| Age | Commit message (Collapse) | Author | Files | Lines |
|---|
|
Make mlx4 compatible with the newly added netlink queue GET APIs.
Signed-off-by: Joe Damato <jdamato@fastly.com>
Tested-by: Martin Karsten <mkarsten@uwaterloo.ca>
Acked-by: Jakub Kicinski <kuba@kernel.org>
Reviewed-by: Tariq Toukan <tariqt@nvidia.com>
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Link: https://lore.kernel.org/r/20240528181139.515070-3-jdamato@fastly.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
mlx4_en_alloc_frags currently returns -ENOMEM when mlx4_alloc_page
fails but does not increment a stat field when this occurs.
A new field called alloc_fail has been added to struct mlx4_en_rx_ring
which is now incremented in mlx4_en_rx_ring when -ENOMEM occurs.
Signed-off-by: Joe Damato <jdamato@fastly.com>
Tested-by: Martin Karsten <mkarsten@uwaterloo.ca>
Reviewed-by: Tariq Toukan <tariqt@nvidia.com>
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Link: https://lore.kernel.org/r/20240528181139.515070-2-jdamato@fastly.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
Use a notifier to implement mlx4_dispatch_event() in preparation to
switch mlx4_en and mlx4_ib to be an auxiliary device.
A problem is that if the mlx4_interface.event callback was replaced with
something as mlx4_adrv.event then the implementation of
mlx4_dispatch_event() would need to acquire a lock on a given device
before executing this callback. That is necessary because otherwise
there is no guarantee that the associated driver cannot get unbound when
the callback is running. However, taking this lock is not possible
because mlx4_dispatch_event() can be invoked from the hardirq context.
Using an atomic notifier allows the driver to accurately record when it
wants to receive these events and solves this problem.
A handler registration is done by both mlx4_en and mlx4_ib at the end of
their mlx4_interface.add callback. This matches the current situation
when mlx4_add_device() would enable events for a given device
immediately after this callback, by adding the device on the
mlx4_priv.list.
Signed-off-by: Petr Pavlu <petr.pavlu@suse.com>
Tested-by: Leon Romanovsky <leonro@nvidia.com>
Acked-by: Tariq Toukan <tariqt@nvidia.com>
Reviewed-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Rename the mlx4_en_dev.nb notifier_block member to netdev_nb in
preparation to add a mlx4 core notifier_block.
Signed-off-by: Petr Pavlu <petr.pavlu@suse.com>
Tested-by: Leon Romanovsky <leonro@nvidia.com>
Reviewed-by: Leon Romanovsky <leonro@nvidia.com>
Acked-by: Tariq Toukan <tariqt@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Conflicts:
tools/testing/selftests/net/config
62199e3f1658 ("selftests: net: Add VXLAN MDB test")
3a0385be133e ("selftests: add the missing CONFIG_IP_SCTP in net config")
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
The RSS hash type specifies what portion of packet data NIC hardware used
when calculating RSS hash value. The RSS types are focused on Internet
traffic protocols at OSI layers L3 and L4. L2 (e.g. ARP) often get hash
value zero and no RSS type. For L3 focused on IPv4 vs. IPv6, and L4
primarily TCP vs UDP, but some hardware supports SCTP.
Hardware RSS types are differently encoded for each hardware NIC. Most
hardware represent RSS hash type as a number. Determining L3 vs L4 often
requires a mapping table as there often isn't a pattern or sorting
according to ISO layer.
The patch introduce a XDP RSS hash type (enum xdp_rss_hash_type) that
contains both BITs for the L3/L4 types, and combinations to be used by
drivers for their mapping tables. The enum xdp_rss_type_bits get exposed
to BPF via BTF, and it is up to the BPF-programmer to match using these
defines.
This proposal change the kfunc API bpf_xdp_metadata_rx_hash() adding
a pointer value argument for provide the RSS hash type.
Change signature for all xmo_rx_hash calls in drivers to make it compile.
The RSS type implementations for each driver comes as separate patches.
Fixes: 3d76a4d3d4e5 ("bpf: XDP metadata RX kfuncs")
Signed-off-by: Jesper Dangaard Brouer <brouer@redhat.com>
Acked-by: Toke Høiland-Jørgensen <toke@redhat.com>
Acked-by: Stanislav Fomichev <sdf@google.com>
Link: https://lore.kernel.org/r/168132892042.340624.582563003880565460.stgit@firesoul
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
|
|
Zero-length arrays as fake flexible arrays are deprecated and we are
moving towards adopting C99 flexible-array members instead.
Transform zero-length array into flexible-array member in struct
mlx4_en_rx_desc.
Address the following warnings found with GCC-13 and
-fstrict-flex-arrays=3 enabled:
drivers/net/ethernet/mellanox/mlx4/en_rx.c:88:30: warning: array subscript i is outside array bounds of ‘struct mlx4_wqe_data_seg[0]’ [-Warray-bounds=]
drivers/net/ethernet/mellanox/mlx4/en_rx.c:149:30: warning: array subscript 0 is outside array bounds of ‘struct mlx4_wqe_data_seg[0]’ [-Warray-bounds=]
drivers/net/ethernet/mellanox/mlx4/en_rx.c:127:30: warning: array subscript i is outside array bounds of ‘struct mlx4_wqe_data_seg[0]’ [-Warray-bounds=]
drivers/net/ethernet/mellanox/mlx4/en_rx.c:128:30: warning: array subscript i is outside array bounds of ‘struct mlx4_wqe_data_seg[0]’ [-Warray-bounds=]
drivers/net/ethernet/mellanox/mlx4/en_rx.c:129:30: warning: array subscript i is outside array bounds of ‘struct mlx4_wqe_data_seg[0]’ [-Warray-bounds=]
drivers/net/ethernet/mellanox/mlx4/en_rx.c:117:30: warning: array subscript i is outside array bounds of ‘struct mlx4_wqe_data_seg[0]’ [-Warray-bounds=]
drivers/net/ethernet/mellanox/mlx4/en_rx.c:119:30: warning: array subscript i is outside array bounds of ‘struct mlx4_wqe_data_seg[0]’ [-Warray-bounds=]
This helps with the ongoing efforts to tighten the FORTIFY_SOURCE
routines on memcpy() and help us make progress towards globally
enabling -fstrict-flex-arrays=3 [1].
Link: https://github.com/KSPP/linux/issues/21
Link: https://github.com/KSPP/linux/issues/264
Link: https://gcc.gnu.org/pipermail/gcc-patches/2022-October/602902.html [1]
Signed-off-by: Gustavo A. R. Silva <gustavoars@kernel.org>
Reviewed-by: Tariq Toukan <tariqt@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
RX timestamp and hash for now. Tested using the prog from the next
patch.
Also enabling xdp metadata support; don't see why it's disabled,
there is enough headroom..
Cc: John Fastabend <john.fastabend@gmail.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Martin KaFai Lau <martin.lau@linux.dev>
Cc: Jakub Kicinski <kuba@kernel.org>
Cc: Willem de Bruijn <willemb@google.com>
Cc: Jesper Dangaard Brouer <brouer@redhat.com>
Cc: Anatoly Burakov <anatoly.burakov@intel.com>
Cc: Alexander Lobakin <alexandr.lobakin@intel.com>
Cc: Magnus Karlsson <magnus.karlsson@gmail.com>
Cc: Maryam Tahhan <mtahhan@redhat.com>
Cc: xdp-hints@xdp-project.net
Cc: netdev@vger.kernel.org
Reviewed-by: Tariq Toukan <tariqt@nvidia.com>
Signed-off-by: Stanislav Fomichev <sdf@google.com>
Link: https://lore.kernel.org/r/20230119221536.3349901-14-sdf@google.com
Signed-off-by: Martin KaFai Lau <martin.lau@kernel.org>
|
|
Google production kernel has increased MAX_SKB_FRAGS to 45
for BIG-TCP rollout.
Unfortunately mlx4 TX bounce buffer is not big enough whenever
an skb has up to 45 page fragments.
This can happen often with TCP TX zero copy, as one frag usually
holds 4096 bytes of payload (order-0 page).
Tested:
Kernel built with MAX_SKB_FRAGS=45
ip link set dev eth0 gso_max_size 185000
netperf -t TCP_SENDFILE
I made sure that "ethtool -G eth0 tx 64" was properly working,
ring->full_size being set to 15.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reported-by: Wei Wang <weiwan@google.com>
Reviewed-by: Tariq Toukan <tariqt@nvidia.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
MAX_DESC_SIZE is really the size of the bounce buffer used
when reaching the right side of TX ring buffer.
MAX_DESC_TXBBS get a MLX4_ prefix.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Tariq Toukan <tariqt@nvidia.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
This patch caches doorbell address directly in struct mlx4_en_tx_ring.
This removes the need to bring in cpu caches whole struct mlx4_uar
in fast path.
Note that mlx4_uar is not guaranteed to be on a local node,
because mlx4_bf_alloc() uses a single free list (priv->bf_list)
regardless of its node parameter.
This kind of change does matter in presence of light/moderate traffic.
In high stress, this read-only line would be kept hot in caches.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Tariq Toukan <tariqt@nvidia.com>
Reviewed-by: Tariq Toukan <tariqt@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Add counters for XDP REDIRECT success and failure. This brings the
redirect path in line with metrics gathered via the other XDP paths.
Signed-off-by: Joshua Roys <roysjosh@gmail.com>
Reviewed-by: Tariq Toukan <tariqt@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Any link state change that's done prior to net device registration
isn't reflected on the state, thus the operational state is left
obsolete, with 'UNKNOWN' status.
To resolve the issue, query link state from FW upon open operations
to ensure operational state is updated.
Fixes: c27a02cd94d6 ("mlx4_en: Add driver for Mellanox ConnectX 10GbE NIC")
Signed-off-by: Lama Kayal <lkayal@nvidia.com>
Signed-off-by: Tariq Toukan <tariqt@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
This patch fixes a bug that the moderation config will not be
applied when calling mlx4_en_reset_config. For example, when
turning on rx timestamping, mlx4_en_reset_config() will be called,
causing the NIC to forget previous moderation config.
This fix is in phase with a previous fix:
commit 79c54b6bbf06 ("net/mlx4_en: Fix TX moderation info loss
after set_ringparam is called")
Tested: Before this patch, on a host with NIC using mlx4, run
netserver and stream TCP to the host at full utilization.
$ sar -I SUM 1
INTR intr/s
14:03:56 sum 48758.00
After rx hwtstamp is enabled:
$ sar -I SUM 1
14:10:38 sum 317771.00
We see the moderation is not working properly and issued 7x more
interrupts.
After the patch, and turned on rx hwtstamp, the rate of interrupts
is as expected:
$ sar -I SUM 1
14:52:11 sum 49332.00
Fixes: 79c54b6bbf06 ("net/mlx4_en: Fix TX moderation info loss after set_ringparam is called")
Signed-off-by: Kevin(Yudong) Yang <yyd@google.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Neal Cardwell <ncardwell@google.com>
CC: Tariq Toukan <tariqt@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
Pull irq updates from Thomas Gleixner:
"This is the second attempt after the first one failed miserably and
got zapped to unblock the rest of the interrupt related patches.
A treewide cleanup of interrupt descriptor (ab)use with all sorts of
racy accesses, inefficient and disfunctional code. The goal is to
remove the export of irq_to_desc() to prevent these things from
creeping up again"
* tag 'irq-core-2020-12-23' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: (30 commits)
genirq: Restrict export of irq_to_desc()
xen/events: Implement irq distribution
xen/events: Reduce irq_info:: Spurious_cnt storage size
xen/events: Only force affinity mask for percpu interrupts
xen/events: Use immediate affinity setting
xen/events: Remove disfunct affinity spreading
xen/events: Remove unused bind_evtchn_to_irq_lateeoi()
net/mlx5: Use effective interrupt affinity
net/mlx5: Replace irq_to_desc() abuse
net/mlx4: Use effective interrupt affinity
net/mlx4: Replace irq_to_desc() abuse
PCI: mobiveil: Use irq_data_get_irq_chip_data()
PCI: xilinx-nwl: Use irq_data_get_irq_chip_data()
NTB/msi: Use irq_has_action()
mfd: ab8500-debugfs: Remove the racy fiddling with irq_desc
pinctrl: nomadik: Use irq_has_action()
drm/i915/pmu: Replace open coded kstat_irqs() copy
drm/i915/lpe_audio: Remove pointless irq_to_desc() usage
s390/irq: Use irq_desc_kstat_cpu() in show_msi_interrupt()
parisc/irq: Use irq_desc_kstat_cpu() in show_interrupts()
...
|
|
No driver has any business with the internals of an interrupt
descriptor. Storing a pointer to it just to use yet another helper at the
actual usage site to retrieve the affinity mask is creative at best. Just
because C does not allow encapsulation does not mean that the kernel has no
limits.
Retrieve a pointer to the affinity mask itself and use that. It's still
using an interface which is usually not for random drivers, but definitely
less hideous than the previous hack.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Tariq Toukan <tariqt@nvidia.com>
Link: https://lore.kernel.org/r/20201210194044.580936243@linutronix.de
|
|
xdp_return_frame_bulk() needs to pass a xdp_buff
to __xdp_return().
strlcpy got converted to strscpy but here it makes no
functional difference, so just keep the right code.
Conflicts:
net/netfilter/nf_tables_api.c
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
In case error CQE was found while polling TX CQ, the QP is in error
state and all posted WQEs will generate error CQEs without any data
transmitted. Fix it by reopening the channels, via same method used for
TX timeout handling.
In addition add some more info on error CQE and WQE for debug.
Fixes: bd2f631d7c60 ("net/mlx4_en: Notify user when TX ring in error state")
Signed-off-by: Moshe Shemesh <moshe@mellanox.com>
Signed-off-by: Tariq Toukan <tariqt@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Add restarting state flag to avoid scheduling another restart task while
such task is already running. Change task name from watchdog_task to
restart_task to better fit the task role.
Fixes: 1e338db56e5a ("mlx4_en: Fix a race at restart task")
Signed-off-by: Moshe Shemesh <moshe@mellanox.com>
Signed-off-by: Tariq Toukan <tariqt@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
linux/netdevice.h is included in very many places, touching any
of its dependecies causes large incremental builds.
Drop the linux/ethtool.h include, linux/netdevice.h just needs
a forward declaration of struct ethtool_ops.
Fix all the places which made use of this implicit include.
Acked-by: Johannes Berg <johannes@sipsolutions.net>
Acked-by: Shannon Nelson <snelson@pensando.io>
Reviewed-by: Jesse Brandeburg <jesse.brandeburg@intel.com>
Link: https://lore.kernel.org/r/20201120225052.1427503-1-kuba@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
Performance analysis counters are maintained under the MLX4_EN_PERF_STAT
definition, which is never set.
Clean them up, with all related structures and logic.
Signed-off-by: Tariq Toukan <tariqt@nvidia.com>
Reviewed-by: Moshe Shemesh <moshe@nvidia.com>
Link: https://lore.kernel.org/r/20201118103427.4314-1-tariqt@nvidia.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
Convert to new infra, make use of the ability to sleep in the callback.
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Acked-by: Tariq Toukan <tariqt@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
In order to benefit from the new napi_defer_hard_irqs feature,
we need to use napi_complete_done() variant in this driver.
RX path is already using it, this patch implements TX completion side.
mlx4_en_process_tx_cq() now returns the amount of retired packets,
instead of a boolean, so that mlx4_en_poll_tx_cq() can pass
this value to napi_complete_done().
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
After the previous patch, all the callers of ndo_select_queue()
provide as a 'fallback' argument netdev_pick_tx.
The only exceptions are nested calls to ndo_select_queue(),
which pass down the 'fallback' available in the current scope
- still netdev_pick_tx.
We can drop such argument and replace fallback() invocation with
netdev_pick_tx(). This avoids an indirect call per xmit packet
in some scenarios (TCP syn, UDP unconnected, XDP generic, pktgen)
with device drivers implementing such ndo. It also clean the code
a bit.
Tested with ixgbe and CONFIG_FCOE=m
With pktgen using queue xmit:
threads vanilla patched
(kpps) (kpps)
1 2334 2428
2 4166 4278
4 7895 8100
v1 -> v2:
- rebased after helper's name change
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
NIC driver minimal MTU size shall be set to ETH_MIN_MTU, as defined in
the RFC791 and in the network stack. Remove old mlx4_en only define for
it, which was set to wrong value.
Fixes: b80f71f5816f ("ethernet/mellanox: use core min/max MTU checking")
Signed-off-by: Eran Ben Elisha <eranbe@mellanox.com>
Signed-off-by: Tariq Toukan <tariqt@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
When memory is limited (on kdump kernel), reduce size of rx and tx rings.
Also reduce the number of rx rings.
Signed-off-by: Alaa Hleihel <alaa@mellanox.com>
Signed-off-by: Tariq Toukan <tariqt@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
This patch makes it so that instead of passing a void pointer as the
accel_priv we instead pass a net_device pointer as sb_dev. Making this
change allows us to pass the subordinate device through to the fallback
function eventually so that we can keep the actual code in the
ndo_select_queue call as focused on possible on the exception cases.
Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
|
|
Add check of coalescing parameters received through ethtool are within
range of values supported by the HW.
Driver gets the coalescing rx/tx-usecs and rx/tx-frames as set by the
users through ethtool. The ethtool support up to 32 bit value for each.
However, mlx4 modify cq limits the coalescing time parameter and
coalescing frames parameters to 16 bits.
Return out of range error if user tries to set these parameters to
higher values.
Change type of sample-interval and adaptive_rx_coal parameters in mlx4
driver to u32 as the ethtool holds them as u32 and these parameters are
not limited due to mlx4 HW.
Fixes: c27a02cd94d6 ('mlx4_en: Add driver for Mellanox ConnectX 10GbE NIC')
Signed-off-by: Moshe Shemesh <moshe@mellanox.com>
Signed-off-by: Tariq Toukan <tariqt@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Add physical RX/TX packets/bytes counters into ethtool output to monitor
all traffic that was received and transmitted on the port. These
counters are available only for none Virtual Function.
Signed-off-by: Eran Ben Elisha <eranbe@mellanox.com>
Signed-off-by: Tariq Toukan <tariqt@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Driver hook points for xdp_rxq_info:
* reg : mlx4_en_create_rx_ring
* unreg: mlx4_en_destroy_rx_ring
Tested on actual hardware.
Cc: Tariq Toukan <tariqt@mellanox.com>
Signed-off-by: Jesper Dangaard Brouer <brouer@redhat.com>
Reviewed-by: Tariq Toukan <tariqt@mellanox.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
|
|
Change the default mapping between TC and TCG as follows:
Prio | TC/TCG
| from to
| (set by FW) (set by SW)
---------+-----------------------------------
0 | 0/0 0/7
1 | 1/0 0/6
2 | 2/0 0/5
3 | 3/0 0/4
4 | 4/0 0/3
5 | 5/0 0/2
6 | 6/0 0/1
7 | 7/0 0/0
These new settings cause that a pause frame for any prio stops
traffic for all prios.
Fixes: 564c274c3df0 ("net/mlx4_en: DCB QoS support")
Signed-off-by: Moni Shoua <monis@mellanox.com>
Signed-off-by: Maor Gottlieb <maorg@mellanox.com>
Signed-off-by: Tariq Toukan <tariqt@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Set the minimal MTU threshold for running loopback selftest.
MTU should be big enough to include packet payload, NET_IP_ALIGN,
Ethernet headers and preamble length.
Fixes: e7c1c2c46201 ("mlx4_en: Added self diagnostics test implementation")
Signed-off-by: Eugenia Emantayev <eugenia@mellanox.com>
Signed-off-by: Tariq Toukan <tariqt@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
In XDP_TX, some fields in tx_info and tx_desc are constants across
all entries of the different XDP_TX rings.
Assign values to these fields on ring creation time, rather than in
data-path.
Patchset performance tests:
Tested on ConnectX3Pro, Intel(R) Xeon(R) CPU E5-2680 v3 @ 2.50GHz
Single queue no-RSS optimization ON.
XDP_TX packet rate:
------------------------------
Before | After | Gain |
13.7 Mpps | 14.0 Mpps | %2.2 |
------------------------------
Signed-off-by: Tariq Toukan <tariqt@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
The struct net_device parameter was passed only to extract
struct mlx4_en_priv out of it.
Here we pass the priv parameter directly.
Signed-off-by: Tariq Toukan <tariqt@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Limit the number of TX rings per UP by the number of cores
in the system.
Signed-off-by: Inbar Karmy <inbark@mellanox.com>
Signed-off-by: Tariq Toukan <tariqt@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
The function mlx4_en_arm_cq always returns zero. So change the
return type of the function mlx4_en_arm_cq to void.
CC: Joe Jin <joe.jin@oracle.com>
CC: Junxiao Bi <junxiao.bi@oracle.com>
Signed-off-by: Zhu Yanjun <yanjun.zhu@oracle.com>
Reviewed-by: Leon Romanovsky <leonro@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Currently the number of TX queues that are allocated doesn't depend
on the number of TCs, the module always loads with max num of UP
per channel.
In order to prevent the allocation of unnecessary memory, the
module will load with minimum number of UPs per channel, and the
user will be able to control the number of TX queues per channel
by changing the number of TC to 8 using the tc command.
The variable num_up will hold the information about the current
number of UPs.
Due to the change, needed to remove the lines that set the value of
UP to be different than zero in the func "mlx4_en_select_queue",
since now the num of TX queues that are allocated is only one per channel
in default.
In order not to force the UP to be zero in case of only one TC, added
a condition before forcing it in the func "mlx4_en_fill_qp_context".
Tested:
After the module is loaded with minimum number of UP per channel, to
increase num of TCs to 8, use:
tc qdisc add dev ens8 root mqprio num_tc 8
In order to decrease the number of TCs to minimum number of UP per channel,
use:
tc qdisc del dev ens8 root
Signed-off-by: Inbar Karmy <inbark@mellanox.com>
Signed-off-by: Tariq Toukan <tariqt@mellanox.com>
Cc: Tarick Bedeir <tarick@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Until this patch, the number of UPs was hard coded for eight.
Replace this with a variable in struct "mlx4_en_port_profile".
Currently, the variable will hold the maximum number of UP,
as before.
The patch creates an infrastructure to add an option for dynamic
change of the actual number of TCs.
Signed-off-by: Inbar Karmy <inbark@mellanox.com>
Signed-off-by: Tariq Toukan <tariqt@mellanox.com>
Cc: Tarick Bedeir <tarick@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Define LOG_TXBB_SIZE, log of TXBB_SIZE, and use it with a shift
operation instead of a multiplication with TXBB_SIZE.
Operations are equivalent as TXBB_SIZE is a power of two.
Performance tests:
Tested on ConnectX3Pro, Intel(R) Xeon(R) CPU E5-2680 v3 @ 2.50GHz
Gain is too small to be measurable, no degradation sensed.
Results are similar for IPv4 and IPv6.
Signed-off-by: Tariq Toukan <tariqt@mellanox.com>
Reviewed-by: Saeed Mahameed <saeedm@mellanox.com>
Cc: kernel-team@fb.com
Cc: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Increase the default TX ring size (from 512 to 1024) to match
the RX ring size.
This gives the XDP TX ring a better chance to keep up with the
rate of its RX ring in case of a high load of XDP_TX actions.
Tested:
Ethtool counter rx_xdp_tx_full used to increase, after applying this
patch it stopped.
Signed-off-by: Tariq Toukan <tariqt@mellanox.com>
Reviewed-by: Saeed Mahameed <saeedm@mellanox.com>
Cc: kernel-team@fb.com
Cc: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Instead of having their own NAPIs, XDP TX completion queues get
polled within the corresponding RX NAPI.
This prevents any possible race on TX ring prod/cons indices,
between the context that issues the transmits (RX NAPI) and the
context that handles the completions (was previously done in
a separate NAPI).
This also improves performance, as it decreases the number
of NAPIs running on a CPU, saving the overhead of syncing
and switching between the contexts.
Performance tests:
Tested on ConnectX3Pro, Intel(R) Xeon(R) CPU E5-2680 v3 @ 2.50GHz
Single queue no-RSS optimization ON.
XDP_TX packet rate:
-------------------------------------
| Before | After | Gain |
IPv4 | 12.0 Mpps | 13.8 Mpps | 15% |
IPv6 | 12.0 Mpps | 13.8 Mpps | 15% |
-------------------------------------
Signed-off-by: Tariq Toukan <tariqt@mellanox.com>
Reviewed-by: Saeed Mahameed <saeedm@mellanox.com>
Cc: kernel-team@fb.com
Cc: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Several performance improvements in XDP TX datapath,
including:
- Ring a single doorbell for XDP TX ring per NAPI budget,
instead of doing it per a lower threshold (was 8).
This includes removing the flow of immediate doorbell ringing
in case of a full TX ring.
- Compiler branch predictor hints.
- Calculate values in compile time rather than in runtime.
Performance tests:
Tested on ConnectX3Pro, Intel(R) Xeon(R) CPU E5-2680 v3 @ 2.50GHz
Single queue no-RSS optimization ON.
XDP_TX packet rate:
-------------------------------------
| Before | After | Gain |
IPv4 | 10.3 Mpps | 12.0 Mpps | 17% |
IPv6 | 10.3 Mpps | 12.0 Mpps | 17% |
-------------------------------------
Signed-off-by: Tariq Toukan <tariqt@mellanox.com>
Reviewed-by: Saeed Mahameed <saeedm@mellanox.com>
Cc: kernel-team@fb.com
Cc: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Avoid touching RX QP RSS context when loading with only
one RX ring, to allow optimized A0 RX steering.
Enable by:
- loading mlx4_core with module param: log_num_mgm_entry_size = -6.
- then: ethtool -L <interface> rx 1
Performance tests:
Tested on ConnectX3Pro, Intel(R) Xeon(R) CPU E5-2680 v3 @ 2.50GHz
XDP_DROP packet rate:
-------------------------------------
| Before | After | Gain |
IPv4 | 20.5 Mpps | 28.1 Mpps | 37% |
IPv6 | 18.4 Mpps | 28.1 Mpps | 53% |
-------------------------------------
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Signed-off-by: Tariq Toukan <tariqt@mellanox.com>
Cc: kernel-team@fb.com
Cc: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Remove owner argument, as it is obsolete and unused.
This also saves the overhead of calculating its value in data-path.
Signed-off-by: Tariq Toukan <tariqt@mellanox.com>
Reviewed-by: Saeed Mahameed <saeedm@mellanox.com>
Cc: kernel-team@fb.com
Cc: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Remove date and bump version for mlx4_en driver.
Signed-off-by: Tariq Toukan <tariqt@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
This new counter tracks number of pages that we allocated for one port.
lpaa24:~# ethtool -S eth0 | egrep 'rx_alloc_pages|rx_packets'
rx_packets: 306755183
rx_alloc_pages: 932897
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Tariq Toukan <tariqt@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Same technique than some Intel drivers, for arches where PAGE_SIZE = 4096
In most cases, pages are reused because they were consumed
before we could loop around the RX ring.
This brings back performance, and is even better,
a single TCP flow reaches 30Gbit on my hosts.
v2: added full memset() in mlx4_en_free_frag(), as Tariq found it was needed
if we switch to large MTU, as priv->log_rx_info can dynamically be changed.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Tariq Toukan <tariqt@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Use of order-3 pages is problematic in some cases.
This patch might add three kinds of regression :
1) a CPU performance regression, but we will add later page
recycling and performance should be back.
2) TCP receiver could grow its receive window slightly slower,
because skb->len/skb->truesize ratio will decrease.
This is mostly ok, we prefer being conservative to not risk OOM,
and eventually tune TCP better in the future.
This is consistent with other drivers using 2048 per ethernet frame.
3) Because we allocate one page per RX slot, we consume more
memory for the ring buffers. XDP already had this constraint anyway.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Tariq Toukan <tariqt@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
We will soon use order-0 pages, and frag truesize will more precisely
match real sizes.
In the new model, we prefer to use <= 2048 bytes fragments, so that
we can use page-recycle technique on PAGE_SIZE=4096 arches.
We will still pack as much frames as possible on arches with big
pages, like PowerPC.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Tariq Toukan <tariqt@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
We only need to store the page and dma address.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Tariq Toukan <tariqt@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
