librte_malloc was long since merged into librte_eal, mop up the
leftover references to it from driver Makefiles.
Fixes: 2f9d47013e ("mem: move librte_malloc to eal/common")
Signed-off-by: Panu Matilainen <pmatilai@redhat.com>
Applied with qede Makefile update too.
Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
On hosts running a non-DPDK PF driver, the VF has no means of changing
the HW CRC strip setting for a RX queue. It's implicitly enabled.
This patch checks if the host is running a non-DPDK PF kernel driver,
and returns an error, if HW CRC stripping was not requested in the port
configuration.
Signed-off-by: Björn Töpel <bjorn.topel@intel.com>
Acked-by: Helin Zhang <helin.zhang@intel.com>
In i40evf_config_vlan_pvid the check for NULL for the dev value is
unnecessary, since this value is passed in from the ethdev API which
will ensure that a valid rte_eth_dev structure is provided.
Furthermore, all code paths leading to this function already use the
dev value.
Issue identified by Coverity.
Coverity ID 13302:
There may be a null pointer dereference, or else the comparison against
null is unnecessary.
In i40evf_config_vlan_pvid: All paths that lead to this null pointer
comparison already dereference the pointer earlier
Fixes: 2b12431b53 ("i40e: add vlan stripping and insertion to VF")
Signed-off-by: Daniel Mrzyglod <danielx.t.mrzyglod@intel.com>
Acked-by: Helin Zhang <helin.zhang@intel.com>
Acked-by: Jingjing Wu <jingjing.wu@intel.com>
In Table 8-16 of the "Intel® Ethernet Controller XL710 Datasheet" it is
stated that when the whole packet is written to a single buffer, the
header length field in the descriptor will be 0. This means that when
extracting the packet/data_len field from the descriptor in the driver
we do not need to mask out the extra header-length bits.
Inside the vector driver, this reduces the need to pull all four pktlen
fields into a single register to work on. Instead of a shift and mask,
we now need to only do a shift. Therefore, we can work on each descriptor
independently, processing each using one shift intrinsic and a blend.
This change makes the code shorter and easier to read, so we can pull it
into the main descriptor processing loop instead of needing its own
function. This in turn makes the descriptor processing in the loop as a
whole slightly easier to read as it's more linear.
In terms of performance, in testing this change shows little effect, with
single-core perf tests showing a very slight improvement.
Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
Acked-by: Zhe Tao <zhe.tao@intel.com>
An analysis of the i40e code using Intel® VTune™ Amplifier 2016 showed
that the code was unexpectedly causing stalls due to "Loads blocked by
Store Forwards". This can occur when a load from memory has to wait
due to the prior store being to the same address, but being of a smaller
size i.e. the stored value cannot be directly returned to the loader.
[See ref: https://software.intel.com/en-us/node/544454]
These stalls are due to the way in which the data_len values are handled
in the driver. The lengths are extracted using vector operations, but those
16-bit lengths are then assigned using scalar operations i.e. 16-bit
stores.
These regular 16-bit stores actually have two effects in the code:
* they cause the "Loads blocked by Store Forwards" issues reported
* they also cause the previous loads in the RX function to actually be a
load followed by a store to an address on the stack, because the 16-bit
assignment can't be done to an xmm register.
By converting the 16-bit store operations into a sequence of SSE blend
operations, we can ensure that the descriptor loads only occur once, and
avoid both the additional stores and loads from the stack, as well as the
stalls due to the blocked loads.
Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
Acked-by: Zhe Tao <zhe.tao@intel.com>
Later commits to improve the driver will make use of the SSE4.1
_mm_blend_epi16 intrinsic, so:
* set the compilation level to always have SSE4.1 support,
* and add in a runtime check for SSE4.1 as part of the condition checks
for vector driver selection.
Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
Acked-by: Zhe Tao <zhe.tao@intel.com>
This patch removes several redundant forward declarations
in i40e_fdir.c.
Fixes: a778a1fa2e ("i40e: set up and initialize flow director")
Signed-off-by: Rami Rosen <rami.rosen@intel.com>
Acked-by: Helin Zhang <helin.zhang@intel.com>
This patch adds LLDP and DCBX capabilities to the qede PMD.
Signed-off-by: Harish Patil <harish.patil@qlogic.com>
Signed-off-by: Rasesh Mody <rasesh.mody@qlogic.com>
Signed-off-by: Sony Chacko <sony.chacko@qlogic.com>
The physical link is handled by the management Firmware.
This patch lays the infrastructure for interrupt/attention handling in
the driver, as link change notifications arrive via async interrupts,
as well as the handling of such notifications. It adds async event
notification handler interfaces to the PMD.
Signed-off-by: Harish Patil <harish.patil@qlogic.com>
Signed-off-by: Rasesh Mody <rasesh.mody@qlogic.com>
Signed-off-by: Sony Chacko <sony.chacko@qlogic.com>
This patch adds the features to supports configuration of various Layer 2
elements, such as channels and filtering options.
Signed-off-by: Harish Patil <harish.patil@qlogic.com>
Signed-off-by: Rasesh Mody <rasesh.mody@qlogic.com>
Signed-off-by: Sony Chacko <sony.chacko@qlogic.com>
The Qlogic Everest Driver for Ethernet(QEDE) Poll Mode Driver(PMD) is
the DPDK specific module for QLogic FastLinQ QL4xxxx 25G/40G CNA family
of adapters as well as their virtual functions (VF) in SR-IOV context.
This patch adds QEDE PMD, which interacts with base driver and
initialises the HW.
This patch content also includes:
- eth_dev_ops callbacks
- Rx/Tx support for the driver
- link default configuration
- change link property
- link up/down/update notifications
- vlan offload and filtering capability
- device/function/port statistics
- qede nic guide and updated overview.rst
Note that the follow on commits contain the code for the features mentioned
in documents but not implemented in this patch.
Signed-off-by: Harish Patil <harish.patil@qlogic.com>
Signed-off-by: Rasesh Mody <rasesh.mody@qlogic.com>
Signed-off-by: Sony Chacko <sony.chacko@qlogic.com>
The base driver is the backend module for the QLogic FastLinQ QL4xxxx
25G/40G CNA family of adapters as well as their virtual functions (VF)
in SR-IOV context.
The purpose of the base module is to:
- provide all the common code that will be shared between the various
drivers that would be used with said line of products. Flows such as
chip initialization and de-initialization fall under this category.
- abstract the protocol-specific HW & FW components, allowing the
protocol drivers to have clean APIs, which are detached in its
slowpath configuration from the actual Hardware Software Interface(HSI).
This patch adds a base module without any protocol-specific bits.
I.e., this adds a basic implementation that almost entirely falls under
the first category.
Signed-off-by: Harish Patil <harish.patil@qlogic.com>
Signed-off-by: Rasesh Mody <rasesh.mody@qlogic.com>
Signed-off-by: Sony Chacko <sony.chacko@qlogic.com>
Fix issue reported by Coverity.
Coverity ID 13193: Bad bit shift operation (BAD_SHIFT)
large_shift: In expression 1 << pool, left shifting by more than 31 bits
has undefined behavior. The shift amount, pool, is at least 32.
This patch is a rework of register addr selection logic and mask
computation to made it more readable and avoid bit overflow when 32 bit
value is shifted over its size for pool > 31.
Fixes: fe3a45fd41 ("ixgbe: add VMDq support")
Signed-off-by: Tomasz Kulasek <tomaszx.kulasek@intel.com>
Acked-by: Bruce Richardson <bruce.richardson@intel.com>
The statistics queried by calling rte_eth_stats_get are zero when
the API is first called on the port. The root cause is because the
offset_loaded flag is not set correctly after device start.
This patch fixes this issue by resetting statistics at initialization
time. The resetting process will set offset_loaded flag.
Fixes: 4861cde461 ("i40e: new poll mode driver")
Signed-off-by: Jingjing Wu <jingjing.wu@intel.com>
Acked-by: Helin Zhang <helin.zhang@intel.com>
When building a chain of mbufs for a multi-segment packet, the
packet_type field resides at the end of the chain. It should be
copied forward to the head of the list.
Also, uses RTE_LIBRTE_FM10K_RX_OLFLAGS_ENABLE to guard packet-type
computation. The mbuf fields are not copied when this define is not set.
Fixes: fe65e1e1ce ("fm10k: add vector scatter Rx")
Signed-off-by: Michael Frasca <michael.frasca@oracle.com>
Acked-by: Jing Chen <jing.d.chen@intel.com>
The masking for the RX/TX enable bit was incorrect in the rx and tx
queue stop functions. Instead of using "& MASK" it used "| MASK" which
would always return true. This error was found by converity scan.
CID 13215 : Wrong operator used (CONSTANT_EXPRESSION_RESULT)
operator_confusion: txdctl | 33554432 is always 1/true regardless of the
values of its operand. This occurs as the logical second operand of
'&&'.
CID 13216 : Wrong operator used (CONSTANT_EXPRESSION_RESULT)
operator_confusion: rxdctl | 33554432 is always 1/true regardless of the
values of its operand. This occurs as the logical second operand of
'&&'.
Coverity issue: 13215
Coverity issue: 13216
Fixes: 029fd06d40 ("ixgbe: queue start and stop")
Signed-off-by: Piotr Azarewicz <piotrx.t.azarewicz@intel.com>
Acked-by: Wenzhuo Lu <wenzhuo.lu@intel.com>
The position of register values within i40e register dumps is
supposed to reflect the register addresses. These were not being
correctly calculated.
Fixes: d9efd0136a ("i40e: add EEPROM and registers dumping")
Signed-off-by: Remy Horton <remy.horton@intel.com>
Acked-by: Jingjing Wu <jingjing.wu@intel.com>
Run ixgbe driver through checkpatch and fix the issues highlighted
Fix line spacing, some bad indentation, and in a couple
of cases use short circuit (already there) return to lessen indentation.
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
Acked-by: Wenzhuo Lu <wenzhuo.lu@intel.com>
Applied with four additional fixes for issues highlighted by checkpatch
Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
The macro RTE_VERIFY always checks a condition.
It is optimized with "unlikely" hint.
While this macro is well suited for test applications, it is preferred
in libraries and examples to enable such check in debug mode.
That's why the macro RTE_ASSERT is introduced to call RTE_VERIFY only
if built with debug logs enabled.
A lot of assert macros were duplicated and enabled with a specific flag.
Removing these #ifdef allows to test these code branches more easily
and avoid dead code pitfalls.
The ENA_ASSERT is kept (in debug mode only) because it has more
parameters to log.
Signed-off-by: Thomas Monjalon <thomas.monjalon@6wind.com>
Some statistics were deprecated since release 2.1 (49f386542a).
The last deprecated counter to be used was imcasts.
The VF loopback statistics are also removed as they are used only
in igb and duplicated in extended statistics.
The new counters should be added to extended statistics.
Signed-off-by: Thomas Monjalon <thomas.monjalon@6wind.com>
Acked-by: Remy Horton <remy.horton@intel.com>
Compilation fails on 32 bits on Xen driver, due to wrong casting:
drivers/net/xenvirt/virtqueue.h: In function ‘virtqueue_enqueue_xmit’:
drivers/net/xenvirt/virtqueue.h:234:24: error: cast from pointer to integer
of different size [-Werror=pointer-to-int-cast]
start_dp[idx].addr = rte_pktmbuf_mtod(cookie, uint64_t);
^
Fixes: d6b324c00f ("mbuf: get DMA address")
Signed-off-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>
VxLAN & NVGRE are supported by x550. As we know HW can parse
the packet and tell SW the type info. For VxLAN & NVGRE packets
there's some change. HW will not tell SW the info of the outer
header but the inner header instead. But we always take the
info as it's for the outer header. So the packet type info is
not right when x550 receives VxLAN & NVGRE packets.
As x550 only supports IPv4 VxLAN & NVGRE packets, we can tell
the outer header of VxLAN is IPv4 + UDP, and the outer header
of NVGRE is IPv4 only. What we don't know is if there's
optional field in the outer IPv4 header.
This patch implement the support of packet type for VxLAN &
NVGRE. And it fixes the wrong packet type issue either.
BTW:
It doesn't fix any existing commit as although it resolve an
issue it's more like a new feature but not a fix.
Reported-by: Konstantin Ananyev <konstantin.ananyev@intel.com>
Signed-off-by: Wenzhuo Lu <wenzhuo.lu@intel.com>
If the vhost PMD were configured with more queues than the guest, the old
code would segfault in rte_vhost_enable_guest_notification due to a NULL
virtqueue pointer.
Fixes: ee584e9710 ("vhost: add driver on top of the library")
Signed-off-by: Rich Lane <rich.lane@bigswitch.com>
Tested-by: Ciara Loftus <ciara.loftus@intel.com>
Acked-by: Yuanhan Liu <yuanhan.liu@linux.intel.com>
After some testing, it was found that retrieving numa information
about a vhost device via a call to get_mempolicy is more
accurate when performed during the new_device callback versus
the vring_state_changed callback, in particular upon initial boot
of the VM. Performing this check during new_device is also
potentially more efficient as this callback is only triggered once
during device initialisation, compared with vring_state_changed
which may be called multiple times depending on the number of
queues assigned to the device.
Reorganise the code to perform this check and assign the correct
socket_id to the device during the new_device callback.
Fixes: ee584e9710 ("vhost: add driver on top of the library")
Signed-off-by: Ciara Loftus <ciara.loftus@intel.com>
Acked-by: Yuanhan Liu <yuanhan.liu@linux.intel.com>
With (ICC) 16.0.2 20160204, getting following warnings:
.../drivers/net/ena/base/ena_com.c(492): error #3656: variable
"flags" may be used before its value is set
ENA_SPINLOCK_LOCK(admin_queue->q_lock, flags);
.../drivers/net/ena/base/ena_com.c(1971): error #3656: variable
"mem_handle" may be used before its value is set
ENA_MEM_ALLOC_COHERENT(ena_dev->dmadev, len,
For both warnings the variable value is ignored, so there is no defect.
To comfort compiler warning, a initial value provided to variables.
Signed-off-by: Ferruh Yigit <ferruh.yigit@intel.com>
Issue:
when using the following CLI in testpmd to enable ipv6 TSO feature
(set --txqflags=0 in the testpmd command)
set verbose 1
csum set ip hw 0
csum set udp hw 0
csum set tcp hw 0
csum set sctp hw 0
csum set outer-ip hw 0
csum parse_tunnel on 0
tso set 800 0
set fwd csum
start
We will not get what we want, the ipv6 packets sent out from IXIA can be
received by i40e, but cannot forward to another port.
The root cause is when HW doing the TSO offload for packets, it does not only
depends on the context descriptor to define the MSS and TSO payload size, it
also need to know whether this packets is ipv4 or ipv6, we use
i40e_txd_enable_checksum to generate the related fields for data descriptor.
But PMD will not call i40e_txd_enable_checksum if only the TSO offload flag is
set. The reason why ipv4 works fine for TSO in testpmd csum mode is csum engine
will set the ip csum flag when the packet is ipv4 and TSO is enabled but
will not set the flag for ipv6 and this flag will cause the
i40e_txd_enable_checksum to be invoked. For both the cases the TSO flag will be
set, so we need to use TSO flag to trigger the i40e_txd_enable_checksum.
The right logic here is we enable csum offload for both ipv4 and ipv6 when TSO
flag is set.
Fixes: e3f0151f ("i40e: enable Tx checksum only for offloaded packets")
Signed-off-by: Zhe Tao <zhe.tao@intel.com>
Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com>
The issue is the VF's link speed kept as 10G and status always was up.
It did not change even the physical link's status changed.
This patch fixes this issue to make VF's link info consistent with
physical link.
Fixes: 4861cde461 ("i40e: new poll mode driver")
Signed-off-by: Jingjing Wu <jingjing.wu@intel.com>
A problem is found on i350 VF. We found TX will happen once
per 4 packets. If only 1~3 packets are received, they will
not be forwarded. But the real problem is on RX side. The
reason is the default RX write-back threshold is changed to
4, so every first 3 packets may be hung there.
This patch checks the RX wthresh when setting up the RX
queue, and forces it to be 1, so every packet can be handled
immediately.
Fixes: 4a41c17dba ("igb: set default thresholds based on MAC type")
Signed-off-by: Wenzhuo Lu <wenzhuo.lu@intel.com>
Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com>
For simple TX the virtio-net header must be zeroed, but it was using memory
that had been initialized with indirect descriptor tables. This resulted in
"unsupported gso type" errors from librte_vhost.
We can use the same memory for every descriptor to save cachelines in the
vswitch.
Fixes: 6dc5de3a ("virtio: use indirect ring elements")
Signed-off-by: Rich Lane <rich.lane@bigswitch.com>
Acked-by: Yuanhan Liu <yuanhan.liu@linux.intel.com>
Acked-by: Jianfeng Tan <jianfeng.tan@intel.com>
The link speed configuration is now done with bitmaps so 100G speed
requires only a new bit flag.
The actual link speed is a number so its size must be increased from
16-bit to 32-bit.
Signed-off-by: Marc Sune <marcdevel@gmail.com>
Signed-off-by: Thomas Monjalon <thomas.monjalon@6wind.com>
Tested-by: Nelio Laranjeiro <nelio.laranjeiro@6wind.com>
Tested-by: Matej Vido <vido@cesnet.cz>
This patch redesigns the API to set the link speed/s configuration
of an ethernet port. Specifically:
- it allows to define a set of advertised speeds for
auto-negociation.
- it allows to disable link auto-negociation (single fixed speed).
- default: auto-negociate all supported speeds.
A flag autoneg in struct rte_eth_link indicates if link speed was a
result of auto-negociation or was fixed by configuration.
Signed-off-by: Marc Sune <marcdevel@gmail.com>
Tested-by: Nelio Laranjeiro <nelio.laranjeiro@6wind.com>
Signed-off-by: Thomas Monjalon <thomas.monjalon@6wind.com>
Tested-by: Beilei Xing <beilei.xing@intel.com>
Tested-by: Adrien Mazarguil <adrien.mazarguil@6wind.com>
The speed capabilities of a device can be retrieved with
rte_eth_dev_info_get().
The new field speed_capa is initialized in the drivers without
taking care of device characteristics in this patch.
When the capabilities of a driver are accurate, the table in
overview.rst must be filled.
Signed-off-by: Marc Sune <marcdevel@gmail.com>
The speed numbers ETH_LINK_SPEED_ are renamed ETH_SPEED_NUM_.
The prefix ETH_LINK_SPEED_ is kept for AUTONEG and will be used
for bit flags in next patch.
Signed-off-by: Marc Sune <marcdevel@gmail.com>
Some duplex values are replaced from 0 to half-duplex when link is down.
Some drivers are still using their own constants for duplex modes.
Signed-off-by: Marc Sune <marcdevel@gmail.com>
Define and use ETH_LINK_UP and ETH_LINK_DOWN where appropriate.
Signed-off-by: Marc Sune <marcdevel@gmail.com>
Signed-off-by: Thomas Monjalon <thomas.monjalon@6wind.com>
Loop that calculates total number of tx descriptors in slave tx queues
should iterate up to nb_tx_queues, not nb_rx_queues.
Fixes: 3ef7955700 ("bonding: fix LACP mempool size")
Signed-off-by: Vladyslav Buslov <vladyslav.buslov@harmonicinc.com>
Stopping then re-starting a bond interface containing slaves that
used polling for link detection caused the bond to think all slave
links were down and inactive.
Move the start of the polling for link from slave_add() to
bond_ethdev_start() and in bond_ethdev_stop() make sure we clear
the last_link_status of the slaves.
Fixes: a45b288ef2 ("bond: support link status polling")
Signed-off-by: Nelson Escobar <neescoba@cisco.com>
Signed-off-by: John Daley <johndale@cisco.com>
Acked-by: Declan Doherty <declan.doherty@intel.com>
These asserts are only for debugging and never fired during
any testing, but they confuse coverity's null tracking.
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
Acked-by: Yong Wang <yongwang@vmware.com>
Silence a compiler warning that this variable may be used uninitialized.
Signed-off-by: Aaron Conole <aconole@redhat.com>
Acked-by: Helin Zhang <helin.zhang@intel.com>
Tell the compiler to use an unsigned constant for the config shifts.
Signed-off-by: Aaron Conole <aconole@redhat.com>
Acked-by: Panu Matilainen <pmatilai@redhat.com>
Acked-by: Helin Zhang <helin.zhang@intel.com>
Tell the compiler to use an unsigned constant for the config shifts.
Signed-off-by: Aaron Conole <aconole@redhat.com>
Acked-by: Panu Matilainen <pmatilai@redhat.com>
Acked-by: Helin Zhang <helin.zhang@intel.com>
The device lsc interrupt check has a misleading whitespace around it which
can be improved by adding appropriate braces to the check. Since the ret
variable was checked after previous assignment, this introduces no functional
change.
Fixes: 921a72008f ("e1000: add Rx interrupt handler")
Signed-off-by: Aaron Conole <aconole@redhat.com>
Acked-by: Panu Matilainen <pmatilai@redhat.com>
The register read/write mphy functions have misleading whitespace around
the `locked` check. This cleanup merely preserves the existing functionality
and suppresses future gcc versions' "misleading indentation" warning.
Suggested-by: Panu Matilainen <pmatilai@redhat.com>
Signed-off-by: Aaron Conole <aconole@redhat.com>
Acked-by: Wenzhuo Lu <wenzhuo.lu@intel.com>
If the provided configuration does not call for RSS, then RSS is
explicitly disabled. Without this change, the device continues to
operate under the previous RSS configuration.
Fixes: 57033cdf8f ("fm10k: add PF RSS")
Signed-off-by: Michael Frasca <michael.frasca@oracle.com>
Acked-by: Jing Chen <jing.d.chen@intel.com>
Now that vmxnet3 supports TCP/UDP checksum offload, let's update
the default txq flags to allow such offloads. Also fixed the tx
queue setup check to allow TCP/UDP checksum and only error out
if SCTP checksum is requested.
Fixes: f598fd063b ("vmxnet3: add Tx L4 checksum offload")
Reported-by: Heng Ding <hengx.ding@intel.com>
Signed-off-by: Yong Wang <yongwang@vmware.com>
The NFP driver (unlike other PCI devices) was not copying the pci info
from the pci_dev to the eth_dev. This would make the driver_name be
null (and other unset fields) when application uses dev_info_get.
This was found by code review; do not have the hardware.
Fixes: d4a27a3b09 ("nfp: add basic features")
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
Acked-by: Alejandro Lucero <alejandro.lucero@netronome.com>
When the number of RX queues is not a power of two,
the RETA table is configured to its maximum size for
better balancing.
Testing showed that limiting its size to 256 improves performance
noticeably with little to no impact on balancing results.
Fixes: ebb30ec64a ("mlx5: increase RETA table size")
Signed-off-by: Yaacov Hazan <yaacovh@mellanox.com>
Acked-by: Adrien Mazarguil <adrien.mazarguil@6wind.com>
