When the enic was disabled, link notification was correctly disabled
in the NIC but the software indicator that it was disabled was not
updated (vdev->notify_pa not set to 0). When the link came back up,
enic did not re-enable notification in the NIC.
This affected bonding when a enic slave device link bounced.
The fix is to unconditionally enable notification when the enic is
enabled.
Fixes: 9913fbb91df0 ("enic/base: common code")
Signed-off-by: John Daley <johndale@cisco.com>
Reviewed-by: Nelson Escobar <neescoba@cisco.com>
If the nb_pkts parameter to rte_eth_tx_burst() was greater than
the TX descriptor count, a completion was not being requested
from the NIC, so descriptors would not be released back to the
host causing a lock-up.
Introduce a limit of how many TX descriptors can be used in a single
call to the enic PMD burst TX function before requesting a completion.
Fixes: d739ba4c6abf ("enic: improve Tx packet rate")
Signed-off-by: John Daley <johndale@cisco.com>
FreeBSD was not defined in ena_plat.h
ETIME is not defined in FreeBSD.
In file included from DPDK/drivers/net/ena/base/ena_com.h:37:0,
from DPDK/drivers/net/ena/ena_ethdev.h:39,
from DPDK/drivers/net/ena/ena_ethdev.c:41:
DPDK/drivers/net/ena/base/ena_plat.h:48:2: error: #error "Invalid platform"
Fixes: 99ecfbf845b3 ("ena: import communication layer")
Fixes: 9ba7981ec992 ("ena: add communication layer for DPDK")
Signed-off-by: Daniel Mrzyglod <danielx.t.mrzyglod@intel.com>
Acked-by: Bruce Richardson <bruce.richardson@intel.com>
Fix for multiple compilation errors for ICC:
error #188: enumerated type mixed with another type
error #592: variable "flags" is used before its value is set
Fixes: 99ecfbf845b3 ("ena: import communication layer")
Signed-off-by: Daniel Mrzyglod <danielx.t.mrzyglod@intel.com>
Build log:
/root/dpdk/app/test-pmd/cmdline.c:6687:45: error: no member named
's6_addr32' in 'struct in6_addr'
rte_be_to_cpu_32(res->ip_value.addr.ipv6.s6_addr32[i]);
This is caused by macro "s6_addr32" not defined on FreeBSD and testpmd
swap big endian parameter to host endian. Move the swap action to i40e
ethdev will fix this issue.
Fixes: 7b1312891b69 ("ethdev: add IP in GRE tunnel")
Signed-off-by: Marvin Liu <yong.liu@intel.com>
Acked-by: Jingjing Wu <jingjing.wu@intel.com>
Tested-by: Bruce Richardson <bruce.richardson@intel.com>
The code uses "entry" in the next loop of LIST_FOREACH after calling free()
on it in i40e_res_pool_destroy().
Change to a safe way to free entry, which is similar with LIST_FOREACH_SAFE
in FreeBSD.
Fixes: 4861cde46116 ("i40e: new poll mode driver")
Signed-off-by: Jiangu Zhao <zhaojg@arraynetworks.com.cn>
Acked-by: Bruce Richardson <bruce.richardson@intel.com>
This solves issues when an active device is added to a bond.
If a device to be enslaved already has transmit and/or receive queues
allocated, use those and then create any additional queues that are
necessary.
Fixes: 2efb58cbab6e ("bond: new link bonding library")
Signed-off-by: Eric Kinzie <ehkinzie@gmail.com>
Acked-by: Bernard Iremonger <bernard.iremonger@intel.com>
On the 82575 chipset, there is a pool of global TX contexts instead of 2
per queues on 82576. See Table A-1 "Changes in Programming Interface
Relative to 82575" of Intel® 82576EB GbE Controller datasheet (*).
In the driver, the contexts are attributed to a TX queue: 0-1 for txq0,
2-3 for txq1, and so on.
In igbe_set_xmit_ctx(), the variable ctx_curr contains the index of the
per-queue context (0 or 1), and ctx_idx contains the index to be given
to the hardware (0 to 7). The size of txq->ctx_cache[] is 2, and must
be indexed with ctx_curr to avoid an out-of-bound access.
Also, the index returned by what_advctx_update() is the per-queue
index (0 or 1), so we need to add txq->ctx_start before sending it
to the hardware.
(*) The datasheets says 16 global contexts, however the IDX fields in TX
descriptors are 3 bits, which gives a total of 8 contexts. The
driver assumes there are 8 contexts on 82575: 2 per queues, 4 txqs.
Fixes: 4c8db5f09a24 ("igb: enable TSO support")
Fixes: af75078fece3 ("first public release")
Signed-off-by: Olivier Matz <olivier.matz@6wind.com>
Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com>
When using RSS, the number of rxqs has to be a power of two.
This is a problem because there is no API in DPDK that makes
the application aware of that.
A good compromise is to allow the application to request a
number of rxqs that is not a power of 2, but having inactive
queues that will never receive packets. In this configuration,
a warning will be issued to users to let them know that
this is not an optimal configuration.
Signed-off-by: Olivier Matz <olivier.matz@6wind.com>
Acked-by: Adrien Mazarguil <adrien.mazarguil@6wind.com>
l2 tunnel and e-tag are not supported on the new x550em_a NICs, due
to missing checks for that mac type in the code.
This patch adds in the necessary conditional checks to enable the features
for x550em_a.
Fixes: 22e77d4501b8 ("ixgbe: support L2 tunnel operations")
Signed-off-by: Wenzhuo Lu <wenzhuo.lu@intel.com>
An issue is found on x550em NICs: ieee1588 is not working, the time is
always reported as 0.
The root cause is that the timer is only supported by the driver for x550,
switch statement entries are missing for x550em_x and x550em_a. This patch
adds those missing entries.
Fixes: a7740dc1303a ("ixgbe: support new devices and MAC types")
Signed-off-by: Wenzhuo Lu <wenzhuo.lu@intel.com>
The current_primary_port is initialised to an invalid value
during bonded device creation.
It must be set to a valid value later.
This fix sets it to a valid value when the first slave port
is added to the bonding device.
Fixes: 2efb58cbab6e ("bond: new link bonding library")
Signed-off-by: Bernard Iremonger <bernard.iremonger@intel.com>
Acked-by: Bruce Richardson <bruce.richardson@intel.com>
icc (icc (ICC) 16.0.1 20151021) is generating following compile error:
CC ixgbe_rxtx.o
.../drivers/net/ixgbe/ixgbe_rxtx.c(153): error #3656: variable
"free" may be used before its value is set
(nb_free > 0 && m->pool != free[0]->pool)) {
^
Indeed this is a false positive and code is correct.
"nb_free" check prevents the free[] access before its value set.
Disabling this icc warning (#3656) for file ixgbe_rxtx.c.
Signed-off-by: Ferruh Yigit <ferruh.yigit@intel.com>
Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com>
Ixgbe HW supports 128 TX queues. However, the full 128 queues are only
available in VT and DCB mode. In normal default "none" mode (VT/DCB off)
the maximum number of available queues is only 64.
The driver doesn't check the mode when reporting the available
number of queues, allowing more that 64 queues to be used in all cases.
If a queue no. >=64 is used in default mode, the TX packets will be dropped
silently.
This change adds a check to forbid using a queue number larger than 64
during device configuration (in default mode), so that the problem is
reported as early as possible.
Fixes: 27b609cbd1c6 ("ethdev: move the multi-queue mode check to specific drivers")
Signed-off-by: Wenzhuo Lu <wenzhuo.lu@intel.com>
Signed-off-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>
Acked-by: John McNamara <john.mcnamara@intel.com>
Internal variable containing the number of TX queues for a device,
was being incorrectly assigned the number of RX queues, instead of TX.
Fixes: 27b609cbd1c6 ("ethdev: move the multi-queue mode check to specific drivers")
Signed-off-by: Wenzhuo Lu <wenzhuo.lu@intel.com>
Acked-by: John McNamara <john.mcnamara@intel.com>
In the function set_rx_mode, the pointer of device data points
to the wrong address as found in ixgbe code, and fixed in commit:
"ixgbe: fix PF promiscuous mode after VF closed"
Fixes: be2d648a2dd3 ("igb: add PF support")
Signed-off-by: Wenzhuo Lu <wenzhuo.lu@intel.com>
Acked-by: Jingjing Wu <jingjing.wu@intel.com>
There's an issue reported. In the scenario DPDK PF + DPDK VF,
if the VF port is closed, PF port cannot receive packets.
I found at that time the promicuous mode is disabled on the PF
port. But it should be enabled.
When VF port is closed, it will send a message to its PF port to
reset it. During this, PF port will also reset its own
promicuous mode. Which promiscuous mode should be set depends on
the parameter stored in the device data. In the function
set_rx_mode, the pointer of device data points to the wrong
address. So, the promiscuous mode is wrong.
Fixes: 00e30184daa0 ("ixgbe: add PF support")
Signed-off-by: Wenzhuo Lu <wenzhuo.lu@intel.com>
Reported-by: Bernard Iremonger <bernard.iremonger@intel.com>
Acked-by: Jingjing Wu <jingjing.wu@intel.com>
Current vector RX can't always set the packet_type properly.
To be more specific:
a) it never sets RTE_PTYPE_L2_ETHER
b) it doesn't handle tunnel ipv4/ipv6 case correctly.
c) it doesn't check is IXGBE_RXDADV_PKTTYPE_ETQF set or not.
While a) is pretty easy to fix, b) and c) are not that straightforward
in terms of SIMD ops (specially b).
So far I wasn't able to make vRX support packet_type properly without
noticeable performance loss.
So for now, just remove that functionality from vector RX and
update dev_supported_ptypes_get().
Fixes: 396254175854 ("mbuf: redefine packet type")
Signed-off-by: Konstantin Ananyev <konstantin.ananyev@intel.com>
Acked-by: Cunming Liang <cunming.liang@intel.com>
Notify user otherwise. A similar check has already been added to mlx5 in
commit "mlx5: check port is configured as ethernet device".
Signed-off-by: Adrien Mazarguil <adrien.mazarguil@6wind.com>
Currently, the maximum value of rx/tx queues are kept by EAL. But,
the value is used like below with different meanings in vhost PMD.
- The maximum value of current enabled queues.
- The maximum value of current supported queues.
This wrong double meaning will cause an issue like below steps.
* Invoke application with below option.
--vdev 'eth_vhost0,iface=<socket path>,queues=4'
* Configure queues like below.
rte_eth_dev_configure(portid, 2, 2, ...);
* Configure queues again like below.
rte_eth_dev_configure(portid, 4, 4, ...);
The second rte_eth_dev_configure() will fail because both
the maximum value of current enabled queues and supported queues
will be '2' after calling first rte_eth_dev_configure().
To fix the issue, the patch adds another variable to keep the maximum
number of supported queues in vhost PMD.
Fixes: 23981fb0d78b ("vhost: Add vhost PMD")
Signed-off-by: Tetsuya Mukawa <mukawa@igel.co.jp>
Acked-by: Ciara Loftus <ciara.loftus@intel.com>
Issue:
When CONFIG_RTE_LIBTRE_I40E_RX_ALLOW_BULK_ALLOC=n in config file, there
will be a build error:
'i40e_recv_pkts_bulk_alloc' undeclared
Now DPDK i40e PMD uses the preprocessor to choose whether or not to define
the bulk recv functions, but for selection of the RX function, PMD only
depends on a C variable. This causes the inconsistency and leads to the
build error due to the bulk recv function not being defined.
Fixes: 8e109464c022 ("i40e: allow vector Rx and Tx usage")
Signed-off-by: Zhe Tao <zhe.tao@intel.com>
Acked-by: Bruce Richardson <bruce.richardson@intel.com>
This patch extends flow director to select vlan id as part of
filter's input set and program the filter rule with vlan id.
Signed-off-by: Jingjing Wu <jingjing.wu@intel.com>
Acked-by: Helin Zhang <helin.zhang@intel.com>
This patch adds missing VLAN bitmask for inner frame in case of
tunneling and fixes VLAN tags bitmasks for single or outer frame
in case of tunneling.
Fixes: 98f055707685 ("i40e: configure input fields for RSS or flow director")
Signed-off-by: Andrey Chilikin <andrey.chilikin@intel.com>
Signed-off-by: Jingjing Wu <jingjing.wu@intel.com>
Acked-by: Helin Zhang <helin.zhang@intel.com>
This patch extends flow director to select more IP Header fields
as filter input set.
Signed-off-by: Jingjing Wu <jingjing.wu@intel.com>
Acked-by: Helin Zhang <helin.zhang@intel.com>
This patch adds a new function to set the fdir input set to default
when initialization.
Signed-off-by: Jingjing Wu <jingjing.wu@intel.com>
Acked-by: Helin Zhang <helin.zhang@intel.com>
In this patch, flex payload is removed from valid fdir input set
values. This is because all flex payload configuration can be set
in struct rte_fdir_conf during device configure phase, which is
a more flexible way of setting this up.
Signed-off-by: Jingjing Wu <jingjing.wu@intel.com>
Acked-by: Helin Zhang <helin.zhang@intel.com>
For the input set selection, Hash filter and Flow director shared
the same function, i.e. i40e_filter_inset_select.
For code readability, this patch replaces i40e_filter_inset_select
with two new functions: i40e_hash_filter_inset_select and
i40e_fdir_filter_inset_select for Hash filter and Flow director
respectively.
Signed-off-by: Jingjing Wu <jingjing.wu@intel.com>
Acked-by: Helin Zhang <helin.zhang@intel.com>
Virtio has an mbuf descriptor ring containing mbufs to be used for
receiving traffic. When the host queues traffic to be sent to the guest, it
consumes these descriptors. If none exist, it discards the packet.
The virtio pmd allocates mbufs to the descriptor ring every time it
successfully receives a packet. However, it never does it if it does not
receive a valid packet. If the descriptor ring is exhausted, and the mbuf
mempool does not have any mbufs free (which can happen for various reasons,
such as queueing along the processing pipeline), then the receive call will
not allocate any mbufs to the descriptor ring, and when it finishes, the
descriptor ring will be empty. The ring being empty means that we will
never receive a packet again, which means we will never allocate mbufs to
the ring: we are stuck.
Ultimately, the problem arises because there is a dependency between
receiving packets and making the descriptor ring not be empty, and a
dependency between the descriptor ring not being empty, and receiving
packets.
To fix the problem, this pakes makes virtio always try to allocate mbufs
to the descriptor ring, if necessary, when polling for packets. Do this by
removing the early exit if no packets were received. Since the packet loop
later will do nothing if there are no packets, this is fine.
I reproduced the problem by pushing packets through a pipelined systems
(such as the client_server sample application) after artificially
decreasing the size of the mbuf pool and introducing a delay in a secondary
stage.
Without the fix, the process stops receiving packets fairly quicky. With
the fix, it continues to receive packets.
Fixes: c1f86306a026 ("virtio: add new driver")
Signed-off-by: Kyle Larose <klarose@sandvine.com>
Acked-by: Huawei Xie <huawei.xie@intel.com>
This structure has immutable function pointers.
Also fix indentation.
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
Acked-by: Bruce Richardson <bruce.richardson@intel.com>
On initialization, the rq descriptor count was set to the limit
of the vic. When the requested number of rx descriptors was
less than this count, enic_alloc_rq() was incorrectly setting
the count to the lower value. This results in later calls to
enic_alloc_rq() incorrectly using the lower value as the adapter
limit.
Fixes: fefed3d1e62c ("enic: new driver")
Signed-off-by: Nelson Escobar <neescoba@cisco.com>
Reviewed-by: John Daley <johndale@cisco.com>
Update function can be called with no key to enable or disable a RSS
protocol, or with a key to be applied to the desired protocols.
Fixes: 2f97422e7759 ("mlx5: support RSS hash update and get")
Signed-off-by: Nelio Laranjeiro <nelio.laranjeiro@6wind.com>
Acked-by: Adrien Mazarguil <adrien.mazarguil@6wind.com>
RSS configuration provided by the application should not be used as storage
by the PMD.
Fixes: 2f97422e7759 ("mlx5: support RSS hash update and get")
Signed-off-by: Nelio Laranjeiro <nelio.laranjeiro@6wind.com>
Acked-by: Adrien Mazarguil <adrien.mazarguil@6wind.com>
For x550 device, the reta table has 512 entries, but in function
ixgbe_dev_rss_reta_query and ixgbe_dev_rss_reta_update we use an
"uint8_t i" to traverse the entries, this will lead the function
to an endless loop.
This patch changes the data type from uint8_t to uint16_t to fix
the issue.
Fixes: 4bee94a6c22f ("ixgbe: support 512 RSS entries on x550")
Signed-off-by: Wang Xiao W <xiao.w.wang@intel.com>
Acked-by: Wenzhuo Lu <wenzhuo.lu@intel.com>
If the packet_error bit in the completion descriptor is set, the
remainder of the descriptor and data are invalid. PKT_RX_MAC_ERR
was set in the mbuf->ol_flags if packet_error was set and used
later to indicate an error packet. But since PKT_RX_MAC_ERR is
defined as 0, mbuf flags and packet types and length were being
misinterpreted.
Make the function enic_cq_rx_to_pkt_err_flags() return true for error
packets and use the return value instead of mbuf->ol_flags to indicate
error packets. Also remove warning for error packets and rely on
rx_error stats.
Fixes: 947d860c821f ("enic: improve Rx performance")
Signed-off-by: John Daley <johndale@cisco.com>
In the receive path, the function to set mbuf ol_flags used the
mbuf packet_type before it was set.
Fixes: 947d860c821f ("enic: improve Rx performance")
Signed-off-by: John Daley <johndale@cisco.com>
Add checks to make sure we don't try to allocate more tx or rx queues
than we support.
Fixes: fefed3d1e62c ("enic: new driver")
Signed-off-by: Nelson Escobar <neescoba@cisco.com>
Reviewed-by: John Daley <johndale@cisco.com>
Add the missing '\n' character to the end of a few print statements.
Fixes: fefed3d1e62c ("enic: new driver")
Signed-off-by: Nelson Escobar <neescoba@cisco.com>
Acked-by: John Daley <johndale@cisco.com>
VLAN insertion can be done in hardware when supported in Verbs. A software
fallback is provided otherwise. The software implementation is also used
when multi-packet send is enabled on a queue, as both features are mutually
exclusive.
Signed-off-by: Yaacov Hazan <yaacovh@mellanox.com>
Signed-off-by: Adrien Mazarguil <adrien.mazarguil@6wind.com>
Environment variable MLX5_PMD_ENABLE_PADDING enables HW packet padding
in PCI bus transactions.
When packet size is cache aligned and CRC stripping is enabled, 4 fewer
bytes are written to the PCI bus. Enabling padding makes such packets
aligned again.
In cases where PCI bandwidth is the bottleneck, padding can improve
performance by 10%.
This is disabled by default since this can also decrease performance for
unaligned packet sizes.
Signed-off-by: Olga Shern <olgas@mellanox.com>
fix packet padding macro check
Signed-off-by: Adrien Mazarguil <adrien.mazarguil@6wind.com>
Secondary processes are expected to use queues and other resources
allocated by the primary, however Verbs resources can only be shared
between processes when inherited through fork().
This limitation can be worked around for TX by configuring separate queues
from secondary processes.
Signed-off-by: Or Ami <ora@mellanox.com>
Add driver functions to set link state up or down.
Burst functions are updated to make sure applications cannot attempt to
send/receive after link is brought down.
Signed-off-by: Or Ami <ora@mellanox.com>
When Linux PF and DPDK VF are used for i40e PMD, when a PF reset occurs,
an interrupt will go via adminq event to inform the VF of the reset.
A callback mechanism is introduced for the VF to allow it to invoke a
registered callback when PF reset happens.
Users can register a callback for this interrupt event using:
rte_eth_dev_callback_register(portid,
RTE_ETH_EVENT_INTR_RESET,
reset_event_callback,
arg);
Signed-off-by: Jingjing Wu <jingjing.wu@intel.com>
Acked-by: Helin Zhang <helin.zhang@intel.com>
Currently, i40evf PMD uses a global static buffer to send virtchnl
commands to host driver. It is shared by multiple VFs.
This patch changed to allocate a virtchnl cmd buffer for each VF.
Signed-off-by: Jingjing Wu <jingjing.wu@intel.com>
Acked-by: Helin Zhang <helin.zhang@intel.com>
The patch introduces a new PMD. This PMD is implemented as thin wrapper
of librte_vhost. It means librte_vhost is also needed to compile the PMD.
The vhost messages will be handled only when a port is started. So start
a port first, then invoke QEMU.
The PMD has 2 parameters.
- iface: The parameter is used to specify a path to connect to a
virtio-net device.
- queues: The parameter is used to specify the number of the queues
virtio-net device has.
(Default: 1)
Here is an example.
$ ./testpmd -c f -n 4 --vdev 'eth_vhost0,iface=/tmp/sock0,queues=1' -- -i
To connect above testpmd, here is qemu command example.
$ qemu-system-x86_64 \
<snip>
-chardev socket,id=chr0,path=/tmp/sock0 \
-netdev vhost-user,id=net0,chardev=chr0,vhostforce,queues=1 \
-device virtio-net-pci,netdev=net0,mq=on
Signed-off-by: Tetsuya Mukawa <mukawa@igel.co.jp>
Acked-by: Ferruh Yigit <ferruh.yigit@intel.com>
Acked-by: Yuanhan Liu <yuanhan.liu@linux.intel.com>
Acked-by: Rich Lane <rich.lane@bigswitch.com>
Tested-by: Rich Lane <rich.lane@bigswitch.com>
Update for queue state event name:
Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
This is a PMD for the Amazon ethernet ENA (Elastic Network Adapters)
family.
The driver operates variety of ENA adapters through feature negotiation
with the adapter and upgradable commands set.
ENA driver handles PCI Physical and Virtual ENA functions.
Signed-off-by: Evgeny Schemeilin <evgenys@amazon.com>
Signed-off-by: Jan Medala <jan@semihalf.com>
Signed-off-by: Jakub Palider <jpa@semihalf.com>
Release Note addition:
Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
Implementation of platform specific code for ENA communication layer.
Signed-off-by: Evgeny Schemeilin <evgenys@amazon.com>
Signed-off-by: Jan Medala <jan@semihalf.com>
Signed-off-by: Jakub Palider <jpa@semihalf.com>
Low level common abstraction for ENA device communication.
Signed-off-by: Netanel Belgazal <netanel@amazon.com>
Signed-off-by: Jan Medala <jan@semihalf.com>
Signed-off-by: Jakub Palider <jpa@semihalf.com>
Add a new API rte_eth_dev_get_supported_ptypes to query what packet types
can be filled by a given device. The device should be already started or
its PMD RX burst function already decided, since the packet types supported
may vary depending on RX function.
Signed-off-by: Jianfeng Tan <jianfeng.tan@intel.com>
Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com>
Acked-by: Adrien Mazarguil <adrien.mazarguil@6wind.com>
