When rxhash is enabled on any ethernet port except the first in each CP
block, traffic flow is prevented. The analysis is below:
I've been investigating this afternoon, and what I've found, comparing
a kernel without 895586d5dc and with 895586d5dc applied is:
- The table programmed into the hardware via mvpp22_rss_fill_table()
appears to be identical with or without the commit.
- When rxhash is enabled on eth2, mvpp2_rss_port_c2_enable() reports
that c2.attr[0] and c2.attr[2] are written back containing:
- with 895586d5dc, failing: 00200000 40000000
- without 895586d5dc, working: 04000000 40000000
- When disabling rxhash, c2.attr[0] and c2.attr[2] are written back as:
04000000 00000000
The second value represents the MVPP22_CLS_C2_ATTR2_RSS_EN bit, the
first value is the queue number, which comprises two fields. The high
5 bits are 24:29 and the low three are 21:23 inclusive. This comes
from:
c2.attr[0] = MVPP22_CLS_C2_ATTR0_QHIGH(qh) |
MVPP22_CLS_C2_ATTR0_QLOW(ql);
So, the working case gives eth2 a queue id of 4.0, or 32 as per
port->first_rxq, and the non-working case a queue id of 0.1, or 1.
The allocation of queue IDs seems to be in mvpp2_port_probe():
if (priv->hw_version == MVPP21)
port->first_rxq = port->id * port->nrxqs;
else
port->first_rxq = port->id * priv->max_port_rxqs;
Where:
if (priv->hw_version == MVPP21)
priv->max_port_rxqs = 8;
else
priv->max_port_rxqs = 32;
Making the port 0 (eth0 / eth1) have port->first_rxq = 0, and port 1
(eth2) be 32. It seems the idea is that the first 32 queues belong to
port 0, the second 32 queues belong to port 1, etc.
mvpp2_rss_port_c2_enable() gets the queue number from it's parameter,
'ctx', which comes from mvpp22_rss_ctx(port, 0). This returns
port->rss_ctx[0].
mvpp22_rss_context_create() is responsible for allocating that, which
it does by looking for an unallocated priv->rss_tables[] pointer. This
table is shared amongst all ports on the CP silicon.
When we write the tables in mvpp22_rss_fill_table(), the RSS table
entry is defined by:
u32 sel = MVPP22_RSS_INDEX_TABLE(rss_ctx) |
MVPP22_RSS_INDEX_TABLE_ENTRY(i);
where rss_ctx is the context ID (queue number) and i is the index in
the table.
If we look at what is written:
- The first table to be written has "sel" values of 00000000..0000001f,
containing values 0..3. This appears to be for eth1. This is table 0,
RX queue number 0.
- The second table has "sel" values of 00000100..0000011f, and appears
to be for eth2. These contain values 0x20..0x23. This is table 1,
RX queue number 0.
- The third table has "sel" values of 00000200..0000021f, and appears
to be for eth3. These contain values 0x40..0x43. This is table 2,
RX queue number 0.
How do queue numbers translate to the RSS table? There is another
table - the RXQ2RSS table, indexed by the MVPP22_RSS_INDEX_QUEUE field
of MVPP22_RSS_INDEX and accessed through the MVPP22_RXQ2RSS_TABLE
register. Before 895586d5dc, it was:
mvpp2_write(priv, MVPP22_RSS_INDEX,
MVPP22_RSS_INDEX_QUEUE(port->first_rxq));
mvpp2_write(priv, MVPP22_RXQ2RSS_TABLE,
MVPP22_RSS_TABLE_POINTER(port->id));
and after:
mvpp2_write(priv, MVPP22_RSS_INDEX, MVPP22_RSS_INDEX_QUEUE(ctx));
mvpp2_write(priv, MVPP22_RXQ2RSS_TABLE, MVPP22_RSS_TABLE_POINTER(ctx));
Before the commit, for eth2, that would've contained '32' for the
index and '1' for the table pointer - mapping queue 32 to table 1.
Remember that this is queue-high.queue-low of 4.0.
After the commit, we appear to map queue 1 to table 1. That again
looks fine on the face of it.
Section 9.3.1 of the A8040 manual seems indicate the reason that the
queue number is separated. queue-low seems to always come from the
classifier, whereas queue-high can be from the ingress physical port
number or the classifier depending on the MVPP2_CLS_SWFWD_PCTRL_REG.
We set the port bit in MVPP2_CLS_SWFWD_PCTRL_REG, meaning that queue-high
comes from the MVPP2_CLS_SWFWD_P2HQ_REG() register... and this seems to
be where our bug comes from.
mvpp2_cls_oversize_rxq_set() sets this up as:
mvpp2_write(port->priv, MVPP2_CLS_SWFWD_P2HQ_REG(port->id),
(port->first_rxq >> MVPP2_CLS_OVERSIZE_RXQ_LOW_BITS));
val = mvpp2_read(port->priv, MVPP2_CLS_SWFWD_PCTRL_REG);
val |= MVPP2_CLS_SWFWD_PCTRL_MASK(port->id);
mvpp2_write(port->priv, MVPP2_CLS_SWFWD_PCTRL_REG, val);
Setting the MVPP2_CLS_SWFWD_PCTRL_MASK bit means that the queue-high
for eth2 is _always_ 4, so only queues 32 through 39 inclusive are
available to eth2. Yet, we're trying to tell the classifier to set
queue-high, which will be ignored, to zero. Hence, the queue-high
field (MVPP22_CLS_C2_ATTR0_QHIGH()) from the classifier will be
ignored.
This means we end up directing traffic from eth2 not to queue 1, but
to queue 33, and then we tell it to look up queue 33 in the RSS table.
However, RSS table has not been programmed for queue 33, and so it ends
up (presumably) dropping the packets.
It seems that mvpp22_rss_context_create() doesn't take account of the
fact that the upper 5 bits of the queue ID can't actually be changed
due to the settings in mvpp2_cls_oversize_rxq_set(), _or_ it seems that
mvpp2_cls_oversize_rxq_set() has been missed in this commit. Either
way, these two functions mutually disagree with what queue number
should be used.
Looking deeper into what mvpp2_cls_oversize_rxq_set() and the MTU
validation is doing, it seems that MVPP2_CLS_SWFWD_P2HQ_REG() is used
for over-sized packets attempting to egress through this port. With
the classifier having had RSS enabled and directing eth2 traffic to
queue 1, we may still have packets appearing on queue 32 for this port.
However, the only way we may end up with over-sized packets attempting
to egress through eth2 - is if the A8040 forwards frames between its
ports. From what I can see, we don't support that feature, and the
kernel restricts the egress packet size to the MTU. In any case, if we
were to attempt to transmit an oversized packet, we have no support in
the kernel to deal with that appearing in the port's receive queue.
So, this patch attempts to solve the issue by clearing the
MVPP2_CLS_SWFWD_PCTRL_MASK() bit, allowing MVPP22_CLS_C2_ATTR0_QHIGH()
from the classifier to define the queue-high field of the queue number.
My testing seems to confirm my findings above - clearing this bit
means that if I enable rxhash on eth2, the interface can then pass
traffic, as we are now directing traffic to RX queue 1 rather than
queue 33. Traffic still seems to work with rxhash off as well.
Reported-by: Matteo Croce <mcroce@redhat.com>
Tested-by: Matteo Croce <mcroce@redhat.com>
Fixes: 895586d5dc ("net: mvpp2: cls: Use RSS contexts to handle RSS tables")
Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk>
Signed-off-by: David S. Miller <davem@davemloft.net>
The "info->fs.location" is a u32 that comes from the user via the
ethtool_set_rxnfc() function. We need to check for invalid values to
prevent a buffer overflow.
I copy and pasted this check from the mvpp2_ethtool_cls_rule_ins()
function.
Fixes: 90b509b39a ("net: mvpp2: cls: Add Classification offload support")
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
flow_action_hw_stats_types_check() helper takes one of the
FLOW_ACTION_HW_STATS_*_BIT values as input. If we align
the arguments to the opening bracket of the helper there
is no way to call this helper and stay under 80 characters.
Remove the "types" part from the new flow_action helpers
and enum values.
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Introduce flow_action_basic_hw_stats_types_check() helper and use it
in drivers. That sanitizes the drivers which do not have support
for action HW stats types.
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Users can specify classification actions based on the 'ether' flow type.
In that case, this will apply to all ethernet traffic, superseeding
flows such as 'udp4' or 'tcp6'.
Add support for this flow type in the PPv2 classifier, by mapping the
ETHER_FLOW value to the corresponding entries in the classifier.
Signed-off-by: Maxime Chevallier <maxime.chevallier@bootlin.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add a missing check to detect flow types that we don't support, so that
user can be informed of this.
Signed-off-by: Maxime Chevallier <maxime.chevallier@bootlin.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This commit allows using the vlan Id and priority as parts of the key
for classification offload. These fields are extracted from the
outermost tag, if multiple tags are present.
Vlan Id and priority are considered as 2 different fields by the
classifier, however the fields are both appended in the Header Extracted
Key in the same layout as they are found in the tags. This means that
when steering only based on the prio, a 16-bit slot is still taken in
the HEK.
The classifier doesn't allow extracting the DEI bit from the tag, so we
explicitly prevent user from using this bit in the key.
This commit adds the vlan priotity as a compatible HEK field for
tagged traffic, meaning that we limit the possibility of extracting this
field only to the flows that contain tagged traffic.
Signed-off-by: Maxime Chevallier <maxime.chevallier@bootlin.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The C2 TCAM used for classification uses a key (Header Extracted Key)
built by concatenating several fields extracted from the packet header.
After a lot of trial-and-error and some guess work, it seems the HEK is
right justified, with the first fields being stored in the MSB, then
concatenated up until the LSB.
Until now, this doesn't cause any issue since all HEK fields we use are
full bytes. However this is an issue for the upcoming VLAN id and pri
extraction, which aren't full bytes.
Rework the way we built that TCAM key, by changing the order in which we
append the fields.
Signed-off-by: Maxime Chevallier <maxime.chevallier@bootlin.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The way we currently handle classification offload and RSS is by having
dedicated lookup sequences in the flow table, each being selected
depending on several fields being present in the packet header.
We need to make sure the classification operation we want to perform can
be done in each flow we want to insert it into. As an example,
classifying on VLAN tag can only be done on flows used for tagged
traffic.
This commit makes sure we don't insert rules in flows we aren't
compatible with.
Signed-off-by: Maxime Chevallier <maxime.chevallier@bootlin.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
When performing a TCAM lookup in the C2 engine, it's possible that
multiple entries match the packet. To make sure the correct entry match
when performing a lookup, the Flow Table can set a lookup type, which
will be used in the TCAM lookup, thus preventing such false-positives.
We need to make sure the RSS match doesn't interfere with other
classification lookups, hence we use a dedicated lookup_type for it.
Signed-off-by: Maxime Chevallier <maxime.chevallier@bootlin.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The phylink conflict was between a bug fix by Russell King
to make sure we have a consistent PHY interface mode, and
a change in net-next to pull some code in phylink_resolve()
into the helper functions phylink_mac_link_{up,down}()
On the dp83867 side it's mostly overlapping changes, with
the 'net' side removing a condition that was supposed to
trigger for RGMII but because of how it was coded never
actually could trigger.
Signed-off-by: David S. Miller <davem@davemloft.net>
Fix smatch warning:
drivers/net/ethernet/marvell/mvpp2/mvpp2_cls.c:1236
mvpp2_ethtool_cls_rule_ins() warn: unsigned 'info->fs.location' is never less than zero.
'info->fs.location' is u32 type, never less than zero.
Signed-off-by: YueHaibing <yuehaibing@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Make sure we don't use an out-of-bound index for the per-port RSS
context array.
As of today, the global context creation in mvpp22_rss_context_create
will prevent us from reaching this case, but we should still make sure
we are using a sane value anyway.
Reported-by: kbuild test robot <lkp@intel.com>
Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Maxime Chevallier <maxime.chevallier@bootlin.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
When steering to an RXQ, we can perform an extra RSS step to assign a
queue from an RSS table.
This is done by setting the RSS_EN attribute in the C2 engine. In that
case, the RXQ that is assigned is the global RSS context id, that is
then translated to an RSS table using the RXQ2RSS table.
An example using ethtool to steer to RXQ 2 and 3 would be :
ethtool -X eth0 weight 0 0 1 1 context new
(This would print the allocated context id, let's say it's 1)
ethtool -N eth0 flow-type udp4 dst-port 1234 context 1 loc 0
The hash parameters are the ones that are globally configured for RSS :
ethtool -N eth0 rx-flow-hash udp4 sdfn
When an RSS context is removed while there are active classification
rules using this context, these rules are removed.
Signed-off-by: Maxime Chevallier <maxime.chevallier@bootlin.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
ethtool_rx_flow_rule_create takes into parameter the ethtool flow spec,
which doesn't contain the rss context id. We therefore need to extract
it ourself before parsing the ethtool rule.
The FLOW_RSS flag is only set in info->fs.flow_type, and not
info->flow_type.
Signed-off-by: Maxime Chevallier <maxime.chevallier@bootlin.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The PPv2 controller has 8 RSS tables that are shared across all ports on
a given PPv2 instance. The previous implementation allocated one table
per port, leaving others unused.
By using RSS contexts, we can make use of multiple RSS tables per
port, one being the default table (always id 0), the other ones being
used as destinations for flow steering, in the same way as rx rings.
This commit introduces RSS contexts management in the PPv2 driver. We
always reserve one table per port, allocated when the port is probed.
The global table list is stored in the struct mvpp2, as it's a global
resource. Each port then maintains a list of indices in that global
table, that way each port can have it's own numbering scheme starting
from 0.
One limitation that seems unavoidable is that the hashing parameters are
shared across all RSS contexts for a given port. Hashing parameters for
ctx 0 will be applied to all contexts.
Signed-off-by: Maxime Chevallier <maxime.chevallier@bootlin.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The C2 TCAM has internal FIFOs that are only useful for the built-in
self-tests. Disable these FIFOS at init, as recommended in the
functionnal specs.
Suggested-by: Alan Winkowski <walan@marvell.com>
Signed-off-by: Maxime Chevallier <maxime.chevallier@bootlin.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
As of today, the classification offload implementation only supports 4
different rules to be offloaded. This number has been hardcoded in the
rule insertion function, and the wrong define is being used elsewhere.
Use the correct #define everywhere to make sure we always check for the
correct number of rules.
Signed-off-by: Maxime Chevallier <maxime.chevallier@bootlin.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The flow_rule is only used when configuring the classification tables,
and should be free'd once we're done using it. The current code only
frees it in the error path.
Fixes: 90b509b39a ("net: mvpp2: cls: Add Classification offload support")
Signed-off-by: Maxime Chevallier <maxime.chevallier@bootlin.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The signed return from the call to mvpp2_cls_c2_port_flow_index is being
assigned to the u32 variable c2.index and then checked for a negative
error condition which is always going to be false. Fix this by assigning
the return to the int variable index and checking this instead.
Addresses-Coverity: ("Unsigned compared against 0")
Fixes: 90b509b39a ("net: mvpp2: cls: Add Classification offload support")
Signed-off-by: Colin Ian King <colin.king@canonical.com>
Reviewed-by: Maxime Chevallier <maxime.chevallier@bootlin.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Fixes gcc '-Wunused-but-set-variable' warning:
drivers/net/ethernet/marvell/mvpp2/mvpp2_cls.c: In function 'mvpp2_cls_c2_build_match':
drivers/net/ethernet/marvell/mvpp2/mvpp2_cls.c:1159:28: warning:
variable 'act' set but not used [-Wunused-but-set-variable]
It is never used since introduction in
commit 90b509b39a ("net: mvpp2: cls: Add Classification offload support")
Signed-off-by: YueHaibing <yuehaibing@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This commit introduces support for the "Drop" action in classification
offload. This corresponds to the "-1" action with ethtool -N.
This is achieved using the color marking actions available in the C2
engine, which associate a color to a packet. These colors can be either
Green, Yellow or Red, Red meaning that the packet should be dropped.
Green and Yellow colors are interpreted by the Policer, which isn't
supported yet.
This method of dropping using the Classifier is different than the
already existing early-drop features, such as VLAN filtering and MAC
UC/MC filtering, which are performed during the Parsing step, and
therefore take precedence over classification actions.
Signed-off-by: Maxime Chevallier <maxime.chevallier@bootlin.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This commit introduces basic classification offloading support for the
PPv2 controller.
The PPv2 classifier has many classification engines, for now we only use
the C2 TCAM match engine.
This engine allows to perform ternary lookups on 64 bits keys (called
Header Extracted Key), that are built by extracting fields from the packet
header and concatenating them. At most 4 fields can be extracted for a
single lookup.
This basic implementation allows to build the HEK from the following
fields :
- L4 source and destination ports (for UDP and TCP)
More fields are to be added in the future.
Classification flows are added through the ethtool interface, using the
newly introduced flow_rule infrastructure as an internal rule
representation, allowing to more easily implement tc flower rules if
need be.
The internal design for now allocates one range of 4 rules per port
due to the internal design of the flow table, which uses 22 sub-flows.
When inserting a classification rule, the rule is created in every
relevant sub-flow.
This low rule-count is a very simple design which reaches quickly the
limitations of the flow table ordering, but guarantees that the rule
ordering will always be respected.
This commit only introduces support for the "steer to rxq" action.
Signed-off-by: Maxime Chevallier <maxime.chevallier@bootlin.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
As of today, the classification code is used only for RSS. We split the
incoming traffic into multiple flows, that correspond to the ethtool
flow_type parameter.
We don't want to use the ethtool flow definitions such as TCP_V4_FLOW,
for several reason :
- We want to decorrelate the driver code from ethtool as much as
possible, so that we can easily use other interfaces such as tc flower,
- We want the flow_type to be a bitfield, so that we can match flows
embedded into each other, such as TCP4 which is a subset of IP4.
This commit does the conversion to the newer type.
Signed-off-by: Maxime Chevallier <maxime.chevallier@bootlin.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Cosmetic patch removing extra whitespaces when writing the flow_table
entries
Signed-off-by: Maxime Chevallier <maxime.chevallier@bootlin.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
When classifying a packet pertaining to a given flow, the classifier
will issue multiple lookup commands until it finds one with the 'last'
bit set. It expects all prorities to be assign continuously (although
not necessarily in an ordered fashion) from 0 to the number of lookups.
We can initialize this once, and make sure unused lookups are given an
empty port map. This avoids having to maintain priorities and the
information of which lookup is the last.
Signed-off-by: Maxime Chevallier <maxime.chevallier@bootlin.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
C2 TCAM entries can be invalidated to avoid unwanted matches. Make sure
all entries are invalidated at init, then validate only the ones we use.
Signed-off-by: Maxime Chevallier <maxime.chevallier@bootlin.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The Flow Table dictates what lookups will be issued for each flow type.
The lookup sequence for each flow is similar, and the index of each
lookup is computed by some macros.
There are similar mechanisms for the C2 TCAM lookups, so in order to
avoid confusion, rename the flow table index computing macros with a
common prefix.
The only difference in behaviour is that we now use the very first entry
in the flow for the RSS lookup (the first entry was previously unused).
Signed-off-by: Maxime Chevallier <maxime.chevallier@bootlin.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The classifier allows to combine multiple lookups in one "sequence" that
is counted as a single lookup to an engine, with a single result.
We don't actually use that feature, so remove any places where we set
this field, so that the classifier doesn't try to interpret these
fields.
Signed-off-by: Maxime Chevallier <maxime.chevallier@bootlin.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This commit renames some of the classifier functions to follow the
naming 'mvpp2_port_*' that's used for function that act on a given port.
This commit is purely cosmetic.
Signed-off-by: Maxime Chevallier <maxime.chevallier@bootlin.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Move C2 read/write helpers higher in the file to ease future work that
rely on these helpers
Signed-off-by: Maxime Chevallier <maxime.chevallier@bootlin.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
When writing a C2 entry to hardware, some registers writes will only
take effect when the TCAM_DATA4 register is written. This includes all
C2 TCAM registers, and the C2 invalidate register.
To make sure we always write C2 entries correctly, document that
behaviour with a comment, and move TCAM writes to the end of the
mvpp2_cls_c2_write helper.
Signed-off-by: Maxime Chevallier <maxime.chevallier@bootlin.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The cls_table is a global read-only table containing the different
parameters that are used by various tables in the classifier. It
describes the links between the Header Parser, the decoding table and
the flow_table.
There are several possible way we want to iterate over that table,
depending on wich classifier engine we want to configure. For the Header
Parser, we want to iterate over each entry. For the Decoding table, we
want to iterate over each entry having a unique flow_id. Finally, when
configuring an ethtool flow, we want to iterate over each entry having a
unique flow_id and that has a given flow_type.
This commit introduces some iterator to both provide syntactic sugar and
also clarify the way we want to iterate over the table.
Signed-off-by: Maxime Chevallier <maxime.chevallier@bootlin.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The cls_flow table represent the overall configuration of the
classifier, used to match the different traffic classes in the Parsing
and Classification engines.
This configuration is static, and applies to all PPv2 instances, we must
therefore keep it const so that no modifications of this table are
performed at runtime.
Signed-off-by: Maxime Chevallier <maxime.chevallier@bootlin.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The macro definition MVPP2_N_FLOWS is ambiguous because it really
represents the number of entries in the Header Parser that are used to
identify the classification flows.
Rename the macro to clearly state that we represent the number of flows
in the Header Parser.
Signed-off-by: Maxime Chevallier <maxime.chevallier@bootlin.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The PPv2 classifier allows to perform multiple lookups on the same
engine when classifying a packet. These lookups can match similar parts
of a packet header, but perform different actions upon matching. To
differentiate these types of lookups, it's possible to specify a Lookup
Type in the flow table entries, which will be part of the key for the
lookup engines.
This commit introduces the use of Lookup Types for C2 matches. Since for
now we only perform C2 lookups to enable RSS, we only need one Lookup
Type.
Signed-off-by: Maxime Chevallier <maxime.chevallier@bootlin.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
PPv2's classifier supports extracting the MAC Destination Address from the
L2 header to perform RSS and flow steering. Add the missing case when
setting the Header Extracted Key fields in the flow table.
Signed-off-by: Maxime Chevallier <maxime.chevallier@bootlin.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The classification operations that are used for RSS make use of several
lookup tables. Having hit counters for these tables is really helpful
to determine what flows were matched by ingress traffic, and see the
path of packets among all the classifier tables.
This commit adds hit counters for the 3 tables used at the moment :
- The decoding table (also called lookup_id table), that links flows
identified by the Header Parser to the flow table.
There's one entry per flow, located at :
.../mvpp2/<controller>/flows/XX/dec_hits
Note that there are 21 flows in the decoding table, whereas there are
52 flows in the Header Parser. That's because there are several kind
of traffic that will match a given flow. Reading the hit counter from
one sub-flow will clear all hit counter that have the same flow_id.
This also applies to the flow_hits.
- The flow table, that contains all the different lookups to be
performed by the classifier for each packet of a given flow. The match
is done on the first entry of the flow sequence.
- The C2 engine entries, that are used to assign the default rx queue,
and enable or disable RSS for a given port.
There's one entry per flow, located at:
.../mvpp2/<controller>/flows/XX/flow_hits
There is one C2 entry per port, so the c2 hit counter is located at :
.../mvpp2/<controller>/ethX/c2_hits
All hit counter values are 16-bits clear-on-read values.
Signed-off-by: Maxime Chevallier <maxime.chevallier@bootlin.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The classifier configuration for RSS is quite complex, with several
lookup tables being used. This commit adds useful info in debugfs to
see how the different tables are configured :
Added 2 new entries in the per-port directory :
- .../eth0/default_rxq : The default rx queue on that port
- .../eth0/rss_enable : Indicates if RSS is enabled in the C2 entry
Added the 'flows' directory :
It contains one entry per sub-flow. a 'sub-flow' is a unique path from
Header Parser to the flow table. Multiple sub-flows can point to the
same 'flow' (each flow has an id from 8 to 29, which is its index in the
Lookup Id table) :
- .../flows/00/...
/01/...
...
/51/id : The flow id. There are 21 unique flows. There's one
flow per combination of the following parameters :
- L4 protocol (TCP, UDP, none)
- L3 protocol (IPv4, IPv6)
- L3 parameters (Fragmented or not)
- L2 parameters (Vlan tag presence or not)
.../type : The flow type. This is an even higher level flow,
that we manipulate with ethtool. It can be :
"udp4" "tcp4" "udp6" "tcp6" "ipv4" "ipv6" "other".
.../eth0/...
.../eth1/engine : The hash generation engine used for this
flow on the given port
.../hash_opts : The hash generation options indicating on
what data we base the hash (vlan tag, src
IP, src port, etc.)
Signed-off-by: Maxime Chevallier <maxime.chevallier@bootlin.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Use the appropriate SPDX license identifiers and drop the license text.
This patch is only cosmetic.
Signed-off-by: Antoine Tenart <antoine.tenart@bootlin.com>
Signed-off-by: Maxime Chevallier <maxime.chevallier@bootlin.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Fixes: f9358e12a0 ("net: mvpp2: split ingress traffic into multiple flows")
Signed-off-by: kbuild test robot <fengguang.wu@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This commit allows setting the RSS hash generation parameters from
ethtool. When setting parameters for a given flow type from ethtool
(e.g. tcp4), all the corresponding flows in the flow table are updated,
according to the supported hash parameters.
For example, when configuring TCP over IPv4 hash parameters to be
src/dst IP + src/dst port ("ethtool -N eth0 rx-flow-hash tcp4 sdfn"),
we only set the "src/dst port" hash parameters on the non-fragmented TCP
over IPv4 flows.
Signed-off-by: Maxime Chevallier <maxime.chevallier@bootlin.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
One of the classification action that can be performed is to compute a
hash of the packet header based on some header fields, and lookup a RSS
table based on this hash to determine the final RxQ.
This is done by adding one lookup entry per flow per port, so that we
can configure the hash generation parameters for each flow and each
port.
There are 2 possible engines that can be used for RSS hash generation :
- C3HA, that generates a hash based on up to 4 header-extracted fields
- C3HB, that does the same as c3HA, but also includes L4 info in the hash
There are a lot of fields that can be extracted from the header. For now,
we only use the ones that we can configure using ethtool :
- DST MAC address
- L3 info
- Source IP
- Destination IP
- Source port
- Destination port
The C3HB engine is selected when we use L4 fields (src/dst port).
Header parser Dec table
Ingress pkt +-------------+ flow id +----------------------------+
------------->| TCAM + SRAM |-------->|TCP IPv4 w/ VLAN, not frag |
+-------------+ |TCP IPv4 w/o VLAN, not frag |
|TCP IPv4 w/ VLAN, frag |--+
|etc. | |
+----------------------------+ |
|
Flow table |
+---------+ +------------+ +--------------------------+ |
| RSS tbl |<--| Classifier |<--------| flow 0: C2 lookup | |
+---------+ +------------+ | C3 lookup port 0 | |
| | | C3 lookup port 1 | |
+-----------+ +-------------+ | ... | |
| C2 engine | | C3H engines | | flow 1: C2 lookup |<--+
+-----------+ +-------------+ | C3 lookup port 0 |
| ... |
| ... |
| flow 51 : C2 lookup |
| ... |
+--------------------------+
The C2 engine also gains the role of enabling and disabling the RSS
table lookup for this packet.
Signed-off-by: Maxime Chevallier <maxime.chevallier@bootlin.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The PPv2 classifier allows to perform classification operations on each
ingress packet, based on the flow the packet is assigned to.
The current code uses only 1 flow per port, and the only classification
action consists of assigning the rx queue to the packet, depending on the
port.
In preparation for adding RSS support, we have to split all incoming
traffic into different flows. Since RSS assigns a rx queue depending on
the hash of some header fields, we have to make sure that the hash is
generated in a consistent way for all packets in the same flow.
What we call a "flow" is actually a set of attributes attached to a
packet that depends on various L2/L3/L4 info.
This patch introduces 52 flows, wich are a combination of various L2, L3
and L4 attributes :
- Whether or not the packet has a VLAN tag
- Whether the packet is IPv4, IPv6 or something else
- Whether the packet is TCP, UDP or something else
- Whether or not the packet is fragmented at L3 level.
The flow is associated to a packet by the Header Parser. Each flow
corresponds to an entry in the decoding table. This entry then points to
the sequence of classification lookups to be performed by the
classifier, represented in the flow table.
For now, the only lookup we perform is a C2 lookup to set the default
rx queue.
Header parser Dec table
Ingress pkt +-------------+ flow id +----------------------------+
------------->| TCAM + SRAM |-------->|TCP IPv4 w/ VLAN, not frag |
+-------------+ |TCP IPv4 w/o VLAN, not frag |
|TCP IPv4 w/ VLAN, frag |--+
|etc. | |
+----------------------------+ |
|
Flow table |
+------------+ +---------------------+ |
To RxQ <---| Classifier |<-------| flow 0: C2 lookup |<--------+
+------------+ | flow 1: C2 lookup |
| | ... |
+------------+ | flow 51 : C2 lookup |
| C2 engine | +---------------------+
+------------+
Signed-off-by: Maxime Chevallier <maxime.chevallier@bootlin.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The PPv2 Controller has a classifier, that can perform multiple lookup
operations for each packet, using different engines.
One of these engines is the C2 engine, which performs TCAM based lookups
on data extracted from the packet header. When a packet matches an
entry, the engine sets various attributes, used to perform
classification operations.
One of these attributes is the rx queue in which the packet should be sent.
The current code uses the lookup_id table (also called decoding table)
to assign the rx queue. However, this only works if we use one entry per
port in the decoding table, which won't be the case once we add RSS
lookups.
This patch uses the C2 engine to assign the rx queue to each packet.
The C2 engine is used through the flow table, which dictates what
classification operations are done for a given flow.
Right now, we have one flow per port, which contains every ingress
packet for this port.
Signed-off-by: Maxime Chevallier <maxime.chevallier@bootlin.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
mvpp22_init_rss function configures the RSS parameters for each port, so
rename it accordingly. Since this function relies on classifier
configuration, move its call right after the classifier config.
Signed-off-by: Maxime Chevallier <maxime.chevallier@bootlin.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
When filling the RSS table, we have to make sure that the rx queue is
attached to an online CPU.
This patch is not a full support for cpu_hotplug, but rather a way to
make sure that we don't break network on system booted with the maxcpus
parameter.
Signed-off-by: Maxime Chevallier <maxime.chevallier@bootlin.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch adds an extra indirection when setting the indirection table
into the RSS hardware table to improve the packets distribution across
CPUs. For example, if 2 queues are used on a multi-core system this new
indirection will choose two queues on two different CPUs instead of the
two first queues which are on the same first CPU.
Signed-off-by: Antoine Tenart <antoine.tenart@bootlin.com>
Signed-off-by: Maxime Chevallier <maxime.chevallier@bootlin.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch adds the RSS indirection table support, allowing to use the
ethtool -x and -X options to dump and set this table.
Signed-off-by: Antoine Tenart <antoine.tenart@bootlin.com>
[Maxime: Small warning fixes, use one table per port]
Signed-off-by: Maxime Chevallier <maxime.chevallier@bootlin.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
PPv2 Controller has 8 RSS Tables, of 32 entries each. A lookup in the
RXQ2RSS_TABLE is performed for each incoming packet, and the RSS Table
to be used is chosen according to the default rx queue that would be
used for the packet.
This default rx queue is set in the Lookup_id Table (also called
Decoding Table), and is equal to the port->first_rxq.
Since the Classifier itself isn't active at any time for the moment,
this doesn't have a direct effect, the default rx queue at the moment is
the one where all packets end-up into.
Signed-off-by: Maxime Chevallier <maxime.chevallier@bootlin.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
