For IA, the AVX2 vector path is only recommended to be used on later
platforms (identified by AVX512 support, like SKL etc.) This is because
performance benchmark shows downgrade when running AVX2 vector path on
early platform (BDW/HSW) in some cases. But we still observe perf gain
with some real work loading.
So this patch introduced the new devarg use-latest-supported-vec to
force the driver always selecting the latest supported vec path. Then
apps are able to take AVX2 path on early platforms. And this logic can
be re-used if we will have AVX512 vec path in future.
This patch only affects IA platforms. The selected vec path would be
like the following:
Without devarg/devarg = 0:
Machine vPMD
AVX512F AVX2
AVX2 SSE4.2
SSE4.2 SSE4.2
<SSE4.2 Not Supported
With devarg = 1
Machine vPMD
AVX512F AVX2
AVX2 AVX2
SSE4.2 SSE4.2
<SSE4.2 Not Supported
Other platforms can also apply the same logic if necessary in future.
Signed-off-by: Xiaoyun Li <xiaoyun.li@intel.com>
Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
Acked-by: Qi Zhang <qi.z.zhang@intel.com>
For i40evf, internal rx interrupt and adminq interrupt share the same
source, that cause a lot cpu cycles be wasted on interrupt handler
on rx path. This is complained by customers which require low latency
(when set I40E_ITR_INTERVAL to small value), but have to be sufferred by
tremendous interrupts handling that eat significant CPU resources.
The patch disable pci interrupt and remove the interrupt handler,
replace it with a low frequency (50ms) interrupt polling daemon
which is implemented by registering a alarm callback periodly, this
save CPU time significently: On a typical x86 server with 2.1GHz CPU,
with low latency configure (32us) we saw CPU usage from top commmand
reduced from 20% to 0% on management core in testpmd).
Also with the new method we can remove compile option: I40E_ITR_INTERVAL
which is used to balance between low latency and low CPU usage previously.
Now we don't need it since we can reach both at same time.
Suggested-by: Jingjing Wu <jingjing.wu@intel.com>
Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
Acked-by: Beilei Xing <beilei.xing@intel.com>
The summarized global register change info will be logged
no matter if there's real global register change. Since
only real changes are logged now, there's no need to
summarize global register change info, otherwise will
cause misunderstanding.
Signed-off-by: Beilei Xing <beilei.xing@intel.com>
Acked-by: Qi Zhang <qi.z.zhang@intel.com>
Currently no matter if there's global change, the global
configuration will be always logged. But there's no value
to log the info if the configuration is not changed.
This patch prints only real global changes.
Also, change log level from DEBUG to WARNING.
Signed-off-by: Beilei Xing <beilei.xing@intel.com>
Acked-by: Qi Zhang <qi.z.zhang@intel.com>
The resetting of stats on a VF involves the setting of an offset
that is subtracted from future calls, rather zeroing of counters.
However doing a stats reset on the port representor was also
adjusting the values forwarded to the VF, which had the effect of
corrupting the VF's counters. The fix is for the port representor
to maintain its own stats offset, so the port representor and VF
maintain independently-resettable counters.
Fixes: e0cb96204b71 ("net/i40e: add support for representor ports")
Signed-off-by: Remy Horton <remy.horton@intel.com>
Acked-by: Qi Zhang <qi.z.zhang@intel.com>
Currently, only new value is printed during global
register change. Add original value to help debugging
facility.
Fixes: bc66b9717c50 ("net/i40e: add debug logs when writing global registers")
Cc: stable@dpdk.org
Signed-off-by: Beilei Xing <beilei.xing@intel.com>
Acked-by: Qi Zhang <qi.z.zhang@intel.com>
Add support for virtual function representor ports to the i40e PF
driver. When SR-IOV virtual functions devices are enabled a
corresponding representor port for each VF can be enabled, in the
process in which the i40e PMD is running, by specifying the
representor devargs with the list of VF ports that representors
are to be created for.
An example of the devargs which would create VF representor for virtual
functions 0,2,4,5,6 and 7 is:
-w DBDF,representor=[0,2,4-7]
and to just specify a single representor on virtual function 3 (switch
port id):
-w DBDF,representor=3
Signed-off-by: Declan Doherty <declan.doherty@intel.com>
Signed-off-by: Mohammad Abdul Awal <mohammad.abdul.awal@intel.com>
Signed-off-by: Remy Horton <remy.horton@intel.com>
Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
Since its inception, the rte_flow RSS action has been relying in part on
external struct rte_eth_rss_conf for compatibility with the legacy RSS API.
This structure lacks parameters such as the hash algorithm to use, and more
recently, a method to tell which layer RSS should be performed on [1].
Given struct rte_eth_rss_conf will never be flexible enough to represent a
complete RSS configuration (e.g. RETA table), this patch supersedes it by
extending the rte_flow RSS action directly.
A subsequent patch will add a field to use a non-default RSS hash
algorithm. To that end, a field named "types" replaces the field formerly
known as "rss_hf" and standing for "RSS hash functions" as it was
confusing. Actual RSS hash function types are defined by enum
rte_eth_hash_function.
This patch updates all PMDs and example applications accordingly.
It breaks ABI compatibility for the following public functions:
- rte_flow_copy()
- rte_flow_create()
- rte_flow_query()
- rte_flow_validate()
[1] commit 676b605182a5 ("doc: announce ethdev API change for RSS
configuration")
Signed-off-by: Adrien Mazarguil <adrien.mazarguil@6wind.com>
Acked-by: Andrew Rybchenko <arybchenko@solarflare.com>
Customized info will be updated when processing DDP package,
including PCYPE/PTYPE/protocol. Previously, the customized info
is updated without any check for package operation - ADD or DEL,
but only covers ADD operation. In this situation, even if a package
is being removed, new PCTYPE/PTYPE/protocol will still be created,
it will cause wrong parsing for SW. This patch cleans new
PCTYPE/PTYPE/protocol created when a package is being removed.
Fixes: e163c18a15b0 ("net/i40e: update ptype and pctype info")
Cc: stable@dpdk.org
Signed-off-by: Beilei Xing <beilei.xing@intel.com>
Acked-by: Jingjing Wu <jingjing.wu@intel.com>
Add support the set_mc_addr_list device operation in the i40evf PMD.
The configured addresses are stored in the device private area, so
they can be flushed before adding new ones.
Signed-off-by: Olivier Matz <olivier.matz@6wind.com>
Acked-by: Qi Zhang <qi.z.zhang@intel.com>
There's interrupt conflict when using DPDK and Linux i40e
on different ports of the same Ethernet controller, this
patch fixes it by switching from IntN to Int0 if multiple
drivers are used.
Fixes: be6c228d4da3 ("i40e: support Rx interrupt")
Cc: stable@dpdk.org
Signed-off-by: Beilei Xing <beilei.xing@intel.com>
Acked-by: Jingjing Wu <jingjing.wu@intel.com>
This patch provides the option to disable writing some global registers
in PMD, in order to avoid affecting other drivers, when multiple drivers
run on the same NIC and control different physical ports. Because there
are few global resources shared among different physical ports.
Fixes: ec246eeb5da1 ("i40e: use default filter input set on init")
Fixes: 98f055707685 ("i40e: configure input fields for RSS or flow director")
Fixes: f05ec7d77e41 ("i40e: initialize flow director flexible payload setting")
Fixes: e536c2e32883 ("net/i40e: fix parsing QinQ packets type")
Fixes: 19b16e2f6442 ("ethdev: add vlan type when setting ether type")
Cc: stable@dpdk.org
Signed-off-by: Beilei Xing <beilei.xing@intel.com>
Acked-by: Jingjing Wu <jingjing.wu@intel.com>
Rte_flow was defined to include RSS, this patch moves i40e
existing RSS to rte_flow. The old RSS configuration is kept
as it was, and can be deprecated in the future.
Signed-off-by: Wei Zhao <wei.zhao1@intel.com>
Acked-by: Qi Zhang <qi.z.zhang@intel.com>
Function i40e_GLQF_reg_init() overwrites global register for
flexible payload, forcing extraction of first 16 bytes of
L2/L3/L4 payload to the field vector even if flexible payload
is not used by an application. Such unconditional turn on of
flexible payload effectively disables ability to use outer
IP Destination address for RSS/FDIR for tunnelled packets,
as flexible payload overwrites outer IP destination address
on the field vector.
Now flexible payload turned on only when flow director is
enabled and configured.
Signed-off-by: Kirill Rybalchenko <kirill.rybalchenko@intel.com>
Acked-by: Beilei Xing <beilei.xing@intel.com>
Current default ITR interval I40E_QUEUE_ITR_INTERVAL_DEFAULT is
used by both DPDK PF and DPDK VF, but this ITR interval value
will cause lots of interrupts when using VF. This patch modifies
VF default ITR interval to reduce interrupts.
ITR interval can be reconfigured by RTE_LIBRTE_I40E_ITR_INTERVAL
if expecting low latency.
Signed-off-by: Beilei Xing <beilei.xing@intel.com>
Acked-by: Jingjing Wu <jingjing.wu@intel.com>
According to loopback mode, setup loopback link or not.
If loopback link is set, packets transmitted will be
looped back to received side directly.
Signed-off-by: Yanglong Wu <yanglong.wu@intel.com>
Acked-by: Beilei Xing <beilei.xing@intel.com>
Repeated occurrences of 'the'.
The change was obtained using the following command:
sed -i "s;the the ;the ;" `git grep -l "the "`
Signed-off-by: Thierry Herbelot <thierry.herbelot@6wind.com>
Replace the BSD license header with the SPDX tag for files
with only an Intel copyright on them.
Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
For complex packets use raw flow type with pre-constructed packet buffer
instead of creating a packet internally in PMD.
Signed-off-by: Kirill Rybalchenko <kirill.rybalchenko@intel.com>
Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
This feature enable queue regions configuration for RSS in PF,
so that different traffic classes or different packet
classification types can be separated to different queues in
different queue regions.This patch can set queue region range,
it include queue number in a region and the index of first queue.
This patch enable mapping between different priorities (UP) and
different traffic classes.It also enable mapping between a region
index and a sepcific flowtype(PCTYPE).It also provide the solution
of flush all configuration about queue region the above described.
Signed-off-by: Wei Zhao <wei.zhao1@intel.com>
This patch sets TEID of GTP-C and GTP-U as filter type
by replacing existed filter types inner_mac and TUNNEL_KEY.
This configuration will be set when adding GTP-C or
GTP-U filter rules, and it will be invalid only by
NIC core reset.
Signed-off-by: Beilei Xing <beilei.xing@intel.com>
Acked-by: Jingjing Wu <jingjing.wu@intel.com>
This patch adds i40e_flow_parse_gtp_filter parsing
function for GTP-C and GTP-U to support cloud filter.
Signed-off-by: Beilei Xing <beilei.xing@intel.com>
Acked-by: Jingjing Wu <jingjing.wu@intel.com>
This patch adds FDIR support for GTP-C and GTP-U. The
input set of GTP-C and GTP-U is TEID.
Signed-off-by: Beilei Xing <beilei.xing@intel.com>
Reviewed-by: Seán Harte <seanbh@gmail.com>
rte_eth_fdir_* structures are still used in FDIR functions.
This patch adds i40e private FDIR related structures and
functions to finish integration FDIR with generic flow API.
Signed-off-by: Beilei Xing <beilei.xing@intel.com>
Update new packet type and new pctype info when downloading
profile.
Signed-off-by: Beilei Xing <beilei.xing@intel.com>
Acked-by: Jingjing Wu <jingjing.wu@intel.com>
Implement dynamic mapping of software flow types to hardware pctypes.
This allows to add new flow types and pctypes for DDP without changing
API of the driver. The mapping table is located in private
data area for particular network adapter and can be individually
modified with set of appropriate functions.
Signed-off-by: Kirill Rybalchenko <kirill.rybalchenko@intel.com>
Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
As no matter the PF host driver is DPDK or other kernel drivers,
they are sharing the same virtchnnl interfaces to communicate to VFs.
To follow the generic interface, DPDK PF need to set Interrupt
Throttling (ITR) index according to the rxitr_idx from virtchnnl
instead of ITR_NONE.
Fixes: 6d59e4ea74a6 ("net/i40e: change version number to support Linux VF")
Cc: stable@dpdk.org
Signed-off-by: Jingjing Wu <jingjing.wu@intel.com>
Add the support of the Traffic Management API,
rte_tm_hierarchy_commit.
When calling this API, the driver tries to enable
the TM configuration on HW.
Signed-off-by: Wenzhuo Lu <wenzhuo.lu@intel.com>
To support QoS scheduler APIs, create a new C file for
the TM (Traffic Management) ops but without any function
implemented.
Signed-off-by: Wenzhuo Lu <wenzhuo.lu@intel.com>
Internal Rx packet bytes is over counted, CRC size is not excluded,
that cause incorrect Rx bytes or even negative value.
Fixes: 0bcdc44510ef ("net/i40e: exclude internal packet's byte count")
Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
Modify the necessary files to be compatible with the new virtchnl.h file
instead of relying on i40e_virtchnl.h variant. This mostly changes
references to VIRTCHNL_ variables by removing prefix of I40E_.
Signed-off-by: Jingjing Wu <jingjing.wu@intel.com>
After adding a fdir rule for IPv6 with input set TC, IPv6 packets
with the specific TC can't be assigned the right queue.
The root cause is that TC is parsed wrongly, this patch fixes
TC parsing problem.
Fixes: 7d83c152a207 ("net/i40e: parse flow director filter")
Cc: stable@dpdk.org
Signed-off-by: Beilei Xing <beilei.xing@intel.com>
Tested-by: Yuan Peng <yuan.peng@intel.com>
Fixing typos across dpdk source code using codespell utility.
Skipped the ethdev driver's base code fixes to keep the base
code intact.
Signed-off-by: Jerin Jacob <jerin.jacob@caviumnetworks.com>
Acked-by: John McNamara <john.mcnamara@intel.com>
This patch supports input set selection for flow
director filter.
Signed-off-by: Beilei Xing <beilei.xing@intel.com>
Acked-by: Wenzhuo Lu <wenzhuo.lu@intel.com>
This patch adds flexible payload parsing support for
flow director filter.
Signed-off-by: Beilei Xing <beilei.xing@intel.com>
Acked-by: Wenzhuo Lu <wenzhuo.lu@intel.com>
Tx/Rx byte counts of internal managed packet should be
excluded from the total rx/tx bytes.
Fixes: 9aace75fc82e ("i40e: fix statistics")
Cc: stable@dpdk.org
Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
Acked-by: Beilei Xing <beilei.xing@intel.com>
Instead of many PMD define their own macro, define a generic one in
ethdev and use that in PMDs.
Signed-off-by: Ferruh Yigit <ferruh.yigit@intel.com>
Acked-by: Allain Legacy <allain.legacy@windriver.com>
Acked-by: Wenzhuo Lu <wenzhuo.lu@intel.com>
Acked-by: Andrew Rybchenko <arybchenko@solarflare.com>
remove __rte_unused instances that are not required.
Signed-off-by: Ferruh Yigit <ferruh.yigit@intel.com>
Acked-by: Allain Legacy <allain.legacy@windriver.com>
Acked-by: Wenzhuo Lu <wenzhuo.lu@intel.com>
When counting max packet length from MTU, count
VLAN tag length twice for QinQ packets.
Signed-off-by: Wenzhuo Lu <wenzhuo.lu@intel.com>
Acked-by: Jingjing Wu <jingjing.wu@intel.com>
The mapping from hardware defined packet type to software defined
packet type is static for i40e device, the patch let each ethdev to
to have their own copy of mapping table, this give the possibility
that different ethdev can be set different PTYPE mapping rule which
is the requirement to support following hardware's dynamic PTYPE
feature.
Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
Acked-by: Jianbo Liu <jianbo.liu@linaro.org>
Acked-by: Chao Zhu <chaozhu@linux.vnet.ibm.com>
Failed to destroy tunnel filter rule if the action of
the tunnel filter is VF, root cause is the wrong vsi
used.
Fixes: c50474f31efe ("net/i40e: support tunnel filter to VF")
Signed-off-by: Beilei Xing <beilei.xing@intel.com>
Acked-by: Bernard Iremonger <bernard.iremonger@intel.com>
