| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
When the admin enables/disables cqe compression, updating
mpwqe stride size is required:
CQE compress ON ==> stride size = 256B
CQE compress OFF ==> stride size = 64B
This is already done on driver load via mlx5e_set_rq_type_params, all we
need is just to call it on arbitrary admin changes of cqe compression
state via priv flags or when changing timestamping state
(as it is mutually exclusive with cqe compression).
This bug introduces no functional damage, it only makes cqe compression
occur less often, since in ConnectX4-LX CQE compression is performed
only on packets smaller than stride size.
Tested:
ethtool --set-priv-flags ethxx rx_cqe_compress on
pktgen with 64 < pkt size < 256 and netperf TCP_STREAM (IPv4/IPv6)
verify `ethtool -S ethxx | grep compress` are advancing more often
(rapidly)
Fixes: 7219ab34f184 ("net/mlx5e: CQE compression")
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Reviewed-by: Tariq Toukan <tariqt@mellanox.com>
Cc: kernel-team@fb.com
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
4K Uar series modified the mlx5e driver to use the new bfreg API,
and mistakenly removed the sq->uar_map iomem data path dedicated
pointer, which was meant to be read from xmit path for cache locality
utilization.
Fix that by returning that pointer to the SQ struct.
Fixes: 7309cb4ad71e ("IB/mlx5: Support 4k UAR for libmlx5")
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Reviewed-by: Tariq Toukan <tariqt@mellanox.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
ConnectX-5 and later HW generations will report min inline mode ==
MLX5_INLINE_MODE_NONE, which means driver is not required to copy packet
headers to inline fields of TX WQE.
Avoid copy to inline segment in XDP TX routine when HW inline mode doesn't
require it.
This will improve CPU utilization and boost XDP TX performance.
Tested with xdp2 single flow:
CPU: Intel(R) Xeon(R) CPU E5-2680 v3 @ 2.50GHz
HCA: Mellanox Technologies MT28800 Family [ConnectX-5 Ex]
Before: 7.4Mpps
After: 7.8Mpps
Improvement: 5%
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Reviewed-by: Tariq Toukan <tariqt@mellanox.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
There is a hardware feature that will pad the start or end of a DMA to
be cache line aligned to avoid RMWs on the last cache line. The default
cache line size setting for this feature is 64B. This change configures
the hardware to use 128B alignment on systems with 128B cache lines.
In addition we lower bound MPWRQ stride by HCA cacheline in mlx5e,
MPWRQ stride should be at least the HCA cacheline, the current default
is 64B and in case HCA_CAP.cach_line_128byte capability is set, MPWRQ RX
stride will automatically be aligned to 128B.
Signed-off-by: Daniel Jurgens <danielj@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
|
|\
| |
| |
| |
| |
| | |
All merge conflicts were simple overlapping changes.
Signed-off-by: David S. Miller <davem@davemloft.net>
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Modifying TIR hash should change selected fields bitmask in addition to
the function and key.
Formerly, Only on ethool mlx5e_set_rxfh "ethtoo -X" we would not set this
field resulting in zeroing of its value, which means no packet fields are
used for RX RSS hash calculation thus causing all traffic to arrive in
RQ[0].
On driver load out of the box we don't have this issue, since the TIR
hash is fully created from scratch.
Tested:
ethtool -X ethX hkey <new key>
ethtool -X ethX hfunc <new func>
ethtool -X ethX equal <new indirection table>
All cases are verified with TCP Multi-Stream traffic over IPv4 & IPv6.
Fixes: bdfc028de1b3 ("net/mlx5e: Fix ethtool RX hash func configuration change")
Signed-off-by: Gal Pressman <galp@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
As ENOTSUPP is specific to NFS, change the return error value to
EOPNOTSUPP in various places in the mlx5 driver.
Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com>
Suggested-by: Yotam Gigi <yotamg@mellanox.com>
Reviewed-by: Matan Barak <matanb@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
|
| |
| |
| |
| |
| |
| |
| |
| | |
This patch is intended for code reuse of mlx5e_modify_rx_cqe_compression
function.
Signed-off-by: Shaker Daibes <shakerd@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
|
| |
| |
| |
| |
| |
| |
| |
| | |
Reduce memory consumption on kdump kernel by decreasing the number of
channels to 1 and the size of RQs and SQs to the minimal values.
Signed-off-by: Kamal Heib <kamalh@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Today when the driver enter to promiscuous mode or vlan
filter is disabled, we add flow rule to receive any c-taggd
packets, therefore s-tagged packets are dropped.
In order to receive s-tagged packets as well we need to add
flow rule to receive any s-tagged packet.
Fixes: 7cb21b794baa ('net/mlx5e: Rename en_flow_table.c to en_fs.c')
Signed-off-by: Mohamad Haj Yahia <mohamad@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
This patch enables the 1PPS IN and 1PPS OUT support according
to the advertised HCA capability. Single pin may be configured
to one of the above mutual exclusive functions via standard
Linux tools and APIs. For example, testptp open source application.
Signed-off-by: Eugenia Emantayev <eugenia@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
This patch adds bpf_xdp_adjust_head() support to mlx5e.
1. rx_headroom is added to struct mlx5e_rq. It uses
an existing 4 byte hole in the struct.
2. The adjusted data length is checked against
MLX5E_XDP_MIN_INLINE and MLX5E_SW2HW_MTU(rq->netdev->mtu).
3. The macro MLX5E_SW2HW_MTU is moved from en_main.c to en.h.
MLX5E_HW2SW_MTU is also moved to en.h for symmetric reason
but it is not a must.
v2:
- Keep the xdp specific logic in mlx5e_xdp_handle()
- Update dma_len after the sanity checks in mlx5e_xmit_xdp_frame()
Signed-off-by: Martin KaFai Lau <kafai@fb.com>
Acked-by: Saeed Mahameed <saeedm@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Add fields to structs to convey to kernel an indication whether the
library supports multi UARs per page and return to the library the size
of a UAR based on the queried value.
Signed-off-by: Eli Cohen <eli@mellanox.com>
Reviewed-by: Matan Barak <matanb@mellanox.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
|
|/
|
|
|
|
|
|
|
|
|
|
|
|
| |
Make use of the blue flame registers allocator at mlx5_ib. Since blue
flame was not really supported we remove all the code that is related to
blue flame and we let all consumers to use the same blue flame register.
Once blue flame is supported we will add the code. As part of this patch
we also move the definition of struct mlx5_bf to mlx5_ib.h as it is only
used by mlx5_ib.
Signed-off-by: Eli Cohen <eli@mellanox.com>
Reviewed-by: Matan Barak <matanb@mellanox.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
|
|\ |
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
When using the negative logic (i.e. FLUSH state), after the RQ/SQ reopen
we will have a time interval that the RQ/SQ is not really ready and the
state indicates that its not in FLUSH state because the initial SQ/RQ struct
memory starts as zeros.
Now we changed the state to indicate if the SQ/RQ is opened and we will
set the READY state after finishing preparing all the SQ/RQ resources.
Fixes: 6e8dd6d6f4bd ("net/mlx5e: Don't wait for SQ completions on close")
Fixes: f2fde18c52a7 ("net/mlx5e: Don't wait for RQ completions on close")
Signed-off-by: Mohamad Haj Yahia <mohamad@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
In Striding RQ implementation, we used a single UMR
(User-Mode Memory Registration) memory key for all RQs.
When the product of RQs number*size gets high, we hit a
limitation of u16 field size in FW.
Here we move to using a UMR memory key per RQ, so we can
scale to any number of rings, with the maximum buffer
size in each.
Signed-off-by: Tariq Toukan <tariqt@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Add new type of struct mlx5_frag_buf which is used to allocate fragmented
buffers rather than contiguous, and make the Completion Queues (CQs) use
it as they are big (default of 2MB per CQ in Striding RQ).
This fixes the failures of type:
"mlx5e_open_locked: mlx5e_open_channels failed, -12"
due to dma_zalloc_coherent insufficient contiguous coherent memory to
satisfy the driver's request when the user tries to setup more or larger
rings.
Signed-off-by: Tariq Toukan <tariqt@mellanox.com>
Reported-by: Sebastian Ott <sebott@linux.vnet.ibm.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
The user can now override the automatic driver decision using the
rx_cqe_compress flag, which is the preference for CQE compression.
The flag is initialized with the automatic driver decision.
Signed-off-by: Shaker Daibes <shakerd@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
pflags is a configuration parameter for the netdev, naturally it belongs
to priv->params.
Also introduce MLX5E_GET_PFLAG
Signed-off-by: Shaker Daibes <shakerd@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Extend the self diagnostic tests to support loopback test.
The loopback test doesn't require the offline flag, it will use the
generic dev_queue_xmit and a dedicated packet_type to capture and verify
mlx5e selftest loopback packets.
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Signed-off-by: Kamal Heib <kamalh@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
The self diagnostics test implementaion include the following features:
1. Link Test: Check that link is in up state.
2. Speed Test: Check that link was negotiated correctly.
3. Health Test: Check the device health.
Signed-off-by: Kamal Heib <kamalh@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
DBCX by default is controlled by firmware where dcbx capability bit
is set. In this mode, firmware is responsible for reading/sending the
TLV packets from/to the remote partner.
This patch sets up the infrastructure to move between HOST/FW DCBX
control mode.
Signed-off-by: Huy Nguyen <huyn@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Issue description:
Current implementation saves the ETS settings from user in
a temporal soft copy and returns this settings when user
queries the ETS settings.
With the new DCBX firmware, the ETS settings can be changed
by firmware when the DCBX is in firmware controlled mode. Therefore,
user will obtain wrong values from the temporal soft copy.
Solution:
1. Read the ETS settings directly from firmware.
2. For tc_tsa:
a. Initialize tc_tsa to vendor IEEE_8021QAZ_TSA_VENDOR at netdev
creation.
b. When reading ETS setting from FW, if the traffic class bandwidth
is less than 100, set tc_tsa to IEEE_8021QAZ_TSA_ETS. This
implementation solves the scenarios when the DCBX is in FW control
and willing bit is on which means the ETS setting is dictated
by remote switch.
Also check ETS capability where needed.
Signed-off-by: Huy Nguyen <huyn@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Add DCBX CEE API interface for ConnectX-4. Configurations are stored in
a temporary structure and are applied to the card's firmware when
the CEE's setall callback function is called.
Note:
priority group in CEE is equivalent to traffic class in ConnectX-4
hardware spec.
bw allocation per priority in CEE is not supported because ConnectX-4
only supports bw allocation per traffic class.
user priority in CEE does not have an equivalent term in ConnectX-4.
Therefore, user priority to priority mapping in CEE is not supported.
Signed-off-by: Huy Nguyen <huyn@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Also move the inline capablities enum to a shared header vport.h
Signed-off-by: Roi Dayan <roid@mellanox.com>
Reviewed-by: Or Gerlitz <ogerlitz@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Switchdev driver net-device port statistics should follow the model introduced
in commit a5ea31f57309 'Merge branch net-offloaded-stats'.
For VF reps we return the SRIOV eswitch vport stats as the usual ones and SW stats
if asked. For the PF, if we're in the switchdev mode, we return the uplink stats
and SW stats if asked, otherwise as before. The uplink stats are implemented using
the PPCNT 802_3 counters which are already being read/cached by the driver.
Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
By implementing this ndo, the host stack will set the vxlan udp port
also to VF representor netdevices. This will allow the TC offload code
in the driver when it gets a tunnel key set action to identify the UDP
port as vxlan, and hence the rule will be a candidate for offloading.
Signed-off-by: Hadar Hen Zion <hadarh@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|\ \
| |/
|/|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
git://git.kernel.org/pub/scm/linux/kernel/git/leon/linux-rdma
Saeed Mahameed says:
====================
Mellanox mlx5 core driver updates 2016-10-25
This series contains some updates and fixes of mlx5 core and
IB drivers with the addition of two features that demand
new low level commands and infrastructure updates.
- SRIOV VF max rate limit support
- mlx5e tc support for FWD rules with counter.
Needed for both net and rdma subsystems.
Updates and Fixes:
From Saeed Mahameed (2):
- mlx5 IB: Skip handling unknown mlx5 events
- Add ConnectX-5 PCIe 4.0 VF device ID
From Artemy Kovalyov (2):
- Update struct mlx5_ifc_xrqc_bits
- Ensure SRQ physical address structure endianness
From Eugenia Emantayev (1):
- Fix length of async_event_mask
New Features:
From Mohamad Haj Yahia (3): mlx5 SRIOV VF max rate limit support
- Introduce TSAR manipulation firmware commands
- Introduce E-switch QoS management
- Add SRIOV VF max rate configuration support
From Mark Bloch (7): mlx5e Tc support for FWD rule with counter
- Don't unlock fte while still using it
- Use fte status to decide on firmware command
- Refactor find_flow_rule
- Group similar rules under the same fte
- Add multi dest support
- Add option to add fwd rule with counter
- mlx5e tc support for FWD rule with counter
Mark here fixed two trivial issues with the flow steering core, and did
some refactoring in the flow steering API to support adding mulit destination
rules to the same hardware flow table entry at once. In the last two patches
added the ability to populate a flow rule with a flow counter to the same flow entry.
V2: Dropped some patches that added new structures without adding any usage of them.
Added SRIOV VF max rate configuration support patch that introduces
the usage of the TSAR infrastructure.
Added flow steering fixes and refactoring in addition to mlx5 tc
support for forward rule with counter.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Currently when calling mlx5_add_flow_rule we accept
only one flow destination, this commit allows to pass
multiple destinations.
This change forces us to change the return structure to a more
flexible one. We introduce a flow handle (struct mlx5_flow_handle),
it holds internally the number for rules created and holds an array
where each cell points the to a flow rule.
From the consumers (of mlx5_add_flow_rule) point of view this
change is only cosmetic and requires only to change the type
of the returned value they store.
From the core point of view, we now need to use a loop when
allocating and deleting rules (e.g given to us a flow handler).
Signed-off-by: Mark Bloch <markb@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
|
|/
|
|
|
|
|
|
|
|
| |
Instead of predicting the index of the wanted LRO timeout value from
hardware capabilities, look for the nearest LRO timeout value.
Fixes: 5c50368f3831 ('net/mlx5e: Light-weight netdev open/stop')
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Signed-off-by: Mohamad Haj Yahia <mohamad@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Many virtualization systems use a policy under which a vlan tag is
pushed to packets sent by guests, and popped before the packet is
forwarded to the VM.
The current generation of the mlx5 HW doesn't fully support that on
a per flow level. As such, we are addressing the above common use
case with the SRIOV e-Switch abilities to push vlan into packets
sent by VFs and pop vlan from packets forwarded to VFs.
The HW can match on the correct vlan being present in packets
forwarded to VFs (eSwitch steering is done before stripping
the tag), so this part is offloaded as is.
A common practice for vlans is to avoid both push vlan and pop vlan
for inter-host VM/VM (east-west) communication because in this case,
push on egress cancels out with pop on ingress.
For supporting that, we use a global eswitch vlan pop policy, hence
allowing guest A to communicate with both remote VM B and local VM C.
This works since the HW pops the vlan only if it exists (e.g for
C --> A packets but not for B --> A packets).
On the slow path, when a VF vport has an offloaded flow which involves
pushing vlans, wheres another flow is not currently offloaded, the
packets from the 2nd flow seen by the VF representor on the host have
vlan. The VF rep driver removes such vlan before calling into the host
networking stack.
Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Previously we rang XDP SQ doorbell on every forwarded XDP packet.
Here we introduce a xmit more like mechanism that will queue up more
than one packet into SQ (up to RX napi budget) w/o notifying the hardware.
Once RX napi budget is consumed and we exit napi RX loop, we will
flush (doorbell) all XDP looped packets in case there are such.
XDP forward packet rate:
Comparing XDP with and w/o xmit more (bulk transmit):
RX Cores XDP TX XDP TX (xmit more)
---------------------------------------------------
1 6.5Mpps 12.4Mpps
2 13.2Mpps 24.2Mpps
4 25.2Mpps 36.3Mpps*
8 36.3Mpps* 36.3Mpps*
*My xmitter was limited to 36.3Mpps, so it is the bottleneck.
It seems that receive side can handle more.
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Signed-off-by: Tariq Toukan <tariqt@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Adding support for XDP_TX forwarding from xdp program.
Using XDP, now user can loop packets out of the same port.
We create a dedicated TX SQ for each channel that will serve
XDP programs that return XDP_TX action to loop packets back to
the wire directly from the channel RQ RX path.
For that RX pages will now need to be mapped bi-directionally,
and on XDP_TX action we will sync the page back to device then
queue it into SQ for transmission. The XDP xmit frame function will
report back to the RX path if the page was consumed (transmitted), if so,
RX path will forget about that page as if it were released to the stack.
Later on, on XDP TX completion, the page will be released back to the
page cache.
For simplicity this patch will hit a doorbell on every XDP TX packet.
Next patch will introduce a xmit more like mechanism that will
queue up more than one packet into SQ w/o notifying the hardware,
once RX napi loop is done we will hit doorbell once for all XDP TX
packets form the previous loop. This should drastically improve
XDP TX performance.
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Signed-off-by: Tariq Toukan <tariqt@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Make a clear separate between Regular SQ (TXQ) and ICO SQ creation,
destruction and union their mutual information structures.
Don't allocate redundant TXQ skb/wqe_info/dma_fifo arrays for ICO SQ.
And have a different SQ edge for ICO SQ than TXQ SQ, to be more
accurate.
In preparation for XDP TX support.
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Signed-off-by: Tariq Toukan <tariqt@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Add support for the BPF_PROG_TYPE_PHYS_DEV hook in mlx5e driver.
When XDP is on we make sure to change channels RQs type to
MLX5_WQ_TYPE_LINKED_LIST rather than "striding RQ" type to
ensure "page per packet".
On XDP set, we fail if HW LRO is set and request from user to turn it
off. Since on ConnectX4-LX HW LRO is always on by default, this will be
annoying, but we prefer not to enforce LRO off from XDP set function.
Full channels reset (close/open) is required only when setting XDP
on/off.
When XDP set is called just to exchange programs, we will update
each RQ xdp program on the fly and for synchronization with current
data path RX activity of that RQ, we temporally disable that RQ and
ensure RX path is not running, quickly update and re-enable that RQ,
for that we do:
- rq.state = disabled
- napi_synnchronize
- xchg(rq->xdp_prg)
- rq.state = enabled
- napi_schedule // Just in case we've missed an IRQ
Packet rate performance testing was done with pktgen 64B packets and on
TX side and, TC drop action on RX side compared to XDP fast drop.
CPU: Intel(R) Xeon(R) CPU E5-2680 v3 @ 2.50GHz
Comparison is done between:
1. Baseline, Before this patch with TC drop action
2. This patch with TC drop action
3. This patch with XDP RX fast drop
RX Cores Baseline(TC drop) TC drop XDP fast Drop
--------------------------------------------------------------
1 5.3Mpps 5.3Mpps 16.5Mpps
2 10.2Mpps 10.2Mpps 31.3Mpps
4 20.5Mpps 19.9Mpps 36.3Mpps*
*My xmitter was limited to 36.3Mpps, so it is the bottleneck.
It seems that receive side can handle more.
Signed-off-by: Rana Shahout <ranas@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Signed-off-by: Tariq Toukan <tariqt@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
We have two types of RX RQs, and they use two separate sets of
info arrays and structures in RX data path function. Today those
structures are mutually exclusive per RQ type, hence one kind is
allocated on RQ creation according to the RQ type.
For better cache locality and to minimalize the
sizeof(struct mlx5e_rq), in this patch we define them as a union.
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Signed-off-by: Tariq Toukan <tariqt@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
For non-striding RQ configuration before this patch we had a ring
with pre-allocated SKBs and mapped the SKB->data buffers for
device.
For robustness and better RX data buffers management, we allocate a
page per packet and build_skb around it.
This patch (which is a prerequisite for XDP) will actually reduce
performance for normal stack usage, because we are now hitting a bottleneck
in the page allocator. We use the page-cache to restore or even improve
performance in comparison to the old RX scheme.
Packet rate performance testing was done with pktgen 64B packets on xmit
side and TC ingress dropping action on RX side.
CPU: Intel(R) Xeon(R) CPU E5-2680 v3 @ 2.50GHz
Comparison is done between:
1.Baseline, before 'net/mlx5e: Build RX SKB on demand'
2.Build SKB with RX page cache (This patch)
RX Cores Baseline Build SKB+page-cache Improvement
-----------------------------------------------------------
1 4.16Mpps 5.33Mpps 28%
2 7.16Mpps 10.24Mpps 43%
4 13.61Mpps 20.51Mpps 51%
8 25.32Mpps 32.00Mpps 26%
All respective cores were 100% utilized.
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Signed-off-by: Tariq Toukan <tariqt@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Instead of reallocating and mapping pages for RX data-path,
recycle already used pages in a per ring cache.
Performance tests:
The following results were measured on a freshly booted system,
giving optimal baseline performance, as high-order pages are yet to
be fragmented and depleted.
We ran pktgen single-stream benchmarks, with iptables-raw-drop:
Single stride, 64 bytes:
* 4,739,057 - baseline
* 4,749,550 - order0 no cache
* 4,786,899 - order0 with cache
1% gain
Larger packets, no page cross, 1024 bytes:
* 3,982,361 - baseline
* 3,845,682 - order0 no cache
* 4,127,852 - order0 with cache
3.7% gain
Larger packets, every 3rd packet crosses a page, 1500 bytes:
* 3,731,189 - baseline
* 3,579,414 - order0 no cache
* 3,931,708 - order0 with cache
5.4% gain
Signed-off-by: Tariq Toukan <tariqt@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
To improve the memory consumption scheme, we omit the flow that
demands and splits high-order pages in Striding RQ, and stay
with a single Striding RQ flow that uses order-0 pages.
Moving to fragmented memory allows the use of larger MPWQEs,
which reduces the number of UMR posts and filler CQEs.
Moving to a single flow allows several optimizations that improve
performance, especially in production servers where we would
anyway fallback to order-0 allocations:
- inline functions that were called via function pointers.
- improve the UMR post process.
This patch alone is expected to give a slight performance reduction.
However, the new memory scheme gives the possibility to use a page-cache
of a fair size, that doesn't inflate the memory footprint, which will
dramatically fix the reduction and even give a performance gain.
Performance tests:
The following results were measured on a freshly booted system,
giving optimal baseline performance, as high-order pages are yet to
be fragmented and depleted.
We ran pktgen single-stream benchmarks, with iptables-raw-drop:
Single stride, 64 bytes:
* 4,739,057 - baseline
* 4,749,550 - this patch
no reduction
Larger packets, no page cross, 1024 bytes:
* 3,982,361 - baseline
* 3,845,682 - this patch
3.5% reduction
Larger packets, every 3rd packet crosses a page, 1500 bytes:
* 3,731,189 - baseline
* 3,579,414 - this patch
4% reduction
Fixes: 461017cb006a ("net/mlx5e: Support RX multi-packet WQE (Striding RQ)")
Fixes: bc77b240b3c5 ("net/mlx5e: Add fragmented memory support for RX multi packet WQE")
Signed-off-by: Tariq Toukan <tariqt@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Needed to support seamless and lightweight PCI/Internal error recovery.
Implement the attach/detach interface callbacks.
In attach callback we only allocate HW resources.
In detach callback we only deallocate HW resources.
All SW/kernel objects initialzing/destroying is kept in add/remove
callbacks.
Signed-off-by: Mohamad Haj Yahia <mohamad@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|\
| |
| |
| |
| |
| |
| | |
All three conflicts were cases of simple overlapping
changes.
Signed-off-by: David S. Miller <davem@davemloft.net>
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Instead of asking the firmware to flush the SQ (Send Queue) via
asynchronous completions when moved to error, we handle SQ flush
manually (mlx5e_free_tx_descs) same as we did when SQ flush got
timed out or on tx_timeout.
This will reduce SQs flush time and speedup interface down procedure.
Moved mlx5e_free_tx_descs to the end of en_tx.c for tx
critical code locality.
Fixes: 29429f3300a3 ('net/mlx5e: Timeout if SQ doesn't flush during close')
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
This will significantly reduce receive queue flush time on interface
down.
Instead of asking the firmware to flush the RQ (Receive Queue) via
asynchronous completions when moved to error, we handle RQ flush
manually (mlx5e_free_rx_descs) same as we did when RQ flush got timed
out.
This will reduce RQs flush time and speedup interface down procedure
(ifconfig down) from 6 sec to 0.3 sec on a 48 cores system.
Moved mlx5e_free_rx_descs en_main.c where it is needed, to keep en_rx.c
free form non critical data path code for better code locality.
Fixes: 6cd392a082de ('net/mlx5e: Handle RQ flush in error cases')
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
ConnectX-4 UMR (User Memory Region) MTT translation table offset in WQE
is limited to U16_MAX, before this patch we ignored that limitation and
requested the maximum possible UMR translation length that the netdev
might need (MAX channels * MAX pages per channel).
In case of a system with #cores > 32 and when linear WQE allocation fails,
falling back to using UMR WQEs will cause the RQ (Receive Queue) to get
stuck.
Here we limit UMR length to min(U16_MAX, max required pages) (while
considering the required alignments) on driver load, by default U16_MAX is
sufficient since the default RX rings value guarantees that we are in
range, dynamically (on set_ringparam/set_channels) we will check if the
new required UMR length (num mtts) is still in range, if not, fail the
request.
Fixes: bc77b240b3c5 ('net/mlx5e: Add fragmented memory support for RX multi packet WQE')
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|/
|
|
|
|
|
|
|
|
| |
The mlx5e_link_mode enumeration will also be used in mlx5_ib for RoCE.
This patch moves the enumeration to the mlx5 driver port header file.
Signed-off-by: Noa Osherovich <noaos@mellanox.com>
Signed-off-by: Eran Ben Elisha <eranbe@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Add support for query the minimum inline mode from the Firmware.
It is required for correct TX steering according to L3/L4 packet
headers.
Each send queue (SQ) has inline mode that defines the minimal required
headers that needs to be copied into the SQ WQE.
The driver asks the Firmware for the wqe_inline_mode device capability
value. In case the device capability defined as "vport context" the
driver must check the reported min inline mode from the vport context
before creating its SQs.
Signed-off-by: Hadar Hen Zion <hadarh@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
|
|
|
|
|
|
|
|
|
| |
Each send queue (SQ) has inline mode that defines the minimal required
inline headers in the SQ WQE.
Before sending each packet check that the minimum required headers
on the WQE are copied.
Signed-off-by: Hadar Hen Zion <hadarh@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|\
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Conflicts:
drivers/net/ethernet/mellanox/mlx5/core/en.h
drivers/net/ethernet/mellanox/mlx5/core/en_main.c
drivers/net/usb/r8152.c
All three conflicts were overlapping changes.
Signed-off-by: David S. Miller <davem@davemloft.net>
|
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Valid weight assigned to ETS TClass values are 1-100
Fixes: 08fb1dacdd76 ('net/mlx5e: Support DCBNL IEEE ETS')
Signed-off-by: Rana Shahout <ranas@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|