In case of QINT interrupt occurrence, SW fails to clear the QINT
line resulting in recursive interrupts because currently interrupt
handler gets the cause of the interrupt by reading
NIX_LF_RQ[SQ/CQ/AURA/POOL]_OP_INT but does not write 1 to clear
RQ[SQ/CQ/ERR]_INT field in respective NIX_LF_RQ[SQ/CQ/AURA/POOL]_OP_INT
registers.
Fixes: dc47ba15f6 ("net/octeontx2: handle queue specific error interrupts")
Fixes: 50b95c3ea7 ("mempool/octeontx2: add NPA IRQ handler")
Signed-off-by: Harman Kalra <hkalra@marvell.com>
Acked-by: Jerin Jacob <jerinj@marvell.com>
The packets transmitting in mlx5 is performed by building
Tx descriptors (WQEs) and sending last ones to the NIC.
The descriptor can contain the special flags, telling the NIC
to generate Tx completion notification (CQEs). At the beginning
of tx_burst() routine PMD checks whether there are some Tx
completions and frees the transmitted packet buffers.
The flags to request completion generation must be set once
per specified amount of packets to provide uniform stream
of completions and freeing the Tx queue in uniform fashion.
The previous implementation sets the completion request
generation once per burst, if burst size if big enough it may
latency in CQE generation and freeing large amount of buffers
in tx_burst routine on multiple completions which also
affects the latency and even causes the Tx queue overflow
and Tx drops.
This patches enforces the completion request will be set
in the exact Tx descriptor if specified amount of packets
is already sent.
Fixes: 18a1c20044 ("net/mlx5: implement Tx burst template")
Signed-off-by: Viacheslav Ovsiienko <viacheslavo@mellanox.com>
Acked-by: Matan Azrad <matan@mellanox.com>
Mellanox ConnectX-4LX NIC in configurations with disabled
E-Switch can operate without minimal required inline data
into Tx descriptor. There was the hardcoded limit set to
18B in PMD, fixed to be no limit (0B).
Fixes: 38b4b397a5 ("net/mlx5: add Tx configuration and setup")
Signed-off-by: Viacheslav Ovsiienko <viacheslavo@mellanox.com>
Acked-by: Matan Azrad <matan@mellanox.com>
This patch limits the amount of fetched and processed
completion descriptors in one tx_burst routine call.
The completion processing involves the buffer freeing
which may be time consuming and introduce the significant
latency, so limiting the amount of processed completions
mitigates the latency issue.
Fixes: 18a1c20044 ("net/mlx5: implement Tx burst template")
Signed-off-by: Viacheslav Ovsiienko <viacheslavo@mellanox.com>
Acked-by: Matan Azrad <matan@mellanox.com>
Enabling LRO offload per queue makes sense because the user will
probably want to allocate different mempool for LRO queues - the LRO
mempool mbuf size may be bigger than non LRO mempool.
Change the LRO offload to be per queue instead of per port.
If one of the queues is with LRO enabled, all the queues will be
configured via DevX.
If RSS flows direct TCP packets to queues with different LRO enabling,
these flows will not be offloaded with LRO.
Signed-off-by: Matan Azrad <matan@mellanox.com>
Acked-by: Viacheslav Ovsiienko <viacheslavo@mellanox.com>
When a user configures LRO in the port offloads, he probably wants each
TCP packet will have a chance to open an LRO session.
The PMD wasn't configure LRO in the flow TIR if the flow is not
explicitly configured TCP item despite the flow included TCP traffic.
For example, the next flows were not LRO offloaded:
pattern eth / end, pattern eth / ip / end, pattern eth / ipv6 / end.
Enable LRO configuration for all the TIRs if LRO is configured in the
port.
No performance impact for non-LRO traffic in these TIRs.
Signed-off-by: Matan Azrad <matan@mellanox.com>
Acked-by: Viacheslav Ovsiienko <viacheslavo@mellanox.com>
When LRO offload is configured in Rx queue, the HW may coalesce TCP
packets from same TCP connection into single packet.
In this case the SW should fix the relevant packet headers because
the HW doesn't update them according to the new created packet
characteristics but provides the update values in the CQE.
Add update header code to the regular Rx burst function to support LRO
feature.
Make sure the first mbuf has enough space to include each TCP header,
otherwise the header update may cross mbufs what complicates the
operation too match.
Signed-off-by: Matan Azrad <matan@mellanox.com>
Acked-by: Viacheslav Ovsiienko <viacheslavo@mellanox.com>
The alignment requested by the FW for WQ buffer allocation is 512.
Change it from cache line alignment to 512.
Fixes: dc9ceff73c ("net/mlx5: create advanced RxQ via DevX")
Signed-off-by: Matan Azrad <matan@mellanox.com>
Acked-by: Viacheslav Ovsiienko <viacheslavo@mellanox.com>
LRO support was only for MPRQ, hence mprq Rx burst was selected when
LRO was configured in the port.
The current support for MPRQ is suffering from bad memory utilization
since an external mempool is allocated by the PMD for the packets data
in addition to the user mempool, besides that, the user may get packet
data addresses which were not configured by him.
Even though MPRQ has the best performance for packet receiving in the
most cases and because of the above facts it is better to remove the
automatic MPRQ select when LRO is configured.
Move MPRQ to be selected only when the user force it by the PMD
arguments including LRO case.
Allow LRO offload using the regular RQ with the regular Rx burst
function.
Signed-off-by: Matan Azrad <matan@mellanox.com>
Acked-by: Viacheslav Ovsiienko <viacheslavo@mellanox.com>
When the Rx queue is not in striding RQ mode it should be configured as
cyclic RQ.
In this case the type remains 0 which means linked-list type.
Set the RQ type to be cyclic when the queue is not in striding RQ mode.
Fixes: dc9ceff73c ("net/mlx5: create advanced RxQ via DevX")
Signed-off-by: Matan Azrad <matan@mellanox.com>
Acked-by: Viacheslav Ovsiienko <viacheslavo@mellanox.com>
The WQ size configuration via DevX didn't take into account the maximum
number of segments per packet what wrongly caused to configure bigger
WQE size than the size expected by the PMD in other places.
The scatter mode stride size should be the size of segment multiplied
by the number of maximum segments per packet.
The number of WQEs per WQ should be the number of descriptors divided by
the number of the maximum segments per packet.
Fix the size calculations to the above rule.
Fixes: dc9ceff73c ("net/mlx5: create advanced RxQ via DevX")
Signed-off-by: Matan Azrad <matan@mellanox.com>
Acked-by: Viacheslav Ovsiienko <viacheslavo@mellanox.com>
Patch [1] zeroes the mbuf headroom when the port is configured with LRO
because when working with more than one stride per packet the HW cannot
guaranty an headroom in the start stride of each packet.
Change the solution to support mbuf headroom by adding an empty buffer
as the first packet segment, scatter mode must be enabled to support it.
[1] http://patches.dpdk.org/patch/56912/
Signed-off-by: Matan Azrad <matan@mellanox.com>
Acked-by: Viacheslav Ovsiienko <viacheslavo@mellanox.com>
When mbuf is allocated by rte_pktmbuf_alloc the offload flag is reset by
it, so data-path function should not do it again.
Remove the above offload flag reset from MPRQ data-path.
Signed-off-by: Matan Azrad <matan@mellanox.com>
Acked-by: Viacheslav Ovsiienko <viacheslavo@mellanox.com>
The field max_rx_pkt_len in Rx configuration indicates the maximum size
for Rx packet to be received.
There was no any field to indicate the maximum size of LRO packet to be
received by the application.
Assuming the user configures max_rx_pkt_len as the maximum LRO packet
length when LRO is configured on the port, the PMD limits the maximum
LRO packet size received from HW to be max_rx_pkt_len.
Signed-off-by: Matan Azrad <matan@mellanox.com>
Acked-by: Viacheslav Ovsiienko <viacheslavo@mellanox.com>
If the mbuf size of the Rx mempool supplied by the user in the Rx setup
is unable to contain the maximum Rx packet length in addition to the
mbuf head-room, the Rx scatter offload must be configured. Otherwise,
there is not enough space in single mbuf to contain a packet with size
of the maximum Rx packet length.
The PMD did not return an error in the above mentioned case.
Return an error in the above case.
Fixes: 7d6bf6b866 ("net/mlx5: add Multi-Packet Rx support")
Fixes: edad38fcd0 ("net/mlx: enhance Rx scatter mode detection")
Cc: stable@dpdk.org
Signed-off-by: Matan Azrad <matan@mellanox.com>
Acked-by: Viacheslav Ovsiienko <viacheslavo@mellanox.com>
The patch fix the issue that LLDP packet can't be forwarded to host.
Fixes: 59d151de66 ("net/ice: stop LLDP by default")
Cc: stable@dpdk.org
Signed-off-by: Ying A Wang <ying.a.wang@intel.com>
Acked-by: Qi Zhang <qi.z.zhang@intel.com>
Fix license from BSD-2-Clause to BSD-3-Clause since the file was
lost on the driver license change.
Bump copyright year.
Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>
Acked-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Prior to the fix, RTE_LOGTYPE_INFO messages were displayed by default.
After the fix, only NOTICE level and higher were displayed by default
and INFO level were not. There are INFO level vNIC config related
messages which customers and tech support currently depend on for
debugging and so on and to suddenly hide these messages is not a good
idea.
This patch changes the default log level to RTE_LOG_INFO for enic so
messages are printed as before the fix.
Fixes: bbd8ecc054 ("net/enic: remove PMD log type references")
Signed-off-by: John Daley <johndale@cisco.com>
Adding support to parse GRE KEY for octeontx2 Flow.
Matching on GRE Key will only work, if checksum and routing
bits in the GRE header are equal to 0.
Signed-off-by: Kiran Kumar K <kirankumark@marvell.com>
Acked-by: Jerin Jacob <jerinj@marvell.com>
This patch implements read clock api whose purpose is to return
raw clock ticks. Using this API real time ticks spent in
processing a packet can be known:
<read_clock val at any time> - mbuf->timestamp
Calling mbox for reading raw clock ticks in fastpath is very
expensive so its value is derived from time stamp counter(tsc)
using freq multiplier (ratio of raw clock ticks and tsc) and clock
delta (by how much tsc is lagging from raw clock value).
Signed-off-by: Harman Kalra <hkalra@marvell.com>
Acked-by: Jerin Jacob <jerinj@marvell.com>
A huge drop in per core MPPS value was observed when PTP stack is
enabled. The reason behind the bottleneck is HW serialises the
transfer of all SQEs, which seeks timestamp capture, on the same
send DMA path. Hence only those packets which requires timestamp
capture should set SETTSTAMP in send mem alg.
With this patch timestamping would be done only for those packets
with PKT_TX_IEEE1588_TMST set.
Fixes: fb3ae0951a ("net/octeontx2: support Tx")
Fixes: 8980a15300 ("event/octeontx2: support PTP for SSO")
Signed-off-by: Harman Kalra <hkalra@marvell.com>
Acked-by: Jerin Jacob <jerinj@marvell.com>
Earlier implementation for enabling ptp via RX offload flag was
causing segmentation fault as it was getting executed in the
device configuration stage where RX and TX queues were not
configured. As in the ptp enable process rx queues are used for
mbuf setup while tx queues are used for send descriptor setup.
Moving the logic in dev start as all the resources will be
configured.
Fixes: b5dc314044 ("net/octeontx2: support base PTP")
Signed-off-by: Harman Kalra <hkalra@marvell.com>
Acked-by: Jerin Jacob <jerinj@marvell.com>
Multi segmented packet may be spliced with indirect mbufs also.
Currently driver causes buffer leak for indirect mbufs as they
were not being freed to packet pool.
Patch fixes handling of indirect mbufs for following use cases
- packet contains all indirect mbufs only.
- packet contains mixed mbufs i.e. direct and indirect both.
Fixes: cbd5710db4 ("net/octeontx2: add Tx multi segment version")
Signed-off-by: Sunil Kumar Kori <skori@marvell.com>
Acked-by: Nithin Dabilpuram <ndabilpuram@marvell.com>
missed_pkts reflects the number of packets that the driver did not manage
to send.
This is a temporary situation, those packets are not freed and the
application can still retry to send them later.
Hence, we can't count them as transmit failed.
Fixes: 5f05e95cd5 ("net/vhost: fix Tx error counting")
Fixes: ee584e9710 ("vhost: add driver on top of the library")
Cc: stable@dpdk.org
Signed-off-by: David Marchand <david.marchand@redhat.com>
Reviewed-by: Tiwei Bie <tiwei.bie@intel.com>
err_pkts reflects the number of packets that the driver did not manage
to send.
This is a temporary situation, those packets are not freed and the
application can still retry to send them later.
Hence, we can't count them as transmit failed.
Fixes: e1e4017751 ("ring: add new driver")
Cc: stable@dpdk.org
Signed-off-by: David Marchand <david.marchand@redhat.com>
Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
n_err reflects the number of packets that the driver did not manage to
send.
This is a temporary situation, those packets are not freed and the
application can still retry to send them later.
Hence, we can't count them as transmit failed.
Fixes: 09c7e63a71 ("net/memif: introduce memory interface PMD")
Signed-off-by: David Marchand <david.marchand@redhat.com>
Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
err_pkts reflects the number of packets that the driver did not manage to
send.
This is a temporary situation, those packets are not freed and the
application can still retry to send them later.
Hence, we can't count them as transmit failed.
Fixes: 75e2bc54c0 ("net/kni: add KNI PMD")
Cc: stable@dpdk.org
Signed-off-by: David Marchand <david.marchand@redhat.com>
Acked-by: Ferruh Yigit <ferruh.yigit@intel.com>
The delta between what the application asked to receive and what was
indeed received, can not be called an error counter.
This counter is not reported anywhere, remove it.
Fixes: 75e2bc54c0 ("net/kni: add KNI PMD")
Cc: stable@dpdk.org
Signed-off-by: David Marchand <david.marchand@redhat.com>
Acked-by: Ferruh Yigit <ferruh.yigit@intel.com>
This Tx counter has never been used.
Fixes: 9658d17da2 ("virtio: maintain stats per queue")
Cc: stable@dpdk.org
Signed-off-by: David Marchand <david.marchand@redhat.com>
Reviewed-by: Tiwei Bie <tiwei.bie@intel.com>
This Tx counter has never been used.
Fixes: c743e50c47 ("null: new poll mode driver")
Cc: stable@dpdk.org
Signed-off-by: David Marchand <david.marchand@redhat.com>
Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
This Tx counter is now unused.
Fixes: 10edf857fd ("net/af_xdp: make reserve/submit peek/release consistent")
Cc: stable@dpdk.org
Signed-off-by: David Marchand <david.marchand@redhat.com>
Reviewed-by: Xiaolong Ye <xiaolong.ye@intel.com>
This Rx counter has never been used.
Fixes: 364e08f2bb ("af_packet: add PMD for AF_PACKET-based virtual devices")
Cc: stable@dpdk.org
Signed-off-by: David Marchand <david.marchand@redhat.com>
Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
Filling correct RETA table size at ixgbevf_dev_info_get,
so RETA table update will be supported for VF port.
For X540_vf and 82599_vf, since they don't support
RETA table update, set RETA size to 0.
Fixes: 2144f6630f ("ixgbe: add redirection table size in device info")
Cc: stable@dpdk.org
Signed-off-by: Wei Zhao <wei.zhao1@intel.com>
Acked-by: Qi Zhang <qi.z.zhang@intel.com>
This commit fixes an issue with the error checking in flow
MARK action. Previously, (ANY + MARK) would fail, as the
(mark_spec == 0) condition would cause an early error return,
however really it is (mark_spec != 0) that should cause the
early error return.
Flipping the binary comparison corrects the behaviour, and
(ANY + MARK) now succeeds, while (MARK + MARK) fails.
Fixes: 0bbcfc706a ("net/i40e: support MARK and RSS flow action")
Suggested-by: Harry van Haaren <harry.van.haaren@intel.com>
Signed-off-by: Mesut Ali Ergin <mesut.a.ergin@intel.com>
Acked-by: Qi Zhang <qi.z.zhang@intel.com>
Change loop index from uint16_t to uint32_t since max
index 65535 could be exceeded when ring size is 2k+.
Fixes: 69dd4c3d08 ("net/avf: enable queue and device")
Cc: stable@dpdk.org
Reported-by: Lei Yao <lei.a.yao@intel.com>
Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
Reviewed-by: Xiaolong Ye <xiaolong.ye@intel.com>
Two cores can send multi segment packets on two different pcap ports.
Because of this, we can't have one single buffer to linearize packets.
Use rte_pktmbuf_read() to copy the packet into a buffer on the stack
and remove eth_pcap_gather_data() when necessary (if the mbuf is
contiguous, rte_pktmbuf_read() just points at the buffer address).
Fixes: 6db141c91e ("pcap: support jumbo frames")
Cc: stable@dpdk.org
Signed-off-by: David Marchand <david.marchand@redhat.com>
Acked-by: Ferruh Yigit <ferruh.yigit@intel.com>
When a packet cannot be transmitted, the driver is supposed to free this
packet and report it as handled.
This is to prevent the application from retrying to send the same packet
and ending up in a liveloop since the driver will never manage to send
it.
Fixes: 49a0a2ffd5 ("net/pcap: fix possible mbuf double freeing")
Fixes: 6db141c91e ("pcap: support jumbo frames")
Cc: stable@dpdk.org
Signed-off-by: David Marchand <david.marchand@redhat.com>
Acked-by: Ferruh Yigit <ferruh.yigit@intel.com>
If the pkt pool contains only buffers smaller than the default headroom,
then the driver will compute an invalid buffer size (negative value cast
to an uint16_t).
Rely on the mbuf api to check how much space is available in the mbuf.
Fixes: 6eb0ae218a ("pcap: fix mbuf allocation")
Cc: stable@dpdk.org
Signed-off-by: David Marchand <david.marchand@redhat.com>
Acked-by: Ferruh Yigit <ferruh.yigit@intel.com>
Change verbosity of a message to DEBUG from ERROR.
This is just debug message.
Fixes: 3e92fd4e4e ("net/bnxt: use dynamic log type")
Cc: stable@dpdk.org
Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com>
Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com>
