802.1ad, also known as QinQ is an extension to the 802.1q standard, which
is concerned with passing possibly 802.1q-tagged packets through another
VLAN-like tunnel. The format of 802.1ad tag is the same as 802.1q, except
it uses the EtherType of 0x88a8, unlike 802.1q's 0x8100.
Add support for 802.1ad protocol. Most of the configuration is the same
as 802.1q. The difference is that before a port joins an 802.1ad bridge it
needs to be configured to recognize 802.1ad packets as tagged and other
packets (e.g., 802.1q) as untagged.
VXLAN is not currently supported with 802.1ad bridge, so return an error
with an appropriate extack message.
Signed-off-by: Amit Cohen <amcohen@nvidia.com>
Reviewed-by: Petr Machata <petrm@nvidia.com>
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The code in mlxsw_sp_bridge_8021q_port_{join, leave}() can be used also
for 802.1ad bridge.
Move the code to functions called
mlxsw_sp_bridge_vlan_aware_port_{join, leave}() and call them from
mlxsw_sp_bridge_8021q_port_{join, leave}() respectively to enable code
reuse.
Signed-off-by: Amit Cohen <amcohen@nvidia.com>
Reviewed-by: Petr Machata <petrm@nvidia.com>
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Currently, when pushing a PVID at ingress, mlxsw always uses 802.1q
EtherType.
Make this EtherType configurable by extending mlxsw_sp_port_pvid_set()
with an EtherType argument.
This is a preparation for QinQ support, that needs to push a PVID with
802.1ad EtherType.
Signed-off-by: Amit Cohen <amcohen@nvidia.com>
Reviewed-by: Petr Machata <petrm@nvidia.com>
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
By default, the device considers both 802.1ad and 802.1q packets as tagged,
but this is not supported by the driver. It only supports VLAN and bridge
devices that use 802.1q protocol.
Instead, configure the device to only treat 802.1q packets as tagged
packets.
Signed-off-by: Amit Cohen <amcohen@nvidia.com>
Reviewed-by: Petr Machata <petrm@nvidia.com>
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
et_vlan field is used to configure which EtherType is used when VLAN is
pushed at ingress (for untagged packets or for QinQ push mode).
It will be used to configure tagging with ether_type1 (i.e., 0x88A8) for
QinQ mode.
Signed-off-by: Amit Cohen <amcohen@nvidia.com>
Reviewed-by: Petr Machata <petrm@nvidia.com>
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
SPVC configures the port to identify packets as untagged / single tagged /
double tagged packets based on the packet EtherTypes.
It will be used to classify 802.1q packets as untagged and 802.1ad packets
as tagged when received by ports member in a 802.1ad bridge.
Signed-off-by: Amit Cohen <amcohen@nvidia.com>
Reviewed-by: Petr Machata <petrm@nvidia.com>
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Huazhong Tan says:
====================
net: hns3: updates for -next
This series includes some updates for the HNS3 ethernet driver.
#1~#6: add some updates related to the checksum offload.
#7: add support for multiple TCs' MAC pauce mode.
====================
Link: https://lore.kernel.org/r/1606535510-44346-1-git-send-email-tanhuazhong@huawei.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Bellow HNAE3_DEVICE_VERSION_V3, MAC pause mode just support one
TC, when enabled multiple TCs, force enable PFC mode.
HNAE3_DEVICE_VERSION_V3 can support MAC pause mode on multiple
TCs, so when enable multiple TCs, just keep MAC pause mode,
and enable PFC mode just according to the user settings.
Signed-off-by: Yonglong Liu <liuyonglong@huawei.com>
Signed-off-by: Huazhong Tan <tanhuazhong@huawei.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
For the device whose version is above V3(include V3), the hardware
can do checksum offload for the non-tunnel udp packet, who has
a dest port as the IANA assigned. So add a check for devcie's verion
in hns3_tunnel_csum_bug().
Signed-off-by: Huazhong Tan <tanhuazhong@huawei.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Since TX hardware checksum and RX completion checksum have been
supported now, so add related information in hns3_dbg_bd_info().
Signed-off-by: Huazhong Tan <tanhuazhong@huawei.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
For the device who has the capability to handle udp tunnel
checksum segmentation, add support for it.
Signed-off-by: Huazhong Tan <tanhuazhong@huawei.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Currently, device V1 and V2 do not support segmentation
offload for UDP based tunnel packet who needs outer UDP
checksum offload, so there is a workaround in the driver
to set the checksum of the outer UDP checksum as zero. This
is not what the user wants, so remove this feature for
device V1 and V2, add support for it later(when the device
has the ability to do that).
Signed-off-by: Huazhong Tan <tanhuazhong@huawei.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
For the device that supports TX hardware checksum, the hardware
can calculate the checksum from the start and fill the checksum
to the offset position, which reduces the operations of
calculating the type and header length of L3/L4. So add this
feature for the HNS3 ethernet driver.
The previous simple BD description is unsuitable, rename it as
HW TX CSUM.
Signed-off-by: Huazhong Tan <tanhuazhong@huawei.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
In some cases (for example ip fragment), hardware will
calculate the checksum of whole packet in RX, and setup
the HNS3_RXD_L2_CSUM_B flag in the descriptor, so add
support to utilize this checksum.
Signed-off-by: Huazhong Tan <tanhuazhong@huawei.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Pull vhost fixes from Michael Tsirkin:
"A couple of minor fixes"
* tag 'for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mst/vhost:
vhost-vdpa: fix page pinning leakage in error path (rework)
vringh: fix vringh_iov_push_*() documentation
vhost scsi: fix lun reset completion handling
For IPV6_2292PKTOPTIONS, do_ipv6_getsockopt() stores the user pointer
optval in the msg_control field of the msghdr.
Hence, sparse rightfully warns at ./net/ipv6/ipv6_sockglue.c:1151:33:
warning: incorrect type in assignment (different address spaces)
expected void *msg_control
got char [noderef] __user *optval
Since commit 1f466e1f15 ("net: cleanly handle kernel vs user buffers for
->msg_control"), user pointers shall be stored in the msg_control_user
field, and kernel pointers in the msg_control field. This allows to
propagate __user annotations nicely through this struct.
Store optval in msg_control_user to properly record and propagate the
memory space annotation of this pointer.
Note that msg_control_is_user is set to true, so the key invariant, i.e.,
use msg_control_user if and only if msg_control_is_user is true, holds.
The msghdr is further used in the six alternative put_cmsg() calls, with
msg_control_is_user being true, put_cmsg() picks msg_control_user
preserving the __user annotation and passes that properly to
copy_to_user().
No functional change. No change in object code.
Signed-off-by: Lukas Bulwahn <lukas.bulwahn@gmail.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Link: https://lore.kernel.org/r/20201127093421.21673-1-lukas.bulwahn@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Functions tfilter_notify_chain() and tcf_get_next_proto() are always called
with rtnl lock held in current implementation. Moreover, attempting to call
them without rtnl lock would cause a warning down the call chain in
function __tcf_get_next_proto() that requires the lock to be held by
callers. Remove the 'rtnl_held' argument in order to simplify the code and
make rtnl lock requirement explicit.
Signed-off-by: Vlad Buslov <vladbu@nvidia.com>
Link: https://lore.kernel.org/r/20201127151205.23492-1-vladbu@nvidia.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Thomas Falcon says:
====================
ibmvnic: Bug fixes for queue descriptor processing
This series resolves a few issues in the ibmvnic driver's
RX buffer and TX completion processing. The first patch
includes memory barriers to synchronize queue descriptor
reads. The second patch fixes a memory leak that could
occur if the device returns a TX completion with an error
code in the descriptor, in which case the respective socket
buffer and other relevant data structures may not be freed
or updated properly.
v3: Correct length of Fixes tags, requested by Jakub Kicinski
v2: Provide more detailed comments explaining specifically what
reads are being ordered, suggested by Michael Ellerman
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
TX completions received with an error return code are not
being processed properly. When an error code is seen, do not
proceed to the next completion before cleaning up the existing
entry's data structures.
Fixes: 032c5e8284 ("Driver for IBM System i/p VNIC protocol")
Signed-off-by: Thomas Falcon <tlfalcon@linux.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Ensure that received Subordinate Command-Response Queue (SCRQ)
entries are properly read in order by the driver. These queues
are used in the ibmvnic device to process RX buffer and TX completion
descriptors. dma_rmb barriers have been added after checking for a
pending descriptor to ensure the correct descriptor entry is checked
and after reading the SCRQ descriptor to ensure the entire
descriptor is read before processing.
Fixes: 032c5e8284 ("Driver for IBM System i/p VNIC protocol")
Signed-off-by: Thomas Falcon <tlfalcon@linux.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The v9fs file operations were missing the splice_read operations, which
breaks sendfile() of files on such a filesystem. I discovered this while
trying to load an eBPF program using iproute2 inside a 'virtme' environment
which uses 9pfs for the virtual file system. iproute2 relies on sendfile()
with an AF_ALG socket to hash files, which was erroring out in the virtual
environment.
Since generic_file_splice_read() seems to just implement splice_read in
terms of the read_iter operation, I simply added the generic implementation
to the file operations, which fixed the error I was seeing. A quick grep
indicates that this is what most other file systems do as well.
Link: http://lkml.kernel.org/r/20201201135409.55510-1-toke@redhat.com
Fixes: 36e2c7421f ("fs: don't allow splice read/write without explicit ops")
Signed-off-by: Toke Høiland-Jørgensen <toke@redhat.com>
Signed-off-by: Dominique Martinet <asmadeus@codewreck.org>
Stephen reported the following build error for !CONFIG_NET_RX_BUSY_POLL
built kernels:
In file included from fs/select.c:32:
include/net/busy_poll.h: In function 'sk_mark_napi_id_once':
include/net/busy_poll.h:150:36: error: 'const struct sk_buff' has no member named 'napi_id'
150 | __sk_mark_napi_id_once_xdp(sk, skb->napi_id);
| ^~
Fix it by wrapping a CONFIG_NET_RX_BUSY_POLL around the helpers.
Fixes: b02e5a0ebb ("xsk: Propagate napi_id to XDP socket Rx path")
Reported-by: Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Cc: Björn Töpel <bjorn.topel@intel.com>
Link: https://lore.kernel.org/linux-next/20201201190746.7d3357fb@canb.auug.org.au
The current ring buffer logic checks to see if the updating of the event
buffer was interrupted, and if it is, it will try to fix up the before stamp
with the write stamp to make them equal again. This logic is flawed, because
if it is not interrupted, the two are guaranteed to be different, as the
current event just updated the before stamp before allocation. This
guarantees that the next event (this one or another interrupting one) will
think it interrupted the time updates of a previous event and inject an
absolute time stamp to compensate.
The correct logic is to always update the timestamps when traversing to a
new sub buffer.
Cc: stable@vger.kernel.org
Fixes: a389d86f7f ("ring-buffer: Have nested events still record running time stamp")
Signed-off-by: Steven Rostedt (VMware) <rostedt@goodmis.org>
Marc Kleine-Budde says:
====================
pull-request: can 2020-11-30
The first patch is by me an target the tcan4x5x bindings for the m_can driver.
It fixes the error path in the tcan4x5x_can_probe() function.
The next two patches are by Jeroen Hofstee and makes the lost of arbitration
error counters of sja1000 and the sun4i drivers consistent with the other
drivers.
Zhang Qilong contributes two patch that clean up the error path in the c_can
and kvaser_pciefd drivers.
* tag 'linux-can-fixes-for-5.10-20201130' of git://git.kernel.org/pub/scm/linux/kernel/git/mkl/linux-can:
can: kvaser_pciefd: kvaser_pciefd_open(): fix error handling
can: c_can: c_can_power_up(): fix error handling
can: sun4i_can: sun4i_can_err(): don't count arbitration lose as an error
can: sja1000: sja1000_err(): don't count arbitration lose as an error
can: m_can: tcan4x5x_can_probe(): fix error path: remove erroneous clk_disable_unprepare()
====================
Link: https://lore.kernel.org/r/20201130125307.218258-1-mkl@pengutronix.de
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Marc Kleine-Budde says:
====================
pull-request: can-next 2020-11-30
Gustavo A. R. Silva's patch for the pcan_usb driver fixes fall-through warnings
for Clang.
The next 5 patches target the mcp251xfd driver and are by Ursula Maplehurst and
me. They optimizie the TEF and RX path by reducing number of SPI core requests
to set the UINC bit.
The remaining 8 patches target the m_can driver. The first 4 are various
cleanups for the SPI binding driver (tcan4x5x) by Sean Nyekjaer, Dan Murphy and
me. Followed by 4 cleanup patches by me for the m_can and m_can_platform
driver.
* tag 'linux-can-next-for-5.11-20201130' of git://git.kernel.org/pub/scm/linux/kernel/git/mkl/linux-can-next:
can: m_can: m_can_class_unregister(): move right after m_can_class_register()
can: m_can: m_can_plat_remove(): remove unneeded platform_set_drvdata()
can: m_can: remove not used variable struct m_can_classdev::freq
can: m_can: Kconfig: convert the into menu
can: tcan4x5x: tcan4x5x_can_probe(): remove probe failed error message
can: tcan4x5x: remove mram_start and reg_offset from struct tcan4x5x_priv
can: tcan4x5x: rename parse_config() function
can: tcan4x5x: tcan4x5x_clear_interrupts(): remove redundant return statement
can: mcp251xfd: tef-path: reduce number of SPI core requests to set UINC bit
can: mcp251xfd: move struct mcp251xfd_tef_ring definition
can: mcp251xfd: struct mcp251xfd_priv::tef to array of length 1
can: mcp25xxfd: rx-path: reduce number of SPI core requests to set UINC bit
can: mcp251xfd: mcp25xxfd_ring_alloc(): add define instead open coding the maximum number of RX objects
can: pcan_usb_core: fix fall-through warnings for Clang
====================
Link: https://lore.kernel.org/r/20201130141432.278219-1-mkl@pengutronix.de
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
On powerpc, kprobe-direct.tc triggered FTRACE_WARN_ON() in
ftrace_get_addr_new() followed by the below message:
Bad trampoline accounting at: 000000004222522f (wake_up_process+0xc/0x20) (f0000001)
The set of steps leading to this involved:
- modprobe ftrace-direct-too
- enable_probe
- modprobe ftrace-direct
- rmmod ftrace-direct <-- trigger
The problem turned out to be that we were not updating flags in the
ftrace record properly. From the above message about the trampoline
accounting being bad, it can be seen that the ftrace record still has
FTRACE_FL_TRAMP set though ftrace-direct module is going away. This
happens because we are checking if any ftrace_ops has the
FTRACE_FL_TRAMP flag set _before_ updating the filter hash.
The fix for this is to look for any _other_ ftrace_ops that also needs
FTRACE_FL_TRAMP.
Link: https://lkml.kernel.org/r/56c113aa9c3e10c19144a36d9684c7882bf09af5.1606412433.git.naveen.n.rao@linux.vnet.ibm.com
Cc: stable@vger.kernel.org
Fixes: a124692b69 ("ftrace: Enable trampoline when rec count returns back to one")
Signed-off-by: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com>
Signed-off-by: Steven Rostedt (VMware) <rostedt@goodmis.org>
With 5.9 kernel on ARM64, I found ftrace_dump output was broken but
it had no problem with normal output "cat /sys/kernel/debug/tracing/trace".
With investigation, it seems coping the data into temporal buffer seems to
break the align binary printf expects if the static buffer is not aligned
with 4-byte. IIUC, get_arg in bstr_printf expects that args has already
right align to be decoded and seq_buf_bprintf says ``the arguments are saved
in a 32bit word array that is defined by the format string constraints``.
So if we don't keep the align under copy to temporal buffer, the output
will be broken by shifting some bytes.
This patch fixes it.
Link: https://lkml.kernel.org/r/20201125225654.1618966-1-minchan@kernel.org
Cc: <stable@vger.kernel.org>
Fixes: 8e99cf91b9 ("tracing: Do not allocate buffer in trace_find_next_entry() in atomic")
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
Signed-off-by: Minchan Kim <minchan@kernel.org>
Signed-off-by: Steven Rostedt (VMware) <rostedt@goodmis.org>
This patch reverts commit 978defee11 ("tracing: Do a WARN_ON()
if start_thread() in hwlat is called when thread exists")
.start hook can be legally called several times if according
tracer is stopped
screen window 1
[root@localhost ~]# echo 1 > /sys/kernel/tracing/events/kmem/kfree/enable
[root@localhost ~]# echo 1 > /sys/kernel/tracing/options/pause-on-trace
[root@localhost ~]# less -F /sys/kernel/tracing/trace
screen window 2
[root@localhost ~]# cat /sys/kernel/debug/tracing/tracing_on
0
[root@localhost ~]# echo hwlat > /sys/kernel/debug/tracing/current_tracer
[root@localhost ~]# echo 1 > /sys/kernel/debug/tracing/tracing_on
[root@localhost ~]# cat /sys/kernel/debug/tracing/tracing_on
0
[root@localhost ~]# echo 2 > /sys/kernel/debug/tracing/tracing_on
triggers warning in dmesg:
WARNING: CPU: 3 PID: 1403 at kernel/trace/trace_hwlat.c:371 hwlat_tracer_start+0xc9/0xd0
Link: https://lkml.kernel.org/r/bd4d3e70-400d-9c82-7b73-a2d695e86b58@virtuozzo.com
Cc: Ingo Molnar <mingo@redhat.com>
Cc: stable@vger.kernel.org
Fixes: 978defee11 ("tracing: Do a WARN_ON() if start_thread() in hwlat is called when thread exists")
Signed-off-by: Vasily Averin <vvs@virtuozzo.com>
Signed-off-by: Steven Rostedt (VMware) <rostedt@goodmis.org>
It was observed that sending data via batadv over vxlan (on top of
wireguard) reduced the performance massively compared to raw ethernet or
batadv on raw ethernet. A check of perf data showed that the
vxlan_build_skb was calling all the time pskb_expand_head to allocate
enough headroom for:
min_headroom = LL_RESERVED_SPACE(dst->dev) + dst->header_len
+ VXLAN_HLEN + iphdr_len;
But the vxlan_config_apply only requested needed headroom for:
lowerdev->hard_header_len + VXLAN6_HEADROOM or VXLAN_HEADROOM
So it completely ignored the needed_headroom of the lower device. The first
caller of net_dev_xmit could therefore never make sure that enough headroom
was allocated for the rest of the transmit path.
Cc: Annika Wickert <annika.wickert@exaring.de>
Signed-off-by: Sven Eckelmann <sven@narfation.org>
Tested-by: Annika Wickert <aw@awlnx.space>
Link: https://lore.kernel.org/r/20201126125247.1047977-1-sven@narfation.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Paolo Abeni says:
====================
mptcp: avoid workqueue usage for data
The current locking schema used to protect the MPTCP data-path
requires the usage of the MPTCP workqueue to process the incoming
data, depending on trylock result.
The above poses scalability limits and introduces random delays
in MPTCP-level acks.
With this series we use a single spinlock to protect the MPTCP
data-path, removing the need for workqueue and delayed ack usage.
This additionally reduces the number of atomic operations required
per packet and cleans-up considerably the poll/wake-up code.
====================
Link: https://lore.kernel.org/r/cover.1606413118.git.pabeni@redhat.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
We have some tasks triggered by the subflow receive path
which require to access the msk socket status, specifically:
mptcp_clean_una() and mptcp_push_pending()
We have almost everything in place to defer to the msk
release_cb such tasks when the msk sock is owned.
Since the worker is no more used to clean the acked data,
for fallback sockets we need to explicitly flush them.
As an added bonus we can move the wake-up code in __mptcp_clean_una(),
simplify a lot mptcp_poll() and move the timer update under
the data lock.
The worker is now used only to process and send DATA_FIN
packets and do the mptcp-level retransmissions.
Acked-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Reviewed-by: Mat Martineau <mathew.j.martineau@linux.intel.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Extending the data_lock scope in mptcp_incoming_option
we can use that to protect both snd_una and wnd_end.
In the typical case, we will have a single atomic op instead of 2
Acked-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Reviewed-by: Mat Martineau <mathew.j.martineau@linux.intel.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Move the TX skbs allocation in mptcp_sendmsg() scope,
and tentatively pre-allocate a skbs number proportional
to the sendmsg() length.
Use the ssk tx skb cache to prevent the subflow allocation.
This allows removing the msk skb extension cache and will
make possible the later patches.
Acked-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Reviewed-by: Mat Martineau <mathew.j.martineau@linux.intel.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Such spinlock is currently used only to protect the 'owned'
flag inside the socket lock itself. With this patch, we extend
its scope to protect the whole msk receive path and
sk_forward_memory.
Given the above, we can always move data into the msk receive
queue (and OoO queue) from the subflow.
We leverage the previous commit, so that we need to acquire the
spinlock in the tx path only when moving fwd memory.
recvmsg() must now explicitly acquire the socket spinlock
when moving skbs out of sk_receive_queue. To reduce the number of
lock operations required we use a second rx queue and splice the
first into the latter in mptcp_lock_sock(). Additionally rmem
allocated memory is bulk-freed via release_cb()
Acked-by: Florian Westphal <fw@strlen.de>
Co-developed-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Reviewed-by: Mat Martineau <mathew.j.martineau@linux.intel.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
This leverages the previous commit to reserve the wmem
required for the sendmsg() operation when the msk socket
lock is first acquired.
Some heuristics are used to get a reasonable [over] estimation of
the whole memory required. If we can't forward alloc such amount
fallback to a reasonable small chunk, otherwise enter the wait
for memory path.
When sendmsg() needs more memory it looks at wmem_reserved
first and if that is exhausted, move more space from
sk_forward_alloc.
The reserved memory is not persistent and is released at the
next socket unlock via the release_cb().
Overall this will simplify the next patch.
Acked-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Reviewed-by: Mat Martineau <mathew.j.martineau@linux.intel.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
This allows invoking an additional callback under the
socket spin lock.
Will be used by the next patches to avoid additional
spin lock contention.
Acked-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Reviewed-by: Mat Martineau <mathew.j.martineau@linux.intel.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Camelia Groza says:
====================
dpaa_eth: add XDP support
Enable XDP support for the QorIQ DPAA1 platforms.
Implement all the current actions (DROP, ABORTED, PASS, TX, REDIRECT). No
Tx batching is added at this time.
Additional XDP_PACKET_HEADROOM bytes are reserved in each frame's headroom.
After transmit, a reference to the xdp_frame is saved in the buffer for
clean-up on confirmation in a newly created structure for software
annotations. DPAA_TX_PRIV_DATA_SIZE bytes are reserved in the buffer for
storing this structure and the XDP program is restricted from accessing
them.
The driver shares the egress frame queues used for XDP with the network
stack. The DPAA driver is a LLTX driver so no explicit locking is required
on transmission.
====================
Link: https://lore.kernel.org/r/cover.1606322126.git.camelia.groza@nxp.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
For XDP TX, even tough we start out with correctly aligned buffers, the
XDP program might change the data's alignment. For REDIRECT, we have no
control over the alignment either.
Create a new workaround for xdp_frame structures to verify the erratum
conditions and move the data to a fresh buffer if necessary. Create a new
xdp_frame for managing the new buffer and free the old one using the XDP
API.
Due to alignment constraints, all frames have a 256 byte headroom that
is offered fully to XDP under the erratum. If the XDP program uses all
of it, the data needs to be move to make room for the xdpf backpointer.
Disable the metadata support since the information can be lost.
Acked-by: Madalin Bucur <madalin.bucur@oss.nxp.com>
Signed-off-by: Camelia Groza <camelia.groza@nxp.com>
Reviewed-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
