dpdk.git
5 years agonet/bnxt: clear HWRM sniffer list for PFs
Randy Schacher [Tue, 1 May 2018 01:06:16 +0000 (18:06 -0700)]
net/bnxt: clear HWRM sniffer list for PFs

Clear HWRM sniffer list for DPDK PFs so that VFs on
DPDK PFs initialize successfully. DPDK PF driver does not
handle HWRM commands from VFs.

Signed-off-by: Randy Schacher <stuart.schacher@broadcom.com>
Signed-off-by: Scott Branden <scott.branden@broadcom.com>
Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
5 years agonet/bnxt: fix usage of vnic id
Ajit Khaparde [Tue, 1 May 2018 01:06:15 +0000 (18:06 -0700)]
net/bnxt: fix usage of vnic id

VNIC ID returned by the FW is a 16-bit field.
We are incorrectly using it as a 32-bit value in few places.
This patch corrects that.

Fixes: daef48efe5e5 ("net/bnxt: support set MTU")
Cc: stable@dpdk.org
Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
Signed-off-by: Scott Branden <scott.branden@broadcom.com>
Reviewed-by: Michael Wildt <michael.wildt@broadcom.com>
Reviewed-by: Randy Schacher <stuart.schacher@broadcom.com>
Reviewed-by: Ray Jui <ray.jui@broadcom.com>
5 years agonet/bnxt: fix Rx mbuf and agg ring leak in dev stop
Xiaoxin Peng [Tue, 1 May 2018 01:06:14 +0000 (18:06 -0700)]
net/bnxt: fix Rx mbuf and agg ring leak in dev stop

In the start/stop_op operation, mbufs allocated for rings were not freed

1) add bnxt_free_tx_mbuf/bnxt_free_rx_mbuf in bnxt_dev_stop_op to free MBUF
   before freeing the rings.
2) MBUF allocation and free routines were not in sync. Allocation uses the
   ring->ring_size including any rounded up and multiple factors while the
   free routine uses the requested queue size.

Fixes: c09f57b49c13 ("net/bnxt: add start/stop/link update operations")
Cc: stable@dpdk.org
Signed-off-by: Jay Ding <jay.ding@broadcom.com>
Signed-off-by: Scott Branden <scott.branden@broadcom.com>
Reviewed-by: Ray Jui <ray.jui@broadcom.com>
Reviewed-by: Randy Schacher <stuart.schacher@broadcom.com>
Signed-off-by: Xiaoxin Peng <xiaoxin.peng@broadcom.com>
Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
5 years agonet/bnxt: fix to reset status of initialization
Ajit Khaparde [Tue, 1 May 2018 01:06:13 +0000 (18:06 -0700)]
net/bnxt: fix to reset status of initialization

clear flag on stop at proper location to avoid race conditions.

Fixes: ed2ced6fe927 ("net/bnxt: check initialization before accessing stats")
Cc: stable@dpdk.org
Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
5 years agonet/bnxt: return error in stats if init is not complete
Jay Ding [Tue, 1 May 2018 01:06:12 +0000 (18:06 -0700)]
net/bnxt: return error in stats if init is not complete

return error if init is not complete before accessing stats.

Fixes: ed2ced6fe927 ("net/bnxt: check initialization before accessing stats")
Cc: stable@dpdk.org
Signed-off-by: Jay Ding <jay.ding@broadcom.com>
Signed-off-by: Scott Branden <scott.branden@broadcom.com>
Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
Reviewed-by: Randy Schacher <stuart.schacher@broadcom.com>
Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
5 years agonet/bnxt: fix MTU calculation
Ajit Khaparde [Tue, 1 May 2018 01:06:11 +0000 (18:06 -0700)]
net/bnxt: fix MTU calculation

We were not considering the case of nested VLANs while
calculating MTU. This patch takes care of the same.

Fixes: daef48efe5e5 ("net/bnxt: support set MTU")
Cc: stable@dpdk.org
Signed-off-by: Qingmin Liu <qingmin.liu@broadcom.com>
Signed-off-by: Scott Branden <scott.branden@broadcom.com>
Reviewed-by: Jay Ding <jay.ding@broadcom.com>
Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
Reviewed-by: Randy Schacher <stuart.schacher@broadcom.com>
Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
5 years agonet/bnxt: set MTU in dev config for jumbo packets
Qingmin Liu [Tue, 1 May 2018 01:06:10 +0000 (18:06 -0700)]
net/bnxt: set MTU in dev config for jumbo packets

MTU setting does not take effect after rte_eth_dev_configure
is called with jumbo enable unless it is configured using the
set_mtu dev_op.

Fixes: daef48efe5e5 ("net/bnxt: support set MTU")
Cc: stable@dpdk.org
Signed-off-by: Qingmin Liu <qingmin.liu@broadcom.com>
Signed-off-by: Scott Branden <scott.branden@broadcom.com>
Reviewed-by: Jay Ding <jay.ding@broadcom.com>
Reviewed-by: Randy Schacher <stuart.schacher@broadcom.com>
Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
5 years agonet/bnxt: validate structs and pointers before use
Rahul Gupta [Tue, 1 May 2018 01:06:09 +0000 (18:06 -0700)]
net/bnxt: validate structs and pointers before use

Validate pointers aren't pointing to uninitialized areas
including txq and rxq before using them to avoid
bnxt driver from crashing.

Signed-off-by: Rahul Gupta <rahul.gupta@broadcom.com>
Signed-off-by: Jay Ding <jay.ding@broadcom.com>
Signed-off-by: Scott Branden <scott.branden@broadcom.com>
Reviewed-by: Ray Jui <ray.jui@broadcom.com>
Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
Reviewed-by: Randy Schacher <stuart.schacher@broadcom.com>
Tested-by: Randy Schacher <stuart.schacher@broadcom.com>
5 years agonet/bnxt: update returned error on invalid max ring
Jay Ding [Tue, 1 May 2018 01:06:08 +0000 (18:06 -0700)]
net/bnxt: update returned error on invalid max ring

Return EINVAL instead of ENOSPC when invalid queue_idx passed in
during rx and tx queue_setup_op routines.

Signed-off-by: Jay Ding <jay.ding@broadcom.com>
Signed-off-by: Scott Branden <scott.branden@broadcom.com>
Reviewed-by: Ray Jui <ray.jui@broadcom.com>
Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
5 years agonet/bnxt: rename driver version from Cumulus to NetXtreme
Scott Branden [Tue, 1 May 2018 01:06:07 +0000 (18:06 -0700)]
net/bnxt: rename driver version from Cumulus to NetXtreme

Rename driver version from "Broadcom Cumulus driver" to
"Broadcom NetXtreme driver" to reflect this driver is applicable to
NetXtreme family beyond Cumulus.

Signed-off-by: Scott Branden <scott.branden@broadcom.com>
Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
5 years agonet/bnxt: rename function checking MAC address
Scott Branden [Tue, 1 May 2018 01:06:06 +0000 (18:06 -0700)]
net/bnxt: rename function checking MAC address

rename check_zero_bytes to bnxt_check_zero_bytes to match proper prefix.

Signed-off-by: Scott Branden <scott.branden@broadcom.com>
Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
5 years agonet/bnxt: add support for LSC interrupt event
Qingmin Liu [Tue, 1 May 2018 01:06:05 +0000 (18:06 -0700)]
net/bnxt: add support for LSC interrupt event

Add support to bnxt driver to register RTE_ETH_EVENT_INTR_LSC
event and monitor physical link status.

Signed-off-by: Qingmin Liu <qingmin.liu@broadcom.com>
Signed-off-by: Scott Branden <scott.branden@broadcom.com>
Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
Reviewed-by: Randy Schacher <stuart.schacher@broadcom.com>
5 years agonet/mlx4: fix UDP flow rule limitation enforcement
Adrien Mazarguil [Fri, 4 May 2018 15:17:44 +0000 (17:17 +0200)]
net/mlx4: fix UDP flow rule limitation enforcement

For some unknown reason, priorities do not have any effect on flow rules
that happen to match UDP destination ports. Those are always matched first
regardless.

This patch is a workaround that enforces this limitation at the PMD level;
such flow rules can only be created at the highest priority level for
correctness.

Fixes: a5171594fc3b ("net/mlx4: expose support for flow rule priorities")
Cc: stable@dpdk.org
Signed-off-by: Adrien Mazarguil <adrien.mazarguil@6wind.com>
5 years agonet/mlx5: fix flow validation
Nélio Laranjeiro [Thu, 3 May 2018 09:31:38 +0000 (11:31 +0200)]
net/mlx5: fix flow validation

Item spec and last are wrongly compared to the NIC capability causing a
validation failure when the mask is null.
This validation function should only verify the user is not configuring
unsupported matching fields.

Fixes: 2097d0d1e2cc ("net/mlx5: support basic flow items and actions")
Cc: stable@dpdk.org
Signed-off-by: Nelio Laranjeiro <nelio.laranjeiro@6wind.com>
5 years agonet/mlx5: fix probe return value polarity
Shahaf Shuler [Thu, 3 May 2018 07:59:37 +0000 (10:59 +0300)]
net/mlx5: fix probe return value polarity

mlx5 prefixed function returns a negative errno value.
the error handler on mlx5_pci_probe is doing the same.

Fixes: a6d83b6a9209 ("net/mlx5: standardize on negative errno values")

Signed-off-by: Shahaf Shuler <shahafs@mellanox.com>
Acked-by: Nelio Laranjeiro <nelio.laranjeiro@6wind.com>
5 years agonet/mlx5: fix socket connection return value
Shahaf Shuler [Thu, 3 May 2018 07:59:36 +0000 (10:59 +0300)]
net/mlx5: fix socket connection return value

Upon success, mlx5_socket_connect should return the fd descriptor of the
primary process

Fixes: a6d83b6a9209 ("net/mlx5: standardize on negative errno values")

Signed-off-by: Yongseok Koh <yskoh@mellanox.com>
Signed-off-by: Shahaf Shuler <shahafs@mellanox.com>
Acked-by: Nelio Laranjeiro <nelio.laranjeiro@6wind.com>
5 years agonet/mlx5: add Rx and Tx tuning parameters
Shahaf Shuler [Tue, 1 May 2018 09:58:49 +0000 (12:58 +0300)]
net/mlx5: add Rx and Tx tuning parameters

A new ethdev API was exposed by
commit 3be82f5cc5e3 ("ethdev: support PMD-tuned Tx/Rx parameters")

Enabling the PMD to provide default parameters in case no strict request
from application in order to improve the out of the box experience.

While the current API lacks the means for the PMD to provide the best
possible value, providing the best default the PMD can guess.
The values are based on Mellanox performance report and depends on the
underlying NIC capabilities.

Signed-off-by: Shahaf Shuler <shahafs@mellanox.com>
Acked-by: Nelio Laranjeiro <nelio.laranjeiro@6wind.com>
5 years agonet/mlx5: fix ethtool link setting call order
Shahaf Shuler [Tue, 1 May 2018 09:58:48 +0000 (12:58 +0300)]
net/mlx5: fix ethtool link setting call order

According to ethtool_link_setting API recommendation ETHTOOL_GLINKSETTINGS
should be called before ETHTOOL_GSET as the later one deprecated.

Fixes: f47ba80080ab ("net/mlx5: remove kernel version check")

Signed-off-by: Shahaf Shuler <shahafs@mellanox.com>
Acked-by: Nelio Laranjeiro <nelio.laranjeiro@6wind.com>
5 years agonet/i40e: fix flow RSS key array error
Wei Zhao [Mon, 7 May 2018 05:34:00 +0000 (13:34 +0800)]
net/i40e: fix flow RSS key array error

This is a bug introduced into RSS key array span.

Fixes: ac8d22de2394 ("ethdev: flatten RSS configuration in flow API")

Signed-off-by: Wei Zhao <wei.zhao1@intel.com>
Acked-by: Qi Zhang <qi.z.zhang@intel.com>
5 years agonet/i40e: fix missing some offload capabilities
Yanglong Wu [Sat, 28 Apr 2018 09:37:29 +0000 (17:37 +0800)]
net/i40e: fix missing some offload capabilities

MULTI_SEGS and JUMBO_FRAME offload capability should be exposed
both VF and PF since i40e does support it.

Fixes: c3ac7c5b0b8a ("net/i40e: convert to new Rx offloads API")
Fixes: 7497d3e2f777 ("net/i40e: convert to new Tx offloads API")

Signed-off-by: Yanglong Wu <yanglong.wu@intel.com>
Acked-by: Qi Zhang <qi.z.zhang@intel.com>
5 years agonet/e1000: add minimum support for Broadcom 54616 PHY
Chas Williams [Wed, 6 Dec 2017 23:55:28 +0000 (18:55 -0500)]
net/e1000: add minimum support for Broadcom 54616 PHY

If we find a Broadcom 54616, handle as a e1000_phy_none assuming that
the NIC reset has initialized the PHY to a sane state.

Signed-off-by: Chas Williams <chas3@att.com>
Acked-by: Wenzhuo Lu <wenzhuo.lu@intel.com>
5 years agonet/i40e: add workaround promiscuous disable
Rosen Xu [Wed, 18 Apr 2018 08:11:41 +0000 (16:11 +0800)]
net/i40e: add workaround promiscuous disable

In scenario of Kernel Driver runs on PF and PMD runs on VF, PMD exit
doesn't disable promiscuous mode, this will cause vlan filter set by
Kernel Driver will not take effect.

This patch will fix it, add promiscuous disable at device disable.

Signed-off-by: Rosen Xu <rosen.xu@intel.com>
Acked-by: Qi Zhang <qi.z.zhang@intel.com>
5 years agonet/ixgbe: add API to update SBP bit
Shweta Choudaha [Tue, 10 Apr 2018 15:42:46 +0000 (16:42 +0100)]
net/ixgbe: add API to update SBP bit

Add ixgbe API to enable SBP bit in FCTRL register
to allow receiving packets that may otherwise be
considered length errors by ixgbe and dropped

Signed-off-by: Shweta Choudaha <shweta.choudaha@att.com>
Reviewed-by: Chas Williams <chas3@att.com>
Reviewed-by: Luca Boccassi <bluca@debian.org>
Acked-by: Wenzhuo Lu <wenzhuo.lu@intel.com>
5 years agonet/virtio-user: fix hugepage files enumeration
Jianfeng Tan [Thu, 26 Apr 2018 15:34:07 +0000 (15:34 +0000)]
net/virtio-user: fix hugepage files enumeration

After the commit 2a04139f66b4 ("eal: add single file segments option"),
one hugepage file could contain multiple hugepages which are further
mapped to different memory regions.

Original enumeration implementation cannot handle this situation.

This patch filters out the duplicated files; and adjust the size after
the enumeration.

Fixes: 6a84c37e3975 ("net/virtio-user: add vhost-user adapter layer")

Signed-off-by: Jianfeng Tan <jianfeng.tan@intel.com>
Acked-by: Maxime Coquelin <maxime.coquelin@redhat.com>
5 years agonet/vhost: initialise device as inactive
Ciara Loftus [Thu, 3 May 2018 13:17:03 +0000 (14:17 +0100)]
net/vhost: initialise device as inactive

rte_eth_vhost_get_vid_from_port_id returns a value of 0 if
called before the first call to the new_device callback.
A vid value >=0 suggests the device is active which is not
the case in this instance. Initialise vid to a negative
value to prevent this.

Fixes: ee584e9710b9 ("vhost: add driver on top of the library")
Cc: stable@dpdk.org
Signed-off-by: Ciara Loftus <ciara.loftus@intel.com>
Reviewed-by: Maxime Coquelin <maxime.coquelin@redhat.com>
5 years agovhost/crypto: fix symmetric ciphering
Fan Zhang [Mon, 30 Apr 2018 10:31:17 +0000 (11:31 +0100)]
vhost/crypto: fix symmetric ciphering

A bracket was misplaced in a condition check, this patch
fixes it.

Coverity issue: 277232, 277237
Fixes: 3bb595ecd682 ("vhost/crypto: add request handler")

Signed-off-by: Fan Zhang <roy.fan.zhang@intel.com>
Reviewed-by: Maxime Coquelin <maxime.coquelin@redhat.com>
5 years agovhost: fix header copy to discontiguous desc buffer
Maxime Coquelin [Mon, 30 Apr 2018 09:35:23 +0000 (11:35 +0200)]
vhost: fix header copy to discontiguous desc buffer

In the loop to copy virtio-net header to the descriptor buffer,
destination pointer was incremented instead of the source
pointer.

Fixes: fb3815cc614d ("vhost: handle virtually non-contiguous buffers in Rx-mrg")
Fixes: 6727f5a739b6 ("vhost: handle virtually non-contiguous buffers in Rx")
Cc: stable@dpdk.org
Signed-off-by: Maxime Coquelin <maxime.coquelin@redhat.com>
5 years agoexamples/vhost: fix header copy to discontiguous desc buffer
Maxime Coquelin [Mon, 30 Apr 2018 09:35:22 +0000 (11:35 +0200)]
examples/vhost: fix header copy to discontiguous desc buffer

In the loop to copy virtio-net header to the descriptor buffer,
destination pointer was incremented instead of the source
pointer.

Coverity issue: 277240
Fixes: 82c93a567d3b ("examples/vhost: move to safe GPA translation API")

Signed-off-by: Maxime Coquelin <maxime.coquelin@redhat.com>
5 years agovhost: fix typo in comment
Tonghao Zhang [Fri, 27 Apr 2018 15:19:45 +0000 (08:19 -0700)]
vhost: fix typo in comment

Fixes: 3670686ab99f ("vhost: fix race for connection fd")
Cc: stable@dpdk.org
Signed-off-by: Tonghao Zhang <xiangxia.m.yue@gmail.com>
Acked-by: Maxime Coquelin <maxime.coquelin@redhat.com>
5 years agovhost: fix crash on closing in client mode
Tonghao Zhang [Fri, 27 Apr 2018 15:19:44 +0000 (08:19 -0700)]
vhost: fix crash on closing in client mode

when rte_vhost_driver_unregister detstroy the vsocket, we
should set it to NULL after freeing it, because in client mode,
the conn may be added to reconnect thread while vsocket is
destroyed. In one case, if qemu create vhostuser port as a
server with the same unix path, the reconnect thread will
reconnect to it while vsocket is destroyed.

To fix this:
1. set vsocket to NULL after free it.
2. remove the reconnection from reconnection thread in suitable
   position.

Cc: stable@dpdk.org
Signed-off-by: Tonghao Zhang <xiangxia.m.yue@gmail.com>
Reviewed-by: Maxime Coquelin <maxime.coquelin@redhat.com>
5 years agovhost: fix dead lock on closing in server mode
Tonghao Zhang [Fri, 27 Apr 2018 15:19:43 +0000 (08:19 -0700)]
vhost: fix dead lock on closing in server mode

When qemu close the unix socket fd of the vhostuser as a
server, and then immediately delete the vhostuser port on
openvswitch. There will be a deadlock.

A thread (fdset event thread):       B thread:
1. fdset_event_dispatch              rte_vhost_driver_unregister
2. set the fd busy to 1.             lock vsocket->conn_mutex
3. vhost_user_read_cb                fdset_del waits busy changed to 0.
4. vhost peer closed, remove the
   conn from vsocket->conn_list:
   lock vsocket->conn_mutex

5. set the fd busy to 0

Fixes: 65388b43f592 ("vhost: fix fd leaks for vhost-user server mode")
Cc: stable@dpdk.org
Signed-off-by: Tonghao Zhang <xiangxia.m.yue@gmail.com>
Acked-by: Maxime Coquelin <maxime.coquelin@redhat.com>
5 years agodoc/vhost: update zero copy performance tip
Junjie Chen [Thu, 26 Apr 2018 08:13:20 +0000 (16:13 +0800)]
doc/vhost: update zero copy performance tip

In VM2NIC case zero copy may need some tuning to get best performance.
This patch describes the zero copy starved case and provides a tuning
tip.

Signed-off-by: Junjie Chen <junjie.j.chen@intel.com>
Acked-by: Maxime Coquelin <maxime.coquelin@redhat.com>
5 years agonet/thunderx: remove deprecated Txq flags
Pavan Nikhilesh [Thu, 3 May 2018 11:32:54 +0000 (17:02 +0530)]
net/thunderx: remove deprecated Txq flags

Signed-off-by: Pavan Nikhilesh <pbhagavatula@caviumnetworks.com>
5 years agonet/octeontx: fix missing offload flags
Pavan Nikhilesh [Thu, 3 May 2018 11:31:15 +0000 (17:01 +0530)]
net/octeontx: fix missing offload flags

Fix missing DEV_RX_OFFLOAD_CHECKSUM flag in RX offloads.
Remove depricated txq_flags field.

Fixes: a92870896b4a ("net/octeontx: use the new offload APIs")

Signed-off-by: Pavan Nikhilesh <pbhagavatula@caviumnetworks.com>
5 years agonet/bnxt: remove unused Txq flags
Ajit Khaparde [Tue, 1 May 2018 22:50:52 +0000 (15:50 -0700)]
net/bnxt: remove unused Txq flags

We are still using the txq_flags which is no longer needed with the
new offload API. Cleaning it up.

Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
5 years agonet/fm10k: remove dependence on Tx queue flags
Qi Zhang [Wed, 2 May 2018 03:16:26 +0000 (11:16 +0800)]
net/fm10k: remove dependence on Tx queue flags

Since we move to new offload APIs, txq_flags is no long needed.
This patch remove the dependence on that.

Fixes: 30f3ce999e6a ("net/fm10k: convert to new Tx offloads API")

Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com>
5 years agonet/e1000: remove dependence on Tx queue flags
Qi Zhang [Wed, 2 May 2018 03:16:25 +0000 (11:16 +0800)]
net/e1000: remove dependence on Tx queue flags

Since we move to new offload APIs, txq_flags is no long needed.
This patch remove the dependence on that.

Fixes: e5c05e6590ea ("net/e1000: convert to new Tx offloads API")

Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com>
5 years agonet/ixgbe: remove dependence on Tx queue flags
Qi Zhang [Wed, 2 May 2018 03:16:24 +0000 (11:16 +0800)]
net/ixgbe: remove dependence on Tx queue flags

Since we move to new offload APIs, txq_flags is no long needed.
This patch remove the dependence on that.

Fixes: 51215925a32f ("net/ixgbe: convert to new Tx offloads API")

Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com>
5 years agonet/sfc: rely on ethdev mapping of Tx offloads to TxQ flags
Andrew Rybchenko [Thu, 3 May 2018 10:17:38 +0000 (11:17 +0100)]
net/sfc: rely on ethdev mapping of Tx offloads to TxQ flags

There is no necessity to fill in TxQ flags since ethdev maps
Tx offloads to TxQ flags on device info get for apps which are
not converted yet to Tx offloads API.

Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>
5 years agoethdev: convert Tx offloads to Tx queue config
Qi Zhang [Thu, 3 May 2018 06:03:25 +0000 (14:03 +0800)]
ethdev: convert Tx offloads to Tx queue config

Tx offload will be converted to txq_flags automatically during
rte_eth_dev_info_get and rte_eth_tx_queue_info_get. So PMD can
clean the code to get rid of txq_flags at all while keep old APP
not be impacted.

Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
Tested-by: Andrew Rybchenko <arybchenko@solarflare.com>
Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com>
5 years agonet/i40e: fix Tx queue info get
Qi Zhang [Thu, 3 May 2018 02:01:14 +0000 (10:01 +0800)]
net/i40e: fix Tx queue info get

Add missing Tx queue offload assignment in i40e_txq_info_get.

Fixes: 7497d3e2f777 ("net/i40e: convert to new Tx offloads API")

Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
5 years agoip_frag: fix some debug logs
Li Han [Wed, 18 Apr 2018 08:51:37 +0000 (04:51 -0400)]
ip_frag: fix some debug logs

In ip_frag_process, some IP_FRAG_LOG content is wrong.

Fixes: 4f1a8f633862 ("ip_frag: add IPv6 reassembly")
Cc: stable@dpdk.org
Signed-off-by: Li Han <han.li1@zte.com.cn>
Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com>
5 years agoapp/testpmd: fix initialization
Anatoly Burakov [Mon, 14 May 2018 15:46:42 +0000 (16:46 +0100)]
app/testpmd: fix initialization

Patch adding configurable locking has reshuffled some init stages, but
did not put them back in correct order. Fix order of init by moving
everything that was before arguments parsing into correct places.

Fixes: e505d84c64ab ("app/testpmd: make locking memory configurable")

Reported-by: Thomas Monjalon <thomas@monjalon.net>
Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
Tested-by: Ali Alnubani <alialnu@mellanox.com>
Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com>
5 years agobpf: add missed experimental tags
Konstantin Ananyev [Mon, 14 May 2018 13:15:42 +0000 (14:15 +0100)]
bpf: add missed experimental tags

- add EXPERIMENTAL tag for the section in MAINTAINERS.
- add EXPERIMENTAL tag to BPF public API files.
- add attribute __rte_experimental to BPF public API declarations.

Fixes: 94972f35a02e ("bpf: add BPF loading and execution framework")
Fixes: 5dba93ae5f2d ("bpf: add ability to load eBPF program from ELF object file")
Fixes: a93ff62a8938 ("bpf: introduce basic Rx/Tx filters")

Reported-by: Thomas Monjalon <thomas@monjalon.net>
Signed-off-by: Konstantin Ananyev <konstantin.ananyev@intel.com>
5 years agotest/pipeline: fix type of table entry parameter
Jasvinder Singh [Mon, 14 May 2018 13:35:24 +0000 (14:35 +0100)]
test/pipeline: fix type of table entry parameter

Fixes the wrong argument in table action miss function.

Fixes: 4c387fcdf777 ("pipeline: add new functions for action handlers")
Cc: stable@dpdk.org
Signed-off-by: Jasvinder Singh <jasvinder.singh@intel.com>
5 years agoexamples/ipsec-secgw: fix IP address parsing
Kirill Rybalchenko [Mon, 14 May 2018 10:00:51 +0000 (11:00 +0100)]
examples/ipsec-secgw: fix IP address parsing

In strlcpy function parameters there was no allowance for
null terminator, so ip address was copied without last character.

Fixes: ae943ebe1ed3 ("examples/ipsec-secgw: replace strncpy with strlcpy")
Cc: stable@dpdk.org
Signed-off-by: Kirill Rybalchenko <kirill.rybalchenko@intel.com>
Acked-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>
5 years agoapp/testpmd: fix log after detach
Zhiyong Yang [Mon, 14 May 2018 03:20:48 +0000 (11:20 +0800)]
app/testpmd: fix log after detach

The second parameter "name" in the function rte_eth_dev_detach
has been already redefined as "char *name __rte_unused",
"port_id" is printed instead of "name" in testpmd.

Fixes: b65ecf199324 ("devargs: rename legacy API")
Cc: stable@dpdk.org
Signed-off-by: Zhiyong Yang <zhiyong.yang@intel.com>
Acked-by: Matan Azrad <matan@mellanox.com>
5 years agoconfig: increase max lcore to 16 for DPAA
Shreyansh Jain [Mon, 14 May 2018 10:05:00 +0000 (15:35 +0530)]
config: increase max lcore to 16 for DPAA

With this change, DPAA and DPAA2 configuration compiled applications
are interoperable without any impact on the functional behavior.

Signed-off-by: Shreyansh Jain <shreyansh.jain@nxp.com>
5 years agoraw/ifpga/base: fix build with icc
Tianfei Zhang [Mon, 14 May 2018 09:58:37 +0000 (17:58 +0800)]
raw/ifpga/base: fix build with icc

fix compile error on icc compiler

Fixes: 56bb54ea1bdf ("raw/ifpga/base: add Intel FPGA OPAE share code")

Signed-off-by: Tianfei Zhang <tianfei.zhang@intel.com>
Tested-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>
5 years agoexamples/performance-thread: fix return type of threads
David Hunt [Mon, 14 May 2018 02:53:15 +0000 (03:53 +0100)]
examples/performance-thread: fix return type of threads

The function pthread_create() expects void *(*func) (void *)
for function pointer, however, lthread_func_t was defined as
void (*func) (void *), so now gcc 8.1 warns that the cast is
incorrect, causing a compilation failure. This patch changes
the declaration of lthread_func_t from returning a void to
returning a void*, and then changes the sample app in the
relevant places that are affected by the typedef change.

Fixes: 116819b9ed0d ("examples/performance-thread: add lthread subsystem")
Cc: stable@dpdk.org
Signed-off-by: David Hunt <david.hunt@intel.com>
Acked-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>
5 years agotest: fix build with icc
Abhinandan Gujjar [Mon, 14 May 2018 06:50:22 +0000 (12:20 +0530)]
test: fix build with icc

This patch provides fix for icc compilation issue with event
crypto adapter test application.
Currently, number of elements in the session mempool is defined
by max_nb_sessions & this is replaced with a macro.

Fixes: 3c2c535ecfc0 ("test: add event crypto adapter auto-test")

Reported-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>
Signed-off-by: Abhinandan Gujjar <abhinandan.gujjar@intel.com>
Acked-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>
5 years agoapp/testpmd: fix asynchronic port removal
Matan Azrad [Thu, 3 May 2018 10:31:48 +0000 (10:31 +0000)]
app/testpmd: fix asynchronic port removal

When a removable device is plugged-out, a RMV interrupt is invoked and
the application can catch the event in order to stop the device
management.

The Testpmd wrong behavior in this case is to detach the removed device
using the EAL detach API.

The EAL API does not invalidate the ethdev port and the port keeps
appearing as valid from the ethdev point of view.

Thus, the next operations for the ethtev port X may trigger an invalid
rte_device access. For example, calling "show port info X" may cause
segfault.

Moreover, the removed port is not removed from the Testpmd data-path
structures. Therefore, the invalid device may still be used by the
Testpmd data-path.

Call the Testpmd detach_port() function which uses the ethdev detach
API, and prepare the Testpmd forward ports database for a new
forwarding session without the detached port.

Fixes: 284c908cc588 ("app/testpmd: request device removal interrupt")
Cc: stable@dpdk.org
Signed-off-by: Matan Azrad <matan@mellanox.com>
5 years agoapp/testpmd: fix removed device link status asking
Matan Azrad [Thu, 3 May 2018 10:31:47 +0000 (10:31 +0000)]
app/testpmd: fix removed device link status asking

In the RMV device event callback, there is a call for the removed
device stop operation which triggers a link status operation for the
removed device.

It may casue an error from the removed device PMD.

Skip the link status operation in the above described case.

Fixes: 284c908cc588 ("app/testpmd: request device removal interrupt")
Cc: stable@dpdk.org
Signed-off-by: Matan Azrad <matan@mellanox.com>
5 years agoapp/testpmd: fix synchronic port hotplug
Matan Azrad [Thu, 3 May 2018 10:31:46 +0000 (10:31 +0000)]
app/testpmd: fix synchronic port hotplug

When the user uses the synchronic hot-plug commands, attach\detach, in
order to insert\remove a port from the system, the forward ports list
update is missed in the current implementation.

Thus, an invalid port may be used for data-path in case of detach
because the detached port was not removed from the forward port list.

In addition, a new port is not used for data-path in case of attach, as the
default behavior of Testpmd, because the attached port was not inserted
to the forward port list.

Update the forward port list in the above cases to allow the correct
port usage for data-path in the next packet forwarding start.

Fixes: edab33b1c01d ("app/testpmd: support port hotplug")
Cc: stable@dpdk.org
Signed-off-by: Matan Azrad <matan@mellanox.com>
5 years agoapp/testpmd: fix forward ports Rx flush
Matan Azrad [Thu, 3 May 2018 10:31:45 +0000 (10:31 +0000)]
app/testpmd: fix forward ports Rx flush

A port Rx queue flush is done when the packet forwarding starts in
order to clean the port statistics for a new traffic session.

The flush operation is wrongly called before the update of the new
forward ports, and may fail due to flush operation for an invalid port
configured by the old session.

Move the new forward port setup to be done before the Rx queue flush.

Fixes: 7741e4cf16c0 ("app/testpmd: VMDq and DCB updates")
Cc: stable@dpdk.org
Signed-off-by: Matan Azrad <matan@mellanox.com>
5 years agoapp/testpmd: fix forward ports update
Matan Azrad [Thu, 3 May 2018 10:31:44 +0000 (10:31 +0000)]
app/testpmd: fix forward ports update

When the forward ports are changed either by new portlist\portmask
configurations or by a port detachment, all the old forward streams
are freed and new streams are allocated to be aligned with the new
forward ports.

If the number of the forward ports drops to 0, there is an attempt
to wrongly allocate 0 memory for the streams.

Skip the streams memory allocation if no forward ports are configured.

Fixes: ce8d561418d4 ("app/testpmd: add port configuration settings")
Cc: stable@dpdk.org
Signed-off-by: Matan Azrad <matan@mellanox.com>
5 years agoapp/testpmd: fix valid ports prints
Matan Azrad [Thu, 3 May 2018 10:31:43 +0000 (10:31 +0000)]
app/testpmd: fix valid ports prints

There are several cases of an invalid port data access that causes the
printing of all the valid ports, for example, when the user asks to
receive a port information of an invalid port.

Wrongly, the port with id 0 is printed in all the above described
cases, regardless of its validity.

Print port 0 only if it is valid as done for the rest of the ports.

Fixes: af75078fece3 ("first public release")
Fixes: b6ea6408fbc7 ("ethdev: store numa_node per device")
Fixes: edab33b1c01d ("app/testpmd: support port hotplug")
Cc: stable@dpdk.org
Signed-off-by: Matan Azrad <matan@mellanox.com>
5 years agoapp/testpmd: fix slave port detection
Matan Azrad [Wed, 28 Mar 2018 11:17:53 +0000 (11:17 +0000)]
app/testpmd: fix slave port detection

Testpmd allows to create and control bonding devices by run time
command lines using the bonding PMD API.

Some bonding device slaves operations (close, stop, etc) should not be
used by the application and must be managed by the bonding PMD.

Thus, Testpmd manages slave flags to prevent the special operations
calls and when a slave is added to bonding device by a run time command
line, the flag is set.

There is one more way to define the slaves for a bonding device using
EAL command line and Testpmd doesn't set the slave flag in this case
what causes to the special operations to be called by Testpmd.

Add one more check to detect bonding slave device.

Fixes: 41b05095c4d1 ("app/testpmd: fix bonding start")
Fixes: 0e545d3047fe ("app/testpmd: check stopping port is not in bonding")
Cc: stable@dpdk.org
Signed-off-by: Matan Azrad <matan@mellanox.com>
Acked-by: Bernard Iremonger <bernard.iremonger@intel.com>
5 years agoapp/testpmd: disable memory locking on FreeBSD
Anatoly Burakov [Thu, 3 May 2018 12:38:20 +0000 (13:38 +0100)]
app/testpmd: disable memory locking on FreeBSD

If mlockall() is called while allocated VA space is bigger than
amount of available RAM, FreeBSD kernel will deadlock and need
a hard reboot. We do allocate big amounts of memory because of
how new memory subsystem works, so calling mlockall() will cause
a deadlock. So, disable mlockall() by default on FreeBSD.

Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com>
5 years agoapp/testpmd: make locking memory configurable
Anatoly Burakov [Thu, 3 May 2018 12:38:19 +0000 (13:38 +0100)]
app/testpmd: make locking memory configurable

Add two new command-line parameters for either enabling or
disabling locking all memory at app startup.

Signed-off-by: Jianfeng Tan <jianfeng.tan@intel.com>
Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
Acked-by: Remy Horton <remy.horton@intel.com>
Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com>
5 years agolib: clear experimental version tag in linker scripts
Ferruh Yigit [Fri, 27 Apr 2018 15:17:55 +0000 (16:17 +0100)]
lib: clear experimental version tag in linker scripts

Remove version tag from experimental block in linker version scripts
(.map files).

That label is not used by linker and information only. It is useful
for version blocks but not useful for experimental block but confusing.
Removing those labels.

Signed-off-by: Ferruh Yigit <ferruh.yigit@intel.com>
Acked-by: Neil Horman <nhorman@tuxdriver.com>
5 years agoipc: fix duplicate string copy in async request
Anatoly Burakov [Thu, 3 May 2018 10:35:26 +0000 (11:35 +0100)]
ipc: fix duplicate string copy in async request

Coverity issue: 272582
Fixes: 2147c0950580 ("ipc: clean up code")

Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
Acked-by: Harry van Haaren <harry.van.haaren@intel.com>
5 years agomem: fix unmapping and marking segments as free
Anatoly Burakov [Thu, 3 May 2018 10:11:27 +0000 (11:11 +0100)]
mem: fix unmapping and marking segments as free

Currently, page deallocation might fail if allocator cannot get page
fd, which will leave VA space still mapped, and will also not mark
page as free.

Fix page deallocation function to always unmap space before trying
to get rid of the page itself, and always mark page as free even if
page deallocation failed.

Fixes: a5ff05d60fc5 ("mem: support unmapping pages at runtime")
Fixes: 1a7dc2252f28 ("mem: revert to using flock and add per-segment lockfiles")

Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com>
5 years agomem: fix return code of freeing segment on failure
Anatoly Burakov [Thu, 3 May 2018 10:11:26 +0000 (11:11 +0100)]
mem: fix return code of freeing segment on failure

Return value should be zero for success, but if unlock and unlink
have succeeded, return value was 1, which triggered failure message
in calling code.

Fixes: a5ff05d60fc5 ("mem: support unmapping pages at runtime")

Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com>
5 years agomem: fix index for unmapping segments on failure
Anatoly Burakov [Thu, 3 May 2018 10:11:25 +0000 (11:11 +0100)]
mem: fix index for unmapping segments on failure

Segment index was calculated incorrectly, causing free_seg to
attempt to free segments that do not exist.

Fixes: a5ff05d60fc5 ("mem: support unmapping pages at runtime")

Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
Tested-by: Yong Liu <yong.liu@intel.com>
Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com>
5 years agomem: fix potential underflow on mem size calculation
Anatoly Burakov [Thu, 3 May 2018 13:03:19 +0000 (14:03 +0100)]
mem: fix potential underflow on mem size calculation

If total memory is already bigger than max memory, an underflow
will occur on subtraction. Fix it by simply stopping whenever
we already have amount of memory that is bigger than maximum.

Fixes: 66cc45e293ed ("mem: replace memseg with memseg lists")

Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
5 years agomemzone: document reserving zero-length memzones
Anatoly Burakov [Thu, 3 May 2018 16:25:58 +0000 (17:25 +0100)]
memzone: document reserving zero-length memzones

Currently, reserving a memzone with length set to 0 will not trigger
any memory allocations, and memzone will instead be looking through
already allocated memory only. Document this limitation.

Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
5 years agomemzone: fix size on reserving biggest memzone
Anatoly Burakov [Thu, 3 May 2018 16:00:47 +0000 (17:00 +0100)]
memzone: fix size on reserving biggest memzone

Size of malloc heap elements include overhead, which should not
be counted as part of memzone.

Fixes: fafcc11985a2 ("mem: rework memzone to be allocated by malloc")
Cc: stable@dpdk.org
Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com>
5 years agomemzone: fix race condition on alloc failure
Anatoly Burakov [Thu, 3 May 2018 08:59:28 +0000 (09:59 +0100)]
memzone: fix race condition on alloc failure

Deallocation used the wrong function, which could have resulted in
race conditions because the function does not use locks internally.

Fixes: 1403f87d4fb8 ("malloc: enable memory hotplug support")

Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com>
5 years agomem: unmap unneeded space
Anatoly Burakov [Mon, 30 Apr 2018 11:21:43 +0000 (12:21 +0100)]
mem: unmap unneeded space

When we ask to reserve virtual areas, we usually include
alignment in the mapping size, and that memory ends up
being wasted. Wasting a gigabyte of VA space while trying to
reserve one gigabyte is pretty expensive on 32-bit, so after
we're done mapping, unmap unneeded space.

Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
Acked-by: Bruce Richardson <bruce.richardson@intel.com>
5 years agomem: check if allocation size is too big
Anatoly Burakov [Mon, 30 Apr 2018 11:21:42 +0000 (12:21 +0100)]
mem: check if allocation size is too big

Mapping size is a 64-bit integer, but mmap() will accept size_t for
size mappings. A user could request a mapping with an alignment, which
would have overflown size_t, so check if (size + alignment) will
overflow size_t.

Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
Acked-by: Bruce Richardson <bruce.richardson@intel.com>
5 years agomem: fix typo in local function name
Thomas Monjalon [Tue, 1 May 2018 19:38:52 +0000 (21:38 +0200)]
mem: fix typo in local function name

Fixes: 582bed1e1d1d ("mem: support mapping hugepages at runtime")

Signed-off-by: Thomas Monjalon <thomas@monjalon.net>
Acked-by: Anatoly Burakov <anatoly.burakov@intel.com>
5 years agoeal: fix typo in doc of pointer offset macro
Thomas Monjalon [Tue, 1 May 2018 19:35:50 +0000 (21:35 +0200)]
eal: fix typo in doc of pointer offset macro

Fixes: af75078fece3 ("first public release")
Cc: stable@dpdk.org
Signed-off-by: Thomas Monjalon <thomas@monjalon.net>
5 years agoeal: fix mempool ops name parsing
Ivan Malov [Fri, 4 May 2018 10:31:54 +0000 (11:31 +0100)]
eal: fix mempool ops name parsing

The code aimed to pick and remember the value of
mempool ops name from EAL command line arguments does not
copy the string and remembers the pointer provided
by getopt_long() directly. The latter could be clobbered
later and result in reading wrong mbuf pool ops name
by rte_mempool library.

Typically, this flaw could be avoided by using strdup()
to remember the string value of the option.

Fixes: a103a97e7191 ("eal: allow user to override default mempool driver")
Cc: stable@dpdk.org
Signed-off-by: Ivan Malov <ivan.malov@oktetlabs.ru>
Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>
Acked-by: Santosh Shukla <santosh.shukla@caviumnetworks.com>
5 years agorawdev: remove dead code
Shreyansh Jain [Wed, 9 May 2018 15:26:10 +0000 (20:56 +0530)]
rawdev: remove dead code

Coverity issue: 260406
Fixes: c88b3f2558ed ("rawdev: introduce raw device library")
Cc: stable@dpdk.org
Signed-off-by: Shreyansh Jain <shreyansh.jain@nxp.com>
5 years agoraw/skeleton: remove dead code
Shreyansh Jain [Wed, 9 May 2018 15:26:09 +0000 (20:56 +0530)]
raw/skeleton: remove dead code

Coverity Issue: 260411
Fixes: 55ca1b0f2151 ("raw/skeleton: add test cases")
Cc: stable@dpdk.org
Signed-off-by: Shreyansh Jain <shreyansh.jain@nxp.com>
5 years agoraw/skeleton: fix resource leak in test
Shreyansh Jain [Wed, 9 May 2018 15:26:08 +0000 (20:56 +0530)]
raw/skeleton: fix resource leak in test

Coverity issue: 260402
Fixes: 55ca1b0f2151 ("raw/skeleton: add test cases")
Cc: stable@dpdk.org
Signed-off-by: Shreyansh Jain <shreyansh.jain@nxp.com>
5 years agobus/fslmc: remove dead code
Shreyansh Jain [Wed, 9 May 2018 15:27:52 +0000 (20:57 +0530)]
bus/fslmc: remove dead code

Coverity issue: 268338
Fixes: 828d51d8fc3e ("bus/fslmc: refactor scan and probe functions")
Cc: stable@dpdk.org
Signed-off-by: Shreyansh Jain <shreyansh.jain@nxp.com>
Acked-by: Hemant Agrawal <hemant.agrawal@nxp.com>
5 years agobus/fslmc: fix memory leak and cleanup
Shreyansh Jain [Wed, 9 May 2018 15:27:51 +0000 (20:57 +0530)]
bus/fslmc: fix memory leak and cleanup

Coverity issue: 268327
Fixes: 828d51d8fc3e ("bus/fslmc: refactor scan and probe functions")
Cc: stable@dpdk.org
Signed-off-by: Shreyansh Jain <shreyansh.jain@nxp.com>
Acked-by: Hemant Agrawal <hemant.agrawal@nxp.com>
5 years agobus/dpaa: improve dynamic logging
Hemant Agrawal [Wed, 25 Apr 2018 12:57:02 +0000 (18:27 +0530)]
bus/dpaa: improve dynamic logging

Signed-off-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Acked-by: Shreyansh Jain <shreyansh.jain@nxp.com>
5 years agonet/dpaa: fix xstats implementation
Hemant Agrawal [Wed, 25 Apr 2018 12:57:01 +0000 (18:27 +0530)]
net/dpaa: fix xstats implementation

Some of the applications are expecting that if n is 0, it shall
return the supported number of stats. e.g. VPP

Fixes: b21ed3e2a16d ("net/dpaa: support extended statistics")
Cc: stable@dpdk.org
Signed-off-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Acked-by: Shreyansh Jain <shreyansh.jain@nxp.com>
5 years agobus/fslmc: improve debug logging
Hemant Agrawal [Wed, 25 Apr 2018 12:57:00 +0000 (18:27 +0530)]
bus/fslmc: improve debug logging

Signed-off-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Acked-by: Shreyansh Jain <shreyansh.jain@nxp.com>
5 years agobus/fslmc: support device blacklisting
Hemant Agrawal [Wed, 25 Apr 2018 12:56:59 +0000 (18:26 +0530)]
bus/fslmc: support device blacklisting

Signed-off-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Acked-by: Shreyansh Jain <shreyansh.jain@nxp.com>
5 years agobus/dpaa: support device blacklisting
Hemant Agrawal [Wed, 25 Apr 2018 12:56:58 +0000 (18:26 +0530)]
bus/dpaa: support device blacklisting

Signed-off-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Acked-by: Shreyansh Jain <shreyansh.jain@nxp.com>
5 years agokni: fix possible mbuf leaks and speed up port release
Yangchao Zhou [Thu, 19 Apr 2018 03:12:20 +0000 (11:12 +0800)]
kni: fix possible mbuf leaks and speed up port release

rx_q fifo can only be released by kernel thread. There may be
mbuf leaks in rx_q because kernel threads are randomly stopped.

When the kni is released and netdev is unregisterd, convert the
physical address mbufs in rx_q to the virtual address in free_q.
By the way, alloc_q can be processed together to speed up the
release rate in userspace.

In my test, it is improved from 300-500ms with a mempool that has
 131072 mbufs to 10ms(regardless of the specifications).

Suggested-by: Ferruh Yigit <ferruh.yigit@intel.com>
Signed-off-by: Yangchao Zhou <zhouyates@gmail.com>
Acked-by: Ferruh Yigit <ferruh.yigit@intel.com>
5 years agoexamples/vhost_scsi: replace strncpy with strlcpy
Reshma Pattan [Thu, 10 May 2018 12:05:44 +0000 (13:05 +0100)]
examples/vhost_scsi: replace strncpy with strlcpy

Use strlcpy instead of strncpy.

Fixes: db75c7af19 ("examples/vhost_scsi: introduce a new sample app")
Cc: stable@dpdk.org
Signed-off-by: Reshma Pattan <reshma.pattan@intel.com>
Acked-by: Bruce Richardson <bruce.richardson@intel.com>
5 years agoexamples/ipsec-secgw: replace strncpy with strlcpy
Reshma Pattan [Wed, 9 May 2018 16:11:06 +0000 (17:11 +0100)]
examples/ipsec-secgw: replace strncpy with strlcpy

Use strlcpy instead of strncpy.
Use strcpy where boundchecks on destination is not needed.

Fixes: 0d547ed037 ("examples/ipsec-secgw: support configuration file")
Fixes: 07b156199f ("examples/ipsec-secgw: fix configuration string termination")
Fixes: a1469c319f ("examples/ipsec-secgw: fix configuration parsing")
Cc: stable@dpdk.org
Signed-off-by: Reshma Pattan <reshma.pattan@intel.com>
Acked-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>
5 years agoexamples/quota_watermark: fix return type of threads
Reshma Pattan [Wed, 9 May 2018 11:35:28 +0000 (12:35 +0100)]
examples/quota_watermark: fix return type of threads

Gcc 8.0.1 reports incompatible cast between types i.e. from
`void (*)(void *)` to `(int (*)(void *)`.

Change the pipeline_stage prototype to retun int type
to fix the issue.

Fixes: a0ffcb257a ("examples/quota_watermark: correct code indentation")
Cc: stable@dpdk.org
Signed-off-by: Reshma Pattan <reshma.pattan@intel.com>
Acked-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>
Signed-off-by: Thomas Monjalon <thomas@monjalon.net>
5 years agotest/pipeline: fix return type of stub miss
David Hunt [Thu, 10 May 2018 07:38:15 +0000 (08:38 +0100)]
test/pipeline: fix return type of stub miss

New warning with gcc 8.1:
The table_action_stub_miss() function was incorrectly
declared as returning a typedef of
rte_pipeline_table_action_handler_miss, instead of what it
should have been, an int.
Changed this to declare the function as returning an int, and
now when the action_handler_miss is set to table_action_stub_miss
(cast as rte_pipeline_table_action_handler_miss), gcc 8.1 no
longer complains about the cast.

Fixes: 5205954791cb ("app/test: packet framework unit tests")
Cc: stable@dpdk.org
Signed-off-by: David Hunt <david.hunt@intel.com>
Reviewed-by: Reshma Pattan <reshma.pattan@intel.com>
Acked-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>
5 years agotest/distributor: fix return type of thread function
David Hunt [Thu, 10 May 2018 01:25:38 +0000 (02:25 +0100)]
test/distributor: fix return type of thread function

New warning with gcc 8.1:
When casting the function pointer passed to
rte_eal_remote_launch() as an lcore_function_t *,
it expects the cast function to return an int, but
instead it was returning void. Fixed to return
an int.

Fixes: c3eabff124e6 ("distributor: add unit tests")
Cc: stable@dpdk.org
Signed-off-by: David Hunt <david.hunt@intel.com>
Acked-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>
5 years agotest/reorder: fix freeing mbuf twice
Reshma Pattan [Fri, 4 May 2018 10:47:06 +0000 (11:47 +0100)]
test/reorder: fix freeing mbuf twice

mbufs are being freed twice in error, once in rte_mempool_put_bulk()
and then in rte_reorder_free(). Refactor the code so that we use
rte_reorder_free() to free mbufs in the reorder buffer, and use
rte_pktmbuf_free() to free any unused or drained mbufs.

Fixes: d0c9b58d71 ("app/test: new reorder unit test")
Cc: stable@dpdk.org
Signed-off-by: Reshma Pattan <reshma.pattan@intel.com>
Reviewed-by: David Hunt <david.hunt@intel.com>
5 years agohash: explicit casts for truncation in CRC32c
Andy Green [Sat, 12 May 2018 02:00:37 +0000 (10:00 +0800)]
hash: explicit casts for truncation in CRC32c

In function 'crc32c_sse42_u64_mimic':
rte_hash_crc.h:402:40:
warning: conversion from 'uint64_t' {aka 'long unsigned int'}
to 'uint32_t' {aka 'unsigned int'} may change value [-Wconversion]
  init_val = crc32c_sse42_u32(d.u32[0], init_val);

Fixes: 00bf774bab0b ("hash: add assembly implementation of CRC32 intrinsics")
Cc: stable@dpdk.org
Signed-off-by: Andy Green <andy@warmcat.com>
Reviewed-by: Stephen Hemminger <stephen@networkplumber.org>
5 years agohash: move stack declaration at top of CRC32c function
Andy Green [Sat, 12 May 2018 02:00:32 +0000 (10:00 +0800)]
hash: move stack declaration at top of CRC32c function

In function 'crc32c_2words':
rte_hash_crc.h:347:2:
warning: ISO C90 forbids mixed declarations and code
[-Wdeclaration-after-statement]
  uint32_t crc, term1, term2;

Fixes: d983cf41698f ("hash: add software CRC32 implementation")
Cc: stable@dpdk.org
Signed-off-by: Andy Green <andy@warmcat.com>
Reviewed-by: Stephen Hemminger <stephen@networkplumber.org>
5 years agoethdev: explicit cast of buffered Tx number
Andy Green [Sat, 12 May 2018 02:00:27 +0000 (10:00 +0800)]
ethdev: explicit cast of buffered Tx number

In function 'rte_eth_tx_buffer_flush':
rte_ethdev.h:4248:55:
warning: conversion from 'int' to 'uint16_t'
{aka 'short unsigned int'} may change value [-Wconversion]
   buffer->error_callback(&buffer->pkts[sent], to_send - sent,

Fixes: d6c99e62c852 ("ethdev: add buffered Tx")
Cc: stable@dpdk.org
Signed-off-by: Andy Green <andy@warmcat.com>
Reviewed-by: Stephen Hemminger <stephen@networkplumber.org>
5 years agonet: move stack variable at top of VLAN strip function
Andy Green [Sat, 12 May 2018 02:00:17 +0000 (10:00 +0800)]
net: move stack variable at top of VLAN strip function

In function 'rte_vlan_strip':
rte_ether.h:357:2:
warning: ISO C90 forbids mixed declarations and code
[-Wdeclaration-after-statement]
  struct vlan_hdr *vh = (struct vlan_hdr *)(eh + 1);

Fixes: c974021a5949 ("ether: add soft vlan encap/decap")
Cc: stable@dpdk.org
Signed-off-by: Andy Green <andy@warmcat.com>
Reviewed-by: Stephen Hemminger <stephen@networkplumber.org>
5 years agospinlock/x86: move stack declaration before code
Andy Green [Sat, 12 May 2018 01:59:17 +0000 (09:59 +0800)]
spinlock/x86: move stack declaration before code

In function 'rte_try_tm':
rte_spinlock.h:82:2:
warning: ISO C90 forbids mixed declarations and code
[-Wdeclaration-after-statement]
  int retries = RTE_RTM_MAX_RETRIES;

Fixes: ba7468997ea6 ("spinlock: add HTM lock elision for x86")
Cc: stable@dpdk.org
Signed-off-by: Andy Green <andy@warmcat.com>
Reviewed-by: Stephen Hemminger <stephen@networkplumber.org>
5 years agoeal: declare trace buffer at top of own block
Andy Green [Sat, 12 May 2018 01:59:32 +0000 (09:59 +0800)]
eal: declare trace buffer at top of own block

rte_dev.h:54:2:
warning: ISO C90 forbids mixed declarations and
code [-Wdeclaration-after-statement]
  char buffer[vsnprintf(NULL, 0, fmt, ap) + 1];

Fixes: b974e4a40cb5 ("ethdev: make error checking macros public")
Cc: stable@dpdk.org
Signed-off-by: Andy Green <andy@warmcat.com>
Reviewed-by: Stephen Hemminger <stephen@networkplumber.org>
5 years agoeal: explicit cast of core id when getting index
Andy Green [Sat, 12 May 2018 01:59:07 +0000 (09:59 +0800)]
eal: explicit cast of core id when getting index

rte_lcore.h: In function 'rte_lcore_index':
rte_lcore.h:122:14:
warning: conversion to 'int' from 'unsigned int' may change
the sign of the result [-Wsign-conversion]
   lcore_id = rte_lcore_id();

Fixes: 5583037a7950 ("eal: get relative core index")
Cc: stable@dpdk.org
Signed-off-by: Andy Green <andy@warmcat.com>
Reviewed-by: Stephen Hemminger <stephen@networkplumber.org>
5 years agoeal: explicit cast of builtin for bsf32
Andy Green [Sat, 12 May 2018 01:58:57 +0000 (09:58 +0800)]
eal: explicit cast of builtin for bsf32

rte_common.h:416:9:
warning: conversion to 'uint32_t' {aka 'unsigned int'} from
'int' may change the sign of the result [-Wsign-conversion]
  return __builtin_ctz(v);
         ^~~~~~~~~~~~~~~~

The builtin is defined to return int, but we want to
return it as uint32_t.  Its only defined valid return
values are positive integers or zero, which is OK for
uint32_t.  So just add an explicit cast.

Fixes: 03f6bced5bba ("eal: use intrinsic function")
Cc: stable@dpdk.org
Signed-off-by: Andy Green <andy@warmcat.com>
Reviewed-by: Stephen Hemminger <stephen@networkplumber.org>
5 years agobus/fpga: use strlcpy instead of strncpy
Pablo de Lara [Sat, 12 May 2018 07:02:17 +0000 (08:02 +0100)]
bus/fpga: use strlcpy instead of strncpy

Gcc 8 reports a build error:

drivers/bus/ifpga/ifpga_bus.c:161:2: error: ‘strncpy’ specified
bound 256 equals destination size [-Werror=stringop-truncation]
  strncpy(afu_pr_conf.bs_path, path, sizeof(afu_pr_conf.bs_path));
  ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Fixes: 05fa3d4a6539 ("bus/ifpga: add Intel FPGA bus library")

Signed-off-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>