- 26 Oct, 2023 2 commits
-
-
Yan Zhai authored
Separate GSO and non-GSO packets handling to make the logic cleaner. For GSO packets, frag_max_size check can be omitted because it is only useful for packets defragmented by netfilter hooks. Both local output and GRO logic won't produce GSO packets when defragment is needed. This also mirrors what IPv4 side code is doing. Suggested-by:
Florian Westphal <fw@strlen.de> Signed-off-by:
Yan Zhai <yan@cloudflare.com> Reviewed-by:
Willem de Bruijn <willemb@google.com> Reviewed-by:
Eric Dumazet <edumazet@google.com> Link: https://lore.kernel.org/r/0e1d4599f858e2becff5c4fe0b5f843236bc3fe8.1698156966.git.yan@cloudflare.comSigned-off-by:
Jakub Kicinski <kuba@kernel.org>
-
Yan Zhai authored
RTAX_FEATURE_ALLFRAG was added before the first git commit: https://www.mail-archive.com/bk-commits-head@vger.kernel.org/msg03399.html The feature would send packets to the fragmentation path if a box receives a PMTU value with less than 1280 byte. However, since commit 9d289715 ("ipv6: stop sending PTB packets for MTU < 1280"), such message would be simply discarded. The feature flag is neither supported in iproute2 utility. In theory one can still manipulate it with direct netlink message, but it is not ideal because it was based on obsoleted guidance of RFC-2460 (replaced by RFC-8200). The feature would always test false at the moment, so remove related code or mark them as unused. Signed-off-by:
-
- 25 Oct, 2023 31 commits
-
-
Jakub Kicinski authored
Mat Martineau says: ==================== mptcp: Features and fixes for v6.7 Patch 1 adds a configurable timeout for the MPTCP connection when all subflows are closed, to support break-before-make use cases. Patch 2 is a fix for a 1-byte error in rx data counters with MPTCP fastopen connections. Patch 3 is a minor code cleanup. Patches 4 & 5 add handling of rcvlowat for MPTCP sockets, with a prerequisite patch to use a common scaling ratio between TCP and MPTCP. Patch 6 improves efficiency of memory copying in MPTCP transmit code. Patch 7 refactors syncing of socket options from the MPTCP socket to its subflows. Patches 8 & 9 help the MPTCP packet scheduler perform well by changing the handling of notsent_lowat in subflows and how available buffer space is calculated for MPTCP-level sends. ==================== Link: https://lore.kernel.org/r/20231023-send-net-next-20231023-2-v1-0-9dc60939d371@kernel.orgSigned-off-by:
-
Paolo Abeni authored
The MPTCP protocol account for the data enqueued on all the subflows to the main socket send buffer, while the send buffer auto-tuning algorithm set the main socket send buffer size as the max size among the subflows. That causes bad performances when at least one subflow is sndbuf limited, e.g. due to very high latency, as the MPTCP scheduler can't even fill such buffer. Change the send-buffer auto-tuning algorithm to compute the main socket send buffer size as the sum of all the subflows buffer size. Reviewed-by:
Mat Martineau <martineau@kernel.org> Signed-off-by:
Paolo Abeni <pabeni@redhat.com> Signed-off-by:
-
Paolo Abeni authored
Any latency related tuning taking action at the subflow level does not really affect the user-space, as only the main MPTCP socket is relevant. Anyway any limiting setting may foul the MPTCP scheduler, not being able to fully use the subflow-level cwin, leading to very poor b/w usage. Enforce notsent_lowat to be a no-op on every subflow. Note that TCP_NOTSENT_LOWAT is currently not supported, and properly dealing with that will require more invasive changes. Reviewed-by:
-
Paolo Abeni authored
Move the socket option synchronization for active subflows at subflow creation time. This allows removing the now unused unlocked variant of such helper. While at that, clean-up a bit the mptcp_subflow_create_socket() errors path. Reviewed-by:
-
Paolo Abeni authored
The perf traces show an high cost for the MPTCP transmit path memcpy. It turn out that the helper currently in use carries quite a bit of unneeded overhead, e.g. to map/unmap the memory pages. Moving to the 'copy_from_iter' variant removes such overhead and additionally gains the no-cache support. Reviewed-by:
-
Paolo Abeni authored
The MPTCP protocol allow setting sk_rcvlowat, but the value there is currently ignored. Additionally, the default subflows sk_rcvlowat basically disables per subflow delayed ack: the MPTCP protocol move the incoming data from the subflows into the msk socket as soon as the TCP stacks invokes the subflow data_ready callback. Later, when __tcp_ack_snd_check() takes action, the subflow-level copied_seq matches rcv_nxt, and that mandate for an immediate ack. Let the mptcp receive path be aware of such threshold, explicitly tracking the amount of data available to be ready and checking vs sk_rcvlowat in mptcp_poll() and before waking-up readers. Additionally implement the set_rcvlowat() callback, to properly handle the rcvbuf auto-tuning on sk_rcvlowat changes. Finally to properly handle delayed ack, force the subflow level threshold to 0 and instead explicitly ask for an immediate ack when the msk level th is not reached. Reviewed-by:
-
Paolo Abeni authored
So that other users could access it. Notably MPTCP will use it in the next patch. No functional change intended. Acked-by:
Matthieu Baerts <matttbe@kernel.org> Signed-off-by:
-
Paolo Abeni authored
The 'data_avail' subflow field is already used as plain boolean, drop the custom binary enum type and switch to bool. No functional changed intended. Reviewed-by:
-
Paolo Abeni authored
Currently the socket level counter aggregating the received data does not take in account the data received via fastopen. Address the issue updating the counter as required. Fixes: 38967f42 ("mptcp: track some aggregate data counters") Reviewed-by:
-
Paolo Abeni authored
The MPTCP protocol allows sockets with no alive subflows to stay in ESTABLISHED status for and user-defined timeout, to allow for later subflows creation. Currently such timeout is constant - TCP_TIMEWAIT_LEN. Let the user-space configure them via a newly added sysctl, to better cope with busy servers and simplify (make them faster) the relevant pktdrill tests. Note that the new know does not apply to orphaned MPTCP socket waiting for the data_fin handshake completion: they always wait TCP_TIMEWAIT_LEN. Reviewed-by:
-
Jakub Kicinski authored
This reverts the following commits: commit 53313ed2 ("dt-bindings: marvell: Add Marvell MV88E6060 DSA schema") commit 0f35369b ("dt-bindings: marvell: Rewrite MV88E6xxx in schema") commit 605a5f5d ("ARM64: dts: marvell: Fix some common switch mistakes") commit bfedd842 ("ARM: dts: nxp: Fix some common switch mistakes") commit 2b83557a ("ARM: dts: marvell: Fix some common switch mistakes") commit ddae07ce ("dt-bindings: net: mvusb: Fix up DSA example") commit b5ef6171 ("dt-bindings: net: dsa: Require ports or ethernet-ports") As repoted by Vladimir, it breaks boot on the Turris MOX board. Link: https://lore.kernel.org/all/20231025093632.fb2qdtunzaznd73z@skbuf/Signed-off-by:
-
Bragatheswaran Manickavel authored
kfree()/vfree() internally perform NULL check on the pointer handed to it and take no action if it indeed is NULL. Hence there is no need for a pre-check of the memory pointer before handing it to kfree()/vfree(). Issue reported by ifnullfree.cocci Coccinelle semantic patch script. Signed-off-by:
Bragatheswaran Manickavel <bragathemanick0908@gmail.com> Reviewed-by:
Shannon Nelson <shannon.nelson@amd.com> Reviewed-by:
Wojciech Drewek <wojciech.drewek@intel.com> Signed-off-by:
David S. Miller <davem@davemloft.net>
-
David S. Miller authored
Linus Walleij says: ==================== Create a binding for the Marvell MV88E6xxx DSA switches The Marvell switches are lacking DT bindings. I need proper schema checking to add LED support to the Marvell switch. Just how it is, it can't go on like this. Some Device Tree fixes are included in the series, these remove the major and most annoying warnings fallout noise: some warnings remain, and these are of more serious nature, such as missing phy-mode. They can be applied individually, or to the networking tree with the rest of the patches. Thanks to Andrew Lunn, Vladimir Oltean and Russell King for excellent review and feedback! --- Changes in v7: - Fix the elaborate spacing to satisfy yamllint in the ports/ethernet-ports requirement. - Link to v6: https://lore.kernel.org/r/20231024-marvell-88e6152-wan-led-v6-0-993ab0949344@linaro.org Changes in v6: - Fix ports/ethernet-ports requirement with proper indenting (hopefully). - Link to v5: https://lore.kernel.org/r/20231023-marvell-88e6152-wan-led-v5-0-0e82952015a7@linaro.org Changes in v5: - Consistently rename switch@n to ethernet-switch@n in all cleanup patches - Consistently rename ports to ethernet-ports in all cleanup patches - Consistently rename all port@n to ethernet-port@n in all cleanup patches - Consistently rename all phy@n to ethernet-phy@n in all cleanup patches - Restore the nodename on the Turris MOX which has a U-Boot binary using the nodename as ABI, put in a blurb warning about this so no-one else tries to change it in the future. - Drop dsa.yaml direct references where we reference dsa.yaml#/$defs/ethernet-ports - Replace the conjured MV88E6xxx example by a better one based on imx6qdl plus strictly named nodes and added reset-gpios for a more complete example, and another example using the interrupt controller based on armada-381-netgear-gs110emx.dts - Bump lineage to 2008 as Vladimir says the code was developed starting 2008. - Link to v4: https://lore.kernel.org/r/20231018-marvell-88e6152-wan-led-v4-0-3ee0c67383be@linaro.org Changes in v4: - Rebase the series on top of Rob's series "dt-bindings: net: Child node schema cleanups" (or the hex numbered ports will not work) - Fix up a whitespacing error corrupting v3... - Add a new patch making the generic DSA binding require ports or ethernet-ports in the switch node. - Drop any corrections of port@a in the patches. - Drop oneOf in the compatible enum for mv88e6xxx - Use ethernet-switch, ethernet-ports and ethernet-phy in the examples - Transclude the dsa.yaml#/$defs/ethernet-ports define for ports - Move the DTS and binding fixes first, before the actual bindings, so they apply without (too many) warnings as fallout. - Drop stray colon in text. - Drop example port in the mveusb binding. - Link to v3: https://lore.kernel.org/r/20231016-marvell-88e6152-wan-led-v3-0-38cd449dfb15@linaro.org Changes in v3: - Fix up a related mvusb example in a different binding that the scripts were complaining about. - Fix up the wording on internal vs external MDIO buses in the mv88e6xxx binding document. - Remove pointless label and put the right rev-mii into the MV88E6060 schema. - Link to v2: https://lore.kernel.org/r/20231014-marvell-88e6152-wan-led-v2-0-7fca08b68849@linaro.org Changes in v2: - Break out a separate Marvell MV88E6060 binding file. I stand corrected. - Drop the idea to rely on nodename mdio-external for the external MDIO bus, keep the compatible, drop patch for the driver. - Fix more Marvell DT mistakes. - Fix NXP DT mistakes in a separate patch. - Fix Marvell ARM64 mistakes in a separate patch. - Link to v1: https://lore.kernel.org/r/20231013-marvell-88e6152-wan-led-v1-0-0712ba99857c@linaro.org ==================== Signed-off-by:
Linus Walleij <linus.walleij@linaro.org> Signed-off-by:
-
Linus Walleij authored
The Marvell MV88E6060 is one of the oldest DSA switches from Marvell, and it has DT bindings used in the wild. Let's define them properly. It is different enough from the rest of the MV88E6xxx switches that it deserves its own binding. Reviewed-by:
Andrew Lunn <andrew@lunn.ch> Reviewed-by:
Vladimir Oltean <vladimir.oltean@nxp.com> Reviewed-by:
Rob Herring <robh@kernel.org> Signed-off-by:
Florian Fainelli <florian.fainelli@broadcom.com> Signed-off-by:
-
Linus Walleij authored
This is an attempt to rewrite the Marvell MV88E6xxx switch bindings in YAML schema. The current text binding says: WARNING: This binding is currently unstable. Do not program it into a FLASH never to be changed again. Once this binding is stable, this warning will be removed. Well that never happened before we switched to YAML markup, we can't have it like this, what about fixing the mess? Reviewed-by:
-
Linus Walleij authored
Fix some errors in the Marvell MV88E6xxx switch descriptions: - The top node had no address size or cells. - switch0@0 is not OK, should be ethernet-switch@0. - ports should be ethernet-ports - port@0 should be ethernet-port@0 - PHYs should be named ethernet-phy@ Reviewed-by:
-
Linus Walleij authored
Fix some errors in the Marvell MV88E6xxx switch descriptions: - switch0@0 is not OK, should be ethernet-switch@0 - ports should be ethernet-ports - port should be ethernet-port - phy should be ethernet-phy Reviewed-by:
-
Linus Walleij authored
Fix some errors in the Marvell MV88E6xxx switch descriptions: - The top node had no address size or cells. - switch0@0 is not OK, should be ethernet-switch@0. - The ports node should be named ethernet-ports - The ethernet-ports node should have port@0 etc children, no plural "ports" in the children. - Ports should be named ethernet-port@0 etc - PHYs should be named ethernet-phy@0 etc This serves as an example of fixes needed for introducing a schema for the bindings, but the patch can simply be applied. Reviewed-by:
-
Linus Walleij authored
When adding a proper schema for the Marvell mx88e6xxx switch, the scripts start complaining about this embedded example: dtschema/dtc warnings/errors: net/marvell,mvusb.example.dtb: switch@0: ports: '#address-cells' is a required property from schema $id: http://devicetree.org/schemas/net/dsa/marvell,mv88e6xxx.yaml# net/marvell,mvusb.example.dtb: switch@0: ports: '#size-cells' is a required property from schema $id: http://devicetree.org/schemas/net/dsa/marvell,mv88e6xxx.yaml# Fix this up by extending the example with those properties in the ports node. While we are at it, rename "ports" to "ethernet-ports" and rename "switch" to "ethernet-switch" as this is recommended practice. Reviewed-by:
-
Linus Walleij authored
Bindings using dsa.yaml#/$defs/ethernet-ports specify that a DSA switch node need to have a ports or ethernet-ports subnode, and that is actually required, so add requirements using oneOf. Suggested-by:
-
Florian Westphal authored
net->ct.labels_used was meant to convey 'number of ip/nftables rules that need the label extension allocated'. act_ct enables this for each net namespace, which voids all attempts to avoid ct->ext allocation when possible. Move this increment to the control plane to request label extension space allocation only when its needed. Signed-off-by:
Pedro Tammela <pctammela@mojatatu.com> Reviewed-by:
Jamal Hadi Salim <jhs@mojatatu.com> Signed-off-by:
-
Hao Chen authored
Add HCLGE_SUPPORT_50G_R1_BIT and HCLGE_SUPPORT_100G_R2_BIT two capability bits and Corresponding link modes. Signed-off-by:
Hao Chen <chenhao418@huawei.com> Signed-off-by:
Jijie Shao <shaojijie@huawei.com> Reviewed-by:
-
David S. Miller authored
-
Oleksij Rempel authored
Add WoL support for KSZ9477 family of switches. This code was tested on KSZ8563 chip. KSZ9477 family of switches supports multiple PHY events: - wake on Link Up - wake on Energy Detect. Since current UAPI can't differentiate between this PHY events, map all of them to WAKE_PHY. Signed-off-by:
Oleksij Rempel <o.rempel@pengutronix.de> Reviewed-by:
-
Oleksij Rempel authored
KSZ switches with WoL support signals wake event over PME pin. If this pin is attached to some external PMIC or System Controller can't be described as GPIO, the only way to describe it in the devicetree is to use wakeup-source property. So, add support for this property and enable PME switch output if this property is present. Signed-off-by:
-
Oleksij Rempel authored
Add wakeup-source property to enable Wake on Lan functionality in the switch. Since PME wake pin is not always attached to the SoC, use wakeup-source instead of wakeup-gpios Signed-off-by:
Conor Dooley <conor.dooley@microchip.com> Reviewed-by:
-
Oleksij Rempel authored
Add the missing offset for the global MAC address register (REG_SW_MAC_ADDR) for the ksz8863 family of switches. Signed-off-by:
Russell King (Oracle) <rmk+kernel@armlinux.org.uk> Reviewed-by:
-
Justin Stitt authored
strncpy() is deprecated for use on NUL-terminated destination strings [1] and as such we should prefer more robust and less ambiguous string interfaces. We expect new_entry->dbf_name to be NUL-terminated based on its use with strcmp(): | if (strcmp(entry->dbf_name, name) == 0) { Moreover, NUL-padding is not required as new_entry is kzalloc'd just before this assignment: | new_entry = kzalloc(sizeof(struct qeth_dbf_entry), GFP_KERNEL); ... rendering any future NUL-byte assignments (like the ones strncpy() does) redundant. Considering the above, a suitable replacement is `strscpy` [2] due to the fact that it guarantees NUL-termination on the destination buffer without unnecessarily NUL-padding. Link: https://www.kernel.org/doc/html/latest/process/deprecated.html#strncpy-on-nul-terminated-strings [1] Link: https://manpages.debian.org/testing/linux-manual-4.8/strscpy.9.en.html [2] Link: https://github.com/KSPP/linux/issues/90Signed-off-by:Justin Stitt <justinstitt@google.com> Reviewed-by:
Thorsten Winkler <twinkler@linux.ibm.com> Tested-by:
Kees Cook <keescook@chromium.org> Link: https://lore.kernel.org/r/20231023-strncpy-drivers-s390-net-qeth_core_main-c-v1-1-e7ce65454446@google.comSigned-off-by:
-
Justin Stitt authored
strncpy() is deprecated for use on NUL-terminated destination strings [1] and as such we should prefer more robust and less ambiguous string interfaces. We expect chid to be NUL-terminated based on its use with format strings: CTCM_DBF_TEXT_(SETUP, CTC_DBF_INFO, "%s(%s) %s", CTCM_FUNTAIL, chid, ok ? "OK" : "failed"); Moreover, NUL-padding is not required as it is _only_ used in this one instance with a format string. Considering the above, a suitable replacement is `strscpy` [2] due to the fact that it guarantees NUL-termination on the destination buffer without unnecessarily NUL-padding. We can also drop the +1 from chid's declaration as we no longer need to be cautious about leaving a spot for a NUL-byte. Let's use the more idiomatic strscpy usage of (dest, src, sizeof(dest)) as this more closely ties the destination buffer to the length. Link: https://www.kernel.org/doc/html/latest/process/deprecated.html#strncpy-on-nul-terminated-strings [1] Link: https://manpages.debian.org/testing/linux-manual-4.8/strscpy.9.en.html [2] Link: https://github.com/KSPP/linux/issues/90Signed-off-by:
-
Lorenzo Bianconi authored
wo pointer is no longer used in wo_r32 and wo_w32 routines so get rid of it. Signed-off-by:
Lorenzo Bianconi <lorenzo@kernel.org> Link: https://lore.kernel.org/r/530537db0872f7523deff21f0a5dfdd9b75fdc9d.1698098459.git.lorenzo@kernel.orgSigned-off-by:
-
Lorenzo Bianconi authored
The WED mcu firmware does not contain all the memory regions defined in the dts reserved_memory node (e.g. MT7986 WED firmware does not contain cpu-boot region). Reverse the mtk_wed_mcu_run_firmware() logic to check all the fw sections are defined in the dts reserved_memory node. Fixes: c6d961ae ("net: ethernet: mtk_wed: move mem_region array out of mtk_wed_mcu_load_firmware") Tested-by:
Frank Wunderlich <frank-w@public-files.de> Signed-off-by:
Jacob Keller <jacob.e.keller@intel.com> Link: https://lore.kernel.org/r/d983cbfe8ea562fef9264de8f0c501f7d5705bd5.1698098381.git.lorenzo@kernel.orgSigned-off-by:
-
- 24 Oct, 2023 7 commits
-
-
Jakub Kicinski authored
Wolfram Sang says: ==================== net: ethernet: renesas: infrastructure preparations for upcoming driver Before we upstream a new driver, Niklas and I thought that a few cleanups for Kconfig/Makefile will help readability and maintainability. Here they are, looking forward to comments. ==================== Link: https://lore.kernel.org/r/20231022205316.3209-1-wsa+renesas@sang-engineering.comSigned-off-by:
-
Wolfram Sang authored
Mentioning SoCs in Kconfig descriptions tends to get stale (e.g. RAVB is missing RZV2M) or imprecise (e.g. SH_ETH is not available on all R8A779x). Drop them instead of providing vague information. Improve the file description a tad while here. Signed-off-by:
Wolfram Sang <wsa+renesas@sang-engineering.com> Reviewed-by:
Sergey Shtylyov <s.shtylyov@omp.ru> Reviewed-by:
Niklas Söderlund <niklas.soderlund+renesas@ragnatech.se> Link: https://lore.kernel.org/r/20231022205316.3209-3-wsa+renesas@sang-engineering.comSigned-off-by:
-
Wolfram Sang authored
A new Renesas driver shall be added soon. Prepare the Makefile by grouping the specific objects to the Kconfig symbol for better readability. Improve the file description a tad while here. Signed-off-by:
-
Swarup Laxman Kotiaklapudi authored
Change ifconfig with ip command, on a system where ifconfig is not used this script will not work correcly. Test result with this patchset: sudo make TARGETS="net" kselftest .... TAP version 13 1..1 timeout set to 1500 selftests: net: route_localnet.sh run arp_announce test net.ipv4.conf.veth0.route_localnet = 1 net.ipv4.conf.veth1.route_localnet = 1 net.ipv4.conf.veth0.arp_announce = 2 net.ipv4.conf.veth1.arp_announce = 2 PING 127.25.3.14 (127.25.3.14) from 127.25.3.4 veth0: 56(84) bytes of data. 64 bytes from 127.25.3.14: icmp_seq=1 ttl=64 time=0.038 ms 64 bytes from 127.25.3.14: icmp_seq=2 ttl=64 time=0.068 ms 64 bytes from 127.25.3.14: icmp_seq=3 ttl=64 time=0.068 ms 64 bytes from 127.25.3.14: icmp_seq=4 ttl=64 time=0.068 ms 64 bytes from 127.25.3.14: icmp_seq=5 ttl=64 time=0.068 ms --- 127.25.3.14 ping statistics --- 5 packets transmitted, 5 received, 0% packet loss, time 4073ms rtt min/avg/max/mdev = 0.038/0.062/0.068/0.012 ms ok run arp_ignore test net.ipv4.conf.veth0.route_localnet = 1 net.ipv4.conf.veth1.route_localnet = 1 net.ipv4.conf.veth0.arp_ignore = 3 net.ipv4.conf.veth1.arp_ignore = 3 PING 127.25.3.14 (127.25.3.14) from 127.25.3.4 veth0: 56(84) bytes of data. 64 bytes from 127.25.3.14: icmp_seq=1 ttl=64 time=0.032 ms 64 bytes from 127.25.3.14: icmp_seq=2 ttl=64 time=0.065 ms 64 bytes from 127.25.3.14: icmp_seq=3 ttl=64 time=0.066 ms 64 bytes from 127.25.3.14: icmp_seq=4 ttl=64 time=0.065 ms 64 bytes from 127.25.3.14: icmp_seq=5 ttl=64 time=0.065 ms --- 127.25.3.14 ping statistics --- 5 packets transmitted, 5 received, 0% packet loss, time 4092ms rtt min/avg/max/mdev = 0.032/0.058/0.066/0.013 ms ok ok 1 selftests: net: route_localnet.sh ... Signed-off-by:
Swarup Laxman Kotiaklapudi <swarupkotikalapudi@gmail.com> Reviewed-by:
David Ahern <dsahern@kernel.org> Link: https://lore.kernel.org/r/20231023123422.2895-1-swarupkotikalapudi@gmail.comSigned-off-by:
-
Jakub Kicinski authored
Florian Fainelli says: ==================== Switch DSA to inclusive terminology One of the action items following Netconf'23 is to switch subsystems to use inclusive terminology. DSA has been making extensive use of the "master" and "slave" words which are now replaced by "conduit" and "user" respectively. ==================== Link: https://lore.kernel.org/r/20231023181729.1191071-1-florian.fainelli@broadcom.comSigned-off-by:
-
Florian Fainelli authored
This preserves the existing IFLA_DSA_MASTER which is part of the uAPI and creates an alias named IFLA_DSA_CONDUIT. Reviewed-by:
-
Florian Fainelli authored
Use more inclusive terms throughout the DSA subsystem by moving away from "master" which is replaced by "conduit" and "slave" which is replaced by "user". No functional changes. Acked-by:
Stephen Hemminger <stephen@networkplumber.org> Reviewed-by:
-
