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>
We should be moving anything that isn't DSA-specific or SoC-specific out
of the felix DSA driver, and into the common mscc_ocelot switch library.
The number of traffic classes is one of the aspects that is common
between all ocelot switches, so it belongs in the library.
This patch also makes seville use 8 TX queues, and therefore enables
prioritization via the QOS_CLASS field in the NPI injection header.
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>
In general it is desirable that cleanup is the reverse process of setup.
In this case I am not seeing any particular issue, but with the
introduction of devlink-sb for felix, a non-obvious decision had to be
made as to where to put its cleanup method. When there's a convention in
place, that decision becomes obvious.
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>
The devlink function pointer names are super long, and they would break
the alignment. So reindent the existing ops now by adding one tab.
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>
Instead of reading these values from the reference manual and writing
them down into the driver, it appears that the hardware gives us the
option of detecting them dynamically.
The number of frame references corresponds to what the reference manual
notes, however it seems that the frame buffers are reported as slightly
less than the books would indicate. On VSC9959 (Felix), the books say it
should have 128KB of packet buffer, but the registers indicate only
129840 bytes (126.79 KB). Also, the unit of measurement for FREECNT from
the documentation of all these devices is incorrect (taken from an older
generation). This was confirmed by Younes Leroul from Microchip support.
Not having anything better to do with these values at the moment* (this
will change soon), let's just print them.
*The frame buffer size is, in fact, used to calculate the tail dropping
watermarks.
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>
As explained in commit 54a0ed0df4 ("net: dsa: provide an option for
drivers to always receive bridge VLANs"), DSA has historically been
skipping VLAN switchdev operations when the bridge wasn't in
vlan_filtering mode, but the reason why it was doing that has never been
clear. So the configure_vlan_while_not_filtering option is there merely
to preserve functionality for existing drivers. It isn't some behavior
that drivers should opt into. Ideally, when all drivers leave this flag
set, we can delete the dsa_port_skip_vlan_configuration() function.
New drivers always seem to omit setting this flag, for some reason. So
let's reverse the logic: the DSA core sets it by default to true before
the .setup() callback, and legacy drivers can turn it off. This way, new
drivers get the new behavior by default, unless they explicitly set the
flag to false, which is more obvious during review.
Remove the assignment from drivers which were setting it to true, and
add the assignment to false for the drivers that didn't previously have
it. This way, it should be easier to see how many we have left.
The following drivers: lan9303, mv88e6060 were skipped from setting this
flag to false, because they didn't have any VLAN offload ops in the
first place.
The Broadcom Starfighter 2 driver calls the common b53_switch_alloc and
therefore also inherits the configure_vlan_while_not_filtering=true
behavior.
Also, print a message through netlink extack every time a VLAN has been
skipped. This is mildly annoying on purpose, so that (a) it is at least
clear that VLANs are being skipped - the legacy behavior in itself is
confusing, and the extack should be much more difficult to miss, unlike
kernel logs - and (b) people have one more incentive to convert to the
new behavior.
No behavior change except for the added prints is intended at this time.
$ ip link add br0 type bridge vlan_filtering 0
$ ip link set sw0p2 master br0
[ 60.315148] br0: port 1(sw0p2) entered blocking state
[ 60.320350] br0: port 1(sw0p2) entered disabled state
[ 60.327839] device sw0p2 entered promiscuous mode
[ 60.334905] br0: port 1(sw0p2) entered blocking state
[ 60.340142] br0: port 1(sw0p2) entered forwarding state
Warning: dsa_core: skipping configuration of VLAN. # This was the pvid
$ bridge vlan add dev sw0p2 vid 100
Warning: dsa_core: skipping configuration of VLAN.
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Reviewed-by: Kurt Kanzenbach <kurt@linutronix.de>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Link: https://lore.kernel.org/r/20210115231919.43834-1-vladimir.oltean@nxp.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
It should be the driver's business to logically separate its VLAN
offloading into a preparation and a commit phase, and some drivers don't
need / can't do this.
So remove the transactional shim from DSA and let drivers propagate
errors directly from the .port_vlan_add callback.
It would appear that the code has worse error handling now than it had
before. DSA is the only in-kernel user of switchdev that offloads one
switchdev object to more than one port: for every VLAN object offloaded
to a user port, that VLAN is also offloaded to the CPU port. So the
"prepare for user port -> check for errors -> prepare for CPU port ->
check for errors -> commit for user port -> commit for CPU port"
sequence appears to make more sense than the one we are using now:
"offload to user port -> check for errors -> offload to CPU port ->
check for errors", but it is really a compromise. In the new way, we can
catch errors from the commit phase that we previously had to ignore.
But we have our hands tied and cannot do any rollback now: if we add a
VLAN on the CPU port and it fails, we can't do the rollback by simply
deleting it from the user port, because the switchdev API is not so nice
with us: it could have simply been there already, even with the same
flags. So we don't even attempt to rollback anything on addition error,
just leave whatever VLANs managed to get offloaded right where they are.
This should not be a problem at all in practice.
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>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
For many drivers, the .port_mdb_prepare callback was not a good opportunity
to avoid any error condition, and they would suppress errors found during
the actual commit phase.
Where a logical separation between the prepare and the commit phase
existed, the function that used to implement the .port_mdb_prepare
callback still exists, but now it is called directly from .port_mdb_add,
which was modified to return an int code.
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: Kurt Kanzenbach <kurt@linutronix.de> # hellcreek
Reviewed-by: Linus Wallei <linus.walleij@linaro.org> # RTL8366
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
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 91cf8eceff ("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 2e554a7a5d
("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>
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 bdced7ef78
("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 47f8328bb1 ("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 41c498b935 ("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 0944d6b5a2 ("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
29ab586c3d ("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 9c86ce2c1a ("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 29ab586c3d 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 6623c60dc2
("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>
Given the following setup:
ip link add br0 type bridge
ip link set eno0 master br0
ip link set swp0 master br0
ip link set swp1 master br0
ip link set swp2 master br0
ip link set swp3 master br0
Currently, packets received on a DSA slave interface (such as swp0)
which should be routed by the software bridge towards a non-switch port
(such as eno0) are also flooded towards the other switch ports (swp1,
swp2, swp3) because the destination is unknown to the hardware switch.
This patch addresses the issue by monitoring the addresses learnt by the
software bridge on eno0, and adding/deleting them as static FDB entries
on the CPU port accordingly.
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
xdp_return_frame_bulk() needs to pass a xdp_buff
to __xdp_return().
strlcpy got converted to strscpy but here it makes no
functional difference, so just keep the right code.
Conflicts:
net/netfilter/nf_tables_api.c
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The current assumption is that the felix DSA driver has flooding knobs
per traffic class, while ocelot switchdev has a single flooding knob.
This was correct for felix VSC9959 and ocelot VSC7514, but with the
introduction of seville VSC9953, we see a switch driven by felix.c which
has a single flooding knob.
So it is clear that we must do what should have been done from the
beginning, which is not to overwrite the configuration done by ocelot.c
in felix, but instead to teach the common ocelot library about the
differences in our switches, and set up the flooding PGIDs centrally.
The effect that the bogus iteration through FELIX_NUM_TC has upon
seville is quite dramatic. ANA_FLOODING is located at 0x00b548, and
ANA_FLOODING_IPMC is located at 0x00b54c. So the bogus iteration will
actually overwrite ANA_FLOODING_IPMC when attempting to write
ANA_FLOODING[1]. There is no ANA_FLOODING[1] in sevile, just ANA_FLOODING.
And when ANA_FLOODING_IPMC is overwritten with a bogus value, the effect
is that ANA_FLOODING_IPMC gets the value of 0x0003CF7D:
MC6_DATA = 61,
MC6_CTRL = 61,
MC4_DATA = 60,
MC4_CTRL = 0.
Because MC4_CTRL is zero, this means that IPv4 multicast control packets
are not flooded, but dropped. An invalid configuration, and this is how
the issue was actually spotted.
Reported-by: Eldar Gasanov <eldargasanov2@gmail.com>
Reported-by: Maxim Kochetkov <fido_max@inbox.ru>
Tested-by: Eldar Gasanov <eldargasanov2@gmail.com>
Fixes: 84705fc165 ("net: dsa: felix: introduce support for Seville VSC9953 switch")
Fixes: 3c7b51bd39 ("net: dsa: felix: allow flooding for all traffic classes")
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Reviewed-by: Alexandre Belloni <alexandre.belloni@bootlin.com>
Link: https://lore.kernel.org/r/20201204175416.1445937-1-vladimir.oltean@nxp.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
After the good discussion with Florian from here:
https://lore.kernel.org/netdev/20200911000337.htwr366ng3nc3a7d@skbuf/
I realized that the VLAN settings on the NPI port (the hardware "CPU port",
in DSA parlance) don't actually make any difference, because that port
is hardcoded in hardware to use what mv88e6xxx would call "unmodified"
egress policy for VLANs.
So earlier patch 183be6f967 ("net: dsa: felix: send VLANs on CPU port
as egress-tagged") was incorrect in the sense that it didn't actually
make the VLANs be sent on the NPI port as egress-tagged. It only made
ocelot_port_set_native_vlan shut up.
Now that we have moved the check from ocelot_port_set_native_vlan to
ocelot_vlan_prepare, we can simply shunt ocelot_vlan_prepare from DSA,
and avoid calling it. This is the correct way to deal with things,
because the NPI port configuration is DSA-specific, so the ocelot switch
library should not have the check for multiple native VLANs refined in
any way, it is correct the way it is.
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
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>
A driver may refuse to enable VLAN filtering for any reason beyond what
the DSA framework cares about, such as:
- having tc-flower rules that rely on the switch being VLAN-aware
- the particular switch does not support VLAN, even if the driver does
(the DSA framework just checks for the presence of the .port_vlan_add
and .port_vlan_del pointers)
- simply not supporting this configuration to be toggled at runtime
Currently, when a driver rejects a configuration it cannot support, it
does this from the commit phase, which triggers various warnings in
switchdev.
So propagate the prepare phase to drivers, to give them the ability to
refuse invalid configurations cleanly and avoid the warnings.
Since we need to modify all function prototypes and check for the
prepare phase from within the drivers, take that opportunity and move
the existing driver restrictions within the prepare phase where that is
possible and easy.
Cc: Florian Fainelli <f.fainelli@gmail.com>
Cc: Martin Blumenstingl <martin.blumenstingl@googlemail.com>
Cc: Hauke Mehrtens <hauke@hauke-m.de>
Cc: Woojung Huh <woojung.huh@microchip.com>
Cc: Microchip Linux Driver Support <UNGLinuxDriver@microchip.com>
Cc: Sean Wang <sean.wang@mediatek.com>
Cc: Landen Chao <Landen.Chao@mediatek.com>
Cc: Andrew Lunn <andrew@lunn.ch>
Cc: Vivien Didelot <vivien.didelot@gmail.com>
Cc: Jonathan McDowell <noodles@earth.li>
Cc: Linus Walleij <linus.walleij@linaro.org>
Cc: Alexandre Belloni <alexandre.belloni@bootlin.com>
Cc: Claudiu Manoil <claudiu.manoil@nxp.com>
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
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>
In the Ocelot switches there are 3 TCAMs: VCAP ES0, IS1 and IS2, which
have the same configuration interface, but different sets of keys and
actions. The driver currently only supports VCAP IS2.
In preparation of VCAP IS1 and ES0 support, the existing code must be
generalized to work with any VCAP.
In that direction, we should move the structures that depend upon VCAP
instantiation, like vcap_is2_keys and vcap_is2_actions, out of struct
ocelot and into struct vcap_props .keys and .actions, a structure that
is replicated 3 times, once per VCAP. We'll pass that structure as an
argument to each function that does the key and action packing - only
the control logic needs to distinguish between ocelot->vcap[VCAP_IS2]
or IS1 or ES0.
Another change is to make use of the newly introduced ocelot_target_read
and ocelot_target_write API, since the 3 VCAPs have the same registers
but put at different addresses.
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
There are 2 goals that we follow:
- Reduce the header size
- Make the header size equal between RX and TX
The issue that required long prefix on RX was the fact that the ocelot
DSA tag, being put before Ethernet as it is, would overlap with the area
that a DSA master uses for RX filtering (destination MAC address
mainly).
Now that we can ask DSA to put the master in promiscuous mode, in theory
we could remove the prefix altogether and call it a day, but it looks
like we can't. Using no prefix on ingress, some packets (such as ICMP)
would be received, while others (such as PTP) would not be received.
This is because the DSA master we use (enetc) triggers parse errors
("MAC rx frame errors") presumably because it sees Ethernet frames with
a bad length. And indeed, when using no prefix, the EtherType (bytes
12-13 of the frame, bits 96-111) falls over the REW_VAL field from the
extraction header, aka the PTP timestamp.
When turning the short (32-bit) prefix on, the EtherType overlaps with
bits 64-79 of the extraction header, which are a reserved area
transmitted as zero by the switch. The packets are not dropped by the
DSA master with a short prefix. Actually, the frames look like this in
tcpdump (below is a PTP frame, with an extra dsa_8021q tag - dadb 0482 -
added by a downstream sja1105).
89:0c:a9:f2:01:00 > 88:80:00:0a:00:1d, 802.3, length 0: LLC, \
dsap Unknown (0x10) Individual, ssap ProWay NM (0x0e) Response, \
ctrl 0x0004: Information, send seq 2, rcv seq 0, \
Flags [Response], length 78
0x0000: 8880 000a 001d 890c a9f2 0100 0000 100f ................
0x0010: 0400 0000 0180 c200 000e 001f 7b63 0248 ............{c.H
0x0020: dadb 0482 88f7 1202 0036 0000 0000 0000 .........6......
0x0030: 0000 0000 0000 0000 0000 001f 7bff fe63 ............{..c
0x0040: 0248 0001 1f81 0500 0000 0000 0000 0000 .H..............
0x0050: 0000 0000 0000 0000 0000 0000 ............
So the short prefix is our new default: we've shortened our RX frames by
12 octets, increased TX by 4, and headers are now equal between RX and
TX. Note that we still need promiscuous mode for the DSA master to not
drop it.
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Remove the ocelot_configure_cpu() function, which was in fact bringing
up 2 ports: the CPU port module, which both switchdev and DSA have, and
the NPI port, which only DSA has.
The (non-Ethernet) CPU port module is at a fixed index in the analyzer,
whereas the NPI port is selected through the "ethernet" property in the
device tree.
Therefore, the function to set up an NPI port is DSA-specific, so we
move it there, simplifying the ocelot switch library a little bit.
Cc: Horatiu Vultur <horatiu.vultur@microchip.com>
Cc: Alexandre Belloni <alexandre.belloni@bootlin.com>
Cc: UNGLinuxDriver <UNGLinuxDriver@microchip.com>
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
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>
Two minor conflicts:
1) net/ipv4/route.c, adding a new local variable while
moving another local variable and removing it's
initial assignment.
2) drivers/net/dsa/microchip/ksz9477.c, overlapping changes.
One pretty prints the port mode differently, whilst another
changes the driver to try and obtain the port mode from
the port node rather than the switch node.
Signed-off-by: David S. Miller <davem@davemloft.net>
Seville does not need to depend on PCI or on the ENETC MDIO controller.
There will also be other compile-time differences in the future.
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Not only does Sevile not have a PTP clock, but with separate modules,
this structure cannot even live in felix.c, due to the .owner =
THIS_MODULE assignment causing this link time error:
drivers/net/dsa/ocelot/felix.o:(.data+0x0): undefined reference to `__this_module'
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Over the time, some patches have introduced structures aligned with
spaces, near structures aligned with tabs. Fix the inconsistencies.
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Currently mscc_ocelot_init_ports() will skip initializing a port when it
doesn't have a phy-handle, so the ocelot->ports[port] pointer will be
NULL. Take this into consideration when tearing down the driver, and add
a new function ocelot_deinit_port() to the switch library, mirror of
ocelot_init_port(), which needs to be called by the driver for all ports
it has initialized.
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.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>
ocelot_init() allocates memory, resets the switch and polls for a status
register, things which can fail. Stop probing the driver in that case,
and propagate the error result.
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>
We got slightly different patches removing a double word
in a comment in net/ipv4/raw.c - picked the version from net.
Simple conflict in drivers/net/ethernet/ibm/ibmvnic.c. Use cached
values instead of VNIC login response buffer (following what
commit 507ebe6444 ("ibmvnic: Fix use-after-free of VNIC login
response buffer") did).
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Use the helper functions introduced by the newly added
Lynx PCS MDIO module in the Felix VSC9959 and Seville VSC9953.
Instead of representing the PCS as a phy_device, a mdio_device structure
will be passed to the Lynx module which is now actually implementing all
the PCS configuration and status reporting.
All code previously used for PCS monitoring and runtime configuration
is removed and replaced will calls to the Lynx PCS operations.
Tested on the following SERDES protocols of LS1028A: 0x7777
(2500Base-X), 0x85bb (QSGMII), 0x9999 (SGMII) and 0x13bb (USXGMII).
Signed-off-by: Ioana Ciornei <ioana.ciornei@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Every iteration of for_each_available_child_of_node() decrements
the reference count of the previous node, however when control
is transferred from the middle of the loop, as in the case of
a return or break or goto, there is no decrement thus ultimately
resulting in a memory leak.
Fix a potential memory leak in felix.c by inserting of_node_put()
before the return statement.
Issue found with Coccinelle.
Signed-off-by: Sumera Priyadarsini <sylphrenadin@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This is another switch from Vitesse / Microsemi / Microchip, that has
10 ports (8 external, 2 internal) and is integrated into the Freescale /
NXP T1040 PowerPC SoC. It is very similar to Felix from NXP LS1028A,
except that this is a platform device and Felix is a PCI device, and it
doesn't support IEEE 1588 and TSN.
Like Felix, this driver configures its own PCS on the internal MDIO bus
using a phy_device abstraction for it (yes, it will be refactored to use
a raw mdio_device, like other phylink drivers do, but let's keep it like
that for now). But unlike Felix, the MDIO bus and the PCS are not from
the same vendor. The PCS is the same QorIQ/Layerscape PCS as found in
Felix/ENETC/DPAA*, but the internal MDIO bus that is used to access it
is actually an instantiation of drivers/net/phy/mdio-mscc-miim.c. But it
would be difficult to reuse that driver (it doesn't even use regmap, and
it's less than 200 lines of code), so we hand-roll here some internal
MDIO bus accessors within seville_vsc9953.c, which serves the purpose of
driving the PCS absolutely fine.
Also, same as Felix, the PCS doesn't support dynamic reconfiguration of
SerDes protocol, so we need to do pre-validation of PHY mode from device
tree and not let phylink change it.
Signed-off-by: Maxim Kochetkov <fido_max@inbox.ru>
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>
Felix is not actually meant to be a DSA driver only for the switch
inside NXP LS1028A, but an umbrella for all Vitesse / Microsemi /
Microchip switches that are register-compatible with Ocelot and that are
using in DSA mode (with an NPI Ethernet port).
For the dsa_switch_ops exported by the felix driver to be generic enough
to be used by other non-PCI switches, we need to move the PCI-specific
probing to the low-level translation module felix_vsc9959.c. This way,
other switches can have their own probing functions, as platform devices
or otherwise.
This patch also removes the "Felix instance table", which did not stand
the test of time and is unnecessary at this point.
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>
With this patch we try to kill 2 birds with 1 stone.
First of all, some switches that use tag_ocelot.c don't have the exact
same bitfield layout for the DSA tags. The destination ports field is
different for Seville VSC9953 for example. So the choices are to either
duplicate tag_ocelot.c into a new tag_seville.c (sub-optimal) or somehow
take into account a supposed ocelot->dest_ports_offset when packing this
field into the DSA injection header (again not ideal).
Secondly, tag_ocelot.c already needs to memset a 128-bit area to zero
and call some packing() functions of dubious performance in the
fastpath. And most of the values it needs to pack are pretty much
constant (BYPASS=1, SRC_PORT=CPU, DEST=port index). So it would be good
if we could improve that.
The proposed solution is to allocate a memory area per port at probe
time, initialize that with the statically defined bits as per chip
hardware revision, and just perform a simpler memcpy in the fastpath.
Other alternatives have been analyzed, such as:
- Create a separate tag_seville.c: too much code duplication for just 1
bit field difference.
- Create a separate DSA_TAG_PROTO_SEVILLE under tag_ocelot.c, just like
tag_brcm.c, which would have a separate .xmit function. Again, too
much code duplication for just 1 bit field difference.
- Allocate the template from the init function of the tag_ocelot.c
module, instead of from the driver: couldn't figure out a method of
accessing the correct port template corresponding to the correct
tagger in the .xmit function.
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>
Currently Felix and Ocelot share the same bit layout in these per-port
registers, but Seville does not. So we need reg_fields for that.
Actually since these are per-port registers, we need to also specify the
number of ports, and register size per port, and use the regmap API for
multiple ports.
There's a more subtle point to be made about the other 2 register
fields:
- QSYS_SWITCH_PORT_MODE_SCH_NEXT_CFG
- QSYS_SWITCH_PORT_MODE_INGRESS_DROP_MODE
which we are not writing any longer, for 2 reasons:
- Using the previous API (ocelot_write_rix), we were only writing 1 for
Felix and Ocelot, which was their hardware-default value, and which
there wasn't any intention in changing.
- In the case of SCH_NEXT_CFG, in fact Seville does not have this
register field at all, and therefore, if we want to have common code
we would be required to not write to it.
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>
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>
Phylink now requires that parameters established through
auto-negotiation be written into the MAC at the time of the
mac_link_up() callback. In the case of felix, that means taking the port
out of reset, setting the correct timers for PAUSE frames, and
enabling/disabling TX flow control.
This patch also splits the inband and noinband configuration of the
vsc9959 PCS (currently found in a function called "init") into 2
different functions, which have a nomenclature closer to phylink:
"config", for inband setup, and "link_up", for noinband (forced) setup.
This is necessary as a preparation step for giving up control of the PCS
to phylink, which will be done in further patch series.
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>
Phylink uses the .mac_an_restart method to offer the user an
implementation of the "ethtool -r" behavior, when the media-side auto
negotiation can be restarted by the local MAC PCS. This is the case for
fiber modes 1000Base-X and 2500Base-X (IEEE clause 37) that don't have
an Ethernet PHY connected locally, and the media is connected to the MAC
PCS directly.
On the other hand, the Cisco SGMII and USXGMII standards also have an
auto negotiation mechanism based on IEEE 802.3 clause 37 (their
respective specs require a MAC PCS and a PHY PCS to implement the same
state machine, which is described in IEEE 802.3 "Auto-Negotiation Figure
37-6"), so the ability to restart auto-negotiation is intrinsically
symmetrical (the MAC PCS can do it too).
However, it appears that not all SGMII/USXGMII PHYs have logic to
restart the MDI-side auto-negotiation process when they detect a
transition of the SGMII link from data mode to configuration mode.
Some do (VSC8234) and some don't (AR8033, MV88E1111). IEEE and/or Cisco
specification wordings to not help to prove whether propagating the "AN
restart" event from MII side ("mr_restart_an") to MDI side
("mr_restart_negotiation") is required behavior - neither of them
specifies any mandatory interaction between the clause 37 AN state
machine from Figure 37-6 and the clause 28 AN state machine from Figure
28-18.
Therefore, even if a certain behavior could be proven as being required,
real-life SGMII/USXGMII PHYs are inconsistent enough that a clause 37 AN
restart cannot be used by phylink to reliably trigger a media-side
renegotiation, when the user requests it via ethtool.
The only remaining use that the .mac_an_restart callback might possibly
have, given what we know now, is to implement some silicon quirks, but
so far that has proven to not be necessary.
So remove this code for now, since it never gets called and we don't
foresee any circumstance in which it might be, either.
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>
state->speed holds a value of 10, 100, 1000 or 2500, but
SYS_MAC_FC_CFG_FC_LINK_SPEED expects a value in the range 0, 1, 2 or 3.
So set the correct speed encoding into this register.
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>
In VSC9959, the PCS is the one who performs rate adaptation (symbol
duplication) to the speed negotiated by the PHY. The MAC is unaware of
that and must remain configured for gigabit. If it is configured at
OCELOT_SPEED_10 or OCELOT_SPEED_100, it'll start transmitting PAUSE
frames out of control and never recover, _even if_ we then reconfigure
it at OCELOT_SPEED_1000 afterwards.
This patch fixes a bug that luckily did not have any functional impact.
We were writing 10, 100, 1000 etc into this 2-bit field in
DEV_CLOCK_CFG, but the hardware expects values in the range 0, 1, 2, 3.
So all speed values were getting truncated to 0, which is
OCELOT_SPEED_2500, and which also appears to be fine.
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>
Ping tested:
[ 11.808455] mscc_felix 0000:00:00.5 swp0: Link is Up - 1Gbps/Full - flow control rx/tx
[ 11.816497] IPv6: ADDRCONF(NETDEV_CHANGE): swp0: link becomes ready
[root@LS1028ARDB ~] # ethtool -s swp0 advertise 0x4
[ 18.844591] mscc_felix 0000:00:00.5 swp0: Link is Down
[ 22.048337] mscc_felix 0000:00:00.5 swp0: Link is Up - 100Mbps/Half - flow control off
[root@LS1028ARDB ~] # ip addr add 192.168.1.1/24 dev swp0
[root@LS1028ARDB ~] # ping 192.168.1.2
PING 192.168.1.2 (192.168.1.2): 56 data bytes
(...)
^C--- 192.168.1.2 ping statistics ---
3 packets transmitted, 3 packets received, 0% packet loss
round-trip min/avg/max = 0.383/0.611/1.051 ms
[root@LS1028ARDB ~] # ethtool -s swp0 advertise 0x10
[ 355.637747] mscc_felix 0000:00:00.5 swp0: Link is Down
[ 358.788034] mscc_felix 0000:00:00.5 swp0: Link is Up - 1Gbps/Half - flow control off
[root@LS1028ARDB ~] # ping 192.168.1.2
PING 192.168.1.2 (192.168.1.2): 56 data bytes
(...)
^C
--- 192.168.1.2 ping statistics ---
16 packets transmitted, 16 packets received, 0% packet loss
round-trip min/avg/max = 0.301/0.384/1.138 ms
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Reviewed-by: Russell King <rmk+kernel@armlinux.org.uk>
Signed-off-by: David S. Miller <davem@davemloft.net>
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>
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>
xdp_umem.c had overlapping changes between the 64-bit math fix
for the calculation of npgs and the removal of the zerocopy
memory type which got rid of the chunk_size_nohdr member.
The mlx5 Kconfig conflict is a case where we just take the
net-next copy of the Kconfig entry dependency as it takes on
the ESWITCH dependency by one level of indirection which is
what the 'net' conflicting change is trying to ensure.
Signed-off-by: David S. Miller <davem@davemloft.net>
As explained in other commits before (b9cd75e668 and 87b0f983f6),
ocelot switches have a single egress-untagged VLAN per port, and the
driver would deny adding a second one while an egress-untagged VLAN
already exists.
But on the CPU port (where the VLAN configuration is implicit, because
there is no net device for the bridge to control), the DSA core attempts
to add a VLAN using the same flags as were used for the front-panel
port. This would make adding any untagged VLAN fail due to the CPU port
rejecting the configuration:
bridge vlan add dev swp0 vid 100 pvid untagged
[ 1865.854253] mscc_felix 0000:00:00.5: Port already has a native VLAN: 1
[ 1865.860824] mscc_felix 0000:00:00.5: Failed to add VLAN 100 to port 5: -16
(note that port 5 is the CPU port and not the front-panel swp0).
So this hardware will send all VLANs as tagged towards the CPU.
Fixes: 5605194877 ("net: dsa: ocelot: add driver for Felix switch family")
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>
The ocelot core library is written with the idea in mind that the VLAN
table is populated by the bridge. Otherwise, not even a sane default
pvid is provided: in standalone mode, the default pvid is 0, and the
core expects the bridge layer to change it to 1.
So without this patch, the VLAN table is completely empty at the end of
the commands below, and traffic is broken as a result:
ip link add dev br0 type bridge vlan_filtering 0 && ip link set dev br0 up
for eth in $(ls /sys/bus/pci/devices/0000\:00\:00.5/net/); do
ip link set dev $eth master br0
ip link set dev $eth up
done
ip link set dev br0 type bridge vlan_filtering 1
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>
The MSCC bug fix in 'net' had to be slightly adjusted because the
register accesses are done slightly differently in net-next.
Signed-off-by: David S. Miller <davem@davemloft.net>
The caller of devm_ioremap_resource(), either accidentally
or by wrong assumption, is writing back derived resource data
to global static resource initialization tables that should
have been constant. Meaning that after it computes the final
physical start address it saves the address for no reason
in the static tables. This doesn't affect the first driver
probing after reboot, but it breaks consecutive driver reloads
(i.e. driver unbind & bind) because the initialization tables
no longer have the correct initial values. So the next probe()
will map the device registers to wrong physical addresses,
causing ARM SError async exceptions.
This patch fixes all of the above.
Fixes: 5605194877 ("net: dsa: ocelot: add driver for Felix switch family")
Signed-off-by: Claudiu Manoil <claudiu.manoil@nxp.com>
Reviewed-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Tested-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Ocelot VSC9959 switch supports time-based egress shaping in hardware
according to IEEE 802.1Qbv. This patch add support for TAS configuration
on egress port of VSC9959 switch.
Felix driver is an instance of Ocelot family, with a DSA front-end. The
patch uses tc taprio hardware offload to setup TAS set function on felix
driver.
Signed-off-by: Xiaoliang Yang <xiaoliang.yang_1@nxp.com>
Reviewed-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Set the default QoS Classification based on PCP and DEI of vlan tag,
after that, frames can be Classified to different Qos based on PCP tag.
If there is no vlan tag or vlan ignored, use port default Qos.
Signed-off-by: Xiaoliang Yang <xiaoliang.yang_1@nxp.com>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
