Minor cleanup in bnxt_mtu_set_op() to move pre-mature
setting of jumbo flag post mtu check and remove
a redundant mtu set operation from rxq vnic configs.
Fixes: daef48efe5e5 ("net/bnxt: support set MTU")
Cc: stable@dpdk.org
Signed-off-by: Damodharam Ammepalli <damodharam.ammepalli@broadcom.com>
Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
With this patch, bnxt_mtu_set_op() will return an error code if the
device has already started. The user application will have to take
care to bring down device before invoking the mtu_set()
Fixes: daef48efe5e5 ("net/bnxt: support set MTU")
Cc: stable@dpdk.org
Signed-off-by: Damodharam Ammepalli <damodharam.ammepalli@broadcom.com>
Reviewed-by: Andy Gospodarek <andrew.gospodarek@broadcom.com>
Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
The "active_fec_signal_mode" in HWRM_PORT_PHY_QCFG response
does not return correct value till the link is up. Driver cannot
rely on active_fec_signal_mode while setting forced speed.
While setting forced speed of 50G/100G/200G, check if PAM4 speeds
are supported for the port first and then populate the HWRM request
accordingly.
Also, If PAM4 speed is supported, use PAM4 supported speed while
reporting speed capabilities.
Fixes: c23f9ded0391 ("net/bnxt: support 200G PAM4 link")
Cc: stable@dpdk.org
Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com>
Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
Currently, we fail the init/probe of PMD if eth_dev->data->nb_tx_queues
or eth_dev->data->nb_rx_queues is 0. We are removing this check.
Fixes: daef48efe5e5 ("net/bnxt: support set MTU")
Cc: stable@dpdk.org
Signed-off-by: Damodharam Ammepalli <damodharam.ammepalli@broadcom.com>
Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
Broadcom HW drops packets when there are no descriptors available.
It does not matter what flag the application specifies in "rx_drop_en"
when configuring the Rx ring.
Reduce the verbosity of the log to print the status of the "rx_drop_en"
when configuring the Rx ring.
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>
Allocate switch domain after the trusted VF capability is queried
from the FW. Currently we are calling the function earlier.
Since the switch domain is allocated only for PFs or trusted VF,
the current location of code fails to allocate the domain during init.
But during cleanup we try to free the domain incorrectly.
Fix the behavior by changing the sequence of function calls.
Fixes: 3127f99274b67 ("net/bnxt: refactor init/uninit")
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>
Reviewed-by: Andy Gospodarek <andrew.gospodarek@broadcom.com>
Currently the PMD tries to detect a potential 0 byte DMA by
using RTE_VERIFY.
But since RTE_VERIFY internally calls rte_panic() it is fatal to
the application and some applications want to avoid that.
So return an error from the bnxt xmit handler if such a bad pkt is
encountered by logging an error message, dumping the pkt header and
dump the current stack as well
Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com>
Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
To read descriptors in expected order, barriers are inserted after each
descriptor read. The excessive use of barriers is unnecessary and could
cause performance drop.
Removed barriers between descriptor reads. And changed counting of valid
packets so as to handle discontinuous valid packets. Because out of
order read could lead to valid descriptors that fetched being
discontinuous.
In VPP L3 routing test, 6% performance gain was observed. The test was
done on a platform with ThunderX2 CPU and Broadcom PS225 NIC.
Signed-off-by: Ruifeng Wang <ruifeng.wang@arm.com>
Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
The requirement of SYS_ADMIN capability in legacy virtio mode
was missing. Add it to the driver-specific page.
Signed-off-by: Dmitry Kozlyuk <dkozlyuk@nvidia.com>
Reference the common guide for generic setup.
Remove excessive capabilities from the recommended list.
Cc: stable@dpdk.org
Signed-off-by: Dmitry Kozlyuk <dkozlyuk@nvidia.com>
The guide to run DPDK applications as non-root in Linux
did not provide specific instructions to configure the required access
and did not explain why each bit is needed.
The latter is important because running as non-root
is one of the ways to tighten security and grant minimal permissions.
Cc: stable@dpdk.org
Signed-off-by: Dmitry Kozlyuk <dkozlyuk@nvidia.com>
Acked-by: Bruce Richardson <bruce.richardson@intel.com>
Per mount(8), the previous owner and mode of the mount point
become invisible as long as this filesystem remains mounted.
Because dpdk-hugepages.py must be run as root,
the new owner would be root.
This is undesirable if the hugepage directory is being set up
by the administrator for an unprivileged user.
HugeTLB filesystem has options to set the mount point owner.
Add --user/-U and --group/-G options to apply this when mounting.
The benefit of performing this in dpdk-hugepages.py
is that the user does not need to care about this detail
of mount command operation.
Signed-off-by: Dmitry Kozlyuk <dkozlyuk@nvidia.com>
Acked-by: Bruce Richardson <bruce.richardson@intel.com>
dpdk-hugepages.py had /dev/hugepages hardcoded as the mount point.
It may be desirable to setup hugepage directory at another path,
for example, when using hugepages of multiple sizes in different
directories or when granting different permissions to mount points.
Add --directory/-d option to the script.
Signed-off-by: Dmitry Kozlyuk <dkozlyuk@nvidia.com>
Acked-by: Bruce Richardson <bruce.richardson@intel.com>
Both the Linux and FreeBSD GSG docs had a "Documentation Roadmap"
section as part of the introduction page, and this contained the same
information, with only the reference to the GSGs themselves being
different. This text can be consolidated into a single text file which
is included by both GSG intro sections - using relative links for the
self reference.
Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
Acked-by: John McNamara <john.mcnamara@intel.com>
This patch fixes the syntax error when using the single-core
for both Rx and distributor functions.
Fixes: 4a7f40c0ff9a ("examples/distributor: add dedicated core")
Cc: stable@dpdk.org
Signed-off-by: Abdullah Ömer Yamaç <omer.yamac@ceng.metu.edu.tr>
Acked-by: Ferruh Yigit <ferruh.yigit@xilinx.com>
Tested-by: David Hunt <david.hunt@intel.com>
To enable input coloring, based on VLAN or DSCP, patch adds
command line interface to configure the following:
- configuring input coloring using VLAN or DSCP while creating
meter i.e. during rte_mtr_create()
- Update VLAN input coloring table at runtime.
- configures protocol priorities.
- retrieve protocol and priority information
Signed-off-by: Sunil Kumar Kori <skori@marvell.com>
Acked-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
The RTE_HNS3_ONLY_1630_FPGA macro is not in use, so delete the code.
Fixes: 2192c428f9a6 ("net/hns3: fix firmware compatibility configuration")
Fixes: bdaf190f8235 ("net/hns3: support link speed autoneg for PF")
Cc: stable@dpdk.org
Signed-off-by: Dongdong Liu <liudongdong3@huawei.com>
The SVE algorithm and NEON algorithm have the same requirements for
nb-desc, but the nb-desc is verified only when using NEON.
Fixes: fa29fe45a7b4 ("net/hns3: support queue start and stop")
Cc: stable@dpdk.org
Signed-off-by: Chengwen Feng <fengchengwen@huawei.com>
Signed-off-by: Dongdong Liu <liudongdong3@huawei.com>
The stats_lock is used to protect statistics update in stats APIs and
periodic task, but current code only protect queue related statistics.
Fixes: a65342d9d5d2 ("net/hns3: fix MAC and queues HW statistics overflow")
Cc: stable@dpdk.org
Signed-off-by: Huisong Li <lihuisong@huawei.com>
Signed-off-by: Dongdong Liu <liudongdong3@huawei.com>
PMD driver will receive a PTP interrupt when receive a PTP packet.
But driver doesn't distinguish it. As a result, many unknown events
are printed when many PTP packets are received on the link. The PTP
interrupt is normal, so this patch doesn't log and ignores it.
Fixes: 38b539d96eb6 ("net/hns3: support IEEE 1588 PTP")
Cc: stable@dpdk.org
Signed-off-by: Huisong Li <lihuisong@huawei.com>
Signed-off-by: Dongdong Liu <liudongdong3@huawei.com>
The purpose of the heartbeat alarm is to keep alive for VF. The mailbox
channel is disabled when VF is reset, and the heartbeat mailbox message
will fail to send. If the reset is not complete, the error information
about the heartbeat sending failure will be printed continuously.
In fact, VF does set alive when VF restore its configuration. So the
heartbeat alarm can be canceled to prepare to start reset and start the
alarm when start service.
Signed-off-by: Huisong Li <lihuisong@huawei.com>
Signed-off-by: Dongdong Liu <liudongdong3@huawei.com>
The 802.11 physical PMA sub-layer defines three media: copper, fiber and
backplane. For PMD, the backplane is similar to the fiber, the main
differences are that backplane doesn't have optical module.
Because the interface of firmware fiber is also applicable to the
backplane, this patch supports the backplane only through simple
extension.
Cc: stable@dpdk.org
Signed-off-by: Chengwen Feng <fengchengwen@huawei.com>
Signed-off-by: Dongdong Liu <liudongdong3@huawei.com>
Currently, the VF LSC capability is obtained from PF driver in
the interrupt mailbox interrupt thread, it is asynchronous.
The VF driver waits for 500ms to get this capability in probe
process.
The primary process will receive a message and do probe in the
interrupt thread context when attach a device in the secondary
process. At this case, VF driver never obtains this capability
from PF.
The root cause is that 'vf->pf_push_lsc_cap' is not updated by
the handling mailbox thread until finishing probe. The reason
this update wouldn't be done is that the handling mailbox interrupt
thread and the probe alarm thread are both in epool thread, and
the probe alarm thread is before the mailbox interrupt thread.
Fixes: 9bc2289fe5ea ("net/hns3: refactor VF LSC event report")
Cc: stable@dpdk.org
Signed-off-by: Huisong Li <lihuisong@huawei.com>
Signed-off-by: Dongdong Liu <liudongdong3@huawei.com>
Lijun Ou and Min Hu currently do not work for the hns3 PMD.
I will do the work, so update the hns3 maintainers.
Signed-off-by: Dongdong Liu <liudongdong3@huawei.com>
Acked-by: Yisen Zhuang <yisen.zhuang@huawei.com>
Use generic GTP PSC header definition in raw handler.
Fixes: 9213c50e36fa ("app/testpmd: support GTP PSC option in raw sets")
Cc: stable@dpdk.org
Signed-off-by: Gregory Etelson <getelson@nvidia.com>
Reviewed-by: Viacheslav Ovsiienko <viacheslavo@nvidia.com>
Acked-by: Aman Singh <aman.deep.singh@intel.com>
Separate flex item destruction function implementation.
Setups with installed JSON development library can use any value in
range [0, FLEX_MAX_PARSERS_NUM - 1] as input flex item ID.
In setups without JSON development library flex item destruction
function is resolved to empty stub.
Fixes: 2d3d84013508 ("app/testpmd: fix flex item flush")
Cc: stable@dpdk.org
Signed-off-by: Gregory Etelson <getelson@nvidia.com>
Reviewed-by: Viacheslav Ovsiienko <viacheslavo@nvidia.com>
Reviewed-by: Andrew Rybchenko <andrew.rybchenko@oktetlabs.ru>
When a port is closed implicitly by the PMD, for example,
if it is a representor port and its master port is detached,
flow indirect actions could remain with their handles no longer valid.
If a newly attached device is assigned the same ID as the closed port,
those indirect actions became accessible again.
Any attempt to use them resulted in an undefined behavior.
Flow flex items had no such issue on close, but had it on detach.
Introduce flush_port_owned_resources() function for consistent
cleanup and call it when a port is closed or detached.
Make it flush flow rules and multicast addresses too
because they logically belong to the port being removed.
Fixes: 55509e3a49fb ("app/testpmd: support shared flow action")
Fixes: 59f3a8acbcdb ("app/testpmd: add flex item commands")
Cc: stable@dpdk.org
Signed-off-by: Dmitry Kozlyuk <dkozlyuk@nvidia.com>
Acked-by: Matan Azrad <matan@nvidia.com>
Acked-by: Andrew Rybchenko <andrew.rybchenko@oktetlabs.ru>
DPDK assumes that users only want AF_XDP socket (XSK) into zero copy
mode when the kernel supports it. However, sometimes kernel driver
doesn't support it well and copy mode is more stable and preferred.
This patch allows using devarg "-a xx:xx.x,force_copy=1" to force the
AF_XDP socket into copy mode.
Signed-off-by: Xiaoyun Li <xiaoyun.li@intel.com>
Reviewed-by: Andrew Rybchenko <andrew.rybchenko@oktetlabs.ru>
Implement NFP3800 card packet transmit function for firmware
with NFDk.
Signed-off-by: Jin Liu <jin.liu@corigine.com>
Signed-off-by: Diana Wang <na.wang@corigine.com>
Signed-off-by: Peng Zhang <peng.zhang@corigine.com>
Signed-off-by: Chaoyong He <chaoyong.he@corigine.com>
Signed-off-by: Niklas Söderlund <niklas.soderlund@corigine.com>
The rte_zmalloc is called in test_crc_calc without null pointer
check. This patch adds null pointer checks on return value of
rte_zmalloc.
Fixes: 9c77b848b1c1 ("test: add CRC computation")
Cc: stable@dpdk.org
Signed-off-by: Hongbo Zheng <zhenghongbo3@huawei.com>
Signed-off-by: Min Hu (Connor) <humin29@huawei.com>
This patch adds support for dump the device private info
from a running application.
It can help developers locate a problem.
Signed-off-by: Min Hu (Connor) <humin29@huawei.com>
Signed-off-by: Dongdong Liu <liudongdong3@huawei.com>
Acked-by: Reshma Pattan <reshma.pattan@intel.com>
Show all non-owned ports when no port mask is specified.
The show-port option, without the mask option,
was showing only the last non-owned port.
Show all the non-owned ports instead.
Fixes: 1dd6cffb6571 ("app/procinfo: provide way to request info on owned ports")
Cc: stable@dpdk.org
Signed-off-by: Subendu Santra <subendu@arista.com>
Acked-by: Stephen Hemminger <stephen@networkplumber.org>
Acked-by: Morten Brørup <mb@smartsharesystems.com>
rwc_non_lf_results->multi_rw, rwc_lf_results->multi_rw, and
rwc_perf_results->multi_rw are accessed at indexes
[0..NUM_TEST-1][0..1][0..NUMTEST-1]. Currently the first index
overflows the array size in struct rwc_perf.
Fixes: c7eb0972e74b ("test/hash: add lock-free r/w concurrency")
Cc: stable@dpdk.org
Signed-off-by: Heinrich Schuchardt <heinrich.schuchardt@canonical.com>
These are functions related to interrupts that have been
in since 20.02 release or earlier.
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
Acked-by: Ray Kinsella <mdr@ashroe.eu>
These API's have been around for a long time and by now are fixed.
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
Acked-by: Ray Kinsella <mdr@ashroe.eu>
The RTE_LOG_REGISTER is not experimental, and the experimental
tag was never enforced on these.
Make rte_log_can_log a fully supported function.
It was introduced nearly 2yrs ago.
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
Acked-by: Ray Kinsella <mdr@ashroe.eu>
The comments in rte_rib6 were cut-and-pasted from rte_rib
and because of that some references to rte_rib_node were
not updated.
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
Acked-by: Vladimir Medvedkin <vladimir.medvedkin@intel.com>
The header files "l3fwd_em.h" and "l3fwd_em_sequential.h" use the
"__rte_always_inline" macro but don't directly include "rte_common.h" to
get the definition of it. This inclusion is not necessary for
compilation, but the lack of it can confuse some indexers - such as
those in eclipse, which reports the lines:
"static __rte_always_inline uint16_t"
as possible definitions of a variable called "uint16_t". This confusion
leads to uint16_t being flagged as an unknown type in all other parts of
the project being indexed, e.g. across all of DPDK code.
Adding in the include of rte_common.h makes it clear to the indexer that
those lines are part of a function definition, and that allows eclipse
to correctly recognise uint16_t as a type from stdint.h
Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
When building without AVX2 support using an older compiler e.g. gcc 4.8
on Centos/RHEL 7, we get build errors due to the use of AVX2 intrinsics.
This is because the compiler does not support
"__attribute__((target(AVX2)))" function attribute. Disable build of
this driver such edge cases.
Generic builds using recent compilers, and all builds with a minimum
baseline of AVX2 are unaffected by this change.
Fixes: aa802b10237c ("dma/idxd: fix AVX2 in non-datapath functions")
Cc: stable@dpdk.org
Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
Tested-by: Yu Jiang <yux.jiang@intel.com>
The build of the raw/ioat driver only occurs when the equivalent dmadev
drivers are disabled. Complications occur when the ioat dmadev is being
built but not the idxd. In this case, only the idxd part of raw/ioat
gets built, but the definition of the logtype is in the ioat part,
causing build errors.
.../raw_ioat_idxd_bus.c.o: In function `idxd_vdev_mmap_wq':
idxd_bus.c:(.text+0x116): undefined reference to `ioat_pmd_logtype'
Fix this by moving the logtype definition to the common C file, and
renaming it to avoid conflicts with a similarly named value in the
dma/ioat driver.
Fixes: ff06fa2cf3ba ("raw/ioat: probe idxd PCI")
Cc: stable@dpdk.org
Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
The inline functions in rte_idxd_rawdev_fns.h make use of rte_errno, but
the header with its definition is not included by that file leading to
build errors.
Fixes: f82c87eb14a4 ("raw/ioat: move idxd functions to separate file")
Cc: stable@dpdk.org
Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
Found by nullfree.cocci.
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
[David: for lpm parts:]
Reviewed-by: Ruifeng Wang <ruifeng.wang@arm.com>
Acked-by: Vladimir Medvedkin <vladimir.medvedkin@intel.com>
[David: for vdpa/mlx5 parts:]
Acked-by: Matan Azrad <matan@nvidia.com>
[David: for dma/dpaa2, raw/ifpga, vdpa/mlx5:]
Acked-by: Tyler Retzlaff <roretzla@linux.microsoft.com>
Reviewed-by: Chengwen Feng <fengchengwen@huawei.com>
[David: reran cocci.sh and updated common/mlx5 and cryptodev asym test]
Signed-off-by: David Marchand <david.marchand@redhat.com>
There are more functions in DPDK which have the semantics
as free() when passed NULL pointer. Also, put the checks
in alphabetical order.
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
Make sure all functions which use the convention that XXX_free(NULL)
is a nop are all documented.
The wording is chosen to match the documentation of free(3).
"If ptr is NULL, no operation is performed."
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
Acked-by: Chengwen Feng <fengchengwen@huawei.com>
[David: squashed with other series updates, unified wording]
Remove extraneous phrase "This API is used to" and use
active instead of passive voice when describing a function.
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
Acked-by: Tyler Retzlaff <roretzla@linux.microsoft.com>
[David: for raw/ioat and dmadev parts:]
Acked-by: Chengwen Feng <fengchengwen@huawei.com>
Acked-by: Conor Walsh <conor.walsh@intel.com>
