,
ocelot_port->phy_mode);
if (err) {
netdev_err(dev, "Could not set mode of SerDes\n");
return err;
}
}
err = phy_connect_direct(dev, priv->phy, &ocelot_port_adjust_link,
ocelot_port->phy_mode);
if (err) {
netdev_err(dev, "Could not attach to PHY\n");
return err;
}
dev->phydev = priv->phy;
phy_attached_info(priv->phy);
phy_start(priv->phy);
ocelot_port_enable(ocelot, port, priv->phy);
return 0;
}
static int ocelot_port_stop(struct net_device *dev)
{
struct ocelot_port_private *priv = netdev_priv(dev);
struct ocelot *ocelot = priv->port.ocelot;
int port = priv->chip_port;
phy_disconnect(priv->phy);
dev->phydev = NULL;
ocelot_port_disable(ocelot, port);
return 0;
}
/* Generate the IFH for frame injection
*
* The IFH is a 128bit-value
* bit 127: bypass the analyzer processing
* bit 56-67: destination mask
* bit 28-29: pop_cnt: 3 disables all rewriting of the frame
* bit 20-27: cpu extraction queue mask
* bit 16: tag type 0: C-tag, 1: S-tag
* bit 0-11: VID
*/
static int ocelot_gen_ifh(u32 *ifh, struct frame_info *info)
{
ifh[0] = IFH_INJ_BYPASS | ((0x1ff & info->rew_op) << 21);
ifh[1] = (0xf00 & info->port) >> 8;
ifh[2] = (0xff & info->port) << 24;
ifh[3] = (info->tag_type << 16) | info->vid;
return 0;
}
static int ocelot_port_xmit(struct sk_buff *skb, struct net_device *dev)
{
struct ocelot_port_private *priv = netdev_priv(dev);
struct skb_shared_info *shinfo = skb_shinfo(skb);
struct ocelot_port *ocelot_port = &priv->port;
struct ocelot *ocelot = ocelot_port->ocelot;
u32 val, ifh[OCELOT_TAG_LEN / 4];
struct frame_info info = {};
u8 grp = 0; /* Send everything on CPU group 0 */
unsigned int i, count, last;
int port = priv->chip_port;
val = ocelot_read(ocelot, QS_INJ_STATUS);
if (!(val & QS_INJ_STATUS_FIFO_RDY(BIT(grp))) ||
(val & QS_INJ_STATUS_WMARK_REACHED(BIT(grp))))
return NETDEV_TX_BUSY;
ocelot_write_rix(ocelot, QS_INJ_CTRL_GAP_SIZE(1) |
QS_INJ_CTRL_SOF, QS_INJ_CTRL, grp);
info.port = BIT(port);
info.tag_type = IFH_TAG_TYPE_C;
info.vid = skb_vlan_tag_get(skb);
/* Check if timestamping is needed */
net: mscc: ocelot: always pass skb clone to ocelot_port_add_txtstamp_skb Currently, ocelot switchdev passes the skb directly to the function that enqueues it to the list of skb's awaiting a TX timestamp. Whereas the felix DSA driver first clones the skb, then passes the clone to this queue. This matters because in the case of felix, the common IRQ handler, which is ocelot_get_txtstamp(), currently clones the clone, and frees the original clone. This is useless and can be simplified by using skb_complete_tx_timestamp() instead of skb_tstamp_tx(). Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Acked-by: Richard Cochran <richardcochran@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2020-09-23 14:24:20 +03:00
if (ocelot->ptp && (shinfo->tx_flags & SKBTX_HW_TSTAMP)) {
info.rew_op = ocelot_port->ptp_cmd;
if (ocelot_port->ptp_cmd == IFH_REW_OP_TWO_STEP_PTP) {
struct sk_buff *clone;
clone = skb_clone_sk(skb);
if (!clone) {
kfree_skb(skb);
return NETDEV_TX_OK;
}
ocelot_port_add_txtstamp_skb(ocelot, port, clone);
info.rew_op |= clone->cb[0] << 3;
}
}
net: mscc: ocelot: fix race condition with TX timestamping The TX-timestampable skb is added late to the ocelot_port->tx_skbs. It is in a race with the TX timestamp IRQ, which checks that queue trying to match the timestamp with the skb by the ts_id. The skb should be added to the queue before the IRQ can fire. Fixes: 4e3b0468e6d7 ("net: mscc: PTP Hardware Clock (PHC) support") Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Reviewed-by: Horatiu Vultur <horatiu.vultur@microchip.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2020-09-18 04:07:23 +03:00
net: mscc: ocelot: move net_device related functions to ocelot_net.c The ocelot hardware library shouldn't contain too much net_device specific code, since it is shared with DSA which abstracts that structure away. So much as much of this code as possible into the mscc_ocelot driver and outside of the common library. We're making an exception for MDB and LAG code. That is not yet exported to DSA, but when it will, most of the code that's already in ocelot.c will remain there. So, there's no point in moving code to ocelot_net.c just to move it back later. We could have moved all net_device code to ocelot_vsc7514.c directly, but let's operate under the assumption that if a new switchdev ocelot driver gets added, it'll define its SoC-specific stuff in a new ocelot_vsc*.c file and it'll reuse the rest of the code. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2020-06-20 18:43:44 +03:00
if (ocelot->ptp && shinfo->tx_flags & SKBTX_HW_TSTAMP) {
info.rew_op = ocelot_port->ptp_cmd;
net: mscc: ocelot: add locking for the port TX timestamp ID The ocelot_port->ts_id is used to: (a) populate skb->cb[0] for matching the TX timestamp in the PTP IRQ with an skb. (b) populate the REW_OP from the injection header of the ongoing skb. Only then is ocelot_port->ts_id incremented. This is a problem because, at least theoretically, another timestampable skb might use the same ocelot_port->ts_id before that is incremented. Normally all transmit calls are serialized by the netdev transmit spinlock, but in this case, ocelot_port_add_txtstamp_skb() is also called by DSA, which has started declaring the NETIF_F_LLTX feature since commit 2b86cb829976 ("net: dsa: declare lockless TX feature for slave ports"). So the logic of using and incrementing the timestamp id should be atomic per port. The solution is to use the global ocelot_port->ts_id only while protected by the associated ocelot_port->ts_id_lock. That's where we populate skb->cb[0]. Note that for ocelot, ocelot_port_add_txtstamp_skb is called for the actual skb, but for felix, it is called for the skb's clone. That is something which will also be changed in the future. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Reviewed-by: Horatiu Vultur <horatiu.vultur@microchip.com> Reviewed-by: Florian Fainelli <f.fainelli@gmail.com> Tested-by: Alexandre Belloni <alexandre.belloni@bootlin.com> Reviewed-by: Alexandre Belloni <alexandre.belloni@bootlin.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2020-09-18 04:07:24 +03:00
if (ocelot_port->ptp_cmd == IFH_REW_OP_TWO_STEP_PTP)
info.rew_op |= skb->cb[0] << 3;
net: mscc: ocelot: move net_device related functions to ocelot_net.c The ocelot hardware library shouldn't contain too much net_device specific code, since it is shared with DSA which abstracts that structure away. So much as much of this code as possible into the mscc_ocelot driver and outside of the common library. We're making an exception for MDB and LAG code. That is not yet exported to DSA, but when it will, most of the code that's already in ocelot.c will remain there. So, there's no point in moving code to ocelot_net.c just to move it back later. We could have moved all net_device code to ocelot_vsc7514.c directly, but let's operate under the assumption that if a new switchdev ocelot driver gets added, it'll define its SoC-specific stuff in a new ocelot_vsc*.c file and it'll reuse the rest of the code. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2020-06-20 18:43:44 +03:00
}
ocelot_gen_ifh(ifh, &info);
for (i = 0; i < OCELOT_TAG_LEN / 4; i++)
ocelot_write_rix(ocelot, (__force u32)cpu_to_be32(ifh[i]),
QS_INJ_WR, grp);
count = (skb->len + 3) / 4;
last = skb->len % 4;
for (i = 0; i < count; i++)
ocelot_write_rix(ocelot, ((u32 *)skb->data)[i], QS_INJ_WR, grp);
/* Add padding */
while (i < (OCELOT_BUFFER_CELL_SZ / 4)) {
ocelot_write_rix(ocelot, 0, QS_INJ_WR, grp);
i++;
}
/* Indicate EOF and valid bytes in last word */
ocelot_write_rix(ocelot, QS_INJ_CTRL_GAP_SIZE(1) |
QS_INJ_CTRL_VLD_BYTES(skb->len < OCELOT_BUFFER_CELL_SZ ? 0 : last) |
QS_INJ_CTRL_EOF,
QS_INJ_CTRL, grp);
/* Add dummy CRC */
ocelot_write_rix(ocelot, 0, QS_INJ_WR, grp);
skb_tx_timestamp(skb);
dev->stats.tx_packets++;
dev->stats.tx_bytes += skb->len;
net: mscc: ocelot: always pass skb clone to ocelot_port_add_txtstamp_skb Currently, ocelot switchdev passes the skb directly to the function that enqueues it to the list of skb's awaiting a TX timestamp. Whereas the felix DSA driver first clones the skb, then passes the clone to this queue. This matters because in the case of felix, the common IRQ handler, which is ocelot_get_txtstamp(), currently clones the clone, and frees the original clone. This is useless and can be simplified by using skb_complete_tx_timestamp() instead of skb_tstamp_tx(). Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Acked-by: Richard Cochran <richardcochran@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2020-09-23 14:24:20 +03:00
kfree_skb(skb);
net: mscc: ocelot: move net_device related functions to ocelot_net.c The ocelot hardware library shouldn't contain too much net_device specific code, since it is shared with DSA which abstracts that structure away. So much as much of this code as possible into the mscc_ocelot driver and outside of the common library. We're making an exception for MDB and LAG code. That is not yet exported to DSA, but when it will, most of the code that's already in ocelot.c will remain there. So, there's no point in moving code to ocelot_net.c just to move it back later. We could have moved all net_device code to ocelot_vsc7514.c directly, but let's operate under the assumption that if a new switchdev ocelot driver gets added, it'll define its SoC-specific stuff in a new ocelot_vsc*.c file and it'll reuse the rest of the code. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2020-06-20 18:43:44 +03:00
return NETDEV_TX_OK;
}
net: mscc: ocelot: install MAC addresses in .ndo_set_rx_mode from process context Currently ocelot_set_rx_mode calls ocelot_mact_learn directly, which has a very nice ocelot_mact_wait_for_completion at the end. Introduced in commit 639c1b2625af ("net: mscc: ocelot: Register poll timeout should be wall time not attempts"), this function uses readx_poll_timeout which triggers a lot of lockdep warnings and is also dangerous to use from atomic context, potentially leading to lockups and panics. Steen Hegelund added a poll timeout of 100 ms for checking the MAC table, a duration which is clearly absurd to poll in atomic context. So we need to defer the MAC table access to process context, which we do via a dynamically allocated workqueue which contains all there is to know about the MAC table operation it has to do. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Reviewed-by: Florian Fainelli <f.fainelli@gmail.com> Link: https://lore.kernel.org/r/20201212191612.222019-1-vladimir.oltean@nxp.com Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2020-12-12 21:16:12 +02:00
enum ocelot_action_type {
OCELOT_MACT_LEARN,
OCELOT_MACT_FORGET,
};
struct ocelot_mact_work_ctx {
struct work_struct work;
struct ocelot *ocelot;
enum ocelot_action_type type;
union {
/* OCELOT_MACT_LEARN */
struct {
unsigned char addr[ETH_ALEN];
u16 vid;
enum macaccess_entry_type entry_type;
int pgid;
} learn;
/* OCELOT_MACT_FORGET */
struct {
unsigned char addr[ETH_ALEN];
u16 vid;
} forget;
};
};
#define ocelot_work_to_ctx(x) \
container_of((x), struct ocelot_mact_work_ctx, work)
static void ocelot_mact_work(struct work_struct *work)
{
struct ocelot_mact_work_ctx *w = ocelot_work_to_ctx(work);
struct ocelot *ocelot = w->ocelot;
switch (w->type) {
case OCELOT_MACT_LEARN:
ocelot_mact_learn(ocelot, w->learn.pgid, w->learn.addr,
w->learn.vid, w->learn.entry_type);
break;
case OCELOT_MACT_FORGET:
ocelot_mact_forget(ocelot, w->forget.addr, w->forget.vid);
break;
default:
break;
net: mscc: ocelot: Remove unneeded semicolon fix semicolon.cocci warnings: drivers/net/ethernet/mscc/ocelot_net.c:460:2-3: Unneeded semicolon Signed-off-by: Xu Wang <vulab@iscas.ac.cn> Link: https://lore.kernel.org/r/20210115095544.33164-1-vulab@iscas.ac.cn Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2021-01-15 09:55:44 +00:00
}
net: mscc: ocelot: install MAC addresses in .ndo_set_rx_mode from process context Currently ocelot_set_rx_mode calls ocelot_mact_learn directly, which has a very nice ocelot_mact_wait_for_completion at the end. Introduced in commit 639c1b2625af ("net: mscc: ocelot: Register poll timeout should be wall time not attempts"), this function uses readx_poll_timeout which triggers a lot of lockdep warnings and is also dangerous to use from atomic context, potentially leading to lockups and panics. Steen Hegelund added a poll timeout of 100 ms for checking the MAC table, a duration which is clearly absurd to poll in atomic context. So we need to defer the MAC table access to process context, which we do via a dynamically allocated workqueue which contains all there is to know about the MAC table operation it has to do. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Reviewed-by: Florian Fainelli <f.fainelli@gmail.com> Link: https://lore.kernel.org/r/20201212191612.222019-1-vladimir.oltean@nxp.com Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2020-12-12 21:16:12 +02:00
kfree(w);
}
static int ocelot_enqueue_mact_action(struct ocelot *ocelot,
const struct ocelot_mact_work_ctx *ctx)
{
struct ocelot_mact_work_ctx *w = kmemdup(ctx, sizeof(*w), GFP_ATOMIC);
if (!w)
return -ENOMEM;
w->ocelot = ocelot;
INIT_WORK(&w->work, ocelot_mact_work);
queue_work(ocelot->owq, &w->work);
return 0;
}
net: mscc: ocelot: move net_device related functions to ocelot_net.c The ocelot hardware library shouldn't contain too much net_device specific code, since it is shared with DSA which abstracts that structure away. So much as much of this code as possible into the mscc_ocelot driver and outside of the common library. We're making an exception for MDB and LAG code. That is not yet exported to DSA, but when it will, most of the code that's already in ocelot.c will remain there. So, there's no point in moving code to ocelot_net.c just to move it back later. We could have moved all net_device code to ocelot_vsc7514.c directly, but let's operate under the assumption that if a new switchdev ocelot driver gets added, it'll define its SoC-specific stuff in a new ocelot_vsc*.c file and it'll reuse the rest of the code. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2020-06-20 18:43:44 +03:00
static int ocelot_mc_unsync(struct net_device *dev, const unsigned char *addr)
{
struct ocelot_port_private *priv = netdev_priv(dev);
struct ocelot_port *ocelot_port = &priv->port;
struct ocelot *ocelot = ocelot_port->ocelot;
net: mscc: ocelot: install MAC addresses in .ndo_set_rx_mode from process context Currently ocelot_set_rx_mode calls ocelot_mact_learn directly, which has a very nice ocelot_mact_wait_for_completion at the end. Introduced in commit 639c1b2625af ("net: mscc: ocelot: Register poll timeout should be wall time not attempts"), this function uses readx_poll_timeout which triggers a lot of lockdep warnings and is also dangerous to use from atomic context, potentially leading to lockups and panics. Steen Hegelund added a poll timeout of 100 ms for checking the MAC table, a duration which is clearly absurd to poll in atomic context. So we need to defer the MAC table access to process context, which we do via a dynamically allocated workqueue which contains all there is to know about the MAC table operation it has to do. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Reviewed-by: Florian Fainelli <f.fainelli@gmail.com> Link: https://lore.kernel.org/r/20201212191612.222019-1-vladimir.oltean@nxp.com Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2020-12-12 21:16:12 +02:00
struct ocelot_mact_work_ctx w;
ether_addr_copy(w.forget.addr, addr);
w.forget.vid = ocelot_port->pvid_vlan.vid;
w.type = OCELOT_MACT_FORGET;
net: mscc: ocelot: move net_device related functions to ocelot_net.c The ocelot hardware library shouldn't contain too much net_device specific code, since it is shared with DSA which abstracts that structure away. So much as much of this code as possible into the mscc_ocelot driver and outside of the common library. We're making an exception for MDB and LAG code. That is not yet exported to DSA, but when it will, most of the code that's already in ocelot.c will remain there. So, there's no point in moving code to ocelot_net.c just to move it back later. We could have moved all net_device code to ocelot_vsc7514.c directly, but let's operate under the assumption that if a new switchdev ocelot driver gets added, it'll define its SoC-specific stuff in a new ocelot_vsc*.c file and it'll reuse the rest of the code. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2020-06-20 18:43:44 +03:00
net: mscc: ocelot: install MAC addresses in .ndo_set_rx_mode from process context Currently ocelot_set_rx_mode calls ocelot_mact_learn directly, which has a very nice ocelot_mact_wait_for_completion at the end. Introduced in commit 639c1b2625af ("net: mscc: ocelot: Register poll timeout should be wall time not attempts"), this function uses readx_poll_timeout which triggers a lot of lockdep warnings and is also dangerous to use from atomic context, potentially leading to lockups and panics. Steen Hegelund added a poll timeout of 100 ms for checking the MAC table, a duration which is clearly absurd to poll in atomic context. So we need to defer the MAC table access to process context, which we do via a dynamically allocated workqueue which contains all there is to know about the MAC table operation it has to do. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Reviewed-by: Florian Fainelli <f.fainelli@gmail.com> Link: https://lore.kernel.org/r/20201212191612.222019-1-vladimir.oltean@nxp.com Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2020-12-12 21:16:12 +02:00
return ocelot_enqueue_mact_action(ocelot, &w);
net: mscc: ocelot: move net_device related functions to ocelot_net.c The ocelot hardware library shouldn't contain too much net_device specific code, since it is shared with DSA which abstracts that structure away. So much as much of this code as possible into the mscc_ocelot driver and outside of the common library. We're making an exception for MDB and LAG code. That is not yet exported to DSA, but when it will, most of the code that's already in ocelot.c will remain there. So, there's no point in moving code to ocelot_net.c just to move it back later. We could have moved all net_device code to ocelot_vsc7514.c directly, but let's operate under the assumption that if a new switchdev ocelot driver gets added, it'll define its SoC-specific stuff in a new ocelot_vsc*.c file and it'll reuse the rest of the code. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2020-06-20 18:43:44 +03:00
}
static int ocelot_mc_sync(struct net_device *dev, const unsigned char *addr)
{
struct ocelot_port_private *priv = netdev_priv(dev);
struct ocelot_port *ocelot_port = &priv->port;
struct ocelot *ocelot = ocelot_port->ocelot;
net: mscc: ocelot: install MAC addresses in .ndo_set_rx_mode from process context Currently ocelot_set_rx_mode calls ocelot_mact_learn directly, which has a very nice ocelot_mact_wait_for_completion at the end. Introduced in commit 639c1b2625af ("net: mscc: ocelot: Register poll timeout should be wall time not attempts"), this function uses readx_poll_timeout which triggers a lot of lockdep warnings and is also dangerous to use from atomic context, potentially leading to lockups and panics. Steen Hegelund added a poll timeout of 100 ms for checking the MAC table, a duration which is clearly absurd to poll in atomic context. So we need to defer the MAC table access to process context, which we do via a dynamically allocated workqueue which contains all there is to know about the MAC table operation it has to do. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Reviewed-by: Florian Fainelli <f.fainelli@gmail.com> Link: https://lore.kernel.org/r/20201212191612.222019-1-vladimir.oltean@nxp.com Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2020-12-12 21:16:12 +02:00
struct ocelot_mact_work_ctx w;
ether_addr_copy(w.learn.addr, addr);
w.learn.vid = ocelot_port->pvid_vlan.vid;
w.learn.pgid = PGID_CPU;
w.learn.entry_type = ENTRYTYPE_LOCKED;
w.type = OCELOT_MACT_LEARN;
net: mscc: ocelot: move net_device related functions to ocelot_net.c The ocelot hardware library shouldn't contain too much net_device specific code, since it is shared with DSA which abstracts that structure away. So much as much of this code as possible into the mscc_ocelot driver and outside of the common library. We're making an exception for MDB and LAG code. That is not yet exported to DSA, but when it will, most of the code that's already in ocelot.c will remain there. So, there's no point in moving code to ocelot_net.c just to move it back later. We could have moved all net_device code to ocelot_vsc7514.c directly, but let's operate under the assumption that if a new switchdev ocelot driver gets added, it'll define its SoC-specific stuff in a new ocelot_vsc*.c file and it'll reuse the rest of the code. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2020-06-20 18:43:44 +03:00
net: mscc: ocelot: install MAC addresses in .ndo_set_rx_mode from process context Currently ocelot_set_rx_mode calls ocelot_mact_learn directly, which has a very nice ocelot_mact_wait_for_completion at the end. Introduced in commit 639c1b2625af ("net: mscc: ocelot: Register poll timeout should be wall time not attempts"), this function uses readx_poll_timeout which triggers a lot of lockdep warnings and is also dangerous to use from atomic context, potentially leading to lockups and panics. Steen Hegelund added a poll timeout of 100 ms for checking the MAC table, a duration which is clearly absurd to poll in atomic context. So we need to defer the MAC table access to process context, which we do via a dynamically allocated workqueue which contains all there is to know about the MAC table operation it has to do. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Reviewed-by: Florian Fainelli <f.fainelli@gmail.com> Link: https://lore.kernel.org/r/20201212191612.222019-1-vladimir.oltean@nxp.com Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2020-12-12 21:16:12 +02:00
return ocelot_enqueue_mact_action(ocelot, &w);
net: mscc: ocelot: move net_device related functions to ocelot_net.c The ocelot hardware library shouldn't contain too much net_device specific code, since it is shared with DSA which abstracts that structure away. So much as much of this code as possible into the mscc_ocelot driver and outside of the common library. We're making an exception for MDB and LAG code. That is not yet exported to DSA, but when it will, most of the code that's already in ocelot.c will remain there. So, there's no point in moving code to ocelot_net.c just to move it back later. We could have moved all net_device code to ocelot_vsc7514.c directly, but let's operate under the assumption that if a new switchdev ocelot driver gets added, it'll define its SoC-specific stuff in a new ocelot_vsc*.c file and it'll reuse the rest of the code. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2020-06-20 18:43:44 +03:00
}
static void ocelot_set_rx_mode(struct net_device *dev)
{
struct ocelot_port_private *priv = netdev_priv(dev);
struct ocelot *ocelot = priv->port.ocelot;
u32 val;
int i;
/* This doesn't handle promiscuous mode because the bridge core is
* setting IFF_PROMISC on all slave interfaces and all frames would be
* forwarded to the CPU port.
*/
val = GENMASK(ocelot->num_phys_ports - 1, 0);
net: mscc: ocelot: introduce macros for iterating over PGIDs The current iterators are impossible to understand at first glance without switching back and forth between the definitions and their actual use in the for loops. So introduce some convenience names to help readability. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2020-06-21 14:46:02 +03:00
for_each_nonreserved_multicast_dest_pgid(ocelot, i)
net: mscc: ocelot: move net_device related functions to ocelot_net.c The ocelot hardware library shouldn't contain too much net_device specific code, since it is shared with DSA which abstracts that structure away. So much as much of this code as possible into the mscc_ocelot driver and outside of the common library. We're making an exception for MDB and LAG code. That is not yet exported to DSA, but when it will, most of the code that's already in ocelot.c will remain there. So, there's no point in moving code to ocelot_net.c just to move it back later. We could have moved all net_device code to ocelot_vsc7514.c directly, but let's operate under the assumption that if a new switchdev ocelot driver gets added, it'll define its SoC-specific stuff in a new ocelot_vsc*.c file and it'll reuse the rest of the code. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2020-06-20 18:43:44 +03:00
ocelot_write_rix(ocelot, val, ANA_PGID_PGID, i);
__dev_mc_sync(dev, ocelot_mc_sync, ocelot_mc_unsync);
}
static int ocelot_port_set_mac_address(struct net_device *dev, void *p)
{
struct ocelot_port_private *priv = netdev_priv(dev);
struct ocelot_port *ocelot_port = &priv->port;
struct ocelot *ocelot = ocelot_port->ocelot;
const struct sockaddr *addr = p;
/* Learn the new net device MAC address in the mac table. */
net: mscc: ocelot: transform the pvid and native vlan values into a structure This is a mechanical patch only. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2020-10-31 12:29:12 +02:00
ocelot_mact_learn(ocelot, PGID_CPU, addr->sa_data,
ocelot_port->pvid_vlan.vid, ENTRYTYPE_LOCKED);
net: mscc: ocelot: move net_device related functions to ocelot_net.c The ocelot hardware library shouldn't contain too much net_device specific code, since it is shared with DSA which abstracts that structure away. So much as much of this code as possible into the mscc_ocelot driver and outside of the common library. We're making an exception for MDB and LAG code. That is not yet exported to DSA, but when it will, most of the code that's already in ocelot.c will remain there. So, there's no point in moving code to ocelot_net.c just to move it back later. We could have moved all net_device code to ocelot_vsc7514.c directly, but let's operate under the assumption that if a new switchdev ocelot driver gets added, it'll define its SoC-specific stuff in a new ocelot_vsc*.c file and it'll reuse the rest of the code. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2020-06-20 18:43:44 +03:00
/* Then forget the previous one. */
net: mscc: ocelot: transform the pvid and native vlan values into a structure This is a mechanical patch only. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2020-10-31 12:29:12 +02:00
ocelot_mact_forget(ocelot, dev->dev_addr, ocelot_port->pvid_vlan.vid);
net: mscc: ocelot: move net_device related functions to ocelot_net.c The ocelot hardware library shouldn't contain too much net_device specific code, since it is shared with DSA which abstracts that structure away. So much as much of this code as possible into the mscc_ocelot driver and outside of the common library. We're making an exception for MDB and LAG code. That is not yet exported to DSA, but when it will, most of the code that's already in ocelot.c will remain there. So, there's no point in moving code to ocelot_net.c just to move it back later. We could have moved all net_device code to ocelot_vsc7514.c directly, but let's operate under the assumption that if a new switchdev ocelot driver gets added, it'll define its SoC-specific stuff in a new ocelot_vsc*.c file and it'll reuse the rest of the code. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2020-06-20 18:43:44 +03:00
ether_addr_copy(dev->dev_addr, addr->sa_data);
return 0;
}
static void ocelot_get_stats64(struct net_device *dev,
struct rtnl_link_stats64 *stats)
{
struct ocelot_port_private *priv = netdev_priv(dev);
struct ocelot *ocelot = priv->port.ocelot;
int port = priv->chip_port;
/* Configure the port to read the stats from */
ocelot_write(ocelot, SYS_STAT_CFG_STAT_VIEW(port),
SYS_STAT_CFG);
/* Get Rx stats */
stats->rx_bytes = ocelot_read(ocelot, SYS_COUNT_RX_OCTETS);
stats->rx_packets = ocelot_read(ocelot, SYS_COUNT_RX_SHORTS) +
ocelot_read(ocelot, SYS_COUNT_RX_FRAGMENTS) +
ocelot_read(ocelot, SYS_COUNT_RX_JABBERS) +
ocelot_read(ocelot, SYS_COUNT_RX_LONGS) +
ocelot_read(ocelot, SYS_COUNT_RX_64) +
ocelot_read(ocelot, SYS_COUNT_RX_65_127) +
ocelot_read(ocelot, SYS_COUNT_RX_128_255) +
ocelot_read(ocelot, SYS_COUNT_RX_256_1023) +
ocelot_read(ocelot, SYS_COUNT_RX_1024_1526) +
ocelot_read(ocelot, SYS_COUNT_RX_1527_MAX);
stats->multicast = ocelot_read(ocelot, SYS_COUNT_RX_MULTICAST);
stats->rx_dropped = dev->stats.rx_dropped;
/* Get Tx stats */
stats->tx_bytes = ocelot_read(ocelot, SYS_COUNT_TX_OCTETS);
stats->tx_packets = ocelot_read(ocelot, SYS_COUNT_TX_64) +
ocelot_read(ocelot, SYS_COUNT_TX_65_127) +
ocelot_read(ocelot, SYS_COUNT_TX_128_511) +
ocelot_read(ocelot, SYS_COUNT_TX_512_1023) +
ocelot_read(ocelot, SYS_COUNT_TX_1024_1526) +
ocelot_read(ocelot, SYS_COUNT_TX_1527_MAX);
stats->tx_dropped = ocelot_read(ocelot, SYS_COUNT_TX_DROPS) +
ocelot_read(ocelot, SYS_COUNT_TX_AGING);
stats->collisions = ocelot_read(ocelot, SYS_COUNT_TX_COLLISION);
}
static int ocelot_port_fdb_add(struct ndmsg *ndm, struct nlattr *tb[],
struct net_device *dev,
const unsigned char *addr,
u16 vid, u16 flags,
struct netlink_ext_ack *extack)
{
struct ocelot_port_private *priv = netdev_priv(dev);
struct ocelot *ocelot = priv->port.ocelot;
int port = priv->chip_port;
return ocelot_fdb_add(ocelot, port, addr, vid);
}
static int ocelot_port_fdb_del(struct ndmsg *ndm, struct nlattr *tb[],
struct net_device *dev,
const unsigned char *addr, u16 vid)
{
struct ocelot_port_private *priv = netdev_priv(dev);
struct ocelot *ocelot = priv->port.ocelot;
int port = priv->chip_port;
return ocelot_fdb_del(ocelot, port, addr, vid);
}
static int ocelot_port_fdb_dump(struct sk_buff *skb,
struct netlink_callback *cb,
struct net_device *dev,
struct net_device *filter_dev, int *idx)
{
struct ocelot_port_private *priv = netdev_priv(dev);
struct ocelot *ocelot = priv->port.ocelot;
struct ocelot_dump_ctx dump = {
.dev = dev,
.skb = skb,
.cb = cb,
.idx = *idx,
};
int port = priv->chip_port;
int ret;
ret = ocelot_fdb_dump(ocelot, port, ocelot_port_fdb_do_dump, &dump);
*idx = dump.idx;
return ret;
}
static int ocelot_vlan_rx_add_vid(struct net_device *dev, __be16 proto,
u16 vid)
{
return ocelot_vlan_vid_add(dev, vid, false, false);
}
static int ocelot_vlan_rx_kill_vid(struct net_device *dev, __be16 proto,
u16 vid)
{
return ocelot_vlan_vid_del(dev, vid);
}
static void ocelot_vlan_mode(struct ocelot *ocelot, int port,
netdev_features_t features)
{
u32 val;
/* Filtering */
val = ocelot_read(ocelot, ANA_VLANMASK);
if (features & NETIF_F_HW_VLAN_CTAG_FILTER)
val |= BIT(port);
else
val &= ~BIT(port);
ocelot_write(ocelot, val, ANA_VLANMASK);
}
static int ocelot_set_features(struct net_device *dev,
netdev_features_t features)
{
netdev_features_t changed = dev->features ^ features;
struct ocelot_port_private *priv = netdev_priv(dev);
struct ocelot *ocelot = priv->port.ocelot;
int port = priv->chip_port;
if ((dev->features & NETIF_F_HW_TC) > (features & NETIF_F_HW_TC) &&
priv->tc.offload_cnt) {
netdev_err(dev,
"Cannot disable HW TC offload while offloads active\n");
return -EBUSY;
}
if (changed & NETIF_F_HW_VLAN_CTAG_FILTER)
ocelot_vlan_mode(ocelot, port, features);
return 0;
}
static int ocelot_ioctl(struct net_device *dev, struct ifreq *ifr, int cmd)
{
struct ocelot_port_private *priv = netdev_priv(dev);
struct ocelot *ocelot = priv->port.ocelot;
int port = priv->chip_port;
/* If the attached PHY device isn't capable of timestamping operations,
* use our own (when possible).
*/
if (!phy_has_hwtstamp(dev->phydev) && ocelot->ptp) {
switch (cmd) {
case SIOCSHWTSTAMP:
return ocelot_hwstamp_set(ocelot, port, ifr);
case SIOCGHWTSTAMP:
return ocelot_hwstamp_get(ocelot, port, ifr);
}
}
return phy_mii_ioctl(dev->phydev, ifr, cmd);
}
static const struct net_device_ops ocelot_port_netdev_ops = {
.ndo_open = ocelot_port_open,
.ndo_stop = ocelot_port_stop,
.ndo_start_xmit = ocelot_port_xmit,
.ndo_set_rx_mode = ocelot_set_rx_mode,
.ndo_set_mac_address = ocelot_port_set_mac_address,
.ndo_get_stats64 = ocelot_get_stats64,
.ndo_fdb_add = ocelot_port_fdb_add,
.ndo_fdb_del = ocelot_port_fdb_del,
.ndo_fdb_dump = ocelot_port_fdb_dump,
.ndo_vlan_rx_add_vid = ocelot_vlan_rx_add_vid,
.ndo_vlan_rx_kill_vid = ocelot_vlan_rx_kill_vid,
.ndo_set_features = ocelot_set_features,
.ndo_setup_tc = ocelot_setup_tc,
.ndo_do_ioctl = ocelot_ioctl,
net: mscc: ocelot: register devlink ports Add devlink integration into the mscc_ocelot switchdev driver. All physical ports (i.e. the unused ones as well) except the CPU port module at ocelot->num_phys_ports are registered with devlink, and that requires keeping the devlink_port structure outside struct ocelot_port_private, since the latter has a 1:1 mapping with a struct net_device (which does not exist for unused ports). Since we use devlink_port_type_eth_set to link the devlink port to the net_device, we can as well remove the .ndo_get_phys_port_name and .ndo_get_port_parent_id implementations, since devlink takes care of retrieving the port name and number automatically, once .ndo_get_devlink_port is implemented. Note that the felix DSA driver is already integrated with devlink by default, since that is a thing that the DSA core takes care of. This is the reason why these devlink stubs were put in ocelot_net.c and not in the common library. It is also the reason why ocelot::devlink is a pointer and not a full structure embedded inside struct ocelot: because the mscc_ocelot driver allocates that by itself (as the container of struct ocelot, in fact), but in the case of felix, it is DSA who allocates the devlink, and felix just propagates the pointer towards struct ocelot. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Reviewed-by: Florian Fainelli <f.fainelli@gmail.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2021-01-15 04:11:18 +02:00
.ndo_get_devlink_port = ocelot_get_devlink_port,
net: mscc: ocelot: move net_device related functions to ocelot_net.c The ocelot hardware library shouldn't contain too much net_device specific code, since it is shared with DSA which abstracts that structure away. So much as much of this code as possible into the mscc_ocelot driver and outside of the common library. We're making an exception for MDB and LAG code. That is not yet exported to DSA, but when it will, most of the code that's already in ocelot.c will remain there. So, there's no point in moving code to ocelot_net.c just to move it back later. We could have moved all net_device code to ocelot_vsc7514.c directly, but let's operate under the assumption that if a new switchdev ocelot driver gets added, it'll define its SoC-specific stuff in a new ocelot_vsc*.c file and it'll reuse the rest of the code. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2020-06-20 18:43:44 +03:00
};
net: mscc: ocelot: introduce conversion helpers between port and netdev Since the mscc_ocelot_switch_lib is common between a pure switchdev and a DSA driver, the procedure of retrieving a net_device for a certain port index differs, as those are registered by their individual front-ends. Up to now that has been dealt with by always passing the port index to the switch library, but now, we're going to need to work with net_device pointers from the tc-flower offload, for things like indev, or mirred. It is not desirable to refactor that, so let's make sure that the flower offload core has the ability to translate between a net_device and a port index properly. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Acked-by: Alexandre Belloni <alexandre.belloni@bootlin.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2020-10-02 15:02:21 +03:00
struct net_device *ocelot_port_to_netdev(struct ocelot *ocelot, int port)
{
struct ocelot_port *ocelot_port = ocelot->ports[port];
struct ocelot_port_private *priv;
if (!ocelot_port)
return NULL;
priv = container_of(ocelot_port, struct ocelot_port_private, port);
return priv->dev;
}
net: mscc: ocelot: remove duplicate ocelot_port_dev_check A helper for checking whether a net_device belongs to mscc_ocelot already existed and did not need to be rewritten. Use it. Fixes: 319e4dd11a20 ("net: mscc: ocelot: introduce conversion helpers between port and netdev") Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Link: https://lore.kernel.org/r/20201011092041.3535101-1-vladimir.oltean@nxp.com Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2020-10-11 12:20:41 +03:00
/* Checks if the net_device instance given to us originates from our driver */
static bool ocelot_netdevice_dev_check(const struct net_device *dev)
net: mscc: ocelot: introduce conversion helpers between port and netdev Since the mscc_ocelot_switch_lib is common between a pure switchdev and a DSA driver, the procedure of retrieving a net_device for a certain port index differs, as those are registered by their individual front-ends. Up to now that has been dealt with by always passing the port index to the switch library, but now, we're going to need to work with net_device pointers from the tc-flower offload, for things like indev, or mirred. It is not desirable to refactor that, so let's make sure that the flower offload core has the ability to translate between a net_device and a port index properly. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Acked-by: Alexandre Belloni <alexandre.belloni@bootlin.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2020-10-02 15:02:21 +03:00
{
return dev->netdev_ops == &ocelot_port_netdev_ops;
}
int ocelot_netdev_to_port(struct net_device *dev)
{
struct ocelot_port_private *priv;
net: mscc: ocelot: remove duplicate ocelot_port_dev_check A helper for checking whether a net_device belongs to mscc_ocelot already existed and did not need to be rewritten. Use it. Fixes: 319e4dd11a20 ("net: mscc: ocelot: introduce conversion helpers between port and netdev") Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Link: https://lore.kernel.org/r/20201011092041.3535101-1-vladimir.oltean@nxp.com Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2020-10-11 12:20:41 +03:00
if (!dev || !ocelot_netdevice_dev_check(dev))
net: mscc: ocelot: introduce conversion helpers between port and netdev Since the mscc_ocelot_switch_lib is common between a pure switchdev and a DSA driver, the procedure of retrieving a net_device for a certain port index differs, as those are registered by their individual front-ends. Up to now that has been dealt with by always passing the port index to the switch library, but now, we're going to need to work with net_device pointers from the tc-flower offload, for things like indev, or mirred. It is not desirable to refactor that, so let's make sure that the flower offload core has the ability to translate between a net_device and a port index properly. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Acked-by: Alexandre Belloni <alexandre.belloni@bootlin.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2020-10-02 15:02:21 +03:00
return -EINVAL;
priv = netdev_priv(dev);
return priv->chip_port;
}
net: mscc: ocelot: move net_device related functions to ocelot_net.c The ocelot hardware library shouldn't contain too much net_device specific code, since it is shared with DSA which abstracts that structure away. So much as much of this code as possible into the mscc_ocelot driver and outside of the common library. We're making an exception for MDB and LAG code. That is not yet exported to DSA, but when it will, most of the code that's already in ocelot.c will remain there. So, there's no point in moving code to ocelot_net.c just to move it back later. We could have moved all net_device code to ocelot_vsc7514.c directly, but let's operate under the assumption that if a new switchdev ocelot driver gets added, it'll define its SoC-specific stuff in a new ocelot_vsc*.c file and it'll reuse the rest of the code. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2020-06-20 18:43:44 +03:00
static void ocelot_port_get_strings(struct net_device *netdev, u32 sset,
u8 *data)
{
struct ocelot_port_private *priv = netdev_priv(netdev);
struct ocelot *ocelot = priv->port.ocelot;
int port = priv->chip_port;
ocelot_get_strings(ocelot, port, sset, data);
}
static void ocelot_port_get_ethtool_stats(struct net_device *dev,
struct ethtool_stats *stats,
u64 *data)
{
struct ocelot_port_private *priv = netdev_priv(dev);
struct ocelot *ocelot = priv->port.ocelot;
int port = priv->chip_port;
ocelot_get_ethtool_stats(ocelot, port, data);
}
static int ocelot_port_get_sset_count(struct net_device *dev, int sset)
{
struct ocelot_port_private *priv = netdev_priv(dev);
struct ocelot *ocelot = priv->port.ocelot;
int port = priv->chip_port;
return ocelot_get_sset_count(ocelot, port, sset);
}
static int ocelot_port_get_ts_info(struct net_device *dev,
struct ethtool_ts_info *info)
{
struct ocelot_port_private *priv = netdev_priv(dev);
struct ocelot *ocelot = priv->port.ocelot;
int port = priv->chip_port;
if (!ocelot->ptp)
return ethtool_op_get_ts_info(dev, info);
return ocelot_get_ts_info(ocelot, port, info);
}
static const struct ethtool_ops ocelot_ethtool_ops = {
.get_strings = ocelot_port_get_strings,
.get_ethtool_stats = ocelot_port_get_ethtool_stats,
.get_sset_count = ocelot_port_get_sset_count,
.get_link_ksettings = phy_ethtool_get_link_ksettings,
.set_link_ksettings = phy_ethtool_set_link_ksettings,
.get_ts_info = ocelot_port_get_ts_info,
};
static void ocelot_port_attr_stp_state_set(struct ocelot *ocelot, int port,
u8 state)
{
ocelot_bridge_stp_state_set(ocelot, port, state);
}
static void ocelot_port_attr_ageing_set(struct ocelot *ocelot, int port,
unsigned long ageing_clock_t)
{
unsigned long ageing_jiffies = clock_t_to_jiffies(ageing_clock_t);
u32 ageing_time = jiffies_to_msecs(ageing_jiffies);
ocelot_set_ageing_time(ocelot, ageing_time);
}
static void ocelot_port_attr_mc_set(struct ocelot *ocelot, int port, bool mc)
{
u32 cpu_fwd_mcast = ANA_PORT_CPU_FWD_CFG_CPU_IGMP_REDIR_ENA |
ANA_PORT_CPU_FWD_CFG_CPU_MLD_REDIR_ENA |
ANA_PORT_CPU_FWD_CFG_CPU_IPMC_CTRL_COPY_ENA;
u32 val = 0;
if (mc)
val = cpu_fwd_mcast;
ocelot_rmw_gix(ocelot, val, cpu_fwd_mcast,
ANA_PORT_CPU_FWD_CFG, port);
}
static int ocelot_port_attr_set(struct net_device *dev,
net: switchdev: remove the transaction structure from port attributes Since the introduction of the switchdev API, port attributes were transmitted to drivers for offloading using a two-step transactional model, with a prepare phase that was supposed to catch all errors, and a commit phase that was supposed to never fail. Some classes of failures can never be avoided, like hardware access, or memory allocation. In the latter case, merely attempting to move the memory allocation to the preparation phase makes it impossible to avoid memory leaks, since commit 91cf8eceffc1 ("switchdev: Remove unused transaction item queue") which has removed the unused mechanism of passing on the allocated memory between one phase and another. It is time we admit that separating the preparation from the commit phase is something that is best left for the driver to decide, and not something that should be baked into the API, especially since there are no switchdev callers that depend on this. This patch removes the struct switchdev_trans member from switchdev port attribute notifier structures, and converts drivers to not look at this member. In part, this patch contains a revert of my previous commit 2e554a7a5d8a ("net: dsa: propagate switchdev vlan_filtering prepare phase to drivers"). For the most part, the conversion was trivial except for: - Rocker's world implementation based on Broadcom OF-DPA had an odd implementation of ofdpa_port_attr_bridge_flags_set. The conversion was done mechanically, by pasting the implementation twice, then only keeping the code that would get executed during prepare phase on top, then only keeping the code that gets executed during the commit phase on bottom, then simplifying the resulting code until this was obtained. - DSA's offloading of STP state, bridge flags, VLAN filtering and multicast router could be converted right away. But the ageing time could not, so a shim was introduced and this was left for a further commit. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Acked-by: Linus Walleij <linus.walleij@linaro.org> Acked-by: Jiri Pirko <jiri@nvidia.com> Reviewed-by: Kurt Kanzenbach <kurt@linutronix.de> # hellcreek Reviewed-by: Linus Walleij <linus.walleij@linaro.org> # RTL8366RB Reviewed-by: Ido Schimmel <idosch@nvidia.com> Reviewed-by: Florian Fainelli <f.fainelli@gmail.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2021-01-09 02:01:50 +02:00
const struct switchdev_attr *attr)
net: mscc: ocelot: move net_device related functions to ocelot_net.c The ocelot hardware library shouldn't contain too much net_device specific code, since it is shared with DSA which abstracts that structure away. So much as much of this code as possible into the mscc_ocelot driver and outside of the common library. We're making an exception for MDB and LAG code. That is not yet exported to DSA, but when it will, most of the code that's already in ocelot.c will remain there. So, there's no point in moving code to ocelot_net.c just to move it back later. We could have moved all net_device code to ocelot_vsc7514.c directly, but let's operate under the assumption that if a new switchdev ocelot driver gets added, it'll define its SoC-specific stuff in a new ocelot_vsc*.c file and it'll reuse the rest of the code. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2020-06-20 18:43:44 +03:00
{
struct ocelot_port_private *priv = netdev_priv(dev);
struct ocelot *ocelot = priv->port.ocelot;
int port = priv->chip_port;
int err = 0;
switch (attr->id) {
case SWITCHDEV_ATTR_ID_PORT_STP_STATE:
net: switchdev: remove the transaction structure from port attributes Since the introduction of the switchdev API, port attributes were transmitted to drivers for offloading using a two-step transactional model, with a prepare phase that was supposed to catch all errors, and a commit phase that was supposed to never fail. Some classes of failures can never be avoided, like hardware access, or memory allocation. In the latter case, merely attempting to move the memory allocation to the preparation phase makes it impossible to avoid memory leaks, since commit 91cf8eceffc1 ("switchdev: Remove unused transaction item queue") which has removed the unused mechanism of passing on the allocated memory between one phase and another. It is time we admit that separating the preparation from the commit phase is something that is best left for the driver to decide, and not something that should be baked into the API, especially since there are no switchdev callers that depend on this. This patch removes the struct switchdev_trans member from switchdev port attribute notifier structures, and converts drivers to not look at this member. In part, this patch contains a revert of my previous commit 2e554a7a5d8a ("net: dsa: propagate switchdev vlan_filtering prepare phase to drivers"). For the most part, the conversion was trivial except for: - Rocker's world implementation based on Broadcom OF-DPA had an odd implementation of ofdpa_port_attr_bridge_flags_set. The conversion was done mechanically, by pasting the implementation twice, then only keeping the code that would get executed during prepare phase on top, then only keeping the code that gets executed during the commit phase on bottom, then simplifying the resulting code until this was obtained. - DSA's offloading of STP state, bridge flags, VLAN filtering and multicast router could be converted right away. But the ageing time could not, so a shim was introduced and this was left for a further commit. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Acked-by: Linus Walleij <linus.walleij@linaro.org> Acked-by: Jiri Pirko <jiri@nvidia.com> Reviewed-by: Kurt Kanzenbach <kurt@linutronix.de> # hellcreek Reviewed-by: Linus Walleij <linus.walleij@linaro.org> # RTL8366RB Reviewed-by: Ido Schimmel <idosch@nvidia.com> Reviewed-by: Florian Fainelli <f.fainelli@gmail.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2021-01-09 02:01:50 +02:00
ocelot_port_attr_stp_state_set(ocelot, port, attr->u.stp_state);
net: mscc: ocelot: move net_device related functions to ocelot_net.c The ocelot hardware library shouldn't contain too much net_device specific code, since it is shared with DSA which abstracts that structure away. So much as much of this code as possible into the mscc_ocelot driver and outside of the common library. We're making an exception for MDB and LAG code. That is not yet exported to DSA, but when it will, most of the code that's already in ocelot.c will remain there. So, there's no point in moving code to ocelot_net.c just to move it back later. We could have moved all net_device code to ocelot_vsc7514.c directly, but let's operate under the assumption that if a new switchdev ocelot driver gets added, it'll define its SoC-specific stuff in a new ocelot_vsc*.c file and it'll reuse the rest of the code. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2020-06-20 18:43:44 +03:00
break;
case SWITCHDEV_ATTR_ID_BRIDGE_AGEING_TIME:
ocelot_port_attr_ageing_set(ocelot, port, attr->u.ageing_time);
break;
case SWITCHDEV_ATTR_ID_BRIDGE_VLAN_FILTERING:
net: switchdev: remove the transaction structure from port attributes Since the introduction of the switchdev API, port attributes were transmitted to drivers for offloading using a two-step transactional model, with a prepare phase that was supposed to catch all errors, and a commit phase that was supposed to never fail. Some classes of failures can never be avoided, like hardware access, or memory allocation. In the latter case, merely attempting to move the memory allocation to the preparation phase makes it impossible to avoid memory leaks, since commit 91cf8eceffc1 ("switchdev: Remove unused transaction item queue") which has removed the unused mechanism of passing on the allocated memory between one phase and another. It is time we admit that separating the preparation from the commit phase is something that is best left for the driver to decide, and not something that should be baked into the API, especially since there are no switchdev callers that depend on this. This patch removes the struct switchdev_trans member from switchdev port attribute notifier structures, and converts drivers to not look at this member. In part, this patch contains a revert of my previous commit 2e554a7a5d8a ("net: dsa: propagate switchdev vlan_filtering prepare phase to drivers"). For the most part, the conversion was trivial except for: - Rocker's world implementation based on Broadcom OF-DPA had an odd implementation of ofdpa_port_attr_bridge_flags_set. The conversion was done mechanically, by pasting the implementation twice, then only keeping the code that would get executed during prepare phase on top, then only keeping the code that gets executed during the commit phase on bottom, then simplifying the resulting code until this was obtained. - DSA's offloading of STP state, bridge flags, VLAN filtering and multicast router could be converted right away. But the ageing time could not, so a shim was introduced and this was left for a further commit. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Acked-by: Linus Walleij <linus.walleij@linaro.org> Acked-by: Jiri Pirko <jiri@nvidia.com> Reviewed-by: Kurt Kanzenbach <kurt@linutronix.de> # hellcreek Reviewed-by: Linus Walleij <linus.walleij@linaro.org> # RTL8366RB Reviewed-by: Ido Schimmel <idosch@nvidia.com> Reviewed-by: Florian Fainelli <f.fainelli@gmail.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2021-01-09 02:01:50 +02:00
ocelot_port_vlan_filtering(ocelot, port, attr->u.vlan_filtering);
net: mscc: ocelot: move net_device related functions to ocelot_net.c The ocelot hardware library shouldn't contain too much net_device specific code, since it is shared with DSA which abstracts that structure away. So much as much of this code as possible into the mscc_ocelot driver and outside of the common library. We're making an exception for MDB and LAG code. That is not yet exported to DSA, but when it will, most of the code that's already in ocelot.c will remain there. So, there's no point in moving code to ocelot_net.c just to move it back later. We could have moved all net_device code to ocelot_vsc7514.c directly, but let's operate under the assumption that if a new switchdev ocelot driver gets added, it'll define its SoC-specific stuff in a new ocelot_vsc*.c file and it'll reuse the rest of the code. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2020-06-20 18:43:44 +03:00
break;
case SWITCHDEV_ATTR_ID_BRIDGE_MC_DISABLED:
ocelot_port_attr_mc_set(ocelot, port, !attr->u.mc_disabled);
break;
default:
err = -EOPNOTSUPP;
break;
}
return err;
}
static int ocelot_port_obj_add_vlan(struct net_device *dev,
net: switchdev: remove the transaction structure from port object notifiers Since the introduction of the switchdev API, port objects were transmitted to drivers for offloading using a two-step transactional model, with a prepare phase that was supposed to catch all errors, and a commit phase that was supposed to never fail. Some classes of failures can never be avoided, like hardware access, or memory allocation. In the latter case, merely attempting to move the memory allocation to the preparation phase makes it impossible to avoid memory leaks, since commit 91cf8eceffc1 ("switchdev: Remove unused transaction item queue") which has removed the unused mechanism of passing on the allocated memory between one phase and another. It is time we admit that separating the preparation from the commit phase is something that is best left for the driver to decide, and not something that should be baked into the API, especially since there are no switchdev callers that depend on this. This patch removes the struct switchdev_trans member from switchdev port object notifier structures, and converts drivers to not look at this member. Where driver conversion is trivial (like in the case of the Marvell Prestera driver, NXP DPAA2 switch, TI CPSW, and Rocker drivers), it is done in this patch. Where driver conversion needs more attention (DSA, Mellanox Spectrum), the conversion is left for subsequent patches and here we only fake the prepare/commit phases at a lower level, just not in the switchdev notifier itself. Where the code has a natural structure that is best left alone as a preparation and a commit phase (as in the case of the Ocelot switch), that structure is left in place, just made to not depend upon the switchdev transactional model. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Reviewed-by: Florian Fainelli <f.fainelli@gmail.com> Acked-by: Linus Walleij <linus.walleij@linaro.org> Acked-by: Jiri Pirko <jiri@nvidia.com> Reviewed-by: Ido Schimmel <idosch@nvidia.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2021-01-09 02:01:48 +02:00
const struct switchdev_obj_port_vlan *vlan)
net: mscc: ocelot: move net_device related functions to ocelot_net.c The ocelot hardware library shouldn't contain too much net_device specific code, since it is shared with DSA which abstracts that structure away. So much as much of this code as possible into the mscc_ocelot driver and outside of the common library. We're making an exception for MDB and LAG code. That is not yet exported to DSA, but when it will, most of the code that's already in ocelot.c will remain there. So, there's no point in moving code to ocelot_net.c just to move it back later. We could have moved all net_device code to ocelot_vsc7514.c directly, but let's operate under the assumption that if a new switchdev ocelot driver gets added, it'll define its SoC-specific stuff in a new ocelot_vsc*.c file and it'll reuse the rest of the code. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2020-06-20 18:43:44 +03:00
{
net: switchdev: remove vid_begin -> vid_end range from VLAN objects The call path of a switchdev VLAN addition to the bridge looks something like this today: nbp_vlan_init | __br_vlan_set_default_pvid | | | | | br_afspec | | | | | | | v | | | br_process_vlan_info | | | | | | | v | | | br_vlan_info | | | / \ / | | / \ / | | / \ / | | / \ / v v v v v nbp_vlan_add br_vlan_add ------+ | ^ ^ | | | / | | | | / / / | \ br_vlan_get_master/ / v \ ^ / / br_vlan_add_existing \ | / / | \ | / / / \ | / / / \ | / / / \ | / / / v | | v / __vlan_add / / | / / | / v | / __vlan_vid_add | / \ | / v v v br_switchdev_port_vlan_add The ranges UAPI was introduced to the bridge in commit bdced7ef7838 ("bridge: support for multiple vlans and vlan ranges in setlink and dellink requests") (Jan 10 2015). But the VLAN ranges (parsed in br_afspec) have always been passed one by one, through struct bridge_vlan_info tmp_vinfo, to br_vlan_info. So the range never went too far in depth. Then Scott Feldman introduced the switchdev_port_bridge_setlink function in commit 47f8328bb1a4 ("switchdev: add new switchdev bridge setlink"). That marked the introduction of the SWITCHDEV_OBJ_PORT_VLAN, which made full use of the range. But switchdev_port_bridge_setlink was called like this: br_setlink -> br_afspec -> switchdev_port_bridge_setlink Basically, the switchdev and the bridge code were not tightly integrated. Then commit 41c498b9359e ("bridge: restore br_setlink back to original") came, and switchdev drivers were required to implement .ndo_bridge_setlink = switchdev_port_bridge_setlink for a while. In the meantime, commits such as 0944d6b5a2fa ("bridge: try switchdev op first in __vlan_vid_add/del") finally made switchdev penetrate the br_vlan_info() barrier and start to develop the call path we have today. But remember, br_vlan_info() still receives VLANs one by one. Then Arkadi Sharshevsky refactored the switchdev API in 2017 in commit 29ab586c3d83 ("net: switchdev: Remove bridge bypass support from switchdev") so that drivers would not implement .ndo_bridge_setlink any longer. The switchdev_port_bridge_setlink also got deleted. This refactoring removed the parallel bridge_setlink implementation from switchdev, and left the only switchdev VLAN objects to be the ones offloaded from __vlan_vid_add (basically RX filtering) and __vlan_add (the latter coming from commit 9c86ce2c1ae3 ("net: bridge: Notify about bridge VLANs")). That is to say, today the switchdev VLAN object ranges are not used in the kernel. Refactoring the above call path is a bit complicated, when the bridge VLAN call path is already a bit complicated. Let's go off and finish the job of commit 29ab586c3d83 by deleting the bogus iteration through the VLAN ranges from the drivers. Some aspects of this feature never made too much sense in the first place. For example, what is a range of VLANs all having the BRIDGE_VLAN_INFO_PVID flag supposed to mean, when a port can obviously have a single pvid? This particular configuration _is_ denied as of commit 6623c60dc28e ("bridge: vlan: enforce no pvid flag in vlan ranges"), but from an API perspective, the driver still has to play pretend, and only offload the vlan->vid_end as pvid. And the addition of a switchdev VLAN object can modify the flags of another, completely unrelated, switchdev VLAN object! (a VLAN that is PVID will invalidate the PVID flag from whatever other VLAN had previously been offloaded with switchdev and had that flag. Yet switchdev never notifies about that change, drivers are supposed to guess). Nonetheless, having a VLAN range in the API makes error handling look scarier than it really is - unwinding on errors and all of that. When in reality, no one really calls this API with more than one VLAN. It is all unnecessary complexity. And despite appearing pretentious (two-phase transactional model and all), the switchdev API is really sloppy because the VLAN addition and removal operations are not paired with one another (you can add a VLAN 100 times and delete it just once). The bridge notifies through switchdev of a VLAN addition not only when the flags of an existing VLAN change, but also when nothing changes. There are switchdev drivers out there who don't like adding a VLAN that has already been added, and those checks don't really belong at driver level. But the fact that the API contains ranges is yet another factor that prevents this from being addressed in the future. Of the existing switchdev pieces of hardware, it appears that only Mellanox Spectrum supports offloading more than one VLAN at a time, through mlxsw_sp_port_vlan_set. I have kept that code internal to the driver, because there is some more bookkeeping that makes use of it, but I deleted it from the switchdev API. But since the switchdev support for ranges has already been de facto deleted by a Mellanox employee and nobody noticed for 4 years, I'm going to assume it's not a biggie. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Reviewed-by: Ido Schimmel <idosch@nvidia.com> # switchdev and mlxsw Reviewed-by: Florian Fainelli <f.fainelli@gmail.com> Reviewed-by: Kurt Kanzenbach <kurt@linutronix.de> # hellcreek Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2021-01-09 02:01:46 +02:00
bool untagged = vlan->flags & BRIDGE_VLAN_INFO_UNTAGGED;
bool pvid = vlan->flags & BRIDGE_VLAN_INFO_PVID;
net: mscc: ocelot: move net_device related functions to ocelot_net.c The ocelot hardware library shouldn't contain too much net_device specific code, since it is shared with DSA which abstracts that structure away. So much as much of this code as possible into the mscc_ocelot driver and outside of the common library. We're making an exception for MDB and LAG code. That is not yet exported to DSA, but when it will, most of the code that's already in ocelot.c will remain there. So, there's no point in moving code to ocelot_net.c just to move it back later. We could have moved all net_device code to ocelot_vsc7514.c directly, but let's operate under the assumption that if a new switchdev ocelot driver gets added, it'll define its SoC-specific stuff in a new ocelot_vsc*.c file and it'll reuse the rest of the code. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2020-06-20 18:43:44 +03:00
int ret;
net: switchdev: remove the transaction structure from port object notifiers Since the introduction of the switchdev API, port objects were transmitted to drivers for offloading using a two-step transactional model, with a prepare phase that was supposed to catch all errors, and a commit phase that was supposed to never fail. Some classes of failures can never be avoided, like hardware access, or memory allocation. In the latter case, merely attempting to move the memory allocation to the preparation phase makes it impossible to avoid memory leaks, since commit 91cf8eceffc1 ("switchdev: Remove unused transaction item queue") which has removed the unused mechanism of passing on the allocated memory between one phase and another. It is time we admit that separating the preparation from the commit phase is something that is best left for the driver to decide, and not something that should be baked into the API, especially since there are no switchdev callers that depend on this. This patch removes the struct switchdev_trans member from switchdev port object notifier structures, and converts drivers to not look at this member. Where driver conversion is trivial (like in the case of the Marvell Prestera driver, NXP DPAA2 switch, TI CPSW, and Rocker drivers), it is done in this patch. Where driver conversion needs more attention (DSA, Mellanox Spectrum), the conversion is left for subsequent patches and here we only fake the prepare/commit phases at a lower level, just not in the switchdev notifier itself. Where the code has a natural structure that is best left alone as a preparation and a commit phase (as in the case of the Ocelot switch), that structure is left in place, just made to not depend upon the switchdev transactional model. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Reviewed-by: Florian Fainelli <f.fainelli@gmail.com> Acked-by: Linus Walleij <linus.walleij@linaro.org> Acked-by: Jiri Pirko <jiri@nvidia.com> Reviewed-by: Ido Schimmel <idosch@nvidia.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2021-01-09 02:01:48 +02:00
ret = ocelot_vlan_vid_prepare(dev, vlan->vid, pvid, untagged);
if (ret)
return ret;
net: mscc: ocelot: move net_device related functions to ocelot_net.c The ocelot hardware library shouldn't contain too much net_device specific code, since it is shared with DSA which abstracts that structure away. So much as much of this code as possible into the mscc_ocelot driver and outside of the common library. We're making an exception for MDB and LAG code. That is not yet exported to DSA, but when it will, most of the code that's already in ocelot.c will remain there. So, there's no point in moving code to ocelot_net.c just to move it back later. We could have moved all net_device code to ocelot_vsc7514.c directly, but let's operate under the assumption that if a new switchdev ocelot driver gets added, it'll define its SoC-specific stuff in a new ocelot_vsc*.c file and it'll reuse the rest of the code. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2020-06-20 18:43:44 +03:00
net: switchdev: remove the transaction structure from port object notifiers Since the introduction of the switchdev API, port objects were transmitted to drivers for offloading using a two-step transactional model, with a prepare phase that was supposed to catch all errors, and a commit phase that was supposed to never fail. Some classes of failures can never be avoided, like hardware access, or memory allocation. In the latter case, merely attempting to move the memory allocation to the preparation phase makes it impossible to avoid memory leaks, since commit 91cf8eceffc1 ("switchdev: Remove unused transaction item queue") which has removed the unused mechanism of passing on the allocated memory between one phase and another. It is time we admit that separating the preparation from the commit phase is something that is best left for the driver to decide, and not something that should be baked into the API, especially since there are no switchdev callers that depend on this. This patch removes the struct switchdev_trans member from switchdev port object notifier structures, and converts drivers to not look at this member. Where driver conversion is trivial (like in the case of the Marvell Prestera driver, NXP DPAA2 switch, TI CPSW, and Rocker drivers), it is done in this patch. Where driver conversion needs more attention (DSA, Mellanox Spectrum), the conversion is left for subsequent patches and here we only fake the prepare/commit phases at a lower level, just not in the switchdev notifier itself. Where the code has a natural structure that is best left alone as a preparation and a commit phase (as in the case of the Ocelot switch), that structure is left in place, just made to not depend upon the switchdev transactional model. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Reviewed-by: Florian Fainelli <f.fainelli@gmail.com> Acked-by: Linus Walleij <linus.walleij@linaro.org> Acked-by: Jiri Pirko <jiri@nvidia.com> Reviewed-by: Ido Schimmel <idosch@nvidia.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2021-01-09 02:01:48 +02:00
return ocelot_vlan_vid_add(dev, vlan->vid, pvid, untagged);
net: mscc: ocelot: move net_device related functions to ocelot_net.c The ocelot hardware library shouldn't contain too much net_device specific code, since it is shared with DSA which abstracts that structure away. So much as much of this code as possible into the mscc_ocelot driver and outside of the common library. We're making an exception for MDB and LAG code. That is not yet exported to DSA, but when it will, most of the code that's already in ocelot.c will remain there. So, there's no point in moving code to ocelot_net.c just to move it back later. We could have moved all net_device code to ocelot_vsc7514.c directly, but let's operate under the assumption that if a new switchdev ocelot driver gets added, it'll define its SoC-specific stuff in a new ocelot_vsc*.c file and it'll reuse the rest of the code. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2020-06-20 18:43:44 +03:00
}
net: dsa: felix: call port mdb operations from ocelot This adds the mdb hooks in felix and exports the mdb functions from ocelot. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2020-06-21 14:46:01 +03:00
static int ocelot_port_obj_add_mdb(struct net_device *dev,
net: switchdev: remove the transaction structure from port object notifiers Since the introduction of the switchdev API, port objects were transmitted to drivers for offloading using a two-step transactional model, with a prepare phase that was supposed to catch all errors, and a commit phase that was supposed to never fail. Some classes of failures can never be avoided, like hardware access, or memory allocation. In the latter case, merely attempting to move the memory allocation to the preparation phase makes it impossible to avoid memory leaks, since commit 91cf8eceffc1 ("switchdev: Remove unused transaction item queue") which has removed the unused mechanism of passing on the allocated memory between one phase and another. It is time we admit that separating the preparation from the commit phase is something that is best left for the driver to decide, and not something that should be baked into the API, especially since there are no switchdev callers that depend on this. This patch removes the struct switchdev_trans member from switchdev port object notifier structures, and converts drivers to not look at this member. Where driver conversion is trivial (like in the case of the Marvell Prestera driver, NXP DPAA2 switch, TI CPSW, and Rocker drivers), it is done in this patch. Where driver conversion needs more attention (DSA, Mellanox Spectrum), the conversion is left for subsequent patches and here we only fake the prepare/commit phases at a lower level, just not in the switchdev notifier itself. Where the code has a natural structure that is best left alone as a preparation and a commit phase (as in the case of the Ocelot switch), that structure is left in place, just made to not depend upon the switchdev transactional model. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Reviewed-by: Florian Fainelli <f.fainelli@gmail.com> Acked-by: Linus Walleij <linus.walleij@linaro.org> Acked-by: Jiri Pirko <jiri@nvidia.com> Reviewed-by: Ido Schimmel <idosch@nvidia.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2021-01-09 02:01:48 +02:00
const struct switchdev_obj_port_mdb *mdb)
net: dsa: felix: call port mdb operations from ocelot This adds the mdb hooks in felix and exports the mdb functions from ocelot. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2020-06-21 14:46:01 +03:00
{
struct ocelot_port_private *priv = netdev_priv(dev);
struct ocelot_port *ocelot_port = &priv->port;
struct ocelot *ocelot = ocelot_port->ocelot;
int port = priv->chip_port;
return ocelot_port_mdb_add(ocelot, port, mdb);
}
static int ocelot_port_obj_del_mdb(struct net_device *dev,
const struct switchdev_obj_port_mdb *mdb)
{
struct ocelot_port_private *priv = netdev_priv(dev);
struct ocelot_port *ocelot_port = &priv->port;
struct ocelot *ocelot = ocelot_port->ocelot;
int port = priv->chip_port;
return ocelot_port_mdb_del(ocelot, port, mdb);
}
net: mscc: ocelot: move net_device related functions to ocelot_net.c The ocelot hardware library shouldn't contain too much net_device specific code, since it is shared with DSA which abstracts that structure away. So much as much of this code as possible into the mscc_ocelot driver and outside of the common library. We're making an exception for MDB and LAG code. That is not yet exported to DSA, but when it will, most of the code that's already in ocelot.c will remain there. So, there's no point in moving code to ocelot_net.c just to move it back later. We could have moved all net_device code to ocelot_vsc7514.c directly, but let's operate under the assumption that if a new switchdev ocelot driver gets added, it'll define its SoC-specific stuff in a new ocelot_vsc*.c file and it'll reuse the rest of the code. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2020-06-20 18:43:44 +03:00
static int ocelot_port_obj_add(struct net_device *dev,
const struct switchdev_obj *obj,
struct netlink_ext_ack *extack)
{
int ret = 0;
switch (obj->id) {
case SWITCHDEV_OBJ_ID_PORT_VLAN:
ret = ocelot_port_obj_add_vlan(dev,
net: switchdev: remove the transaction structure from port object notifiers Since the introduction of the switchdev API, port objects were transmitted to drivers for offloading using a two-step transactional model, with a prepare phase that was supposed to catch all errors, and a commit phase that was supposed to never fail. Some classes of failures can never be avoided, like hardware access, or memory allocation. In the latter case, merely attempting to move the memory allocation to the preparation phase makes it impossible to avoid memory leaks, since commit 91cf8eceffc1 ("switchdev: Remove unused transaction item queue") which has removed the unused mechanism of passing on the allocated memory between one phase and another. It is time we admit that separating the preparation from the commit phase is something that is best left for the driver to decide, and not something that should be baked into the API, especially since there are no switchdev callers that depend on this. This patch removes the struct switchdev_trans member from switchdev port object notifier structures, and converts drivers to not look at this member. Where driver conversion is trivial (like in the case of the Marvell Prestera driver, NXP DPAA2 switch, TI CPSW, and Rocker drivers), it is done in this patch. Where driver conversion needs more attention (DSA, Mellanox Spectrum), the conversion is left for subsequent patches and here we only fake the prepare/commit phases at a lower level, just not in the switchdev notifier itself. Where the code has a natural structure that is best left alone as a preparation and a commit phase (as in the case of the Ocelot switch), that structure is left in place, just made to not depend upon the switchdev transactional model. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Reviewed-by: Florian Fainelli <f.fainelli@gmail.com> Acked-by: Linus Walleij <linus.walleij@linaro.org> Acked-by: Jiri Pirko <jiri@nvidia.com> Reviewed-by: Ido Schimmel <idosch@nvidia.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2021-01-09 02:01:48 +02:00
SWITCHDEV_OBJ_PORT_VLAN(obj));
net: mscc: ocelot: move net_device related functions to ocelot_net.c The ocelot hardware library shouldn't contain too much net_device specific code, since it is shared with DSA which abstracts that structure away. So much as much of this code as possible into the mscc_ocelot driver and outside of the common library. We're making an exception for MDB and LAG code. That is not yet exported to DSA, but when it will, most of the code that's already in ocelot.c will remain there. So, there's no point in moving code to ocelot_net.c just to move it back later. We could have moved all net_device code to ocelot_vsc7514.c directly, but let's operate under the assumption that if a new switchdev ocelot driver gets added, it'll define its SoC-specific stuff in a new ocelot_vsc*.c file and it'll reuse the rest of the code. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2020-06-20 18:43:44 +03:00
break;
case SWITCHDEV_OBJ_ID_PORT_MDB:
net: switchdev: remove the transaction structure from port object notifiers Since the introduction of the switchdev API, port objects were transmitted to drivers for offloading using a two-step transactional model, with a prepare phase that was supposed to catch all errors, and a commit phase that was supposed to never fail. Some classes of failures can never be avoided, like hardware access, or memory allocation. In the latter case, merely attempting to move the memory allocation to the preparation phase makes it impossible to avoid memory leaks, since commit 91cf8eceffc1 ("switchdev: Remove unused transaction item queue") which has removed the unused mechanism of passing on the allocated memory between one phase and another. It is time we admit that separating the preparation from the commit phase is something that is best left for the driver to decide, and not something that should be baked into the API, especially since there are no switchdev callers that depend on this. This patch removes the struct switchdev_trans member from switchdev port object notifier structures, and converts drivers to not look at this member. Where driver conversion is trivial (like in the case of the Marvell Prestera driver, NXP DPAA2 switch, TI CPSW, and Rocker drivers), it is done in this patch. Where driver conversion needs more attention (DSA, Mellanox Spectrum), the conversion is left for subsequent patches and here we only fake the prepare/commit phases at a lower level, just not in the switchdev notifier itself. Where the code has a natural structure that is best left alone as a preparation and a commit phase (as in the case of the Ocelot switch), that structure is left in place, just made to not depend upon the switchdev transactional model. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Reviewed-by: Florian Fainelli <f.fainelli@gmail.com> Acked-by: Linus Walleij <linus.walleij@linaro.org> Acked-by: Jiri Pirko <jiri@nvidia.com> Reviewed-by: Ido Schimmel <idosch@nvidia.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2021-01-09 02:01:48 +02:00
ret = ocelot_port_obj_add_mdb(dev, SWITCHDEV_OBJ_PORT_MDB(obj));
net: mscc: ocelot: move net_device related functions to ocelot_net.c The ocelot hardware library shouldn't contain too much net_device specific code, since it is shared with DSA which abstracts that structure away. So much as much of this code as possible into the mscc_ocelot driver and outside of the common library. We're making an exception for MDB and LAG code. That is not yet exported to DSA, but when it will, most of the code that's already in ocelot.c will remain there. So, there's no point in moving code to ocelot_net.c just to move it back later. We could have moved all net_device code to ocelot_vsc7514.c directly, but let's operate under the assumption that if a new switchdev ocelot driver gets added, it'll define its SoC-specific stuff in a new ocelot_vsc*.c file and it'll reuse the rest of the code. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2020-06-20 18:43:44 +03:00
break;
default:
return -EOPNOTSUPP;
}
return ret;
}
static int ocelot_port_obj_del(struct net_device *dev,
const struct switchdev_obj *obj)
{
int ret = 0;
switch (obj->id) {
case SWITCHDEV_OBJ_ID_PORT_VLAN:
net: switchdev: remove vid_begin -> vid_end range from VLAN objects The call path of a switchdev VLAN addition to the bridge looks something like this today: nbp_vlan_init | __br_vlan_set_default_pvid | | | | | br_afspec | | | | | | | v | | | br_process_vlan_info | | | | | | | v | | | br_vlan_info | | | / \ / | | / \ / | | / \ / | | / \ / v v v v v nbp_vlan_add br_vlan_add ------+ | ^ ^ | | | / | | | | / / / | \ br_vlan_get_master/ / v \ ^ / / br_vlan_add_existing \ | / / | \ | / / / \ | / / / \ | / / / \ | / / / v | | v / __vlan_add / / | / / | / v | / __vlan_vid_add | / \ | / v v v br_switchdev_port_vlan_add The ranges UAPI was introduced to the bridge in commit bdced7ef7838 ("bridge: support for multiple vlans and vlan ranges in setlink and dellink requests") (Jan 10 2015). But the VLAN ranges (parsed in br_afspec) have always been passed one by one, through struct bridge_vlan_info tmp_vinfo, to br_vlan_info. So the range never went too far in depth. Then Scott Feldman introduced the switchdev_port_bridge_setlink function in commit 47f8328bb1a4 ("switchdev: add new switchdev bridge setlink"). That marked the introduction of the SWITCHDEV_OBJ_PORT_VLAN, which made full use of the range. But switchdev_port_bridge_setlink was called like this: br_setlink -> br_afspec -> switchdev_port_bridge_setlink Basically, the switchdev and the bridge code were not tightly integrated. Then commit 41c498b9359e ("bridge: restore br_setlink back to original") came, and switchdev drivers were required to implement .ndo_bridge_setlink = switchdev_port_bridge_setlink for a while. In the meantime, commits such as 0944d6b5a2fa ("bridge: try switchdev op first in __vlan_vid_add/del") finally made switchdev penetrate the br_vlan_info() barrier and start to develop the call path we have today. But remember, br_vlan_info() still receives VLANs one by one. Then Arkadi Sharshevsky refactored the switchdev API in 2017 in commit 29ab586c3d83 ("net: switchdev: Remove bridge bypass support from switchdev") so that drivers would not implement .ndo_bridge_setlink any longer. The switchdev_port_bridge_setlink also got deleted. This refactoring removed the parallel bridge_setlink implementation from switchdev, and left the only switchdev VLAN objects to be the ones offloaded from __vlan_vid_add (basically RX filtering) and __vlan_add (the latter coming from commit 9c86ce2c1ae3 ("net: bridge: Notify about bridge VLANs")). That is to say, today the switchdev VLAN object ranges are not used in the kernel. Refactoring the above call path is a bit complicated, when the bridge VLAN call path is already a bit complicated. Let's go off and finish the job of commit 29ab586c3d83 by deleting the bogus iteration through the VLAN ranges from the drivers. Some aspects of this feature never made too much sense in the first place. For example, what is a range of VLANs all having the BRIDGE_VLAN_INFO_PVID flag supposed to mean, when a port can obviously have a single pvid? This particular configuration _is_ denied as of commit 6623c60dc28e ("bridge: vlan: enforce no pvid flag in vlan ranges"), but from an API perspective, the driver still has to play pretend, and only offload the vlan->vid_end as pvid. And the addition of a switchdev VLAN object can modify the flags of another, completely unrelated, switchdev VLAN object! (a VLAN that is PVID will invalidate the PVID flag from whatever other VLAN had previously been offloaded with switchdev and had that flag. Yet switchdev never notifies about that change, drivers are supposed to guess). Nonetheless, having a VLAN range in the API makes error handling look scarier than it really is - unwinding on errors and all of that. When in reality, no one really calls this API with more than one VLAN. It is all unnecessary complexity. And despite appearing pretentious (two-phase transactional model and all), the switchdev API is really sloppy because the VLAN addition and removal operations are not paired with one another (you can add a VLAN 100 times and delete it just once). The bridge notifies through switchdev of a VLAN addition not only when the flags of an existing VLAN change, but also when nothing changes. There are switchdev drivers out there who don't like adding a VLAN that has already been added, and those checks don't really belong at driver level. But the fact that the API contains ranges is yet another factor that prevents this from being addressed in the future. Of the existing switchdev pieces of hardware, it appears that only Mellanox Spectrum supports offloading more than one VLAN at a time, through mlxsw_sp_port_vlan_set. I have kept that code internal to the driver, because there is some more bookkeeping that makes use of it, but I deleted it from the switchdev API. But since the switchdev support for ranges has already been de facto deleted by a Mellanox employee and nobody noticed for 4 years, I'm going to assume it's not a biggie. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Reviewed-by: Ido Schimmel <idosch@nvidia.com> # switchdev and mlxsw Reviewed-by: Florian Fainelli <f.fainelli@gmail.com> Reviewed-by: Kurt Kanzenbach <kurt@linutronix.de> # hellcreek Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2021-01-09 02:01:46 +02:00
ret = ocelot_vlan_vid_del(dev,
SWITCHDEV_OBJ_PORT_VLAN(obj)->vid);
net: mscc: ocelot: move net_device related functions to ocelot_net.c The ocelot hardware library shouldn't contain too much net_device specific code, since it is shared with DSA which abstracts that structure away. So much as much of this code as possible into the mscc_ocelot driver and outside of the common library. We're making an exception for MDB and LAG code. That is not yet exported to DSA, but when it will, most of the code that's already in ocelot.c will remain there. So, there's no point in moving code to ocelot_net.c just to move it back later. We could have moved all net_device code to ocelot_vsc7514.c directly, but let's operate under the assumption that if a new switchdev ocelot driver gets added, it'll define its SoC-specific stuff in a new ocelot_vsc*.c file and it'll reuse the rest of the code. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2020-06-20 18:43:44 +03:00
break;
case SWITCHDEV_OBJ_ID_PORT_MDB:
ret = ocelot_port_obj_del_mdb(dev, SWITCHDEV_OBJ_PORT_MDB(obj));
break;
default:
return -EOPNOTSUPP;
}
return ret;
}
net: mscc: ocelot: rename ocelot_netdevice_port_event to ocelot_netdevice_changeupper ocelot_netdevice_port_event treats a single event, NETDEV_CHANGEUPPER. So we can remove the check for the type of event, and rename the function to be more suggestive, since there already is a function with a very similar name of ocelot_netdevice_event. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2021-02-06 00:02:10 +02:00
static int ocelot_netdevice_changeupper(struct net_device *dev,
struct netdev_notifier_changeupper_info *info)
net: mscc: ocelot: move net_device related functions to ocelot_net.c The ocelot hardware library shouldn't contain too much net_device specific code, since it is shared with DSA which abstracts that structure away. So much as much of this code as possible into the mscc_ocelot driver and outside of the common library. We're making an exception for MDB and LAG code. That is not yet exported to DSA, but when it will, most of the code that's already in ocelot.c will remain there. So, there's no point in moving code to ocelot_net.c just to move it back later. We could have moved all net_device code to ocelot_vsc7514.c directly, but let's operate under the assumption that if a new switchdev ocelot driver gets added, it'll define its SoC-specific stuff in a new ocelot_vsc*.c file and it'll reuse the rest of the code. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2020-06-20 18:43:44 +03:00
{
struct ocelot_port_private *priv = netdev_priv(dev);
struct ocelot_port *ocelot_port = &priv->port;
struct ocelot *ocelot = ocelot_port->ocelot;
int port = priv->chip_port;
int err = 0;
net: mscc: ocelot: rename ocelot_netdevice_port_event to ocelot_netdevice_changeupper ocelot_netdevice_port_event treats a single event, NETDEV_CHANGEUPPER. So we can remove the check for the type of event, and rename the function to be more suggestive, since there already is a function with a very similar name of ocelot_netdevice_event. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2021-02-06 00:02:10 +02:00
if (netif_is_bridge_master(info->upper_dev)) {
if (info->linking) {
err = ocelot_port_bridge_join(ocelot, port,
net: mscc: ocelot: move net_device related functions to ocelot_net.c The ocelot hardware library shouldn't contain too much net_device specific code, since it is shared with DSA which abstracts that structure away. So much as much of this code as possible into the mscc_ocelot driver and outside of the common library. We're making an exception for MDB and LAG code. That is not yet exported to DSA, but when it will, most of the code that's already in ocelot.c will remain there. So, there's no point in moving code to ocelot_net.c just to move it back later. We could have moved all net_device code to ocelot_vsc7514.c directly, but let's operate under the assumption that if a new switchdev ocelot driver gets added, it'll define its SoC-specific stuff in a new ocelot_vsc*.c file and it'll reuse the rest of the code. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2020-06-20 18:43:44 +03:00
info->upper_dev);
net: mscc: ocelot: rename ocelot_netdevice_port_event to ocelot_netdevice_changeupper ocelot_netdevice_port_event treats a single event, NETDEV_CHANGEUPPER. So we can remove the check for the type of event, and rename the function to be more suggestive, since there already is a function with a very similar name of ocelot_netdevice_event. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2021-02-06 00:02:10 +02:00
} else {
err = ocelot_port_bridge_leave(ocelot, port,
info->upper_dev);
net: mscc: ocelot: move net_device related functions to ocelot_net.c The ocelot hardware library shouldn't contain too much net_device specific code, since it is shared with DSA which abstracts that structure away. So much as much of this code as possible into the mscc_ocelot driver and outside of the common library. We're making an exception for MDB and LAG code. That is not yet exported to DSA, but when it will, most of the code that's already in ocelot.c will remain there. So, there's no point in moving code to ocelot_net.c just to move it back later. We could have moved all net_device code to ocelot_vsc7514.c directly, but let's operate under the assumption that if a new switchdev ocelot driver gets added, it'll define its SoC-specific stuff in a new ocelot_vsc*.c file and it'll reuse the rest of the code. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2020-06-20 18:43:44 +03:00
}
net: mscc: ocelot: rename ocelot_netdevice_port_event to ocelot_netdevice_changeupper ocelot_netdevice_port_event treats a single event, NETDEV_CHANGEUPPER. So we can remove the check for the type of event, and rename the function to be more suggestive, since there already is a function with a very similar name of ocelot_netdevice_event. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2021-02-06 00:02:10 +02:00
}
if (netif_is_lag_master(info->upper_dev)) {
net: mscc: ocelot: don't refuse bonding interfaces we can't offload Since switchdev/DSA exposes network interfaces that fulfill many of the same user space expectations that dedicated NICs do, it makes sense to not deny bonding interfaces with a bonding policy that we cannot offload, but instead allow the bonding driver to select the egress interface in software. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Reviewed-by: Alexandre Belloni <alexandre.belloni@bootlin.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2021-02-06 00:02:12 +02:00
if (info->linking) {
net: mscc: ocelot: rename ocelot_netdevice_port_event to ocelot_netdevice_changeupper ocelot_netdevice_port_event treats a single event, NETDEV_CHANGEUPPER. So we can remove the check for the type of event, and rename the function to be more suggestive, since there already is a function with a very similar name of ocelot_netdevice_event. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2021-02-06 00:02:10 +02:00
err = ocelot_port_lag_join(ocelot, port,
net: mscc: ocelot: don't refuse bonding interfaces we can't offload Since switchdev/DSA exposes network interfaces that fulfill many of the same user space expectations that dedicated NICs do, it makes sense to not deny bonding interfaces with a bonding policy that we cannot offload, but instead allow the bonding driver to select the egress interface in software. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Reviewed-by: Alexandre Belloni <alexandre.belloni@bootlin.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2021-02-06 00:02:12 +02:00
info->upper_dev,
info->upper_info);
if (err == -EOPNOTSUPP) {
NL_SET_ERR_MSG_MOD(info->info.extack,
"Offloading not supported");
err = 0;
}
} else {
net: mscc: ocelot: rename ocelot_netdevice_port_event to ocelot_netdevice_changeupper ocelot_netdevice_port_event treats a single event, NETDEV_CHANGEUPPER. So we can remove the check for the type of event, and rename the function to be more suggestive, since there already is a function with a very similar name of ocelot_netdevice_event. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2021-02-06 00:02:10 +02:00
ocelot_port_lag_leave(ocelot, port,
info->upper_dev);
net: mscc: ocelot: don't refuse bonding interfaces we can't offload Since switchdev/DSA exposes network interfaces that fulfill many of the same user space expectations that dedicated NICs do, it makes sense to not deny bonding interfaces with a bonding policy that we cannot offload, but instead allow the bonding driver to select the egress interface in software. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Reviewed-by: Alexandre Belloni <alexandre.belloni@bootlin.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2021-02-06 00:02:12 +02:00
}
net: mscc: ocelot: move net_device related functions to ocelot_net.c The ocelot hardware library shouldn't contain too much net_device specific code, since it is shared with DSA which abstracts that structure away. So much as much of this code as possible into the mscc_ocelot driver and outside of the common library. We're making an exception for MDB and LAG code. That is not yet exported to DSA, but when it will, most of the code that's already in ocelot.c will remain there. So, there's no point in moving code to ocelot_net.c just to move it back later. We could have moved all net_device code to ocelot_vsc7514.c directly, but let's operate under the assumption that if a new switchdev ocelot driver gets added, it'll define its SoC-specific stuff in a new ocelot_vsc*.c file and it'll reuse the rest of the code. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2020-06-20 18:43:44 +03:00
}
net: mscc: ocelot: use a switch-case statement in ocelot_netdevice_event Make ocelot's net device event handler more streamlined by structuring it in a similar way with others. The inspiration here was dsa_slave_netdevice_event. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Reviewed-by: Alexandre Belloni <alexandre.belloni@bootlin.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2021-02-06 00:02:11 +02:00
return notifier_from_errno(err);
}
static int
ocelot_netdevice_lag_changeupper(struct net_device *dev,
struct netdev_notifier_changeupper_info *info)
{
struct net_device *lower;
struct list_head *iter;
int err = NOTIFY_DONE;
netdev_for_each_lower_dev(dev, lower, iter) {
err = ocelot_netdevice_changeupper(lower, info);
if (err)
return notifier_from_errno(err);
}
return NOTIFY_DONE;
net: mscc: ocelot: move net_device related functions to ocelot_net.c The ocelot hardware library shouldn't contain too much net_device specific code, since it is shared with DSA which abstracts that structure away. So much as much of this code as possible into the mscc_ocelot driver and outside of the common library. We're making an exception for MDB and LAG code. That is not yet exported to DSA, but when it will, most of the code that's already in ocelot.c will remain there. So, there's no point in moving code to ocelot_net.c just to move it back later. We could have moved all net_device code to ocelot_vsc7514.c directly, but let's operate under the assumption that if a new switchdev ocelot driver gets added, it'll define its SoC-specific stuff in a new ocelot_vsc*.c file and it'll reuse the rest of the code. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2020-06-20 18:43:44 +03:00
}
net: mscc: ocelot: rebalance LAGs on link up/down events At present there is an issue when ocelot is offloading a bonding interface, but one of the links of the physical ports goes down. Traffic keeps being hashed towards that destination, and of course gets dropped on egress. Monitor the netdev notifier events emitted by the bonding driver for changes in the physical state of lower interfaces, to determine which ports are active and which ones are no longer. Then extend ocelot_get_bond_mask to return either the configured bonding interfaces, or the active ones, depending on a boolean argument. The code that does rebalancing only needs to do so among the active ports, whereas the bridge forwarding mask and the logical port IDs still need to look at the permanently bonded ports. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2021-02-06 00:02:19 +02:00
static int
ocelot_netdevice_changelowerstate(struct net_device *dev,
struct netdev_lag_lower_state_info *info)
{
struct ocelot_port_private *priv = netdev_priv(dev);
bool is_active = info->link_up && info->tx_enabled;
struct ocelot_port *ocelot_port = &priv->port;
struct ocelot *ocelot = ocelot_port->ocelot;
int port = priv->chip_port;
if (!ocelot_port->bond)
return NOTIFY_DONE;
if (ocelot_port->lag_tx_active == is_active)
return NOTIFY_DONE;
ocelot_port_lag_change(ocelot, port, is_active);
return NOTIFY_OK;
}
net: mscc: ocelot: move net_device related functions to ocelot_net.c The ocelot hardware library shouldn't contain too much net_device specific code, since it is shared with DSA which abstracts that structure away. So much as much of this code as possible into the mscc_ocelot driver and outside of the common library. We're making an exception for MDB and LAG code. That is not yet exported to DSA, but when it will, most of the code that's already in ocelot.c will remain there. So, there's no point in moving code to ocelot_net.c just to move it back later. We could have moved all net_device code to ocelot_vsc7514.c directly, but let's operate under the assumption that if a new switchdev ocelot driver gets added, it'll define its SoC-specific stuff in a new ocelot_vsc*.c file and it'll reuse the rest of the code. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2020-06-20 18:43:44 +03:00
static int ocelot_netdevice_event(struct notifier_block *unused,
unsigned long event, void *ptr)
{
struct net_device *dev = netdev_notifier_info_to_dev(ptr);
net: mscc: ocelot: use a switch-case statement in ocelot_netdevice_event Make ocelot's net device event handler more streamlined by structuring it in a similar way with others. The inspiration here was dsa_slave_netdevice_event. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Reviewed-by: Alexandre Belloni <alexandre.belloni@bootlin.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2021-02-06 00:02:11 +02:00
switch (event) {
case NETDEV_CHANGEUPPER: {
struct netdev_notifier_changeupper_info *info = ptr;
net: mscc: ocelot: move net_device related functions to ocelot_net.c The ocelot hardware library shouldn't contain too much net_device specific code, since it is shared with DSA which abstracts that structure away. So much as much of this code as possible into the mscc_ocelot driver and outside of the common library. We're making an exception for MDB and LAG code. That is not yet exported to DSA, but when it will, most of the code that's already in ocelot.c will remain there. So, there's no point in moving code to ocelot_net.c just to move it back later. We could have moved all net_device code to ocelot_vsc7514.c directly, but let's operate under the assumption that if a new switchdev ocelot driver gets added, it'll define its SoC-specific stuff in a new ocelot_vsc*.c file and it'll reuse the rest of the code. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2020-06-20 18:43:44 +03:00
net: mscc: ocelot: use a switch-case statement in ocelot_netdevice_event Make ocelot's net device event handler more streamlined by structuring it in a similar way with others. The inspiration here was dsa_slave_netdevice_event. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Reviewed-by: Alexandre Belloni <alexandre.belloni@bootlin.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2021-02-06 00:02:11 +02:00
if (ocelot_netdevice_dev_check(dev))
return ocelot_netdevice_changeupper(dev, info);
net: mscc: ocelot: move net_device related functions to ocelot_net.c The ocelot hardware library shouldn't contain too much net_device specific code, since it is shared with DSA which abstracts that structure away. So much as much of this code as possible into the mscc_ocelot driver and outside of the common library. We're making an exception for MDB and LAG code. That is not yet exported to DSA, but when it will, most of the code that's already in ocelot.c will remain there. So, there's no point in moving code to ocelot_net.c just to move it back later. We could have moved all net_device code to ocelot_vsc7514.c directly, but let's operate under the assumption that if a new switchdev ocelot driver gets added, it'll define its SoC-specific stuff in a new ocelot_vsc*.c file and it'll reuse the rest of the code. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2020-06-20 18:43:44 +03:00
net: mscc: ocelot: use a switch-case statement in ocelot_netdevice_event Make ocelot's net device event handler more streamlined by structuring it in a similar way with others. The inspiration here was dsa_slave_netdevice_event. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Reviewed-by: Alexandre Belloni <alexandre.belloni@bootlin.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2021-02-06 00:02:11 +02:00
if (netif_is_lag_master(dev))
return ocelot_netdevice_lag_changeupper(dev, info);
break;
}
net: mscc: ocelot: rebalance LAGs on link up/down events At present there is an issue when ocelot is offloading a bonding interface, but one of the links of the physical ports goes down. Traffic keeps being hashed towards that destination, and of course gets dropped on egress. Monitor the netdev notifier events emitted by the bonding driver for changes in the physical state of lower interfaces, to determine which ports are active and which ones are no longer. Then extend ocelot_get_bond_mask to return either the configured bonding interfaces, or the active ones, depending on a boolean argument. The code that does rebalancing only needs to do so among the active ports, whereas the bridge forwarding mask and the logical port IDs still need to look at the permanently bonded ports. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2021-02-06 00:02:19 +02:00
case NETDEV_CHANGELOWERSTATE: {
struct netdev_notifier_changelowerstate_info *info = ptr;
if (!ocelot_netdevice_dev_check(dev))
break;
return ocelot_netdevice_changelowerstate(dev,
info->lower_state_info);
}
net: mscc: ocelot: use a switch-case statement in ocelot_netdevice_event Make ocelot's net device event handler more streamlined by structuring it in a similar way with others. The inspiration here was dsa_slave_netdevice_event. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Reviewed-by: Alexandre Belloni <alexandre.belloni@bootlin.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2021-02-06 00:02:11 +02:00
default:
break;
net: mscc: ocelot: move net_device related functions to ocelot_net.c The ocelot hardware library shouldn't contain too much net_device specific code, since it is shared with DSA which abstracts that structure away. So much as much of this code as possible into the mscc_ocelot driver and outside of the common library. We're making an exception for MDB and LAG code. That is not yet exported to DSA, but when it will, most of the code that's already in ocelot.c will remain there. So, there's no point in moving code to ocelot_net.c just to move it back later. We could have moved all net_device code to ocelot_vsc7514.c directly, but let's operate under the assumption that if a new switchdev ocelot driver gets added, it'll define its SoC-specific stuff in a new ocelot_vsc*.c file and it'll reuse the rest of the code. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2020-06-20 18:43:44 +03:00
}
net: mscc: ocelot: use a switch-case statement in ocelot_netdevice_event Make ocelot's net device event handler more streamlined by structuring it in a similar way with others. The inspiration here was dsa_slave_netdevice_event. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Reviewed-by: Alexandre Belloni <alexandre.belloni@bootlin.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2021-02-06 00:02:11 +02:00
return NOTIFY_DONE;
net: mscc: ocelot: move net_device related functions to ocelot_net.c The ocelot hardware library shouldn't contain too much net_device specific code, since it is shared with DSA which abstracts that structure away. So much as much of this code as possible into the mscc_ocelot driver and outside of the common library. We're making an exception for MDB and LAG code. That is not yet exported to DSA, but when it will, most of the code that's already in ocelot.c will remain there. So, there's no point in moving code to ocelot_net.c just to move it back later. We could have moved all net_device code to ocelot_vsc7514.c directly, but let's operate under the assumption that if a new switchdev ocelot driver gets added, it'll define its SoC-specific stuff in a new ocelot_vsc*.c file and it'll reuse the rest of the code. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2020-06-20 18:43:44 +03:00
}
struct notifier_block ocelot_netdevice_nb __read_mostly = {
.notifier_call = ocelot_netdevice_event,
};
static int ocelot_switchdev_event(struct notifier_block *unused,
unsigned long event, void *ptr)
{
struct net_device *dev = switchdev_notifier_info_to_dev(ptr);
int err;
switch (event) {
case SWITCHDEV_PORT_ATTR_SET:
err = switchdev_handle_port_attr_set(dev, ptr,
ocelot_netdevice_dev_check,
ocelot_port_attr_set);
return notifier_from_errno(err);
}
return NOTIFY_DONE;
}
struct notifier_block ocelot_switchdev_nb __read_mostly = {
.notifier_call = ocelot_switchdev_event,
};
static int ocelot_switchdev_blocking_event(struct notifier_block *unused,
unsigned long event, void *ptr)
{
struct net_device *dev = switchdev_notifier_info_to_dev(ptr);
int err;
switch (event) {
/* Blocking events. */
case SWITCHDEV_PORT_OBJ_ADD:
err = switchdev_handle_port_obj_add(dev, ptr,
ocelot_netdevice_dev_check,
ocelot_port_obj_add);
return notifier_from_errno(err);
case SWITCHDEV_PORT_OBJ_DEL:
err = switchdev_handle_port_obj_del(dev, ptr,
ocelot_netdevice_dev_check,
ocelot_port_obj_del);
return notifier_from_errno(err);
case SWITCHDEV_PORT_ATTR_SET:
err = switchdev_handle_port_attr_set(dev, ptr,
ocelot_netdevice_dev_check,
ocelot_port_attr_set);
return notifier_from_errno(err);
}
return NOTIFY_DONE;
}
struct notifier_block ocelot_switchdev_blocking_nb __read_mostly = {
.notifier_call = ocelot_switchdev_blocking_event,
};
net: mscc: ocelot: convert port registers to regmap At the moment, there are some minimal register differences between VSC7514 Ocelot and VSC9959 Felix. To be precise, the PCS1G registers are missing from Felix because it was integrated with an NXP PCS. But with VSC9953 Seville (not yet introduced), the register differences are more pronounced. The MAC registers are located at different offsets within the DEV_GMII target. So we need to refactor the driver to keep a regmap even for per-port registers. The callers of the ocelot_port_readl and ocelot_port_writel were kept unchanged, only the implementation is now more generic. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Reviewed-by: Florian Fainelli <f.fainelli@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2020-07-13 19:57:01 +03:00
int ocelot_probe_port(struct ocelot *ocelot, int port, struct regmap *target,
net: mscc: ocelot: move net_device related functions to ocelot_net.c The ocelot hardware library shouldn't contain too much net_device specific code, since it is shared with DSA which abstracts that structure away. So much as much of this code as possible into the mscc_ocelot driver and outside of the common library. We're making an exception for MDB and LAG code. That is not yet exported to DSA, but when it will, most of the code that's already in ocelot.c will remain there. So, there's no point in moving code to ocelot_net.c just to move it back later. We could have moved all net_device code to ocelot_vsc7514.c directly, but let's operate under the assumption that if a new switchdev ocelot driver gets added, it'll define its SoC-specific stuff in a new ocelot_vsc*.c file and it'll reuse the rest of the code. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2020-06-20 18:43:44 +03:00
struct phy_device *phy)
{
struct ocelot_port_private *priv;
struct ocelot_port *ocelot_port;
struct net_device *dev;
int err;
dev = alloc_etherdev(sizeof(struct ocelot_port_private));
if (!dev)
return -ENOMEM;
SET_NETDEV_DEV(dev, ocelot->dev);
priv = netdev_priv(dev);
priv->dev = dev;
priv->phy = phy;
priv->chip_port = port;
ocelot_port = &priv->port;
ocelot_port->ocelot = ocelot;
net: mscc: ocelot: convert port registers to regmap At the moment, there are some minimal register differences between VSC7514 Ocelot and VSC9959 Felix. To be precise, the PCS1G registers are missing from Felix because it was integrated with an NXP PCS. But with VSC9953 Seville (not yet introduced), the register differences are more pronounced. The MAC registers are located at different offsets within the DEV_GMII target. So we need to refactor the driver to keep a regmap even for per-port registers. The callers of the ocelot_port_readl and ocelot_port_writel were kept unchanged, only the implementation is now more generic. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Reviewed-by: Florian Fainelli <f.fainelli@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2020-07-13 19:57:01 +03:00
ocelot_port->target = target;
net: mscc: ocelot: move net_device related functions to ocelot_net.c The ocelot hardware library shouldn't contain too much net_device specific code, since it is shared with DSA which abstracts that structure away. So much as much of this code as possible into the mscc_ocelot driver and outside of the common library. We're making an exception for MDB and LAG code. That is not yet exported to DSA, but when it will, most of the code that's already in ocelot.c will remain there. So, there's no point in moving code to ocelot_net.c just to move it back later. We could have moved all net_device code to ocelot_vsc7514.c directly, but let's operate under the assumption that if a new switchdev ocelot driver gets added, it'll define its SoC-specific stuff in a new ocelot_vsc*.c file and it'll reuse the rest of the code. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2020-06-20 18:43:44 +03:00
ocelot->ports[port] = ocelot_port;
dev->netdev_ops = &ocelot_port_netdev_ops;
dev->ethtool_ops = &ocelot_ethtool_ops;
dev->hw_features |= NETIF_F_HW_VLAN_CTAG_FILTER | NETIF_F_RXFCS |
NETIF_F_HW_TC;
dev->features |= NETIF_F_HW_VLAN_CTAG_FILTER | NETIF_F_HW_TC;
memcpy(dev->dev_addr, ocelot->base_mac, ETH_ALEN);
dev->dev_addr[ETH_ALEN - 1] += port;
net: mscc: ocelot: transform the pvid and native vlan values into a structure This is a mechanical patch only. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2020-10-31 12:29:12 +02:00
ocelot_mact_learn(ocelot, PGID_CPU, dev->dev_addr,
ocelot_port->pvid_vlan.vid, ENTRYTYPE_LOCKED);
net: mscc: ocelot: move net_device related functions to ocelot_net.c The ocelot hardware library shouldn't contain too much net_device specific code, since it is shared with DSA which abstracts that structure away. So much as much of this code as possible into the mscc_ocelot driver and outside of the common library. We're making an exception for MDB and LAG code. That is not yet exported to DSA, but when it will, most of the code that's already in ocelot.c will remain there. So, there's no point in moving code to ocelot_net.c just to move it back later. We could have moved all net_device code to ocelot_vsc7514.c directly, but let's operate under the assumption that if a new switchdev ocelot driver gets added, it'll define its SoC-specific stuff in a new ocelot_vsc*.c file and it'll reuse the rest of the code. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2020-06-20 18:43:44 +03:00
ocelot_init_port(ocelot, port);
err = register_netdev(dev);
if (err) {
dev_err(ocelot->dev, "register_netdev failed\n");
free_netdev(dev);
net: mscc: ocelot: fix error handling bugs in mscc_ocelot_init_ports() There are several error handling bugs in mscc_ocelot_init_ports(). I went through the code, and carefully audited it and made fixes and cleanups. 1) The ocelot_probe_port() function didn't have a mirror release function so it was hard to follow. I created the ocelot_release_port() function. 2) In the ocelot_probe_port() function, if the register_netdev() call failed, then it lead to a double free_netdev(dev) bug. Fix this by setting "ocelot->ports[port] = NULL" on the error path. 3) I was concerned that the "port" which comes from of_property_read_u32() might be out of bounds so I added a check for that. 4) In the original code if ocelot_regmap_init() failed then the driver tried to continue but I think that should be a fatal error. 5) If ocelot_probe_port() failed then the most recent devlink was leaked. The fix for mostly came Vladimir Oltean. Get rid of "registered_ports" and just set a bit in "devlink_ports_registered" to say when the devlink port has been registered (and needs to be unregistered on error). There are fewer than 32 ports so a u32 is large enough for this purpose. 6) The error handling if the final ocelot_port_devlink_init() failed had two problems. The "while (port-- >= 0)" loop should have been "--port" pre-op instead of a post-op to avoid a buffer underflow. The "if (!registered_ports[port])" condition was reversed leading to resource leaks and double frees. Fixes: 6c30384eb1de ("net: mscc: ocelot: register devlink ports") Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com> Reviewed-by: Vladimir Oltean <vladimir.oltean@nxp.com> Tested-by: Vladimir Oltean <vladimir.oltean@nxp.com> Link: https://lore.kernel.org/r/YBkXhqRxHtRGzSnJ@mwanda Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2021-02-02 12:12:38 +03:00
ocelot->ports[port] = NULL;
return err;
net: mscc: ocelot: move net_device related functions to ocelot_net.c The ocelot hardware library shouldn't contain too much net_device specific code, since it is shared with DSA which abstracts that structure away. So much as much of this code as possible into the mscc_ocelot driver and outside of the common library. We're making an exception for MDB and LAG code. That is not yet exported to DSA, but when it will, most of the code that's already in ocelot.c will remain there. So, there's no point in moving code to ocelot_net.c just to move it back later. We could have moved all net_device code to ocelot_vsc7514.c directly, but let's operate under the assumption that if a new switchdev ocelot driver gets added, it'll define its SoC-specific stuff in a new ocelot_vsc*.c file and it'll reuse the rest of the code. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2020-06-20 18:43:44 +03:00
}
net: mscc: ocelot: fix error handling bugs in mscc_ocelot_init_ports() There are several error handling bugs in mscc_ocelot_init_ports(). I went through the code, and carefully audited it and made fixes and cleanups. 1) The ocelot_probe_port() function didn't have a mirror release function so it was hard to follow. I created the ocelot_release_port() function. 2) In the ocelot_probe_port() function, if the register_netdev() call failed, then it lead to a double free_netdev(dev) bug. Fix this by setting "ocelot->ports[port] = NULL" on the error path. 3) I was concerned that the "port" which comes from of_property_read_u32() might be out of bounds so I added a check for that. 4) In the original code if ocelot_regmap_init() failed then the driver tried to continue but I think that should be a fatal error. 5) If ocelot_probe_port() failed then the most recent devlink was leaked. The fix for mostly came Vladimir Oltean. Get rid of "registered_ports" and just set a bit in "devlink_ports_registered" to say when the devlink port has been registered (and needs to be unregistered on error). There are fewer than 32 ports so a u32 is large enough for this purpose. 6) The error handling if the final ocelot_port_devlink_init() failed had two problems. The "while (port-- >= 0)" loop should have been "--port" pre-op instead of a post-op to avoid a buffer underflow. The "if (!registered_ports[port])" condition was reversed leading to resource leaks and double frees. Fixes: 6c30384eb1de ("net: mscc: ocelot: register devlink ports") Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com> Reviewed-by: Vladimir Oltean <vladimir.oltean@nxp.com> Tested-by: Vladimir Oltean <vladimir.oltean@nxp.com> Link: https://lore.kernel.org/r/YBkXhqRxHtRGzSnJ@mwanda Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2021-02-02 12:12:38 +03:00
return 0;
}
void ocelot_release_port(struct ocelot_port *ocelot_port)
{
struct ocelot_port_private *priv = container_of(ocelot_port,
struct ocelot_port_private,
port);
unregister_netdev(priv->dev);
free_netdev(priv->dev);
net: mscc: ocelot: move net_device related functions to ocelot_net.c The ocelot hardware library shouldn't contain too much net_device specific code, since it is shared with DSA which abstracts that structure away. So much as much of this code as possible into the mscc_ocelot driver and outside of the common library. We're making an exception for MDB and LAG code. That is not yet exported to DSA, but when it will, most of the code that's already in ocelot.c will remain there. So, there's no point in moving code to ocelot_net.c just to move it back later. We could have moved all net_device code to ocelot_vsc7514.c directly, but let's operate under the assumption that if a new switchdev ocelot driver gets added, it'll define its SoC-specific stuff in a new ocelot_vsc*.c file and it'll reuse the rest of the code. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2020-06-20 18:43:44 +03:00
}
Reference in a new issue Copy permalink