The mlx5_rxq_obj_hairpin_new function defines a pointer named tmpl and
allocates memory for it using the rte_zmalloc_socket function.
Later, this function allocates memory to a variable inside tmpl using
the mlx5_devx_cmd_create_rq function.
In both cases, if the allocation fails, the code jumps to the error
label and frees allocated resources. However, in the first jump there
are still no resources to free and the jump only for the line return
NULL is unnecessary. Even worse, when it jumps to error label with
invalid tmpl it actually does dereference to a null pointer.
In contrast, the second jump needs to free the tmpl variable but the
function instead of freeing, tries to free the variable that it just
failed to allocate.
In addition, for another error, the function returns NULL without
freeing the tmpl variable before, causing a memory leak.
Delete the error label and replace each jump with local return NULL and
free tmpl variable if needed.
Fixes: e79c9be915 ("net/mlx5: support Rx hairpin queues")
Cc: stable@dpdk.org
Signed-off-by: Michael Baum <michaelba@mellanox.com>
Acked-by: Matan Azrad <matan@mellanox.com>
The mlx5_txq_obj_hairpin_new function defines a pointer named tmpl and
allocates memory for it using the rte_zmalloc_socket function.
Later, this function allocates memory to a variable inside tmpl using
the mlx5_devx_cmd_create_sq function.
In both cases, if the allocation fails, the code jumps to the error
label and frees allocated resources. However, in the first jump there
are still no resources to free and the jump only for the line return
NULL is unnecessary. Even worse, when it jumps to error label with
invalid tmpl it actually does dereference to a null pointer.
In contrast, the second jump needs to free the tmpl variable but the
function instead of freeing, tries to free the variable that it just
failed to allocate, and another variable that has never been allocated.
In addition, for another error, the function returns NULL without
freeing the tmpl variable before, causing a memory leak.
Delete the error label and replace each jump with local return NULL and
free tmpl variable if needed.
Fixes: ae18a1ae96 ("net/mlx5: support Tx hairpin queues")
Cc: stable@dpdk.org
Signed-off-by: Michael Baum <michaelba@mellanox.com>
Acked-by: Matan Azrad <matan@mellanox.com>
The PCI DSN (device serial number) to format package file name should be
lowercase values.
Fixes: d1c91179e9 ("net/ice: check DSN package file firstly")
Cc: stable@dpdk.org
Signed-off-by: Haiyue Wang <haiyue.wang@intel.com>
Acked-by: Xiaolong Ye <xiaolong.ye@intel.com>
In the ingress path, the cfa_code field in Rx completion identifies the
CFA action rule that was used for the incoming packet. It is possible
that the packet could hit the rule at index 0 in the table.
The mark action code was too restrictive by disallowing a cfa_code of
zero.
This code loosens the requirement and allows zero.
Fixes: b87abb2e55 ("net/bnxt: support marking packet")
Cc: stable@dpdk.org
Signed-off-by: Mike Baucom <michael.baucom@broadcom.com>
Reviewed-by: Kishore Padmanabha <kishore.padmanabha@broadcom.com>
Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
When closing VF device, the process of shutdown adminq should be after
the process of uninit the flow, since the VF might still need to use the
adminq to uninit flow.
Fixes: 9e03acd726 ("net/iavf: fix flow access")
Fixes: ff2d0c345c ("net/iavf: support generic flow API")
Cc: stable@dpdk.org
Signed-off-by: Jeff Guo <jia.guo@intel.com>
Acked-by: Xiaolong Ye <xiaolong.ye@intel.com>
Because of bugs in driver or host a reply to a request might
never occur. Better to give an error than spin forever.
Fixes: 4e9c73e96e ("net/netvsc: add Hyper-V network device")
Cc: stable@dpdk.org
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
The original code would deadlock itself if a link change event
happened with link state interrupt enabled. The problem is that
the link state changed message would be seen while reading
the host to guest ring (under lock) and then the driver would
send a query to the host to see the new link state. The response
would never be seen (stuck in a while loop) waiting for the
response.
The solution is to use the link change indication to trigger
a DPDK alarm. The alarm will happen in a different thread and
in that context it can send request for new link state and
also do interrupt callback. This is similar to how the bonding
driver is handling the same thing.
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
When the primary device link state is queried, there is no
need to query the VF state as well. The application only sees
the state of the synthetic device.
Fixes: dc7680e859 ("net/netvsc: support integrated VF")
Cc: stable@dpdk.org
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
The code to unset owner of VF device was changing port to invalid
value before calling unset.
Fixes: 4a9efcddad ("net/netvsc: fix VF support with secondary process")
Cc: stable@dpdk.org
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
The PMD_TX_LOG and PMD_RX_LOG can hide errors since this
debug log is typically disabled. Change the code to use
PMD_DRV_LOG for errors.
Under load, the ring buffer to the host can fill.
Add some statistics to estimate the impact and see other errors.
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
These functions are useful for applications and debugging.
The netvsc PMD also transparently handles the rx/tx descriptor
functions for underlying VF device.
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
There is not a lot of info here from this driver.
But worth supporting these additional info queries.
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
We are freeing flow_stats a little early. This results in a
segfault when the driver accesses the members during cleanup.
Move the call to bnxt_free_flow_stats_info() to prevent this.
Fixes: 02a95625fe ("net/bnxt: add flow stats in extended stats")
Cc: stable@dpdk.org
Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
bnxt PMD uses the macro BNXT_SUPPORTED_SPEEDS to validate
the user requested speed. But this has all the speed values
supported by the PMD and is not chip specific.
The check against this macro returns success when the user
tries set the speed to 100G on a port even if the chip does
not support 100G speed.
Fixed it to use bnxt_get_speed_capabilities() to check the
supported speeds by the chip.
Fixes: 1d0704f4d7 ("net/bnxt: add device configure operation")
Cc: stable@dpdk.org
Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com>
Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
When users set the length of RSS hash key greater than the supported
length by hardware, the driver should intercept and can not configure
the wrong key into the hardware.
Fixes: c37ca66f2b ("net/hns3: support RSS")
Cc: stable@dpdk.org
Signed-off-by: Lijun Ou <oulijun@huawei.com>
Signed-off-by: Wei Hu (Xavier) <xavier.huwei@huawei.com>
Rx offload flag `DEV_RX_OFFLOAD_RSS_HASH` which can be used to
enable/disable PMDs write to `rte_mbuf:#️⃣:rss`. The hns3 PMD driver
already can notify the validity of `rte_mbuf:#️⃣rss` to the
application by enabling `PKT_RX_RSS_HASH` flag in `rte_mbuf::ol_flags`.
Fixes: 19a3ca4c99 ("net/hns3: add start/stop and configure operations")
Fixes: c37ca66f2b ("net/hns3: support RSS")
Cc: stable@dpdk.org
Signed-off-by: Lijun Ou <oulijun@huawei.com>
Signed-off-by: Wei Hu (Xavier) <xavier.huwei@huawei.com>
Currently, when running testpmd application based on hns3 network engine
with csum fwd mode by "set fwd csum" command in the prompt line, sending
42 consecutive bytes of ARP packets to network port with packets
generator. But in fact hardware can't send the ARP packets and the
related logs as below:
"Preparing packet burst to failed: Invalid argument"
The hardware doesn't support transmit packets less than 60 bytes, and in
the '.tx_pkt_burst' ops implementation function named hns3_xmit_pkts
appending operation has been added for less than 60 bytes packets. So
the interception needs to be removed in the '.tx_pkt_prepare' ops
implementation function named hns3_prep_pkts.
Fixes: de620754a1 ("net/hns3: fix sending packets less than 60 bytes")
Fixes: bba6366983 ("net/hns3: support Rx/Tx and related operations")
Cc: stable@dpdk.org
Signed-off-by: Wei Hu (Xavier) <xavier.huwei@huawei.com>
Signed-off-by: Hao Chen <chenhao164@huawei.com>
Signed-off-by: Chengchang Tang <tangchengchang@huawei.com>
Currently, promiscuous mode configuration are not cleared during
uninstallation based on hns3 PF device. The residual entries may cause
unnecessary bandwidth usage.
So, we need clear the PF's promisc mode status during the uninit.
Fixes: a45fd0aa0e ("net/hns3: fix Rx queue search with broadcast packet")
Fixes: d51867db65 ("net/hns3: add initialization")
Cc: stable@dpdk.org
Signed-off-by: Chengchang Tang <tangchengchang@huawei.com>
Signed-off-by: Wei Hu (Xavier) <xavier.huwei@huawei.com>
In hns3 PMD driver, the vport id 0 denote PF, and the vport id 1 denote
the first VF device of the port.
This patch adds two macros named HNS3_PF_FUNC_ID and
HNS3_1ST_VF_FUNC_ID, and replaces this two numbers to improve code
readability.
Signed-off-by: Chengchang Tang <tangchengchang@huawei.com>
Signed-off-by: Wei Hu (Xavier) <xavier.huwei@huawei.com>
When VFs configure the rss rule by virtchnl, it need to set bit mask
into the field selector for the protocol, then PF got the configure
massage and parse the field selector to the corresponding protocol
field.
Fixes: 7be10c3004 ("net/iavf: add RSS configuration for VF")
Cc: stable@dpdk.org
Signed-off-by: Jeff Guo <jia.guo@intel.com>
Acked-by: Qi Zhang <qi.z.zhang@intel.com>
Tested-by: Zhiwei He <zhiwei.he@intel.com>
The action number for switch filter should be 1, any
other such as 0 or more than 1 is invalid.
Fixes: 3428c6b6ec ("net/ice: add action number check for switch")
Cc: stable@dpdk.org
Signed-off-by: Wei Zhao <wei.zhao1@intel.com>
Tested-by: Qimai Xiao <qimaix.xiao@intel.com>
Acked-by: Qi Zhang <qi.z.zhang@intel.com>
The driver type need to be checked in private API.
Fixes: 50556c8810 ("net/ixgbe: fix MACsec setting")
Cc: stable@dpdk.org
Signed-off-by: Guinan Sun <guinanx.sun@intel.com>
Acked-by: Kevin Traynor <ktraynor@redhat.com>
Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
When we flush flow FDIR, all queues are disabled for FDIR.
If FDIR rule is created again, then the flow list is empty,
as it is the first time to create rule after flush fdir filter,
so we need to enable FDIR for all queues.
And also, disable FDIR for queues should be done in function
i40e_flow_flush_fdir_filter().
Fixes: 1491f63c75 ("net/i40e: fix flush of flow director filter")
Fixes: 6ae9b2b5e8 ("net/i40e: cache flow director enable value in Rx queue")
Cc: stable@dpdk.org
Signed-off-by: Wei Zhao <wei.zhao1@intel.com>
Reviewed-by: Jeff Guo <jia.guo@intel.com>
Add invalid flow checking func in iavf generic flow to avoid the error
of "Cannot access memory at address 0xXXXXXX" occur.
When hash init, the default RSS rules would be added, while hash uninit,
the default RSS rules should be deleted. Add the missing part in the
hash uninit process.
Fixes: 5ea6142543 ("net/iavf: fix VF reset for RSS")
Fixes: ff2d0c345c ("net/iavf: support generic flow API")
Signed-off-by: Jeff Guo <jia.guo@intel.com>
Acked-by: Qi Zhang <qi.z.zhang@intel.com>
Tested-by: Yuan Peng <yuan.peng@intel.com>
Base on HW, if a packet is split into multiple segments, the L2TAG
should only be valid on the last Rx descriptor. So fix it by setting
L2TAG into mbuf when processing the last split packet.
Fixes: c68a52b8b3 ("net/ice: support vector SSE in Rx")
Cc: stable@dpdk.org
Signed-off-by: Jeff Guo <jia.guo@intel.com>
Acked-by: Qi Zhang <qi.z.zhang@intel.com>
Base on HW, if a packet is split into multiple segments, the L2TAG
should only be valid on the last Rx descriptor. So fix it by setting
L2TAG into mbuf when processing the last split packet.
Fixes: 319c421f38 ("net/avf: enable SSE Rx Tx")
Cc: stable@dpdk.org
Signed-off-by: Jeff Guo <jia.guo@intel.com>
Acked-by: Qi Zhang <qi.z.zhang@intel.com>
Base on HW, if a packet is split into multiple segments, the L2TAG
should only be valid on the last Rx descriptor. So fix it by setting
L2TAG into mbuf when processing the last split packet.
Fixes: ca74903b75 ("net/i40e: extract non-x86 specific code from vector driver")
Cc: stable@dpdk.org
Signed-off-by: Jeff Guo <jia.guo@intel.com>
Acked-by: Qi Zhang <qi.z.zhang@intel.com>
Recent patch exposed a minor performance issue,
so it is reverted.
Fixes: d237d22fbe ("net/mlx5: prefer DevX API to create Rx objects")
Signed-off-by: Dekel Peled <dekelp@mellanox.com>
Acked-by: Matan Azrad <matan@mellanox.com>
When user requests a filter deletion only filter with
manual priority must be deleted. When an automatic filter has
the same specification, it must be skipped.
Fixes: 585c22edb2 ("net/sfc/base: handle manual and auto filter clashes in EF10")
Cc: stable@dpdk.org
Signed-off-by: Igor Romanov <igor.romanov@oktetlabs.ru>
Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>
When TSO MSS is smaller than 80, and the sum length of continuous
sge num is larger than a MSS, which may cause hardware failed,
so in this scenarios pmd driver should adjust the tso_segsz with
the same with the value of hardware supported.
Fixes: 076221c8fe ("net/hinic: add Rx/Tx")
Cc: stable@dpdk.org
Signed-off-by: Xiaoyun Wang <cloud.wangxiaoyun@huawei.com>
The elt_size field in mempool holds the size of one packet buffer.
It can be used to set the lpm_sizem1 field in rq context.
The lpb_sizem1 field in rq context is 12 bit, direct assignment
to it was causing overflow of value. Because of this errors
were observed while trying inline inbound with large packets.
This patch resolves the errors.
Fixes: 094fc8a3a1 ("net/octeontx2: add Rx queue setup and release")
Cc: stable@dpdk.org
Signed-off-by: Ankur Dwivedi <adwivedi@marvell.com>
Reviewed-by: Nithin Dabilpuram <ndabilpuram@marvell.com>
Reviewed-by: Jerin Jacob <jerinj@marvell.com>
When a queue is started after deferred_start, then increment raw_prod
irrespective of new mbuf is allocated or old mbufs are used.
Fixes: d256c73c11 ("net/bnxt: fix memory leak during queue restart")
Cc: stable@dpdk.org
Signed-off-by: Rahul Gupta <rahul.gupta@broadcom.com>
Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com>
Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
FDIR ID parsing will not be handled correctly after queue reconfigured,
enable FDIR ID parsing per Q regardless of fdir_ref_cnt to fix it.
Fixes: f71dbf852d ("net/iavf: add flow director enabled switch value")
Signed-off-by: Leyi Rong <leyi.rong@intel.com>
Acked-by: Xiaolong Ye <xiaolong.ye@intel.com>
Use a common function when selecting the proper word and mask match for
a tunnel type when programming switch rules.
Store switch recipe field mask as little endian, which avoids needing to
convert back to big endian after reading recipe from FW.
Obtain word mask from FW recipe.
Fix word matching element and index pairing.
Fixes: fed0c5ca5f ("net/ice/base: support programming a new switch recipe")
Cc: stable@dpdk.org
Signed-off-by: Dan Nowlin <dan.nowlin@intel.com>
Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
Acked-by: Qiming Yang <qiming.yang@intel.com>
GCC 10 is detecting the enum mismatch when assigning UDP variables
with MUSDK constants for TCP.
drivers/net/mvpp2/mrvl_flow.c:2521:47: error: implicit conversion
from 'enum mv_net_tcp_fields' to 'enum mv_net_udp_fields'
[-Werror=enum-conversion]
An assigned field is also fixed from "tcp" to "udp".
Fixes: 7235341d75 ("net/mrvl: support classifier")
Cc: stable@dpdk.org
Signed-off-by: Thomas Monjalon <thomas@monjalon.net>
Acked-by: Liron Himi <lironh@marvell.com>
Acked-by: David Marchand <david.marchand@redhat.com>
Acked-by: Kevin Traynor <ktraynor@redhat.com>
Since there are some default rss configure in kernel PF/VF but not DPDK
IAVF, if these configurations be modified by VF and then VF reset, this
default rss configurations can not be reset to default by IAVF. So need
to add default rss set in IAVF hash initial process.
Fixes: 7be10c3004 ("net/iavf: add RSS configuration for VF")
Cc: stable@dpdk.org
Signed-off-by: Jeff Guo <jia.guo@intel.com>
Acked-by: Qi Zhang <qi.z.zhang@intel.com>
Tested-by: Zhiwei He <zhiwei.he@intel.com>
There should have different behavior in queue start fail and stop fail
case. When queue start fail, all the next actions should be terminated
and then started queues should be cleared. But for queue stop stage, one
queue stop fail should not end other queues stop. This patch fixed that
issue in PF and VF.
Fixes: b6583ee402 ("i40e: full VMDQ pools support")
Fixes: 3f6a696f10 ("i40evf: queue start and stop")
Cc: stable@dpdk.org
Signed-off-by: Qiming Yang <qiming.yang@intel.com>
Acked-by: Xiaolong Ye <xiaolong.ye@intel.com>
All supported pattern for GTPU include extend header:
pattern_eth_ipv4_gtpu_eh_ipv4
pattern_eth_ipv4_gtpu_eh_ipv4_udp
pattern_eth_ipv4_gtpu_eh_ipv4_tcp
So the RSS rule should only take effect on GTPU packet that contains
extend header. The patch fix above issue and also allow inner l4 port
as input set.
Fixes: c08a72c79c ("net/ice: fix pattern name of GTPU with extension header")
Cc: stable@dpdk.org
Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
Acked-by: Qiming Yang <qiming.yang@intel.com>
Tested-by: Simei Su <simei.su@intel.com>
In i40e PMD code of function i40e_res_pool_free(), if valid_entry
is freed by "rte_free(valid_entry);" in the code, then the following
code for pool update may still use the wild pointer "valid_entry"
for pool info update. It seems has the risk of core dump for
using wild pointer operation, we should avoid this risk.
Fixes: 4861cde461 ("i40e: new poll mode driver")
Cc: stable@dpdk.org
Signed-off-by: Wei Zhao <wei.zhao1@intel.com>
Reviewed-by: Jeff Guo <jia.guo@intel.com>
Reviewed-by: Xiaolong Ye <xiaolong.ye@intel.com>
Add device arguments to lock NPA aura and pool contexts in NDC cache.
The device args take hexadecimal bitmask where each bit represent the
corresponding aura/pool id.
Example:
-w 0002:02:00.0,npa_lock_mask=0xf // Lock first 4 aura/pool ctx
Signed-off-by: Pavan Nikhilesh <pbhagavatula@marvell.com>
Acked-by: Jerin Jacob <jerinj@marvell.com>
Removed the typing error in doc/guides/eventdevs/index.rst,
drivers/net/mlx5/mlx5.c and in lib/librte_vhost/rte_vhost.h
Bugzilla ID: 477
Fixes: 0857b94211 ("doc: add event device and software eventdev")
Fixes: 039253166a ("vhost: add device op when notification to guest is sent")
Fixes: ad74bc6195 ("net/mlx5: support multiport IB device during probing")
Cc: stable@dpdk.org
Signed-off-by: Muhammad Bilal <m.bilal@emumba.com>
Currently rte_mcp_ptr_list is being shared as a variable
across libs. This is only used in control path.
This patch change it to a exported function based access.
Signed-off-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Acked-by: Ray Kinsella <mdr@ashroe.eu>
This patch changes the export of fman port config
as function call instead of direct variable access.
This is in control path, so it will not impact perf.
Signed-off-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Acked-by: Ray Kinsella <mdr@ashroe.eu>
This is to reduce the number of variables getting exposed
from the dpaa bus. They are not required to be in bus.
Signed-off-by: Hemant Agrawal <hemant.agrawal@nxp.com>
This patch moves the internal symbols to INTERNAL sections
so that any change in them is not reported as ABI breakage.
Signed-off-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Acked-by: Ray Kinsella <mdr@ashroe.eu>
Clarify Intel copyright and update the date to 2020.
Fixes: 8cb7c57d9b ("net/igc: support device initialization")
Cc: stable@dpdk.org
Signed-off-by: Xiaoyun Li <xiaoyun.li@intel.com>
Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
Clarify Intel copyright and update the date to 2020.
Fixes: 547be3f01f ("net/i40e/base: replace license text with SPDX tag")
Cc: stable@dpdk.org
Signed-off-by: Xiaoyun Li <xiaoyun.li@intel.com>
Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
Clarify Intel copyright and update the date to 2020.
Fixes: 9db3087f4f ("net/ixgbe/base: update the license")
Cc: stable@dpdk.org
Signed-off-by: Xiaoyun Li <xiaoyun.li@intel.com>
Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
After memory optimization, the organization of some resources are
changed from pointer based LIST to the index based ILIST. A lot of
code parts are touched due to such change.
Some static code checking and analysis tool will complain and raise
a false warning on the uninitialized value using. E.g. in the port
action registering function, the stack variable will be used as the
right value with some uninitialized field to initialize variable
allocated from heap. But indeed, it is not an error because all the
fields set with the uninitialized value will be overwritten in the
following code part and the macros. All the fields will be used as
the left value explicitly.
It makes no sense to clear the stack variable to 0 in this case,
and the extra memset will introduce some cycles overhead. It just
needs to ignore the false warning from the tool, if any.
Fixes: f3faf9ea11 ("net/mlx5: convert port id action to indexed")
Signed-off-by: Bing Zhao <bingz@mellanox.com>
Reviewed-by: Matan Azrad <matan@mellanox.com>
Reviewed-by: Suanming Mou <suanmingm@mellanox.com>
Acked-by: Viacheslav Ovsiienko <viacheslavo@mellanox.com>
After memory optimization, some action object handles are changed to
index to save the overhead. Assertion in debug mode will be helpful
for trouble shooting.
In the current implementation, only one port action is supported in
switchdev mode for one device flow. In debug mode, an assertion will
be used to check the if the port action is none, and it should
locate before the port action resource registration but not after
it. The action index in the handle should be 0 before registration.
Or else it will always cause a failure because the port action is
registered and the index is not 0.
Fixes: f3faf9ea11 ("net/mlx5: convert port id action to indexed")
Signed-off-by: Bing Zhao <bingz@mellanox.com>
Reviewed-by: Matan Azrad <matan@mellanox.com>
Reviewed-by: Suanming Mou <suanmingm@mellanox.com>
Acked-by: Viacheslav Ovsiienko <viacheslavo@mellanox.com>
Currently, the shared counter search uses the wrong nested index which
is used by the pool index. The incorrect nested index using causes the
search go to incorrect counter pool is not existed.
Add the counter index to fix the incorrect nested use case.
Fixes: 4001d7ad26 ("net/mlx5: change Direct Verbs counter to indexed")
Signed-off-by: Suanming Mou <suanmingm@mellanox.com>
Acked-by: Viacheslav Ovsiienko <viacheslavo@mellanox.com>
Acked-by: Matan Azrad <matan@mellanox.com>
The doorbell record is organized with page and bitmap. When some new
doorbell needs to be associated with a queue, the bit will be set
in the bitmap to indicate the corresponding doorbell occupied. A
counter is used to record the number of doorbell occupied to speed
up the searching.
If the number reaches the maximal value of a pre-defined number of a
page, a new page will be allocated. If not, then the bitmap will be
checked to find a free one.
The LSHIFT and OR (AND NOT) operations are used to update the bitmap
of a page. But 1 will be treated as a signed integer when compiling.
When the shift number is 31, the shifted value will be considered as
negative. Then a wrong extension will be done when setting it to a
64-bits variable. All the upper 32-bits will be set to 1 by such
extension.
Then a wrong offset value will be calculated because of this. The
next 64 bits will be also treated as the bitmap and get corrupted
through the bit set operation.
The immediate value 1 needs to be used as 64 bits width explicitly.
Fixes: 21cae8580f ("net/mlx5: allocate door-bells via DevX")
Cc: stable@dpdk.org
Signed-off-by: Bing Zhao <bingz@mellanox.com>
Acked-by: Viacheslav Ovsiienko <viacheslavo@mellanox.com>
Acked-by: Matan Azrad <matan@mellanox.com>
When create the Verbs flows with counter, randomly SEGSEV will also
comes. The reason is that the counter pool memory is not allocated
sufficiently and initialized correctly in Verbs case.
As the mlx5_flow_counter array member is moved out of the counter pool
struct, the counter pool memory layout currently contain implicitly
with mlx5_flow_counter, mlx5_age_param(if the pool is an age pool),
mlx5_flow_counter_ext(if the pool is a none batch pool). When allocate
the pool memory, the pool size should be calculated based on the pool
type accordingly.
Currently, for Verbs counter pool, both mlx5_flow_counter and
mlx5_flow_counter_ext need to be taken into account in the pool size.
And the pool type should also be initialized as CNT_POOL_TYPE_EXT.
This patch add the missing size and type for the Verbs counter pool.
Fixes: 8d93c830e4 ("net/mlx5: modify ext-counter memory allocation")
Signed-off-by: Suanming Mou <suanmingm@mellanox.com>
Acked-by: Matan Azrad <matan@mellanox.com>
CVL kernel PF configures all reserved queues for VF, including
Rx queue RXDID. The number of reserved queues is the maximum
between Tx and Rx queues. If the number of the enabled Rx queues
is less than that of reserved queues, required RXDID will only
be set for those enabled, but default value (0) is set for others.
However, RXDID 0 (legacy 16byte descriptor) is not supported now,
PF will return error when configuring those disabled VF queues.
In this patch, required RXDID is set for all reserved Rx queues,
no matter enabled or not. In this way, PF will configure Rx
queues correctly without reporting error.
Fixes: b8b4c54ef9 ("net/iavf: support flexible Rx descriptor in normal path")
Signed-off-by: Ting Xu <ting.xu@intel.com>
Tested-by: Xiaoxiao Zeng <xiaoxiaox.zeng@intel.com>
Acked-by: Xiaolong Ye <xiaolong.ye@intel.com>
This patch fixes the issue that the queue region does not
take effect due to incorrectly setting the flow type.
Fixes: ecad87d223 ("net/i40e: move RSS to flow API")
Cc: stable@dpdk.org
Signed-off-by: Shougang Wang <shougangx.wang@intel.com>
Reviewed-by: Jeff Guo <jia.guo@intel.com>
Tested-by: Hailin Xu <hailinx.xu@intel.com>
Tested-by: Lunyuan Cui <lunyuanx.cui@intel.com>
This patch adds support for FDIR passthrough mode. When FDIR rule
hits, FDIR just forward this packet to the next stage filter.
Fixes: 55daca4e45 ("net/ice/base: change function to static")
Signed-off-by: Simei Su <simei.su@intel.com>
Tested-by: Xiaoxiao Zeng <xiaoxiaox.zeng@intel.com>
Acked-by: Qi Zhang <qi.z.zhang@intel.com>
Fix to allocate bnxt ulp context when TRUFLOW is enabled.
This patch reduces the size of struct bnxt.
Fixes: 313ac35ac7 ("net/bnxt: support ULP session manager init")
Signed-off-by: Mike Baucom <michael.baucom@broadcom.com>
Reviewed-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com>
Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
We are maintaining MAC address twice which is unnecessary.
This patch removes the copy which helps reduce size of bnxt structure.
Fixes: b7778e8a1c ("net/bnxt: refactor to properly allocate resources for PF/VF")
Cc: stable@dpdk.org
Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com>
Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
There is no need to maintain local copy.
This helps reduce the size of the bnxt structure.
Fixes: fcc0aa1edc ("net/bnxt: add RSS hash configuration")
Cc: stable@dpdk.org
Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com>
Consolidate flow stat related structs for performance improvement.
The intention of this patch is to reduce the size struct bnxt which
had grown because of recent changes and was impacting performance.
Fixes: 02a95625fe ("net/bnxt: add flow stats in extended stats")
Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com>
Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
Log the command sequence number to aid debug in case of a
FW command timeout.
Fixes: 804e746c7b ("net/bnxt: add hardware resource manager init code")
Cc: stable@dpdk.org
Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com>
This patch fixes deallocation of all fastpath resources unconditionally,
when re-configuring the device. When re-allocating resources PMD depends
on application to explicitly setup the Rx/Tx queue.
Deallocation of all the resources is only required if the Rx/Tx queue
configuration changes. For other scenarios like KNI MTU change we'd keep
existing configuration.
Fixes: 8de0c42019 ("net/qede: fix odd number of queues usage in 100G mode")
Fixes: dd28bc8c6e ("net/qede: fix VF port creation sequence")
Cc: stable@dpdk.org
Signed-off-by: Rasesh Mody <rmody@marvell.com>
Signed-off-by: Igor Russkikh <irusskikh@marvell.com>
Fix to assign dummy Rx/Tx handlers in dev_stop.
For MTU set, assignment of the appropriate Rx/Tx handlers will be
handled by dev_start/dev_stop.
Fixes: 81f8804992 ("net/qede: enhance Rx CPU utilization")
Fixes: 8de0c42019 ("net/qede: fix odd number of queues usage in 100G mode")
Cc: stable@dpdk.org
Signed-off-by: Rasesh Mody <rmody@marvell.com>
Signed-off-by: Igor Russkikh <irusskikh@marvell.com>
Event notification should be disabled after virtqueue initialization and
enabled by calling rte_eth_dev_rx_intr_enable later. When virtio user
device reconnecting to vhost, virtqueue_disable_intr should be called to
disable event notification.
Fixes: 6ebbf4109f ("net/virtio-user: fix packed ring server mode")
Cc: stable@dpdk.org
Signed-off-by: Marvin Liu <yong.liu@intel.com>
Acked-by: Xiao Wang <xiao.w.wang@intel.com>
Reviewed-by: Maxime Coquelin <maxime.coquelin@redhat.com>
Runtime checks for vectorized datapath are organized into device
configuration function. Remove duplicated check in device init
function.
Signed-off-by: Marvin Liu <yong.liu@intel.com>
Reviewed-by: Maxime Coquelin <maxime.coquelin@redhat.com>
Before updating RSS indirection table, firstly determine whether
rq num in RETA table is legal, if it is invalid(such as exceeding
the maximum rxq num), driver will not update hw indirection
table and return fail.
Signed-off-by: Xiaoyun Wang <cloud.wangxiaoyun@huawei.com>
The patch modifies the comments of structures or functions, and adds
space for comments, removes extra empty lines and adjusts the
print level for VF branches in one sdk interface.
Signed-off-by: Xiaoyun Wang <cloud.wangxiaoyun@huawei.com>
When copy a mbuf to a new dst_mbuf, the pkt_len member of
dst_mbuf needs to be updated.
Fixes: 076221c8fe ("net/hinic: add Rx/Tx")
Cc: stable@dpdk.org
Signed-off-by: Xiaoyun Wang <cloud.wangxiaoyun@huawei.com>
Adds tx_queues and rx_queues non-null judgment before free tx or
rx resources, because some app may set tx_queues or rx_queues to
be null before call free resource interfaces, which may cause
a segfault.
Fixes: 64727024d2 ("net/hinic: add device initialization")
Cc: stable@dpdk.org
Signed-off-by: Xiaoyun Wang <cloud.wangxiaoyun@huawei.com>
The VMXNet3 protocol has a start-of-packet (SOP) and end-of-packet (EOP)
marker. If there was a bug where mbuf arrived without SOP the code that
chains the mbuf would dereference a null pointer.
Also, record any mbuf's dropped in statistics.
Although did the initial code no longer have access to VMware.
Compile tested only!
Coverity issue: 124563
Fixes: 8ee787ce80 ("vmxnet3: remove asserts that confuse coverity")
Cc: stable@dpdk.org
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
Acked-by: Yong Wang <yongwang@vmware.com>
Previous patch added support of VLAN item without VLAN ID value,
i.e. using wildcard VLAN item, to match VLAN with any VLAN ID.
The implication on VLAN actions was not taken into consideration.
VLAN actions (e.g. push vlan) use the VLAN ID value in the VLAN item,
and expect it to be valid.
This patch updates function flow_dev_get_vlan_info_from_items() to
check the VLAN item contents before trying to use it.
Fixes: 92818d839e ("net/mlx5: fix match on empty VLAN item in DV mode")
Cc: stable@dpdk.org
Signed-off-by: Dekel Peled <dekelp@mellanox.com>
Acked-by: Viacheslav Ovsiienko <viacheslavo@mellanox.com>
The design of counter container resize used double buffer algorithm in
order to synchronize between the query thread to the control thread.
When the control thread detected resize need, it created new bigger
buffer for the counter pools in a new container and change the container
index atomically.
In case the query thread had not detect the previous resize before a new
one need was detected by the control thread, the control thread returned
EAGAIN to the flow creation API used a COUNT action.
The rte_flow API doesn't allow unblocked commands and doesn't expect to
get EAGAIN error type.
So, when a lot of flows were created between 2 different periodic
queries, 2 different resizes might try to be created and caused EAGAIN
error.
This behavior may blame flow creations.
Change the synchronization way to use lock instead of double buffer
algorithm.
The critical section of this lock is very small, so flow insertion
rate should not be decreased.
Fixes: ebbac312e4 ("net/mlx5: resize a full counter container")
Cc: stable@dpdk.org
Signed-off-by: Matan Azrad <matan@mellanox.com>
Acked-by: Viacheslav Ovsiienko <viacheslavo@mellanox.com>
Move link state enable/disable to dev_start() and dev_stop()
respectively. This will ensure when devices are stopped,
link status will be appropriately shown as down.
Fixes: dd28bc8c6e ("net/qede: fix VF port creation sequence")
Cc: stable@dpdk.org
Signed-off-by: Shahed Shaikh <shshaikh@marvell.com>
Signed-off-by: Rasesh Mody <rmody@marvell.com>
Signed-off-by: Igor Russkikh <irusskikh@marvell.com>
This patch fixes the issue that failed to create an RSS rule with type
L2-payload.
Fixes: feaae285b3 ("net/i40e: support hash configuration in RSS flow")
Signed-off-by: Shougang Wang <shougangx.wang@intel.com>
Tested-by: Hailin Xu <hailinx.xu@intel.com>
Reviewed-by: Jeff Guo <jia.guo@intel.com>
When configure RSS rule, the etherdev rss hash function type should be
mapped to the corresponding virtchnl rss algorithm type.
Fixes: 7be10c3004 ("net/iavf: add RSS configuration for VF")
Signed-off-by: Jeff Guo <jia.guo@intel.com>
Acked-by: Qi Zhang <qi.z.zhang@intel.com>
When the driver is loaded on a 200G NIC, the port speed is not
displayed correctly. Parse the 200G speed before displaying it.
Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com>
Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
Relaxed ordering is a PCI optimization that enables reordering
reads/writes in order to improve performance.
Relaxed ordering was enabled for all processors causing
a degradation in performance in Haswell and Broadwell processors
that don't support this optimization.
In order to avoid that we check if the processor is Haswell
or Broadwell and if so we disable relaxed ordering.
Signed-off-by: Shiri Kuzin <shirik@mellanox.com>
Acked-by: Matan Azrad <matan@mellanox.com>
In order to improve performance relaxed ordering was enabled
when creating count object using Devx.
Currently rte enables this optimization by default when using
Devx.
This causes an issue when using firmware that does not have this
capability causing a count object failure.
In order to fix this issue a check of firmware capabilities was
added before enabling relaxed ordering.
Fixes: 53ac93f71a ("net/mlx5: create relaxed ordering memory regions")
Signed-off-by: Shiri Kuzin <shirik@mellanox.com>
Acked-by: Matan Azrad <matan@mellanox.com>
This patch fixes the issue that the flow which both source MAC mask
and destination MAC mask are all zeros can not be created.
Fixes: ea0c22fd82 ("net/i40e: enable MAC address as flow director input set")
Signed-off-by: Shougang Wang <shougangx.wang@intel.com>
Tested-by: Hailin Xu <hailinx.xu@intel.com>
Reviewed-by: Jeff Guo <jia.guo@intel.com>
The register autoneg can't be updated synchronously with flow control
mode setting in the state of port start, so NIC statistics error occurs.
AUTO_NEG Advt register should be updated by ixgbe_setup_fc() when
enabling flow control.
The patch fixes the issue.
Fixes: a524f550da ("net/ixgbe: fix flow control mode setting")
Cc: stable@dpdk.org
Signed-off-by: Guinan Sun <guinanx.sun@intel.com>
Reviewed-by: Wei Zhao <wei.zhao1@intel.com>
Currently, DevX API is used to create Rx objects (RQ, RQT, TIR) only
if LRO or hairpin features are enabled on this RQ.
This patch uses DevX API by default, if DevX is supported and can be
used. Otherwise, Verbs API is used.
Signed-off-by: Dekel Peled <dekelp@mellanox.com>
Acked-by: Matan Azrad <matan@mellanox.com>
Fix coverity defects of unused value.
Coverity issue: 357745, 357769
Fixes: b8b4c54ef9 ("net/iavf: support flexible Rx descriptor in normal path")
Signed-off-by: Leyi Rong <leyi.rong@intel.com>
Acked-by: Xiaolong Ye <xiaolong.ye@intel.com>
Fix some out-of-bounds memory issues, they may lead to wrong results
or affect application stability.
Coverity issue: 357759, 357713
Fixes: bd3fcf0d0f ("net/igc: support RSS")
Signed-off-by: Alvin Zhang <alvinx.zhang@intel.com>
Reviewed-by: Jeff Guo <jia.guo@intel.com>
Update max packet length supported by Inline IPsec Rx to
NIX_MAX_FRS
Signed-off-by: Anoob Joseph <anoobj@marvell.com>
Signed-off-by: Tejasree Kondoj <ktejasree@marvell.com>
The AVX512 packed ring datapath selection was only done
at build time, but it should also be checked at runtime
that the CPU supports it.
This patch add a CPU flags check so that non-vectorized
path is selected at runtime if AVX512 is not supported.
Also in meson build enable vectorization only for relevant file, not for
all driver.
Fixes: ccb10995c2 ("net/virtio: add election for vectorized path")
Signed-off-by: Maxime Coquelin <maxime.coquelin@redhat.com>
Signed-off-by: Ferruh Yigit <ferruh.yigit@intel.com>
Acked-by: Maxime Coquelin <maxime.coquelin@redhat.com>
If hardware mode for this card does not support nic,
the network features for this card will not be supported
and driver will be initialized failed.
Signed-off-by: Xiaoyun Wang <cloud.wangxiaoyun@huawei.com>
This patch adds to MLX5 PMD the support of matching on
GTP header item v_pt_rsv_flags.
This item is contained in 1 byte of the format:
-------------------------------------------
| bit | 0 - 2 | 3 | 4 | 5 | 6 | 7 |
|-----------------------------------------|
| value | Version | PT | Res | E | S | PN |
-------------------------------------------
Matching is supported only for GTP flags E, S, PN.
Therefore values 0 to 7 are supported.
Mask must be set accordingly:
... gtp v_pt_rsv_flags is 1 v_pt_rsv_flags mask 0x07 ...
Signed-off-by: Dekel Peled <dekelp@mellanox.com>
Acked-by: Matan Azrad <matan@mellanox.com>
Program received signal SIGSEGV, Segmentation fault.
0x00000000008ef7c4 in mlx5_tx_queue_release (dpdk_txq=0x17ce01680) at
drivers/net/mlx5/mlx5_txq.c:302
301 mlx5_txq_release(ETH_DEV(priv), i);
302 DRV_LOG(DEBUG, "port %u removing Tx queue %u from list",
303 PORT_ID(priv), txq->idx);
The problem is txq is freed inside the mlx5_txq_release() function
and no longer valid in the debug log right after this invocation.
Move the debug log before the mlx5_txq_release() function to fix this.
Fixes: a6d83b6a92 ("net/mlx5: standardize on negative errno values")
Cc: stable@dpdk.org
Signed-off-by: Alexander Kozyrev <akozyrev@mellanox.com>
Acked-by: Matan Azrad <matan@mellanox.com>
The function mlx4_drop_get() creates pointer to a struct mlx4_drop and
if needed allocates by rte_malloc.
If the allocation is failed the function goes to label “error”, and
there does dereference to a null pointer.
Skip resources cleaning when the memory allocation is failed.
Coverity issue: 146206
Coverity issue: 146146
Fixes: d3a7e09234 ("net/mlx4: allocate drop flow resources on demand")
Cc: stable@dpdk.org
Signed-off-by: Michael Baum <michaelba@mellanox.com>
Acked-by: Matan Azrad <matan@mellanox.com>
The mlx5_flow_get_reg_id() function translates tag ID to register
from the registers that are supported and available for use. The
user does not know which register is available at a time and therefore
there is an array that represents mapping to the available registers.
Usually the free registers are continuous in the flow_mreg_c array but
sometimes the mtr_color_reg register is between them and it must be
skipped and the next register returned, in which case the function
returns the mapping of the next entity in the array.
When the function reads from the next entity in the array, it does not
check whether such an entity exists and in some situation invalid access
to memory occurs beyond the array boundaries.
So, when all the registers are valid from HW perspective and the meter
color register is not the default, the tag id 5 causes an out of bound
access.
Validate registers availability when meter color register is not the
default.
Coverity issue: 146355
Fixes: 792e749e92 ("net/mlx5: fix register usage in meter")
Cc: stable@dpdk.org
Signed-off-by: Michael Baum <michaelba@mellanox.com>
Acked-by: Matan Azrad <matan@mellanox.com>
When creating flow rule with zero specs it will cause
matching all UDP packets like following:
eth / ipv4 / udp / vxlan / end
Such rule will match all udp packets.
This change the behavior to match the dv flow engine
which will automatically set the match on relative
outer UDP port if the user didn't specify any.
Fixes: 84c406e745 ("net/mlx5: add flow translate function")
Cc: stable@dpdk.org
Signed-off-by: Raslan Darawsheh <rasland@mellanox.com>
Acked-by: Matan Azrad <matan@mellanox.com>
In existing implementation, using wild card VLAN item is not allowed.
A VLAN item in flow pattern must include VLAN ID (vid) value.
This obligation contradict the flow API specification [1].
This patch updates the VLAN item validation and translation, to allow
wild card VLAN item, without VLAN ID value.
User guide and release notes are updated accordingly.
[1]
commit 40513808b165 ("doc: refine ethernet and VLAN flow rule items")
Fixes: 00f75a4057 ("net/mlx5: fix VLAN match for DV mode")
Cc: stable@dpdk.org
Signed-off-by: Dekel Peled <dekelp@mellanox.com>
Acked-by: Viacheslav Ovsiienko <viacheslavo@mellanox.com>
When a net_ring device is allocated, its device pointer is not set
before calling rte_eth_dev_probing_finish, which is incorrect.
The following:
commit 96cb195211 ("net/ring: use EAL APIs in PMD specific API")
commit a6992e9610 ("net/ring: set ethernet device field")
already fixed the same issue in 17.08, which was fine at the time.
Adding the hook rte_eth_dev_probing_finish() however created this bug,
as the eth_dev exposed when this hook is executed is expected to be
complete.
Remove the prior attempts to fix the issue in rte_pmd_ring_probe() and
write the pointer properly in do_eth_dev_ring_create().
Fixes: fbe90cdd77 ("ethdev: add probing finish function")
Cc: stable@dpdk.org
Signed-off-by: Gaetan Rivet <grive@u256.net>
Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
When trying to compile with glibc < 2.24 that doesn't
support SOL_NETLINK it will cause compilation failure:
drivers/net/tap/tap_netlink.c:70:17: error:
'SOL_NETLINK' undeclared (first use in this function)
setsockopt(fd, SOL_NETLINK, NETLINK_EXT_ACK, &one, sizeof(one));
The glibc commits adds the SOL_NETLINK support:
https://github.com/bminor/glibc/commit/f9b437d5efce93800b51ad2a437c8b1c9
Fixes: 647909bcf3 ("net/tap: use netlink extended ack support")
Signed-off-by: Raslan Darawsheh <rasland@mellanox.com>
Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
NIX exposes NIX_AF_SMQ(0..511)_CFG to configure minimum length
of the packet which is being used for zero padding if packet is
less than configured value.
Setting it to default minimum length i.e. 60 bytes.
Fixes: ec8ddd4fb1 ("net/octeontx2: restructure TM helper functions")
Signed-off-by: Sunil Kumar Kori <skori@marvell.com>
Acked-by: Jerin Jacob <jerinj@marvell.com>
Add Rx descriptor limit for number of segments per MTU.
PMD doesn't support Jumbo Rx scatter gather hence set 1 segment per
MTU. Some applications can adjust mbuf_size based on this value.
For others PMD detects the condition where Rx packet length cannot
be held by configured mbuf size and logs the message.
Signed-off-by: Rasesh Mody <rmody@marvell.com>
Signed-off-by: Igor Russkikh <irusskikh@marvell.com>
Some log will always be printed when the device is initialized
and shut down, which will mislead users and should be removed.
Fixes: 746664d546 ("net/igc: support flow API")
Signed-off-by: Alvin Zhang <alvinx.zhang@intel.com>
Acked-by: Xiaolong Ye <xiaolong.ye@intel.com>
The dummy packet for pppoe-ipv6 has an error, we should
use 0x3b for next header in ipv6 header in the dummy packet,
or some case can not be download, such as:
"eth / pppoes seid is 3 / pppoe_proto_id is 0x0057 /
end actions vf id 1 / end"
Fixes: 55d61fb27a ("net/ice/base: add PPPoE IPv6 dummy packet")
Signed-off-by: Wei Zhao <wei.zhao1@intel.com>
Tested-by: Nannan Lu <nannan.lu@intel.com>
Acked-by: Qi Zhang <qi.z.zhang@intel.com>
When processing a rte flow, such as creating a parse engine, or
creating or destroying a RSS rule, if they are failed, they all
need to construct the flow error structure before return the error
message back to app. If not so, it will cause app crash when
app printing the message out of a flow error.
Fixes: 7be10c3004 ("net/iavf: add RSS configuration for VF")
Fixes: ff2d0c345c ("net/iavf: support generic flow API")
Signed-off-by: Jeff Guo <jia.guo@intel.com>
Acked-by: Qi Zhang <qi.z.zhang@intel.com>
In ixgbe_dev_link_update_share(), if the media type is fiber and the
link is down, a flag (IXGBE_FLAG_NEED_LINK_CONFIG) is set. A callback
to ixgbe_dev_setup_link_thread_handler() is scheduled which should
try to set up the link and clear the flag afterwards. This flag works
as a guard variable between threads.
To avoid potential race condition between threads, set the
IXGBE_FLAG_NEED_LINK_CONFIG flag only when there is no link thread
running.
Bugzilla ID: 388
Fixes: 819d0d1d57 ("net/ixgbe: fix blocking system events")
Cc: stable@dpdk.org
Signed-off-by: Phil Yang <phil.yang@arm.com>
Reviewed-by: Lijian Zhang <lijian.zhang@arm.com>
Acked-by: Xiaolong Ye <xiaolong.ye@intel.com>
Since some specific RSS hash type need to combine with the protocol
hash type when configure a RSS hash rule, so add the corresponding
input set to support these case for iavf hash.
Fixes: 7be10c3004 ("net/iavf: add RSS configuration for VF")
Signed-off-by: Jeff Guo <jia.guo@intel.com>
Acked-by: Qi Zhang <qi.z.zhang@intel.com>
When using a single TC, flow id is same as congestion group id.
However in case of multiple traffic classes, same flow id values
are used within each traffic class, which causes incorrect
traffic behavior e.g. halting of traffic.
This patches changes to use the cgid as the index which works
for single as well as multiple traffic classes.
Fixes: 13b856ac02 ("net/dpaa2: support taildrop on frame count basis")
Cc: stable@dpdk.org
Signed-off-by: Jun Yang <jun.yang@nxp.com>
Acked-by: Hemant Agrawal <hemant.agrawal@nxp.com>
This change adds DPDMUX support to bifurcate traffic on
the basis of UDP destination port.
Signed-off-by: Nipun Gupta <nipun.gupta@nxp.com>
Acked-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Update the portal allocation failure log to print the thread id
as well.
Signed-off-by: Nipun Gupta <nipun.gupta@nxp.com>
Acked-by: Hemant Agrawal <hemant.agrawal@nxp.com>
This patch adds default/preferred rx/tx params in dev info,
specially the advertised burst size.
Signed-off-by: Hemant Agrawal <hemant.agrawal@nxp.com>
When IOVA is physical address do not prefetch the annotation
of the next frame, as there is a cost involved there to convert
the physical address to virtual address.
Signed-off-by: Nipun Gupta <nipun.gupta@nxp.com>
Acked-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Fixed 10G port negotiation issue with another 10G/non 10G port.
When running testpmd with 10G interfaces on 10BaseT interface
on LS2088ARDB, the ports were showing link as down.
This was identified to be caused by the setting of link as down
during config.
Also, the line rate was not being updated in device link params,
thus having the incorrect link speed in status (as 0).
Fixes: c5acbb5ea2 ("net/dpaa2: support link status event")
Cc: stable@dpdk.org
Signed-off-by: Rohit Raj <rohit.raj@nxp.com>
Acked-by: Hemant Agrawal <hemant.agrawal@nxp.com>
IAVF FDIR needs to check program status from FD completion descriptors
for both successful and failed case, rather than only ask for completion
on error. This patch corrects completion report value so that the called
can determine whether to report completion on error or other cases.
Fixes: 55daca4e45 ("net/ice/base: change function to static")
Signed-off-by: Simei Su <simei.su@intel.com>
Acked-by: Qi Zhang <qi.z.zhang@intel.com>
bnxt_fw_version_get is not indicating the complete FW version.
Fix it to indicate complete version string.
Fixes: e2652b0a20 ("net/bnxt: support get FW version")
Cc: stable@dpdk.org
Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
Fix a buffer overrun issue spotted by coverity while accessing
the array ulp_device_params.
Note that the issue was observed in an internal Coverity scan.
Fixes: 313ac35ac7 ("net/bnxt: support ULP session manager init")
Signed-off-by: Kishore Padmanabha <kishore.padmanabha@broadcom.com>
Reviewed-by: Mike Baucom <michael.baucom@broadcom.com>
Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
Clang 6.0.0 will undefine function _mm512_maskz_set1_epi64 on i686
target. Fix it by replacing the function with _mm512_set4_epi64 when
doing 32-bit build.
Warning message during build:
../drivers/net/virtio/virtio_rxtx_packed_avx.c:385:19: warning:
implicit declaration of function '_mm512_maskz_set1_epi64' is invalid
in C99 [-Wimplicit-function-declaration]
Fixes: 77d66da838 ("net/virtio: add vectorized packed ring Rx")
Signed-off-by: Marvin Liu <yong.liu@intel.com>
Reviewed-by: Maxime Coquelin <maxime.coquelin@redhat.com>
When doing virtio device initialization, virtqueues will be reset in
server mode if ring type is packed. It will cause issue because queues
have been freed in the beginning of device initialization.
Fix this issue by checking whether device has been initialized before
reset. If device hasn't been initialized, there's no need to reset
queues.
Fixes: 6ebbf4109f ("net/virtio-user: fix packed ring server mode")
Cc: stable@dpdk.org
Signed-off-by: Marvin Liu <yong.liu@intel.com>
Reviewed-by: Maxime Coquelin <maxime.coquelin@redhat.com>
The current formulas to calculate the TQM slow path and fast path ring
context memory sizes are not quite correct. TQM slow path entry is
array index 0 of ctx->tqm_mem[]. The other array entries are for fast
path. Fix these sizes according to firmware spec. for 57500 and newer
chips.
Fixes: cc5e26b8ef ("net/bnxt: increase TQM entry allocation")
Cc: stable@dpdk.org
Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com>
Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
Newer firmware advertises the number of TQM rings to allocate
context memory for. Use the firmware specified value and fall back
to the old value derived from "bp->max_q" if it is not available.
Fixes: f8168ca0e6 ("net/bnxt: support thor controller")
Cc: stable@dpdk.org
Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com>
Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
In recent Linux kernels, there is support for extended acknowledgment
to netlink messages. This is quite useful for diagnosing errors
in configuration in the kernel with TAP.
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
Acked-by: Keith Wiles <keith.wiles@intel.com>
The tap_nl_recv() function does not need to use the full
complex recvmsg() system call, basic recv() will work here.
Ditto for tap_nl_send() full sendmsg is not needed.
Add logic to retry in case EINTR rather than forcing
error handling back in driver or worse to ethdev API.
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
Acked-by: Keith Wiles <keith.wiles@intel.com>
The TAP driver does not initialize all the elements of the rte_flow
structure. This can lead to crash in rte_flow_destroy.
(gdb) where
flow=0x100e99280, error=0x0)
at drivers/net/tap/tap_flow.c:1514
(gdb) p remote_flow
$1 = (struct rte_flow *) 0x6b6b6b6b6b6b6b6b
Which is here:
static int
tap_flow_destroy_pmd(struct pmd_internals *pmd,
struct rte_flow *flow,
struct rte_flow_error *error)
{
struct rte_flow *remote_flow = flow->remote_flow;
...
if (remote_flow) {
remote_flow->msg.nh.nlmsg_flags = NLM_F_REQUEST | NLM_F_ACK;
Simplest fix is to use rte_zmalloc() so remote_flow and other fields
are always set at zero.
Fixes: 2bc06869cd ("net/tap: add remote netdevice traffic capture")
Cc: stable@dpdk.org
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
The number of queues in queue group should be checked before
using it. This patch fixed the issue.
Fixes: 47d460d632 ("net/ice: rework switch filter")
Cc: stable@dpdk.org
Signed-off-by: Junyu Jiang <junyux.jiang@intel.com>
Tested-by: Qimai Xiao <qimaix.xiao@intel.com>
Acked-by: Qiming Yang <qiming.yang@intel.com>
This patch fixes issue that doesn't support mark only case.
Mark only action is equal to mark + passthru action.
Fixes: f5cafa961f ("net/ice: add flow director create and destroy")
Cc: stable@dpdk.org
Signed-off-by: Simei Su <simei.su@intel.com>
Acked-by: Qi Zhang <qi.z.zhang@intel.com>
DPDK does not implement interrupt mechanism on BSD,
so force NIC status synchronization.
Fixes: dc66e5fd01 ("net/ixgbe: improve link state check on VF")
Cc: stable@dpdk.org
Signed-off-by: Zhihong Peng <zhihongx.peng@intel.com>
Tested-by: Zhimin Huang <zhiminx.huang@intel.com>
Acked-by: Xiaolong Ye <xiaolong.ye@intel.com>
When we download a switch rule for ipv6 with esp payload
"eth / ipv6 / esp spi is 1 / end actions queue index 2 / end"
if we don't add bm bit set check for tun_type, then a packet of
ipv4 with esp payload
"sendp([Ether(dst="00:00:00:00:01:00")/IP(proto=50)/ESP(spi=1)/
("X"*480)], iface="ens5f0", count=10)"
Will also go to queue index 2. And also, we need to do tun_type
check, or the second rule of following can not be download because
of rejection from switch rule download function ice_aq_sw_rules().
"eth / ipv4 / esp spi is 1 / end actions queue index 5 / end"
"eth / ipv6 / esp spi is 1 / end actions queue index 2 / end"
Fixes: 4f11962fce ("net/ice/base: support AH ESP and NAT-T on switch")
Fixes: 99d8ba79ef ("net/ice/base: force switch to use different recipe")
Signed-off-by: Wei Zhao <wei.zhao1@intel.com>
Tested-by: Qi Fu <qi.fu@intel.com>
Acked-by: Qi Zhang <qi.z.zhang@intel.com>
If the PF driver does not support the new speed reporting capabilities
then use link_event instead of link_event_adv to get the speed.
Fixes: 48de41ca11 ("net/avf: enable link status update")
Cc: stable@dpdk.org
Signed-off-by: Alvin Zhang <alvinx.zhang@intel.com>
Acked-by: Beilei Xing <beilei.xing@intel.com>
We see a stack smashing as a result of defensive code missing. Once the
nb_pkts is less than RTE_BNXT_DESCS_PER_LOOP, it will be modified to
zero after doing a floor align, and we can not exit the following
receiving packets loop. And the buffers will be overwrite, then the
stack frame was ruined.
Fix the problem by adding defensive code, once the nb_pkts is zero, just
directly return with no packets.
Fixes: bc4a000f2f ("net/bnxt: implement SSE vector mode")
Cc: stable@dpdk.org
Signed-off-by: Linsi Yuan <yuanlinsi01@baidu.com>
Signed-off-by: Dongsheng Rong <rongdongsheng@baidu.com>
Acked-by: Lance Richardson <lance.richardson@broadcom.com>
Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
This fixes the problem where driver would not start if only
have a single Rx queue and multiple Txq. In that case, RSS
should stay disabled.
Fixes: 92d23a57ca ("net/netvsc: support configuring RSS parameters")
Cc: stable@dpdk.org
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
If number of tx queues is greater than the number of rx queues;
the driver ends up allocating more channels than rx queues.
The problem is that the RSS indirection table is programmed such
that some packets will end up on a channel that would never be
polled. The fix is to limit the RSS indirection table by number
of rx queues not channels.
Fixes: 92d23a57ca ("net/netvsc: support configuring RSS parameters")
Cc: stable@dpdk.org
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
With multiple channels, the primary channel may receive notification
that VF has been added or removed while secondary channel is in
process of doing receive or transmit. Resolve this race by converting
existing vf_lock to a reader/writer lock.
Users of lock (tx/rx/stats) acquire for read, and actions like
add/remove acquire it for write.
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
Since VF notifications are handled as VMBUS notifications on the
primary channel (and not as hotplug). The channel should be checked
before deciding to use VF for Rx or Tx.
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
Some PMD do not respect the eth_dev API when allocating their
rte_eth_dev. As a result, on device add event resulting from
rte_eth_dev_probing_finish() call, the eth_dev processed is incomplete.
The segfault is a good way to focus the developer on the issue, but does
not inspire confidence. Instead, warn the user of the error repeatedly.
The failsafe PMD can warn of the issue and continue. It will repeatedly
attempt to initialize the failed port and complain about it, which
should result in the same developer focus but with less crashing.
Signed-off-by: Gaetan Rivet <grive@u256.net>
Zero is a valid fd. The fd won't be closed thus leading fd leak,
when it is zero.
Also the service proxy is initialized at 0. This is assuming that all of
its fields are invalid at 0. The issue is that a file descriptor at 0 is
a valid one.
The value -1 is used as sentinel during cleanup. Initialize the RX proxy
file descriptor to -1.
Fixes: f234e5bd99 ("net/failsafe: register slaves Rx interrupts")
Fixes: 9e0360aebf ("net/failsafe: register as Rx interrupt mode")
Cc: stable@dpdk.org
Signed-off-by: Yunjian Wang <wangyunjian@huawei.com>
Signed-off-by: Gaetan Rivet <grive@u256.net>
Tested-by: Ali Alnubani <alialnu@mellanox.com>
Can be reproduced with "make EXTRA_CFLAGS='-O1'" command using
gcc (GCC) 9.3.1 20200408 (Red Hat 9.3.1-2)
Build error:
.../drivers/net/ena/ena_ethdev.c: In function ‘eth_ena_dev_init’:
.../drivers/net/ena/ena_ethdev.c:1815:20:
error: ‘wd_state’ may be used uninitialized in this function
[-Werror=maybe-uninitialized]
1815 | adapter->wd_state = wd_state;
| ~~~~~~~~~~~~~~~~~~^~~~~~~~~~
This looks like false positive, fixing by assigning initial value to
'wd_state' variable.
Signed-off-by: Ferruh Yigit <ferruh.yigit@intel.com>
Acked-by: Michal Krawczyk <mk@semihalf.com>
rte_pci_probe() is private to the PCI bus.
Clean the remaining references in the documentation and comments.
Fixes: c752998b5e ("pci: introduce library and driver")
Cc: stable@dpdk.org
Signed-off-by: David Marchand <david.marchand@redhat.com>
Reviewed-by: Gaetan Rivet <grive@u256.net>
Adding support for rx checksum offload. In case of wrong
checksum received (inner/outer l3/l4) it reports the
corresponding layer which has bad checksum. It also adds
rx burst function pointer hook for rx checksum offload to
event PMD.
Signed-off-by: Harman Kalra <hkalra@marvell.com>
Acked-by: Jerin Jacob <jerinj@marvell.com>
Adding macro based framework to hook dequeue/enqueue function
pointers to the appropriate function based on rx/tx offloads.
Signed-off-by: Harman Kalra <hkalra@marvell.com>
Acked-by: Jerin Jacob <jerinj@marvell.com>
The flow_verbs_translate() function accumulates hash fields while
iterating through the flow items (SRC_IPV4, DST_IPV4, SRC_IPV6,
DST_IPV6, SRC_PORT_TCP, DST_PORT_TCP, SRC_PORT_UDP, DST_PORT_UDP).
Before this commit the dev_flow handle structure was reused in each new
flow_verbs_translate() call, however the dev_flow->hash_fields variable
was not reset before each call. As a result hash_fields from previous
calls remained present in the current flow which lead to invalid
combinations (e.g. simultaneous IPv4 and IPv6 specs). This scenario
happens for example in the next flows sequence, when running in verbs
mode (dv_flow_en=0).
flow create 0 ingress group 0 pattern eth / ipv4 / end <rss actions>
flow create 0 ingress group 0 pattern eth / ipv6 / end <rss actions>
The fix is to reset dev_flow->hash_fields in flow_verbs_prepare().
Fixes: e7bfa3596a ("net/mlx5: separate the flow handle resource")
Signed-off-by: Ophir Munk <ophirmu@mellanox.com>
Acked-by: Matan Azrad <matan@mellanox.com>
The HW is optimized for IPv4/IPv6. For such cases avoid matching on
ethertype, and use ip_version field instead.
Signed-off-by: Eli Britstein <elibr@mellanox.com>
Acked-by: Viacheslav Ovsiienko <viacheslavo@mellanox.com>
Acked-by: Matan Azrad <matan@mellanox.com>
Introduce a helper function to set the ip_version match.
Signed-off-by: Eli Britstein <elibr@mellanox.com>
Acked-by: Viacheslav Ovsiienko <viacheslavo@mellanox.com>
Acked-by: Matan Azrad <matan@mellanox.com>
gcc 10.0.1 reports warnings when using mlx5_rte_flow enums
with rte_flow type enums. For example:
../drivers/net/mlx5/mlx5_flow.c: In function ‘flow_hairpin_split’:
../drivers/net/mlx5/mlx5_flow.c:3406:19:
warning: implicit conversion from ‘enum mlx5_rtedflow_action_type’ to
‘enum rte_flow_action_type’ [-Wenum-conversion]
3406 | tag_action->type = MLX5_RTE_FLOW_ACTION_TYPE_TAG;
| ^
../drivers/net/mlx5/mlx5_flow.c:3419:13:
warning: implicit conversion from ‘enum mlx5_rte_flow_item_type’
to ‘enum rte_flow_item_type’ [-Wenum-conversion]
3419 | item->type = MLX5_RTE_FLOW_ITEM_TYPE_TAG;
| ^
Fix by casting to the correct enum.
Signed-off-by: Kevin Traynor <ktraynor@redhat.com>
Acked-by: Viacheslav Ovsiienko <viacheslavo@mellanox.com>
All comparison should be done in CPU endianness, otherwise
it will not give right results.
for example:
255 after converting into RTE_BE16 will be biger than 4096 after
converting into RTE_BE16.
Fixes: a5f2da0b81 ("net/mlx5: support modify VLAN ID on new VLAN header")
Cc: stable@dpdk.org
Signed-off-by: Wisam Jaddo <wisamm@mellanox.com>
Acked-by: Viacheslav Ovsiienko <viacheslavo@mellanox.com>
The assertion was added incorrectly in converting the modify actions
into the format of low layer driver.
There is no mask specified in the rte_flow actions, and PMD driver
will give a mask of all 1s to the field to be modified. For each
field, the mask could not be zero. But for the whole header which
contains this field, the masks of other fields could be zero. The
assertion needs to be removed for debug mode.
Fixes: 72a944dba1 ("net/mlx5: fix header modify action validation")
Cc: stable@dpdk.org
Signed-off-by: Bing Zhao <bingz@mellanox.com>
Acked-by: Viacheslav Ovsiienko <viacheslavo@mellanox.com>
Currently, there is no flow aging check and age-out event callback
mechanism for mlx5 driver, this patch implements it. It's included:
- Splitting the current counter container to aged or no-aged container
since reducing memory consumption. Aged container will allocate extra
memory to save the aging parameter from user configuration.
- Aging check and age-out event callback mechanism based on current
counter. When a flow be checked aged-out, RTE_ETH_EVENT_FLOW_AGED
event will be triggered to applications.
- Implement the new API: rte_flow_get_aged_flows, applications can use
this API to get aged flows.
Signed-off-by: Dong Zhou <dongz@mellanox.com>
Acked-by: Matan Azrad <matan@mellanox.com>
Currently, the counter pool needs 512 ext-counter memory for no batch
counters, it's allocated separately by once, behind the 512
basic-counter memory. This is not easy to get ext-counter pointer by
corresponding basic-counter pointer. This is also no easy for expanding
some other potential additional type of counter memory.
So, need allocate every one of ext-counter and basic-counter together,
as a single piece of memory. It's will be same for further additional
type of counter memory. In this case, one piece of memory contains all
type of memory for one counter, it's easy to get each type memory by
using offsetting.
Signed-off-by: Dong Zhou <dongz@mellanox.com>
Acked-by: Matan Azrad <matan@mellanox.com>
PMD create some default control rules with RSS action
if it's not isolated mode.
However whether default control rules need to do RSS or not should be
controlled by device configuration, the mq_mode of rxmode configuration
in specific.
In another word, only when mq_mode is configured with ETH_MQ_RX_RSS_FLAG
set, then RSS is needed for default rules.
Fixes: c64ccc0eca ("mlx5: fix overwritten RSS configuration")
Cc: stable@dpdk.org
Signed-off-by: Xiaoyu Min <jackmin@mellanox.com>
Signed-off-by: Viacheslav Ovsiienko <viacheslavo@mellanox.com>
Acked-by: Viacheslav Ovsiienko <viacheslavo@mellanox.com>
Acked-by: Matan Azrad <matan@mellanox.com>
The ULP mark manager originally assumed that zero was an invalid
mark and used it for invalidation and deletion. The mark manager
now supports adding zero as a mark, flags for validity and type,
and adds explicit bounds checking instead of relying on mask.
Signed-off-by: Kishore Padmanabha <kishore.padmanabha@broadcom.com>
Reviewed-by: Mike Baucom <michael.baucom@broadcom.com>
Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
The current mark handling uses the meta data field of the rxcmp as the
first level check for determining gfid vs lfid. When the meta data is
zero due to only the lowest 16bits of the gfid being set, the cfa code
is incorrectly interpreted as being an lfid. Changing code to look at
meta fmt instead of the meta data directly for the determination.
Fixes: b87abb2e55 ("net/bnxt: support marking packet")
Signed-off-by: Mike Baucom <michael.baucom@broadcom.com>
Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
Reviewed-by: Lance Richardson <lance.richardson@broadcom.com>
Due to an errata red algo needs to be set to discard instead of stall
for 96XX C0 silicon for two rate shaping. This workaround is being
already handled for newly created hierarchy but not for dynamic
shaper update cases. This patch hence applies the workaround
even when for shaper dynamic update.
Signed-off-by: Nithin Dabilpuram <ndabilpuram@marvell.com>
On detecting outer L4 checksum as bad, both outer and
inner checksums are marked as bad. No need to explicitly
check inner L4 checksum in this case.
Outer L4 UDP checksum error => PKT_RX_OUTER_L4_CKSUM_BAD
and PKT_RX_L4_CKSUM_BAD
Inner L4 UDP checksum error => PKT_RX_L4_CKSUM_BAD
Fixes: 41fe7a3a11 ("net/octeontx2: offload bad L2/L3/L4 UDP lengths detection")
Signed-off-by: Amit Gupta <agupta3@marvell.com>
Acked-by: Jerin Jacob <jerinj@marvell.com>
When octeontx_create() is cleaning up, it does not correctly set
the mac_addrs variable to NULL, which will lead to a double free.
Fixes: 9e399b88ce ("net/octeontx: fix memory leak of MAC address table")
Cc: stable@dpdk.org
Signed-off-by: Yunjian Wang <wangyunjian@huawei.com>
Acked-by: Harman Kalra <hkalra@marvell.com>
Add support for get firmware version operation.
Get and dump multi boot image (MBI) version as part of get
firmware version string along with Management firmware (MFW) version.
Use qede_fw_version_get() for PMD info logs.
Signed-off-by: Yash Sharma <ysharma@marvell.com>
Signed-off-by: Rasesh Mody <rmody@marvell.com>
Signed-off-by: Igor Russkikh <irusskikh@marvell.com>
In case VIRTIO_F_ORDER_PLATFORM(36) is not negotiated, then the frontend
and backend are assumed to be implemented in software, that is they can
run on identical CPUs in an SMP configuration.
Thus a weak form of memory barriers like rte_smp_r/wmb, other than
rte_cio_r/wmb, is sufficient for this case(vq->hw->weak_barriers == 1)
and yields better performance.
For the above case, this patch helps yielding even better performance
by replacing the two-way barriers with C11 one-way barriers for avail
index in split ring.
Signed-off-by: Joyce Kong <joyce.kong@arm.com>
Reviewed-by: Gavin Hu <gavin.hu@arm.com>
Reviewed-by: Maxime Coquelin <maxime.coquelin@redhat.com>
In case VIRTIO_F_ORDER_PLATFORM(36) is not negotiated, then the frontend
and backend are assumed to be implemented in software, that is they can
run on identical CPUs in an SMP configuration.
Thus a weak form of memory barriers like rte_smp_r/wmb, other than
rte_cio_r/wmb, is sufficient for this case(vq->hw->weak_barriers == 1)
and yields better performance.
For the above case, this patch helps yielding even better performance
by replacing the two-way barriers with C11 one-way barriers for used
index in split ring.
Signed-off-by: Joyce Kong <joyce.kong@arm.com>
Reviewed-by: Gavin Hu <gavin.hu@arm.com>
Reviewed-by: Maxime Coquelin <maxime.coquelin@redhat.com>
Since the return value of the '.stats_reset' and '.xstats_reset'
callback function is int, when failing to issue command to firmware to
execute clear statistics, the relevant callback function should return
non-zero value.
Fixes: 8839c5e202 ("net/hns3: support device stats")
Cc: stable@dpdk.org
Signed-off-by: Wei Hu (Xavier) <xavier.huwei@huawei.com>
Currently, based on hns3 VF device error may occur during initialization.
The root cause as below:
When the following formula is executed during initialization, the
private variable named hw->tqps_num has not been obtained from PF driver
through mailbox, further causes failure when mapping interrupt and
queues.
hw->num_msi = (num_msi > hw->tqps_num + 1) ? hw->tqps_num + 1 : num_msi;
We need to use hw->tqp_num after it is correctly assigned.
On the other hand, because the private variable named hw->num_msi, which
represents the number of MSI-x interrupt of hns3 PF/VF device, is used in
the '.get_reg' ops implementation function to dump all interrupt related
registers, it should be obtained from firmware directly and we'd better
not modify it in the driver.
Fixes: ef2e785c36 ("net/hns3: fix Tx interrupt when enabling Rx interrupt")
Fixes: 02a7b55657 ("net/hns3: support Rx interrupt")
Cc: stable@dpdk.org
Signed-off-by: Wei Hu (Xavier) <xavier.huwei@huawei.com>
Signed-off-by: Hao Chen <chenhao164@huawei.com>
Signed-off-by: Min Hu (Connor) <humin29@huawei.com>
In current version, when upper level application calls the
rte_eth_dev_configure API function, if pvid config is not set of the
input parameter which struct type is rte_eth_conf, hns3 pmd driver also
sets the VLAN pvid related configuration to hardware, and this is not
reasonable. For example, As pvid is set to 100 by
rte_eth_dev_set_vlan_pvid, when pvid config is not set in rte_eth_conf,
rte_eth_dev_configure will tell driver to delete pvid 0, and that is
meaningless.
This patch fixes it to ensure that driver does not set VLAN pvid related
configuration to hardware when pvid config is not set in rte_eth_conf.
Fixes: 411d23b9ea ("net/hns3: support VLAN")
Cc: stable@dpdk.org
Signed-off-by: Min Hu (Connor) <humin29@huawei.com>
Signed-off-by: Wei Hu (Xavier) <xavier.huwei@huawei.com>
Signed-off-by: Chengchang Tang <tangchengchang@huawei.com>
The hns3 network engine is built-in multiple SoCs, such as kunpeng 920,
kunpeng 930, etc. The PCI revision id is 0x21 in kunpeng 920, and the PCI
revision id is 0x30 in kunpeng 930.
This patch gets PCI revision to identify different version of hardware
network engine.
Signed-off-by: Wei Hu (Xavier) <xavier.huwei@huawei.com>
Here adds some prints for return value when the relative function fails
and enter the exception branch.
Signed-off-by: Lijun Ou <oulijun@huawei.com>
Signed-off-by: Wei Hu (Xavier) <xavier.huwei@huawei.com>
When upper level application calls the rte_eth_tx_burst API function to
send multiple packets at a time with burst mode based on hns3 network
engine, there are some abnormal conditions that cause the driver to fail
to operate the hardware to send packets correctly.
This patch adds some statistic counts for the abnormal errors of Tx data
path to the extend device statistics. The upper level application can
get them by calling the rte_eth_xstats_get API function.
Note: When using burst mode to call the rte_eth_tx_burst API function to
send multiple packets at a time. When the first abnormal error is
detected, add one to the relevant error statistics item, and then exit
the loop of sending multiple packets of the function. That is to say,
even if there are multiple packets in which abnormal errors may be
detected in the burst, the relevant error statistics in the driver will
only be increased by one.
The detail description of the Tx abnormal errors statistic items as
below:
- TX_OVER_LENGTH_PKT_CNT Total number of greater than
HNS3_MAX_FRAME_LEN the driver supported.
- TX_EXCEED_LIMITED_BD_PKT_CNT
Total number of exceeding the hardware limited bd which process a
packet needed bd numbers.
- TX_EXCEED_LIMITED_BD_PKT_REASSEMBLE_FAIL_CNT
Total number of exceeding the hardware limited bd fail which
process a packet needed bd numbers and reassemble fail.
- TX_UNSUPPORTED_TUNNEL_PKT_CNT
Total number of unsupported tunnel packet. The unsupported tunnel
type: vxlan_gpe, gtp, ipip and MPLSINUDP, MPLSINUDP is a packet
with MPLS-in-UDP RFC 7510 header.
- TX_QUEUE_FULL_CNT
Total count which the available bd numbers in current bd queue is
less than the bd numbers with the pkt process needed.
- TX_SHORT_PKT_PAD_FAIL_CNT
Total count which the packet length is less than minimum packet
size HNS3_MIN_PKT_SIZE and fail to be appended with 0.
Signed-off-by: Lijun Ou <oulijun@huawei.com>
Signed-off-by: Wei Hu (Xavier) <xavier.huwei@huawei.com>
Signed-off-by: Huisong Li <lihuisong@huawei.com>
Signed-off-by: Chengwen Feng <fengchengwen@huawei.com>
Signed-off-by: Hao Chen <chenhao164@huawei.com>
For sending request messages to data plane threads, the
caller invokes thread_msg_send_recv() function which never
returns null response. Thus, removed redundant check on
the returned response.
Coverity issue: 357717, 357772
Fixes: 70709c78fd ("net/softnic: add command to enable/disable pipeline")
Cc: stable@dpdk.org
Signed-off-by: Jasvinder Singh <jasvinder.singh@intel.com>
Acked-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
Make local MCDI helper functions static.
Fixes: f0bda0cd68 ("net/sfc/base: add MCDI wrappers for vPort and vSwitch in EVB")
Fixes: ea94d14dbe ("net/sfc/base: provide APIs to configure and reset vPort")
Cc: stable@dpdk.org
Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>
The maximal supported header modifications number of a single modify
context on the root table cannot be queried from firmware directly.
It is a fixed value of 16 in the latest releases. In the validation
stage, PMD driver should ensure that no more than 16 header modify
actions exist in a single context.
In some old firmware releases, the supported value is 8. PMD driver
should try its best to create the flow. Firmware will return error
and refuse to create the flow if the actions number exceeds the
maximal value.
Fixes: 72a944dba1 ("net/mlx5: fix header modify action validation")
Cc: stable@dpdk.org
Signed-off-by: Bing Zhao <bingz@mellanox.com>
Acked-by: Viacheslav Ovsiienko <viacheslavo@mellanox.com>
Currently, the indexed memory pool bitmap start address is not aligned
to cacheline size explicitly. The bitmap initialization requires the
address should be cacheline aligned. In that case, the initialization
maybe failed if the address is not cacheline aligned.
Add RTE_CACHE_LINE_ROUNDUP() to the trunk size calculation to make sure
the bitmap offset address will start with cacheline aligned.
Fixes: a3cf59f56c ("net/mlx5: add indexed memory pool")
Signed-off-by: Suanming Mou <suanmingm@mellanox.com>
Tested-by: Lijian Zhang <lijian.zhang@arm.com>
Acked-by: Viacheslav Ovsiienko <viacheslavo@mellanox.com>
Reviewed-by: Ruifeng Wang <ruifeng.wang@arm.com>
The assert that checks if there is a enough room for the
whole packet minus headroom data is written incorrectly.
The check should be negated in order to work properly.
Fixes: bd0d5930bf ("net/mlx5: enable MPRQ multi-stride operations")
Signed-off-by: Alexander Kozyrev <akozyrev@mellanox.com>
Acked-by: Viacheslav Ovsiienko <viacheslavo@mellanox.com>
The assert in dynamic flow metadata handling is wrong after the
fix for the performance degradation. The assert meant to check
the metadata mask but was updated with the metadata offset instead.
Fix this assert and restore proper metadata mask checking.
Fixes: 6c55b622a9 ("net/mlx5: set dynamic flow metadata in Rx queues")
Cc: stable@dpdk.org
Signed-off-by: Alexander Kozyrev <akozyrev@mellanox.com>
Acked-by: Viacheslav Ovsiienko <viacheslavo@mellanox.com>
Rewrite vectorized path selection logic. Default setting comes from
vectorized devarg, then checks each criteria.
Packed ring vectorized path need:
AVX512F and required extensions are supported by compiler and host
VERSION_1 and IN_ORDER features are negotiated
mergeable feature is not negotiated
LRO offloading is disabled
Split ring vectorized rx path need:
mergeable and IN_ORDER features are not negotiated
LRO, chksum and vlan strip offloadings are disabled
Signed-off-by: Marvin Liu <yong.liu@intel.com>
Reviewed-by: Maxime Coquelin <maxime.coquelin@redhat.com>
Optimize packed ring Tx path like Rx path. Split Tx path into batch and
single Tx functions. Batch function is further optimized by AVX512
instructions.
Signed-off-by: Marvin Liu <yong.liu@intel.com>
Reviewed-by: Maxime Coquelin <maxime.coquelin@redhat.com>
Optimize packed ring Rx path with SIMD instructions. Solution of
optimization is pretty like vhost, is that split path into batch and
single functions. Batch function is further optimized by AVX512
instructions.
Signed-off-by: Marvin Liu <yong.liu@intel.com>
Reviewed-by: Maxime Coquelin <maxime.coquelin@redhat.com>
Move offload, xmit cleanup and packed xmit enqueue function to header
file. These functions will be reused by packed ring vectorized path.
Signed-off-by: Marvin Liu <yong.liu@intel.com>
Reviewed-by: Maxime Coquelin <maxime.coquelin@redhat.com>
Add new devarg for virtio user device vectorized path selection.
By default vectorized path is disabled.
Signed-off-by: Marvin Liu <yong.liu@intel.com>
Reviewed-by: Maxime Coquelin <maxime.coquelin@redhat.com>
Previously, virtio split ring vectorized path was enabled by default.
This is not suitable for everyone because that path does not follow
virtio spec. Add new devarg for virtio vectorized path selection. By
default vectorized path is disabled.
Signed-off-by: Marvin Liu <yong.liu@intel.com>
Reviewed-by: Maxime Coquelin <maxime.coquelin@redhat.com>
Ring initialization is different when inorder feature negotiated. This
action should dependent on negotiated feature bits.
Signed-off-by: Marvin Liu <yong.liu@intel.com>
Reviewed-by: Maxime Coquelin <maxime.coquelin@redhat.com>
Introduce free threshold setting in Rx queue, its default value is 32.
Limit the threshold size to multiple of four as only vectorized packed
Rx function will utilize it. Virtio driver will rearm Rx queue when
more than rx_free_thresh descs were dequeued.
Signed-off-by: Marvin Liu <yong.liu@intel.com>
Reviewed-by: Maxime Coquelin <maxime.coquelin@redhat.com>
Currently, while creating the flow with meter, meter id is saved to the
rte flow. While destroying the flow, the meter object will be found by
the meter id, so the meter object will be released accordingly. But as
the meter id is configured by user, while the meter id is set to 0, it
doesn't make any sense to flow destroy since 0 means flow doesn't have
meter. The meter object with id 0 will be leaked.
As meter object is allocated from indexed memory, and the index starts
from 1, save the internal generated index instead of user defined meter
id will never meet the issue as above.
This patch saves meter index instead of meter id in rte flow.
Signed-off-by: Suanming Mou <suanmingm@mellanox.com>
Acked-by: Viacheslav Ovsiienko <viacheslavo@mellanox.com>
The asserts makes sure that 'i' doesn't exceed the expected value.
This to prevent an out of bound access to dbr_bitmap.
The current location of the assert protects the assignment of
dbr_bitmap, but not the access to it.
Moved the assert to the correct place, to protect both cases.
Also, used an existing define for the assert.
Fixes: 21cae8580f ("net/mlx5: allocate door-bells via DevX")
Cc: stable@dpdk.org
Signed-off-by: Asaf Penso <asafp@mellanox.com>
Reviewed-by: Dekel Peled <dekelp@mellanox.com>
Acked-by: Viacheslav Ovsiienko <viacheslavo@mellanox.com>
The output flow error parameter is used to indicate the detailed
reason of the failure when calling a rte_flow_* interface. Even
though sometimes the application will not check it or use it, the PMD
must fill it in the failure branch before returning. Or else, some
dirty value in the stack, heap will be accessed as a pointer and then
cause a crash.
In this case, when a port is stopped, it is not allowed to insert a
flow from application. The detailed error information should be
filled. If the application needs to check the detailed error reason,
it will get the information but not result in any crash.
Fixes: 40b9e7f65f ("net/mlx5: check device status before creating flow")
Signed-off-by: Bing Zhao <bingz@mellanox.com>
Acked-by: Ori Kam <orika@mellanox.com>
After inserting an offload flow, the software flag information will
be updated based on the flow. When receiving a packet on this queue,
the hardware packet type bits and the software flag will be used
together to get the inner packet and tunnel header type (if any) from
the global packet type table.
When destroying a flow, the corresponding Rx queue flag needs to be
updated. All flags should be cleared when closing a device because
all control flows and application flows are invalid anymore.
Such behavior is missed when implementing the non-cached mode.
Fixes: 8db7e3b698 ("net/mlx5: change operations for non-cached flows")
Signed-off-by: Bing Zhao <bingz@mellanox.com>
Acked-by: Viacheslav Ovsiienko <viacheslavo@mellanox.com>
1. ln_en bit should not be turned on, since we only support Rx VEB.
2. lan_en bit need to be turned on for a DCF switch rule, otherwise
any Tx packet that hit on a rule will be dropped.
Fixes: fed0c5ca5f ("net/ice/base: support programming a new switch recipe")
Cc: stable@dpdk.org
Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
Acked-by: Qiming Yang <qiming.yang@intel.com>
After VF reset, FDIR rule still takes effect. To solve the issue,
this patch adds to flush all flows before flow uninit. VIRTCHNL
sends message to PF by Admin Queue, so flow flush should be implemented
before Admin Queue shut down.
Fixes: ff2d0c345c ("net/iavf: support generic flow API")
Signed-off-by: Simei Su <simei.su@intel.com>
Acked-by: Xiaolong Ye <xiaolong.ye@intel.com>
When we flush FDIR filter, we can not call i40e_fdir_teardown()
function as it will free vsi used for FDIR, then the vsi->base_queue
will be freed from pf->qp_pool, but vsi->base_queue can only get
once when do dev init in i40e_pf_setup(). If we free it, it will
never be alloc again.
Bugzilla ID: 404
Fixes: 2e67a7fbf3 ("net/i40e: config flow director automatically")
Cc: stable@dpdk.org
Signed-off-by: Wei Zhao <wei.zhao1@intel.com>
Acked-by: Beilei Xing <beilei.xing@intel.com>
