- 08 Mar, 2024 20 commits
-
-
Russell King (Oracle) authored
Update the Marvell 88e6185 PCS driver to use neg_mode rather than the mode argument to match the other updated PCS drivers. Signed-off-by:
Russell King (Oracle) <rmk+kernel@armlinux.org.uk> Reviewed-by:
Andrew Lunn <andrew@lunn.ch> Reviewed-by:
Florian Fainelli <florian.fainelli@broadcom.com> Link: https://lore.kernel.org/r/E1rhosE-003yuc-FM@rmk-PC.armlinux.org.ukSigned-off-by:
Jakub Kicinski <kuba@kernel.org>
-
Russell King (Oracle) authored
Update the RZN1-MIIC PCS driver to use neg_mode rather than the mode argument to match the other updated PCS drivers. Signed-off-by:
-
Russell King (Oracle) authored
The comment in m88e1111_config_init_1000basex() is wrong - it claims that Autoneg will be enabled, but this doesn't actually happen. Signed-off-by:
-
Eric Dumazet authored
After commit b5a89915 ("netlink: handle EMSGSIZE errors in the core"), we can remove some code that was not 100 % correct anyway. Signed-off-by:
Eric Dumazet <edumazet@google.com> Reviewed-by:
Simon Horman <horms@kernel.org> Reviewed-by:
Ido Schimmel <idosch@nvidia.com> Reviewed-by:
David Ahern <dsahern@kernel.org> Link: https://lore.kernel.org/r/20240306102426.245689-1-edumazet@google.comSigned-off-by:
-
Kui-Feng Lee authored
ipv6_gc fails occasionally. According to the study, fib6_run_gc() using jiffies_round() to round the GC interval could increase the waiting time up to 750ms (3/4 seconds). The timer has a granularity of 512ms at the range 4s to 32s. That means a route with an expiration time E seconds can wait for more than E * 2 + 1 seconds if the GC interval is also E seconds. E * 2 + 2 seconds should be enough for waiting for removing routes. Also remove a check immediately after replacing 5 routes since it is very likely to remove some of routes before completing the last route with a slow environment. Signed-off-by:
Kui-Feng Lee <thinker.li@gmail.com> Link: https://lore.kernel.org/r/20240305183949.258473-1-thinker.li@gmail.comSigned-off-by:
-
Christoph Paasch authored
We observed that TCP-pacing was falling back to the TCP-layer pacing instead of utilizing sch_fq for the pacing. This causes significant CPU-usage due to the hrtimer running on a per-TCP-connection basis. The issue is that mpls_xmit() calls skb_orphan() and thus sets skb->sk to NULL. Which implies that many of the goodies of TCP won't work. Pacing falls back to TCP-layer pacing. TCP Small Queues does not work, ... It is safe to remove this call to skb_orphan() in mpls_xmit() as there really is not reason for it to be there. It appears that this call to skb_orphan comes from the very initial implementation of MPLS. Cc: Roopa Prabhu <roopa@nvidia.com> Reported-by:
Craig Taylor <cmtaylor@apple.com> Signed-off-by:
Christoph Paasch <cpaasch@apple.com> Link: https://lore.kernel.org/r/20240306181117.77419-1-cpaasch@apple.comSigned-off-by:
-
Florian Fainelli authored
With commit 34d21de9 ("net: Move {l,t,d}stats allocation to core and convert veth & vrf"), stats allocation could be done on net core instead of in this driver. With this new approach, the driver doesn't have to bother with error handling (allocation failure checking, making sure free happens in the right spot, etc). This is core responsibility now. Remove the allocation in the DSA user network device code and leverage the network core allocation instead. Signed-off-by:
Breno Leitao <leitao@debian.org> Link: https://lore.kernel.org/r/20240306200416.2973179-1-florian.fainelli@broadcom.comSigned-off-by:
-
Uwe Kleine-König authored
The .remove() callback for a platform driver returns an int which makes many driver authors wrongly assume it's possible to do error handling by returning an error code. However the value returned is ignored (apart from emitting a warning) and this typically results in resource leaks. To improve here there is a quest to make the remove callback return void. In the first step of this quest all drivers are converted to .remove_new(), which already returns void. Eventually after all drivers are converted, .remove_new() will be renamed to .remove(). Trivially convert this driver from always returning zero in the remove callback to the void returning variant. Signed-off-by:
Uwe Kleine-König <u.kleine-koenig@pengutronix.de> Reviewed-by:
-
Jakub Kicinski authored
Donald Hunter says: ==================== tools/net/ynl: Add support for nlctrl netlink family This series adds a new YNL spec for the nlctrl family, plus some fixes and enhancements for ynl. Patch 1 fixes an extack decoding bug Patch 2 gives cleaner netlink error reporting Patch 3 fixes an array-nest codegen bug Patch 4 adds nest-type-value support to ynl Patch 5 fixes the ynl schemas to allow empty enum-name attrs Patch 6 contains the nlctrl spec ==================== Link: https://lore.kernel.org/r/20240306231046.97158-1-donald.hunter@gmail.comSigned-off-by:
-
Donald Hunter authored
Add a spec for the nlctrl family. Example usage: ./tools/net/ynl/cli.py \ --spec Documentation/netlink/specs/nlctrl.yaml \ --do getfamily --json '{"family-name": "nlctrl"}' ./tools/net/ynl/cli.py \ --spec Documentation/netlink/specs/nlctrl.yaml \ --dump getpolicy --json '{"family-name": "nlctrl"}' Signed-off-by:Donald Hunter <donald.hunter@gmail.com> Link: https://lore.kernel.org/r/20240306231046.97158-7-donald.hunter@gmail.comSigned-off-by:
-
Donald Hunter authored
Update the ynl schemas to allow the specification of empty enum names for all enum code generation. Signed-off-by:
-
Donald Hunter authored
The nlctrl genetlink-legacy family uses nest-type-value encoding as described in Documentation/userspace-api/netlink/genetlink-legacy.rst Add nest-type-value decoding to ynl. Signed-off-by:
-
Donald Hunter authored
ynl-gen-c generates e.g. 'calloc(mcast_groups, sizeof(*dst->mcast_groups))' for array-nest attrs when it should be 'n_mcast_groups'. Add a 'n_' prefix in the generated code for array-nests. Signed-off-by:
-
Donald Hunter authored
ynl does not handle NlError exceptions so they get reported like program failures. Handle the NlError exceptions and report the netlink errors more cleanly. Example now: Netlink error: No such file or directory nl_len = 44 (28) nl_flags = 0x300 nl_type = 2 error: -2 extack: {'bad-attr': '.op'} Example before: Traceback (most recent call last): File "/home/donaldh/net-next/./tools/net/ynl/cli.py", line 81, in <module> main() File "/home/donaldh/net-next/./tools/net/ynl/cli.py", line 69, in main reply = ynl.dump(args.dump, attrs) ^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/home/donaldh/net-next/tools/net/ynl/lib/ynl.py", line 906, in dump return self._op(method, vals, [], dump=True) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/home/donaldh/net-next/tools/net/ynl/lib/ynl.py", line 872, in _op raise NlError(nl_msg) lib.ynl.NlError: Netlink error: No such file or directory nl_len = 44 (28) nl_flags = 0x300 nl_type = 2 error: -2 extack: {'bad-attr': '.op'} Signed-off-by: -
Donald Hunter authored
Extack decoding was using a hard-coded msg header size of 20 but netlink-raw has a header size of 16. Use a protocol specific msghdr_size() when decoding the attr offssets. Signed-off-by:
-
Jakub Kicinski authored
Ricardo B. Marliere says: ==================== isdn: constify struct class usage This is a simple and straight forward cleanup series that aims to make the class structures in isdn constant. This has been possible since 2023 [1]. [1]: https://lore.kernel.org/all/2023040248-customary-release-4aec@gregkh/ ==================== Link: https://lore.kernel.org/r/20240305-class_cleanup-isdn-v1-0-6f0edca75b61@marliere.netSigned-off-by:
-
Ricardo B. Marliere authored
Since commit 43a7206b ("driver core: class: make class_register() take a const *"), the driver core allows for struct class to be in read-only memory, so move the capi_class structure to be declared at build time placing it into read-only memory, instead of having to be dynamically allocated at boot time. Suggested-by:
Greg Kroah-Hartman <gregkh@linuxfoundation.org> Signed-off-by:
Ricardo B. Marliere <ricardo@marliere.net> Reviewed-by:
-
Ricardo B. Marliere authored
Since commit 43a7206b ("driver core: class: make class_register() take a const *"), the driver core allows for struct class to be in read-only memory, so move the elements_class structure to be declared at build time placing it into read-only memory, instead of having to be dynamically allocated at boot time. Suggested-by:
-
Jérémie Dautheribes authored
Drop reference to the 25MHz clock as it has nothing to do with connecting the PHY and the MAC. Add info about the reference clock direction between the PHY and the MAC as it depends on the selected rmii mode. Suggested-by:
Jérémie Dautheribes <jeremie.dautheribes@bootlin.com> Reviewed-by:
Conor Dooley <conor.dooley@microchip.com> Link: https://lore.kernel.org/r/20240305141309.127669-1-jeremie.dautheribes@bootlin.comSigned-off-by:
-
Justin Swartz authored
Remove the "You can read more about X.25 at" links provided in Kconfig as they have not pointed at any relevant pages for quite a while. An old copy of https://www.sangoma.com/tutorials/x25/ can be retrieved via https://archive.org/web/ but nothing useful seems to have been preserved for http://docwiki.cisco.com/wiki/X.25 For the sake of necromancy and those who really did want to read more about X.25, a previous incarnation of Kconfig included a link to: http://www.cisco.com/univercd/cc/td/doc/product/software/ios11/cbook/cx25.htm Which can still be read at: https://web.archive.org/web/20071013101232/http://cisco.com/en/US/docs/ios/11_0/router/configuration/guide/cx25.htmlSigned-off-by:
Justin Swartz <justin.swartz@risingedge.co.za> Acked-by:
Martin Schiller <ms@dev.tdt.de> Link: https://lore.kernel.org/r/20240306112659.25375-1-justin.swartz@risingedge.co.zaSigned-off-by:
-
- 07 Mar, 2024 20 commits
-
-
Jakub Kicinski authored
Donald points out that we don't check for overflows. Stash the length of the message on nlmsg_pid (nlmsg_seq would do as well). This allows the attribute helpers to remain self-contained (no extra arguments). Also let the put helpers continue to return nothing. The error is checked only in (newly introduced) ynl_msg_end(). Reviewed-by:
-
git://git.kernel.org/pub/scm/linux/kernel/git/netdev/netJakub Kicinski authored
Cross-merge networking fixes after downstream PR. No conflicts. Adjacent changes: net/core/page_pool_user.c 0b11b1c5 ("netdev: let netlink core handle -EMSGSIZE errors") 429679dc ("page_pool: fix netlink dump stop/resume") Signed-off-by:
-
git://git.kernel.org/pub/scm/linux/kernel/git/netdev/netLinus Torvalds authored
Pull networking fixes from Paolo Abeni: "Including fixes from bpf, ipsec and netfilter. No solution yet for the stmmac issue mentioned in the last PR, but it proved to be a lockdep false positive, not a blocker. Current release - regressions: - dpll: move all dpll<>netdev helpers to dpll code, fix build regression with old compilers Current release - new code bugs: - page_pool: fix netlink dump stop/resume Previous releases - regressions: - bpf: fix verifier to check bpf_func_state->callback_depth when pruning states as otherwise unsafe programs could get accepted - ipv6: avoid possible UAF in ip6_route_mpath_notify() - ice: reconfig host after changing MSI-X on VF - mlx5: - e-switch, change flow rule destination checking - add a memory barrier to prevent a possible null-ptr-deref - switch to using _bh variant of of spinlock where needed Previous releases - always broken: - netfilter: nf_conntrack_h323: add protection for bmp length out of range - bpf: fix to zero-initialise xdp_rxq_info struct before running XDP program in CPU map which led to random xdp_md fields - xfrm: fix UDP encapsulation in TX packet offload - netrom: fix data-races around sysctls - ice: - fix potential NULL pointer dereference in ice_bridge_setlink() - fix uninitialized dplls mutex usage - igc: avoid returning frame twice in XDP_REDIRECT - i40e: disable NAPI right after disabling irqs when handling xsk_pool - geneve: make sure to pull inner header in geneve_rx() - sparx5: fix use after free inside sparx5_del_mact_entry - dsa: microchip: fix register write order in ksz8_ind_write8() Misc: - selftests: mptcp: fixes for diag.sh" * tag 'net-6.8-rc8' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net: (63 commits) net: pds_core: Fix possible double free in error handling path netrom: Fix data-races around sysctl_net_busy_read netrom: Fix a data-race around sysctl_netrom_link_fails_count netrom: Fix a data-race around sysctl_netrom_routing_control netrom: Fix a data-race around sysctl_netrom_transport_no_activity_timeout netrom: Fix a data-race around sysctl_netrom_transport_requested_window_size netrom: Fix a data-race around sysctl_netrom_transport_busy_delay netrom: Fix a data-race around sysctl_netrom_transport_acknowledge_delay netrom: Fix a data-race around sysctl_netrom_transport_maximum_tries netrom: Fix a data-race around sysctl_netrom_transport_timeout netrom: Fix data-races around sysctl_netrom_network_ttl_initialiser netrom: Fix a data-race around sysctl_netrom_obsolescence_count_initialiser netrom: Fix a data-race around sysctl_netrom_default_path_quality netfilter: nf_conntrack_h323: Add protection for bmp length out of range netfilter: nf_tables: mark set as dead when unbinding anonymous set with timeout netfilter: nft_ct: fix l3num expectations with inet pseudo family netfilter: nf_tables: reject constant set with timeout netfilter: nf_tables: disallow anonymous set with timeout flag net/rds: fix WARNING in rds_conn_connect_if_down net: dsa: microchip: fix register write order in ksz8_ind_write8() ... -
Paolo Abeni authored
Jason Xing says: ==================== tcp: add two missing addresses when using trace When I reviewed other people's patch [1], I noticed that similar things also happen in tcp_event_skb class and tcp_event_sk_skb class. They don't print those two addrs of skb/sk which already exist. In this patch, I just do as other trace functions do, like trace_net_dev_start_xmit(), to know the exact flow or skb we would like to know in case some systems doesn't support BPF programs well or we have to use /sys/kernel/debug/tracing only for some reasons. [1] Link: https://lore.kernel.org/netdev/CAL+tcoAhvFhXdr1WQU8mv_6ZX5nOoNpbOLAB6=C+DB-qXQ11Ew@mail.gmail.com/ v2 Link: https://lore.kernel.org/netdev/CANn89iJcScraKAUk1GzZFoOO20RtC9iXpiJ4LSOWT5RUAC_QQA@mail.gmail.com/ 1. change the description. ==================== Link: https://lore.kernel.org/r/20240304092934.76698-1-kerneljasonxing@gmail.comSigned-off-by:
Paolo Abeni <pabeni@redhat.com>
-
Jason Xing authored
Use the existing parameter and print the address of skbaddr as other trace functions do. Signed-off-by:
Jason Xing <kernelxing@tencent.com> Reviewed-by:
-
Jason Xing authored
Printing the addresses can help us identify the exact skb/sk for those system in which it's not that easy to run BPF program. As we can see, it already fetches those, then use it directly and it will print like below: ...tcp_retransmit_skb: skbaddr=XXX skaddr=XXX family=AF_INET... Signed-off-by:
-
Paolo Abeni authored
Maxime Chevallier says: ==================== doc: sfp-phylink: update the porting guide Here's a V3 for an update on the phylink porting guide. The only difference with V2 is a whitespace fix along with a line-wrap. The main point of the update is the description of a basic process to follow to expose one or more PCS to phylink. Let me know if you spot any inaccuracies in the guide. The second patch is a simple fixup on some in-code doc that was spotted while updating the guide. Link to V2: https://lore.kernel.org/netdev/20240228095755.1499577-1-maxime.chevallier@bootlin.com/ Link to V1: https://lore.kernel.org/netdev/20240220160406.3363002-1-maxime.chevallier@bootlin.com/ ==================== Link: https://lore.kernel.org/r/20240301164309.3643849-1-maxime.chevallier@bootlin.comSigned-off-by:
-
Maxime Chevallier authored
commit 4d72c3bb ("net: phylink: strip out pre-March 2020 legacy code") dropped the mac_pcs_get_state ops in phylink_mac_ops in favor of dedicated PCS operation pcs_get_state. However, the documentation for the pcs_get_state ops was incorrectly converted and now self-references. Drop the extra comment. Signed-off-by:
Maxime Chevallier <maxime.chevallier@bootlin.com> Signed-off-by:
-
Maxime Chevallier authored
Now that phylink has a comprehensive PCS support, update the porting guide to explain the process of supporting the PCS configuration. This also removed outdated references to phylink_config fields that no longer exists. Signed-off-by:
-
Yongzhi Liu authored
When auxiliary_device_add() returns error and then calls auxiliary_device_uninit(), Callback function pdsc_auxbus_dev_release calls kfree(padev) to free memory. We shouldn't call kfree(padev) again in the error handling path. Fix this by cleaning up the redundant kfree() and putting the error handling back to where the errors happened. Fixes: 4569cce4 ("pds_core: add auxiliary_bus devices") Signed-off-by:
Yongzhi Liu <hyperlyzcs@gmail.com> Reviewed-by:
Wojciech Drewek <wojciech.drewek@intel.com> Reviewed-by:
Shannon Nelson <shannon.nelson@amd.com> Link: https://lore.kernel.org/r/20240306105714.20597-1-hyperlyzcs@gmail.comSigned-off-by:
-
git://git.kernel.org/pub/scm/linux/kernel/git/netfilter/nfPaolo Abeni authored
Pablo Neira Ayuso says: ==================== Netfilter fixes for net The following patchset contains fixes for net: Patch #1 disallows anonymous sets with timeout, except for dynamic sets. Anonymous sets with timeouts using the pipapo set backend makes no sense from userspace perspective. Patch #2 rejects constant sets with timeout which has no practical usecase. This kind of set, once bound, contains elements that expire but no new elements can be added. Patch #3 restores custom conntrack expectations with NFPROTO_INET, from Florian Westphal. Patch #4 marks rhashtable anonymous set with timeout as dead from the commit path to avoid that async GC collects these elements. Rules that refers to the anonymous set get released with no mutex held from the commit path. Patch #5 fixes a UBSAN shift overflow in H.323 conntrack helper, from Lena Wang. netfilter pull request 24-03-07 * tag 'nf-24-03-07' of git://git.kernel.org/pub/scm/linux/kernel/git/netfilter/nf: netfilter: nf_conntrack_h323: Add protection for bmp length out of range netfilter: nf_tables: mark set as dead when unbinding anonymous set with timeout netfilter: nft_ct: fix l3num expectations with inet pseudo family netfilter: nf_tables: reject constant set with timeout netfilter: nf_tables: disallow anonymous set with timeout flag ==================== Link: https://lore.kernel.org/r/20240307021545.149386-1-pablo@netfilter.orgSigned-off-by: -
Paolo Abeni authored
Jason Xing says: ==================== netrom: Fix all the data-races around sysctls As the title said, in this patchset I fix the data-race issues because the writer and the reader can manipulate the same value concurrently. ==================== Link: https://lore.kernel.org/r/20240304082046.64977-1-kerneljasonxing@gmail.comSigned-off-by:
-
Jason Xing authored
We need to protect the reader reading the sysctl value because the value can be changed concurrently. Fixes: 1da177e4 ("Linux-2.6.12-rc2") Signed-off-by:
-
Jason Xing authored
We need to protect the reader reading the sysctl value because the value can be changed concurrently. Fixes: 1da177e4 ("Linux-2.6.12-rc2") Signed-off-by:
-
Jason Xing authored
We need to protect the reader reading the sysctl value because the value can be changed concurrently. Fixes: 1da177e4 ("Linux-2.6.12-rc2") Signed-off-by:
-
Jason Xing authored
We need to protect the reader reading the sysctl value because the value can be changed concurrently. Fixes: 1da177e4 ("Linux-2.6.12-rc2") Signed-off-by:
-
Jason Xing authored
We need to protect the reader reading the sysctl value because the value can be changed concurrently. Fixes: 1da177e4 ("Linux-2.6.12-rc2") Signed-off-by:
-
Jason Xing authored
We need to protect the reader reading the sysctl value because the value can be changed concurrently. Fixes: 1da177e4 ("Linux-2.6.12-rc2") Signed-off-by:
-
Jason Xing authored
We need to protect the reader reading the sysctl value because the value can be changed concurrently. Fixes: 1da177e4 ("Linux-2.6.12-rc2") Signed-off-by:
-
Jason Xing authored
We need to protect the reader reading the sysctl value because the value can be changed concurrently. Fixes: 1da177e4 ("Linux-2.6.12-rc2") Signed-off-by:
-
