Beilei Xing [Tue, 28 Mar 2017 09:28:43 +0000 (17:28 +0800)]
net/i40e: support tunnel filter to VF
Previously, only tunnel filter to PF is supported.
This patch adds i40e_dev_consistent_tunnel_filter_set
function for consistent filter API to support tunnel
filter to VF.
Signed-off-by: Yong Liu <yong.liu@intel.com> Signed-off-by: Bernard Iremonger <bernard.iremonger@intel.com> Signed-off-by: Beilei Xing <beilei.xing@intel.com> Acked-by: Jingjing Wu <jingjing.wu@intel.com>
Zhiyong Yang [Wed, 29 Mar 2017 07:16:14 +0000 (15:16 +0800)]
net/ixgbe: remove limit of Tx burst size
To add a wrapper function to remove the limit of tx burst size and
implement the "make an best effort to transmit the pkts" policy.
The patch makes ixgbe vec function work in a consistent behavior
like ixgbe_xmit_pkts_simple and ixgbe_xmit_pkts do that.
Cc: Helin Zhang <helin.zhang@intel.com> Cc: Konstantin Ananyev <konstantin.ananyev@intel.com> Signed-off-by: Zhiyong Yang <zhiyong.yang@intel.com> Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com>
Zhiyong Yang [Wed, 29 Mar 2017 07:16:13 +0000 (15:16 +0800)]
net/i40e: remove limit of Tx burst size
To add a wrapper function to remove the limit of tx burst size. The patch
makes i40e vec function an best effort to transmit the pkts in the
consistent behavior like i40e_xmit_pkts_simple and i40e_xmit_pkts do that.
Cc: Helin Zhang <helin.zhang@intel.com> Cc: Jingjing Wu <jingjing.wu@intel.com> Signed-off-by: Zhiyong Yang <zhiyong.yang@intel.com> Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com>
Zhiyong Yang [Wed, 29 Mar 2017 07:16:12 +0000 (15:16 +0800)]
net/fm10k: remove limit of Tx burst size
To add a wrapper function to remove the limit of tx burst size.
The patch makes fm10k vec function an best effort to transmit
pkts in the consistent behavior like fm10k_xmit_pkts does that.
Cc: Jing Chen <jing.d.chen@intel.com> Signed-off-by: Zhiyong Yang <zhiyong.yang@intel.com> Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com>
Harish Patil [Wed, 29 Mar 2017 20:37:01 +0000 (13:37 -0700)]
net/qede: support LRO/TSO offloads
This patch includes slowpath configuration and fastpath changes
to support LRO and TSO. A bit of revamping is needed in order
to make use of existing packet classification schemes in Rx fastpath
and for SG element processing in Tx.
Harish Patil [Wed, 29 Mar 2017 20:37:00 +0000 (13:37 -0700)]
net/qede: support ntuple and flow director filter
Add limited support for ntuple filter and flow director configuration.
The filtering is based on 4-tuples viz src-ip, dst-ip, src-port,
dst-port. The mask fields, tcp_flags, flex masks, priority fields,
Rx queue drop etc are not supported.
Rasesh Mody [Wed, 29 Mar 2017 20:36:57 +0000 (13:36 -0700)]
net/qede/base: prevent race condition during unload
Merge hw_stop and hw_reset into one function.
Prevent race condition between MFW attentions and pf stop command during
unload flow that causes an ASSERT.
Rasesh Mody [Wed, 29 Mar 2017 20:36:56 +0000 (13:36 -0700)]
net/qede/base: multi-Txq support on same queue-zone for VFs
A step toward having multi-Txq support on same queue-zone for VFs.
This change takes care of:
- VFs assume a single CID per-queue, where queue X receives CID X.
Switch to a model similar to that of PF - I.e., Use different CIDs
for Rx/Tx, and use mapping to acquire/release those. Each VF
currently will have 32 CIDs available for it [for its possible 16
Rx & 16 Tx queues].
- To retain the same interface for PFs/VFs when initializing queues,
the base driver would have to retain a unique number per-each queue
that would be communicated in some extended TLV [current TLV
interface allows the PF to send only the queue-id]. The new TLV isn't
part of the current change but base driver would now start adding
such unique keys internally to queue_cids. This would also force
us to start having alloc/setup/free for L2 [we've refrained from
doing so until now]
The limit would be no-more than 64 queues per qzone [This could be
changed if needed, but hopefully no one needs so many queues]
- In IOV, Add infrastructure for up to 64 qids per-qzone, although
at the moment hard-code '0' for Rx and '1' for Tx [Since VF still
isn't communicating via new TLV which index to associate with a
given queue in its queue-zone].
Rasesh Mody [Wed, 29 Mar 2017 20:36:31 +0000 (13:36 -0700)]
net/qede/base: revise tunnel APIs/structs
Revise tunnel APIs/structs.
- Unite tunnel start and update params in single struct
"ecore_tunnel_info"
- Remove A0 chip tunnelling support.
- Added per tunnel info - removed bitmasks.
Rasesh Mody [Wed, 29 Mar 2017 20:36:29 +0000 (13:36 -0700)]
net/qede/base: optimize cache-line access
Optimize cache-line access in ecore_chain -
re-arrange fields so that fields that are needed for fastpath
[mostly produce/consume and their derivatives] are in the first cache
line, and the rest are in the second.
This is true for both PBL and NEXT_PTR kind of chains.
Advancing a page in a SINGLE_PAGE chain would still require the 2nd
cacheline as well, but afaik only SPQ uses it and so it isn't
considered as 'fastpath'.
Rasesh Mody [Wed, 29 Mar 2017 20:36:27 +0000 (13:36 -0700)]
net/qede/base: make L2 queues handle based
L2 handler changes:
This is change to remove the queue-id/qzone difference for Tx queues.
It does that by mainly doing:
a. VFs queues are no longer determined by the SBs they're using.
Instead, the ecore-client needs to maintain those and choose the values
to be used by VF when initializing it.
b. Eliminate the HW-cid array in the hw-function.
To do that, have all the rx/tx functionality turn into 'handle' base -
when queue would be started the caller would get a (void*) handle,
which it would later use with ecore for configuring various
queue-related stop [update, close].
Rasesh Mody [Wed, 29 Mar 2017 20:36:26 +0000 (13:36 -0700)]
net/qede/base: refactor code of SP queues
Maintain slowpath event queue and consumer queue within HW function
structure, update corresponding alloc and free APIs accordingly.
Cleanup unused code under CONFIG_ECORE_LL2 ifdef.
Rasesh Mody [Wed, 29 Mar 2017 20:36:19 +0000 (13:36 -0700)]
net/qede/base: allow only trusted VFs to be promisc
Allow only trusted VFs to be promisc/multi-promisc. The reasonable
thing is to use the 'trusted' node instead of simply allowing VFs to
become promiscuous.
Rasesh Mody [Wed, 29 Mar 2017 20:36:05 +0000 (13:36 -0700)]
net/qede/base: utilize FW 8.18.9.0
This change is in preparation to work with new FW 8.18.9.0.
Rename the defines to use E4_ and structs to use e4_. This renaming
is to add support for future chipsets.
Allain Legacy [Tue, 28 Mar 2017 11:54:08 +0000 (07:54 -0400)]
net/avp: handle interrupt migration
This commit introduces changes required to support VM live-migration. This
is done by registering and responding to interrupts coming from the host to
signal that the memory is about to be made invalid and replaced with a new
memory zone on the destination compute node.
Enabling and disabling of the interrupts are maintained outside of the
start/stop functions because they must be enabled for the lifetime of the
device. This is so that host interrupts are serviced and acked even in
cases where the app may have stopped the device.
Signed-off-by: Allain Legacy <allain.legacy@windriver.com> Signed-off-by: Matt Peters <matt.peters@windriver.com> Acked-by: Vincent Jardin <vincent.jardin@6wind.com>
Allain Legacy [Tue, 28 Mar 2017 11:54:07 +0000 (07:54 -0400)]
net/avp: add device start and stop operations
Adds support for device start and stop functions. This allows an
application to control the administrative state of an AVP device. Stopping
the device will notify the host application to stop sending packets on that
device's receive queues.
Signed-off-by: Allain Legacy <allain.legacy@windriver.com> Signed-off-by: Matt Peters <matt.peters@windriver.com> Acked-by: Vincent Jardin <vincent.jardin@6wind.com>
Allain Legacy [Tue, 28 Mar 2017 11:54:06 +0000 (07:54 -0400)]
net/avp: add device promiscuous functions
Adds support for setting and clearing promiscuous mode on an AVP device.
When enabled the _mac_filter function will allow packets destined to any
MAC address to be processed by the receive functions.
Signed-off-by: Allain Legacy <allain.legacy@windriver.com> Signed-off-by: Matt Peters <matt.peters@windriver.com> Acked-by: Vincent Jardin <vincent.jardin@6wind.com>
Allain Legacy [Tue, 28 Mar 2017 11:54:04 +0000 (07:54 -0400)]
net/avp: add packet transmit functions
Adds support for packet transmit functions so that an application can send
packets to the host application via an AVP device queue. Both the simple
and scattered functions are supported.
Signed-off-by: Allain Legacy <allain.legacy@windriver.com> Signed-off-by: Matt Peters <matt.peters@windriver.com> Acked-by: Vincent Jardin <vincent.jardin@6wind.com>
Allain Legacy [Tue, 28 Mar 2017 11:53:57 +0000 (07:53 -0400)]
net/avp: add public header files
Adds public/exported header files for the AVP PMD. The AVP device is a
shared memory based device. The structures and constants that define the
method of operation of the device must be visible by both the PMD and the
host DPDK application. They must not change without proper version
controls and updates to both the hypervisor DPDK application and the PMD.
The hypervisor DPDK application is a Wind River Systems proprietary
virtual switch.
Signed-off-by: Allain Legacy <allain.legacy@windriver.com> Signed-off-by: Matt Peters <matt.peters@windriver.com> Acked-by: Vincent Jardin <vincent.jardin@6wind.com>
Vasily Philipov [Tue, 28 Mar 2017 10:18:49 +0000 (13:18 +0300)]
net/mlx: enhance Rx scatter mode detection
Toggle Rx scatter mode based on the scatter_enable flag and the maximum
packet size only instead of deriving this information from the jumbo_frame
setting and the MTU configuration.
Wenzhuo Lu [Fri, 24 Feb 2017 03:24:29 +0000 (11:24 +0800)]
net/i40e: allocate VF TC bandwidth from PF
Allocate all TCs' relative bandwidth (percentage) on
a specific VF.
It can be taken as relative min bandwidth setting.
User can call the API to set VF TC's min bandwidth
from PF.
Signed-off-by: Wenzhuo Lu <wenzhuo.lu@intel.com> Acked-by: Jingjing Wu <jingjing.wu@intel.com>
git.droids-corp.org / dpdk.git / log