The ASO connection tracking structure is initialized once for sharing
device context.
Its release takes place in the close function which is called for each
ethdev individually. i.e. when there is more than one ethdev under the
same sharing device context, it will be destroyed when one of them is
closed. If the other wants to use it later, it may cause it to crash.
In addition, the creation of this structure is performed in the spawn
function. If one of the creations of the objects following it fails, it
is supposed to be destroyed but this does not happen.
This patch moves its release to the sharing device context free function
and thus solves both problems.
Fixes: 0af8a2298a42 ("net/mlx5: release connection tracking management")
Fixes: ee9e5fad03eb ("net/mlx5: initialize connection tracking management")
Cc: stable@dpdk.org
Signed-off-by: Michael Baum <michaelba@nvidia.com>
Acked-by: Matan Azrad <matan@nvidia.com>
In "dv_xmeta_en" devarg there is an option of dv_xmeta_en=3 which
engages tunnel offload mode. In E-Switch configuration, that mode
implicitly activates dv_xmeta_en=1.
The update according to E-switch support is done immediately after the
first parsing of the devargs, but there is another adjustment later.
This patch moves the adjustment after the second parsing.
Fixes: 4ec6360de37d ("net/mlx5: implement tunnel offload")
Cc: stable@dpdk.org
Signed-off-by: Michael Baum <michaelba@nvidia.com>
Acked-by: Matan Azrad <matan@nvidia.com>
The MLX5 net driver supports "probe again". In probing again, it
creates a new ethdev under an existing infiniband device context.
Sibling devices sharing infiniband device context should have compatible
configurations, so some of the devargs given in the probe again, the
ones that are mainly relevant to the sharing device context are sent to
the mlx5_dev_check_sibling_config function which makes sure that they
compatible its siblings.
However, the arguments are adjusted according to the capability of the
device, and the function compares the arguments of the probe again
before the adjustment with the arguments of the siblings after the
adjustment. A user who sends the same values to all siblings may fail in
this comparison if he requested something that the device does not
support and adjusted.
This patch moves the call to the mlx5_dev_check_sibling_config function
after the relevant adjustments.
Fixes: 92d5dd483450 ("net/mlx5: check sibling device configurations mismatch")
Fixes: 2d241515ebaf ("net/mlx5: add devarg for extensive metadata support")
Cc: stable@dpdk.org
Signed-off-by: Michael Baum <michaelba@nvidia.com>
Acked-by: Matan Azrad <matan@nvidia.com>
Currently max frame size is set at queue level, which makes the values
of the following counters wrong when a jumbo frame is received.
The expected value:
rx_good_bytes: 0
rx_errors: 1
rx_oversize_errors: 1
The actual value:
rx_good_bytes: 1626
rx_errors: 0
rx_oversize_errors: 0
This patch enables setting max frame size at port level, and makes the
values above right.
Cc: stable@dpdk.org
Signed-off-by: Dapeng Yu <dapengx.yu@intel.com>
Tested-by: Peng Zhang <peng1x.zhang@intel.com>
Reviewed-by: Qi Zhang <qi.z.zhang@intel.com>
This fix commit ff8b8bcd2ebe, which resulted in incorrect buffer size
being computed for non IPses TSO packets.
Fixes: ff8b8bcd2ebe ("net/iavf: fix segmentation offload condition")
Cc: stable@dpdk.org
Signed-off-by: Radu Nicolau <radu.nicolau@intel.com>
Acked-by: Qi Zhang <qi.z.zhang@intel.com>
After enabling DCF on a VF, the ice driver stops receiving
link updates on it's Admin Receive Queue. During the init
of DCF ice_aqc_opc_get_link_status command is send to the
firmware without LSE (Link Status Event) bit set. This prevents
the ice driver from receiving up/down events, and correspondingly
updating netdev.
Fixes: 0b02c9519432 ("net/ice: handle PF initialization by DCF")
Cc: stable@dpdk.org
Signed-off-by: Michal Wilczynski <michal.wilczynski@intel.com>
Acked-by: Haiyue Wang <haiyue.wang@intel.com>
Some zero copy AF_XDP drivers eg. ice require that there are addresses
already in the fill queue before the socket is created. Otherwise you may
see log messages such as:
XSK buffer pool does not provide enough addresses to fill 2047 buffers on
Rx ring 0
This commit ensures that the addresses are available before creating the
socket, instead of after.
Signed-off-by: Ciara Loftus <ciara.loftus@intel.com>
Tested-by: Ferruh Yigit <ferruh.yigit@intel.com>
The Rx queue setup can fail for many reasons eg. failure to setup the
custom program, failure to allocate or reserve fill queue buffers,
failure to configure busy polling etc. When a failure like one of these
occurs, if the xsk is already set up it should be deleted before
returning. This commit ensures this happens.
Fixes: d8a210774e1d ("net/af_xdp: support unaligned umem chunks")
Fixes: 288a85aef192 ("net/af_xdp: enable custom XDP program loading")
Fixes: 055a393626ed ("net/af_xdp: prefer busy polling")
Fixes: 01fa83c94d7e ("net/af_xdp: workaround custom program loading")
Cc: stable@dpdk.org
Signed-off-by: Ciara Loftus <ciara.loftus@intel.com>
The size of unit cache should be sizeof(**cache) instead of
sizeof(*cache). Memory reallocation is inadequate by sizeof(*cache)
for the platform whose size of pointer is 32-bits. Found by coccinelle
(see https://coccinelle.gitlabpages.inria.fr/website) script.
Fixes: 63abf8d29225 ("net/sfc: support SW stats groups")
Cc: stable@dpdk.org
Signed-off-by: Weiguo Li <liwg06@foxmail.com>
Reviewed-by: Andrew Rybchenko <andrew.rybchenko@oktetlabs.ru>
The condition for that must use the new MAE admin status.
Fixes: 2f577f0ea1a3 ("net/sfc: allow ports without MAE privilege")
Cc: stable@dpdk.org
Signed-off-by: Ivan Malov <ivan.malov@oktetlabs.ru>
Reviewed-by: Andrew Rybchenko <andrew.rybchenko@oktetlabs.ru>
Reviewed-by: Andy Moreton <amoreton@xilinx.com>
One may set RECIRC_ID in a MAE action rule specification.
This field is not a network one, and its handling goes
to the code snippet which does not recognise field
sizes other than 4 bytes. Add the missing handler.
Fixes: 3a73dcfdb255 ("common/sfc_efx/base: match on recirc ID in action rules")
Cc: stable@dpdk.org
Signed-off-by: Ivan Malov <ivan.malov@oktetlabs.ru>
Reviewed-by: Andrew Rybchenko <andrew.rybchenko@oktetlabs.ru>
Reviewed-by: Andy Moreton <amoreton@xilinx.com>
Recirculation ID field of MAE outer rule insert MCDI is
part of the lookup control structure and it has non-zero
bit offset relative to the byte offset of the structure.
Fixes: 5cf153e79c6c ("common/sfc_efx/base: support recirculation ID in outer rules")
Cc: stable@dpdk.org
Signed-off-by: Ivan Malov <ivan.malov@oktetlabs.ru>
Reviewed-by: Andy Moreton <amoreton@xilinx.com>
Except for the cn*_rte_flow* files, none of the other
file has rte in the file name, remove the
rte to unify the file name across the directory.
Signed-off-by: Jerin Jacob <jerinj@marvell.com>
Acked-by: Kiran Kumar K <kirankumark@marvell.com>
Adding changes to enable keep flow rule device capability.
With this change, flow rules will be kept across device restart.
Signed-off-by: Kiran Kumar K <kirankumark@marvell.com>
Reviewed-by: Satheesh Paul <psatheesh@marvell.com>
Acked-by: Ray Kinsella <mdr@ashroe.eu>
Fix uninitialized local variable as reported by coverity scan.
Coverity issue: 373630
Fixes: 6af19a9d89bf ("net/cnxk: support meter action to flow destroy")
Cc: stable@dpdk.org
Signed-off-by: Gowrishankar Muthukrishnan <gmuthukrishn@marvell.com>
Acked-by: Jerin Jacob <jerinj@marvell.com>
Adds telemetry function to get information about inline
outbound and inline inbound SA's. The function takes port id
as input. Some fields in the structures roc_ot_ipsec_outb_sa
and roc_ot_ipsec_inb_sa are returned as output.
Signed-off-by: Ankur Dwivedi <adwivedi@marvell.com>
Reviewed-by: Jerin Jacob <jerinj@marvell.com>
Adds telemetry function to get information about inline
outbound and inline inbound SA's. The function takes port id
as input. Some fields in the structures roc_onf_ipsec_outb_sa
and roc_onf_ipsec_inb_sa are returned as output.
Signed-off-by: Ankur Dwivedi <adwivedi@marvell.com>
Reviewed-by: Jerin Jacob <jerinj@marvell.com>
libbpf v0.7.0 deprecates the bpf_prog_load function. Use meson to detect
if libbpf >= v0.7.0 is linked and if so, use the recommended replacement
functions bpf_object__open_file and bpf_object__load.
Signed-off-by: Ciara Loftus <ciara.loftus@intel.com>
Caught while trying --in-memory mode, some log messages in this driver
are not terminated with a newline:
rte_pmd_af_xdp_probe(): net_af_xdp: Failed to register multi-process IPC
callback: Operation not supportedvdev_probe(): failed to initialize
net_af_xdp device
Other locations in this driver had the same issue, fix all at once.
Fixes: f1debd77efaf ("net/af_xdp: introduce AF_XDP PMD")
Fixes: d8a210774e1d ("net/af_xdp: support unaligned umem chunks")
Fixes: 9876cf8316b3 ("net/af_xdp: re-enable secondary process support")
Cc: stable@dpdk.org
Signed-off-by: David Marchand <david.marchand@redhat.com>
Acked-by: Ciara Loftus <ciara.loftus@intel.com>
If EAL multiprocess feature has been disabled via rte_mp_disable()
function, AF_XDP driver may not be able to register its IPC callback.
Previously this leads to probe failure.
This commit adds a check for this condition so that AF_XDP can still be
used even if multiprocess is disabled.
Fixes: 9876cf8316b3 ("net/af_xdp: re-enable secondary process support")
Signed-off-by: Junxiao Shi <git@mail1.yoursunny.com>
Acked-by: Ciara Loftus <ciara.loftus@intel.com>
The x->len is passed as argument to npc_prep_mcam_ldata().
In the function, the len is used to reference elements of
int_info and int_info_mask array.
The arrays are of length NPC_MAX_EXTRACT_DATA_LEN.
Validating the x->len value so that it is not greater than
NPC_MAX_EXTRACT_DATA_LEN.
This patch also resolves warning observed with gcc 12 compiler.
log:
../drivers/common/cnxk/roc_npc_utils.c:13:26: warning: writing 16 bytes
into a region of size 0 [-Wstringop-overflow=]
ptr[idx] = data[len - 1 - idx];
../drivers/common/cnxk/roc_npc_utils.c:163:17: note: at offset 64 into
destination object ‘int_info’ of size 64
uint8_t int_info[NPC_MAX_EXTRACT_DATA_LEN];
Bugzilla ID: 854
Fixes: 665b6a7400bf ("common/cnxk: add NPC helper API")
Cc: stable@dpdk.org
Signed-off-by: Ankur Dwivedi <adwivedi@marvell.com>
Reviewed-by: Kiran Kumar K <kirankumark@marvell.com>
Buffer for MCDI channel is allocated using rte_memzone_reserve_aligned
with zone name 'mcdi'. Since multiple MCDI channels are needed to
support multiple VF(s) and rte_memzone_reserve_aligned expects unique
zone names, append PCI address to zone name to make it unique.
Signed-off-by: Abhimanyu Saini <asaini@xilinx.com>
Reviewed-by: Maxime Coquelin <maxime.coquelin@redhat.com>
An errata exists whereby NIX may incorrectly overwrite the value in
NIX_SQ_CTX_S[SQ_INT]. This may cause set interrupts to get cleared or
causing an QINT when no error is outstanding.
As a workaround, software should always read all SQ debug registers
and not just rely on NIX_SQINT_E bits set in NIX_SQ_CTX_S[SQ_INT].
Also for detecting SQB faults software must read SQ context and
check id next_sqb is NULL.
Signed-off-by: Harman Kalra <hkalra@marvell.com>
Acked-by: Jerin Jacob <jerinj@marvell.com>
When Rx/Tx functions are updated such as in a rte flow
rule creation with VLAN stripping or marking action,
update the fastpath ops table as well.
Fixes: b951c2efcb15 ("net/cnxk: add Rx burst for CN9K")
Fixes: f742a9a66813 ("net/cnxk: add Rx burst for CN10K")
Cc: stable@dpdk.org
Signed-off-by: Satheesh Paul <psatheesh@marvell.com>
Reviewed-by: Jerin Jacob <jerinj@marvell.com>
Egress rules do not have a base steering rule and the
current base steering rule is applicable only for ingress
rules. Hence, when creating a flow rule, merge base steering
rule only for ingress rules.
Fixes: f9af90807466 ("common/cnxk: add mcam utility API")
Cc: stable@dpdk.org
Signed-off-by: Satheesh Paul <psatheesh@marvell.com>
Reviewed-by: Kiran Kumar K <kirankumark@marvell.com>
Changed log level from info to debug for a couple of log
messages printed during MCAM allocation.
Fixes: 1f66919817ee ("common/cnxk: improve MCAM entries management")
Cc: stable@dpdk.org
Signed-off-by: Satheesh Paul <psatheesh@marvell.com>
Reviewed-by: Kiran Kumar K <kirankumark@marvell.com>
When a flow is deleted, the counters are deleted without
being cleared first. This results in counter values being
retained and shown in flows created later. Hence, counter
is cleared before being deleted.
Updated driver to clear and delete counters upon flow deletion.
Fixes: a07f7ced436d ("common/cnxk: add NPC init and fini")
Cc: stable@dpdk.org
Signed-off-by: Satheesh Paul <psatheesh@marvell.com>
Reviewed-by: Jerin Jacob <jerinj@marvell.com>
Added null check in IPv6 flow pattern parsing.
Fixes: 474e275b1bc6 ("common/cnxk: support extensions attributes in IPv6 item")
Signed-off-by: Satheesh Paul <psatheesh@marvell.com>
Reviewed-by: Jerin Jacob <jerinj@marvell.com>
If a initial process for the bonding device is like:
rte_eth_dev_configure
rte_eth_dev_set_mtu
queue setup and start, etc.
Pass the vdev args to application, and init bonding device only.
-a 0000:af:00.0 --vdev="net_bonding0,mode=2,slave=0000:af:00.0"
It will fail and complain for the slave device
"Port 0 must be configured before MTU set"
Test can be reproduced with ovs.
Fixes: b26bee10ee37 ("ethdev: forbid MTU set before device configure")
Cc: stable@dpdk.org
Signed-off-by: Junjie Wan <wanjunjie@bytedance.com>
Tested-by: Ferruh Yigit <ferruh.yigit@intel.com>
Acked-by: Min Hu (Connor) <humin29@huawei.com>
When invalid args exist, application exits with no error hint.
Adding a log message here will help users to know the reason.
Signed-off-by: Junjie Wan <wanjunjie@bytedance.com>
Acked-by: Ferruh Yigit <ferruh.yigit@intel.com>
Acked-by: Min Hu (Connor) <humin29@huawei.com>
If multi-segment is enabled and single segment/packet
is received, then mbuf data_len is not updated in
cn9k_nix_cqe_to_mbuf() function.
Also, in case of timestamp is enabled, mbuf data_len
and pkt_len will be updated for all packets including
multi segmented packets.
Fixes: 7c6bee34064f ("net/cnxk: enable PTP processing in vector Rx")
Cc: stable@dpdk.org
Signed-off-by: Rahul Bhansali <rbhansali@marvell.com>
Acked-by: Jerin Jacob <jerinj@marvell.com>
When iavf_security_ctx_destroy() is cleaning up, it does not correctly
set the security_ctx variable to NULL, which will lead to wild pointer.
Fixes: 6bc987ecb860 ("net/iavf: support IPsec inline crypto")
Cc: stable@dpdk.org
Signed-off-by: Yunjian Wang <wangyunjian@huawei.com>
Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
When txgbe_dev_close() is cleaning up, it does not correctly set
the security_ctx variable to NULL, which will lead to wild pointer.
Fixes: f437d97c3df1 ("net/txgbe: add IPsec context creation")
Cc: stable@dpdk.org
Signed-off-by: Yunjian Wang <wangyunjian@huawei.com>
Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
When ixgbe_dev_close() is cleaning up, it does not correctly set
the security_ctx variable to NULL, which will lead to wild pointer.
Fixes: 9a0752f498d2 ("net/ixgbe: enable inline IPsec")
Cc: stable@dpdk.org
Signed-off-by: Yunjian Wang <wangyunjian@huawei.com>
Acked-by: Haiyue Wang <haiyue.wang@intel.com>
Add support for chunk DMA transfers.
Chunk mpu transfer use 64 objects (512 byte) to maintain memory
read alignment.
Align mpu memory allocation to be at 512 byte boundaries.
Reduce force-close allocation from 10000 objects to 64 objects.
Signed-off-by: John Miller <john.miller@atomicrules.com>
Added arbitrary mbuf size per queue capability.
Updated ARK_UDM_CONST3 value to reflect the version number
read from the HW that is required to support this change.
Signed-off-by: John Miller <john.miller@atomicrules.com>
Add a device capabilities record for supported features.
Certain variants require that PCIe read-requests be correctly
throttled. This is called "rqpacing" in Arkville, and has to do
with credit and flow control on certain Arkville implementations.
Signed-off-by: John Miller <john.miller@atomicrules.com>
In the firmware, the promisc mode overrides the multicast mode.
So when the promisc mode is turned off, driver must check if the
multicast mode was active before and conditionally reactivate it.
Fixes: 6435f9a0ac22 ("net/nfb: add new netcope driver")
Cc: stable@dpdk.org
Signed-off-by: Martin Spinler <spinler@cesnet.cz>
The indexes in the for cycle were wrongly used and
the code accessed outside of the rxmac/txmac array.
Fixes: 6435f9a0ac22 ("net/nfb: add new netcope driver")
Cc: stable@dpdk.org
Signed-off-by: Martin Spinler <spinler@cesnet.cz>
This patch fixes an issue found by coverity. It checks for a possible
null value of "error" before dereferencing it.
CID: 375064: Dereference after null check
Either the check against null is unnecessary, or there may be a null
pointer dereference.
Coverity issue: 375064
Fixes: ee806eea59fe ("net/enic: support GENEVE flow item")
Signed-off-by: John Daley <johndale@cisco.com>
Reviewed-by: Hyong Youb Kim <hyonkim@cisco.com>
