kernel/linux.git/drivers/net/ethernet/broadcom, branch v5.18.2

net: systemport: Fix an error handling path in bcm_sysport_probe()

2022-05-17T08:38:52+00:00

if devm_clk_get_optional() fails, we still need to go through the error handling path. Add the missing goto. Fixes: 6328a126896ea ("net: systemport: Manage Wake-on-LAN clock") Signed-off-by: Christophe JAILLET Acked-by: Florian Fainelli Link: https://lore.kernel.org/r/99d70634a81c229885ae9e4ee69b2035749f7edc.1652634040.git.christophe.jaillet@wanadoo.fr Signed-off-by: Paolo Abeni

net: bcmgenet: Check for Wake-on-LAN interrupt probe deferral

2022-05-12T07:58:58+00:00

The interrupt controller supplying the Wake-on-LAN interrupt line maybe modular on some platforms (irq-bcm7038-l1.c) and might be probed at a later time than the GENET driver. We need to specifically check for -EPROBE_DEFER and propagate that error to ensure that we eventually fetch the interrupt descriptor. Fixes: 9deb48b53e7f ("bcmgenet: add WOL IRQ check") Fixes: 5b1f0e62941b ("net: bcmgenet: Avoid touching non-existent interrupt") Signed-off-by: Florian Fainelli Reviewed-by: Stefan Wahren Link: https://lore.kernel.org/r/20220511031752.2245566-1-f.fainelli@gmail.com Signed-off-by: Paolo Abeni

bnxt_en: Fix unnecessary dropping of RX packets

2022-05-04T00:41:32+00:00

In bnxt_poll_p5(), we first check cpr->has_more_work. If it is true, we are in NAPI polling mode and we will call __bnxt_poll_cqs() to continue polling. It is possible to exhanust the budget again when __bnxt_poll_cqs() returns. We then enter the main while loop to check for new entries in the NQ. If we had previously exhausted the NAPI budget, we may call __bnxt_poll_work() to process an RX entry with zero budget. This will cause packets to be dropped unnecessarily, thinking that we are in the netpoll path. Fix it by breaking out of the while loop if we need to process an RX NQ entry with no budget left. We will then exit NAPI and stay in polling mode. Fixes: 389a877a3b20 ("bnxt_en: Process the NQ under NAPI continuous polling.") Reviewed-by: Andy Gospodarek Signed-off-by: Michael Chan Signed-off-by: Jakub Kicinski

bnxt_en: Initiallize bp->ptp_lock first before using it

2022-05-04T00:41:32+00:00

bnxt_ptp_init() calls bnxt_ptp_init_rtc() which will acquire the ptp_lock spinlock. The spinlock is not initialized until later. Move the bnxt_ptp_init_rtc() call after the spinlock is initialized. Fixes: 24ac1ecd5240 ("bnxt_en: Add driver support to use Real Time Counter for PTP") Reviewed-by: Pavan Chebbi Reviewed-by: Saravanan Vajravel Reviewed-by: Andy Gospodarek Reviewed-by: Somnath Kotur Reviewed-by: Damodharam Ammepalli Signed-off-by: Michael Chan Signed-off-by: Jakub Kicinski

bnxt_en: Fix possible bnxt_open() failure caused by wrong RFS flag

2022-05-04T00:41:31+00:00

bnxt_open() can fail in this code path, especially on a VF when it fails to reserve default rings: bnxt_open() __bnxt_open_nic() bnxt_clear_int_mode() bnxt_init_dflt_ring_mode() RX rings would be set to 0 when we hit this error path. It is possible for a subsequent bnxt_open() call to potentially succeed with a code path like this: bnxt_open() bnxt_hwrm_if_change() bnxt_fw_init_one() bnxt_fw_init_one_p3() bnxt_set_dflt_rfs() bnxt_rfs_capable() bnxt_hwrm_reserve_rings() On older chips, RFS is capable if we can reserve the number of vnics that is equal to RX rings + 1. But since RX rings is still set to 0 in this code path, we may mistakenly think that RFS is supported for 0 RX rings. Later, when the default RX rings are reserved and we try to enable RFS, it would fail and cause bnxt_open() to fail unnecessarily. We fix this in 2 places. bnxt_rfs_capable() will always return false if RX rings is not yet set. bnxt_init_dflt_ring_mode() will call bnxt_set_dflt_rfs() which will always clear the RFS flags if RFS is not supported. Fixes: 20d7d1c5c9b1 ("bnxt_en: reliably allocate IRQ table on reset to avoid crash") Signed-off-by: Somnath Kotur Signed-off-by: Michael Chan Signed-off-by: Jakub Kicinski

bnx2x: fix napi API usage sequence

2022-04-28T00:50:32+00:00

While handling PCI errors (AER flow) driver tries to disable NAPI [napi_disable()] after NAPI is deleted [__netif_napi_del()] which causes unexpected system hang/crash. System message log shows the following: ======================================= [ 3222.537510] EEH: Detected PCI bus error on PHB#384-PE#800000 [ 3222.537511] EEH: This PCI device has failed 2 times in the last hour and will be permanently disabled after 5 failures. [ 3222.537512] EEH: Notify device drivers to shutdown [ 3222.537513] EEH: Beginning: 'error_detected(IO frozen)' [ 3222.537514] EEH: PE#800000 (PCI 0384:80:00.0): Invoking bnx2x->error_detected(IO frozen) [ 3222.537516] bnx2x: [bnx2x_io_error_detected:14236(eth14)]IO error detected [ 3222.537650] EEH: PE#800000 (PCI 0384:80:00.0): bnx2x driver reports: 'need reset' [ 3222.537651] EEH: PE#800000 (PCI 0384:80:00.1): Invoking bnx2x->error_detected(IO frozen) [ 3222.537651] bnx2x: [bnx2x_io_error_detected:14236(eth13)]IO error detected [ 3222.537729] EEH: PE#800000 (PCI 0384:80:00.1): bnx2x driver reports: 'need reset' [ 3222.537729] EEH: Finished:'error_detected(IO frozen)' with aggregate recovery state:'need reset' [ 3222.537890] EEH: Collect temporary log [ 3222.583481] EEH: of node=0384:80:00.0 [ 3222.583519] EEH: PCI device/vendor: 168e14e4 [ 3222.583557] EEH: PCI cmd/status register: 00100140 [ 3222.583557] EEH: PCI-E capabilities and status follow: [ 3222.583744] EEH: PCI-E 00: 00020010 012c8da2 00095d5e 00455c82 [ 3222.583892] EEH: PCI-E 10: 10820000 00000000 00000000 00000000 [ 3222.583893] EEH: PCI-E 20: 00000000 [ 3222.583893] EEH: PCI-E AER capability register set follows: [ 3222.584079] EEH: PCI-E AER 00: 13c10001 00000000 00000000 00062030 [ 3222.584230] EEH: PCI-E AER 10: 00002000 000031c0 000001e0 00000000 [ 3222.584378] EEH: PCI-E AER 20: 00000000 00000000 00000000 00000000 [ 3222.584416] EEH: PCI-E AER 30: 00000000 00000000 [ 3222.584416] EEH: of node=0384:80:00.1 [ 3222.584454] EEH: PCI device/vendor: 168e14e4 [ 3222.584491] EEH: PCI cmd/status register: 00100140 [ 3222.584492] EEH: PCI-E capabilities and status follow: [ 3222.584677] EEH: PCI-E 00: 00020010 012c8da2 00095d5e 00455c82 [ 3222.584825] EEH: PCI-E 10: 10820000 00000000 00000000 00000000 [ 3222.584826] EEH: PCI-E 20: 00000000 [ 3222.584826] EEH: PCI-E AER capability register set follows: [ 3222.585011] EEH: PCI-E AER 00: 13c10001 00000000 00000000 00062030 [ 3222.585160] EEH: PCI-E AER 10: 00002000 000031c0 000001e0 00000000 [ 3222.585309] EEH: PCI-E AER 20: 00000000 00000000 00000000 00000000 [ 3222.585347] EEH: PCI-E AER 30: 00000000 00000000 [ 3222.586872] RTAS: event: 5, Type: Platform Error (224), Severity: 2 [ 3222.586873] EEH: Reset without hotplug activity [ 3224.762767] EEH: Beginning: 'slot_reset' [ 3224.762770] EEH: PE#800000 (PCI 0384:80:00.0): Invoking bnx2x->slot_reset() [ 3224.762771] bnx2x: [bnx2x_io_slot_reset:14271(eth14)]IO slot reset initializing... [ 3224.762887] bnx2x 0384:80:00.0: enabling device (0140 -> 0142) [ 3224.768157] bnx2x: [bnx2x_io_slot_reset:14287(eth14)]IO slot reset --> driver unload Uninterruptible tasks ===================== crash> ps | grep UN 213 2 11 c000000004c89e00 UN 0.0 0 0 [eehd] 215 2 0 c000000004c80000 UN 0.0 0 0 [kworker/0:2] 2196 1 28 c000000004504f00 UN 0.1 15936 11136 wickedd 4287 1 9 c00000020d076800 UN 0.0 4032 3008 agetty 4289 1 20 c00000020d056680 UN 0.0 7232 3840 agetty 32423 2 26 c00000020038c580 UN 0.0 0 0 [kworker/26:3] 32871 4241 27 c0000002609ddd00 UN 0.1 18624 11648 sshd 32920 10130 16 c00000027284a100 UN 0.1 48512 12608 sendmail 33092 32987 0 c000000205218b00 UN 0.1 48512 12608 sendmail 33154 4567 16 c000000260e51780 UN 0.1 48832 12864 pickup 33209 4241 36 c000000270cb6500 UN 0.1 18624 11712 sshd 33473 33283 0 c000000205211480 UN 0.1 48512 12672 sendmail 33531 4241 37 c00000023c902780 UN 0.1 18624 11648 sshd EEH handler hung while bnx2x sleeping and holding RTNL lock =========================================================== crash> bt 213 PID: 213 TASK: c000000004c89e00 CPU: 11 COMMAND: "eehd" #0 [c000000004d477e0] __schedule at c000000000c70808 #1 [c000000004d478b0] schedule at c000000000c70ee0 #2 [c000000004d478e0] schedule_timeout at c000000000c76dec #3 [c000000004d479c0] msleep at c0000000002120cc #4 [c000000004d479f0] napi_disable at c000000000a06448 ^^^^^^^^^^^^^^^^ #5 [c000000004d47a30] bnx2x_netif_stop at c0080000018dba94 [bnx2x] #6 [c000000004d47a60] bnx2x_io_slot_reset at c0080000018a551c [bnx2x] #7 [c000000004d47b20] eeh_report_reset at c00000000004c9bc #8 [c000000004d47b90] eeh_pe_report at c00000000004d1a8 #9 [c000000004d47c40] eeh_handle_normal_event at c00000000004da64 And the sleeping source code ============================ crash> dis -ls c000000000a06448 FILE: ../net/core/dev.c LINE: 6702 6697 { 6698 might_sleep(); 6699 set_bit(NAPI_STATE_DISABLE, &n->state); 6700 6701 while (test_and_set_bit(NAPI_STATE_SCHED, &n->state)) * 6702 msleep(1); 6703 while (test_and_set_bit(NAPI_STATE_NPSVC, &n->state)) 6704 msleep(1); 6705 6706 hrtimer_cancel(&n->timer); 6707 6708 clear_bit(NAPI_STATE_DISABLE, &n->state); 6709 } EEH calls into bnx2x twice based on the system log above, first through bnx2x_io_error_detected() and then bnx2x_io_slot_reset(), and executes the following call chains: bnx2x_io_error_detected() +-> bnx2x_eeh_nic_unload() +-> bnx2x_del_all_napi() +-> __netif_napi_del() bnx2x_io_slot_reset() +-> bnx2x_netif_stop() +-> bnx2x_napi_disable() +->napi_disable() Fix this by correcting the sequence of NAPI APIs usage, that is delete the NAPI after disabling it. Fixes: 7fa6f34081f1 ("bnx2x: AER revised") Reported-by: David Christensen Tested-by: David Christensen Signed-off-by: Manish Chopra Signed-off-by: Ariel Elior Link: https://lore.kernel.org/r/20220426153913.6966-1-manishc@marvell.com Signed-off-by: Jakub Kicinski

net: bcmgenet: hide status block before TX timestamping

2022-04-26T08:45:46+00:00

The hardware checksum offloading requires use of a transmit status block inserted before the outgoing frame data, this was updated in '9a9ba2a4aaaa ("net: bcmgenet: always enable status blocks")' However, skb_tx_timestamp() assumes that it is passed a raw frame and PTP parsing chokes on this status block. Fix this by calling __skb_pull(), which hides the TSB before calling skb_tx_timestamp(), so an outgoing PTP packet is parsed correctly. As the data in the skb has already been set up for DMA, and the dma_unmap_* calls use a separately stored address, there is no no effective change in the data transmission. Signed-off-by: Jonathan Lemon Acked-by: Florian Fainelli Link: https://lore.kernel.org/r/20220424165307.591145-1-jonathan.lemon@gmail.com Fixes: d03825fba459 ("net: bcmgenet: add skb_tx_timestamp call") Signed-off-by: Paolo Abeni

net: bcmgenet: Revert "Use stronger register read/writes to assure ordering"

2022-04-14T07:13:07+00:00

It turns out after digging deeper into this bug, that it was being triggered by GCC12 failing to call the bcmgenet_enable_dma() routine. Given that a gcc12 fix has been merged [1] and the genet driver now works properly when built with gcc12, this commit should be reverted. [1] https://gcc.gnu.org/bugzilla/show_bug.cgi?id=105160 https://gcc.gnu.org/git/?p=gcc.git;a=commit;h=aabb9a261ef060cf24fd626713f1d7d9df81aa57 Fixes: 8d3ea3d402db ("net: bcmgenet: Use stronger register read/writes to assure ordering") Signed-off-by: Jeremy Linton Acked-by: Florian Fainelli Link: https://lore.kernel.org/r/20220412210420.1129430-1-jeremy.linton@arm.com Signed-off-by: Paolo Abeni

bnxt_en: Prevent XDP redirect from running when stopping TX queue

2022-04-04T11:44:50+00:00

Add checks in the XDP redirect callback to prevent XDP from running when the TX ring is undergoing shutdown. Also remove redundant checks in the XDP redirect callback to validate the txr and the flag that indicates the ring supports XDP. The modulo arithmetic on 'tx_nr_rings_xdp' already guarantees the derived TX ring is an XDP ring. txr is also guaranteed to be valid after checking BNXT_STATE_OPEN and within RCU grace period. Fixes: f18c2b77b2e4 ("bnxt_en: optimized XDP_REDIRECT support") Reviewed-by: Vladimir Olovyannikov Signed-off-by: Ray Jui Signed-off-by: Michael Chan Signed-off-by: David S. Miller

bnxt_en: reserve space inside receive page for skb_shared_info

2022-04-04T11:44:50+00:00

Insufficient space was being reserved in the page used for packet reception, so the interface MTU could be set too large to still have room for the contents of the packet when doing XDP redirect. This resulted in the following message when redirecting a packet between 3520 and 3822 bytes with an MTU of 3822: [311815.561880] XDP_WARN: xdp_update_frame_from_buff(line:200): Driver BUG: missing reserved tailroom Fixes: f18c2b77b2e4 ("bnxt_en: optimized XDP_REDIRECT support") Reviewed-by: Somnath Kotur Reviewed-by: Pavan Chebbi Signed-off-by: Andy Gospodarek Signed-off-by: Michael Chan Signed-off-by: David S. Miller