Jiawei Wang [Tue, 13 Oct 2020 14:11:46 +0000 (17:11 +0300)]
net/mlx5: split sample flow into two sub-flows
The flow with sample action will be split into two sub flows:
the prefix sub flow with the all actions preceding the sample
action and sample action itself, and the suffix sub flow with
the actions following the sample action.
The original items remain in the prefix sub flow, add the
implicit tag action with unique id to set in metadata register,
and suffix sub flow uses the tag item to match with that unique id.
The flow split as below:
Original flow: items / actions pre / sample / actions sfx ->
prefix sub flow -
items / actions pre / set_tag action / sample
suffix sub flow -
tag_item / actions sfx
Signed-off-by: Jiawei Wang <jiaweiw@nvidia.com> Acked-by: Viacheslav Ovsiienko <viacheslavo@nvidia.com>
Jiawei Wang [Tue, 13 Oct 2020 14:11:45 +0000 (17:11 +0300)]
net/mlx5: validate sample action
Add sample action validate function.
Sample Flow is supported in NIC-RX and FDB domains. For the NIC-RX
the Sample Flow action list must include the destination queue action.
Only NIC-RX domain supports the optional actions list. FDB doesn't
support any optional actions, the sampled packets is always forwarded
to the E-Switch manager port.
Signed-off-by: Jiawei Wang <jiaweiw@nvidia.com> Acked-by: Viacheslav Ovsiienko <viacheslavo@nvidia.com>
Jiawei Wang [Tue, 13 Oct 2020 14:11:44 +0000 (17:11 +0300)]
common/mlx5: query sampler object capability via DevX
Update function mlx5_devx_cmd_query_hca_attr() to add the NIC Flow
Table attributes query, then get the log_max_flow_sampler_num from
flow table properties.
Add the related structs definition in mlx5_prm.h.
Signed-off-by: Jiawei Wang <jiaweiw@nvidia.com> Acked-by: Ori Kam <orika@nvidia.com> Acked-by: Viacheslav Ovsiienko <viacheslavo@nvidia.com>
Andrew Rybchenko [Tue, 13 Oct 2020 13:45:48 +0000 (14:45 +0100)]
net/sfc: support per-queue Rx prefix for EF100
Riverhead FW supports Rx prefix choice based on required fields in Rx
prefix. The feature is generalized in libefx to provide Rx prefixes
layout for other NICs and firmware variants. Now driver can get
the prefix layout after Rx queue start and use the layout details to
check its expectations or simply in run-time.
Rx prefix choice and query interface is defined in SF-119689-TC
EF100 host interface.
Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com> Reviewed-by: Andy Moreton <amoreton@xilinx.com>
Andrew Rybchenko [Tue, 13 Oct 2020 13:45:46 +0000 (14:45 +0100)]
common/sfc_efx/base: provide helper to check Rx prefix
A new function allows to check if used Rx prefix layout matches
available Rx prefix layout. Length check is out-of-scope of the
function and caller should ensure length is either checked or
different length with everything required in place is handled
properly.
Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>
Ivan Malov [Tue, 13 Oct 2020 13:45:38 +0000 (14:45 +0100)]
net/sfc: add header segments check for EF100 Tx
EF100 native Tx datapath demands that packet header be contiguous
when partial checksum offloads are used since helper function is
used to calculate pseudo-header checksum (and the function requires
contiguous header).
Add an explicit check for this assumption and restructure the code
to avoid TSO header linearisation check since TSO header
linearisation is not done on EF100 native Tx datapath.
Signed-off-by: Ivan Malov <ivan.malov@oktetlabs.ru> Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>
Andrew Rybchenko [Tue, 13 Oct 2020 13:45:32 +0000 (14:45 +0100)]
net/sfc: support datapath logs which may be compiled out
Add datapath log level which limits logs included in build since
on datapath it is too expensive to dive into rte_log() function
even if it does nothing.
Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>
These callbacks are used when event queue is polled via libefx.
The libefx polling is used for management event queue, but we do not
expect any Tx complete events on it, and for datapath event queue at
flushing.
Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>
These callbacks are used when event queue is polled via libefx.
The libefx polling is used for management event queue, but we do not
expect any Rx events on it, and for datapath event queue at flushing
(when these events are typically ignored, since queue is being stopped).
Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>
Igor Romanov [Tue, 13 Oct 2020 13:45:28 +0000 (14:45 +0100)]
net/sfc: use BAR layout discovery to find control window
Control window is required to talk to NIC.
Signed-off-by: Igor Romanov <igor.romanov@oktetlabs.ru> Reviewed-by: Andy Moreton <amoreton@xilinx.com> Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>
Andrew Rybchenko [Tue, 13 Oct 2020 13:45:26 +0000 (14:45 +0100)]
net/sfc: add capabilities for Rx/Tx support in libefx
libefx usage may be limited to control path only and its
implementation of datapath may not support NIC family or
PMD efx Rx/Tx datapaths implementation may be not yet ported
to updated libefx.
Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>
Andrew Rybchenko [Tue, 13 Oct 2020 13:45:23 +0000 (14:45 +0100)]
common/sfc_efx/base: add max number of Rx scatter buffers
Riverhead QDMA has limitation on maximum number of Rx scatter
buffers to be used by a packet. If the limitation is violated,
the datapath is dead. FW should ensure that it is OK, but
drivers need to know the limitation anyway to check parameters
when Rx queues are configured and MTU is set.
Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com> Reviewed-by: Andy Moreton <amoreton@xilinx.com>
Andrew Rybchenko [Tue, 13 Oct 2020 13:45:22 +0000 (14:45 +0100)]
common/sfc_efx/base: fix PHY config failure on Riverhead
Riverhead does not support LED control yet. It is perfectly
fine to ignore LED set failure because of no support if
configured LED mode is the default.
Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com> Reviewed-by: Andy Moreton <amoreton@xilinx.com>
Ciara Loftus [Tue, 13 Oct 2020 13:10:08 +0000 (13:10 +0000)]
net/af_xdp: forbid umem sharing for xsks with same context
AF_XDP PMDs who wish to share a UMEM must have a unique context
(ctx) ie. netdev,qid tuple. For instance, the following will not
work since both PMDs' contexts are identical.
Gaetan Rivet [Mon, 12 Oct 2020 14:19:04 +0000 (16:19 +0200)]
net/failsafe: fix state synchro cleanup
During a hotplug attempt, failsafe will try to bring a subdevice that
just appeared to its internal state. On error, the subdevice is marked
for removal and will be cleaned up.
However failsafe_dev_remove() only remove active devices. Devices that
failed during probe will be stuck in DEV_PARSED state repeatedly.
Consider all devices when doing a removal round, but limit burst control
and stats saving to active devices.
Wei Hu (Xavier) [Tue, 13 Oct 2020 11:50:53 +0000 (19:50 +0800)]
ethdev: extract checking queue id into common functions
This patch extract checking rx_queue_id or tx_queue_id into two separate
common functions named eth_dev_validate_rx_queue and
eth_dev_validate_tx_queue.
Signed-off-by: Chengchang Tang <tangchengchang@huawei.com> Signed-off-by: Wei Hu (Xavier) <xavier.huwei@huawei.com> Signed-off-by: Chengwen Feng <fengchengwen@huawei.com> Acked-by: Stephen Hemminger <stephen@networkplumber.org> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Reviewed-by: Andrew Rybchenko <andrew.rybchenko@oktetlabs.ru> Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
Dekel Peled [Wed, 7 Oct 2020 13:28:42 +0000 (16:28 +0300)]
ethdev: support query of age action
Existing API supports AGE action to monitor the aging of a flow.
This patch implements RFC [1], introducing the response format for query
of an AGE action.
Application will be able to query the AGE action state.
The response will be returned in the format implemented here.
Junfeng Guo [Tue, 13 Oct 2020 06:25:55 +0000 (14:25 +0800)]
net/ice: refactor RSS config wrap and fix potential bugs
Current implementation for PF RSS config wrap function has some
potential bugs about GTPU, e.g., same input set for GTPU inner and
non-TUN have different hash values, which should be same. Thus, we
use extra pre and post processing to re-config GTPU rules.
Jiawei Wang [Fri, 9 Oct 2020 13:46:06 +0000 (16:46 +0300)]
app/testpmd: support port and encap for sample action
Use sample action with ratio is 1 for mirroring flow, add
supports to set the different port or encap action for mirrored
packets.
The example of test-pmd command:
1. set sample_actions 1 port_id id 1 / end
flow create 0 ... pattern eth / end actions
sample ratio 1 index 1 / port_id id 2...
The flow will result in all the matched ingress packets will be sent to
port 2, and also mirrored the packets and sent to port 1.
2. set raw_encap 0 eth src.../ ipv4.../...
set raw_encap 1 eth src.../ ipv4.../...
set sample_actions 2 raw_encap index 0 / port_id id 0 / end
flow create 0 ... pattern eth / end actions
sample ratio 1 index 2 / raw_encap index 1 / port_id id 0...
The flow will result in all the matched egress packets will be
encapsulated and sent to wire, and also mirrored the packets and with
the different encapsulated data and sent to wire.
Signed-off-by: Jiawei Wang <jiaweiw@nvidia.com> Acked-by: Viacheslav Ovsiienko <viacheslavo@nvidia.com>
Jiawei Wang [Fri, 9 Oct 2020 13:46:05 +0000 (16:46 +0300)]
app/testpmd: add command for sample action
Add a new testpmd command 'set sample_actions' that supports the
multiple sample actions list configuration by using the index:
set sample_actions <index> <actions list>
The examples for the sample flow use case and result as below:
1. set sample_actions 0 mark id 0x8 / queue index 2 / end
.. pattern eth / end actions sample ratio 2 index 0 / jump group 2 ...
This flow will result in all the matched ingress packets will be
jumped to next flow table, and the each second packet will be
marked and sent to queue 2 of the control application.
2. ...pattern eth / end actions sample ratio 2 / port_id id 2 ...
The flow will result in all the matched ingress packets will be
duplicated and sent to the representor peer (VF or wire) on DPDK port 2,
and the each second packet will also be sent to E-Switch manager vport.
Signed-off-by: Jiawei Wang <jiaweiw@nvidia.com> Acked-by: Ori Kam <orika@nvidia.com> Acked-by: Viacheslav Ovsiienko <viacheslavo@nvidia.com>
Jiawei Wang [Fri, 9 Oct 2020 13:46:04 +0000 (16:46 +0300)]
ethdev: introduce flow sample action
When using full offload, all traffic will be handled by the HW, and
forwarded to the requested VF or wire and the control application does
not see this traffic anymore. So there's a need for an action that
enables the control application some forwarded traffic visibility.
The solution introduces a new action that will sample the incoming
traffic and send a duplicated traffic with the specified ratio to the
application, while the original packet will continue to the target
destination.
The packets sampled equals is '1/ratio', the ratio value set to 1
means that the packets will be completely mirrored. The sample packet
can be assigned with different set of actions from the original packet.
In order to support the sample packet in rte_flow, new rte_flow action
definition RTE_FLOW_ACTION_TYPE_SAMPLE and structure rte_flow_action_sample
will be introduced.
Signed-off-by: Jiawei Wang <jiaweiw@nvidia.com> Acked-by: Ori Kam <orika@nvidia.com> Acked-by: Jerin Jacob <jerinj@marvell.com> Acked-by: Andrew Rybchenko <arybchenko@solarflare.com> Acked-by: Viacheslav Ovsiienko <viacheslavo@nvidia.com> Acked-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
Ferruh Yigit [Fri, 9 Oct 2020 15:50:39 +0000 (16:50 +0100)]
net/pcap: fix crash on exit for infinite Rx
If the infinite Rx argument ('infinite_rx') is provided a ring is
allocated and filled in the '.rx_queue_setup' dev_ops.
Later this ring freed in the '.dev_close' dev_ops.
If the 'infinite_rx' provided and '.dev_close' called before
'.rx_queue_setup', the ring will be NULL and trying to empty/free it
will cause a crash.
This is fixed by adding ring NULL check before trying to empty/free it.
Added support for parent child flow database apis. This
feature adds support to enable vxlan decap support where
flows needs to maintain parent-child flow relationship.
Mike Baucom [Fri, 9 Oct 2020 11:11:25 +0000 (16:41 +0530)]
net/bnxt: support runtime EM selection
This patch adds support to select internal Exact Match vs
External Exact Match support while loading the PMD.
- Added new mem type conditional opcode for internal/external
- Adapted the flowdb resource counts based on selected mode
- Template changes to use the new opcode
- The decision for internal/external EM support is based on the
devargs parameter max_num_kflows. If this is set, external EM
is used.
The default and regular flows are stored in the same flow table
instead of different flow tables. This should help code reuse
and reducing the number of allocations.
So combine default and regular flows in flow database.
Venkat Duvvuru [Fri, 9 Oct 2020 11:11:21 +0000 (16:41 +0530)]
net/bnxt: fix PF support in SR-IOV mode
1. Implement HWRM_FUNC_VF_RESOURCE_CFG command and use it to
reserve resources for VFs when NEW RM is enabled.
2. Invoke PF’s FUNC_CFG before configuring VFs resources.
3. Don’t consider max_rx_em_flows in max_l2_ctx calculation
when VFs are configured.
4. Issue HWRM_FUNC_QCFG instead of HWRM_FUNC_QCAPS to find
out the actual allocated resources for VF.
5. Don’t add random mac to the VF.
6. Handle completion type CMPL_BASE_TYPE_HWRM_FWD_REQ instead
of CMPL_BASE_TYPE_HWRM_FWD_RESP.
7. Don't enable HWRM_FUNC_DRV_RGTR_INPUT_FLAGS_FWD_NONE_MODE
when the list of HWRM commands that needs to be forwarded
to the PF is specified in HWRM_FUNC_DRV_RGTR.
8. Update the HWRM commands list that can be forwarded to the
PF.
Mike Baucom [Fri, 9 Oct 2020 11:11:19 +0000 (16:41 +0530)]
net/bnxt: add multi-device infrastructure
In order to support multiple devices this patch:
- Breaks the template into device specific files
- Changes template list retrieval to use device id
- Determines the software device id using the bp pointer
- Determines the TRUFLOW resources based on device id
Signed-off-by: Mike Baucom <michael.baucom@broadcom.com> Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
The fast mbuf free offload for non-vector mode requires
additional checks in order to handle long tx buffer
descriptors, so dedicated functions are needed for
vector- and non-vector-modes.
Fixes: 369f6077c515 ("net/bnxt: support fast mbuf free") Signed-off-by: Lance Richardson <lance.richardson@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
Li Zhang [Fri, 9 Oct 2020 06:11:42 +0000 (09:11 +0300)]
net/mlx5: support ICMP identifier matching
PRM expose fields "Icmp_header_data" in IPv4 ICMP.
Update ICMP mask parameter with ICMP identifier and sequence number
fields.
ICMP sequence number spec with mask, Icmp_header_data low 16 bits are
set.
ICMP identifier spec with mask, Icmp_header_data high 16 bits are set.
Signed-off-by: Li Zhang <lizh@nvidia.com> Acked-by: Ori Kam <orika@nvidia.com>
Kiran Kumar K [Sat, 10 Oct 2020 06:33:10 +0000 (12:03 +0530)]
net/octeontx2: fix RSS flow create
While creating flow with action type RSS, action type is not being
set to RSS, and action type is being set to unicast. Therefore it breaks
RSS functionality. This patch add changes to program the RSS action
properly.
Thomas Monjalon [Wed, 7 Oct 2020 21:48:48 +0000 (23:48 +0200)]
ethdev: fix xstat name of basic stats per queue
As described in doc/guides/prog_guide/poll_mode_drv.rst,
the naming scheme for the xstats is parts separated with underscore:
* direction
* detail 1
* detail 2
* detail n
* unit
where detail 1 can be "q" followed with a queue number.
It means the name of the stats per queue should be rx_qN_* or tx_qN_*.
The second underscore was missing so far.
Fixing the basic xstat names may be considered an API change,
that's why it should not be backported.
While fixing this mistake, some examples of the naming scheme
are given as part of the API documentation of rte_eth_xstat_name.
More proposals about standardizing statistics:
http://fast.dpdk.org/events/slides/DPDK-2019-09-Ethernet_Statistics.pdf
Fixes: bd6aa172cf35 ("ethdev: fetch extended statistics with integer ids") Signed-off-by: Thomas Monjalon <thomas@monjalon.net> Reviewed-by: Bruce Richardson <bruce.richardson@intel.com> Acked-by: Kevin Traynor <ktraynor@redhat.com> Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com> Acked-by: Ciara Power <ciara.power@intel.com>
load-acquire memory order has a barrier. Loading it inside
the loop will result in a barrier in every iteration. Hence,
load the variable once outside the loop.
Signed-off-by: Honnappa Nagarahalli <honnappa.nagarahalli@arm.com> Reviewed-by: Phil Yang <phil.yang@arm.com> Reviewed-by: Ruifeng Wang <ruifeng.wang@arm.com> Reviewed-by: Jakub Grajciar <jgrajcia@cisco.com>
net/memif: relax load of ring head for S2M zc ring
For zero-copy S2M rings, ring->head is updated by the sender and
eth_memif_tx_zc function is called in the context of sending thread.
The loads in the sender do not need to synchronize with its own stores.
Signed-off-by: Honnappa Nagarahalli <honnappa.nagarahalli@arm.com> Reviewed-by: Phil Yang <phil.yang@arm.com> Reviewed-by: Ruifeng Wang <ruifeng.wang@arm.com> Reviewed-by: Jakub Grajciar <jgrajcia@cisco.com>
eth_memif_tx_zc function is called only for S2M rings. Remove
additional code for M2S rings in this function.
Signed-off-by: Honnappa Nagarahalli <honnappa.nagarahalli@arm.com> Reviewed-by: Phil Yang <phil.yang@arm.com> Reviewed-by: Ruifeng Wang <ruifeng.wang@arm.com> Reviewed-by: Jakub Grajciar <jgrajcia@cisco.com>
net/memif: relax load of ring head for M2S zc ring
For zero-copy M2S rings, ring->head is updated by the receiver
and eth_memif_rx_zc function is called in the context of
receiving thread. The loads in the receiver do not need to
synchronize with its own stores.
Signed-off-by: Honnappa Nagarahalli <honnappa.nagarahalli@arm.com> Reviewed-by: Phil Yang <phil.yang@arm.com> Reviewed-by: Ruifeng Wang <ruifeng.wang@arm.com> Reviewed-by: Jakub Grajciar <jgrajcia@cisco.com>
For S2M rings, ring->head is updated by the sender and eth_memif_tx
function is called in the context of sending thread. The loads in
the sender do not need to synchronize with its own stores.
Fixes: a2aafb9aa651 ("net/memif: optimize with one-way barrier") Cc: stable@dpdk.org Signed-off-by: Honnappa Nagarahalli <honnappa.nagarahalli@arm.com> Reviewed-by: Phil Yang <phil.yang@arm.com> Reviewed-by: Ruifeng Wang <ruifeng.wang@arm.com> Reviewed-by: Jakub Grajciar <jgrajcia@cisco.com>
For M2S rings, ring->head is updated by the receiver and eth_memif_rx
function is called in the context of receiving thread. The loads in
the receiver do not need to synchronize with its own stores.
Fixes: a2aafb9aa651 ("net/memif: optimize with one-way barrier") Cc: stable@dpdk.org Signed-off-by: Honnappa Nagarahalli <honnappa.nagarahalli@arm.com> Reviewed-by: Phil Yang <phil.yang@arm.com> Reviewed-by: Ruifeng Wang <ruifeng.wang@arm.com> Reviewed-by: Jakub Grajciar <jgrajcia@cisco.com>
For M2S rings, ring->tail is updated by the sender and eth_memif_tx
function is called in the context of sending thread. The loads in
the sender do not need to synchronize with its own stores.
Fixes: a2aafb9aa651 ("net/memif: optimize with one-way barrier") Cc: stable@dpdk.org Signed-off-by: Honnappa Nagarahalli <honnappa.nagarahalli@arm.com> Reviewed-by: Phil Yang <phil.yang@arm.com> Reviewed-by: Ruifeng Wang <ruifeng.wang@arm.com> Reviewed-by: Jakub Grajciar <jgrajcia@cisco.com>
In the case of S2M queues, the receiver synchronizes with the sender
(i.e. informs of the packets it has received) using ring->tail.
Hence, the sender does not need to update last_tail.
In the case of M2S queues, the receiver uses last_tail to
keep track of the descriptors it has received. The
sender is not required to update the last_tail. Updating
the last_tail makes it a shared variable between the
transmitter and receiver affecting the performance.
Fixes: 09c7e63a71f9 ("net/memif: introduce memory interface PMD") Cc: stable@dpdk.org Signed-off-by: Honnappa Nagarahalli <honnappa.nagarahalli@arm.com> Reviewed-by: Phil Yang <phil.yang@arm.com> Reviewed-by: Ruifeng Wang <ruifeng.wang@arm.com> Reviewed-by: Jakub Grajciar <jgrajcia@cisco.com>
Ferruh Yigit [Thu, 8 Oct 2020 11:34:25 +0000 (12:34 +0100)]
net/pcap: fix input only Rx
When input only Rx interface argument 'rx_iface_in' provided, current
code assigns 'eth_null_rx' burst function by mistake and no packet
received as a result.
Like in following usage no packets received from physical interface:
"--vdev net_pcap0,rx_iface_in=eth0,tx_iface=eth0"
Fixing the burst function assignment when 'rx_iface_in' argument is used
RongQing Li [Tue, 22 Sep 2020 10:29:31 +0000 (18:29 +0800)]
net/bonding: fix possible unbalanced packet receiving
Current Rx round robin policy for the slaves has two issue:
1. active_slave in bond_dev_private is shared by multiple PMDS which
maybe cause some slave Rx hungry, for example, there is two PMD and
two slave port, both PMDs start to receive, and see that active_slave
is 0, and receive from slave 0, after complete, they increase
active_slave by one, totally active_slave are increased by two, next
time, they will start to receive from slave 0 again, at last, slave 1
maybe drop packets during to not be polled by PMD
2. active_slave is shared and written by multiple PMD in RX path for
every time RX, this is a kind of cache false share, low performance.
So move active_slave from bond_dev_private to bond_rx_queue make it as
per queue variable
Fixes: ae2a04864a9a ("net/bonding: reduce slave starvation on Rx poll") Cc: stable@dpdk.org Signed-off-by: RongQing Li <lirongqing@baidu.com> Signed-off-by: Dongsheng Rong <rongdongsheng@baidu.com> Reviewed-by: Wei Hu (Xavier) <xavier.huwei@huawei.com>
Ophir Munk [Thu, 8 Oct 2020 20:16:56 +0000 (20:16 +0000)]
app/testpmd: cleanup tunnel protocols parsing
This is a cleanup commit.
It assembles all tunnel outer updates into one function call to avoid
code duplications.
It defines RTE_VXLAN_GPE_DEFAULT_PORT (4790) in accordance with all
other tunnel protocol definitions.
It replaces all numeric values 4789 in their corresponding definition
RTE_VXLAN_GPE_DEFAULT_PORT.
It updates the 'csum parse-tunnel' documentation.
Ophir Munk [Thu, 8 Oct 2020 20:16:55 +0000 (20:16 +0000)]
app/testpmd: enable configuring GENEVE port
IANA has assigned port 6081 as the fixed well-known destination port for
GENEVE. Nevertheless draft-ietf-nvo3-geneve-09 recommends that
implementations make this configurable. This commit enables specifying
any positive UDP destination port number for GENEVE protocol parsing.
Ophir Munk [Thu, 8 Oct 2020 20:16:54 +0000 (20:16 +0000)]
app/testpmd: add GENEVE parsing
GENEVE is a widely used tunneling protocol in modern Virtualized
Networks. testpmd already supports parsing of several tunneling
protocols including VXLAN, VXLAN-GPE, GRE. This commit adds GENEVE
parsing of inner protocols (IPv4-0x0800, IPv6-0x86dd, Ethernet-0x6558)
based on IETF draft-ietf-nvo3-geneve-09. GENEVE is considered more
flexible than the other protocols. In terms of protocol format GENEVE
header has a variable length options as opposed to other tunneling
protocols which have a fixed header size.
Yunjian Wang [Wed, 7 Oct 2020 02:06:25 +0000 (10:06 +0800)]
net/netvsc: fix Tx queue leak in error path
In hn_dev_tx_queue_setup() allocated memory for txq, we don't free it
when error happens and it will lead to memory leak.
We can check for tx_free_thresh at the beginning of the function to
fix it, before calling txq = rte_zmalloc_socket().
Fixes: cc0251813277 ("net/netvsc: split send buffers from Tx descriptors") Cc: stable@dpdk.org Signed-off-by: Yunjian Wang <wangyunjian@huawei.com> Acked-by: Long Li <longli@microsoft.com>
Chenxu Di [Tue, 6 Oct 2020 17:16:37 +0000 (18:16 +0100)]
net/dpaa2: fix build with timesync functions
Some timesync related source files are built only when
'RTE_LIBRTE_IEEE1588' config is set, which is missing in meson.
Building with '-DRTE_LIBRTE_IEEE1588' cflag will enable timesync
functionality in the files that are build, but it won't build files
that require 'RTE_LIBRTE_IEEE1588' config.
This causes the build error.
Fixing by removing config check in meson file and build all .c files by
default, but wrap relevant code part with 'RTE_LIBRTE_IEEE1588' macro.
Also removing 'RTE_LIBRTE_IEEE1588' ifdef around some fields of the data
structures, to not require finer grained macro wraps on the functions
using them.
Since the registration of the function disabled with macro check, having
functions compiled shouldn't affect the functionality.
Maxime Coquelin [Thu, 1 Oct 2020 10:11:55 +0000 (12:11 +0200)]
vhost: use fixed virtio-net header length packed ring
This small optimization uses the static the Virtio-net
header len in packed datapath, since Virtio-net header
cannot be the legacy one in case of packed ring.
Signed-off-by: Maxime Coquelin <maxime.coquelin@redhat.com> Reviewed-by: Chenbo Xia <chenbo.xia@intel.com>
Maxime Coquelin [Thu, 1 Oct 2020 10:11:54 +0000 (12:11 +0200)]
vhost: fix virtio-net header length with packed ring
In case packed ring layout has been negotiated, but neither
Version 1 nor mergeable buffers, the Virtio-net header len
is assigned to the legacy devices value, which is wrong.
This patch fixes this with using the proper len as devices
using packed ring are not legacy devices.
Fixes: a922401f35cc ("vhost: add Rx support for packed ring") Fixes: ae999ce49dcb ("vhost: add Tx support for packed ring") Cc: stable@dpdk.org Reported-by: Marvin Liu <yong.liu@intel.com> Signed-off-by: Maxime Coquelin <maxime.coquelin@redhat.com> Reviewed-by: Chenbo Xia <chenbo.xia@intel.com>
Olivier Matz [Wed, 7 Oct 2020 12:53:18 +0000 (14:53 +0200)]
vhost: fix external mbuf creation
In virtio_dev_extbuf_alloc(), the shinfo structure used to store
the reference counter and the free callback of the external buffer
is by default stored inside the mbuf data.
This is wrong because the mbuf (and its data) can be freed before
the external buffer, for instance in the following situation:
Fan Zhang [Fri, 2 Oct 2020 15:36:01 +0000 (16:36 +0100)]
vhost/crypto: fix feature negotiation
This patch fixes the feature negotiation for vhost crypto during
initialization. The patch uses the newly created driver start
function to inform the driver type with the fixed vhost features.
In addition the patch provides a new API specifically used by
the application to start a vhost-crypto driver.
Fixes: 939066d96563 ("vhost/crypto: add public function implementation") Cc: stable@dpdk.org Signed-off-by: Fan Zhang <roy.fan.zhang@intel.com> Reviewed-by: Maxime Coquelin <maxime.coquelin@redhat.com>
Bruce Richardson [Mon, 12 Oct 2020 14:51:48 +0000 (15:51 +0100)]
eal/x86: fix memcpy AVX-512 enablement
When testing on some x86 platforms, code compiled with meson was observed
running at a different power-license level to that compiled with make. This
is due to the fact that meson auto-detects the instruction sets available
on the system and enabled AVX512 rte_memcpy when AVX512 was available,
while on make, a build time AVX-512 flag needed to be explicitly set to
enable that AVX512 rte_memcpy code path.
In the absence of runtime path selection for rte_memcpy - which is
complicated by it being a static inline function in a header file - we can
fix this behaviour regression by similarly having a build-time option which
must be set to enable the AVX-512 memcpy path.
Fixes: a25a650be5f0 ("build: add infrastructure for meson and ninja builds") Fixes: 3e1bb55fd6ef ("build/x86: add SSE flags") Cc: stable@dpdk.org Signed-off-by: Bruce Richardson <bruce.richardson@intel.com> Tested-by: Yingya Han <yingyax.han@intel.com>
git.droids-corp.org / dpdk.git / log