idpf_compl_queue uses a union for comp, comp_4b, and desc_ring. The
release path should check complq->desc_ring to determine whether the DMA
descriptor ring is allocated. The current check against comp works but is
leftover from a previous commit and is misleading in this context.
Switching the check to desc_ring improves readability and more directly
reflects the intended meaning, since desc_ring is the field representing
the allocated DMA-backed descriptor ring.
No functional change.
Signed-off-by: Alok Tiwari <alok.a.tiwari@oracle.com>
Reviewed-by: Aleksandr Loktionov <aleksandr.loktionov@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Link: https://patch.msgid.link/20251125223632.1857532-9-anthony.l.nguyen@intel.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
ixgbe_non_sfp_link_config() is called twice in ixgbe_open()
once to assign its return value to err and again in the
conditional check. This patch uses the stored err value
instead of calling the function a second time. This avoids
redundant work and ensures consistent error reporting.
Also fix a small typo in the ixgbe_remove() comment:
"The could be caused" -> "This could be caused".
Signed-off-by: Alok Tiwari <alok.a.tiwari@oracle.com>
Reviewed-by: Jedrzej Jagielski <jedrzej.jagielski@intel.com>
Reviewed-by: Paul Menzel <pmenzel@molgen.mpg.de>
Tested-by: Rinitha S <sx.rinitha@intel.com> (A Contingent worker at Intel)
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Link: https://patch.msgid.link/20251125223632.1857532-8-anthony.l.nguyen@intel.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Convert vport state to a bitmap and remove the DOWN state which is
redundant in the existing logic. There are no functional changes aside
from the use of bitwise operations when setting and checking the states.
Removed the double underscore to be consistent with the naming of other
bitmaps in the header and renamed current_state to vport_is_up to match
the meaning of the new variable.
Reviewed-by: Przemek Kitszel <przemyslaw.kitszel@intel.com>
Reviewed-by: Aleksandr Loktionov <aleksandr.loktionov@intel.com>
Reviewed-by: Chittim Madhu <madhu.chittim@intel.com>
Signed-off-by: Emil Tantilov <emil.s.tantilov@intel.com>
Tested-by: Samuel Salin <Samuel.salin@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Link: https://patch.msgid.link/20251125223632.1857532-6-anthony.l.nguyen@intel.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Flex array should be at the end of the structure and use [] syntax
Remove unused fields of ixgbevf_q_vector.
They aren't used since busy poll was moved to core code in commit
508aac6dee ("ixgbevf: get rid of custom busy polling code").
Reviewed-by: Przemek Kitszel <przemyslaw.kitszel@intel.com>
Reviewed-by: Aleksandr Loktionov <aleksandr.loktionov@intel.com>
Signed-off-by: Natalia Wochtman <natalia.wochtman@intel.com>
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Rafal Romanowski <rafal.romanowski@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Link: https://patch.msgid.link/20251125223632.1857532-5-anthony.l.nguyen@intel.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The KMSG_COMPONENT macro is a leftover of the s390 specific "kernel message
catalog" from 2008 [1] which never made it upstream.
The macro was added to s390 code to allow for an out-of-tree patch which
used this to generate unique message ids. Also this out-of-tree doesn't
exist anymore.
The pattern of how the KMSG_COMPONENT is used was partially also used for
non s390 specific code, for whatever reasons.
Remove the macro in order to get rid of a pointless indirection.
[1] https://lwn.net/Articles/292650/
Signed-off-by: Heiko Carstens <hca@linux.ibm.com>
Acked-by: Alexandra Winter <wintera@linux.ibm.com>
Link: https://patch.msgid.link/20251126142242.2124317-1-hca@linux.ibm.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Convert the Cavium Thunder NIC VF driver to use the new .get_rx_ring_count
ethtool operation instead of implementing .get_rxnfc solely for handling
ETHTOOL_GRXRINGS command. This simplifies the code by removing the
switch statement and replacing it with a direct return of the queue
count.
The new callback provides the same functionality in a more direct way,
following the ongoing ethtool API modernization.
Signed-off-by: Breno Leitao <leitao@debian.org>
Link: https://patch.msgid.link/20251126-gxring_cavium-v1-1-a066c0c9e0c6@debian.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The extra "count >= limit" check in stmmac_rx_zc() is redundant and
has no effect because the value of "count" doesn't change after the
while condition at this point.
However, it can change after "read_again:" label:
while (count < limit) {
...
if (count >= limit)
break;
read_again:
...
/* XSK pool expects RX frame 1:1 mapped to XSK buffer */
if (likely(status & rx_not_ls)) {
xsk_buff_free(buf->xdp);
buf->xdp = NULL;
dirty++;
count++;
goto read_again;
}
...
This patch addresses the same issue previously resolved in stmmac_rx()
by commit fa02de9e75 ("net: stmmac: fix rx budget limit check").
The fix is the same: move the check after the label to ensure that it
bounds the goto loop.
Fixes: bba2556efa ("net: stmmac: Enable RX via AF_XDP zero-copy")
Signed-off-by: Alexey Kodanev <aleksei.kodanev@bell-sw.com>
Reviewed-by: Russell King (Oracle) <rmk+kernel@armlinux.org.uk>
Link: https://patch.msgid.link/20251126104327.175590-1-aleksei.kodanev@bell-sw.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Extract MIB constants into the header file to improve code style. This
patch will not change the behavior of the function.
Signed-off-by: David Yang <mmyangfl@gmail.com>
Link: https://patch.msgid.link/20251126084024.2843851-3-mmyangfl@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
There are hard-to-find unused fields in the MIB table I didn't notice in
the example driver code, causing wrong interpretation of the MIB data.
For some 64-bit attributes, the current (wrong) implementation took the
correct lower 32 bits, but messed up the upper 32 bits, so it would work
accidentally until 32-bit overflows happen. Fix that too.
Fixes: 186623f4aa ("net: dsa: yt921x: Add support for Motorcomm YT921x")
Signed-off-by: David Yang <mmyangfl@gmail.com>
Link: https://patch.msgid.link/20251126084024.2843851-2-mmyangfl@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
This adds DASH support for chip RTL8127AP. Its mac version is
RTL_GIGA_MAC_VER_80. DASH is a standard for remote management of network
device, allowing out-of-band control.
Signed-off-by: Javen Xu <javen_xu@realsil.com.cn>
Link: https://patch.msgid.link/20251126055950.2050-1-javen_xu@realsil.com.cn
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Neither sock4 nor sock6 pointers are guaranteed to be non-NULL in
vxlan_xmit_one, e.g. if the iface is brought down. This can lead to the
following NULL dereference:
BUG: kernel NULL pointer dereference, address: 0000000000000010
Oops: Oops: 0000 [#1] SMP NOPTI
RIP: 0010:vxlan_xmit_one+0xbb3/0x1580
Call Trace:
vxlan_xmit+0x429/0x610
dev_hard_start_xmit+0x55/0xa0
__dev_queue_xmit+0x6d0/0x7f0
ip_finish_output2+0x24b/0x590
ip_output+0x63/0x110
Mentioned commits changed the code path in vxlan_xmit_one and as a side
effect the sock4/6 pointer validity checks in vxlan(6)_get_route were
lost. Fix this by adding back checks.
Since both commits being fixed were released in the same version (v6.7)
and are strongly related, bundle the fixes in a single commit.
Reported-by: Liang Li <liali@redhat.com>
Fixes: 6f19b2c136 ("vxlan: use generic function for tunnel IPv4 route lookup")
Fixes: 2aceb896ee ("vxlan: use generic function for tunnel IPv6 route lookup")
Cc: Beniamino Galvani <b.galvani@gmail.com>
Signed-off-by: Antoine Tenart <atenart@kernel.org>
Reviewed-by: Ido Schimmel <idosch@nvidia.com>
Tested-by: Ido Schimmel <idosch@nvidia.com>
Link: https://patch.msgid.link/20251126102627.74223-1-atenart@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
ptp_clock_settime() assumes every ptp_clock has implemented settime64().
Stub it with -EOPNOTSUPP to prevent a NULL dereference.
The fix is similar to commit 329d050bbe ("gve: Implement settime64
with -EOPNOTSUPP").
Fixes: d734223b2f ("iavf: add initial framework for registering PTP clock")
Signed-off-by: Michal Schmidt <mschmidt@redhat.com>
Reviewed-by: Aleksandr Loktionov <aleksandr.loktionov@intel.com>
Reviewed-by: Tim Hostetler <thostet@google.com>
Link: https://patch.msgid.link/20251126094850.2842557-1-mschmidt@redhat.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
This 0x88C3 is registered to Infineon Technologies Corporate Research ST
and are used by MaxLinear.
Infineon made a spin off called Lantiq.
Lantiq was acquired by Intel
MaxLinear acquired Intels Connected Home division.
The product FAQ from MaxLinear describes it's history from the F24S.
The driver for the gsw1xx is based on Lantiq showing it's similarities.
Ref https://standards-oui.ieee.org/ethertype/eth.txt
Signed-off-by: Peter Enderborg <Peter.Enderborg@axis.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Use DEFINE_RAW_FLEX() to avoid a -Wflex-array-member-not-at-end warning.
Remove fixed-size array struct usb_cdc_ncm_dpe16 dpe16[2]; from struct
mbim_tx_hdr, so that flex-array member struct mbim_tx_hdr::ndp16.dpe16[]
ends last in this structure.
Compensate for this by using the DEFINE_RAW_FLEX() helper to declare the
on-stack struct instance that contains struct usb_cdc_ncm_ndp16 as a
member. Adjust the rest of the code, accordingly.
So, with these changes fix the following warning:
drivers/net/wwan/mhi_wwan_mbim.c:81:34: warning: structure containing a flexible array member is not at the end of another structure [-Wflex-array-member-not-at-end]
Reviewed-by: Loic Poulain <loic.poulain@oss.qualcomm.com>
Signed-off-by: Gustavo A. R. Silva <gustavoars@kernel.org>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The tx->dropped_pkt counter is a 64-bit integer that is incremented
directly. On 32-bit architectures, this operation is not atomic and
can lead to read/write tearing if a reader accesses the counter during
the update. This can result in incorrect values being reported for
dropped packets.
To prevent this potential data corruption, wrap the increment
operation with u64_stats_update_begin() and u64_stats_update_end().
This ensures that updates to the 64-bit counter are atomic, even on
32-bit systems, by using a sequence lock.
The u64_stats_sync API requires the writer to have exclusive access,
which is already provided in this context by the network stack's
serialization of the transmit path (net_device_ops::ndo_start_xmit
[1]) for a given queue.
[1]: https://www.kernel.org/doc/Documentation/networking/netdevices.txt
Signed-off-by: Max Yuan <maxyuan@google.com>
Reviewed-by: Jordan Rhee <jordanrhee@google.com>
Signed-off-by: Harshitha Ramamurthy <hramamurthy@google.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Commit e20dfbad8a ("net: fix napi_consume_skb() with alien skbs")
added a skb->cpu check to napi_consume_skb(), before the point where
napi_consume_skb() validated skb is not NULL.
Add an explicit check to the early exit condition.
Reviewed-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
As those following recent changes from Eric know very well
using NAPI skb cache is crucial to achieve good perf, at
least on recent AMD platforms. Make sure bnxt feeds the skb
cache with Tx skbs.
Reviewed-by: Michael Chan <michael.chan@broadcom.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Convert all the legacy code directly accessing the pp fields in net_iov
to access them through @desc in net_iov.
Signed-off-by: Byungchul Park <byungchul@sk.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
- two last minute fixes for the recently modified DMA API infrastructure:
a proper handling of DMA_ATTR_MMIO in dma_iova_unlink() function (me) and
a regression fix for the code refactoring related to P2PDMA (Pranjal
Shrivastava)
-----BEGIN PGP SIGNATURE-----
iHUEABYIAB0WIQSrngzkoBtlA8uaaJ+Jp1EFxbsSRAUCaSh6UAAKCRCJp1EFxbsS
RCYGAP94GdoTEBmQL1dXLooASbHHAyB5xrd8Di1H/Aqd7maCvAEAgT2+BRPyEtf+
wiz4mwKMG1RrF9Cj4XqPqJ2NDU1zdQ8=
=ngH/
-----END PGP SIGNATURE-----
Merge tag 'dma-mapping-6.18-2025-11-27' of git://git.kernel.org/pub/scm/linux/kernel/git/mszyprowski/linux
Pull dma-mapping fixes from Marek Szyprowski:
"Two last minute fixes for the recently modified DMA API infrastructure:
- proper handling of DMA_ATTR_MMIO in dma_iova_unlink() function (me)
- regression fix for the code refactoring related to P2PDMA (Pranjal
Shrivastava)"
* tag 'dma-mapping-6.18-2025-11-27' of git://git.kernel.org/pub/scm/linux/kernel/git/mszyprowski/linux:
dma-direct: Fix missing sg_dma_len assignment in P2PDMA bus mappings
iommu/dma: add missing support for DMA_ATTR_MMIO for dma_iova_unlink()
There is one more commit that needs to be reverted after reverting
problematic commit 7a8c994cbb ("ACPI: processor: idle: Optimize
ACPI idle driver registration"), so revert it.
-----BEGIN PGP SIGNATURE-----
iQFGBAABCAAwFiEEcM8Aw/RY0dgsiRUR7l+9nS/U47UFAmkorhYSHHJqd0Byand5
c29ja2kubmV0AAoJEO5fvZ0v1OO1HmQH/1mAIwpYlmtXtW8LmGuZsrvZx4PTl8OQ
QyATMm/0+/gSWXYs79vLikwoRfT0rSEOdZcNk6AwdqeP2ciHLwCAzNnOE7vw80/s
xo7yj1SswzEZDydEhbxqRFNsJ8hl5VCU7DdsScVtuDFb/vJ0GHkQEu1DsSFJQoSL
CidC/2RCn0n5egOMdxL/GHemEbdUDGDXr2JilrcNLmcnnArN43+JqdiV8p102oSz
i8ssS/LIMUMrIMZCuJA1298pZ6Ir1WWgaodyTOYDcmq3u25SSabF5WlS2u0KGgm/
Ldgl6/lCuMxt+BBJ1r3uq9uHRAWeqY+DpHcZ7q1yFB+rOyRA9za5ERc=
=X5FV
-----END PGP SIGNATURE-----
Merge tag 'acpi-6.18-rc8-2' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm
Pull ACPI fix from Rafael Wysocki:
"One more urgent ACPI support fix for 6.18
There is one more commit that needs to be reverted after reverting
problematic commit 7a8c994cbb ("ACPI: processor: idle: Optimize ACPI
idle driver registration"), so revert it"
* tag 'acpi-6.18-rc8-2' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm:
Revert "ACPI: processor: Update cpuidle driver check in __acpi_processor_start()"
The trace_marker_raw file in tracefs takes a buffer from user space that
contains an id as well as a raw data string which is usually a binary
structure. The structure used has the following:
struct raw_data_entry {
struct trace_entry ent;
unsigned int id;
char buf[];
};
Since the passed in "cnt" variable is both the size of buf as well as the
size of id, the code to allocate the location on the ring buffer had:
size = struct_size(entry, buf, cnt - sizeof(entry->id));
Which is quite ugly and hard to understand. Instead, add a helper macro
called struct_offset() which then changes the above to a simple and easy
to understand:
size = struct_offset(entry, id) + cnt;
This will likely come in handy for other use cases too.
Link: https://lore.kernel.org/all/CAHk-=whYZVoEdfO1PmtbirPdBMTV9Nxt9f09CK0k6S+HJD3Zmg@mail.gmail.com/
Cc: Masami Hiramatsu <mhiramat@kernel.org>
Cc: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
Cc: "Gustavo A. R. Silva" <gustavoars@kernel.org>
Link: https://patch.msgid.link/20251126145249.05b1770a@gandalf.local.home
Suggested-by: Linus Torvalds <torvalds@linux-foundation.org>
Reviewed-by: Kees Cook <kees@kernel.org>
Signed-off-by: Steven Rostedt (Google) <rostedt@goodmis.org>
In include/uapi/linux/netfilter/nf_tables.h,
correct the kernel-doc comments for mistyped enum names and enum values to
avoid these kernel-doc warnings and improve the documentation:
nf_tables.h:896: warning: Enum value 'NFT_EXTHDR_OP_TCPOPT' not described
in enum 'nft_exthdr_op'
nf_tables.h:896: warning: Excess enum value 'NFT_EXTHDR_OP_TCP' description
in 'nft_exthdr_op'
nf_tables.h:1210: warning: expecting prototype for enum
nft_flow_attributes. Prototype was for enum nft_offload_attributes instead
nf_tables.h:1428: warning: expecting prototype for enum nft_reject_code.
Prototype was for enum nft_reject_inet_code instead
(add beginning '@' to each enum value description:)
nf_tables.h:1493: warning: Enum value 'NFTA_TPROXY_FAMILY' not described
in enum 'nft_tproxy_attributes'
nf_tables.h:1493: warning: Enum value 'NFTA_TPROXY_REG_ADDR' not described
in enum 'nft_tproxy_attributes'
nf_tables.h:1493: warning: Enum value 'NFTA_TPROXY_REG_PORT' not described
in enum 'nft_tproxy_attributes'
nf_tables.h:1796: warning: expecting prototype for enum
nft_device_attributes. Prototype was for enum
nft_devices_attributes instead
Signed-off-by: Randy Dunlap <rdunlap@infradead.org>
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Fix the kernel-doc format for struct members to be "@member" instead of
"@ member" to avoid kernel-doc warnings.
Warning: ip6t_srh.h:60 struct member 'next_hdr' not described in 'ip6t_srh'
Warning: ip6t_srh.h:60 struct member 'hdr_len' not described in 'ip6t_srh'
Warning: ip6t_srh.h:60 struct member 'segs_left' not described
in 'ip6t_srh'
Warning: ip6t_srh.h:60 struct member 'last_entry' not described
in 'ip6t_srh'
Warning: ip6t_srh.h:60 struct member 'tag' not described in 'ip6t_srh'
Warning: ip6t_srh.h:60 struct member 'mt_flags' not described in 'ip6t_srh'
Warning: ip6t_srh.h:60 struct member 'mt_invflags' not described
in 'ip6t_srh'
Warning: ip6t_srh.h:93 struct member 'next_hdr' not described
in 'ip6t_srh1'
Warning: ip6t_srh.h:93 struct member 'hdr_len' not described in 'ip6t_srh1'
Warning: ip6t_srh.h:93 struct member 'segs_left' not described
in 'ip6t_srh1'
Warning: ip6t_srh.h:93 struct member 'last_entry' not described
in 'ip6t_srh1'
Warning: ip6t_srh.h:93 struct member 'tag' not described in 'ip6t_srh1'
Warning: ip6t_srh.h:93 struct member 'psid_addr' not described
in 'ip6t_srh1'
Warning: ip6t_srh.h:93 struct member 'nsid_addr' not described
in 'ip6t_srh1'
Warning: ip6t_srh.h:93 struct member 'lsid_addr' not described
in 'ip6t_srh1'
Warning: ip6t_srh.h:93 struct member 'psid_msk' not described
in 'ip6t_srh1'
Warning: ip6t_srh.h:93 struct member 'nsid_msk' not described
in 'ip6t_srh1'
Warning: ip6t_srh.h:93 struct member 'lsid_msk' not described
in 'ip6t_srh1'
Warning: ip6t_srh.h:93 struct member 'mt_flags' not described
in 'ip6t_srh1'
Warning: ip6t_srh.h:93 struct member 'mt_invflags' not described
in 'ip6t_srh1'
Signed-off-by: Randy Dunlap <rdunlap@infradead.org>
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Introduce the capability to send TCP traffic over IPv6 to
nft_flowtable netfilter selftest.
Reviewed-by: Simon Horman <horms@kernel.org>
Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
This is useful to update the limit or flags without clearing the
connections tracked. Use READ_ONCE() on packetpath as it can be modified
on controlplane.
Signed-off-by: Fernando Fernandez Mancera <fmancera@suse.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Connlimit expression can be used for all kind of packets and not only
for packets with connection state new. See this ruleset as example:
table ip filter {
chain input {
type filter hook input priority filter; policy accept;
tcp dport 22 ct count over 4 counter
}
}
Currently, if the connection count goes over the limit the counter will
count the packets. When a connection is closed, the connection count
won't decrement as it should because it is only updated for new
connections due to an optimization on __nf_conncount_add() that prevents
updating the list if the connection is duplicated.
To solve this problem, check whether the connection was skipped and if
so, update the list. Adjust count_tree() too so the same fix is applied
for xt_connlimit.
Fixes: 976afca1ce ("netfilter: nf_conncount: Early exit in nf_conncount_lookup() and cleanup")
Closes: https://lore.kernel.org/netfilter/trinity-85c72a88-d762-46c3-be97-36f10e5d9796-1761173693813@3c-app-mailcom-bs12/
Signed-off-by: Fernando Fernandez Mancera <fmancera@suse.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
For convenience when performing GC over the connection list, make
nf_conncount_gc_list() to disable BH. This unifies the behavior with
nf_conncount_add() and nf_conncount_count().
Signed-off-by: Fernando Fernandez Mancera <fmancera@suse.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
When using nf_conncount infrastructure for non-confirmed connections a
duplicated track is possible due to an optimization introduced since
commit d265929930 ("netfilter: nf_conncount: reduce unnecessary GC").
In order to fix this introduce a new conncount API that receives
directly an sk_buff struct. It fetches the tuple and zone and the
corresponding ct from it. It comes with both existing conncount variants
nf_conncount_count_skb() and nf_conncount_add_skb(). In addition remove
the old API and adjust all the users to use the new one.
This way, for each sk_buff struct it is possible to check if there is a
ct present and already confirmed. If so, skip the add operation.
Fixes: d265929930 ("netfilter: nf_conncount: reduce unnecessary GC")
Signed-off-by: Fernando Fernandez Mancera <fmancera@suse.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Introduce specific selftest for IPIP flowtable SW acceleration in
nft_flowtable.sh
Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Introduce sw acceleration for tx path of IPIP tunnels relying on the
netfilter flowtable infrastructure.
This patch introduces basic infrastructure to accelerate other tunnel
types (e.g. IP6IP6).
IPIP sw tx acceleration can be tested running the following scenario where
the traffic is forwarded between two NICs (eth0 and eth1) and an IPIP
tunnel is used to access a remote site (using eth1 as the underlay device):
ETH0 -- TUN0 <==> ETH1 -- [IP network] -- TUN1 (192.168.100.2)
$ip addr show
6: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP group default qlen 1000
link/ether 00:00:22:33:11:55 brd ff:ff:ff:ff:ff:ff
inet 192.168.0.2/24 scope global eth0
valid_lft forever preferred_lft forever
7: eth1: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP group default qlen 1000
link/ether 00:11:22:33:11:55 brd ff:ff:ff:ff:ff:ff
inet 192.168.1.1/24 scope global eth1
valid_lft forever preferred_lft forever
8: tun0@NONE: <POINTOPOINT,NOARP,UP,LOWER_UP> mtu 1480 qdisc noqueue state UNKNOWN group default qlen 1000
link/ipip 192.168.1.1 peer 192.168.1.2
inet 192.168.100.1/24 scope global tun0
valid_lft forever preferred_lft forever
$ip route show
default via 192.168.100.2 dev tun0
192.168.0.0/24 dev eth0 proto kernel scope link src 192.168.0.2
192.168.1.0/24 dev eth1 proto kernel scope link src 192.168.1.1
192.168.100.0/24 dev tun0 proto kernel scope link src 192.168.100.1
$nft list ruleset
table inet filter {
flowtable ft {
hook ingress priority filter
devices = { eth0, eth1 }
}
chain forward {
type filter hook forward priority filter; policy accept;
meta l4proto { tcp, udp } flow add @ft
}
}
Reproducing the scenario described above using veths I got the following
results:
- TCP stream trasmitted into the IPIP tunnel:
- net-next: (baseline) ~ 85Gbps
- net-next + IPIP flowtable support: ~102Gbps
Co-developed-by: Pablo Neira Ayuso <pablo@netfilter.org>
Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Introduce sw acceleration for rx path of IPIP tunnels relying on the
netfilter flowtable infrastructure. Subsequent patches will add sw
acceleration for IPIP tunnels tx path.
This series introduces basic infrastructure to accelerate other tunnel
types (e.g. IP6IP6).
IPIP rx sw acceleration can be tested running the following scenario where
the traffic is forwarded between two NICs (eth0 and eth1) and an IPIP
tunnel is used to access a remote site (using eth1 as the underlay device):
ETH0 -- TUN0 <==> ETH1 -- [IP network] -- TUN1 (192.168.100.2)
$ip addr show
6: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP group default qlen 1000
link/ether 00:00:22:33:11:55 brd ff:ff:ff:ff:ff:ff
inet 192.168.0.2/24 scope global eth0
valid_lft forever preferred_lft forever
7: eth1: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP group default qlen 1000
link/ether 00:11:22:33:11:55 brd ff:ff:ff:ff:ff:ff
inet 192.168.1.1/24 scope global eth1
valid_lft forever preferred_lft forever
8: tun0@NONE: <POINTOPOINT,NOARP,UP,LOWER_UP> mtu 1480 qdisc noqueue state UNKNOWN group default qlen 1000
link/ipip 192.168.1.1 peer 192.168.1.2
inet 192.168.100.1/24 scope global tun0
valid_lft forever preferred_lft forever
$ip route show
default via 192.168.100.2 dev tun0
192.168.0.0/24 dev eth0 proto kernel scope link src 192.168.0.2
192.168.1.0/24 dev eth1 proto kernel scope link src 192.168.1.1
192.168.100.0/24 dev tun0 proto kernel scope link src 192.168.100.1
$nft list ruleset
table inet filter {
flowtable ft {
hook ingress priority filter
devices = { eth0, eth1 }
}
chain forward {
type filter hook forward priority filter; policy accept;
meta l4proto { tcp, udp } flow add @ft
}
}
Reproducing the scenario described above using veths I got the following
results:
- TCP stream received from the IPIP tunnel:
- net-next: (baseline) ~ 71Gbps
- net-next + IPIP flowtbale support: ~101Gbps
Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
This simplifies IPIP tunnel support coming in follow up patches.
No function changes are intended.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
hw_ifidx was originally introduced to store the real netdevice as a
requirement for the hardware offload support in:
73f97025a9 ("netfilter: nft_flow_offload: use direct xmit if hardware offload is enabled")
Since ("netfilter: flowtable: consolidate xmit path"), ifidx and
hw_ifidx points to the real device in the xmit path, remove it.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Push the pppoe header from the flowtable xmit path, inlining is faster
than the original xmit path because it can avoid some locking.
This is based on a patch originally written by wenxu.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Push the vlan header from the flowtable xmit path, instead of passing
the packet to the vlan device.
This is based on a patch originally written by wenxu.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Use dev_queue_xmit() for the XMIT_NEIGH case. Store the interface index
of the real device behind the vlan/pppoe device, this introduces an
extra lookup for the real device in the xmit path because rt->dst.dev
provides the vlan/pppoe device.
XMIT_NEIGH now looks more similar to XMIT_DIRECT but the check for stale
dst and the neighbour lookup still remain in place which is convenient
to deal with network topology changes.
Note that nft_flow_route() needs to relax the check for _XMIT_NEIGH so
the existing basic xfrm offload (which only works in one direction) does
not break.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
This file contains the path discovery that is run from the forward chain
for the packet offloading the flow into the flowtable. This consists
of a series of calls to dev_fill_forward_path() for each device stack.
More topologies may be supported in the future, so move this code to its
own file to separate it from the nftables flow_offload expression.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Add a sanity check to skip path discovery if the maximum number of
encapsulation is reached. While at it, check for underflow too.
Fixes: 26267bf9bb ("netfilter: flowtable: bridge vlan hardware offload and switchdev")
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
- Restrict the pointer size of flush pages to prevent a regression.
-----BEGIN PGP SIGNATURE-----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=9d7X
-----END PGP SIGNATURE-----
Merge tag 'drm-misc-next-fixes-2025-11-26' of https://gitlab.freedesktop.org/drm/misc/kernel into drm-next
drm-misc-next-fixes for v6.19:
- Restrict the pointer size of flush pages to prevent a regression.
Signed-off-by: Dave Airlie <airlied@redhat.com>
From: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Link: https://patch.msgid.link/0090a4fc-9cc4-4c03-bfe5-d1b1f0cc7e05@linux.intel.com
This reverts commit 5a9b0c7418.
The switch from AUX interrupts to pollign was very hand-wavy.
Yes, there have been some situations in CI on a few platforms
where the AUX hardware seemingly forgets to signal the timeout,
but those have been happening after we switched to polling as
well. So I don't think we have any conclusive evidence that
polling actually helps here.
Someone really should root cause the actual problem, and see
if there is a proper workaround we could implemnt (eg. disabling
clock gating/etc.). In the meantime just go back to using the
interrupt for AUX completion.
If the hardware fails to signal the timeout we will just hit
the wait_event_timeout() software timeout instead. I suppose
we could try to tune the software timeout to more closely
match the expected hardware timeout. Might need to use
wait_event_hrtimeout() or something to avoid jiffies
granularity issues...
The AUX polling is also a hinderance towards using poll_timeout_us()
because we have a very long timeout, but would need a fairly short
polling interval to keep AUX transfer reasonably fast. Someone would
need to come up with good numbers in a somewhat scientific way.
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patch.msgid.link/20251119185310.10428-3-ville.syrjala@linux.intel.com
Acked-by: Jani Nikula <jani.nikula@intel.com>
Apparently the DDI A/B AUX interrupts move onto the PICA side
on LNL. Unmask them properly so that we actually get the
interrupts. The interrupt handler was already trying to handle
them despite the interrupts remaining masked.
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patch.msgid.link/20251119185310.10428-2-ville.syrjala@linux.intel.com
Reviewed-by: Jani Nikula <jani.nikula@intel.com>
Another (somewhat expensive) drm_format_info() call has
appeared in intel_plane_can_async_flip(). That one may get
called several times per commit so we need to get rid of
it.
Fortunately most callers already have the framebuffer at
hand, so we can just grab the format info from there.
The one exception is intel_plane_format_mod_supported_async()
where we have to do the lookup. But that only gets called
(a bunch of times) during driver init to build the
IN_FORMATS_ASYNC blob, and afterwards there is no runtime
cost.
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patch.msgid.link/20251112233030.24117-4-ville.syrjala@linux.intel.com
Reviewed-by: Jouni Högander <jouni.hogander@intel.com>
Since old kernel versions wouldn't expose the IN_FORMATS_ASYNC blob,
userspace can't really use the absence of the blob to determine
that async flips aren't supported. Thus it seems better to always
expose the blob on all planes, whether they support async flips
or not. The blob will simply not indicate any format+modifier
combinations as supported on planes that aren't async flip capable.
Currently we expose the blob for all skl+ universal planes (even
though we implement async flips only for the first plane on each
pipe), and i9xx primary planes (for ilk+ we have async flips support,
for pre-ilk we do not). Complete the full set by also expsosing
the blob on pre-skl sprite planes, and cursors.
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patch.msgid.link/20251112233030.24117-3-ville.syrjala@linux.intel.com
Reviewed-by: Jouni Högander <jouni.hogander@intel.com>
Use the standard variable names for things, and get rid of any
annoying aliasing variables. And sprinkle the consts in while at
it.
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patch.msgid.link/20251119181606.17129-7-ville.syrjala@linux.intel.com
Reviewed-by: Jani Nikula <jani.nikula@intel.com>
