| Age | Commit message (Collapse) | Author | Files | Lines |
|---|
|
Make use of the "shared port table" to convert i40e to the new
infra.
i40e did not have any reference tracking, locking is also dodgy
because rtnl gets released while talking to FW, so port may get
removed from the table while it's getting added etc.
On the good side i40e does not seem to be using the ports for
TX so we can remove the table from the driver state completely.
Signed-off-by: Jakub Kicinski <[email protected]>
Tested-by: Aaron Brown <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
|
|
The i40e NIC supports two flavors of HW descriptors, 16 and 32
byte. The latter has, obviously, room for more offloading
information. However, the only fields of the 32B HW descriptor that is
being used by the driver, is also available in the 16B descriptor.
In other words; Reading and writing 32 bytes instead of 16 byte is a
waste of bus bandwidth.
This commit starts using 16 byte descriptors instead of 32 byte
descriptors.
For AF_XDP the rx_drop benchmark was improved by 2%.
Signed-off-by: Björn Töpel <[email protected]>
Tested-by: Aaron Brown <[email protected]>
Signed-off-by: Tony Nguyen <[email protected]>
|
|
After OS requests to down a link on a physical network port, the
traffic is no longer being processed but the physical link with
a link partner is still established.
Currently there is a feature (Link down on close) which allows
to physically bring the link down (after OS request).
With this patch new feature with similar capability is introduced:
TOTAL_PORT_SHUTDOWN
Allows to physically disable the link on the NIC's port.
If enabled, (after link down request from the OS)
no link, traffic or led activity is possible on that port.
If I40E_FLAG_TOTAL_PORT_SHUTDOWN is enabled, the
I40E_FLAG_LINK_DOWN_ON_CLOSE_ENABLED must be explicitly forced to
true and cannot be disabled at that time.
The functionalities are exclusive in terms of configuration, but
they also have similar behavior (allowing to disable physical link
of the port), with following differences:
- LINK_DOWN_ON_CLOSE_ENABLED is configurable at host OS run-time
and is supported by whole family of 7xx Intel Ethernet Controllers
- TOTAL_PORT_SHUTDOWN may be enabled only before OS loads (in BIOS)
only if motherboard's BIOS and NIC's FW has support of it
- when LINK_DOWN_ON_CLOSE_ENABLED is used, the link is being brought
down by sending phy_type=0 to NIC's FW
- when TOTAL_PORT_SHUTDOWN is used, phy_type is not altered, instead
the link is being brought down by clearing bit
(I40E_AQ_PHY_ENABLE_LINK) in abilities field of
i40e_aq_set_phy_config structure
Introduced changes:
- new private flag I40E_FLAG_TOTAL_PORT_SHUTDOWN for handling the
feature
- probe of NVM if the feature was enabled at driver's port
initialization
- special handling on link-down procedure to let FW physically
shutdown the port if the feature was enabled
Signed-off-by: Arkadiusz Kubalewski <[email protected]>
Signed-off-by: Aleksandr Loktionov <[email protected]>
Tested-by: Andrew Bowers <[email protected]>
Signed-off-by: Tony Nguyen <[email protected]>
|
|
As with other networking drivers, remove the unnecessary driver version
from the Intel drivers. The ethtool driver information and module version
will then report the kernel version instead.
For ixgbe, i40e and ice drivers, the driver passes the driver version to
the firmware to confirm that we are up and running. So we now pass the
value of UTS_RELEASE to the firmware. This adminq call is required per
the HAS document. The Device then sends an indication to the BMC that the
PF driver is present. This is done using Host NC Driver Status Indication
in NC-SI Get Link command or via the Host Network Controller Driver Status
Change AEN.
What the BMC may do with this information is implementation-dependent, but
this is a standard NC-SI 1.1 command we honor per the HAS.
CC: Bruce Allan <[email protected]>
CC: Jesse Brandeburg <[email protected]>
CC: Alek Loktionov <[email protected]>
CC: Kevin Liedtke <[email protected]>
CC: Aaron Rowden <[email protected]>
Signed-off-by: Jeff Kirsher <[email protected]>
Co-developed-by: Jacob Keller <[email protected]>
Tested-by: Aaron Brown <[email protected]>
|
|
Remove all the unused defines as they are just dead weight.
Signed-off-by: Jesse Brandeburg <[email protected]>
Tested-by: Andrew Bowers <[email protected]>
Signed-off-by: Jeff Kirsher <[email protected]>
|
|
Move i40e_client.h to include/linux/net/intel/*
since its shared between i40iw and i40e.
Signed-off-by: Shiraz Saleem <[email protected]>
Tested-by: Andrew Bowers <[email protected]>
Signed-off-by: Jeff Kirsher <[email protected]>
|
|
The current codebase makes use of the zero-length array language
extension to the C90 standard, but the preferred mechanism to declare
variable-length types such as these ones is a flexible array member[1][2],
introduced in C99:
struct foo {
int stuff;
struct boo array[];
};
By making use of the mechanism above, we will get a compiler warning
in case the flexible array does not occur last in the structure, which
will help us prevent some kind of undefined behavior bugs from being
inadvertently introduced[3] to the codebase from now on.
Also, notice that, dynamic memory allocations won't be affected by
this change:
"Flexible array members have incomplete type, and so the sizeof operator
may not be applied. As a quirk of the original implementation of
zero-length arrays, sizeof evaluates to zero."[1]
This issue was found with the help of Coccinelle.
[1] https://gcc.gnu.org/onlinedocs/gcc/Zero-Length.html
[2] https://github.com/KSPP/linux/issues/21
[3] commit 76497732932f ("cxgb3/l2t: Fix undefined behaviour")
Signed-off-by: Gustavo A. R. Silva <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
|
|
Use synchronize_rcu to wait until the XSK wakeup function finishes
before destroying the resources it uses:
1. i40e_down already calls synchronize_rcu. On i40e_down either
__I40E_VSI_DOWN or __I40E_CONFIG_BUSY is set. Check the latter in
i40e_xsk_wakeup (the former is already checked there).
2. After switching the XDP program, call synchronize_rcu to let
i40e_xsk_wakeup exit before the XDP program is freed.
3. Changing the number of channels brings the interface down (see
i40e_prep_for_reset and i40e_pf_quiesce_all_vsi).
4. Disabling UMEM sets __I40E_CONFIG_BUSY, too.
Signed-off-by: Maxim Mikityanskiy <[email protected]>
Signed-off-by: Björn Töpel <[email protected]>
Signed-off-by: Daniel Borkmann <[email protected]>
Link: https://lore.kernel.org/bpf/[email protected]
|
|
This patch prepares ground for the next VF MAC address change fix.
It lets untrusted VF to delete any VF mac filter, but it still
doesn't let untrusted VF to add mac filter not setup by PF.
It removes information duplication in num_mac mac filters counter.
And improves exact h/w mac filters usage checking in the
i40e_check_vf_permission() function by counting mac2add_cnt.
It also improves logging because now all mac addresses will be validated
first and corresponding messages will be logged.
Signed-off-by: Aleksandr Loktionov <[email protected]>
Tested-by: Andrew Bowers <[email protected]>
|
|
The macros used to specify the cloud filter fields are intended to be
individual bits. Declare them using the BIT() macro to make their
intention a little more clear.
Signed-off-by: Jacob Keller <[email protected]>
Tested-by: Andrew Bowers <[email protected]>
Signed-off-by: Jeff Kirsher <[email protected]>
|
|
Since commit
'5098850c9b9b ("i40e/i40evf: i40e_register.h updates")'
it is no longer possible to trigger an EMP Reset from debugfs, but it's
possible to request it either way, to end up with a bad reset request:
echo empr > /sys/kernel/debug/i40e/0002\:01\:00.1/command
i40e 0002:01:00.1: debugfs: forcing EMPR
i40e 0002:01:00.1: bad reset request 0x00010000
So let's remove this piece of code and show the available valid commands
as it is when any invalid command is issued.
Signed-off-by: "Mauro S. M. Rodrigues" <[email protected]>
Tested-by: Andrew Bowers <[email protected]>
Signed-off-by: Jeff Kirsher <[email protected]>
|
|
This patch fixes missing call of i40e_update_veb_stats() in function
i40e_get_ethtool_stats() to update stats data of VEB and VEB TC
counters before they are written into ethtool buffer.
Before the patch ethtool counters may fell behind interface counters.
Signed-off-by: Dmitrii Golovanov <[email protected]>
Tested-by: Andrew Bowers <[email protected]>
Signed-off-by: Jeff Kirsher <[email protected]>
|
|
This patch enables macvlan offloads for i40e. The idea is to use
channels as macvlan interfaces. The channels are VSIs of
type VMDQ. When the first macvlan is created, the maximum number of
channels possible are created. From then on, as a macvlan interface
is created, a macvlan filter is added to these already created
channels (VSIs).
This patch utilizes subordinate device traffic classes to make queue
groups(channels) available for an upper device like a macvlan.
Steps to configure macvlan offloads:
1. ethtool -K ethx l2-fwd-offload on
2. ip link add link ethx name macvlan1 type macvlan
3. ip addr add <address> dev macvlan1
4. ip link set macvlan1 up
Signed-off-by: Harshitha Ramamurthy <[email protected]>
Tested-by: Andrew Bowers <[email protected]>
Signed-off-by: Jeff Kirsher <[email protected]>
|
|
Changing descriptor count via 'ethtool -G' is not persistent across resets.
When PF reset occurs, we roll back to the default value of vsi->num_desc,
which is used then in i40e_alloc_rings to set descriptor count. XDP does a
PF reset so when user has changed the descriptor count and load XDP
program, the default count will be back there.
To fix this:
* introduce new VSI members - num_tx_desc and num_rx_desc in favour of
num_desc
* set them in i40e_set_ringparam to user's values
* set them to default values in i40e_set_num_rings_in_vsi only when they
don't have previous values
Signed-off-by: Maciej Fijalkowski <[email protected]>
Tested-by: Andrew Bowers <[email protected]>
Signed-off-by: Jeff Kirsher <[email protected]>
|
|
Remove references to I40E_ETH_P_LLDP and use ETH_P_LLDP instead.
Signed-off-by: Anirudh Venkataramanan <[email protected]>
Tested-by: Andrew Bowers <[email protected]>
Signed-off-by: Jeff Kirsher <[email protected]>
|
|
This patch introduces "recovery mode" to the i40e driver. It is
part of a new Any2Any idea of upgrading the firmware. In this
approach, it is required for the driver to have support for
"transition firmware", that is used for migrating from structured
to flat firmware image. In this new, very basic mode, i40e driver
must be able to handle particular IOCTL calls from the NVM Update
Tool and run a small set of AQ commands.
These additional AQ commands are part of the interface used by
the NVMUpdate tool. The NVMUpdate tool contains all of the
necessary logic to reference these new AQ commands. The end user
experience remains the same, they are using the NVMUpdate tool to
update the NVM contents.
Signed-off-by: Alice Michael <[email protected]>
Signed-off-by: Piotr Marczak <[email protected]>
Tested-by: Don Buchholz <[email protected]>
Signed-off-by: Jeff Kirsher <[email protected]>
|
|
In the case where PTP is running on the hardware clock, but the kernel
system time is not being synced, a device reset can mess up the clock
time.
This occurs because we reset the clock time based on the kernel time
every reset. This causes us to potentially completely reset the PTP
time, and can cause unexpected behavior in programs like ptp4l.
Avoid this by saving the PTP time prior to device reset, and then
restoring using that time after the reset.
Directly restoring the PTP time we saved isn't perfect, because time
should have continued running, but the clock will essentially be stopped
during the reset. This is still better than the current solution of
assuming that the PTP HW clock is synced to the CLOCK_REALTIME.
We can do even better, by saving the ktime and calculating
a differential, using ktime_get(). This is based on CLOCK_MONOTONIC, and
allows us to get a fairly precise measure of the time difference between
saving and restoring the time.
Using this, we can update the saved PTP time, and use that as the value
to write to the hardware clock registers. This, of course is not perfect.
However, it does help ensure that the PTP time is restored as close as
feasible to the time it should have been if the reset had not occurred.
During device initialization, continue using the system time as the
source for the creation of the PTP clock, since this is the best known
current time source at driver load.
Signed-off-by: Jacob Keller <[email protected]>
Signed-off-by: Jeff Kirsher <[email protected]>
|
|
This patch introduces DDP (Dynamic Device Personalization) which allows
loading profiles that change the way internal parser interprets processed
frames. To load DDP profiles it utilizes ethtool flash feature. The files
with recipes must be located in /var/lib/firmware directory. Afterwards
the recipe can be loaded by invoking:
ethtool -f <if_name> <file_name> 100
ethtool -f <if_name> - 100
See further details of this feature in the i40e documentation, or
visit
https://www.intel.com/content/www/us/en/architecture-and-technology/ethernet/dynamic-device-personalization-brief.html
The driver shall verify DDP profile can be loaded in accordance with
the rules:
* Package with Group ID 0 are exclusive and can only be loaded the first.
* Packages with Group ID 0x01-0xFE can only be loaded simultaneously
with the packages from the same group.
* Packages with Group ID 0xFF are compatible with all other packages.
Signed-off-by: Aleksandr Loktionov <[email protected]>
Tested-by: Andrew Bowers <[email protected]>
Signed-off-by: Jeff Kirsher <[email protected]>
|
|
In commit f3fef2b6e1cc ("i40e: Remove umem from VSI") a regression was
introduced; When the VSI was reset, the setup code would try to enable
AF_XDP ZC unconditionally (as long as there was a umem placed in the
netdev._rx struct). Here, we add a bitmap to the VSI that tracks if a
certain queue pair has been "zero-copy enabled" via the ndo_bpf. The
bitmap is used in i40e_xsk_umem, and enables zero-copy if and only if
XDP is enabled, the corresponding qid in the bitmap is set and the
umem is non-NULL.
Fixes: f3fef2b6e1cc ("i40e: Remove umem from VSI")
Signed-off-by: Björn Töpel <[email protected]>
Tested-by: Aaron Brown <[email protected]>
Signed-off-by: Jeff Kirsher <[email protected]>
|
|
The i40e_xsk_umem function was explicitly inlined in i40e.h. There is
no reason for that, so move it to i40e_main.c instead.
Signed-off-by: Björn Töpel <[email protected]>
Tested-by: Aaron Brown <[email protected]>
Signed-off-by: Jeff Kirsher <[email protected]>
|
|
This patch adds support for setting/getting FEC configuration
using ethtool options:
set/show-priv-flags rs-fec/base-r-fec
set/show-fec off/rs/baser/auto for kernels version >= 4.14
Signed-off-by: Damian Dybek <[email protected]>
Tested-by: Andrew Bowers <[email protected]>
Signed-off-by: Jeff Kirsher <[email protected]>
|
|
As current implementation of netdev already contains and provides
umems for us, we no longer have the need to contain these
structures in i40e_vsi.
Refactor the code to operate on netdev-provided umems.
Signed-off-by: Jan Sokolowski <[email protected]>
Acked-by: Björn Töpel <[email protected]>
Tested-by: Andrew Bowers <[email protected]>
Signed-off-by: Jeff Kirsher <[email protected]>
|
|
If a TX hang occurs, we attempt to recover by incrementally resetting.
If we're starved for CPU time, it's possible the reset doesn't actually
complete (or even fire) before another tx_timeout fires causing us to
fly through the different resets without actually doing them.
This adds a bit to set and check if a timeout recovery is already
pending and, if so, bail out of tx_timeout. The bit will get cleared at
the end of i40e_rebuild when reset is complete.
Signed-off-by: Alan Brady <[email protected]>
Tested-by: Andrew Bowers <[email protected]>
Signed-off-by: Jeff Kirsher <[email protected]>
|
|
This patch allows disabling FW LLDP agent on X722 devices.
It also changes a source of information for this feature from
pf->hw_features to pf->hw.flags which are set in i40e_init_adminq.
Signed-off-by: Patryk Małek <[email protected]>
Tested-by: Andrew Bowers <[email protected]>
Signed-off-by: Jeff Kirsher <[email protected]>
|
|
A scenario has been found in which simultaneous
addition/removal and modification of VF's might cause
unstable behaviour, up to and including kernel panics.
Protect the methods that create/modify/destroy VF's
by locking them behind an atomically set bit in PF status
bitfield.
Signed-off-by: Jan Sokolowski <[email protected]>
Tested-by: Andrew Bowers <[email protected]>
Signed-off-by: Jeff Kirsher <[email protected]>
|
|
This patch adds zero-copy Rx support for AF_XDP sockets. Instead of
allocating buffers of type MEM_TYPE_PAGE_SHARED, the Rx frames are
allocated as MEM_TYPE_ZERO_COPY when AF_XDP is enabled for a certain
queue.
All AF_XDP specific functions are added to a new file, i40e_xsk.c.
Note that when AF_XDP zero-copy is enabled, the XDP action XDP_PASS
will allocate a new buffer and copy the zero-copy frame prior passing
it to the kernel stack.
Signed-off-by: Björn Töpel <[email protected]>
Signed-off-by: Alexei Starovoitov <[email protected]>
|
|
The expectation of the ops VIRTCHNL_OP_ENABLE_QUEUES and
VIRTCHNL_OP_DISABLE_QUEUES is that the queue map sent by
the VF is taken into account when enabling/disabling
queues in the VF VSI. This patch makes sure that happens.
By breaking out the individual queue set up functions so
that they can be called directly from the i40e_virtchnl_pf.c
file, only the queues as specified by the queue bit map that
accompanies the enable/disable queues ops will be handled.
Signed-off-by: Harshitha Ramamurthy <[email protected]>
Tested-by: Andrew Bowers <[email protected]>
Signed-off-by: Jeff Kirsher <[email protected]>
|
|
When operating at 1GbE, the base incval for the PTP clock is so large
that multiplying it by numbers close to the max_adj can overflow the
u64.
Rather than attempting to limit the max_adj to a value small enough to
avoid overflow, instead calculate the incvalue adjustment based on the
40GbE incvalue, and then multiply that by the scaling factor for the
link speed.
This sacrifices a small amount of precision in the adjustment but we
avoid erratic behavior of the clock due to the overflow caused if ppb is
very near the maximum adjustment.
Signed-off-by: Jacob Keller <[email protected]>
Tested-by: Andrew Bowers <[email protected]>
Signed-off-by: Jeff Kirsher <[email protected]>
|
|
This fixes at least 2 issues I have found with the UDP tunnel filter
configuration.
The first issue is the fact that the tunnels didn't have any sort of mutual
exclusion in place to prevent an update from racing with a user request to
add/remove a port. As such you could request to add and remove a port
before the port update code had a chance to respond which would result in a
very confusing result. To address it I have added 2 changes. First I added
the RTNL mutex wrapper around our updating of the pending, port, and
filter_index bits. Second I added logic so that we cannot use a port that
has a pending deletion since we need to free the space in hardware before
we can allow software to reuse it.
The second issue addressed is the fact that we were not recording the
actual filter index provided to us by the admin queue. As a result we were
deleting filters that were not associated with the actual filter we wanted
to delete. To fix that I added a filter_index member to the UDP port
tracking structure.
Signed-off-by: Alexander Duyck <[email protected]>
Tested-by: Andrew Bowers <[email protected]>
Signed-off-by: Jeff Kirsher <[email protected]>
|
|
After many years of having a ~30 line copyright and license header to our
source files, we are finally able to reduce that to one line with the
advent of the SPDX identifier.
Also caught a few files missing the SPDX license identifier, so fixed
them up.
Signed-off-by: Jeff Kirsher <[email protected]>
Acked-by: Shannon Nelson <[email protected]>
Acked-by: Richard Cochran <[email protected]>
Tested-by: Andrew Bowers <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
|
|
Remove the gaps created by the recent refactor of various feature flags
that have moved to the state field. Use only a u32 now that we have
fewer than 32 flags in the field.
Signed-off-by: Jacob Keller <[email protected]>
Tested-by: Andrew Bowers <[email protected]>
Signed-off-by: Jeff Kirsher <[email protected]>
|
|
The iWarp client flags are all potentially changed when the RTNL lock is
not held, so they should not be part of the pf->flags variable. Instead,
move them into the state field so that we can use atomic bit operations.
This is part of a larger effort to remove cmpxchg64 in
i40e_set_priv_flags()
Signed-off-by: Jacob Keller <[email protected]>
Tested-by: Andrew Bowers <[email protected]>
Signed-off-by: Jeff Kirsher <[email protected]>
|
|
This flag is modified outside of the RTNL lock and thus should not be
part of the pf->flags variable.
Use a state bit instead, so that we can use atomic bit operations.
This is part of a larger effort to remove cmpxchg64 in
i40e_set_priv_flags()
Signed-off-by: Jacob Keller <[email protected]>
Tested-by: Andrew Bowers <[email protected]>
Signed-off-by: Jeff Kirsher <[email protected]>
|
|
The two Flow Directory auto disable flags are used at run time to mark
when the flow director features needed to be disabled. Thus the flags
could change even when the RTNL lock is not held.
They also have some code constructions which really should be
test_and_set or test_and_clear using atomic bit operations.
Create new state fields to mark this, and stop including them in
pf->flags.
This is part of a larger effort to remove the need for cmpxchg64 in
i40e_set_priv_flags().
Signed-off-by: Jacob Keller <[email protected]>
Tested-by: Andrew Bowers <[email protected]>
Signed-off-by: Jeff Kirsher <[email protected]>
|
|
This flag is modified during run time, possibly even when the RTNL lock
is not held. Additionally it has a few places which should be using
test_and_set or test_and_clear atomic bit operations.
Create a new state bit __I40E_UDP_SYNC_PENDING and use it instead of the
ole I40E_FLAG_UDP_FILTER_SYNC flag.
This is part of a larger effort to remove the need for using cmpxchg64
in i40e_set_priv_flags.
Signed-off-by: Jacob Keller <[email protected]>
Tested-by: Andrew Bowers <[email protected]>
Signed-off-by: Jeff Kirsher <[email protected]>
|
|
The I40E_FLAG_FILTER_SYNC flag is modified during run time possibly when
the RTNL lock is not held. Thus, it should not be part of pf->flags, and
instead should be using atomic bit operations in the pf->state field.
Create a __I40E_MACVLAN_SYNC_PENDING state bit, and use it instead of
the old I40E_FLAG_FILTER_SYNC flag.
This is part of a larger effort to remove the need for cmpxchg64 in
i40e_set_priv_flags().
Signed-off-by: Jacob Keller <[email protected]>
Tested-by: Andrew Bowers <[email protected]>
Signed-off-by: Jeff Kirsher <[email protected]>
|
|
Add the SPDX identifiers to all the Intel wired LAN driver files, as
outlined in Documentation/process/license-rules.rst.
Signed-off-by: Jeff Kirsher <[email protected]>
Tested-by: Aaron Brown <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
|
|
During suspend client MSIx vectors are freed while they are still
in use causing a crash on entering S3.
Fix this calling client close before freeing up its MSIx vectors.
Also update the client MSIx vectors on resume before client
open is called.
Fixes commit b980c0634fe5 ("i40e: shutdown all IRQs and disable MSI-X
when suspended")
Reported-by: Stefan Assmann <[email protected]>
Signed-off-by: Shiraz Saleem <[email protected]>
Tested-by: Andrew Bowers <[email protected]>
Signed-off-by: Jeff Kirsher <[email protected]>
|
|
This patch provides support to add or delete cloud filter for queue
channels created for ADq on VF.
We are using the HW's cloud filter feature and programming it to act
as a TC filter applied to a group of queues.
There are two possible modes for a VF when applying a cloud filter
1. Basic Mode: Intended to apply filters that don't need a VF to be
Trusted. This would include the following
Dest MAC + L4 port
Dest MAC + VLAN + L4 port
2. Advanced Mode: This mode is only for filters with combination that
requires VF to be Trusted.
Dest IP + L4 port
When cloud filters are applied on a trusted VF and for some reason
the same VF is later made as untrusted then all cloud filters
will be deleted. All cloud filters has to be re-applied in
such a case.
Cloud filters are also deleted when queue channel is deleted.
Testing-Hints:
=============
1. Adding Basic Mode filter should be possible on a VF in
Non-Trusted mode.
2. In Advanced mode all filters should be able to be created.
Steps:
======
1. Enable ADq and create TCs using TC mqprio command
2. Apply cloud filter.
3. Turn-off the spoof check.
4. Pass traffic.
Example:
========
1. tc qdisc add dev enp4s2 root mqprio num_tc 4 map 0 0 0 0 1 2 2 3\
queues 2@0 2@2 1@4 1@5 hw 1 mode channel
2. tc qdisc add dev enp4s2 ingress
3. ethtool -K enp4s2 hw-tc-offload on
4. ip link set ens261f0 vf 0 spoofchk off
5. tc filter add dev enp4s2 protocol ip parent ffff: prio 1 flower\
dst_ip 192.168.3.5/32 ip_proto udp dst_port 25 skip_sw hw_tc 2
Signed-off-by: Avinash Dayanand <[email protected]>
Tested-by: Andrew Bowers <[email protected]>
Signed-off-by: Jeff Kirsher <[email protected]>
|
|
The MAC, FW Version and NPAR check used to determine
if shutting off the FW LLDP engine is supported is not
using the usual feature check mechanism.
This patch fixes the problem by moving the feature check
to i40e_sw_init in order to set a flag in pf->hw_features
that ethtool will use for priv_flags disable operation.
Signed-off-by: Dave Ertman <[email protected]>
Tested-by: Andrew Bowers <[email protected]>
Signed-off-by: Jeff Kirsher <[email protected]>
|
|
This patch replaces the existing mechanism for determining the correct
value to program for adaptive ITR with yet another new and more
complicated approach.
The basic idea from a 30K foot view is that this new approach will push the
Rx interrupt moderation up so that by default it starts in low latency and
is gradually pushed up into a higher latency setup as long as doing so
increases the number of packets processed, if the number of packets drops
to 4 to 1 per packet we will reset and just base our ITR on the size of the
packets being received. For Tx we leave it floating at a high interrupt
delay and do not pull it down unless we start processing more than 112
packets per interrupt. If we start exceeding that we will cut our interrupt
rates in half until we are back below 112.
The side effect of these patches are that we will be processing more
packets per interrupt. This is both a good and a bad thing as it means we
will not be blocking processing in the case of things like pktgen and XDP,
but we will also be consuming a bit more CPU in the cases of things such as
network throughput tests using netperf.
One delta from this versus the ixgbe version of the changes is that I have
made the interrupt moderation a bit more aggressive when we are in bulk
mode by moving our "goldilocks zone" up from 48 to 96 to 56 to 112. The
main motivation behind moving this is to address the fact that we need to
update less frequently, and have more fine grained control due to the
separate Tx and Rx ITR times.
Signed-off-by: Alexander Duyck <[email protected]>
Tested-by: Andrew Bowers <[email protected]>
Signed-off-by: Jeff Kirsher <[email protected]>
|
|
As we have added more flags, we need to now use more
bits and have over flooded the 32 bit size. So
make it 64.
Also change all the existing bits to unsigned long long
bits.
Signed-off-by: Alice Michael <[email protected]>
Tested-by: Andrew Bowers <[email protected]>
Signed-off-by: Jeff Kirsher <[email protected]>
|
|
It is safe to remove the upper limit of 64 queues on a channel
VSI. The upper bound is determined by the VSI's num_queue_pairs
and gets validated when the queue mapping info through mqprio
interface is subject to bound checking in the driver.
Signed-off-by: Amritha Nambiar <[email protected]>
Tested-by: Andrew Bowers <[email protected]>
Signed-off-by: Jeff Kirsher <[email protected]>
|
|
This patch enables tc-flower based hardware offloads. tc flower
filter provided by the kernel is configured as driver specific
cloud filter. The patch implements functions and admin queue
commands needed to support cloud filters in the driver and
adds cloud filters to configure these tc-flower filters.
The classification function of the filter is to direct matched
packets to a traffic class. The hardware traffic class is set
based on the the classid reserved in the range :ffe0 - :ffef.
Match Dst MAC and route to TC0:
prio 1 flower dst_mac 3c:fd:fe:a0:d6:70 skip_sw\
hw_tc 1
Match Dst IPv4,Dst Port and route to TC1:
prio 2 flower dst_ip 192.168.3.5/32\
ip_proto udp dst_port 25 skip_sw\
hw_tc 2
Match Dst IPv6,Dst Port and route to TC1:
prio 3 flower dst_ip fe8::200:1\
ip_proto udp dst_port 66 skip_sw\
hw_tc 2
Delete tc flower filter:
Example:
Flow Director Sideband is disabled while configuring cloud filters
via tc-flower and until any cloud filter exists.
Unsupported matches when cloud filters are added using enhanced
big buffer cloud filter mode of underlying switch include:
1. source port and source IP
2. Combined MAC address and IP fields.
3. Not specifying L4 port
These filter matches can however be used to redirect traffic to
the main VSI (tc 0) which does not require the enhanced big buffer
cloud filter support.
Signed-off-by: Amritha Nambiar <[email protected]>
Signed-off-by: Kiran Patil <[email protected]>
Signed-off-by: Anjali Singhai Jain <[email protected]>
Signed-off-by: Jingjing Wu <[email protected]>
Acked-by: Shannon Nelson <[email protected]>
Tested-by: Andrew Bowers <[email protected]>
Signed-off-by: Jeff Kirsher <[email protected]>
|
|
Introduce the cloud filter data structure and cleanup of cloud
filters associated with the device.
Signed-off-by: Amritha Nambiar <[email protected]>
Acked-by: Shannon Nelson <[email protected]>
Tested-by: Andrew Bowers <[email protected]>
Signed-off-by: Jeff Kirsher <[email protected]>
|
|
Add mapping of TCs with the seids of the channel VSIs. TC0
will be mapped to the main VSI seid and all other TCs are
mapped to the seid of the corresponding channel VSI.
Signed-off-by: Amritha Nambiar <[email protected]>
Acked-by: Shannon Nelson <[email protected]>
Tested-by: Andrew Bowers <[email protected]>
Signed-off-by: Jeff Kirsher <[email protected]>
|
|
Commit 52eb1ff93e98 ("i40e: Add support setting TC max bandwidth rates")
and commit 1ea6f21ae530 ("i40e: Refactor VF BW rate limiting") add some
needed functionality for TC bandwidth rate limiting. Unfortunately they
introduce several usages of unsigned 64-bit division which needs to be
handled special by the kernel to support all architectures.
Fixes: 52eb1ff93e98 ("i40e: Add support setting TC max bandwidth
rates")
Fixes: 1ea6f21ae530 ("i40e: Refactor VF BW rate limiting")
Signed-off-by: Alan Brady <[email protected]>
Tested-by: Andrew Bowers <[email protected]>
Signed-off-by: Jeff Kirsher <[email protected]>
|
|
This patch enables setting up maximum Tx rates for the traffic
classes in i40e. The maximum rate is offloaded to the hardware through
the mqprio framework by specifying the mode option as 'channel' and
shaper option as 'bw_rlimit' and is configured for the VSI. Configuring
minimum Tx rate limit is not supported in the device. The minimum
usable value for Tx rate is 50Mbps.
Example:
# tc qdisc add dev eth0 root mqprio num_tc 2 map 0 0 0 0 1 1 1 1\
queues 4@0 4@4 hw 1 mode channel shaper bw_rlimit\
max_rate 4Gbit 5Gbit
To dump the bandwidth rates:
# tc qdisc show dev eth0
qdisc mqprio 804a: root tc 2 map 0 0 0 0 1 1 1 1 0 0 0 0 0 0 0 0
queues:(0:3) (4:7)
mode:channel
shaper:bw_rlimit max_rate:4Gbit 5Gbit
Signed-off-by: Amritha Nambiar <[email protected]>
Tested-by: Andrew Bowers <[email protected]>
Signed-off-by: Jeff Kirsher <[email protected]>
|
|
This patch refactors the BW rate limiting for Tx traffic
on the VF to be reused in the next patch for rate limiting Tx
traffic for the VSIs on the PF as well.
Signed-off-by: Amritha Nambiar <[email protected]>
Tested-by: Andrew Bowers <[email protected]>
Signed-off-by: Jeff Kirsher <[email protected]>
|
|
The i40e driver is modified to enable the new mqprio hardware
offload mode and factor the TCs and queue configuration by
creating channel VSIs. In this mode, the priority to traffic
class mapping and the user specified queue ranges are used
to configure the traffic classes by setting the mode option to
'channel'.
Example:
map 0 0 0 0 1 2 2 3 queues 2@0 2@2 1@4 1@5\
hw 1 mode channel
qdisc mqprio 8038: root tc 4 map 0 0 0 0 1 2 2 3 0 0 0 0 0 0 0 0
queues:(0:1) (2:3) (4:4) (5:5)
mode:channel
shaper:dcb
The HW channels created are removed and all the queue configuration
is set to default when the qdisc is detached from the root of the
device.
This patch also disables setting up channels via ethtool (ethtool -L)
when the TCs are configured using mqprio scheduler.
The patch also limits setting ethtool Rx flow hash indirection
(ethtool -X eth0 equal N) to max queues configured via mqprio.
The Rx flow hash indirection input through ethtool should be
validated so that it is within in the queue range configured via
tc/mqprio. The bound checking is achieved by reporting the current
rss size to the kernel when queues are configured via mqprio.
Example:
map 0 0 0 1 0 2 3 0 queues 2@0 4@2 8@6 11@14\
hw 1 mode channel
Cannot set RX flow hash configuration: Invalid argument
Signed-off-by: Amritha Nambiar <[email protected]>
Tested-by: Andrew Bowers <[email protected]>
Signed-off-by: Jeff Kirsher <[email protected]>
|