The Rx rings are filled with Rx buffers. Which are supposed to fit
packet headers (or MTU if HW-GRO is disabled). The aggregation buffers
are filled with "device pages". Adjust the sizes of the page pool
recycling ring appropriately, based on ratio of the size of the
buffer on given ring vs system page size. Otherwise on a system
with 64kB pages we end up with >700MB of memory sitting in every
single page pool cache.
Correct the size calculation for the head_pool. Since the buffers
there are always small I'm pretty sure I meant to cap the size
at 1k, rather than make it the lowest possible size. With 64k pages
1k cache with a 1k ring is 64x larger than we need.
Reviewed-by: Michael Chan <michael.chan@broadcom.com>
Link: https://patch.msgid.link/20250626165441.4125047-1-kuba@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Test DSACK behavior with non contiguous ranges.
Without prior fix (tcp: fix tcp_ofo_queue() to avoid including
too much DUP SACK range) this would fail with:
tcp_dsack_mult.pkt:37: error handling packet: bad value outbound TCP option 5
script packet: 0.100682 . 1:1(0) ack 6001 <nop,nop,sack 1001:3001 7001:8001>
actual packet: 0.100679 . 1:1(0) ack 6001 win 1097 <nop,nop,sack 1001:6001 7001:8001>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: xin.guo <guoxin0309@gmail.com>
Link: https://patch.msgid.link/20250626123420.1933835-3-edumazet@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
If the new coming segment covers more than one skbs in the ofo queue,
and which seq is equal to rcv_nxt, then the sequence range
that is duplicated will be sent as DUP SACK, the detail as below,
in step6, the {501,2001} range is clearly including too much
DUP SACK range, in violation of RFC 2883 rules.
1. client > server: Flags [.], seq 501:1001, ack 1325288529, win 20000, length 500
2. server > client: Flags [.], ack 1, [nop,nop,sack 1 {501:1001}], length 0
3. client > server: Flags [.], seq 1501:2001, ack 1325288529, win 20000, length 500
4. server > client: Flags [.], ack 1, [nop,nop,sack 2 {1501:2001} {501:1001}], length 0
5. client > server: Flags [.], seq 1:2001, ack 1325288529, win 20000, length 2000
6. server > client: Flags [.], ack 2001, [nop,nop,sack 1 {501:2001}], length 0
After this fix, the final ACK is as below:
6. server > client: Flags [.], ack 2001, options [nop,nop,sack 1 {501:1001}], length 0
[edumazet] added a new packetdrill test in the following patch.
Fixes: 1da177e4c3 ("Linux-2.6.12-rc2")
Signed-off-by: xin.guo <guoxin0309@gmail.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Link: https://patch.msgid.link/20250626123420.1933835-2-edumazet@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Linus Walleij says:
====================
net: dsa: ks8995: Fix up bindings
After looking at the datasheets for KS8995 I realized this is
a DSA switch and need to have DT bindings as such and be implemented
as such.
This series just fixes up the bindings and the offending device tree.
The existing kernel driver which is in drivers/net/phy/spi_ks8995.c
does not implement DSA. It can be forgiven for this because it was
merged in 2011 and the DSA framework was not widely established
back then. It continues to probe fine but needs to be rewritten
to use the special DSA tag and moved to drivers/net/dsa as time
permits. (I hope I can do this.)
It's fine for the networking tree to merge both patches, I maintain
ixp4xx as well. But I can also carry the second patch through the
SoC tree if so desired.
v1: https://lore.kernel.org/20250624-ks8995-dsa-bindings-v1-0-71a8b4f63315@linaro.org
====================
Link: https://patch.msgid.link/20250625-ks8995-dsa-bindings-v2-0-ce71dce9be0b@linaro.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Fix up the KS8995 switch and PHYs the way that is most likely:
- Phy 1-4 is certainly the PHYs of the KS8995 (mask 0x1e in
the outoftree code masks PHYs 1,2,3,4).
- Phy 5 is the MII-P5 separate WAN phy of the KS8995 directly
connected to EthC.
- The EthB MII is probably connected as CPU interface to the
KS8995.
Properly integrate the KS8995 switch using the new bindings.
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Link: https://patch.msgid.link/20250625-ks8995-dsa-bindings-v2-2-ce71dce9be0b@linaro.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
After studying the datasheets for some of the KS8995 variants
it becomes pretty obvious that this is a straight-forward
and simple MII DSA switch with one port in (CPU) and four outgoing
ports, and it even supports custom tags by setting a bit in
a special register, and elaborate VLAN handling as all DSA
switches do.
What is a bit odd with KS8995 is that it uses an extra MII-P5
port to access one of the PHYs separately, on the side of the
switch fabric, such as when using a WAN port separately from
a LAN switch in a home router.
Rewrite the terse bindings to YAML, and move to the proper
subdirectory. Include a verbose example to make things clear.
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Reviewed-by: Rob Herring (Arm) <robh@kernel.org>
Link: https://patch.msgid.link/20250625-ks8995-dsa-bindings-v2-1-ce71dce9be0b@linaro.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Paul Geurts says:
====================
NFC: trf7970a: Add option to reduce antenna gain
The TRF7970a device is sensitive to RF disturbances, which can make it
hard to pass some EMC immunity tests. By reducing the RX antenna gain,
the device becomes less sensitive to EMC disturbances, as a trade-off
against antenna performance.
====================
Link: https://patch.msgid.link/20250626141242.3749958-1-paul.geurts@prodrive-technologies.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The TRF7970a device is sensitive to RF disturbances, which can make it
hard to pass some EMC immunity tests. By reducing the RX antenna gain,
the device becomes less sensitive to EMC disturbances, as a trade-off
against antenna performance.
Add a device tree option to select RX gain reduction to improve EMC
performance.
Selecting a communication standard in the ISO control register resets
the RX antenna gain settings. Therefore set the RX gain reduction
everytime the ISO control register changes, when the option is used.
Signed-off-by: Paul Geurts <paul.geurts@prodrive-technologies.com>
Reviewed-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Link: https://patch.msgid.link/20250626141242.3749958-3-paul.geurts@prodrive-technologies.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Merge a fix from Jeff from a stable commit ID:
* ref_tracker: do xarray and workqueue job initializations earlier
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Airoha AN7583 SoC have 2 dedicated MDIO bus controller in the SCU
register map. To driver register an MDIO controller based on the DT
reg property and access the register by accessing the parent syscon.
The MDIO bus logic is similar to the MT7530 internal MDIO bus but
deviates of some setting and some HW bug.
On Airoha AN7583 the MDIO clock is set to 25MHz by default and needs to
be correctly setup to 2.5MHz to correctly work (by setting the divisor
to 10x).
There seems to be Hardware bug where AN7583_MII_RWDATA
is not wiped in the context of unconnected PHY and the
previous read value is returned.
Example: (only one PHY on the BUS at 0x1f)
- read at 0x1f report at 0x2 0x7500
- read at 0x0 report 0x7500 on every address
To workaround this, we reset the Mdio BUS at every read
to have consistent values on read operation.
Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
Airoha AN7583 SoC have 3 different MDIO Controller. One comes from
the intergated Switch based on MT7530. The other 2 live under the SCU
register and expose 2 dedicated MDIO controller.
Document the schema for the 2 dedicated MDIO controller.
Each MDIO controller can be independently reset with the SoC reset line.
Each MDIO controller have a dedicated clock configured to 2.5MHz by
default to follow MDIO bus IEEE 802.3 standard.
Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Thomas Gleixner says:
====================
ptp: Belated spring cleaning of the chardev driver
When looking into supporting auxiliary clocks in the PTP ioctl, the
inpenetrable ptp_ioctl() letter soup bothered me enough to clean it up.
The code (~400 lines!) is really hard to follow due to a gazillion of
local variables, which are only used in certain case scopes, and a
mixture of gotos, breaks and direct error return paths.
Clean it up by splitting out the IOCTL functionality into seperate
functions, which contain only the required local variables and are trivial
to follow. Complete the cleanup by converting the code to lock guards and
get rid of all gotos.
That reduces the code size by 48 lines and also the binary text size is
80 bytes smaller than the current maze.
The series is split up into one patch per IOCTL command group for easy
review.
v1: https://lore.kernel.org/20250620130144.351492917@linutronix.de
====================
Link: https://patch.msgid.link/20250625114404.102196103@linutronix.de
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The mixture of gotos and direct return codes is inconsistent and just makes
the code harder to read. Let it consistently return error codes directly and
tidy the code flow up accordingly.
No functional change intended.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://patch.msgid.link/20250625115133.486953538@linutronix.de
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Convert the various spin_lock_irqsave() protected critical regions to
scoped guards. Use spinlock_irq instead of spinlock_irqsave as all the
functions are invoked in thread context with interrupts enabled.
No functional change intended.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://patch.msgid.link/20250625115133.425029269@linutronix.de
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
ptp_ioctl() is an inpenetrable letter soup with a gazillion of case (scope)
specific variables defined at the top of the function and pointless breaks
and gotos.
Start cleaning it up by splitting out the PTP_CLOCK_GETCAPS ioctl code into
a helper function. Use a argument pointer with a single sparse compliant
type cast instead of proliferating the type cast all over the place.
No functional change intended.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Vadim Fedorenko <vadim.fedorenko@linux.dev>
Link: https://patch.msgid.link/20250625115132.733409073@linutronix.de
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
setup_wait() takes an optional argument and then is called from the top
level of the test script. That confuses shellcheck, which thinks that maybe
the intention is to pass $1 of the script to the function, which is never
the case. To avoid having to annotate every single new test with a SC
disable, split the function in two: one that takes a mandatory argument,
and one that takes no argument at all.
Convert the two existing users of that optional argument, both in Spectrum
resource selftest, to use the new form. Clean up vxlan_bridge_1q_mc_ul.sh
to not pass a now-unused argument.
Signed-off-by: Petr Machata <petrm@nvidia.com>
Link: https://patch.msgid.link/8e13123236fe3912ae29bc04a1528bdd8551da1f.1750847794.git.petrm@nvidia.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Daniel Borkmann says:
====================
pull-request: bpf-next 2025-06-27
We've added 6 non-merge commits during the last 8 day(s) which contain
a total of 6 files changed, 120 insertions(+), 20 deletions(-).
The main changes are:
1) Fix RCU usage in task_cls_state() for BPF programs using helpers like
bpf_get_cgroup_classid_curr() outside of networking, from Charalampos
Mitrodimas.
2) Fix a sockmap race between map_update and a pending workqueue from
an earlier map_delete freeing the old psock where both pointed to the
same psock->sk, from Jiayuan Chen.
3) Fix a data corruption issue when using bpf_msg_pop_data() in kTLS which
failed to recalculate the ciphertext length, also from Jiayuan Chen.
4) Remove xdp_redirect_map{,_err} trace events since they are unused and
also hide XDP trace events under CONFIG_BPF_SYSCALL, from Steven Rostedt.
* tag 'for-netdev' of https://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf-next:
xdp: tracing: Hide some xdp events under CONFIG_BPF_SYSCALL
xdp: Remove unused events xdp_redirect_map and xdp_redirect_map_err
net, bpf: Fix RCU usage in task_cls_state() for BPF programs
selftests/bpf: Add test to cover ktls with bpf_msg_pop_data
bpf, ktls: Fix data corruption when using bpf_msg_pop_data() in ktls
bpf, sockmap: Fix psock incorrectly pointing to sk
====================
Link: https://patch.msgid.link/20250626230111.24772-1-daniel@iogearbox.net
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
fbnic_fw_clear_cmpl() does the inverse of fbnic_mbx_set_cmpl().
It removes the completion from the mailbox table.
It also calls fbnic_mbx_set_cmpl_slot() internally.
It should have fbnic_mbx prefix, not fbnic_fw.
I'm not very clear on what the distinction is between the two
prefixes but the matching "set" and "clear" functions should
use the same prefix.
While at it move the "clear" function closer to the "set".
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Reviewed-by: Simon Horman <horms@kernel.org>
Link: https://patch.msgid.link/20250624142834.3275164-6-kuba@kernel.org
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
