The ks8851_rdreg32() is used only in one place, to read two registers
using a single read. To make it easier to support 16-bit accesses via
parallel bus later on, replace this single read with two 16-bit reads
from each of the registers and drop the ks8851_rdreg32() altogether.
If this has noticeable performance impact on the SPI variant of KS8851,
then we should consider using regmap to abstract the SPI and parallel
bus options and in case of SPI, permit regmap to merge register reads
of neighboring registers into single, longer, read.
Signed-off-by: Marek Vasut <marex@denx.de>
Cc: David S. Miller <davem@davemloft.net>
Cc: Lukas Wunner <lukas@wunner.de>
Cc: Petr Stetiar <ynezz@true.cz>
Cc: YueHaibing <yuehaibing@huawei.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
Replace spi_{get,set}_drvdata() with dev_{get,set}_drvdata(), which
works for both SPI and platform drivers. This is done in preparation
for unifying the KS8851 SPI and parallel bus drivers.
There should be no functional change.
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: Marek Vasut <marex@denx.de>
Cc: David S. Miller <davem@davemloft.net>
Cc: Lukas Wunner <lukas@wunner.de>
Cc: Petr Stetiar <ynezz@true.cz>
Cc: YueHaibing <yuehaibing@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Use device managed version of alloc_etherdev() to simplify the code.
No functional change intended.
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: Marek Vasut <marex@denx.de>
Cc: David S. Miller <davem@davemloft.net>
Cc: Lukas Wunner <lukas@wunner.de>
Cc: Petr Stetiar <ynezz@true.cz>
Cc: YueHaibing <yuehaibing@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Since the driver probe function already has a struct device *dev pointer
and can easily derive of_node pointer from it, pass the of_node pointer as
a parameter to ks8851_init_mac() to avoid fishing it out from ks->spidev.
This is the only reference to spidev in the function, so get rid of it.
This is done in preparation for unifying the KS8851 SPI and parallel bus
drivers.
No functional change.
Signed-off-by: Marek Vasut <marex@denx.de>
Cc: David S. Miller <davem@davemloft.net>
Cc: Lukas Wunner <lukas@wunner.de>
Cc: Petr Stetiar <ynezz@true.cz>
Cc: YueHaibing <yuehaibing@huawei.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
Use netdev_err() instead of dev_err() to avoid accessing the spidev->dev
in the interrupt handler. This is the only place which uses the spidev
in this function, so replace it with netdev_err() to get rid of it. This
is done in preparation for unifying the KS8851 SPI and parallel drivers.
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: Marek Vasut <marex@denx.de>
Cc: David S. Miller <davem@davemloft.net>
Cc: Lukas Wunner <lukas@wunner.de>
Cc: Petr Stetiar <ynezz@true.cz>
Cc: YueHaibing <yuehaibing@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Rename ndev variable to netdev for the sake of consistency.
No functional change.
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: Marek Vasut <marex@denx.de>
Cc: David S. Miller <davem@davemloft.net>
Cc: Lukas Wunner <lukas@wunner.de>
Cc: Petr Stetiar <ynezz@true.cz>
Cc: YueHaibing <yuehaibing@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Pull out the spi->dev into one common place in the function instead of
having it repeated over and over again. This is done in preparation for
unifying ks8851 and ks8851-mll drivers. No functional change.
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: Marek Vasut <marex@denx.de>
Cc: David S. Miller <davem@davemloft.net>
Cc: Lukas Wunner <lukas@wunner.de>
Cc: Petr Stetiar <ynezz@true.cz>
Cc: YueHaibing <yuehaibing@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Ronak Doshi says:
====================
vmxnet3: upgrade to version 4
vmxnet3 emulation has recently added several new features which includes
offload support for tunnel packets, support for new commands the driver
can issue to emulation, change in descriptor fields, etc. This patch
series extends the vmxnet3 driver to leverage these new features.
Compatibility is maintained using existing vmxnet3 versioning mechanism as
follows:
- new features added to vmxnet3 emulation are associated with new vmxnet3
version viz. vmxnet3 version 4.
- emulation advertises all the versions it supports to the driver.
- during initialization, vmxnet3 driver picks the highest version number
supported by both the emulation and the driver and configures emulation
to run at that version.
In particular, following changes are introduced:
Patch 1:
This patch introduces utility macros for vmxnet3 version 4 comparison
and updates Copyright information.
Patch 2:
This patch implements get_rss_hash_opts and set_rss_hash_opts methods
to allow querying and configuring different Rx flow hash configurations
which can be used to support UDP/ESP RSS.
Patch 3:
This patch introduces segmentation and checksum offload support for
encapsulated packets. This avoids segmenting and calculating checksum
for each segment and hence gives performance boost.
Patch 4:
With all vmxnet3 version 4 changes incorporated in the vmxnet3 driver,
with this patch, the driver can configure emulation to run at vmxnet3
version 4.
Changes in v3 -> v4:
- Replaced BUG_ON() with WARN_ON_ONCE()
Changes in v2 -> v3:
- fixed get_rss_hash_opts to return correct values for udp rss
Changes in v2:
- Fixed compilation issue due to missing closed brace
- added fallthrough comment
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
With all vmxnet3 version 4 changes incorporated in the vmxnet3 driver,
the driver can configure emulation to run at vmxnet3 version 4, provided
the emulation advertises support for version 4.
Signed-off-by: Ronak Doshi <doshir@vmware.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Vmxnet3 version 3 device supports checksum/TSO offload. Thus, vNIC to
pNIC traffic can leverage hardware checksum/TSO offloads. However,
vmxnet3 does not support checksum/TSO offload for Geneve/VXLAN
encapsulated packets. Thus, for a vNIC configured with an overlay, the
guest stack must first segment the inner packet, compute the inner
checksum for each segment and encapsulate each segment before
transmitting the packet via the vNIC. This results in significant
performance penalty.
This patch will enhance vmxnet3 to support Geneve/VXLAN TSO as well as
checksum offload.
Signed-off-by: Ronak Doshi <doshir@vmware.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
With vmxnet3 version 4, the emulation supports multiqueue(RSS) for
UDP and ESP traffic. A guest can enable/disable RSS for UDP/ESP over
IPv4/IPv6 by issuing commands introduced in this patch. ESP ipv6 is
not yet supported in this patch.
This patch implements get_rss_hash_opts and set_rss_hash_opts
methods to allow querying and configuring different Rx flow hash
configurations.
Signed-off-by: Ronak Doshi <doshir@vmware.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
vmxnet3 is currently at version 3 and this patch initiates the
preparation to accommodate changes for version 4. Introduced utility
macros for vmxnet3 version 4 comparison and update Copyright
information.
Signed-off-by: Ronak Doshi <doshir@vmware.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
'nic_data' is no longer used outside of the #ifdef block
in efx_ef10_set_mac_address:
drivers/net/ethernet/sfc/ef10.c:3231:28: error: unused variable 'nic_data' [-Werror,-Wunused-variable]
struct efx_ef10_nic_data *nic_data = efx->nic_data;
Move the variable into a local scope.
Fixes: dfcabb0788 ("sfc: move vport_id to struct efx_nic")
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
Jeff Kirsher says:
====================
100GbE Intel Wired LAN Driver Updates 2020-05-27
This series contains updates to the ice driver only.
Jesse fixes a number of issues, starting with fixing the remaining
signed versus unsigned comparison issues. Cleaned up an unused code
define. Fixed the implementation of the manage MAC write command, to
simplify it by using a simple array to represent the MAC address when
writing it.
Paul fixes the setting of the VF default LAN address, by removing a
check that assumed that the address had been deleted and zeroed.
Surabhi prevents a memory leak on filter management initialization
failures and during queue initialization and buffer allocation failures.
Brett adds additional receive error counters that are reported by
ethtool. Fixed the enabling and disabling of VLAN stripping when the
PVID has been set.
Evan fixes a race condition between the firmware and software, which can
occur between the admin queue setup and the first command sent.
Marta fixes the driver when XDP transmit rings are destroyed, also make
sure the XDP transmit queues are also destroyed. Update the statistics
when XDP transmit programs are loaded and packets are sent. Changed the
number of XDP transmit queues to match the number of receive queues,
instead of matching the number of transmit queues.
Bruce avoids undefined behavior by not writing the 8-bit element
init_q_state with the associated internal-to-hardware field which is
122-bits.
Anirudh (Ani) refactors the receive checksum checks.
Krzysztof notifies the user if the fill queue is not long enough to
prepare all buffers before packet processing starts and allocates the
buffers during the NAPI poll.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Christoph Hellwig says:
====================
remove most callers of kernel_setsockopt v3
this series removes most callers of the kernel_setsockopt functions, and
instead switches their users to small functions that implement setting a
sockopt directly using a normal kernel function call with type safety and
all the other benefits of not having a function call.
In some cases these functions seem pretty heavy handed as they do
a lock_sock even for just setting a single variable, but this mirrors
the real setsockopt implementation unlike a few drivers that just set
set the fields directly.
Changes since v2:
- drop the separately merged kernel_getopt_removal
- drop the sctp patches, as there is conflicting cleanup going on
- add an additional ACK for the rxrpc changes
Changes since v1:
- use ->getname for sctp sockets in dlm
- add a new ->bind_add struct proto method for dlm/sctp
- switch the ipv6 and remaining sctp helpers to inline function so that
the ipv6 and sctp modules are not pulled in by any module that could
potentially use ipv6 or sctp connections
- remove arguments to various sock_* helpers that are always used with
the same constant arguments
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Avoid using kernel_setsockopt for the TIPC_IMPORTANCE option when we can
just use the internal helper. The only change needed is to pass a struct
sock instead of tipc_sock, which is private to socket.c
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add a helper to directly set the RXRPC_MIN_SECURITY_LEVEL sockopt from
kernel space without going through a fake uaccess.
Thanks to David Howells for the documentation updates.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Acked-by: David Howells <dhowells@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add a helper to directly set the IPV6_RECVPKTINFO sockopt from kernel
space without going through a fake uaccess.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add a helper to directly set the IPV6_ADD_PREFERENCES sockopt from kernel
space without going through a fake uaccess.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add a helper to directly set the IPV6_RECVERR sockopt from kernel space
without going through a fake uaccess.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: David Howells <dhowells@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add a helper to directly set the IPV6_V6ONLY sockopt from kernel space
without going through a fake uaccess.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add a helper to directly set the IP_PKTINFO sockopt from kernel
space without going through a fake uaccess.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add a helper to directly set the IP_MTU_DISCOVER sockopt from kernel
space without going through a fake uaccess.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: David Howells <dhowells@redhat.com> [rxrpc bits]
Signed-off-by: David S. Miller <davem@davemloft.net>
Add a helper to directly set the IP_RECVERR sockopt from kernel space
without going through a fake uaccess.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: David Howells <dhowells@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add a helper to directly set the IP_FREEBIND sockopt from kernel space
without going through a fake uaccess.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add a helper to directly set the IP_TOS sockopt from kernel space without
going through a fake uaccess.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Acked-by: Sagi Grimberg <sagi@grimberg.me>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add a helper to directly set the TCP_KEEPCNT sockopt from kernel space
without going through a fake uaccess.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add a helper to directly set the TCP_KEEPINTVL sockopt from kernel space
without going through a fake uaccess.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add a helper to directly set the TCP_KEEP_IDLE sockopt from kernel
space without going through a fake uaccess.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add a helper to directly set the TCP_USER_TIMEOUT sockopt from kernel
space without going through a fake uaccess.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add a helper to directly set the TCP_SYNCNT sockopt from kernel space
without going through a fake uaccess.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Acked-by: Sagi Grimberg <sagi@grimberg.me>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add a helper to directly set the TCP_QUICKACK sockopt from kernel space
without going through a fake uaccess. Cleanup the callers to avoid
pointless wrappers now that this is a simple function call.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add a helper to directly set the TCP_NODELAY sockopt from kernel space
without going through a fake uaccess. Cleanup the callers to avoid
pointless wrappers now that this is a simple function call.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Acked-by: Sagi Grimberg <sagi@grimberg.me>
Acked-by: Jason Gunthorpe <jgg@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add a helper to directly set the TCP_CORK sockopt from kernel space
without going through a fake uaccess. Cleanup the callers to avoid
pointless wrappers now that this is a simple function call.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add a helper to directly set the SO_REUSEPORT sockopt from kernel space
without going through a fake uaccess.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add a helper to directly set the SO_RCVBUFFORCE sockopt from kernel space
without going through a fake uaccess.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add a helper to directly set the SO_KEEPALIVE sockopt from kernel space
without going through a fake uaccess.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add a helper to directly enable timestamps instead of setting the
SO_TIMESTAMP* sockopts from kernel space and going through a fake
uaccess.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add a helper to directly set the SO_BINDTOIFINDEX sockopt from kernel
space without going through a fake uaccess.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add a helper to directly set the SO_SNDTIMEO_NEW sockopt from kernel
space without going through a fake uaccess. The interface is
simplified to only pass the seconds value, as that is the only
thing needed at the moment.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add a helper to directly set the SO_PRIORITY sockopt from kernel space
without going through a fake uaccess.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Acked-by: Sagi Grimberg <sagi@grimberg.me>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add a helper to directly set the SO_LINGER sockopt from kernel space
with onoff set to true and a linger time of 0 without going through a
fake uaccess.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Acked-by: Sagi Grimberg <sagi@grimberg.me>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add a helper to directly set the SO_REUSEADDR sockopt from kernel space
without going through a fake uaccess.
For this the iscsi target now has to formally depend on inet to avoid
a mostly theoretical compile failure. For actual operation it already
did depend on having ipv4 or ipv6 support.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Acked-by: Sagi Grimberg <sagi@grimberg.me>
Signed-off-by: David S. Miller <davem@davemloft.net>
Updates highlights:
1) From Vu Pham (8): Support VM traffics failover with bonded VF
representors and e-switch egress/ingress ACLs
This series introduce the support for Virtual Machine running I/O
traffic over direct/fast VF path and failing over to slower
paravirtualized path using the following features:
__________________________________
| VM _________________ |
| |FAILOVER device | |
| |________________| |
| | |
| ____|_____ |
| | | |
| ______ |___ ____|_______ |
| | VF PT | |VIRTIO-NET | |
| | device | | device | |
| |_________| |___________| |
|___________|______________|________|
| |
| HYPERVISOR |
| ____|______
| | macvtap |
| |virtio BE |
| |___________|
| |
| ____|_____
| |host VF |
| |_________|
| |
_____|______ _____|_____
| PT VF | | host VF |
|representor| |representor|
|___________| |___________|
\ /
\ /
\ /
\ / _________________
\_______/ | |
_______|________ | V-SWITCH |
|VF representors |________________| (OVS) |
| bond | |________________|
|________________| |
________|________
| Uplink |
| representor |
|_________________|
Summary:
--------
Problem statement:
------------------
Currently in above topology, when netfailover device is configured using
VFs and eswitch VF representors, and when traffic fails over to stand-by
VF which is exposed using macvtap device to guest VM, eswitch fails to
switch the traffic to the stand-by VF representor. This occurs because
there is no knowledge at eswitch level of the stand-by representor
device.
Solution:
---------
Using standard bonding driver, a bond netdevice is created over VF
representor device which is used for offloading tc rules.
Two VF representors are bonded together, one for the passthrough VF
device and another one for the stand-by VF device.
With this solution, mlx5 driver listens to the failover events
occuring at the bond device level to failover traffic to either of
the active VF representor of the bond.
a. VM with netfailover device of VF pass-thru (PT) device and virtio-net
paravirtualized device with same MAC-address to handle failover
traffics at VM level.
b. Host bond is active-standby mode, with the lower devices being the VM
VF PT representor, and the representor of the 2nd VF to handle
failover traffics at Hypervisor/V-Switch OVS level.
- During the steady state (fast datapath): set the bond active
device to be the VM PT VF representor.
- During failover: apply bond failover to the second VF representor
device which connects to the VM non-accelerated path.
c. E-Switch ingress/egress ACL tables to support failover traffics at
E-Switch level
I. E-Switch egress ACL with forward-to-vport rule:
- By default, eswitch vport egress acl forward packets to its
counterpart NIC vport.
- During port failover, the egress acl forward-to-vport rule will
be added to e-switch vport of passive/in-active slave VF
representor
to forward packets to other e-switch vport ie. the active slave
representor's e-switch vport to handle egress "failover"
traffics.
- Using lower change netdev event to detect a representor is a
lower
dev (slave) of bond and becomes active, adding egress acl
forward-to-vport rule of all other slave netdevs to forward to
this
representor's vport.
- Using upper change netdev event to detect a representor unslaving
from bond device to delete its vport's egress acl forward-to-vport
rule.
II. E-Switch ingress ACL metadata reg_c for match
- Bonded representors' vorts sharing tc block have the same
root ingress acl table and a unique metadata for match.
- Traffics from both representors's vports will be tagged with same
unique metadata reg_c.
- Using upper change netdev event to detect a representor
enslaving/unslaving from bond device to setup shared root ingress
acl and unique metadata.
2) From Alex Vesker (2): Slpit RX and TX lock for parallel rule insertion in
software steering
3) Eli Britstein (2): Optimize performance for IPv4/IPv6 ethertype use the HW
ip_version register rather than parsing eth frames for ethertype.
-----BEGIN PGP SIGNATURE-----
iQEzBAABCAAdFiEEGhZs6bAKwk/OTgTpSD+KveBX+j4FAl7PEFAACgkQSD+KveBX
+j4Z5Af+NYwihYZpQYBBN00K7Wu10XZ65u5MbGSDmzpdN62w0kKfjsJ70bb9aiws
h8LC7lspdMLRMMn9pWwFKshyF6RoSD9Ku3ZYhUbtj+hJLElAd9IwGt6pPKr8hPDd
9h+ZcBkacdhNwWKf7CKThic0c/0PLdVyzRysHxcQWKSMPCTdgiL5Z3PQHA0TM6J3
6Excs2z7kSuuyyxQ1cyWCaqSz4rqCrYyd8Ws4HOPhXgSbX14Q3mtMsBDayx2gHNW
rdVbaNN6s2o0TxbrCwd0AaNP3UWcnjNqu1ohxgJiSe8y+MHMoB0OMoO+6vQJnwNI
bzpZEioswV1zdgK3qNmXqbHOiHRSVQ==
=xM1D
-----END PGP SIGNATURE-----
Merge tag 'mlx5-updates-2020-05-26' of git://git.kernel.org/pub/scm/linux/kernel/git/saeed/linux
Saeed Mahameed says:
====================
mlx5-updates-2020-05-26
Updates highlights:
1) From Vu Pham (8): Support VM traffics failover with bonded VF
representors and e-switch egress/ingress ACLs
This series introduce the support for Virtual Machine running I/O
traffic over direct/fast VF path and failing over to slower
paravirtualized path using the following features:
__________________________________
| VM _________________ |
| |FAILOVER device | |
| |________________| |
| | |
| ____|_____ |
| | | |
| ______ |___ ____|_______ |
| | VF PT | |VIRTIO-NET | |
| | device | | device | |
| |_________| |___________| |
|___________|______________|________|
| |
| HYPERVISOR |
| ____|______
| | macvtap |
| |virtio BE |
| |___________|
| |
| ____|_____
| |host VF |
| |_________|
| |
_____|______ _____|_____
| PT VF | | host VF |
|representor| |representor|
|___________| |___________|
\ /
\ /
\ /
\ / _________________
\_______/ | |
_______|________ | V-SWITCH |
|VF representors |________________| (OVS) |
| bond | |________________|
|________________| |
________|________
| Uplink |
| representor |
|_________________|
Summary:
--------
Problem statement:
------------------
Currently in above topology, when netfailover device is configured using
VFs and eswitch VF representors, and when traffic fails over to stand-by
VF which is exposed using macvtap device to guest VM, eswitch fails to
switch the traffic to the stand-by VF representor. This occurs because
there is no knowledge at eswitch level of the stand-by representor
device.
Solution:
---------
Using standard bonding driver, a bond netdevice is created over VF
representor device which is used for offloading tc rules.
Two VF representors are bonded together, one for the passthrough VF
device and another one for the stand-by VF device.
With this solution, mlx5 driver listens to the failover events
occuring at the bond device level to failover traffic to either of
the active VF representor of the bond.
a. VM with netfailover device of VF pass-thru (PT) device and virtio-net
paravirtualized device with same MAC-address to handle failover
traffics at VM level.
b. Host bond is active-standby mode, with the lower devices being the VM
VF PT representor, and the representor of the 2nd VF to handle
failover traffics at Hypervisor/V-Switch OVS level.
- During the steady state (fast datapath): set the bond active
device to be the VM PT VF representor.
- During failover: apply bond failover to the second VF representor
device which connects to the VM non-accelerated path.
c. E-Switch ingress/egress ACL tables to support failover traffics at
E-Switch level
I. E-Switch egress ACL with forward-to-vport rule:
- By default, eswitch vport egress acl forward packets to its
counterpart NIC vport.
- During port failover, the egress acl forward-to-vport rule will
be added to e-switch vport of passive/in-active slave VF
representor
to forward packets to other e-switch vport ie. the active slave
representor's e-switch vport to handle egress "failover"
traffics.
- Using lower change netdev event to detect a representor is a
lower
dev (slave) of bond and becomes active, adding egress acl
forward-to-vport rule of all other slave netdevs to forward to
this
representor's vport.
- Using upper change netdev event to detect a representor unslaving
from bond device to delete its vport's egress acl forward-to-vport
rule.
II. E-Switch ingress ACL metadata reg_c for match
- Bonded representors' vorts sharing tc block have the same
root ingress acl table and a unique metadata for match.
- Traffics from both representors's vports will be tagged with same
unique metadata reg_c.
- Using upper change netdev event to detect a representor
enslaving/unslaving from bond device to setup shared root ingress
acl and unique metadata.
2) From Alex Vesker (2): Slpit RX and TX lock for parallel rule insertion in
software steering
3) Eli Britstein (2): Optimize performance for IPv4/IPv6 ethertype use the HW
ip_version register rather than parsing eth frames for ethertype.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Make tcp_ld_RTO_revert() helper available to IPv6, and
implement RFC 6069 :
Quoting this RFC :
3. Connectivity Disruption Indication
For Internet Protocol version 6 (IPv6) [RFC2460], the counterpart of
the ICMP destination unreachable message of code 0 (net unreachable)
and of code 1 (host unreachable) is the ICMPv6 destination
unreachable message of code 0 (no route to destination) [RFC4443].
As with IPv4, a router should generate an ICMPv6 destination
unreachable message of code 0 in response to a packet that cannot be
delivered to its destination address because it lacks a matching
entry in its routing table.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
SJA1105, being AVB/TSN switches, provide hardware assist for the
Credit-Based Shaper as described in the IEEE 8021Q-2018 document.
First generation has 10 shapers, freely assignable to any of the 4
external ports and 8 traffic classes, and second generation has 16
shapers.
The Credit-Based Shaper tables are accessed through the dynamic
reconfiguration interface, so we have to restore them manually after a
switch reset. The tables are backed up by the static config only on
P/Q/R/S, and we don't want to add custom code only for that family,
since the procedure that is in place now works for both.
Tested with the following commands:
data_rate_kbps=67000
port_transmit_rate_kbps=1000000
idleslope=$data_rate_kbps
sendslope=$(($idleslope - $port_transmit_rate_kbps))
locredit=$((-0x80000000))
hicredit=$((0x7fffffff))
tc qdisc add dev swp2 root handle 1: mqprio hw 0 num_tc 8 \
map 0 1 2 3 4 5 6 7 \
queues 1@0 1@1 1@2 1@3 1@4 1@5 1@6 1@7
tc qdisc replace dev swp2 parent 1:1 cbs \
idleslope $idleslope \
sendslope $sendslope \
hicredit $hicredit \
locredit $locredit \
offload 1
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add Nik's torture tests as a new set to stress the replace and cleanup
paths.
Torture test created by Nikolay Aleksandrov and then I adapted to
selftest and added IPv6 version.
Signed-off-by: David Ahern <dsahern@kernel.org>
Signed-off-by: Nikolay Aleksandrov <nikolay@cumulusnetworks.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
If a UMEM is present on a queue when an interface/queue pair is being
enabled, the driver will try to prepare the Rx buffers in advance to
improve performance. However, if fill queue is shorter than HW Rx ring,
the driver will report failure after getting the last address from the
fill queue.
This still lets the driver process the packets correctly during the NAPI
poll, but leads to a constant NAPI rescheduling. Not allocating the
buffers in advance would result in a potential performance decrease.
Commit d57d76428a ("xsk: Add API to check for available entries in FQ")
provides an API that lets drivers check the number of addresses that the
fill queue holds.
Notify the user if fill queue is not long enough to prepare all buffers
before packet processing starts, and allocate the buffers during the
NAPI poll. If the fill queue size is sufficient, prepare Rx buffers in
advance.
Signed-off-by: Krzysztof Kazimierczak <krzysztof.kazimierczak@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Change the locking flow to support RX and TX locks, splitting
the single lock to two will allow inserting rules in parallel
for RX and TX parts of the FDB.
Locking the dr_domain will be done by locking the RX domain
and the TX domain locks, this is mostly used for control operations
on the dr_domain. When inserting rules for RX or TX the single
nic_doamin RX or TX lock will be used. Splitting the lock is safe since
RX and TX domains are logically separated from each other, shared
objects such the send-ring and memory pool are protected by locks.
Signed-off-by: Alex Vesker <valex@mellanox.com>
Reviewed-by: Mark Bloch <markb@mellanox.com>
Reviewed-by: Erez Shitrit <erezsh@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Adding this lock will allow writing steering entries without
locking the dr_domain and allow parallel insertion.
Signed-off-by: Alex Vesker <valex@mellanox.com>
Reviewed-by: Mark Bloch <markb@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
