When MLX PMD registers memory for DMA, it accesses the global memseg list
of DPDK to maximize the range of registration so that LKey search can be
more efficient. Granularity of MR registration is per page.
Externally allocated memory shouldn't be used for DMA because it can't be
searched in the memseg list and free event can't be tracked by DPDK. If it
is used, the following error will occur:
net_mlx5: port 0 unable to find virtually contiguous chunk for
address (0x5600017587c0). rte_memseg_contig_walk() failed.
There's a pending patchset [1] which enables externally allocated memory.
Once it is merged, users can register their own memory out of EAL then that
will resolve this issue.
Meanwhile, if the external memory is static (allocated on startup and never
freed), such memory can also be registered by little tweak in the code.
[1] http://patches.dpdk.org/project/dpdk/list/?series=1415
This patch is not a bug fix but needs to be included in stable versions.
Fixes: 9797bfcce1 ("net/mlx4: add new memory region support")
Cc: stable@dpdk.org
Signed-off-by: Yongseok Koh <yskoh@mellanox.com>
This patch fixes the issue that representor port shows xstats of PF.
Fixes: 2b73026388 ("net/mlx5: probe all port representors")
Cc: stable@dpdk.org
Signed-off-by: Xueming Li <xuemingl@mellanox.com>
Acked-by: Yongseok Koh <yskoh@mellanox.com>
Current code uses PF links status for representor port, not the
representor interface itself.
This caused wrong representor port link status when toggling
interface up or down.
Fixes: 2b73026388 ("net/mlx5: probe all port representors")
Cc: stable@dpdk.org
Signed-off-by: Xueming Li <xuemingl@mellanox.com>
Acked-by: Yongseok Koh <yskoh@mellanox.com>
Flows having 'transfer' attribute have to be inserted to E-Switch on the
NIC and the control path uses Linux TC flower interface via Netlink
socket.
This patch adds the flow driver on top of the new flow engine.
Signed-off-by: Yongseok Koh <yskoh@mellanox.com>
Netlink based E-Switch flow engine will be migrated to the new flow
engine.
nl_flow will be renamed to flow_tcf as it goes through Linux TC flower
interface.
Signed-off-by: Yongseok Koh <yskoh@mellanox.com>
Flow engine has to support multiple driver paths. Verbs/DV for NIC flow
steering and Linux TC flower for E-Switch flow steering. In the future,
another flow driver could be added (devX).
Signed-off-by: Yongseok Koh <yskoh@mellanox.com>
DV flow API is based on new kernel API and is
missing some functionality like counter but add other functionality
like encap.
In order not to affect current users even if the kernel supports
the new DV API it should be enabled only manually.
Signed-off-by: Ori Kam <orika@mellanox.com>
Acked-by: Yongseok Koh <yskoh@mellanox.com>
This commits add the missing function which are apply, remove, and
destroy.
Signed-off-by: Ori Kam <orika@mellanox.com>
Acked-by: Yongseok Koh <yskoh@mellanox.com>
This commit adds all Direct Verbs required functions to the glue lib.
Signed-off-by: Ori Kam <orika@mellanox.com>
Acked-by: Yongseok Koh <yskoh@mellanox.com>
In this commit we add the translation of flow actions.
Unlike the Verbs API actions are separeted from the items and are passed
to the API in array structure.
Since the target action like RSS require the QP information those
actions are handled both in the translate action and in the apply.
Signed-off-by: Ori Kam <orika@mellanox.com>
Acked-by: Yongseok Koh <yskoh@mellanox.com>
This commit handles the translation of the requested flow into Direct
Verbs API.
The Direct Verbs introduce the matcher object which acts as shared mask
for all flows that are using the same mask. So in this commit we
translate the item and get in return a matcher and the value that should
be matched.
Signed-off-by: Ori Kam <orika@mellanox.com>
Acked-by: Yongseok Koh <yskoh@mellanox.com>
This function allocates the Direct Verbs device flow, and
introduce the relevant PRM structures.
This commit also adds the matcher object. The matcher object acts as a
mask and should be shared between flows. For example all rules that
should match source IP with full mask should use the same matcher. A
flow that should match dest IP or source IP but without full mask should
have a new matcher allocated.
Signed-off-by: Ori Kam <orika@mellanox.com>
Acked-by: Yongseok Koh <yskoh@mellanox.com>
This is commit introduce the Direct Verbs driver API.
The Direct Verbs is an API adds new features like encapsulation, match
on metatdata.
In this commit the validation function was added, most of the validation
is done with functions that are also in use for the Verbs API.
Signed-off-by: Ori Kam <orika@mellanox.com>
Acked-by: Yongseok Koh <yskoh@mellanox.com>
This commit modify the conversion of the input parameters into Verbs
spec, in order to support all previous changes.
Some of those changes are:
removing the use of the parser,
storing each flow in its own flow structure.
Signed-off-by: Ori Kam <orika@mellanox.com>
Acked-by: Yongseok Koh <yskoh@mellanox.com>
In current implementation the calculation of the flow size is done
during the validation stage, and the same function is also used to
translate the input parameters into verbs spec. This is hard to
maintain and error prone.
Another issue is dev-flows (flows that are created implicitly in order
to support the requested flow for example when the user request RSS on
UDP 2 rules need to be created one for IPv4 and one for IPv6).
In current implementation the dev-flows are created on the same
memory allocation. This will be harder to implement in future drivers.
The commits extract the calculation and creation of the dev-flow from
the translation part (the part that converts the parameters into the
format required by the driver). This results in that the prepare
function only function is to allocate the dev-flow.
Signed-off-by: Ori Kam <orika@mellanox.com>
Acked-by: Yongseok Koh <yskoh@mellanox.com>
In current implementation the validation logic reside in the same
function that calculates the size of the verbs spec and also create the
verbs spec.
This approach results in hard to maintain code which can't be shared.
also in current logic there is a use of parser entity that holds the
information between function calls. The main problem with this parser is
that it assumes the connection between different functions. For example
it assumes that the validation function was called and relevant values
were set.
This may result in an issue if and when we only call the validation
function, or call the apply function without the validation (Currently
according to RTE flow we must call validation before creating flow, but
if we want to change that to save time during flow creation, for example
the user validated some rule and just want to change the IP there is no
true reason the validate the rule again).
This commit address both of those issues by extracting the validation
logic into detected functions and remove the use of the parser object.
The side effect of those changes is that in some cases there will be a
need to traverse the item list again.
Signed-off-by: Ori Kam <orika@mellanox.com>
Acked-by: Yongseok Koh <yskoh@mellanox.com>
Add the vectorized version of the no-scatter Rx handler. It aims to
process 8 descriptors per loop using AVX2 SIMD instructions. This
handler is in its own file enic_rxtx_vec_avx2.c, and makefile and
meson.build are modified to compile it when the compiler supports
AVX2. Under ideal conditions, the vectorized handler reduces
cycles/packet by more than 30%, when compared against the no-scatter
Rx handler. Most implementation ideas come from i40e's AVX2 based
handler, so credit goes to its authors.
At this point, the new handler is meant for field trials, and is not
selected by default. So add a new devarg enable-avx2-rx to allow the
user to request the use of the new handler. When enable-avx2-rx=1, the
driver will consider using the new handler.
Also update the guide doc and introduce the vectorized handler.
Signed-off-by: Hyong Youb Kim <hyonkim@cisco.com>
Reviewed-by: John Daley <johndale@cisco.com>
Move a number of Rx functions to the header file so that the avx2
based Rx handler can use them.
Signed-off-by: Hyong Youb Kim <hyonkim@cisco.com>
Reviewed-by: John Daley <johndale@cisco.com>
Signed-off-by: Igor Romanov <igor.romanov@oktetlabs.ru>
Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>
Reviewed-by: Ivan Malov <ivan.malov@oktetlabs.ru>
In the case of scattered packet one Rx buffer is just a segment
of the whole packet.
Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>
Reviewed-by: Ivan Malov <ivan.malov@oktetlabs.ru>
Number of prepared packets is good when one Rx descriptor is one packet.
Introduce pending Rx descriptor pointer which points to the first not
processed Rx descriptors. Rx descriptors from completed to pending have
buffers ready to be passed to application.
Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>
Reviewed-by: Ivan Malov <ivan.malov@oktetlabs.ru>
If there is no packets to be processed, it does not make sense
to write the same values back to Rx queue structure.
Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>
Reviewed-by: Ivan Malov <ivan.malov@oktetlabs.ru>
mbuf raw free adds debug build checks to ensure that mbuf is
really OK to be returned to mempool.
Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>
Reviewed-by: Ivan Malov <ivan.malov@oktetlabs.ru>
Make sure that number of prepared packets, completed and added
Rx ring pointers are reset to zeros on queue purge at stop.
Fixes: 638bddc99f ("net/sfc: implement EF10 native Rx datapath")
Cc: stable@dpdk.org
Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>
Reviewed-by: Ivan Malov <ivan.malov@oktetlabs.ru>
Add support for resetting of driver statistics.
Signed-off-by: Natalie Samsonov <nsamsono@marvell.com>
Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
Add callback for updating information about link status/info.
Signed-off-by: Natalie Samsonov <nsamsono@marvell.com>
Signed-off-by: Zyta Szpak <zr@semihalf.com>
Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
Add part of PMD for actual reception/transmission.
Signed-off-by: Yelena Krivosheev <yelena@marvell.com>
Signed-off-by: Dmitri Epshtein <dima@marvell.com>
Signed-off-by: Zyta Szpak <zr@semihalf.com>
Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
Add APIs for dscp priority map configuration. APIs added are
ecore_dcbx_get_dscp_priority(), ecore_dcbx_set_dscp_priority().
These base driver APIs can be used for dscp-map query/config.
Configure the doorbell queue (DORQ) to use vlan-id/priority.
Signed-off-by: Rasesh Mody <rasesh.mody@cavium.com>
Add 'rl_bc_stage_th','rl_timer_stage_th' and 'dcqcn_reset_alpha_on_idle'
to RL update param as well as logs.
Signed-off-by: Rasesh Mody <rasesh.mody@cavium.com>
Change details:
- Get engine affinity from the management FW and configure accordingly
- Add an LLH filter with the primary MAC address in QPAR/NPAR
- Move some of the LLH APIs around
- Add PPFID APIs
- Update all allocated ppfids with the same value for the
following PORT_PF registers:
NIG_REG_DSCP_TO_TC_MAP_ENABLE
- Add port_id, src_pfid and dst_pfid to DMA engine params
Signed-off-by: Rasesh Mody <rasesh.mody@cavium.com>
Request management FW for OEM values, which are negotiated prior to
the driver load by sending the GET_OEM_UPDATES command after both
engines are initialized.
Signed-off-by: Rasesh Mody <rasesh.mody@cavium.com>
Add a pretend function for port/PF, pretend to another port and another
function when accessing the ptt window
Signed-off-by: Rasesh Mody <rasesh.mody@cavium.com>
