kernel/drivers/net/ethernet/microchip/lan966x, branch linux-6.2.y

net: lan966x: Fix port police support using tc-matchall

2023-03-17T07:57:53Z

[ Upstream commit 81563d8548b0478075c720666be348d4199b8591 ] When the police was removed from the port, then it was trying to remove the police from the police id and not from the actual police index. The police id represents the id of the police and police index represents the position in HW where the police is situated. The port police id can be any number while the port police index is a number based on the port chip port. Fix this by deleting the police from HW that is situated at the police index and not police id. Fixes: 5390334b59a3 ("net: lan966x: Add port police support using tc-matchall") Signed-off-by: Horatiu Vultur Reviewed-by: Simon Horman Reviewed-by: Vladimir Oltean Signed-off-by: David S. Miller Signed-off-by: Sasha Levin

net: lan966x: Fix possible deadlock inside PTP

2023-03-10T08:28:24Z

[ Upstream commit 3a70e0d4c9d74cb00f7c0ec022f5599f9f7ba07d ] When doing timestamping in lan966x and having PROVE_LOCKING enabled the following warning is shown. ======================================================== WARNING: possible irq lock inversion dependency detected 6.2.0-rc7-01749-gc54e1f7f7e36 #2786 Tainted: G N -------------------------------------------------------- swapper/0/0 just changed the state of lock: c2609f50 (_xmit_ETHER#2){+.-.}-{2:2}, at: sch_direct_xmit+0x16c/0x2e8 but this lock took another, SOFTIRQ-unsafe lock in the past: (&lan966x->ptp_ts_id_lock){+.+.}-{2:2} and interrupts could create inverse lock ordering between them. other info that might help us debug this: Possible interrupt unsafe locking scenario: CPU0 CPU1 ---- ---- lock(&lan966x->ptp_ts_id_lock); local_irq_disable(); lock(_xmit_ETHER#2); lock(&lan966x->ptp_ts_id_lock); lock(_xmit_ETHER#2); *** DEADLOCK *** 5 locks held by swapper/0/0: #0: c1001e18 ((&ndev->rs_timer)){+.-.}-{0:0}, at: call_timer_fn+0x0/0x33c #1: c105e7c4 (rcu_read_lock){....}-{1:2}, at: ndisc_send_skb+0x134/0x81c #2: c105e7d8 (rcu_read_lock_bh){....}-{1:2}, at: ip6_finish_output2+0x17c/0xc64 #3: c105e7d8 (rcu_read_lock_bh){....}-{1:2}, at: __dev_queue_xmit+0x4c/0x1224 #4: c3056174 (dev->qdisc_tx_busylock ?: &qdisc_tx_busylock){+...}-{2:2}, at: __dev_queue_xmit+0x354/0x1224 the shortest dependencies between 2nd lock and 1st lock: -> (&lan966x->ptp_ts_id_lock){+.+.}-{2:2} { HARDIRQ-ON-W at: lock_acquire.part.0+0xb0/0x248 _raw_spin_lock+0x38/0x48 lan966x_ptp_irq_handler+0x164/0x2a8 irq_thread_fn+0x1c/0x78 irq_thread+0x130/0x278 kthread+0xec/0x110 ret_from_fork+0x14/0x28 SOFTIRQ-ON-W at: lock_acquire.part.0+0xb0/0x248 _raw_spin_lock+0x38/0x48 lan966x_ptp_irq_handler+0x164/0x2a8 irq_thread_fn+0x1c/0x78 irq_thread+0x130/0x278 kthread+0xec/0x110 ret_from_fork+0x14/0x28 INITIAL USE at: lock_acquire.part.0+0xb0/0x248 _raw_spin_lock_irqsave+0x4c/0x68 lan966x_ptp_txtstamp_request+0x128/0x1cc lan966x_port_xmit+0x224/0x43c dev_hard_start_xmit+0xa8/0x2f0 sch_direct_xmit+0x108/0x2e8 __dev_queue_xmit+0x41c/0x1224 packet_sendmsg+0xdb4/0x134c __sys_sendto+0xd0/0x154 sys_send+0x18/0x20 ret_fast_syscall+0x0/0x1c } ... key at: [] __key.2+0x0/0x8 ... acquired at: _raw_spin_lock_irqsave+0x4c/0x68 lan966x_ptp_txtstamp_request+0x128/0x1cc lan966x_port_xmit+0x224/0x43c dev_hard_start_xmit+0xa8/0x2f0 sch_direct_xmit+0x108/0x2e8 __dev_queue_xmit+0x41c/0x1224 packet_sendmsg+0xdb4/0x134c __sys_sendto+0xd0/0x154 sys_send+0x18/0x20 ret_fast_syscall+0x0/0x1c -> (_xmit_ETHER#2){+.-.}-{2:2} { HARDIRQ-ON-W at: lock_acquire.part.0+0xb0/0x248 _raw_spin_lock+0x38/0x48 netif_freeze_queues+0x38/0x68 dev_deactivate_many+0xac/0x388 dev_deactivate+0x38/0x6c linkwatch_do_dev+0x70/0x8c __linkwatch_run_queue+0xd4/0x1e8 linkwatch_event+0x24/0x34 process_one_work+0x284/0x744 worker_thread+0x28/0x4bc kthread+0xec/0x110 ret_from_fork+0x14/0x28 IN-SOFTIRQ-W at: lock_acquire.part.0+0xb0/0x248 _raw_spin_lock+0x38/0x48 sch_direct_xmit+0x16c/0x2e8 __dev_queue_xmit+0x41c/0x1224 ip6_finish_output2+0x5f4/0xc64 ndisc_send_skb+0x4cc/0x81c addrconf_rs_timer+0xb0/0x2f8 call_timer_fn+0xb4/0x33c expire_timers+0xb4/0x10c run_timer_softirq+0xf8/0x2a8 __do_softirq+0xd4/0x5fc __irq_exit_rcu+0x138/0x17c irq_exit+0x8/0x28 __irq_svc+0x90/0xbc arch_cpu_idle+0x30/0x3c default_idle_call+0x44/0xac do_idle+0xc8/0x138 cpu_startup_entry+0x18/0x1c rest_init+0xcc/0x168 arch_post_acpi_subsys_init+0x0/0x8 INITIAL USE at: lock_acquire.part.0+0xb0/0x248 _raw_spin_lock+0x38/0x48 netif_freeze_queues+0x38/0x68 dev_deactivate_many+0xac/0x388 dev_deactivate+0x38/0x6c linkwatch_do_dev+0x70/0x8c __linkwatch_run_queue+0xd4/0x1e8 linkwatch_event+0x24/0x34 process_one_work+0x284/0x744 worker_thread+0x28/0x4bc kthread+0xec/0x110 ret_from_fork+0x14/0x28 } ... key at: [] netdev_xmit_lock_key+0x8/0x1c8 ... acquired at: __lock_acquire+0x978/0x2978 lock_acquire.part.0+0xb0/0x248 _raw_spin_lock+0x38/0x48 sch_direct_xmit+0x16c/0x2e8 __dev_queue_xmit+0x41c/0x1224 ip6_finish_output2+0x5f4/0xc64 ndisc_send_skb+0x4cc/0x81c addrconf_rs_timer+0xb0/0x2f8 call_timer_fn+0xb4/0x33c expire_timers+0xb4/0x10c run_timer_softirq+0xf8/0x2a8 __do_softirq+0xd4/0x5fc __irq_exit_rcu+0x138/0x17c irq_exit+0x8/0x28 __irq_svc+0x90/0xbc arch_cpu_idle+0x30/0x3c default_idle_call+0x44/0xac do_idle+0xc8/0x138 cpu_startup_entry+0x18/0x1c rest_init+0xcc/0x168 arch_post_acpi_subsys_init+0x0/0x8 stack backtrace: CPU: 0 PID: 0 Comm: swapper/0 Tainted: G N 6.2.0-rc7-01749-gc54e1f7f7e36 #2786 Hardware name: Generic DT based system unwind_backtrace from show_stack+0x10/0x14 show_stack from dump_stack_lvl+0x58/0x70 dump_stack_lvl from mark_lock.part.0+0x59c/0x93c mark_lock.part.0 from __lock_acquire+0x978/0x2978 __lock_acquire from lock_acquire.part.0+0xb0/0x248 lock_acquire.part.0 from _raw_spin_lock+0x38/0x48 _raw_spin_lock from sch_direct_xmit+0x16c/0x2e8 sch_direct_xmit from __dev_queue_xmit+0x41c/0x1224 __dev_queue_xmit from ip6_finish_output2+0x5f4/0xc64 ip6_finish_output2 from ndisc_send_skb+0x4cc/0x81c ndisc_send_skb from addrconf_rs_timer+0xb0/0x2f8 addrconf_rs_timer from call_timer_fn+0xb4/0x33c call_timer_fn from expire_timers+0xb4/0x10c expire_timers from run_timer_softirq+0xf8/0x2a8 run_timer_softirq from __do_softirq+0xd4/0x5fc __do_softirq from __irq_exit_rcu+0x138/0x17c __irq_exit_rcu from irq_exit+0x8/0x28 irq_exit from __irq_svc+0x90/0xbc Exception stack(0xc1001f20 to 0xc1001f68) 1f20: ffffffff ffffffff 00000001 c011f840 c100e000 c100e000 c1009314 c1009370 1f40: c10f0c1a c0d5e564 c0f5da8c 00000000 00000000 c1001f70 c010f0bc c010f0c0 1f60: 600f0013 ffffffff __irq_svc from arch_cpu_idle+0x30/0x3c arch_cpu_idle from default_idle_call+0x44/0xac default_idle_call from do_idle+0xc8/0x138 do_idle from cpu_startup_entry+0x18/0x1c cpu_startup_entry from rest_init+0xcc/0x168 rest_init from arch_post_acpi_subsys_init+0x0/0x8 Fix this by using spin_lock_irqsave/spin_lock_irqrestore also inside lan966x_ptp_irq_handler. Fixes: e85a96e48e33 ("net: lan966x: Add support for ptp interrupts") Signed-off-by: Horatiu Vultur Link: https://lore.kernel.org/r/20230217210917.2649365-1-horatiu.vultur@microchip.com Signed-off-by: Jakub Kicinski Signed-off-by: Sasha Levin

lan966x: execute xdp_do_flush() before napi_complete_done()

2023-01-28T06:27:59Z

Make sure that xdp_do_flush() is always executed before napi_complete_done(). This is important for two reasons. First, a redirect to an XSKMAP assumes that a call to xdp_do_redirect() from napi context X on CPU Y will be followed by a xdp_do_flush() from the same napi context and CPU. This is not guaranteed if the napi_complete_done() is executed before xdp_do_flush(), as it tells the napi logic that it is fine to schedule napi context X on another CPU. Details from a production system triggering this bug using the veth driver can be found following the first link below. The second reason is that the XDP_REDIRECT logic in itself relies on being inside a single NAPI instance through to the xdp_do_flush() call for RCU protection of all in-kernel data structures. Details can be found in the second link below. Fixes: a825b611c7c1 ("net: lan966x: Add support for XDP_REDIRECT") Signed-off-by: Magnus Karlsson Acked-by: Toke Høiland-Jørgensen Acked-by: Steen Hegelund Link: https://lore.kernel.org/r/20221220185903.1105011-1-sbohrer@cloudflare.com Link: https://lore.kernel.org/all/20210624160609.292325-1-toke@redhat.com/ Signed-off-by: Jakub Kicinski

net: lan966x: add missing fwnode_handle_put() for ports node

2023-01-14T05:53:28Z

Since the "ethernet-ports" node is retrieved using device_get_named_child_node(), it should be release after using it. Add missing fwnode_handle_put() and move the code that retrieved the node from device-tree to avoid complicated handling in case of error. Fixes: db8bcaad5393 ("net: lan966x: add the basic lan966x driver") Signed-off-by: Clément Léger Reviewed-by: Alexander Duyck Link: https://lore.kernel.org/r/20230112161311.495124-1-clement.leger@bootlin.com Signed-off-by: Jakub Kicinski

net: lan966x: check for ptp to be enabled in lan966x_ptp_deinit()

2023-01-11T08:31:36Z

If ptp was not enabled due to missing IRQ for instance, lan966x_ptp_deinit() will dereference NULL pointers. Fixes: d096459494a8 ("net: lan966x: Add support for ptp clocks") Signed-off-by: Clément Léger Reviewed-by: Horatiu Vultur Reviewed-by: Jiri Pirko Signed-off-by: David S. Miller

net: lan966x: Allow to add rules in TCAM even if not enabled

2023-01-09T08:11:44Z

The blamed commit implemented the vcap_operations to allow to add an entry in the TCAM. One of the callbacks is to validate the supported keysets. If the TCAM lookup was not enabled, then this will return failure so no entries could be added. This doesn't make much sense, as you can enable at a later point the TCAM. Therefore change it such to allow entries in TCAM even it is not enabled. Fixes: 4426b78c626d ("net: lan966x: Add port keyset config and callback interface") Signed-off-by: Horatiu Vultur Signed-off-by: David S. Miller

net: lan966x: Fix configuration of the PCS

2022-12-22T11:21:05Z

When the PCS was taken out of reset, we were changing by mistake also the speed to 100 Mbit. But in case the link was going down, the link up routine was setting correctly the link speed. If the link was not getting down then the speed was forced to run at 100 even if the speed was something else. On lan966x, to set the speed link to 1G or 2.5G a value of 1 needs to be written in DEV_CLOCK_CFG_LINK_SPEED. This is similar to the procedure in lan966x_port_init. The issue was reproduced using 1000base-x sfp module using the commands: ip link set dev eth2 up ip link addr add 10.97.10.2/24 dev eth2 ethtool -s eth2 speed 1000 autoneg off Fixes: d28d6d2e37d1 ("net: lan966x: add port module support") Signed-off-by: Horatiu Vultur Reviewed-by: Piotr Raczynski Link: https://lore.kernel.org/r/20221221093315.939133-1-horatiu.vultur@microchip.com Signed-off-by: Paolo Abeni

net: lan966x: Remove a useless test in lan966x_ptp_add_trap()

2022-12-13T03:39:24Z

vcap_alloc_rule() can't return NULL. So remove some dead-code Signed-off-by: Christophe JAILLET Link: https://lore.kernel.org/r/27992ffcee47fc865ce87274d6dfcffe7a1e69e0.1670873784.git.christophe.jaillet@wanadoo.fr Signed-off-by: Jakub Kicinski

net: lan966x: Add ptp trap rules

2022-12-06T12:26:43Z

Currently lan966x, doesn't allow to run PTP over interfaces that are part of the bridge. The reason is when the lan966x was receiving a PTP frame (regardless if L2/IPv4/IPv6) the HW it would flood this frame. Now that it is possible to add VCAP rules to the HW, such to trap these frames to the CPU, it is possible to run PTP also over interfaces that are part of the bridge. Signed-off-by: Horatiu Vultur Signed-off-by: Paolo Abeni

net: lan966x: Add port keyset config and callback interface

2022-11-29T12:08:23Z

Implement vcap_operations and enable default port keyset configuration for each port. Now it is possible actually write/read/move entries in the VCAP. Signed-off-by: Horatiu Vultur Signed-off-by: Paolo Abeni