- 11 May, 2024 9 commits
-
-
Ilya Maximets authored
OVS_PACKET_CMD_EXECUTE has 3 main attributes: - OVS_PACKET_ATTR_KEY - Packet metadata in a netlink format. - OVS_PACKET_ATTR_PACKET - Binary packet content. - OVS_PACKET_ATTR_ACTIONS - Actions to execute on the packet. OVS_PACKET_ATTR_KEY is parsed first to populate sw_flow_key structure with the metadata like conntrack state, input port, recirculation id, etc. Then the packet itself gets parsed to populate the rest of the keys from the packet headers. Whenever the packet parsing code starts parsing the ICMPv6 header, it first zeroes out fields in the key corresponding to Neighbor Discovery information even if it is not an ND packet. It is an 'ipv6.nd' field. However, the 'ipv6' is a union that shares the space between 'nd' and 'ct_orig' that holds the original tuple conntrack metadata parsed from the OVS_PACKET_ATTR_KEY. ND packets should not normally have conntrack state, so it's fine to share the space, but normal ICMPv6 Echo packets or maybe other types of ICMPv6 can have the state attached and it should not be overwritten. The issue results in all but the last 4 bytes of the destination address being wiped from the original conntrack tuple leading to incorrect packet matching and potentially executing wrong actions in case this packet recirculates within the datapath or goes back to userspace. ND fields should not be accessed in non-ND packets, so not clearing them should be fine. Executing memset() only for actual ND packets to avoid the issue. Initializing the whole thing before parsing is needed because ND packet may not contain all the options. The issue only affects the OVS_PACKET_CMD_EXECUTE path and doesn't affect packets entering OVS datapath from network interfaces, because in this case CT metadata is populated from skb after the packet is already parsed. Fixes: 9dd7f890 ("openvswitch: Add original direction conntrack tuple to sw_flow_key.") Reported-by:
Antonin Bas <antonin.bas@broadcom.com> Closes: https://github.com/openvswitch/ovs-issues/issues/327Signed-off-by:
Ilya Maximets <i.maximets@ovn.org> Acked-by:
Aaron Conole <aconole@redhat.com> Acked-by:
Eelco Chaudron <echaudro@redhat.com> Link: https://lore.kernel.org/r/20240509094228.1035477-1-i.maximets@ovn.orgSigned-off-by:
Jakub Kicinski <kuba@kernel.org>
-
Breno Leitao authored
A data-race condition has been identified in af_unix. In one data path, the write function unix_release_sock() atomically writes to sk->sk_shutdown using WRITE_ONCE. However, on the reader side, unix_stream_sendmsg() does not read it atomically. Consequently, this issue is causing the following KCSAN splat to occur: BUG: KCSAN: data-race in unix_release_sock / unix_stream_sendmsg write (marked) to 0xffff88867256ddbb of 1 bytes by task 7270 on cpu 28: unix_release_sock (net/unix/af_unix.c:640) unix_release (net/unix/af_unix.c:1050) sock_close (net/socket.c:659 net/socket.c:1421) __fput (fs/file_table.c:422) __fput_sync (fs/file_table.c:508) __se_sys_close (fs/open.c:1559 fs/open.c:1541) __x64_sys_close (fs/open.c:1541) x64_sys_call (arch/x86/entry/syscall_64.c:33) do_syscall_64 (arch/x86/entry/common.c:?) entry_SYSCALL_64_after_hwframe (arch/x86/entry/entry_64.S:130) read to 0xffff88867256ddbb of 1 bytes by task 989 on cpu 14: unix_stream_sendmsg (net/unix/af_unix.c:2273) __sock_sendmsg (net/socket.c:730 net/socket.c:745) ____sys_sendmsg (net/socket.c:2584) __sys_sendmmsg (net/socket.c:2638 net/socket.c:2724) __x64_sys_sendmmsg (net/socket.c:2753 net/socket.c:2750 net/socket.c:2750) x64_sys_call (arch/x86/entry/syscall_64.c:33) do_syscall_64 (arch/x86/entry/common.c:?) entry_SYSCALL_64_after_hwframe (arch/x86/entry/entry_64.S:130) value changed: 0x01 -> 0x03 The line numbers are related to commit dd5a440a ("Linux 6.9-rc7"). Commit e1d09c2c ("af_unix: Fix data races around sk->sk_shutdown.") addressed a comparable issue in the past regarding sk->sk_shutdown. However, it overlooked resolving this particular data path. This patch only offending unix_stream_sendmsg() function, since the other reads seem to be protected by unix_state_lock() as discussed in Link: https://lore.kernel.org/all/20240508173324.53565-1-kuniyu@amazon.com/ Fixes: 1da177e4 ("Linux-2.6.12-rc2") Signed-off-by:
Breno Leitao <leitao@debian.org> Reviewed-by:
Kuniyuki Iwashima <kuniyu@amazon.com> Link: https://lore.kernel.org/r/20240509081459.2807828-1-leitao@debian.orgSigned-off-by:
-
Linus Walleij authored
This fixes a probably long standing problem in the Cortina Gemini ethernet driver: there are some paths in the code where the IRQ registers are written without taking the proper locks. Fixes: 4d5ae32f ("net: ethernet: Add a driver for Gemini gigabit ethernet") Signed-off-by:
Linus Walleij <linus.walleij@linaro.org> Reviewed-by:
Simon Horman <horms@kernel.org> Link: https://lore.kernel.org/r/20240509-gemini-ethernet-locking-v1-1-afd00a528b95@linaro.orgSigned-off-by:
-
Dan Nowlin authored
Previously, the driver assumed that all signature segments would contain one or more buffers to download. In the future, there will be signature segments that will contain no buffers to download. Correct download flow to allow for signature segments that have zero download buffers and skip the download in this case. Fixes: 3cbdb034 ("ice: Add support for E830 DDP package segment") Reviewed-by:
Przemek Kitszel <przemyslaw.kitszel@intel.com> Signed-off-by:
Dan Nowlin <dan.nowlin@intel.com> Signed-off-by:
Paul Greenwalt <paul.greenwalt@intel.com> Reviewed-by:
Paul Menzel <pmenzel@molgen.mpg.de> Tested-by: Pucha Himasekhar Reddy <himasekharx.reddy.pucha@intel.com> (A Contingent worker at Intel) Signed-off-by:
Tony Nguyen <anthony.l.nguyen@intel.com> Link: https://lore.kernel.org/r/20240508171908.2760776-1-anthony.l.nguyen@intel.comSigned-off-by:
-
Daniel Golle authored
ADMAv2.0 is plagued by RX hangs which can't easily detected and happen upon receival of a corrupted Ethernet frame. Use ADMAv1 instead which is also still present and usable, and doesn't suffer from that problem. Fixes: 197c9e9b ("net: ethernet: mtk_eth_soc: introduce support for mt7986 chipset") Co-developed-by:
Lorenzo Bianconi <lorenzo@kernel.org> Signed-off-by:
Daniel Golle <daniel@makrotopia.org> Link: https://lore.kernel.org/r/57cef74bbd0c243366ad1ff4221e3f72f437ec80.1715164770.git.daniel@makrotopia.orgSigned-off-by:
-
Lorenzo Bianconi authored
Split tx and rx fields in mtk_soc_data struct. This is a preliminary patch to roll back to ADMAv1 for MT7986 and MT7981 SoC in order to fix a hw hang if the device receives a corrupted packet when using ADMAv2.0. Fixes: 197c9e9b ("net: ethernet: mtk_eth_soc: introduce support for mt7986 chipset") Signed-off-by:
-
Jakub Kicinski authored
The test seems to expect that nc will exit after the first received message. This is not the case with Ncat 7.94. There are multiple versions of nc out there, switch to socat for better compatibility. Tell socat to exit after 128 bytes and pad the message. Since the test sets -e make sure we don't set exit code (|| true) and print the pass / fail rather then silently moving over the test and just setting non-zero exit code with no output indicating what failed. Fixes: c08e8bae ("selftests: add amt interface selftest script") Acked-by: Paolo Abeni<pabeni@redhat.com> Tested-by:
Taehee Yoo <ap420073@gmail.com> Link: https://lore.kernel.org/r/20240509161952.3940476-1-kuba@kernel.orgSigned-off-by:
-
Jakub Kicinski authored
Test needs IPv6 multicast. smcroute currently crashes when trying to install a route in a kernel without IPv6 multicast. Fixes: c08e8bae ("selftests: add amt interface selftest script") Link: https://lore.kernel.org/r/20240509161919.3939966-1-kuba@kernel.orgSigned-off-by:
-
Jakub Kicinski authored
Erhard reports netpoll warnings from sungem: netpoll_send_skb_on_dev(): eth0 enabled interrupts in poll (gem_start_xmit+0x0/0x398) WARNING: CPU: 1 PID: 1 at net/core/netpoll.c:370 netpoll_send_skb+0x1fc/0x20c gem_poll_controller() disables interrupts, which may sleep. We can't sleep in netpoll, it has interrupts disabled completely. Strangely, gem_poll_controller() doesn't even poll the completions, and instead acts as if an interrupt has fired so it just schedules NAPI and exits. None of this has been necessary for years, since netpoll invokes NAPI directly. Fixes: fe09bb61 ("sungem: Spring cleaning and GRO support") Reported-and-tested-by:
Erhard Furtner <erhard_f@mailbox.org> Link: https://lore.kernel.org/all/20240428125306.2c3080ef@legionReviewed-by:
Eric Dumazet <edumazet@google.com> Link: https://lore.kernel.org/r/20240508134504.3560956-1-kuba@kernel.orgSigned-off-by:
-
- 10 May, 2024 1 commit
-
-
Vadim Fedorenko authored
In ptp_ocp driver pin actions assume sma_nr starts with 1, but for DPLL subsystem callback 0-based index was used. Fix it providing proper index. Fixes: 09eeb3ae ("ptp_ocp: implement DPLL ops") Signed-off-by:
Vadim Fedorenko <vadim.fedorenko@linux.dev> Link: https://lore.kernel.org/r/20240508132111.11545-1-vadim.fedorenko@linux.devSigned-off-by:
-
- 09 May, 2024 18 commits
-
-
git://git.kernel.org/pub/scm/linux/kernel/git/netdev/netLinus Torvalds authored
Pull networking fixes from Paolo Abeni: "Including fixes from bluetooth and IPsec. The bridge patch is actually a follow-up to a recent fix in the same area. We have a pending v6.8 AF_UNIX regression; it should be solved soon, but not in time for this PR. Current release - regressions: - eth: ks8851: Queue RX packets in IRQ handler instead of disabling BHs - net: bridge: fix corrupted ethernet header on multicast-to-unicast Current release - new code bugs: - xfrm: fix possible bad pointer derferencing in error path Previous releases - regressionis: - core: fix out-of-bounds access in ops_init - ipv6: - fix potential uninit-value access in __ip6_make_skb() - fib6_rules: avoid possible NULL dereference in fib6_rule_action() - tcp: use refcount_inc_not_zero() in tcp_twsk_unique(). - rtnetlink: correct nested IFLA_VF_VLAN_LIST attribute validation - rxrpc: fix congestion control algorithm - bluetooth: - l2cap: fix slab-use-after-free in l2cap_connect() - msft: fix slab-use-after-free in msft_do_close() - eth: hns3: fix kernel crash when devlink reload during initialization - eth: dsa: mv88e6xxx: add phylink_get_caps for the mv88e6320/21 family Previous releases - always broken: - xfrm: preserve vlan tags for transport mode software GRO - tcp: defer shutdown(SEND_SHUTDOWN) for TCP_SYN_RECV sockets - eth: hns3: keep using user config after hardware reset" * tag 'net-6.9-rc8' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net: (47 commits) net: dsa: mv88e6xxx: read cmode on mv88e6320/21 serdes only ports net: dsa: mv88e6xxx: add phylink_get_caps for the mv88e6320/21 family net: hns3: fix kernel crash when devlink reload during initialization net: hns3: fix port vlan filter not disabled issue net: hns3: use appropriate barrier function after setting a bit value net: hns3: release PTP resources if pf initialization failed net: hns3: change type of numa_node_mask as nodemask_t net: hns3: direct return when receive a unknown mailbox message net: hns3: using user configure after hardware reset net/smc: fix neighbour and rtable leak in smc_ib_find_route() ipv6: prevent NULL dereference in ip6_output() hsr: Simplify code for announcing HSR nodes timer setup ipv6: fib6_rules: avoid possible NULL dereference in fib6_rule_action() dt-bindings: net: mediatek: remove wrongly added clocks and SerDes rxrpc: Only transmit one ACK per jumbo packet received rxrpc: Fix congestion control algorithm selftests: test_bridge_neigh_suppress.sh: Fix failures due to duplicate MAC ipv6: Fix potential uninit-value access in __ip6_make_skb() net: phy: marvell-88q2xxx: add support for Rev B1 and B2 appletalk: Improve handling of broadcast packets ... -
git://git.kernel.org/pub/scm/linux/kernel/git/rmk/linuxLinus Torvalds authored
Pull ARM fix from Russell King: - clear stale KASan stack poison when a CPU resumes * tag 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/rmk/linux: ARM: 9381/1: kasan: clear stale stack poison
-
git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfsLinus Torvalds authored
Pull dentry leak fix from Al Viro: "Dentry leak fix in the qibfs driver that I forgot to send a pull request for ;-/ My apologies - it actually sat in vfs.git#fixes for more than two months..." * tag 'pull-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs: qibfs: fix dentry leak
-
Steffen Bätz authored
On the mv88e6320 and 6321 switch family, port 0/1 are serdes only ports. Modified the mv88e6352_get_port4_serdes_cmode function to pass a port number since the register set of the 6352 is equal on the 6320/21. Signed-off-by:
Steffen Bätz <steffen@innosonix.de> Reviewed-by:
Andrew Lunn <andrew@lunn.ch> Reviewed-by:
Fabio Estevam <festevam@gmail.com> Link: https://lore.kernel.org/r/20240508072944.54880-3-steffen@innosonix.deSigned-off-by:
Paolo Abeni <pabeni@redhat.com>
-
Steffen Bätz authored
As of commit de5c9bf4 ("net: phylink: require supported_interfaces to be filled") Marvell 88e6320/21 switches fail to be probed: ... mv88e6085 30be0000.ethernet-1:00: phylink: error: empty supported_interfaces error creating PHYLINK: -22 ... The problem stems from the use of mv88e6185_phylink_get_caps() to get the device capabilities. Since there are serdes only ports 0/1 included, create a new dedicated phylink_get_caps for the 6320 and 6321 to properly support their set of capabilities. Fixes: de5c9bf4 ("net: phylink: require supported_interfaces to be filled") Signed-off-by:
-
Paolo Abeni authored
Jijie Shao says: ==================== There are some bugfix for the HNS3 ethernet driver ==================== Link: https://lore.kernel.org/r/20240507134224.2646246-1-shaojijie@huawei.comSigned-off-by:
-
Yonglong Liu authored
The devlink reload process will access the hardware resources, but the register operation is done before the hardware is initialized. So, processing the devlink reload during initialization may lead to kernel crash. This patch fixes this by registering the devlink after hardware initialization. Fixes: cd624299 ("net: hns3: add support for registering devlink for VF") Fixes: 93305b77 ("net: hns3: fix kernel crash when devlink reload during pf initialization") Signed-off-by:
Yonglong Liu <liuyonglong@huawei.com> Signed-off-by:
Jijie Shao <shaojijie@huawei.com> Signed-off-by:
-
Yonglong Liu authored
According to hardware limitation, for device support modify VLAN filter state but not support bypass port VLAN filter, it should always disable the port VLAN filter. but the driver enables port VLAN filter when initializing, if there is no VLAN(except VLAN 0) id added, the driver will disable it in service task. In most time, it works fine. But there is a time window before the service task shceduled and net device being registered. So if user adds VLAN at this time, the driver will not update the VLAN filter state, and the port VLAN filter remains enabled. To fix the problem, if support modify VLAN filter state but not support bypass port VLAN filter, set the port vlan filter to "off". Fixes: 184cd221 ("net: hns3: disable port VLAN filter when support function level VLAN filter control") Fixes: 2ba30662 ("net: hns3: add support for modify VLAN filter state") Signed-off-by:
-
Peiyang Wang authored
There is a memory barrier in followed case. When set the port down, hclgevf_set_timmer will set DOWN in state. Meanwhile, the service task has different behaviour based on whether the state is DOWN. Thus, to make sure service task see DOWN, use smp_mb__after_atomic after calling set_bit(). CPU0 CPU1 ========================== =================================== hclgevf_set_timer_task() hclgevf_periodic_service_task() set_bit(DOWN,state) test_bit(DOWN,state) pf also has this issue. Fixes: ff200099 ("net: hns3: remove unnecessary work in hclgevf_main") Fixes: 1c6dfe6f ("net: hns3: remove mailbox and reset work in hclge_main") Signed-off-by:Peiyang Wang <wangpeiyang1@huawei.com> Signed-off-by:
-
Peiyang Wang authored
During the PF initialization process, hclge_update_port_info may return an error code for some reason. At this point, the ptp initialization has been completed. To void memory leaks, the resources that are applied by ptp should be released. Therefore, when hclge_update_port_info returns an error code, hclge_ptp_uninit is called to release the corresponding resources. Fixes: eaf83ae5 ("net: hns3: add querying fec ability from firmware") Signed-off-by:
Hariprasad Kelam <hkelam@marvell.com> Signed-off-by:
-
Peiyang Wang authored
It provides nodemask_t to describe the numa node mask in kernel. To improve transportability, change the type of numa_node_mask as nodemask_t. Fixes: 38caee9d ("net: hns3: Add support of the HNAE3 framework") Signed-off-by:
-
Jian Shen authored
Currently, the driver didn't return when receive a unknown mailbox message, and continue checking whether need to generate a response. It's unnecessary and may be incorrect. Fixes: bb5790b7 ("net: hns3: refactor mailbox response scheme between PF and VF") Signed-off-by:
Jian Shen <shenjian15@huawei.com> Signed-off-by:
-
Peiyang Wang authored
When a reset occurring, it's supposed to recover user's configuration. Currently, the port info(speed, duplex and autoneg) is stored in hclge_mac and will be scheduled updated. Consider the case that reset was happened consecutively. During the first reset, the port info is configured with a temporary value cause the PHY is reset and looking for best link config. Second reset start and use pervious configuration which is not the user's. The specific process is as follows: +------+ +----+ +----+ | USER | | PF | | HW | +---+--+ +-+--+ +-+--+ | ethtool --reset | | +------------------->| reset command | | ethtool --reset +-------------------->| +------------------->| +---+ | +---+ | | | | |reset currently | | HW RESET | | |and wait to do | | | |<--+ | | | | send pervious cfg |<--+ | | (1000M FULL AN_ON) | | +-------------------->| | | read cfg(time task) | | | (10M HALF AN_OFF) +---+ | |<--------------------+ | cfg take effect | | reset command |<--+ | +-------------------->| | | +---+ | | send pervious cfg | | HW RESET | | (10M HALF AN_OFF) |<--+ | +-------------------->| | | read cfg(time task) | | | (10M HALF AN_OFF) +---+ | |<--------------------+ | cfg take effect | | | | | | read cfg(time task) |<--+ | | (10M HALF AN_OFF) | | |<--------------------+ | | | v v v To avoid aboved situation, this patch introduced req_speed, req_duplex, req_autoneg to store user's configuration and it only be used after hardware reset and to recover user's configuration Fixes: f5f2b3e4 ("net: hns3: add support for imp-controlled PHYs") Signed-off-by: -
Wen Gu authored
In smc_ib_find_route(), the neighbour found by neigh_lookup() and rtable resolved by ip_route_output_flow() are not released or put before return. It may cause the refcount leak, so fix it. Link: https://lore.kernel.org/r/20240506015439.108739-1-guwen@linux.alibaba.com Fixes: e5c4744c ("net/smc: add SMC-Rv2 connection establishment") Signed-off-by:
Wen Gu <guwen@linux.alibaba.com> Link: https://lore.kernel.org/r/20240507125331.2808-1-guwen@linux.alibaba.comSigned-off-by:
-
Eric Dumazet authored
According to syzbot, there is a chance that ip6_dst_idev() returns NULL in ip6_output(). Most places in IPv6 stack deal with a NULL idev just fine, but not here. syzbot reported: general protection fault, probably for non-canonical address 0xdffffc00000000bc: 0000 [#1] PREEMPT SMP KASAN PTI KASAN: null-ptr-deref in range [0x00000000000005e0-0x00000000000005e7] CPU: 0 PID: 9775 Comm: syz-executor.4 Not tainted 6.9.0-rc5-syzkaller-00157-g6a30653b #0 Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 03/27/2024 RIP: 0010:ip6_output+0x231/0x3f0 net/ipv6/ip6_output.c:237 Code: 3c 1e 00 49 89 df 74 08 4c 89 ef e8 19 58 db f7 48 8b 44 24 20 49 89 45 00 49 89 c5 48 8d 9d e0 05 00 00 48 89 d8 48 c1 e8 03 <42> 0f b6 04 38 84 c0 4c 8b 74 24 28 0f 85 61 01 00 00 8b 1b 31 ff RSP: 0018:ffffc9000927f0d8 EFLAGS: 00010202 RAX: 00000000000000bc RBX: 00000000000005e0 RCX: 0000000000040000 RDX: ffffc900131f9000 RSI: 0000000000004f47 RDI: 0000000000004f48 RBP: 0000000000000000 R08: ffffffff8a1f0b9a R09: 1ffffffff1f51fad R10: dffffc0000000000 R11: fffffbfff1f51fae R12: ffff8880293ec8c0 R13: ffff88805d7fc000 R14: 1ffff1100527d91a R15: dffffc0000000000 FS: 00007f135c6856c0(0000) GS:ffff8880b9400000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 0000000020000080 CR3: 0000000064096000 CR4: 00000000003506f0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 Call Trace: <TASK> NF_HOOK include/linux/netfilter.h:314 [inline] ip6_xmit+0xefe/0x17f0 net/ipv6/ip6_output.c:358 sctp_v6_xmit+0x9f2/0x13f0 net/sctp/ipv6.c:248 sctp_packet_transmit+0x26ad/0x2ca0 net/sctp/output.c:653 sctp_packet_singleton+0x22c/0x320 net/sctp/outqueue.c:783 sctp_outq_flush_ctrl net/sctp/outqueue.c:914 [inline] sctp_outq_flush+0x6d5/0x3e20 net/sctp/outqueue.c:1212 sctp_side_effects net/sctp/sm_sideeffect.c:1198 [inline] sctp_do_sm+0x59cc/0x60c0 net/sctp/sm_sideeffect.c:1169 sctp_primitive_ASSOCIATE+0x95/0xc0 net/sctp/primitive.c:73 __sctp_connect+0x9cd/0xe30 net/sctp/socket.c:1234 sctp_connect net/sctp/socket.c:4819 [inline] sctp_inet_connect+0x149/0x1f0 net/sctp/socket.c:4834 __sys_connect_file net/socket.c:2048 [inline] __sys_connect+0x2df/0x310 net/socket.c:2065 __do_sys_connect net/socket.c:2075 [inline] __se_sys_connect net/socket.c:2072 [inline] __x64_sys_connect+0x7a/0x90 net/socket.c:2072 do_syscall_x64 arch/x86/entry/common.c:52 [inline] do_syscall_64+0xf5/0x240 arch/x86/entry/common.c:83 entry_SYSCALL_64_after_hwframe+0x77/0x7f Fixes: 778d80be ("ipv6: Add disable_ipv6 sysctl to disable IPv6 operaion on specific interface.") Reported-by:
syzbot <syzkaller@googlegroups.com> Signed-off-by:
Larysa Zaremba <larysa.zaremba@intel.com> Link: https://lore.kernel.org/r/20240507161842.773961-1-edumazet@google.comSigned-off-by:
-
Lukasz Majewski authored
Up till now the code to start HSR announce timer, which triggers sending supervisory frames, was assuming that hsr_netdev_notify() would be called at least twice for hsrX interface. This was required to have different values for old and current values of network device's operstate. This is problematic for a case where hsrX interface is already in the operational state when hsr_netdev_notify() is called, so timer is not configured to trigger and as a result the hsrX is not sending supervisory frames to HSR ring. This error has been discovered when hsr_ping.sh script was run. To be more specific - for the hsr1 and hsr2 the hsr_netdev_notify() was called at least twice with different IF_OPER_{LOWERDOWN|DOWN|UP} states assigned in hsr_check_carrier_and_operstate(hsr). As a result there was no issue with sending supervisory frames. However, with hsr3, the notify function was called only once with operstate set to IF_OPER_UP and timer responsible for triggering supervisory frames was not fired. The solution is to use netif_oper_up() and netif_running() helper functions to assess if network hsrX device is up. Only then, when the timer is not already pending, it is started. Otherwise it is deactivated. Fixes: f421436a ("net/hsr: Add support for the High-availability Seamless Redundancy protocol (HSRv0)") Signed-off-by:Lukasz Majewski <lukma@denx.de> Reviewed-by:
-
Eric Dumazet authored
syzbot is able to trigger the following crash [1], caused by unsafe ip6_dst_idev() use. Indeed ip6_dst_idev() can return NULL, and must always be checked. [1] Oops: general protection fault, probably for non-canonical address 0xdffffc0000000000: 0000 [#1] PREEMPT SMP KASAN PTI KASAN: null-ptr-deref in range [0x0000000000000000-0x0000000000000007] CPU: 0 PID: 31648 Comm: syz-executor.0 Not tainted 6.9.0-rc4-next-20240417-syzkaller #0 Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 03/27/2024 RIP: 0010:__fib6_rule_action net/ipv6/fib6_rules.c:237 [inline] RIP: 0010:fib6_rule_action+0x241/0x7b0 net/ipv6/fib6_rules.c:267 Code: 02 00 00 49 8d 9f d8 00 00 00 48 89 d8 48 c1 e8 03 42 80 3c 20 00 74 08 48 89 df e8 f9 32 bf f7 48 8b 1b 48 89 d8 48 c1 e8 03 <42> 80 3c 20 00 74 08 48 89 df e8 e0 32 bf f7 4c 8b 03 48 89 ef 4c RSP: 0018:ffffc9000fc1f2f0 EFLAGS: 00010246 RAX: 0000000000000000 RBX: 0000000000000000 RCX: 1a772f98c8186700 RDX: 0000000000000003 RSI: ffffffff8bcac4e0 RDI: ffffffff8c1f9760 RBP: ffff8880673fb980 R08: ffffffff8fac15ef R09: 1ffffffff1f582bd R10: dffffc0000000000 R11: fffffbfff1f582be R12: dffffc0000000000 R13: 0000000000000080 R14: ffff888076509000 R15: ffff88807a029a00 FS: 00007f55e82ca6c0(0000) GS:ffff8880b9400000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 0000001b31d23000 CR3: 0000000022b66000 CR4: 00000000003506f0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 Call Trace: <TASK> fib_rules_lookup+0x62c/0xdb0 net/core/fib_rules.c:317 fib6_rule_lookup+0x1fd/0x790 net/ipv6/fib6_rules.c:108 ip6_route_output_flags_noref net/ipv6/route.c:2637 [inline] ip6_route_output_flags+0x38e/0x610 net/ipv6/route.c:2649 ip6_route_output include/net/ip6_route.h:93 [inline] ip6_dst_lookup_tail+0x189/0x11a0 net/ipv6/ip6_output.c:1120 ip6_dst_lookup_flow+0xb9/0x180 net/ipv6/ip6_output.c:1250 sctp_v6_get_dst+0x792/0x1e20 net/sctp/ipv6.c:326 sctp_transport_route+0x12c/0x2e0 net/sctp/transport.c:455 sctp_assoc_add_peer+0x614/0x15c0 net/sctp/associola.c:662 sctp_connect_new_asoc+0x31d/0x6c0 net/sctp/socket.c:1099 __sctp_connect+0x66d/0xe30 net/sctp/socket.c:1197 sctp_connect net/sctp/socket.c:4819 [inline] sctp_inet_connect+0x149/0x1f0 net/sctp/socket.c:4834 __sys_connect_file net/socket.c:2048 [inline] __sys_connect+0x2df/0x310 net/socket.c:2065 __do_sys_connect net/socket.c:2075 [inline] __se_sys_connect net/socket.c:2072 [inline] __x64_sys_connect+0x7a/0x90 net/socket.c:2072 do_syscall_x64 arch/x86/entry/common.c:52 [inline] do_syscall_64+0xf5/0x240 arch/x86/entry/common.c:83 entry_SYSCALL_64_after_hwframe+0x77/0x7f Fixes: 5e5f3f0f ("[IPV6] ADDRCONF: Convert ipv6_get_saddr() to ipv6_dev_get_saddr().") Signed-off-by:
David Ahern <dsahern@kernel.org> Link: https://lore.kernel.org/r/20240507163145.835254-1-edumazet@google.comSigned-off-by:
-
Daniel Golle authored
Several clocks as well as both sgmiisys phandles were added by mistake to the Ethernet bindings for MT7988. Also, the total number of clocks didn't match with the actual number of items listed. This happened because the vendor driver which served as a reference uses a high number of syscon phandles to access various parts of the SoC which wasn't acceptable upstream. Hence several parts which have never previously been supported (such SerDes PHY and USXGMII PCS) are going to be implemented by separate drivers. As a result the device tree will look much more sane. Quickly align the bindings with the upcoming reality of the drivers actually adding support for the remaining Ethernet-related features of the MT7988 SoC. Fixes: c94a9aab ("dt-bindings: net: mediatek,net: add mt7988-eth binding") Signed-off-by:
Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org> Link: https://lore.kernel.org/r/1569290b21cc787a424469ed74456a7e976b102d.1715084326.git.daniel@makrotopia.orgSigned-off-by:
-
- 08 May, 2024 12 commits
-
-
git://git.samba.org/ksmbdLinus Torvalds authored
Pull smb server fixes from Steve French: "Five ksmbd server fixes, all also for stable - Three fixes related to SMB3 leases (fixes two xfstests, and a locking issue) - Unitialized variable fix - Socket creation fix when bindv6only is set" * tag '6.9-rc7-ksmbd-fixes' of git://git.samba.org/ksmbd: ksmbd: do not grant v2 lease if parent lease key and epoch are not set ksmbd: use rwsem instead of rwlock for lease break ksmbd: avoid to send duplicate lease break notifications ksmbd: off ipv6only for both ipv4/ipv6 binding ksmbd: fix uninitialized symbol 'share' in smb2_tree_connect() -
git://git.kernel.org/pub/scm/linux/kernel/git/mszeredi/fuseLinus Torvalds authored
Pull fuse fixes from Miklos Szeredi: "Two one-liner fixes for issues introduced in -rc1" * tag 'fuse-fixes-6.9-final' of git://git.kernel.org/pub/scm/linux/kernel/git/mszeredi/fuse: virtiofs: include a newline in sysfs tag fuse: verify zero padding in fuse_backing_map
-
git://git.kernel.org/pub/scm/linux/kernel/git/linkinjeon/exfatLinus Torvalds authored
Pull exfat fixes from Namjae Jeon: - Fix xfstests generic/013 test failure with dirsync mount option - Initialize the reserved fields of deleted file and stream extension dentries to zero * tag 'exfat-for-6.9-rc8' of git://git.kernel.org/pub/scm/linux/kernel/git/linkinjeon/exfat: exfat: zero the reserved fields of file and stream extension dentries exfat: fix timing of synchronizing bitmap and inode
-
https://evilpiepirate.org/git/bcachefsLinus Torvalds authored
Pull bcachefs fixes from Kent Overstreet: - Various syzbot fixes; mainly small gaps in validation - Fix an integer overflow in fiemap() which was preventing filefrag from returning the full list of extents - Fix a refcounting bug on the device refcount, turned up by new assertions in the development branch - Fix a device removal/readd bug; write_super() was repeatedly dropping and retaking bch_dev->io_ref references * tag 'bcachefs-2024-05-07.2' of https://evilpiepirate.org/git/bcachefs: bcachefs: Add missing sched_annotate_sleep() in bch2_journal_flush_seq_async() bcachefs: Fix race in bch2_write_super() bcachefs: BCH_SB_LAYOUT_SIZE_BITS_MAX bcachefs: Add missing skcipher_request_set_callback() call bcachefs: Fix snapshot_t() usage in bch2_fs_quota_read_inode() bcachefs: Fix shift-by-64 in bformat_needs_redo() bcachefs: Guard against unknown k.k->type in __bkey_invalid() bcachefs: Add missing validation for superblock section clean bcachefs: Fix assert in bch2_alloc_v4_invalid() bcachefs: fix overflow in fiemap bcachefs: Add a better limit for maximum number of buckets bcachefs: Fix lifetime issue in device iterator helpers bcachefs: Fix bch2_dev_lookup() refcounting bcachefs: Initialize bch_write_op->failed in inline data path bcachefs: Fix refcount put in sb_field_resize error path bcachefs: Inodes need extra padding for varint_decode_fast() bcachefs: Fix early error path in bch2_fs_btree_key_cache_exit() bcachefs: bucket_pos_to_bp_noerror() bcachefs: don't free error pointers bcachefs: Fix a scheduler splat in __bch2_next_write_buffer_flush_journal_buf()
-
git://git.kernel.org/pub/scm/linux/kernel/git/soc/socLinus Torvalds authored
Pull ARM SoC fixes from Arnd Bergmann: "These are a couple of last minute fixes that came in over the previous week, addressing: - A pin configuration bug on a qualcomm board that caused issues with ethernet and mmc - Two minor code fixes for misleading console output in the microchip firmware driver - A build warning in the sifive cache driver" * tag 'soc-fixes-6.9-3' of git://git.kernel.org/pub/scm/linux/kernel/git/soc/soc: firmware: microchip: clarify that sizes and addresses are in hex firmware: microchip: don't unconditionally print validation success arm64: dts: qcom: sa8155p-adp: fix SDHC2 CD pin configuration cache: sifive_ccache: Silence unused variable warning -
git://git.kernel.org/pub/scm/linux/kernel/git/pci/pciLinus Torvalds authored
Pull pci fixes from Bjorn Helgaas: - Update kernel-parameters doc to describe "pcie_aspm=off" more accurately (Bjorn Helgaas) - Restore the parent's (not the child's) ASPM state to the parent during resume, which fixes a reboot during resume (Kai-Heng Feng) * tag 'pci-v6.9-fixes-2' of git://git.kernel.org/pub/scm/linux/kernel/git/pci/pci: PCI/ASPM: Restore parent state to parent, child state to child PCI/ASPM: Clarify that pcie_aspm=off means leave ASPM untouched
-
Jakub Kicinski authored
David Howells says: ==================== rxrpc: Miscellaneous fixes (part) Here some miscellaneous fixes for AF_RXRPC: (1) Fix the congestion control algorithm to start cwnd at 4 and to not cut ssthresh when the peer cuts its rwind size. (2) Only transmit a single ACK for all the DATA packets glued together into a jumbo packet to reduce the number of ACKs being generated. ==================== Link: https://lore.kernel.org/r/20240503150749.1001323-1-dhowells@redhat.comSigned-off-by: -
David Howells authored
Only generate one ACK packet for all the subpackets in a jumbo packet. If we would like to generate more than one ACK, we prioritise them base on their reason code, in the order, highest first: OutOfSeq > NoSpace > ExceedsWin > Duplicate > Requested > Delay > Idle For the first four, we reference the lowest offending subpacket; for the last three, the highest. This reduces the number of ACKs we end up transmitting to one per UDP packet transmitted to reduce network loading and packet parsing. Fixes: 5d7edbc9 ("rxrpc: Get rid of the Rx ring") Signed-off-by:
David Howells <dhowells@redhat.com> cc: Marc Dionne <marc.dionne@auristor.com> cc: linux-afs@lists.infradead.org Reviewed-by:
Jeffrey Altman <jaltman@auristor.com <mailto:jaltman@auristor.com>> Link: https://lore.kernel.org/r/20240503150749.1001323-3-dhowells@redhat.comSigned-off-by:
-
David Howells authored
Make the following fixes to the congestion control algorithm: (1) Don't vary the cwnd starting value by the size of RXRPC_TX_SMSS since that's currently held constant - set to the size of a jumbo subpacket payload so that we can create jumbo packets on the fly. The current code invariably picks 3 as the starting value. Further, the starting cwnd needs to be an even number because we ack every other packet, so set it to 4. (2) Don't cut ssthresh when we see an ACK come from the peer with a receive window (rwind) less than ssthresh. ssthresh keeps track of characteristics of the connection whereas rwind may be reduced by the peer for any reason - and may be reduced to 0. Fixes: 1fc4fa2a ("rxrpc: Fix congestion management") Fixes: 08511150 ("rxrpc: Reduce ssthresh to peer's receive window") Signed-off-by:Simon Wilkinson <sxw@auristor.com> cc: Marc Dionne <marc.dionne@auristor.com> cc: linux-afs@lists.infradead.org Reviewed-by:
-
Ido Schimmel authored
When creating the topology for the test, three veth pairs are created in the initial network namespace before being moved to one of the network namespaces created by the test. On systems where systemd-udev uses MACAddressPolicy=persistent (default since systemd version 242), this will result in some net devices having the same MAC address since they were created with the same name in the initial network namespace. In turn, this leads to arping / ndisc6 failing since packets are dropped by the bridge's loopback filter. Fix by creating each net device in the correct network namespace instead of moving it there from the initial network namespace. Reported-by:
Ido Schimmel <idosch@nvidia.com> Link: https://lore.kernel.org/r/20240507113033.1732534-1-idosch@nvidia.comSigned-off-by:
-
Shigeru Yoshida authored
As it was done in commit fc1092f5 ("ipv4: Fix uninit-value access in __ip_make_skb()") for IPv4, check FLOWI_FLAG_KNOWN_NH on fl6->flowi6_flags instead of testing HDRINCL on the socket to avoid a race condition which causes uninit-value access. Fixes: ea30388b ("ipv6: Fix an uninit variable access bug in __ip6_make_skb()") Signed-off-by:
Shigeru Yoshida <syoshida@redhat.com> Signed-off-by:
David S. Miller <davem@davemloft.net>
-
Gregor Herburger authored
Different revisions of the Marvell 88q2xxx phy needs different init sequences. Add init sequence for Rev B1 and Rev B2. Rev B2 init sequence skips one register write. Tested-by:
Dimitri Fedrau <dima.fedrau@gmail.com> Signed-off-by:
Gregor Herburger <gregor.herburger@ew.tq-group.com> Signed-off-by:
-
