This patchset adds support for Innova IPSec network interface card.
About Innova device:
--------------------
Innova is a network card with a ConnectX chip and an FPGA chip as a
bump-on-the-wire.
Internal
+----------+ Link +-----------------+
| +--------------+ FPGA | +------+
| ConnectX | | Shell +--+ QSFP |
| +--------------+ +-------+ | | Port |
+----------+ I2C | | SBU | | +------+
| +-------+ |
+--+----------+---+
| |
+--+--+ +---+---+
| DDR | | Flash |
+-----+ +-------+
The FPGA synthesized logic is loaded from dedicated flash storage and has
access to its own dedicated DDR RAM.
The ConnectX chip firmware programs the FPGA by accessing its configuration
space over either the slow internal I2C link or the high-speed internal link.
The FPGA logic is divided into a "Shell" and a "Sandbox Unit" (SBU).
mlx5_core driver (with CONFIG_MLX5_FPGA) handles all shell functionality,
while other components may handle the various SBU functionalities.
The driver opens high-speed reliable communication channels with the shell and
the SBU over the internal link.
These channels may be used for high-bandwidth configuration or for SBU-specific
out-of-band data paths.
About Innova IPSec device:
--------------------------
Innova IPSec is a network card that allows offloading IPSec cryptography operations
from the host CPU to the NIC. It is an Innova card with an IPSec SBU.
The hardware keeps the database of IPSec Security Associations (SADB) in the FPGA's
DDR memory.
Internal
+----------+ Link +-----------------+
| +--------------+ FPGA | +------+
| ConnectX | | Shell +--+ QSFP |
| +--------------+ +-------+ | | Port |
+----------+ Internal I2C | | IPSec | | +------+
| | SBU | |
| +-------+ |
+--+----------+---+
| |
+--+--+ +---+---+
| DDR | | |
| | | Flash |
|SADB | | |
+-----+ +-------+
Modes and ciphers:
Currently the following modes and ciphers are supported:
IPv4 and IPv6
ESP tunnel and transport modes
AES 128 and 256 bit encryption, with GCM authentication (RFC4106)
IV is generated using seqiv, in sync with Linux's geniv.
More modes and ciphers may be added later.
Notes:
In the future similar functionality will be included in a single-chip NIC.
About the driver:
-----------------
Patches 1-4 prepare some existing driver code for the new feature:
* Add support for reserved GIDs in the hardware GID table
* Allow multiple modules to enable hardware RoCE support independently
Patches 5-6 define structs and helper functions for QP work-queues.
Patches 7-11 add various FPGA-related features required for Innova.
IPSec.
Patch 12 adds abstraction layer for Mellanox IPSec-offload capable devices.
atches 13-16 add IPSec offload support to the mlx5 netdevice.
This driver services the new IPSec offload API introduced in commit
d77e38e612 ("xfrm: Add an IPsec hardware offloading API")
Configuration Path:
If Innova IPSec device is detected, the mlx5e netdevice gets the new
NETIF_F_HW_ESP feature and the xdo callbacks, indicating ESP offload
capabilities, and also the matching TX checksum and GSO features.
The driver configures offloaded Security Associations (SAs) by sending
an ADD_SA or DEL_SA message to the IPSec SBU, which updates the SADB in DDR.
These messages and their responses are sent over a high-speed channel.
Counters for ethtool are retrieved by the driver from the SBU.
Data path:
On receive path, the SBU decrypts ESP packets which match the offloaded SADB,
but keeps them encapsulated.
The SBU injects metadata (Mellanox owned ethertype) indicating that crypto-offload
has taken place, the SA with which it was done, and the authentication result.
The ConnectX chip performs RX checksum offload on the packet, and RSS using the
ESP SPI value. The driver detects the special ethertype, and attaches a struct
secpath to the RX SKB, including flags to indicate that crypto offload took place,
the authentication result, and which xfrm_state was used for decryption, in the
olen and ovec members. The RX SKB may have useful CHECKSUM_COMPLETE. A separate
patchset will add support for that in the xfrm stack.
On transmit path, the stack encapsulates the packet but does not encrypt it, and
indicates in the SKB's secpath that crypto offload is to be performed and the SA
to use to do so.
The driver avoids performing crypto-offload for ESP fragments, and packets with
IP options, as the SBU cannot currently do that. For eligible packets, the driver
prepends a special ethertype with metadata instructing the hardware to perform crypto offload.
The stack builds regular (non-GSO) SKBs so that they contain a placeholder for the ESP trailer.
The driver trims it off, because the SBU automatically appends the trailer for offloaded packets.
The ConnectX chip performs TX checksum offload on inner UDP or TCP packets,
and GSO for TCP packets (duplicating the prepended metadata).
The segmented packets then undergo encryption in the SBU before going on the wire.
Performance:
We measure single stream of TCP on Intel(R) Xeon(R) CPU E5-2643 v2 @3.50GHz
Using AES-NI with ESP GSO we get constant 4.1 Gbps.
Using crypto offload we get constant 18 Gbps.
Note that these numbers require CHECKSUM_COMPLETE support in XFRM, which we submit separately.
- Ilan Tayari
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1
iQEcBAABAgAGBQJZUmf1AAoJEEg/ir3gV/o+ukIIALp/5+E1W0cC9xvY1X9dTETW
cKsHvDJ7G1CxUy18W8Mf9z+WOqC6hGCqS+yicOb+umfIqkTcLHDb2irlqprYLC+F
oYl1HqgHTaiAYByqL90qiyPcFbfsaNIqA9KOsED2qdZ1yxjoYBiJnSDZDAdO/0lN
Lt1czNswFc5ovnEUGn8bkjLZZH2pJoJWEI4g4hN9cq33BLLq8A795F/ZjwCJTQ1X
qXdKcEmktBrgZiSiTVFxxpQVhO/uB0HmzaZzrY1k1P5e6yhHEr422mcOcF9KcSL4
aeyRYHjoIh51vPMbScPjvfbO/PwooU3LWLlxLVNLG0MmkSaGyJeUXg/wHsGI910=
=JN0A
-----END PGP SIGNATURE-----
Merge tag 'mlx5-updates-2017-06-27' of git://git.kernel.org/pub/scm/linux/kernel/git/saeed/linux
Saeed Mahameed says:
====================
mlx5-updates-2017-06-27 (Innova IPsec offload support)
This patchset adds support for Innova IPSec network interface card.
About Innova device:
--------------------
Innova is a network card with a ConnectX chip and an FPGA chip as a
bump-on-the-wire.
Internal
+----------+ Link +-----------------+
| +--------------+ FPGA | +------+
| ConnectX | | Shell +--+ QSFP |
| +--------------+ +-------+ | | Port |
+----------+ I2C | | SBU | | +------+
| +-------+ |
+--+----------+---+
| |
+--+--+ +---+---+
| DDR | | Flash |
+-----+ +-------+
The FPGA synthesized logic is loaded from dedicated flash storage and has
access to its own dedicated DDR RAM.
The ConnectX chip firmware programs the FPGA by accessing its configuration
space over either the slow internal I2C link or the high-speed internal link.
The FPGA logic is divided into a "Shell" and a "Sandbox Unit" (SBU).
mlx5_core driver (with CONFIG_MLX5_FPGA) handles all shell functionality,
while other components may handle the various SBU functionalities.
The driver opens high-speed reliable communication channels with the shell and
the SBU over the internal link.
These channels may be used for high-bandwidth configuration or for SBU-specific
out-of-band data paths.
About Innova IPSec device:
--------------------------
Innova IPSec is a network card that allows offloading IPSec cryptography operations
from the host CPU to the NIC. It is an Innova card with an IPSec SBU.
The hardware keeps the database of IPSec Security Associations (SADB) in the FPGA's
DDR memory.
Internal
+----------+ Link +-----------------+
| +--------------+ FPGA | +------+
| ConnectX | | Shell +--+ QSFP |
| +--------------+ +-------+ | | Port |
+----------+ Internal I2C | | IPSec | | +------+
| | SBU | |
| +-------+ |
+--+----------+---+
| |
+--+--+ +---+---+
| DDR | | |
| | | Flash |
|SADB | | |
+-----+ +-------+
Modes and ciphers:
Currently the following modes and ciphers are supported:
IPv4 and IPv6
ESP tunnel and transport modes
AES 128 and 256 bit encryption, with GCM authentication (RFC4106)
IV is generated using seqiv, in sync with Linux's geniv.
More modes and ciphers may be added later.
Notes:
In the future similar functionality will be included in a single-chip NIC.
About the driver:
-----------------
Patches 1-4 prepare some existing driver code for the new feature:
* Add support for reserved GIDs in the hardware GID table
* Allow multiple modules to enable hardware RoCE support independently
Patches 5-6 define structs and helper functions for QP work-queues.
Patches 7-11 add various FPGA-related features required for Innova.
IPSec.
Patch 12 adds abstraction layer for Mellanox IPSec-offload capable devices.
atches 13-16 add IPSec offload support to the mlx5 netdevice.
This driver services the new IPSec offload API introduced in commit
d77e38e612 ("xfrm: Add an IPsec hardware offloading API")
Configuration Path:
If Innova IPSec device is detected, the mlx5e netdevice gets the new
NETIF_F_HW_ESP feature and the xdo callbacks, indicating ESP offload
capabilities, and also the matching TX checksum and GSO features.
The driver configures offloaded Security Associations (SAs) by sending
an ADD_SA or DEL_SA message to the IPSec SBU, which updates the SADB in DDR.
These messages and their responses are sent over a high-speed channel.
Counters for ethtool are retrieved by the driver from the SBU.
Data path:
On receive path, the SBU decrypts ESP packets which match the offloaded SADB,
but keeps them encapsulated.
The SBU injects metadata (Mellanox owned ethertype) indicating that crypto-offload
has taken place, the SA with which it was done, and the authentication result.
The ConnectX chip performs RX checksum offload on the packet, and RSS using the
ESP SPI value. The driver detects the special ethertype, and attaches a struct
secpath to the RX SKB, including flags to indicate that crypto offload took place,
the authentication result, and which xfrm_state was used for decryption, in the
olen and ovec members. The RX SKB may have useful CHECKSUM_COMPLETE. A separate
patchset will add support for that in the xfrm stack.
On transmit path, the stack encapsulates the packet but does not encrypt it, and
indicates in the SKB's secpath that crypto offload is to be performed and the SA
to use to do so.
The driver avoids performing crypto-offload for ESP fragments, and packets with
IP options, as the SBU cannot currently do that. For eligible packets, the driver
prepends a special ethertype with metadata instructing the hardware to perform crypto offload.
The stack builds regular (non-GSO) SKBs so that they contain a placeholder for the ESP trailer.
The driver trims it off, because the SBU automatically appends the trailer for offloaded packets.
The ConnectX chip performs TX checksum offload on inner UDP or TCP packets,
and GSO for TCP packets (duplicating the prepended metadata).
The segmented packets then undergo encryption in the SBU before going on the wire.
Performance:
We measure single stream of TCP on Intel(R) Xeon(R) CPU E5-2643 v2 @3.50GHz
Using AES-NI with ESP GSO we get constant 4.1 Gbps.
Using crypto offload we get constant 18 Gbps.
Note that these numbers require CHECKSUM_COMPLETE support in XFRM, which we submit separately.
- Ilan Tayari
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Ivan Khoronzhuk says:
====================
net: fix sw timestamping for non PTP packets
This series contains several corrections connected with timestamping
for cpsw and netcp drivers based on same cpts module.
Based on net/next
====================
Reviewed-by: Grygorii Strashko <grygorii.strashko@ti.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
There is cpts function to check if packet can be timstamped with cpts.
Seems that ptp_classify_raw cover all cases listed with "case".
Signed-off-by: Ivan Khoronzhuk <ivan.khoronzhuk@linaro.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
The cpts can timestmap only ptp packets at this moment, so driver
cannot mark every packet as though it's going to be timestamped,
only because h/w timestamping for given skb is enabled with
SKBTX_HW_TSTAMP. It doesn't allow to use sw timestamping, as result
outgoing packet is not timestamped at all if it's not PTP and h/w
timestamping is enabled. So, fix it by setting SKBTX_IN_PROGRESS
only for PTP packets.
Signed-off-by: Ivan Khoronzhuk <ivan.khoronzhuk@linaro.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Move sw timestamp function close to channel submit function.
Signed-off-by: Ivan Khoronzhuk <ivan.khoronzhuk@linaro.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Using netdev_<level>(netdev, "%s: ...", netdev->name) duplicates the
name in the output. Remove those uses.
Miscellanea:
o Use the netif_<level> convenience macros at the same time
Signed-off-by: Joe Perches <joe@perches.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Trivial fix to spelling mistake in mlx4_dbg debug message
Signed-off-by: Colin Ian King <colin.king@canonical.com>
Acked-by: Tariq Toukan <tariqt@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Trivial fix to spelling mistake in netif_info message
Signed-off-by: Colin Ian King <colin.king@canonical.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Since the PHY used is internal, simply set phy-mode as internal.
Signed-off-by: Corentin Labbe <clabbe.montjoie@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Since the PHY used is internal, simply set phy-mode as internal.
Signed-off-by: Corentin Labbe <clabbe.montjoie@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Since the PHY used is internal, simply set phy-mode as internal.
Signed-off-by: Corentin Labbe <clabbe.montjoie@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Since the PHY used is internal, simply set phy-mode as internal.
Signed-off-by: Corentin Labbe <clabbe.montjoie@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Since the PHY used is internal, simply set phy-mode as internal.
Signed-off-by: Corentin Labbe <clabbe.montjoie@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The current way to find if the phy is internal is to compare DT phy-mode
and emac_variant/internal_phy.
But it will negate a possible future SoC where an external PHY use the
same phy mode than the internal one.
By using phy-mode = "internal" we permit to have an external PHY with
the same mode than the internal one.
Reported-by: André Przywara <andre.przywara@arm.com>
Signed-off-by: Corentin Labbe <clabbe.montjoie@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The bond_options.c file contains multiple netdev_info statements that clutter kernel output.
This patch replaces all netdev_info with netdev_dbg and adds a netdev_dbg statement for the
packets per slave parameter. Also fixes misalignment at line 467.
Suggested-by: Joe Perches <joe@perches.com>
Signed-off-by: Michael J Dilmore <michael.j.dilmore@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Jakub Kicinski says:
====================
nfp: get_phys_port_name for representors and SR-IOV reorder
This series starts by making the error message if FW cannot be located
easier to understand. Then I move some functions from PCI probe files
into library code (nfpcore) where they belong, and remove one function
which is never used.
Next few patches equip representors with nfp_port structure and make
their NDOs fully shared (not defined in apps), thanks to which we can
easily determine which netdevs are NFP's by comparing the NDO pointers.
10th patch makes use of the shared NDOs and nfp_ports to deliver
netdev-type independent .ndo_get_phys_port_name() implementation.
Patches 11 and 12 reorder the nfp_app SR-IOV callbacks with enabling
SR-IOV VFs. Unfortunately due to how PCI subsystem works we can't
guarantee being able to disable SR-IOV at exit or that it will be
disabled when we first probe... We must therefore make sure FW is
able to deal with being loaded while SR-IOV is already on.
Patch 13 fixes potential deadlock when enabling SR-IOV happens at
the same time as port state refresh. Note that this can't happen
at this point, since Flower doesn't refresh ports... but lockdep
doesn't know about such details and we will have to deal with this
sooner or later anyway.
Last but not least a new Kconfig is added to make sure those who
don't care about flower offloads have a way of not including the
code in their kernels. Thanks to nfp_app separation this costs us
a single ifdef and excluding flower files from the build.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Give users an option not to build the flower-offload related code.
Signed-off-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Reviewed-by: Simon Horman <simon.horman@netronome.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Since we grab pf->lock around pci_enable_sriov() we can no longer
safely queue work which may also grab that lock onto system workqueue.
pci_enable_sriov() will flush system workqueue as part to wait for VF
probing.
Signed-off-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Reviewed-by: Simon Horman <simon.horman@netronome.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
We previously assumed that app callback can be guaranteed to be
executed before SR-IOV is actually enabled. Given that we can't
guarantee that SR-IOV will be disabled during probe or that we
will be able to disable it on remove, we should reorder the callbacks.
We should also call the app's sriov_enable if SR-IOV was enabled
during probe.
Application FW must be able to disable VFs internally and not depend
on them being removed at PCIe level.
Signed-off-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Reviewed-by: Simon Horman <simon.horman@netronome.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
We assumed that when we probe number of enabled VFs will be at 0.
This doesn't have to be the case for example if previous driver left
SR-IOV enabled due to some VFs being assigned. Read the number of VFs
enabled. Fail probe if it's above current FWs limit.
Signed-off-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Reviewed-by: Simon Horman <simon.horman@netronome.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Make nfp_port_get_phys_port_name() support new port types and
wire it up to representors' struct net_device_ops.
Signed-off-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Reviewed-by: Simon Horman <simon.horman@netronome.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Based on struct net_device_ops figure out if netdev is a nfp_repr.
Use this knowledge to convert netdev directly to nfp_port.
Signed-off-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Reviewed-by: Simon Horman <simon.horman@netronome.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Apps shouldn't declare their own struct net_device_ops for
representors, this makes sharing code harder. Add necessary
nfp_app callbacks and move the definition of representors'
struct net_device_ops to common code.
Signed-off-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Reviewed-by: Simon Horman <simon.horman@netronome.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Thanks to the fact that all representors will now have an nfp_port,
we can depend on information there to provide a app-independent
.ndo_get_stats64().
Signed-off-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Reviewed-by: Simon Horman <simon.horman@netronome.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
nfp_port is an abstraction which is supposed to allow us sharing
code between different netdev types (vNIC vs repr). Spawn ports
for PFs and VFs to enable this sharing.
Signed-off-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Reviewed-by: Simon Horman <simon.horman@netronome.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add a cleanup callback for undoing what app init callback did.
Make flower allocate its private structure on init and free
it from the new callback.
While at it remember to set the app pointer to NULL on the
error path to avoid any races while probe path unwinds.
Signed-off-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Reviewed-by: Simon Horman <simon.horman@netronome.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Remove unused nfp_cpp_area_check_range() function.
Signed-off-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Reviewed-by: Simon Horman <simon.horman@netronome.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Move most of the helper for mapping RTsyms from nfp_net_main.c
to nfpcore. Use the new helper directly for mapping MAC statistics,
since they don't need to include the PCIe interface ID in the symbol
name.
Signed-off-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Reviewed-by: Simon Horman <simon.horman@netronome.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
nfp_net_map_area() is a helper for mapping areas of NFP memory
defined in nfp_net_main.c. Move it to nfpcore to allow reuse
and rename accordingly. Create an additional helper -
nfp_cpp_area_alloc_acquire() the opposite of already existing
nfp_cpp_area_release_free().
Signed-off-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Reviewed-by: Simon Horman <simon.horman@netronome.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
We support application FW being either loaded automatically at
boot from flash or (more commonly) by the driver from disk.
If FW is not found on disk and nothing is preloaded users are
faced with this unintuitive error:
nfp 0000:04:00.0: nfp: Failed to find PF symbol _pf0_net_bar0
We can do better. Since we rely on symbol table being present -
check early if it could be correctly read out of from the device
and if not print a more informative message.
Signed-off-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Reviewed-by: Simon Horman <simon.horman@netronome.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Paolo Abeni says:
====================
ipv6: udp: exploit dev_scratch helpers
When bringing in the recent cache optimization for the UDP protocol, I forgot
to leverage the newly introduced scratched area helpers in the UDPv6 code path.
As a result, the UDPv6 implementation suffers some unnecessary performance
penality when compared to v4.
This series aim to bring back UDPv6 on equal footing in respect to v4.
The first patch moves the shared helpers to the common include files, while
the second uses them in the UDPv6 code.
This gives 5-8% performance improvement for a system under flood with small
UDPv6 packets. The performance delta is less than the one reported on the
original patch set because the UDPv6 code path already leveraged some of the
optimization.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
The commit b65ac44674 ("udp: try to avoid 2 cache miss on dequeue")
leveraged the scratched area helpers for UDP v4 but I forgot to
update accordingly the IPv6 code path.
This change extends the scratch area usage to the IPv6 code, synching
the two implementations and giving some performance benefit.
IPv6 is again almost on the same level of IPv4, performance-wide.
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
So that they can be later used by the IPv6 code, too.
Also lift the comments a bit.
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
If icsk_ulp_ops is unset, it dereferences a null ptr.
Add a null ptr check.
BUG: KASAN: null-ptr-deref in copy_to_user include/linux/uaccess.h:168 [inline]
BUG: KASAN: null-ptr-deref in do_tcp_getsockopt.isra.33+0x24f/0x1e30 net/ipv4/tcp.c:3057
Read of size 4 at addr 0000000000000020 by task syz-executor1/15452
Signed-off-by: Dave Watson <davejwatson@fb.com>
Reported-by: "Levin, Alexander (Sasha Levin)" <alexander.levin@verizon.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The access to the wrong variable could lead to a NULL dereference and
possibly other invalid memory reads in vxlan newlink/changelink requests
with a IFLA_MTU attribute.
Fixes: a985343ba9 "vxlan: refactor verification and application of configuration"
Signed-off-by: Matthias Schiffer <mschiffer@universe-factory.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
It dates back from 2.1.16 and is obsolete since 2.1.68 when the current
rule system has been introduced.
Signed-off-by: Vincent Bernat <vincent@bernat.im>
Signed-off-by: David S. Miller <davem@davemloft.net>
In the TX data path, prepend a special metadata ethertype which
instructs the hardware to perform cryptography.
In addition, fill Software-Parser segment in TX descriptor so
that the hardware may parse the ESP protocol, and perform TX
checksum offload on the inner payload.
Support GSO, by providing the inverse of gso_size in the metadata.
This allows the FPGA to update the ESP header (seqno and seqiv) on the
resulting packets, by calculating the packet number within the GSO
back from the TCP sequence number.
Note that for GSO SKBs, the stack does not include an ESP trailer,
unlike the non-GSO case.
Signed-off-by: Ilan Tayari <ilant@mellanox.com>
Signed-off-by: Yossi Kuperman <yossiku@mellanox.com>
Signed-off-by: Yevgeny Kliteynik <kliteyn@mellanox.com>
Signed-off-by: Boris Pismenny <borisp@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
In RX data path, the hardware prepends a special metadata ethertype
which indicates that the packet underwent decryption, and the result of
the authentication check.
Communicate this to the stack in skb->sp.
Make wqe_size large enough to account for the injected metadata.
Support only Linked-list RQ type.
IPSec offload RX packets may have useful CHECKSUM_COMPLETE information,
which the stack may not be able to use yet.
Signed-off-by: Ilan Tayari <ilant@mellanox.com>
Signed-off-by: Yossi Kuperman <yossiku@mellanox.com>
Signed-off-by: Yevgeny Kliteynik <kliteyn@mellanox.com>
Signed-off-by: Boris Pismenny <borisp@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Add Innova IPSec ESP crypto offload configuration paths.
Detect Innova IPSec device and set the NETIF_F_HW_ESP flag.
Configure Security Associations using the API introduced in a previous
patch.
Add Software-parser hardware descriptor layout
Software-Parser (swp) is a hardware feature in ConnectX which allows the
host software to specify protocol header offsets in the TX path, thus
overriding the hardware parser.
This is useful for protocols that the ASIC may not be able to parse on
its own.
Note that due to inline metadata, XDP is not supported in Innova IPSec.
Signed-off-by: Ilan Tayari <ilant@mellanox.com>
Signed-off-by: Yossi Kuperman <yossiku@mellanox.com>
Signed-off-by: Yevgeny Kliteynik <kliteyn@mellanox.com>
Signed-off-by: Boris Pismenny <borisp@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Add routines for manipulating the hardware IPSec SA database (SADB).
In Innova IPSec, a Security Association (SA) is added or deleted
via a command message over the SBU connection.
The HW then sends a response message over the same connection.
Add implementation for Innova IPSec (FPGA-based) hardware.
These routines will be used by the IPSec offload support in a later patch
However they may also be used by others such as RDMA and RoCE IPSec.
mlx5/accel is a middle acceleration layer to allow mlx5e and other ULPs
to work directly with mlx5_core rather than Innova FPGA or other mlx5
acceleration providers.
In this patchset we add Innova IPSec support and mlx5/accel delegates
IPSec offloads to Innova routines.
In the future, when IPSec/TLS or any other acceleration gets integrated
into ConnectX chip, mlx5/accel layer will provide the integrated
acceleration, rather than the Innova one.
Signed-off-by: Ilan Tayari <ilant@mellanox.com>
Signed-off-by: Boris Pismenny <borisp@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Add interface to initialize and interact with Innova FPGA SBU
connections.
A client driver may use these functions to set up a high-speed DMA
connection with its SBU hardware logic, and send/receive messages
over this connection.
A later patch in this patchset will make use of these functions for
Innova IPSec offload in mlx5 Ethernet driver.
Add commands to retrieve Innova FPGA SBU capabilities, and to
read/write Innova FPGA configuration space registers and memory,
over internal I2C.
At high level, the FPGA configuration space is divided such:
0x00000000 - 0x007fffff is reserved for the SBU
0x00800000 - 0xffffffff is reserved for the Shell
0x400000000 - ... is DDR memory
A later patchset will add support for accessing FPGA CrSpace and memory
over a high-speed connection. This is the reason for the ACCESS_TYPE
enumeration, which currently only supports I2C.
Signed-off-by: Ilan Tayari <ilant@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
The Innova FPGA includes shell hardware and Sandbox-Unit (SBU) hardware.
The shell hardware is handled by mlx5_core itself, while the SBU is
handled by a client driver.
Reset the SBU to a well-known initial state when initializing a new
device, and set the FPGA to bypass mode when uninitializing a device.
This allows the client driver to assume that its device has been
reset when a new device is detected.
During SBU reset, the FPGA is put into SBU-bypass mode. In this mode
packets do not pass through the SBU, so it cannot affect the network
data stream at all.
A factory-image does not have an SBU, so skip these flows.
Signed-off-by: Ilan Tayari <ilant@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
An FPGA high-speed connection has two endpoints, an FPGA QP and a
ConnectX QP.
Add library routines to create and connect the endpoints of an
FPGA high-speed connection.
These routines allow creating and interacting with both types of
connections: Shell and Sandbox Unit (SBU).
Shell connection provides an interface to the FPGA's address space,
which includes the configuration space and the DDR.
Use of the shell connection will be introduced in a later patchset.
SBU connection provides a command and/or data interface to the
application-specific logic within the FPGA.
Use of the SBU connection will be introduced in a later patch in
this patchset.
Some struct definitions are added to a new header file sdk.h, which
will be extended in later patches in the patchset.
This header file will contain the in-kernel FPGA client driver API.
Signed-off-by: Ilan Tayari <ilant@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
The FPGA QP is a high-bandwidth communication channel between the host
CPU and the FPGA device. It allows performing DMA operations between
host memory and the FPGA logic via the ConnectX chip.
Add ConnectX FW commands which create and manipulate FPGA QPs.
Signed-off-by: Ilan Tayari <ilant@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
The FPGA init and cleanup routines should be called just once per
device.
Move them to the init_once and cleanup_once routines.
Signed-off-by: Ilan Tayari <ilant@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
A QP in ConnectX is a concatenation of RQ and SQ which share a QP-number
and work together.
Add support for allocating and managing the work-queue buffer for a QP, in
a similar way to how SQs and RQs are already supported.
Signed-off-by: Ilan Tayari <ilant@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Move mlx5e_get_cqe routine to wq.h and rename it to
mlx5_cqwq_get_cqe.
This allows it to be used by other CQ users outside of the
ethernet driver code.
A later patch in this patchset will make use of it from
FPGA code for the FPGA high-speed connection.
Signed-off-by: Ilan Tayari <ilant@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Reserved gids are taken by the mlx5_core, report smaller GID table
size to IB core.
Set mlx5_query_roce_port's return value back to int. In case of
error, return an indication. This rolls back some of the change
in commit 50f22fd8ec ("IB/mlx5: Set mlx5_query_roce_port's return value to void")
Change set_roce_addr to use gid_set function, instead of directly
sending the command.
Signed-off-by: Ilan Tayari <ilant@mellanox.com>
Reviewed-by: Leon Romanovsky <leonro@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Previously, only mlx5_ib enabled RoCE on the port, but FPGA needs it as
well.
Add support for counting number of enables, so that FPGA and IB can work
in parallel and independently.
Program the HW to enable RoCE on the first enable call, and program to
disable RoCE on the last disable call.
Signed-off-by: Ilan Tayari <ilant@mellanox.com>
Reviewed-by: Boris Pismenny <borisp@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
