Add bond_ether_setup helper which is used to fix ether_setup() calls in the
bonding driver. It takes care of both IFF_MASTER and IFF_SLAVE flags, the
former is always restored and the latter only if it was set.
If the bond enslaves non-ARPHRD_ETHER device (changes its type), then
releases it and enslaves ARPHRD_ETHER device (changes back) then we
use ether_setup() to restore the bond device type but it also resets its
flags and removes IFF_MASTER and IFF_SLAVE[1]. Use the bond_ether_setup
helper to restore both after such transition.
[1] reproduce (nlmon is non-ARPHRD_ETHER):
$ ip l add nlmon0 type nlmon
$ ip l add bond2 type bond mode active-backup
$ ip l set nlmon0 master bond2
$ ip l set nlmon0 nomaster
$ ip l add bond1 type bond
(we use bond1 as ARPHRD_ETHER device to restore bond2's mode)
$ ip l set bond1 master bond2
$ ip l sh dev bond2
37: bond2: <BROADCAST,MULTICAST> mtu 1500 qdisc noop state DOWN mode DEFAULT group default qlen 1000
link/ether be:d7:c5:40:5b:cc brd ff:ff:ff:ff:ff:ff promiscuity 0 minmtu 68 maxmtu 1500
(notice bond2's IFF_MASTER is missing)
Fixes: e36b9d16c6 ("bonding: clean muticast addresses when device changes type")
Signed-off-by: Nikolay Aleksandrov <razor@blackwall.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Since the GWCA has the TX timestamp feature, this driver
should not disable it if one of ports is opened. So, fix it.
Reported-by: Phong Hoang <phong.hoang.wz@renesas.com>
Fixes: 33f5d733b5 ("net: renesas: rswitch: Improve TX timestamp accuracy")
Signed-off-by: Yoshihiro Shimoda <yoshihiro.shimoda.uh@renesas.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
If the RX descriptor doesn't have any data, the output value of quote
from rswitch_rx() will be increased unexpectedily. So, fix it.
Reported-by: Volodymyr Babchuk <volodymyr_babchuk@epam.com>
Fixes: 3590918b5d ("net: ethernet: renesas: Add support for "Ethernet Switch"")
Signed-off-by: Yoshihiro Shimoda <yoshihiro.shimoda.uh@renesas.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
In vnet_port_probe() and vsw_port_probe(), we should
check the return value of mdesc_grab() as it may
return NULL which can caused NPD bugs.
Fixes: 5d01fa0c6b ("ldmvsw: Add ldmvsw.c driver code")
Fixes: 43fdf27470 ("[SPARC64]: Abstract out mdesc accesses for better MD update handling.")
Signed-off-by: Liang He <windhl@126.com>
Reviewed-by: Piotr Raczynski <piotr.raczynski@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
A recent commit defined HW_PARAM_4 as a GSI register ID but did not
add it to gsi_reg_id_valid() to indicate it's valid (for IPA v5.0+).
Add version checks for the HW_PARAM_2 and INTER_EE IRQ GSI registers
there as well.
IPA v5.0 supports up to 8 source and destination resource groups.
Update the validity check (and the comments where the register IDs
are defined) to reflect that. Similarly update comments and
validity checks for the hash/cache-related registers.
Note that this patch fixes an omission and constrains things
further, but these don't technically represent bugs.
Fixes: f651334e1e ("net: ipa: add HW_PARAM_4 GSI register")
Signed-off-by: Alex Elder <elder@linaro.org>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
A recent commit defined a few IPA registers used for IPA v5.0+.
One of those was a mistake. Although the filter and router caches
get *flushed* using a single register, they use distinct registers
(ENDP_FILTER_CACHE_CFG and ENDP_ROUTER_CACHE_CFG) for configuration.
And although there *exists* a FILT_ROUT_CACHE_CFG register, it is
not needed in upstream code. So get rid of definitions related to
FILT_ROUT_CACHE_CFG, because they are not needed.
Fixes: 8ba59716d1 ("net: ipa: define IPA v5.0+ registers")
Signed-off-by: Alex Elder <elder@linaro.org>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
When gsi_reg_init() got added, its declaration was added to
"gsi_reg.h" without declaring the two struct pointer types it uses.
Add these struct declarations to "gsi_reg.h".
Fixes: 3c506add35 ("net: ipa: introduce gsi_reg_init()")
Signed-off-by: Alex Elder <elder@linaro.org>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
When "reg.h" got created, it included calls to WARN() and WARN_ON().
Those macros are defined via <linux/bug.h>. In addition, it uses
is_power_of_2(), which is defined in <linux/log2.h>. Include those
files so IPA "reg.h" has access to all definitions it requires.
Meanwhile, <linux/bits.h> is included but nothing defined therein
is required directly in "reg.h", so get rid of that.
Fixes: 81772e444d ("net: ipa: start generalizing "ipa_reg"")
Signed-off-by: Alex Elder <elder@linaro.org>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The blamed commit has replaced a ksz_write8() call to address
REG_PORT_5_CTRL_6 (0x56) with a ksz_set_xmii() -> ksz_pwrite8() call to
regs[P_XMII_CTRL_1], which is also defined as 0x56 for ksz8795_regs[].
The trouble is that, when compared to ksz_write8(), ksz_pwrite8() also
adjusts the register offset with the port base address. So in reality,
ksz_pwrite8(offset=0x56) accesses register 0x56 + 0x50 = 0xa6, which in
this switch appears to be unmapped, and the RGMII delay configuration on
the CPU port does nothing.
So if the switch wasn't fine with the RGMII delay configuration done
through pin strapping and relied on Linux to apply a different one in
order to pass traffic, this is now broken.
Using the offset translation logic imposed by ksz_pwrite8(), the correct
value for regs[P_XMII_CTRL_1] should have been 0x6 on ksz8795_regs[], in
order to really end up accessing register 0x56.
Static code analysis shows that, despite there being multiple other
accesses to regs[P_XMII_CTRL_1] in this driver, the only code path that
is applicable to ksz8795_regs[] and ksz8_dev_ops is ksz_set_xmii().
Therefore, the problem is isolated to RGMII delays.
In its current form, ksz8795_regs[] contains the same value for
P_XMII_CTRL_0 and for P_XMII_CTRL_1, and this raises valid suspicions
that writes made by the driver to regs[P_XMII_CTRL_0] might overwrite
writes made to regs[P_XMII_CTRL_1] or vice versa.
Again, static analysis shows that the only accesses to P_XMII_CTRL_0
from the driver are made from code paths which are not reachable with
ksz8_dev_ops. So the accesses made by ksz_set_xmii() are safe for this
switch family.
[ vladimiroltean: rewrote commit message ]
Fixes: c476bede4b ("net: dsa: microchip: ksz8795: use common xmii function")
Signed-off-by: Marek Vasut <marex@denx.de>
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Acked-by: Arun Ramadoss <arun.ramadoss@microchip.com>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Link: https://lore.kernel.org/r/20230315231916.2998480-1-vladimir.oltean@nxp.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Saeed Mahameed says:
====================
mlx5 fixes 2023-03-15
This series provides bug fixes to mlx5 driver.
* tag 'mlx5-fixes-2023-03-15' of git://git.kernel.org/pub/scm/linux/kernel/git/saeed/linux:
net/mlx5e: TC, Remove error message log print
net/mlx5e: TC, fix cloned flow attribute
net/mlx5e: TC, fix missing error code
net/sched: TC, fix raw counter initialization
net/mlx5e: Lower maximum allowed MTU in XSK to match XDP prerequisites
net/mlx5: Set BREAK_FW_WAIT flag first when removing driver
net/mlx5e: kTLS, Fix missing error unwind on unsupported cipher type
net/mlx5e: Fix cleanup null-ptr deref on encap lock
net/mlx5: E-switch, Fix missing set of split_count when forward to ovs internal port
net/mlx5: E-switch, Fix wrong usage of source port rewrite in split rules
net/mlx5: Disable eswitch before waiting for VF pages
net/mlx5: Fix setting ec_function bit in MANAGE_PAGES
net/mlx5e: Don't cache tunnel offloads capability
net/mlx5e: Fix macsec ASO context alignment
====================
Link: https://lore.kernel.org/r/20230315225847.360083-1-saeed@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The aq_xdp_run_prog() function falls back to the XDP_ABORTED action
handler (using a goto) if the operations for any of the other actions fail.
The XDP_ABORTED handler in turn calls the bpf_warn_invalid_xdp_action()
tracepoint. However, the function also jumps into the XDP_PASS helper if no
XDP program is loaded on the device, which means the XDP_ABORTED handler
can be run with a NULL program pointer. This results in a NULL pointer
deref because the tracepoint dereferences the 'prog' pointer passed to it.
This situation can happen in multiple ways:
- If a packet arrives between the removal of the program from the interface
and the static_branch_dec() in aq_xdp_setup()
- If there are multiple devices using the same driver in the system and
one of them has an XDP program loaded and the other does not.
Fix this by refactoring the aq_xdp_run_prog() function to remove the 'goto
pass' handling if there is no XDP program loaded. Instead, factor out the
skb building in a separate small helper function.
Fixes: 26efaef759 ("net: atlantic: Implement xdp data plane")
Reported-by: Freysteinn Alfredsson <Freysteinn.Alfredsson@kau.se>
Tested-by: Freysteinn Alfredsson <Freysteinn.Alfredsson@kau.se>
Signed-off-by: Toke Høiland-Jørgensen <toke@redhat.com>
Link: https://lore.kernel.org/r/20230315125539.103319-1-toke@redhat.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
SH_ETH doesn't need mdiobus suspend/resume, that's why it sets
'mac_managed_pm'. However, setting it needs to be moved from init to
probe, so mdiobus PM functions will really never be called (e.g. when
the interface is not up yet during suspend/resume).
Fixes: 6a1dbfefda ("net: sh_eth: Fix PHY state warning splat during system resume")
Suggested-by: Heiner Kallweit <hkallweit1@gmail.com>
Signed-off-by: Wolfram Sang <wsa+renesas@sang-engineering.com>
Reviewed-by: Michal Kubiak <michal.kubiak@intel.com>
Reviewed-by: Sergey Shtylyov <s.shtylyov@omp.ru>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
RAVB doesn't need mdiobus suspend/resume, that's why it sets
'mac_managed_pm'. However, setting it needs to be moved from init to
probe, so mdiobus PM functions will really never be called (e.g. when
the interface is not up yet during suspend/resume).
Fixes: 4924c0cdce ("net: ravb: Fix PHY state warning splat during system resume")
Suggested-by: Heiner Kallweit <hkallweit1@gmail.com>
Signed-off-by: Wolfram Sang <wsa+renesas@sang-engineering.com>
Reviewed-by: Michal Kubiak <michal.kubiak@intel.com>
Reviewed-by: Sergey Shtylyov <s.shtylyov@omp.ru>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
It is preferred to use typed property access functions (i.e.
of_property_read_<type> functions) rather than low-level
of_get_property/of_find_property functions for reading properties.
Convert reading boolean properties to of_property_read_bool().
Reviewed-by: Simon Horman <simon.horman@corigine.com>
Acked-by: Marc Kleine-Budde <mkl@pengutronix.de> # for net/can
Acked-by: Kalle Valo <kvalo@kernel.org>
Acked-by: Nicolas Ferre <nicolas.ferre@microchip.com>
Acked-by: Francois Romieu <romieu@fr.zoreil.com>
Reviewed-by: Wei Fang <wei.fang@nxp.com>
Signed-off-by: Rob Herring <robh@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
There are 3 classes of switch families that the driver is aware of, as
far as mv88e6xxx_change_mtu() is concerned:
- MTU configuration is available per port. Here, the
chip->info->ops->port_set_jumbo_size() method will be present.
- MTU configuration is global to the switch. Here, the
chip->info->ops->set_max_frame_size() method will be present.
- We don't know how to change the MTU. Here, none of the above methods
will be present.
Switch families MV88E6165, MV88E6191, MV88E6220, MV88E6250 and MV88E6290
fall in category 3.
The blamed commit has adjusted the MTU for all 3 categories by EDSA_HLEN
(8 bytes), resulting in a new maximum MTU of 1492 being reported by the
driver for these switches.
I don't have the hardware to test, but I do have a MV88E6390 switch on
which I can simulate this by commenting out its .port_set_jumbo_size
definition from mv88e6390_ops. The result is this set of messages at
probe time:
mv88e6085 d0032004.mdio-mii:10: nonfatal error -34 setting MTU to 1500 on port 1
mv88e6085 d0032004.mdio-mii:10: nonfatal error -34 setting MTU to 1500 on port 2
mv88e6085 d0032004.mdio-mii:10: nonfatal error -34 setting MTU to 1500 on port 3
mv88e6085 d0032004.mdio-mii:10: nonfatal error -34 setting MTU to 1500 on port 4
mv88e6085 d0032004.mdio-mii:10: nonfatal error -34 setting MTU to 1500 on port 5
mv88e6085 d0032004.mdio-mii:10: nonfatal error -34 setting MTU to 1500 on port 6
mv88e6085 d0032004.mdio-mii:10: nonfatal error -34 setting MTU to 1500 on port 7
mv88e6085 d0032004.mdio-mii:10: nonfatal error -34 setting MTU to 1500 on port 8
It is highly implausible that there exist Ethernet switches which don't
support the standard MTU of 1500 octets, and this is what the DSA
framework says as well - the error comes from dsa_slave_create() ->
dsa_slave_change_mtu(slave_dev, ETH_DATA_LEN).
But the error messages are alarming, and it would be good to suppress
them.
As a consequence of this unlikeliness, we reimplement mv88e6xxx_get_max_mtu()
and mv88e6xxx_change_mtu() on switches from the 3rd category as follows:
the maximum supported MTU is 1500, and any request to set the MTU to a
value larger than that fails in dev_validate_mtu().
Fixes: b9c587fed6 ("dsa: mv88e6xxx: Include tagger overhead when setting MTU for DSA and CPU ports")
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Reviewed-by: Simon Horman <simon.horman@corigine.com>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Since the blamed commit, phy_ethtool_get_wol() and phy_ethtool_set_wol()
acquire phydev->lock, but the mscc phy driver implementations,
vsc85xx_wol_get() and vsc85xx_wol_set(), acquire the same lock as well,
resulting in a deadlock.
$ ip link set swp3 down
============================================
WARNING: possible recursive locking detected
mscc_felix 0000:00:00.5 swp3: Link is Down
--------------------------------------------
ip/375 is trying to acquire lock:
ffff3d7e82e987a8 (&dev->lock){+.+.}-{4:4}, at: vsc85xx_wol_get+0x2c/0xf4
but task is already holding lock:
ffff3d7e82e987a8 (&dev->lock){+.+.}-{4:4}, at: phy_ethtool_get_wol+0x3c/0x6c
other info that might help us debug this:
Possible unsafe locking scenario:
CPU0
----
lock(&dev->lock);
lock(&dev->lock);
*** DEADLOCK ***
May be due to missing lock nesting notation
2 locks held by ip/375:
#0: ffffd43b2a955788 (rtnl_mutex){+.+.}-{4:4}, at: rtnetlink_rcv_msg+0x144/0x58c
#1: ffff3d7e82e987a8 (&dev->lock){+.+.}-{4:4}, at: phy_ethtool_get_wol+0x3c/0x6c
Call trace:
__mutex_lock+0x98/0x454
mutex_lock_nested+0x2c/0x38
vsc85xx_wol_get+0x2c/0xf4
phy_ethtool_get_wol+0x50/0x6c
phy_suspend+0x84/0xcc
phy_state_machine+0x1b8/0x27c
phy_stop+0x70/0x154
phylink_stop+0x34/0xc0
dsa_port_disable_rt+0x2c/0xa4
dsa_slave_close+0x38/0xec
__dev_close_many+0xc8/0x16c
__dev_change_flags+0xdc/0x218
dev_change_flags+0x24/0x6c
do_setlink+0x234/0xea4
__rtnl_newlink+0x46c/0x878
rtnl_newlink+0x50/0x7c
rtnetlink_rcv_msg+0x16c/0x58c
Removing the mutex_lock(&phydev->lock) calls from the driver restores
the functionality.
Fixes: 2f987d4866 ("net: phy: Add locks to ethtool functions")
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Reviewed-by: Simon Horman <simon.horman@corigine.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Link: https://lore.kernel.org/r/20230314153025.2372970-1-vladimir.oltean@nxp.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The cited commit attempts to update the hw stats when dumping tc actions.
However, the driver may be called to update the stats of a police action
that may not be in hardware. In such cases the driver will fail to lookup
the police action object and will output an error message both to extack
and dmesg. The dmesg error is confusing as it may not indicate an actual
error.
Remove the dmesg error.
Fixes: 2b68d659a7 ("net/mlx5e: TC, support per action stats")
Signed-off-by: Oz Shlomo <ozsh@nvidia.com>
Reviewed-by: Paul Blakey <paulb@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
Currently the cloned flow attr resets the original tc action cookies
count.
Fix that by resetting the cloned flow attribute.
Fixes: cca7eac138 ("net/mlx5e: TC, store tc action cookies per attr")
Signed-off-by: Oz Shlomo <ozsh@nvidia.com>
Reviewed-by: Paul Blakey <paulb@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
Freed counters may be reused by fs core.
As such, raw counters may not be initialized to zero.
Cache the counter values when the action stats object is initialized to
have a proper base value for calculating the difference from the previous
query.
Fixes: 2b68d659a7 ("net/mlx5e: TC, support per action stats")
Signed-off-by: Oz Shlomo <ozsh@nvidia.com>
Reviewed-by: Paul Blakey <paulb@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
XSK redirecting XDP programs require linearity, hence applies
restrictions on the MTU. For PAGE_SIZE=4K, MTU shouldn't exceed 3498.
Features that contradict with XDP such HW-LRO and HW-GRO are enforced
by the driver in advance, during XSK params validation, except for MTU,
which was not enforced before this patch.
This has been spotted during test scenario described below:
Attaching xdpsock program (PAGE_SIZE=4K), with MTU < 3498, detaching
XDP program, changing the MTU to arbitrary value in the range
[3499, 3754], attaching XDP program again, which ended up with failure
since MTU is > 3498.
This commit lowers the XSK MTU limitation to be aligned with XDP MTU
limitation, since XSK socket is meaningless without XDP program.
Signed-off-by: Adham Faris <afaris@nvidia.com>
Reviewed-by: Tariq Toukan <tariqt@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
Currently, BREAK_FW_WAIT flag is set after syncing with fw_reset.
However, fw_reset can call mlx5_load_one() which is waiting for fw
init bit and BREAK_FW_WAIT flag is intended to stop. e.g.: the driver
might wait on a loop it should exit.
Fix it by setting the flag before syncing with fw_reset.
Fixes: 8324a02c34 ("net/mlx5: Add exit route when waiting for FW")
Signed-off-by: Shay Drory <shayd@nvidia.com>
Reviewed-by: Moshe Shemesh <moshe@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
Do proper error unwinding when adding an unsupported TX/RX cipher type.
Move the switch case prior to key creation so there's less to unwind,
and change the goto label name to describe the action performed instead
of what failed.
Fixes: 4960c414db ("net/mlx5e: Support 256 bit keys with kTLS device offload")
Signed-off-by: Gal Pressman <gal@nvidia.com>
Reviewed-by: Tariq Toukan <tariqt@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
During module is unloaded while a peer tc flow is still offloaded,
first the peer uplink rep profile is changed to a nic profile, and so
neigh encap lock is destroyed. Next during unload, the VF reps netdevs
are unregistered which causes the original non-peer tc flow to be deleted,
which deletes the peer flow. The peer flow deletion detaches the encap
entry and try to take the already destroyed encap lock, causing the
below trace.
Fix this by clearing peer flows during tc eswitch cleanup
(mlx5e_tc_esw_cleanup()).
Relevant trace:
[ 4316.837128] BUG: kernel NULL pointer dereference, address: 00000000000001d8
[ 4316.842239] RIP: 0010:__mutex_lock+0xb5/0xc40
[ 4316.851897] Call Trace:
[ 4316.852481] <TASK>
[ 4316.857214] mlx5e_rep_neigh_entry_release+0x93/0x790 [mlx5_core]
[ 4316.858258] mlx5e_rep_encap_entry_detach+0xa7/0xf0 [mlx5_core]
[ 4316.859134] mlx5e_encap_dealloc+0xa3/0xf0 [mlx5_core]
[ 4316.859867] clean_encap_dests.part.0+0x5c/0xe0 [mlx5_core]
[ 4316.860605] mlx5e_tc_del_fdb_flow+0x32a/0x810 [mlx5_core]
[ 4316.862609] __mlx5e_tc_del_fdb_peer_flow+0x1a2/0x250 [mlx5_core]
[ 4316.863394] mlx5e_tc_del_flow+0x(/0x630 [mlx5_core]
[ 4316.864090] mlx5e_flow_put+0x5f/0x100 [mlx5_core]
[ 4316.864771] mlx5e_delete_flower+0x4de/0xa40 [mlx5_core]
[ 4316.865486] tc_setup_cb_reoffload+0x20/0x80
[ 4316.865905] fl_reoffload+0x47c/0x510 [cls_flower]
[ 4316.869181] tcf_block_playback_offloads+0x91/0x1d0
[ 4316.869649] tcf_block_unbind+0xe7/0x1b0
[ 4316.870049] tcf_block_offload_cmd.isra.0+0x1ee/0x270
[ 4316.879266] tcf_block_offload_unbind+0x61/0xa0
[ 4316.879711] __tcf_block_put+0xa4/0x310
Fixes: 04de7dda73 ("net/mlx5e: Infrastructure for duplicated offloading of TC flows")
Fixes: 1418ddd96a ("net/mlx5e: Duplicate offloaded TC eswitch rules under uplink LAG")
Signed-off-by: Paul Blakey <paulb@nvidia.com>
Reviewed-by: Chris Mi <cmi@nvidia.com>
Reviewed-by: Roi Dayan <roid@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
Rules with mirror actions are split to two FTEs when the actions after the mirror
action contains pedit, vlan push/pop or ct. Forward to ovs internal port adds
implicit header rewrite (pedit) but missing trigger to do split.
Fix by setting split_count when forwarding to ovs internal port which
will trigger split in mirror rules.
Fixes: 27484f7170 ("net/mlx5e: Offload tc rules that redirect to ovs internal port")
Signed-off-by: Maor Dickman <maord@nvidia.com>
Reviewed-by: Roi Dayan <roid@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
In few cases, rules with mirror use case are split to two FTEs, one which
do the mirror action and forward to second FTE which do the rest of the rule
actions and the second redirect action.
In case of mirror rules which do split and forward to ovs internal port or
VF stack devices, source port rewrite should be used in the second FTE but
it is wrongly also set in the first FTE which break the offload.
Fix this issue by removing the wrong check if source port rewrite is needed to
be used on the first FTE of the split and instead return EOPNOTSUPP which will
block offload of rules which mirror to ovs internal port or VF stack devices
which isn't supported.
Fixes: 10742efc20 ("net/mlx5e: VF tunnel TX traffic offloading")
Fixes: a508728a4c ("net/mlx5e: VF tunnel RX traffic offloading")
Signed-off-by: Maor Dickman <maord@nvidia.com>
Reviewed-by: Roi Dayan <roid@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
The offending commit changed the ordering of moving to legacy mode and
waiting for the VF pages. Moving to legacy mode is important in
bluefield, because it sends the host driver into error state, and frees
its pages. Without this transition we end up waiting 2 minutes for
pages that aren't coming before carrying on with the unload process.
Fixes: f019679ea5 ("net/mlx5: E-switch, Remove dependency between sriov and eswitch mode")
Signed-off-by: Daniel Jurgens <danielj@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
When ECPF is a page supplier, reclaim pages missed to honor the
ec_function bit provided by the firmware. It always used the ec_function
to true during driver unload flow for ECPF. This is incorrect.
Honor the ec_function bit provided by device during page allocation
request event.
Fixes: d6945242f4 ("net/mlx5: Hold pages RB tree per VF")
Signed-off-by: Parav Pandit <parav@nvidia.com>
Signed-off-by: Daniel Jurgens <danielj@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
When mlx5e attaches again after device health recovery, the device
capabilities might have changed by the eswitch manager.
For example in one flow when ECPF changes the eswitch mode between
legacy and switchdev, it updates the flow table tunnel capability.
The cached value is only used in one place, so just check the capability
there instead.
Fixes: 5bef709d76 ("net/mlx5: Enable host PF HCA after eswitch is initialized")
Signed-off-by: Parav Pandit <parav@nvidia.com>
Signed-off-by: Daniel Jurgens <danielj@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
Currently mlx5e_macsec_umr struct does not satisfy hardware memory
alignment requirement. Hence the result of querying advanced steering
operation (ASO) is not copied to the memory region as expected.
Fix by satisfying hardware memory alignment requirement and move
context to be first field in struct for better readability.
Fixes: 1f53da6764 ("net/mlx5e: Create advanced steering operation (ASO) object for MACsec")
Signed-off-by: Emeel Hakim <ehakim@nvidia.com>
Reviewed-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
Only restart auto-negotiation and write link timer if actually
necessary. This prevents losing the link in case of minor
changes.
Fixes: 7e53837269 ("net: ethernet: mediatek: Re-add support SGMII")
Reviewed-by: Russell King (Oracle) <rmk+kernel@armlinux.org.uk>
Tested-by: Bjørn Mork <bjorn@mork.no>
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Reset the internal PCS state machine when changing interface mode.
This prevents confusing the state machine when changing interface
modes, e.g. from SGMII to 2500Base-X or vice-versa.
Fixes: 7e53837269 ("net: ethernet: mediatek: Re-add support SGMII")
Reviewed-by: Russell King (Oracle) <rmk+kernel@armlinux.org.uk>
Tested-by: Bjørn Mork <bjorn@mork.no>
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Packet length retrieved from skb data may be larger than
the actual socket buffer length (up to 9026 bytes). In such
case the cloned skb passed up the network stack will leak
kernel memory contents.
Fixes: d0cad87170 ("smsc75xx: SMSC LAN75xx USB gigabit ethernet adapter driver")
Signed-off-by: Szymon Heidrich <szymon.heidrich@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Spectrum ASICs have a configurable limit on how deep into the packet
they parse. By default, the limit is 96 bytes.
There are several cases where this parsing depth is not enough and there
is a need to increase it. For example, timestamping of PTP packets and a
FIB multipath hash policy that requires hashing on inner fields. The
driver therefore maintains a reference count that reflects the number of
consumers that require an increased parsing depth.
During reload_down() the parsing depth reference count does not
necessarily drop to zero, but the parsing depth itself is restored to
the default during reload_up() when the firmware is reset. It is
therefore possible to end up in situations where the driver thinks that
the parsing depth was increased (reference count is non-zero), when it
is not.
Fix by making sure that all the consumers that increase the parsing
depth reference count also decrease it during reload_down().
Specifically, make sure that when the routing code is de-initialized it
drops the reference count if it was increased because of a FIB multipath
hash policy that requires hashing on inner fields.
Add a warning if the reference count is not zero after the driver was
de-initialized and explicitly reset it to zero during initialization for
good measures.
Fixes: 2d91f0803b ("mlxsw: spectrum: Add infrastructure for parsing configuration")
Reported-by: Maksym Yaremchuk <maksymy@nvidia.com>
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Signed-off-by: Petr Machata <petrm@nvidia.com>
Link: https://lore.kernel.org/r/9c35e1b3e6c1d8f319a2449d14e2b86373f3b3ba.1678727526.git.petrm@nvidia.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
A recent commit eliminated a hack that adjusted the offset used for
many GSI registers. It became possible because we now specify all
GSI register offsets explicitly for every version of IPA.
Unfortunately, a large number of register offsets were *not* updated
as they should have been in that commit. For IPA v4.5+, the offset
for every GSI register *except* the two inter-EE interrupt masking
registers were supposed to have been reduced by 0xd000.
Tested-by: Luca Weiss <luca.weiss@fairphone.com>
Tested-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org> # SM8350-HDK
Fixes: 59b12b1d27 ("net: ipa: kill gsi->virt_raw")
Signed-off-by: Alex Elder <elder@linaro.org>
Link: https://lore.kernel.org/r/20230310193709.1477102-1-elder@linaro.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
As my testing on the MCM MT7530 switch on MT7621 SoC shows, setting the PLL
frequency does not affect MII modes other than trgmii on port 5 and port 6.
So the assumption is that the operation here called "setting the PLL
frequency" actually sets the frequency of the TRGMII TX clock.
Make it so that it and the rest of the trgmii setup run only when the
trgmii mode is used.
Tested rgmii and trgmii modes of port 6 on MCM MT7530 on MT7621AT Unielec
U7621-06 and standalone MT7530 on MT7623NI Bananapi BPI-R2.
Fixes: b8f126a8d5 ("net-next: dsa: add dsa support for Mediatek MT7530 switch")
Tested-by: Arınç ÜNAL <arinc.unal@arinc9.com>
Signed-off-by: Arınç ÜNAL <arinc.unal@arinc9.com>
Link: https://lore.kernel.org/r/20230310073338.5836-2-arinc.unal@arinc9.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Previously we would divide total_left_rate by zero if num_vports
happened to be 1 because non_requested_count is calculated as
num_vports - req_count. Guard against this by validating num_vports at
the beginning and returning an error otherwise.
Found by Linux Verification Center (linuxtesting.org) with the SVACE
static analysis tool.
Fixes: bcd197c81f ("qed: Add vport WFQ configuration APIs")
Signed-off-by: Daniil Tatianin <d-tatianin@yandex-team.ru>
Reviewed-by: Simon Horman <simon.horman@corigine.com>
Link: https://lore.kernel.org/r/20230309201556.191392-1-d-tatianin@yandex-team.ru
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
When using a PHC in shared between multiple hosts, the previous
frequency value may not be reset and could lead to host being unable to
compensate the offset with timecounter adjustments. To avoid such state
reset the hardware frequency of PHC to zero on init. Some refactoring is
needed to make code readable.
Fixes: 85036aee19 ("bnxt_en: Add a non-real time mode to access NIC clock")
Signed-off-by: Vadim Fedorenko <vadfed@meta.com>
Reviewed-by: Pavan Chebbi <pavan.chebbi@broadcom.com>
Link: https://lore.kernel.org/r/20230310151356.678059-1-vadfed@meta.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
