ath11k: move function ath11k_dp_rx_process_mon_status

Move the function below ath11k_dp_rx_mon_dest_process() and remove
the forward declaration.

Tested-on: IPQ8074 hw2.0 AHB WLAN.HK.2.5.0.1-01179-QCAHKSWPL_SILICONZ-1
Signed-off-by: Aloka Dixit <quic_alokad@quicinc.com>
Signed-off-by: Kalle Valo <quic_kvalo@quicinc.com>
Link: https://lore.kernel.org/r/20220111032224.14093-2-quic_alokad@quicinc.com

ath11k: enable RX PPDU stats in monitor co-exist mode

RX PPDU statistics collection is missing when monitor mode co-exists
with other modes. This commit combines the processing of the destination
ring with the status ring to fix the issue.

Tested-on: IPQ8074 hw2.0 AHB WLAN.HK.2.5.0.1-01179-QCAHKSWPL_SILICONZ-1
Signed-off-by: Miles Hu <milehu@codeaurora.org>
Signed-off-by: Aloka Dixit <quic_alokad@quicinc.com>
Signed-off-by: Kalle Valo <quic_kvalo@quicinc.com>
Link: https://lore.kernel.org/r/20220111032224.14093-1-quic_alokad@quicinc.com

ath5k: fix OOB in ath5k_eeprom_read_pcal_info_5111

The bug was found during fuzzing. Stacktrace locates it in
ath5k_eeprom_convert_pcal_info_5111.
When none of the curve is selected in the loop, idx can go
up to AR5K_EEPROM_N_PD_CURVES. The line makes pd out of bound.
pd = &chinfo[pier].pd_curves[idx];

There are many OOB writes using pd later in the code. So I
added a sanity check for idx. Checks for other loops involving
AR5K_EEPROM_N_PD_CURVES are not needed as the loop index is not
used outside the loops.

The patch is NOT tested with real device.

The following is the fuzzing report

BUG: KASAN: slab-out-of-bounds in ath5k_eeprom_read_pcal_info_5111+0x126a/0x1390 [ath5k]
Write of size 1 at addr ffff8880174a4d60 by task modprobe/214

CPU: 0 PID: 214 Comm: modprobe Not tainted 5.6.0 #1
Call Trace:
 dump_stack+0x76/0xa0
 print_address_description.constprop.0+0x16/0x200
 ? ath5k_eeprom_read_pcal_info_5111+0x126a/0x1390 [ath5k]
 ? ath5k_eeprom_read_pcal_info_5111+0x126a/0x1390 [ath5k]
 __kasan_report.cold+0x37/0x7c
 ? ath5k_eeprom_read_pcal_info_5111+0x126a/0x1390 [ath5k]
 kasan_report+0xe/0x20
 ath5k_eeprom_read_pcal_info_5111+0x126a/0x1390 [ath5k]
 ? apic_timer_interrupt+0xa/0x20
 ? ath5k_eeprom_init_11a_pcal_freq+0xbc0/0xbc0 [ath5k]
 ? ath5k_pci_eeprom_read+0x228/0x3c0 [ath5k]
 ath5k_eeprom_init+0x2513/0x6290 [ath5k]
 ? ath5k_eeprom_init_11a_pcal_freq+0xbc0/0xbc0 [ath5k]
 ? usleep_range+0xb8/0x100
 ? apic_timer_interrupt+0xa/0x20
 ? ath5k_eeprom_read_pcal_info_2413+0x2f20/0x2f20 [ath5k]
 ath5k_hw_init+0xb60/0x1970 [ath5k]
 ath5k_init_ah+0x6fe/0x2530 [ath5k]
 ? kasprintf+0xa6/0xe0
 ? ath5k_stop+0x140/0x140 [ath5k]
 ? _dev_notice+0xf6/0xf6
 ? apic_timer_interrupt+0xa/0x20
 ath5k_pci_probe.cold+0x29a/0x3d6 [ath5k]
 ? ath5k_pci_eeprom_read+0x3c0/0x3c0 [ath5k]
 ? mutex_lock+0x89/0xd0
 ? ath5k_pci_eeprom_read+0x3c0/0x3c0 [ath5k]
 local_pci_probe+0xd3/0x160
 pci_device_probe+0x23f/0x3e0
 ? pci_device_remove+0x280/0x280
 ? pci_device_remove+0x280/0x280
 really_probe+0x209/0x5d0

Reported-by: Brendan Dolan-Gavitt <brendandg@nyu.edu>
Signed-off-by: Zekun Shen <bruceshenzk@gmail.com>
Signed-off-by: Kalle Valo <quic_kvalo@quicinc.com>
Link: https://lore.kernel.org/r/YckvDdj3mtCkDRIt@a-10-27-26-18.dynapool.vpn.nyu.edu

ath11k: free peer for station when disconnect from AP for QCA6390/WCN6855

Commit b4a0f54156ac ("ath11k: move peer delete after vdev stop of station
for QCA6390 and WCN6855") is to fix firmware crash by changing the WMI
command sequence, but actually skip all the peer delete operation, then
it lead commit 58595c9874c6 ("ath11k: Fixing dangling pointer issue upon
peer delete failure") not take effect, and then happened a use-after-free
warning from KASAN. because the peer->sta is not set to NULL and then used
later.

Change to only skip the WMI_PEER_DELETE_CMDID for QCA6390/WCN6855.

log of user-after-free:

[  534.888665] BUG: KASAN: use-after-free in ath11k_dp_rx_update_peer_stats+0x912/0xc10 [ath11k]
[  534.888696] Read of size 8 at addr ffff8881396bb1b8 by task rtcwake/2860

[  534.888705] CPU: 4 PID: 2860 Comm: rtcwake Kdump: loaded Tainted: G        W         5.15.0-wt-ath+ #523
[  534.888712] Hardware name: Intel(R) Client Systems NUC8i7HVK/NUC8i7HVB, BIOS HNKBLi70.86A.0067.2021.0528.1339 05/28/2021
[  534.888716] Call Trace:
[  534.888720]  <IRQ>
[  534.888726]  dump_stack_lvl+0x57/0x7d
[  534.888736]  print_address_description.constprop.0+0x1f/0x170
[  534.888745]  ? ath11k_dp_rx_update_peer_stats+0x912/0xc10 [ath11k]
[  534.888771]  kasan_report.cold+0x83/0xdf
[  534.888783]  ? ath11k_dp_rx_update_peer_stats+0x912/0xc10 [ath11k]
[  534.888810]  ath11k_dp_rx_update_peer_stats+0x912/0xc10 [ath11k]
[  534.888840]  ath11k_dp_rx_process_mon_status+0x529/0xa70 [ath11k]
[  534.888874]  ? ath11k_dp_rx_mon_status_bufs_replenish+0x3f0/0x3f0 [ath11k]
[  534.888897]  ? check_prev_add+0x20f0/0x20f0
[  534.888922]  ? __lock_acquire+0xb72/0x1870
[  534.888937]  ? find_held_lock+0x33/0x110
[  534.888954]  ath11k_dp_rx_process_mon_rings+0x297/0x520 [ath11k]
[  534.888981]  ? rcu_read_unlock+0x40/0x40
[  534.888990]  ? ath11k_dp_rx_pdev_alloc+0xd90/0xd90 [ath11k]
[  534.889026]  ath11k_dp_service_mon_ring+0x67/0xe0 [ath11k]
[  534.889053]  ? ath11k_dp_rx_process_mon_rings+0x520/0x520 [ath11k]
[  534.889075]  call_timer_fn+0x167/0x4a0
[  534.889084]  ? add_timer_on+0x3b0/0x3b0
[  534.889103]  ? lockdep_hardirqs_on_prepare.part.0+0x18c/0x370
[  534.889117]  __run_timers.part.0+0x539/0x8b0
[  534.889123]  ? ath11k_dp_rx_process_mon_rings+0x520/0x520 [ath11k]
[  534.889157]  ? call_timer_fn+0x4a0/0x4a0
[  534.889164]  ? mark_lock_irq+0x1c30/0x1c30
[  534.889173]  ? clockevents_program_event+0xdd/0x280
[  534.889189]  ? mark_held_locks+0xa5/0xe0
[  534.889203]  run_timer_softirq+0x97/0x180
[  534.889213]  __do_softirq+0x276/0x86a
[  534.889230]  __irq_exit_rcu+0x11c/0x180
[  534.889238]  irq_exit_rcu+0x5/0x20
[  534.889244]  sysvec_apic_timer_interrupt+0x8e/0xc0
[  534.889251]  </IRQ>
[  534.889254]  <TASK>
[  534.889259]  asm_sysvec_apic_timer_interrupt+0x12/0x20
[  534.889265] RIP: 0010:_raw_spin_unlock_irqrestore+0x38/0x70
[  534.889271] Code: 74 24 10 e8 ea c2 bf fd 48 89 ef e8 12 53 c0 fd 81 e3 00 02 00 00 75 25 9c 58 f6 c4 02 75 2d 48 85 db 74 01 fb bf 01 00 00 00 <e8> 13 a7 b5 fd 65 8b 05 cc d9 9c 5e 85 c0 74 0a 5b 5d c3 e8 a0 ee
[  534.889276] RSP: 0018:ffffc90002e5f880 EFLAGS: 00000206
[  534.889284] RAX: 0000000000000006 RBX: 0000000000000200 RCX: ffffffff9f256f10
[  534.889289] RDX: 0000000000000000 RSI: ffffffffa1c6e420 RDI: 0000000000000001
[  534.889293] RBP: ffff8881095e6200 R08: 0000000000000001 R09: ffffffffa40d2b8f
[  534.889298] R10: fffffbfff481a571 R11: 0000000000000001 R12: ffff8881095e6e68
[  534.889302] R13: ffffc90002e5f908 R14: 0000000000000246 R15: 0000000000000000
[  534.889316]  ? mark_lock+0xd0/0x14a0
[  534.889332]  klist_next+0x1d4/0x450
[  534.889340]  ? dpm_wait_for_subordinate+0x2d0/0x2d0
[  534.889350]  device_for_each_child+0xa8/0x140
[  534.889360]  ? device_remove_class_symlinks+0x1b0/0x1b0
[  534.889370]  ? __lock_release+0x4bd/0x9f0
[  534.889378]  ? dpm_suspend+0x26b/0x3f0
[  534.889390]  dpm_wait_for_subordinate+0x82/0x2d0
[  534.889400]  ? dpm_for_each_dev+0xa0/0xa0
[  534.889410]  ? dpm_suspend+0x233/0x3f0
[  534.889427]  __device_suspend+0xd4/0x10c0
[  534.889440]  ? wait_for_completion_io+0x270/0x270
[  534.889456]  ? async_suspend_late+0xe0/0xe0
[  534.889463]  ? async_schedule_node_domain+0x468/0x640
[  534.889482]  dpm_suspend+0x25a/0x3f0
[  534.889491]  ? dpm_suspend_end+0x1a0/0x1a0
[  534.889497]  ? ktime_get+0x214/0x2f0
[  534.889502]  ? lockdep_hardirqs_on+0x79/0x100
[  534.889509]  ? recalibrate_cpu_khz+0x10/0x10
[  534.889516]  ? ktime_get+0x119/0x2f0
[  534.889528]  dpm_suspend_start+0xab/0xc0
[  534.889538]  suspend_devices_and_enter+0x1ca/0x350
[  534.889546]  ? suspend_enter+0x850/0x850
[  534.889566]  enter_state+0x27c/0x3d7
[  534.889575]  pm_suspend.cold+0x42/0x189
[  534.889583]  state_store+0xab/0x160
[  534.889595]  ? sysfs_file_ops+0x160/0x160
[  534.889601]  kernfs_fop_write_iter+0x2b5/0x450
[  534.889615]  new_sync_write+0x36a/0x600
[  534.889625]  ? new_sync_read+0x600/0x600
[  534.889639]  ? rcu_read_unlock+0x40/0x40
[  534.889668]  vfs_write+0x619/0x910
[  534.889681]  ksys_write+0xf4/0x1d0
[  534.889689]  ? __ia32_sys_read+0xa0/0xa0
[  534.889699]  ? lockdep_hardirqs_on_prepare.part.0+0x18c/0x370
[  534.889707]  ? syscall_enter_from_user_mode+0x1d/0x50
[  534.889719]  do_syscall_64+0x3b/0x90
[  534.889725]  entry_SYSCALL_64_after_hwframe+0x44/0xae
[  534.889731] RIP: 0033:0x7f0b9bc931e7
[  534.889736] Code: 64 89 02 48 c7 c0 ff ff ff ff eb bb 0f 1f 80 00 00 00 00 f3 0f 1e fa 64 8b 04 25 18 00 00 00 85 c0 75 10 b8 01 00 00 00 0f 05 <48> 3d 00 f0 ff ff 77 51 c3 48 83 ec 28 48 89 54 24 18 48 89 74 24
[  534.889741] RSP: 002b:00007ffd9d34cc88 EFLAGS: 00000246 ORIG_RAX: 0000000000000001
[  534.889749] RAX: ffffffffffffffda RBX: 0000000000000004 RCX: 00007f0b9bc931e7
[  534.889753] RDX: 0000000000000004 RSI: 0000561cd023c5f0 RDI: 0000000000000004
[  534.889757] RBP: 0000561cd023c5f0 R08: 0000000000000000 R09: 0000000000000004
[  534.889761] R10: 0000561ccef842a6 R11: 0000000000000246 R12: 0000000000000004
[  534.889765] R13: 0000561cd0239590 R14: 00007f0b9bd6f4a0 R15: 00007f0b9bd6e8a0
[  534.889789]  </TASK>

[  534.889796] Allocated by task 2711:
[  534.889800]  kasan_save_stack+0x1b/0x40
[  534.889805]  __kasan_kmalloc+0x7c/0x90
[  534.889810]  sta_info_alloc+0x98/0x1ef0 [mac80211]
[  534.889874]  ieee80211_prep_connection+0x30b/0x11e0 [mac80211]
[  534.889950]  ieee80211_mgd_auth+0x529/0xe00 [mac80211]
[  534.890024]  cfg80211_mlme_auth+0x332/0x6f0 [cfg80211]
[  534.890090]  nl80211_authenticate+0x839/0xcf0 [cfg80211]
[  534.890147]  genl_family_rcv_msg_doit+0x1f4/0x2f0
[  534.890154]  genl_rcv_msg+0x280/0x500
[  534.890160]  netlink_rcv_skb+0x11c/0x340
[  534.890165]  genl_rcv+0x1f/0x30
[  534.890170]  netlink_unicast+0x42b/0x700
[  534.890176]  netlink_sendmsg+0x71b/0xc60
[  534.890181]  sock_sendmsg+0xdf/0x110
[  534.890187]  ____sys_sendmsg+0x5c0/0x850
[  534.890192]  ___sys_sendmsg+0xe4/0x160
[  534.890197]  __sys_sendmsg+0xb2/0x140
[  534.890202]  do_syscall_64+0x3b/0x90
[  534.890207]  entry_SYSCALL_64_after_hwframe+0x44/0xae

[  534.890215] Freed by task 2825:
[  534.890218]  kasan_save_stack+0x1b/0x40
[  534.890223]  kasan_set_track+0x1c/0x30
[  534.890227]  kasan_set_free_info+0x20/0x30
[  534.890232]  __kasan_slab_free+0xce/0x100
[  534.890237]  slab_free_freelist_hook+0xf0/0x1a0
[  534.890242]  kfree+0xe5/0x370
[  534.890248]  __sta_info_flush+0x333/0x4b0 [mac80211]
[  534.890308]  ieee80211_set_disassoc+0x324/0xd20 [mac80211]
[  534.890382]  ieee80211_mgd_deauth+0x537/0xee0 [mac80211]
[  534.890472]  cfg80211_mlme_deauth+0x349/0x810 [cfg80211]
[  534.890526]  cfg80211_mlme_down+0x1ce/0x270 [cfg80211]
[  534.890578]  cfg80211_disconnect+0x4f5/0x7b0 [cfg80211]
[  534.890631]  cfg80211_leave+0x24/0x40 [cfg80211]
[  534.890677]  wiphy_suspend+0x23d/0x2f0 [cfg80211]
[  534.890723]  dpm_run_callback+0xf4/0x1b0
[  534.890728]  __device_suspend+0x648/0x10c0
[  534.890733]  async_suspend+0x16/0xe0
[  534.890737]  async_run_entry_fn+0x90/0x4f0
[  534.890741]  process_one_work+0x866/0x1490
[  534.890747]  worker_thread+0x596/0x1010
[  534.890751]  kthread+0x35d/0x420
[  534.890756]  ret_from_fork+0x22/0x30

[  534.890763] The buggy address belongs to the object at ffff8881396ba000
                which belongs to the cache kmalloc-8k of size 8192
[  534.890767] The buggy address is located 4536 bytes inside of
                8192-byte region [ffff8881396ba000, ffff8881396bc000)
[  534.890772] The buggy address belongs to the page:
[  534.890775] page:ffffea0004e5ae00 refcount:1 mapcount:0 mapping:0000000000000000 index:0x0 pfn:0x1396b8
[  534.890780] head:ffffea0004e5ae00 order:3 compound_mapcount:0 compound_pincount:0
[  534.890784] flags: 0x200000000010200(slab|head|node=0|zone=2)
[  534.890791] raw: 0200000000010200 ffffea000562be08 ffffea0004b04c08 ffff88810004e340
[  534.890795] raw: 0000000000000000 0000000000010001 00000001ffffffff 0000000000000000
[  534.890798] page dumped because: kasan: bad access detected

[  534.890804] Memory state around the buggy address:
[  534.890807]  ffff8881396bb080: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
[  534.890811]  ffff8881396bb100: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
[  534.890814] >ffff8881396bb180: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
[  534.890817]                                         ^
[  534.890821]  ffff8881396bb200: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
[  534.890824]  ffff8881396bb280: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
[  534.890827] ==================================================================
[  534.890830] Disabling lock debugging due to kernel taint

Tested-on: WCN6855 hw2.0 PCI WLAN.HSP.1.1-01720.1-QCAHSPSWPL_V1_V2_SILICONZ_LITE-1

Fixes: b4a0f54156ac ("ath11k: move peer delete after vdev stop of station for QCA6390 and WCN6855")
Signed-off-by: Wen Gong <quic_wgong@quicinc.com>
Signed-off-by: Kalle Valo <quic_kvalo@quicinc.com>
Link: https://lore.kernel.org/r/20211222070431.29595-1-quic_wgong@quicinc.com

ath11k: add LDPC FEC type in 802.11 radiotap header

LDPC is one the FEC type advertised in msdu_start info2 for HT packet
type. Hence, add hardware specific callback for fetching LDPC
support from msdu start and enable RX_ENC_FLAG_LDPC flag while passing
rx status to mac80211.

Tested-on: IPQ8074 WLAN.HK.2.4.0.1-01467-QCAHKSWPL_SILICONZ-1

Signed-off-by: P Praneesh <quic_ppranees@quicinc.com>
Signed-off-by: Kalle Valo <quic_kvalo@quicinc.com>
Link: https://lore.kernel.org/r/1638294648-844-3-git-send-email-quic_ppranees@quicinc.com

ath11k: Refactor the fallback routine when peer create fails

When there is an error in peer create process from
ath11k_peer_find(), the code attempts to handle a fallback
for peer create. When this fallback fails, the driver returns
the fallback return code rather than actual error code
(-ENOENT). So refactor the fallback routine to return
the actual error code.

Tested-on: QCN9074 hw1.0 PCI WLAN.HK.2.5.0.1-01067-QCAHKSWPL_SILICONZ-1

Signed-off-by: Karthikeyan Periyasamy <periyasa@codeaurora.org>
Signed-off-by: Kalle Valo <quic_kvalo@quicinc.com>
Link: https://lore.kernel.org/r/1640244819-21183-1-git-send-email-quic_periyasa@quicinc.com

ath11k: fix workqueue not getting destroyed after rmmod

Currently, ath11k_core_alloc() creates a single thread workqueue.
This workqueue is not detroyed during clean up when ath11k modules
are unloaded from the kernel and is left as it is.
If workqueue is not destroyed, it could lead to kernel
memory scarcity in a longer run. This could affect self and
other drivers workability as well.

Add destroy workqueue in ath11k_core_free().

Tested on: IPQ8074 WLAN.HK.2.4.0.1-01746-QCAHKSWPL_SILICONZ-1

Signed-off-by: Aditya Kumar Singh <quic_adisi@quicinc.com>
Signed-off-by: Kalle Valo <quic_kvalo@quicinc.com>
Link: https://lore.kernel.org/r/1640231787-17408-1-git-send-email-quic_adisi@quicinc.com

ath10k: fix memory overwrite of the WoWLAN wakeup packet pattern

In function ath10k_wow_convert_8023_to_80211(), it will do memcpy for
the new->pattern, and currently the new->pattern and new->mask is same
with the old, then the memcpy of new->pattern will also overwrite the
old->pattern, because the header format of new->pattern is 802.11,
its length is larger than the old->pattern which is 802.3. Then the
operation of "Copy frame body" will copy a mistake value because the
body memory has been overwrite when memcpy the new->pattern.

Assign another empty value to new_pattern to avoid the overwrite issue.

Tested-on: QCA6174 hw3.2 SDIO WLAN.RMH.4.4.1-00049

Fixes: fa3440fa2fa1 ("ath10k: convert wow pattern from 802.3 to 802.11")
Signed-off-by: Wen Gong <quic_wgong@quicinc.com>
Signed-off-by: Kalle Valo <quic_kvalo@quicinc.com>
Link: https://lore.kernel.org/r/20211222031347.25463-1-quic_wgong@quicinc.com

ath11k: add missing of_node_put() to avoid leak

The node pointer is returned by of_find_node_by_type()
or of_parse_phandle() with refcount incremented. Calling
of_node_put() to aovid the refcount leak.

Fixes: 6ac04bdc5edb ("ath11k: Use reserved host DDR addresses from DT for PCI devices")
Reported-by: Hulk Robot <hulkci@huawei.com>
Signed-off-by: Yang Yingliang <yangyingliang@huawei.com>
Signed-off-by: Kalle Valo <quic_kvalo@quicinc.com>
Link: https://lore.kernel.org/r/20211221114003.335557-1-yangyingliang@huawei.com

Merge git://git./linux/kernel/git/netdev/net

Merge in fixes directly in prep for the 5.17 merge window.
No conflicts.

Signed-off-by: Jakub Kicinski <kuba@kernel.org>

tcp: tcp_send_challenge_ack delete useless param `skb`

After this parameter is passed in, there is no usage, and deleting it will
 not bring any impact.

Reviewed-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Benjamin Yim <yan2228598786@gmail.com>
Link: https://lore.kernel.org/r/20220109130824.2776-1-yan2228598786@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

net/qla3xxx: Remove useless DMA-32 fallback configuration

As stated in [1], dma_set_mask() with a 64-bit mask never fails if
dev->dma_mask is non-NULL.
So, if it fails, the 32 bits case will also fail for the same reason.

So, if dma_set_mask_and_coherent() succeeds, 'pci_using_dac' is known to be
1.

Simplify code and remove some dead code accordingly.

[1]: https://lkml.org/lkml/2021/6/7/398

Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
Link: https://lore.kernel.org/r/3011689e8c77d49d7e44509d5a8241320ec408c5.1641754134.git.christophe.jaillet@wanadoo.fr
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

rocker: Remove useless DMA-32 fallback configuration

As stated in [1], dma_set_mask() with a 64-bit mask never fails if
dev->dma_mask is non-NULL.
So, if it fails, the 32 bits case will also fail for the same reason.

Simplify code and remove some dead code accordingly.

[1]: https://lkml.org/lkml/2021/6/7/398

Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
Link: https://lore.kernel.org/r/9ba2d13099d216f3df83e50ad33a05504c90fe7c.1641744274.git.christophe.jaillet@wanadoo.fr
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

hinic: Remove useless DMA-32 fallback configuration

As stated in [1], dma_set_mask() with a 64-bit mask never fails if
dev->dma_mask is non-NULL.
So, if it fails, the 32 bits case will also fail for the same reason.

Simplify code and remove some dead code accordingly.

[1]: https://lkml.org/lkml/2021/6/7/398

Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
Link: https://lore.kernel.org/r/23541c28df8d0dcd3663b5dbe0f76af71e70e9cc.1641743855.git.christophe.jaillet@wanadoo.fr
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

lan743x: Remove useless DMA-32 fallback configuration

As stated in [1], dma_set_mask() with a 64-bit mask never fails if
dev->dma_mask is non-NULL.
So, if it fails, the 32 bits case will also fail for the same reason.

Simplify code and remove some dead code accordingly.

[1]: https://lkml.org/lkml/2021/6/7/398

Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
Link: https://lore.kernel.org/r/ef548716606f257939df9738a801f15b6edf2568.1641743405.git.christophe.jaillet@wanadoo.fr
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

net: enetc: Remove useless DMA-32 fallback configuration

As stated in [1], dma_set_mask() with a 64-bit mask never fails if
dev->dma_mask is non-NULL.
So, if it fails, the 32 bits case will also fail for the same reason.

Simplify code and remove some dead code accordingly.

[1]: https://lkml.org/lkml/2021/6/7/398

Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
Link: https://lore.kernel.org/r/dbecd4eb49a9586ee343b5473dda4b84c42112e9.1641742884.git.christophe.jaillet@wanadoo.fr
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

cxgb4vf: Remove useless DMA-32 fallback configuration

As stated in [1], dma_set_mask() with a 64-bit mask never fails if
dev->dma_mask is non-NULL.
So, if it fails, the 32 bits case will also fail for the same reason.

So, if dma_set_mask_and_coherent() succeeds, 'pci_using_dac' is known to be
1.

Simplify code and remove some dead code accordingly.

[1]: https://lkml.org/lkml/2021/6/7/398

Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
Link: https://lore.kernel.org/r/b14986ea39cea2ca9a6cd0476a3fc167c853ee67.1641736772.git.christophe.jaillet@wanadoo.fr
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

cxgb4: Remove useless DMA-32 fallback configuration

As stated in [1], dma_set_mask() with a 64-bit mask never fails if
dev->dma_mask is non-NULL.
So, if it fails, the 32 bits case will also fail for the same reason.

So, if dma_set_mask_and_coherent() succeeds, 'highdma' is known to be true.

Simplify code and remove some dead code accordingly.

[1]: https://lkml.org/lkml/2021/6/7/398

Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
Link: https://lore.kernel.org/r/56db10d53be0897ff1be5f37d64b91cb7e1d932c.1641736387.git.christophe.jaillet@wanadoo.fr
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

cxgb3: Remove useless DMA-32 fallback configuration

As stated in [1], dma_set_mask() with a 64-bit mask never fails if
dev->dma_mask is non-NULL.
So, if it fails, the 32 bits case will also fail for the same reason.

So, if dma_set_mask_and_coherent() succeeds, 'pci_using_dac' is known to be
1.

Simplify code and remove some dead code accordingly.

[1]: https://lkml.org/lkml/2021/6/7/398

Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
Link: https://lore.kernel.org/r/a0e2539aefb0034091aca02c98440ea9459f1258.1641736234.git.christophe.jaillet@wanadoo.fr
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

bnx2x: Remove useless DMA-32 fallback configuration

As stated in [1], dma_set_mask() with a 64-bit mask never fails if
dev->dma_mask is non-NULL.
So, if it fails, the 32 bits case will also fail for the same reason.

Moreover, dma_set_mask_and_coherent() returns 0 or -EIO, so the return
code of the function can be used directly.

Finally, inline bnx2x_set_coherency_mask() because it is now only a wrapper
for a single dma_set_mask_and_coherent() call.

Simplify code and remove some dead code accordingly.

[1]: https://lkml.org/lkml/2021/6/7/398

Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
Link: https://lore.kernel.org/r/29608a525876afddceabf8f11b2ba606da8748fc.1641730747.git.christophe.jaillet@wanadoo.fr
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

et131x: Remove useless DMA-32 fallback configuration

As stated in [1], dma_set_mask() with a 64-bit mask never fails if
dev->dma_mask is non-NULL.
So, if it fails, the 32 bits case will also fail for the same reason.

Moreover, dma_set_mask_and_coherent() returns 0 or -EIO, so the return
code of the function can be used directly. There is no need to 'rc = -EIO'
explicitly.

Simplify code and remove some dead code accordingly.

[1]: https://lkml.org/lkml/2021/6/7/398

Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
Link: https://lore.kernel.org/r/b9aa46e7e5a5aa61f56aac5ea439930f41ad9946.1641726804.git.christophe.jaillet@wanadoo.fr
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

be2net: Remove useless DMA-32 fallback configuration

As stated in [1], dma_set_mask() with a 64-bit mask never fails if
dev->dma_mask is non-NULL.
So, if it fails, the 32 bits case will also fail for the same reason.

So if dma_set_mask_and_coherent() succeeds, 'netdev->features' will have
NETIF_F_HIGHDMA in all cases. Move the assignment of this feature in
be_netdev_init() instead be_probe() which is a much logical place.

Simplify code and remove some dead code accordingly.

[1]: https://lkml.org/lkml/2021/6/7/398

Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
Link: https://lore.kernel.org/r/637696d7141faa68c29fc34b70f9aa67d5e605f0.1641718999.git.christophe.jaillet@wanadoo.fr
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

vmxnet3: Remove useless DMA-32 fallback configuration

As stated in [1], dma_set_mask() with a 64-bit mask never fails if
dev->dma_mask is non-NULL.
So, if it fails, the 32 bits case will also fail for the same reason.

So if dma_set_mask_and_coherent() succeeds, 'dma64' is know to be 'true'.

Simplify code and remove some dead code accordingly.

[1]: https://lkml.org/lkml/2021/6/7/398

Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
Link: https://lore.kernel.org/r/43e5dcf1a5e9e9c5d2d86f87810d6e93e3d22e32.1641718188.git.christophe.jaillet@wanadoo.fr
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

bna: Simplify DMA setting

As stated in [1], dma_set_mask() with a 64-bit mask will never fail if
dev->dma_mask is non-NULL.
So, if it fails, the 32 bits case will also fail for the same reason.

So, if dma_set_mask_and_coherent() succeeds, 'using_dac' is known to be
'true'. This variable can be removed.

Simplify code and remove some dead code accordingly.

[1]: https://lkml.org/lkml/2021/6/7/398

Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
Link: https://lore.kernel.org/r/1d5a7b3f4fa735f1233c3eb3fa07e71df95fad75.1641658516.git.christophe.jaillet@wanadoo.fr
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

net: alteon: Simplify DMA setting

As stated in [1], dma_set_mask() with a 64-bit mask will never fail if
dev->dma_mask is non-NULL.
So, if it fails, the 32 bits case will also fail for the same reason.

If dma_set_mask_and_coherent() succeeds, 'ap->pci_using_dac' is known to be
1. So 'pci_using_dac' can be removed from the 'struct ace_private'.

Simplify code and remove some dead code accordingly.

[1]: https://lkml.org/lkml/2021/6/7/398

Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
Link: https://lore.kernel.org/r/1a414c05c27b21c661aef61dffe1adcd1578b1f5.1641651917.git.christophe.jaillet@wanadoo.fr
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

myri10ge: Simplify DMA setting

As stated in [1], dma_set_mask() with a 64-bit mask will never fail if
dev->dma_mask is non-NULL.
So, if it fails, the 32 bits case will also fail for the same reason.

If dma_set_mask_and_coherent() succeeds, 'dac_enabled' is known to be 1.

Simplify code and remove some dead code accordingly.

[1]: https://lkml.org/lkml/2021/6/7/398

Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
Link: https://lore.kernel.org/r/e92b0c3a3c1574a97a4e6fd0c30225f10fa59d18.1641651693.git.christophe.jaillet@wanadoo.fr
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

qlcnic: Simplify DMA setting

As stated in [1], dma_set_mask() with a 64-bit mask will never fail if
dev->dma_mask is non-NULL.
So, if it fails, the 32 bits case will also fail for the same reason.

So qlcnic_set_dma_mask(), (in qlcnic_main.c) can be simplified a lot and
inlined directly in its only caller.

If dma_set_mask_and_coherent() succeeds, 'pci_using_dac' is known to be 1.
So it can be removed from all the calling chain.

qlcnic_setup_netdev() can finally be simplified as-well.

[1]: https://lkml.org/lkml/2021/6/7/398

Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
Link: https://lore.kernel.org/r/4996ab0337d62ec6a54b2edf234cd5ced4b4d7ad.1641649611.git.christophe.jaillet@wanadoo.fr
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

net: allwinner: Fix print format

Kees reports quoted commit introduced the following warning on arm64:

drivers/net/ethernet/allwinner/sun4i-emac.c:922:60: error: format '%x' expects argument of type 'unsigned int', but argument 3 has type 'resource_size_t' {aka 'long long unsigned int'} [-Werror=format=]
  922 |         netdev_info(ndev, "get io resource from device: 0x%x, size = %u\n",
      |                                                           ~^
      |                                                            |                                      |                                                            unsigned int
      |                                                           %llx
  923 |                     regs->start, resource_size(regs));
      |                     ~~~~~~~~~~~
      |                         |
      |                         resource_size_t {aka long long unsigned int}

.. and another one like that for resource_size().

Switch to %pa and a cast.

Reported-by: Kees Cook <keescook@chromium.org>
Fixes: 47869e82c8b8 ("sun4i-emac.c: add dma support")
Link: https://lore.kernel.org/r/20220108034438.2227343-1-kuba@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

page_pool: remove spinlock in page_pool_refill_alloc_cache()

As page_pool_refill_alloc_cache() is only called by
__page_pool_get_cached(), which assumes non-concurrent access
as suggested by the comment in __page_pool_get_cached(), and
ptr_ring allows concurrent access between consumer and producer,
so remove the spinlock in page_pool_refill_alloc_cache().

Signed-off-by: Yunsheng Lin <linyunsheng@huawei.com>
Acked-by: Jesper Dangaard Brouer <brouer@redhat.com>
Link: https://lore.kernel.org/r/20220107090042.13605-1-linyunsheng@huawei.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

amt: fix wrong return type of amt_send_membership_update()

amt_send_membership_update() would return -1 but it's return type is bool.
So, it should be used TRUE instead of -1.

Fixes: cbc21dc1cfe9 ("amt: add data plane of amt interface")
Reported-by: kernel test robot <lkp@intel.com>
Signed-off-by: Taehee Yoo <ap420073@gmail.com>
Link: https://lore.kernel.org/r/20220109163702.6331-1-ap420073@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

net: mcs7830: handle usb read errors properly

Syzbot reported uninit value in mcs7830_bind(). The problem was in
missing validation check for bytes read via usbnet_read_cmd().

usbnet_read_cmd() internally calls usb_control_msg(), that returns
number of bytes read. Code should validate that requested number of bytes
was actually read.

So, this patch adds missing size validation check inside
mcs7830_get_reg() to prevent uninit value bugs

Reported-and-tested-by: syzbot+003c0a286b9af5412510@syzkaller.appspotmail.com
Fixes: 2a36d7083438 ("USB: driver for mcs7830 (aka DeLOCK) USB ethernet adapter")
Signed-off-by: Pavel Skripkin <paskripkin@gmail.com>
Reviewed-by: Arnd Bergmann <arnd@arndb.de>
Link: https://lore.kernel.org/r/20220106225716.7425-1-paskripkin@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

Merge branch 'net-skb-introduce-kfree_skb_with_reason'

Menglong Dong says:

====================
net: skb: introduce kfree_skb_with_reason()

In this series patch, the interface kfree_skb_with_reason() is
introduced(), which is used to collect skb drop reason, and pass
it to 'kfree_skb' tracepoint. Therefor, 'drop_monitor' or eBPF is
able to monitor abnormal skb with detail reason.

In fact, this series patches are out of the intelligence of David
and Steve, I'm just a truck man :/

Previous discussion is here:

https://lore.kernel.org/netdev/20211118105752.1d46e990@gandalf.local.home/
https://lore.kernel.org/netdev/67b36bd8-2477-88ac-83a0-35a1eeaf40c9@gmail.com/

In the first patch, kfree_skb_with_reason() is introduced and
the 'reason' field is added to 'kfree_skb' tracepoint. In the
second patch, 'kfree_skb()' in replaced with 'kfree_skb_with_reason()'
in tcp_v4_rcv(). In the third patch, 'kfree_skb_with_reason()' is
used in __udp4_lib_rcv().

Changes since v3:
- fix some code style problems in skb.h

Changes since v2:
- rename kfree_skb_with_reason() to kfree_skb_reason()
- make kfree_skb() static inline, as Jakub suggested

Changes since v1:
- rename some drop reason, as David suggested
- add the third patch
====================

Link: https://lore.kernel.org/r/20220109063628.526990-1-imagedong@tencent.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

net: skb: use kfree_skb_reason() in __udp4_lib_rcv()

Replace kfree_skb() with kfree_skb_reason() in __udp4_lib_rcv.
New drop reason 'SKB_DROP_REASON_UDP_CSUM' is added for udp csum
error.

Signed-off-by: Menglong Dong <imagedong@tencent.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

net: skb: use kfree_skb_reason() in tcp_v4_rcv()

Replace kfree_skb() with kfree_skb_reason() in tcp_v4_rcv(). Following
drop reasons are added:

SKB_DROP_REASON_NO_SOCKET
SKB_DROP_REASON_PKT_TOO_SMALL
SKB_DROP_REASON_TCP_CSUM
SKB_DROP_REASON_TCP_FILTER

After this patch, 'kfree_skb' event will print message like this:

$           TASK-PID     CPU#  |||||  TIMESTAMP  FUNCTION
$              | |         |   |||||     |         |
          <idle>-0       [000] ..s1.    36.113438: kfree_skb: skbaddr=(____ptrval____) protocol=2048 location=(____ptrval____) reason: NO_SOCKET

The reason of skb drop is printed too.

Signed-off-by: Menglong Dong <imagedong@tencent.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

net: skb: introduce kfree_skb_reason()

Introduce the interface kfree_skb_reason(), which is able to pass
the reason why the skb is dropped to 'kfree_skb' tracepoint.

Add the 'reason' field to 'trace_kfree_skb', therefor user can get
more detail information about abnormal skb with 'drop_monitor' or
eBPF.

All drop reasons are defined in the enum 'skb_drop_reason', and
they will be print as string in 'kfree_skb' tracepoint in format
of 'reason: XXX'.

( Maybe the reasons should be defined in a uapi header file, so that
user space can use them? )

Signed-off-by: Menglong Dong <imagedong@tencent.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

net/mlx5e: Fix build error in fec_set_block_stats()

Build bot reports:

drivers/net/ethernet/mellanox/mlx5/core/en_stats.c: In function 'fec_set_block_stats':
drivers/net/ethernet/mellanox/mlx5/core/en_stats.c:1235:48: error: 'outl' undeclared (first use in this function); did you mean 'out'?
    1235 |         if (mlx5_core_access_reg(mdev, in, sz, outl, sz, MLX5_REG_PPCNT, 0, 0))
         |                                                ^~~~
         |                                                out

Reported-by: kernel test robot <lkp@intel.com>
Suggested-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Link: https://lore.kernel.org/r/20220109213321.2292830-1-kuba@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

Merge branch 'bnxt_en-update-for-net-next'

Michael Chan says:

====================
bnxt_en: Update for net-next

This series adds better error and debug logging for firmware messages.
We now also use the firmware provided timeout value for long running
commands instead of capping it to 40 seconds.
====================

Link: https://lore.kernel.org/r/1641772485-10421-1-git-send-email-michael.chan@broadcom.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

bnxt_en: improve firmware timeout messaging

While it has always been possible to infer that an HWRM command was
abandoned due to an unhealthy firmware status by the shortened timeout
reported, this change improves the log messaging to account for this
case explicitly. In the interests of further clarity, the firmware
status is now also reported in these new messages.

v2: Remove inline keyword for hwrm_wait_must_abort() in .c file.

Reviewed-by: Andy Gospodarek <gospo@broadcom.com>
Signed-off-by: Edwin Peer <edwin.peer@broadcom.com>
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

bnxt_en: use firmware provided max timeout for messages

Some older devices cannot accommodate the 40 seconds timeout
cap for long running commands (such as NVRAM commands) due to
hardware limitations. Allow these devices to request more time for
these long running commands, but print a warning, since the longer
timeout may cause the hung task watchdog to trigger. In the case of a
firmware update operation, this is preferable to failing outright.

v2: Use bp->hwrm_cmd_max_timeout directly without the constants.

Fixes: 881d8353b05e ("bnxt_en: Add an upper bound for all firmware command timeouts.")
Signed-off-by: Edwin Peer <edwin.peer@broadcom.com>
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

bnxt_en: improve VF error messages when PF is unavailable

The current driver design relies on the PF netdev being open in order
to intercept the following HWRM commands from a VF:
    - HWRM_FUNC_VF_CFG
    - HWRM_CFA_L2_FILTER_ALLOC
    - HWRM_PORT_PHY_QCFG (only if FW_CAP_LINK_ADMIN is not supported)

If the PF is closed, then VFs are subjected to rather inscrutable error
messages in response to any configuration requests involving the above
command types. Recent firmware distinguishes this problem case from
other errors by returning HWRM_ERR_CODE_PF_UNAVAILABLE. In most cases,
the appropriate course of action is still to fail, but this can now be
accomplished with the aid of more user informative log messages. For L2
filter allocations that are already asynchronous, an automatic retry
seems more appropriate.

v2: Delete extra newline.

Signed-off-by: Edwin Peer <edwin.peer@broadcom.com>
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

bnxt_en: add dynamic debug support for HWRM messages

Add logging of firmware messages. These can be useful for diagnosing
issues in the field, but due to their verbosity are only appropriate
at a debug message level.

Signed-off-by: Edwin Peer <edwin.peer@broadcom.com>
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

iwlwifi: mvm: Use div_s64 instead of do_div in iwl_mvm_ftm_rtt_smoothing()

When building ARCH=arm allmodconfig:

drivers/net/wireless/intel/iwlwifi/mvm/ftm-initiator.c: In function ‘iwl_mvm_ftm_rtt_smoothing’:
./include/asm-generic/div64.h:222:35: error: comparison of distinct pointer types lacks a cast [-Werror]
  222 |         (void)(((typeof((n)) *)0) == ((uint64_t *)0));  \
      |                                   ^~
drivers/net/wireless/intel/iwlwifi/mvm/ftm-initiator.c:1070:9: note: in expansion of macro ‘do_div’
 1070 |         do_div(rtt_avg, 100);
      |         ^~~~~~

do_div() has to be used with an unsigned 64-bit integer dividend but
rtt_avg is a signed 64-bit integer.

div_s64() expects a signed 64-bit integer dividend and signed 32-bit
divisor, which fits this scenario, so use that function here to fix the
warning.

Fixes: 8b0f92549f2c ("iwlwifi: mvm: fix 32-bit build in FTM")
Signed-off-by: Nathan Chancellor <nathan@kernel.org>
Link: https://lore.kernel.org/r/20211227191757.2354329-1-nathan@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

net: openvswitch: Fix ct_state nat flags for conns arriving from tc

Netfilter conntrack maintains NAT flags per connection indicating
whether NAT was configured for the connection. Openvswitch maintains
NAT flags on the per packet flow key ct_state field, indicating
whether NAT was actually executed on the packet.

When a packet misses from tc to ovs the conntrack NAT flags are set.
However, NAT was not necessarily executed on the packet because the
connection's state might still be in NEW state. As such, openvswitch
wrongly assumes that NAT was executed and sets an incorrect flow key
NAT flags.

Fix this, by flagging to openvswitch which NAT was actually done in
act_ct via tc_skb_ext and tc_skb_cb to the openvswitch module, so
the packet flow key NAT flags will be correctly set.

Fixes: b57dc7c13ea9 ("net/sched: Introduce action ct")
Signed-off-by: Paul Blakey <paulb@nvidia.com>
Acked-by: Jamal Hadi Salim <jhs@mojatatu.com>
Link: https://lore.kernel.org/r/20220106153804.26451-1-paulb@nvidia.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

Merge git://git./linux/kernel/git/pablo/nf-next

Pablo Neira Ayuso says:

====================
Netfilter updates for net-next

The following patchset contains Netfilter updates for net-next. This
includes one patch to update ovs and act_ct to use nf_ct_put() instead
of nf_conntrack_put().

1) Add netns_tracker to nfnetlink_log and masquerade, from Eric Dumazet.

2) Remove redundant rcu read-size lock in nf_tables packet path.

3) Replace BUG() by WARN_ON_ONCE() in nft_payload.

4) Consolidate rule verdict tracing.

5) Replace WARN_ON() by WARN_ON_ONCE() in nf_tables core.

6) Make counter support built-in in nf_tables.

7) Add new field to conntrack object to identify locally generated
   traffic, from Florian Westphal.

8) Prevent NAT from shadowing well-known ports, from Florian Westphal.

9) Merge nf_flow_table_{ipv4,ipv6} into nf_flow_table_inet, also from
   Florian.

10) Remove redundant pointer in nft_pipapo AVX2 support, from Colin Ian King.

11) Replace opencoded max() in conntrack, from Jiapeng Chong.

12) Update conntrack to use refcount_t API, from Florian Westphal.

13) Move ip_ct_attach indirection into the nf_ct_hook structure.

14) Constify several pointer object in the netfilter codebase,
    from Florian Westphal.

15) Tree-wide replacement of nf_conntrack_put() by nf_ct_put(), also
    from Florian.

16) Fix egress splat due to incorrect rcu notation, from Florian.

17) Move stateful fields of connlimit, last, quota, numgen and limit
    out of the expression data area.

18) Build a blob to represent the ruleset in nf_tables, this is a
    requirement of the new register tracking infrastructure.

19) Add NFT_REG32_NUM to define the maximum number of 32-bit registers.

20) Add register tracking infrastructure to skip redundant
    store-to-register operations, this includes support for payload,
    meta and bitwise expresssions.

* git://git.kernel.org/pub/scm/linux/kernel/git/pablo/nf-next: (32 commits)
  netfilter: nft_meta: cancel register tracking after meta update
  netfilter: nft_payload: cancel register tracking after payload update
  netfilter: nft_bitwise: track register operations
  netfilter: nft_meta: track register operations
  netfilter: nft_payload: track register operations
  netfilter: nf_tables: add register tracking infrastructure
  netfilter: nf_tables: add NFT_REG32_NUM
  netfilter: nf_tables: add rule blob layout
  netfilter: nft_limit: move stateful fields out of expression data
  netfilter: nft_limit: rename stateful structure
  netfilter: nft_numgen: move stateful fields out of expression data
  netfilter: nft_quota: move stateful fields out of expression data
  netfilter: nft_last: move stateful fields out of expression data
  netfilter: nft_connlimit: move stateful fields out of expression data
  netfilter: egress: avoid a lockdep splat
  net: prefer nf_ct_put instead of nf_conntrack_put
  netfilter: conntrack: avoid useless indirection during conntrack destruction
  netfilter: make function op structures const
  netfilter: core: move ip_ct_attach indirection to struct nf_ct_hook
  netfilter: conntrack: convert to refcount_t api
  ...
====================

Link: https://lore.kernel.org/r/20220109231640.104123-1-pablo@netfilter.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

netfilter: nft_meta: cancel register tracking after meta update

The meta expression might mangle the packet metadata, cancel register
tracking since any metadata in the registers is stale.

Finer grain register tracking cancellation by inspecting the meta type
on the register is also possible.

Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>

netfilter: nft_payload: cancel register tracking after payload update

The payload expression might mangle the packet, cancel register tracking
since any payload data in the registers is stale.

Finer grain register tracking cancellation by inspecting the payload
base, offset and length on the register is also possible.

Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>

netfilter: nft_bitwise: track register operations

Check if the destination register already contains the data that this
bitwise expression performs. This allows to skip this redundant
operation.

If the destination contains a different bitwise operation, cancel the
register tracking information. If the destination contains no bitwise
operation, update the register tracking information.

Update the payload and meta expression to check if this bitwise
operation has been already performed on the register. Hence, both the
payload/meta and the bitwise expressions are reduced.

There is also a special case: If source register != destination register
and source register is not updated by a previous bitwise operation, then
transfer selector from the source register to the destination register.

Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>

netfilter: nft_meta: track register operations

Check if the destination register already contains the data that this
meta store expression performs. This allows to skip this redundant
operation. If the destination contains a different selector, update
the register tracking information.

Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>

netfilter: nft_payload: track register operations

Check if the destination register already contains the data that this
payload store expression performs. This allows to skip this redundant
operation. If the destination contains a different selector, update
the register tracking information.

Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>

netfilter: nf_tables: add register tracking infrastructure

This patch adds new infrastructure to skip redundant selector store
operations on the same register to achieve a performance boost from
the packet path.

This is particularly noticeable in pure linear rulesets but it also
helps in rulesets which are already heaving relying in maps to avoid
ruleset linear inspection.

The idea is to keep data of the most recurrent store operations on
register to reuse them with cmp and lookup expressions.

This infrastructure allows for dynamic ruleset updates since the ruleset
blob reduction happens from the kernel.

Userspace still needs to be updated to maximize register utilization to
cooperate to improve register data reuse / reduce number of store on
register operations.

Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>

netfilter: nf_tables: add NFT_REG32_NUM

Add a definition including the maximum number of 32-bits registers that
are used a scratchpad memory area to store data.

Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>

netfilter: nf_tables: add rule blob layout

This patch adds a blob layout per chain to represent the ruleset in the
packet datapath.

size (unsigned long)
struct nft_rule_dp
  struct nft_expr
  ...
        struct nft_rule_dp
          struct nft_expr
          ...
        struct nft_rule_dp (is_last=1)

The new structure nft_rule_dp represents the rule in a more compact way
(smaller memory footprint) compared to the control-plane nft_rule
structure.

The ruleset blob is a read-only data structure. The first field contains
the blob size, then the rules containing expressions. There is a trailing
rule which is used by the tracing infrastructure which is equivalent to
the NULL rule marker in the previous representation. The blob size field
does not include the size of this trailing rule marker.

The ruleset blob is generated from the commit path.

This patch reuses the infrastructure available since 0cbc06b3faba
("netfilter: nf_tables: remove synchronize_rcu in commit phase") to
build the array of rules per chain.

Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>

netfilter: nft_limit: move stateful fields out of expression data

In preparation for the rule blob representation.

Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>

netfilter: nft_limit: rename stateful structure

From struct nft_limit to nft_limit_priv.

Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>

netfilter: nft_numgen: move stateful fields out of expression data

In preparation for the rule blob representation.

Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>

netfilter: nft_quota: move stateful fields out of expression data

In preparation for the rule blob representation.

Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>

netfilter: nft_last: move stateful fields out of expression data

In preparation for the rule blob representation.

Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>

netfilter: nft_connlimit: move stateful fields out of expression data

In preparation for the rule blob representation.

Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>

netfilter: egress: avoid a lockdep splat

include/linux/netfilter_netdev.h:97 suspicious rcu_dereference_check() usage!
2 locks held by sd-resolve/1100:
 0: ..(rcu_read_lock_bh){1:3}, at: ip_finish_output2
 1: ..(rcu_read_lock_bh){1:3}, at: __dev_queue_xmit
 __dev_queue_xmit+0 ..

The helper has two callers, one uses rcu_read_lock, the other
rcu_read_lock_bh().  Annotate the dereference to reflect this.

Fixes: 42df6e1d221dd ("netfilter: Introduce egress hook")
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>

net: prefer nf_ct_put instead of nf_conntrack_put

Its the same as nf_conntrack_put(), but without the
need for an indirect call.  The downside is a module dependency on
nf_conntrack, but all of these already depend on conntrack anyway.

Cc: Paul Blakey <paulb@mellanox.com>
Cc: dev@openvswitch.org
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>

netfilter: conntrack: avoid useless indirection during conntrack destruction

nf_ct_put() results in a usesless indirection:

nf_ct_put -> nf_conntrack_put -> nf_conntrack_destroy -> rcu readlock +
indirect call of ct_hooks->destroy().

There are two _put helpers:
nf_ct_put and nf_conntrack_put.  The latter is what should be used in
code that MUST NOT cause a linker dependency on the conntrack module
(e.g. calls from core network stack).

Everyone else should call nf_ct_put() instead.

A followup patch will convert a few nf_conntrack_put() calls to
nf_ct_put(), in particular from modules that already have a conntrack
dependency such as act_ct or even nf_conntrack itself.

Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>

netfilter: make function op structures const

No functional changes, these structures should be const.

Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>

netfilter: core: move ip_ct_attach indirection to struct nf_ct_hook

ip_ct_attach predates struct nf_ct_hook, we can place it there and
remove the exported symbol.

Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>

netfilter: conntrack: convert to refcount_t api

Convert nf_conn reference counting from atomic_t to refcount_t based api.
refcount_t api provides more runtime sanity checks and will warn on
certain constructs, e.g. refcount_inc() on a zero reference count, which
usually indicates use-after-free.

For this reason template allocation is changed to init the refcount to
1, the subsequenct add operations are removed.

Likewise, init_conntrack() is changed to set the initial refcount to 1
instead refcount_inc().

This is safe because the new entry is not (yet) visible to other cpus.

Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>

netfilter: conntrack: Use max() instead of doing it manually

Fix following coccicheck warning:

./include/net/netfilter/nf_conntrack.h:282:16-17: WARNING opportunity
for max().

Reported-by: Abaci Robot <abaci@linux.alibaba.com>
Signed-off-by: Jiapeng Chong <jiapeng.chong@linux.alibaba.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>

Merge tag 'for-net-next-2022-01-07' of git://git./linux/kernel/git/bluetooth/bluetooth-next

Luiz Augusto von Dentz says:

====================
bluetooth-next pull request for net-next:

 - Add support for Foxconn QCA 0xe0d0
 - Fix HCI init sequence on MacBook Air 8,1 and 8,2
 - Fix Intel firmware loading on legacy ROM devices

* tag 'for-net-next-2022-01-07' of git://git.kernel.org/pub/scm/linux/kernel/git/bluetooth/bluetooth-next:
  Bluetooth: hci_sock: fix endian bug in hci_sock_setsockopt()
  Bluetooth: L2CAP: uninitialized variables in l2cap_sock_setsockopt()
  Bluetooth: btqca: sequential validation
  Bluetooth: btusb: Add support for Foxconn QCA 0xe0d0
  Bluetooth: btintel: Fix broken LED quirk for legacy ROM devices
  Bluetooth: hci_event: Rework hci_inquiry_result_with_rssi_evt
  Bluetooth: btbcm: disable read tx power for MacBook Air 8,1 and 8,2
  Bluetooth: hci_qca: Fix NULL vs IS_ERR_OR_NULL check in qca_serdev_probe
  Bluetooth: hci_bcm: Check for error irq
====================

Link: https://lore.kernel.org/r/20220107210942.3750887-1-luiz.dentz@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

Merge tag 'linux-can-fixes-for-5.16-20220109' of git://git./linux/kernel/git/mkl/linux-can

Marc Kleine-Budde says:

====================
pull-request: can 2022-01-09

The first patch is by Johan Hovold and fixes a mem leak in the error
path of the softing_cs driver.

The next patch is by me and fixes a set but not used variable warning
in the softing driver.

Jiasheng Jiang's patch for the xilinx_can driver adds the missing
error checking when getting the IRQ.

Lad Prabhakar contributes a patch for the rcar_canfd driver to fix a
mem leak in the error path.

The last patch is by Brian Silverman and properly initializes the send
USB messages to avoid spurious CAN error frames.

* tag 'linux-can-fixes-for-5.16-20220109' of git://git.kernel.org/pub/scm/linux/kernel/git/mkl/linux-can:
  can: gs_usb: gs_can_start_xmit(): zero-initialize hf->{flags,reserved}
  can: rcar_canfd: rcar_canfd_channel_probe(): make sure we free CAN network device
  can: xilinx_can: xcan_probe(): check for error irq
  can: softing: softing_startstop(): fix set but not used variable warning
  can: softing_cs: softingcs_probe(): fix memleak on registration failure
====================

Link: https://lore.kernel.org/r/20220109134040.1945428-1-mkl@pengutronix.de
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

Merge tag 'linux-can-next-for-5.17-20220108' of git://git./linux/kernel/git/mkl/linux-can-next

Marc Kleine-Budde says:

====================
pull-request: can-next 2022-01-08

this is a pull request of 22 patches for net-next/master.

The first patch is by Tom Rix and fixes an uninitialized variable in
the janz-ican3 driver (introduced in linux-can-next-for-5.17-20220105).

The next 13 patches are by my and target the mcp251xfd driver. First
several cleanup patches, then the driver is prepared for the upcoming
ethtool ring parameter and IRQ coalescing support, which is added in a
later pull request.

The remaining 8 patches are by Dario Binacchi and me and enhance the
flexcan driver. The driver is moved into a sub directory. An ethtool
private flag is added to optionally disable CAN RTR frame reception,
to make use of more RX buffers. The resulting RX buffer configuration
can be read by ethtool ring parameter support. Finally documentation
for the ethtool private flag is added to the
Documentation/networking/device_drivers/can directory.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>

can: gs_usb: gs_can_start_xmit(): zero-initialize hf->{flags,reserved}

No information is deliberately sent in hf->flags in host -> device
communications, but the open-source candleLight firmware echoes it
back, which can result in the GS_CAN_FLAG_OVERFLOW flag being set and
generating spurious ERRORFRAMEs.

While there also initialize the reserved member with 0.

Fixes: d08e973a77d1 ("can: gs_usb: Added support for the GS_USB CAN devices")
Link: https://lore.kernel.org/all/20220106002952.25883-1-brian.silverman@bluerivertech.com
Link: https://github.com/candle-usb/candleLight_fw/issues/87
Cc: stable@vger.kernel.org
Signed-off-by: Brian Silverman <brian.silverman@bluerivertech.com>
[mkl: initialize the reserved member, too]
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>

can: rcar_canfd: rcar_canfd_channel_probe(): make sure we free CAN network device

Make sure we free CAN network device in the error path. There are
several jumps to fail label after allocating the CAN network device
successfully. This patch places the free_candev() under fail label so
that in failure path a jump to fail label frees the CAN network
device.

Fixes: 76e9353a80e9 ("can: rcar_canfd: Add support for RZ/G2L family")
Link: https://lore.kernel.org/all/20220106114801.20563-1-prabhakar.mahadev-lad.rj@bp.renesas.com
Reported-by: Pavel Machek <pavel@denx.de>
Signed-off-by: Lad Prabhakar <prabhakar.mahadev-lad.rj@bp.renesas.com>
Reviewed-by: Kieran Bingham <kieran.bingham+renesas@ideasonboard.com>
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>

can: xilinx_can: xcan_probe(): check for error irq

For the possible failure of the platform_get_irq(), the returned irq
could be error number and will finally cause the failure of the
request_irq().

Consider that platform_get_irq() can now in certain cases return
-EPROBE_DEFER, and the consequences of letting request_irq()
effectively convert that into -EINVAL, even at probe time rather than
later on. So it might be better to check just now.

Fixes: b1201e44f50b ("can: xilinx CAN controller support")
Link: https://lore.kernel.org/all/20211224021324.1447494-1-jiasheng@iscas.ac.cn
Signed-off-by: Jiasheng Jiang <jiasheng@iscas.ac.cn>
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>

can: softing: softing_startstop(): fix set but not used variable warning

In the function softing_startstop() the variable error_reporting is
assigned but not used. The code that uses this variable is commented
out. Its stated that the functionality is not finally verified.

To fix the warning:

| drivers/net/can/softing/softing_fw.c:424:9: error: variable 'error_reporting' set but not used [-Werror,-Wunused-but-set-variable]

remove the comment, activate the code, but add a "0 &&" to the if
expression and rely on the optimizer rather than the preprocessor to
remove the code.

Link: https://lore.kernel.org/all/20220109103126.1872833-1-mkl@pengutronix.de
Fixes: 03fd3cf5a179 ("can: add driver for Softing card")
Cc: Kurt Van Dijck <dev.kurt@vandijck-laurijssen.be>
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>

can: softing_cs: softingcs_probe(): fix memleak on registration failure

In case device registration fails during probe, the driver state and
the embedded platform device structure needs to be freed using
platform_device_put() to properly free all resources (e.g. the device
name).

Fixes: 0a0b7a5f7a04 ("can: add driver for Softing card")
Link: https://lore.kernel.org/all/20211222104843.6105-1-johan@kernel.org
Cc: stable@vger.kernel.org # 2.6.38
Signed-off-by: Johan Hovold <johan@kernel.org>
Reviewed-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>

docs: networking: device drivers: can: add flexcan

Add initial documentation for Flexcan driver.

Link: https://lore.kernel.org/all/20220107193105.1699523-8-mkl@pengutronix.de
Signed-off-by: Dario Binacchi <dario.binacchi@amarulasolutions.com>
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>

docs: networking: device drivers: add can sub-folder

Add the container for CAN drivers documentation.

Link: https://lore.kernel.org/all/20220107193105.1699523-7-mkl@pengutronix.de
Signed-off-by: Dario Binacchi <dario.binacchi@amarulasolutions.com>
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>

can: flexcan: add ethtool support to get rx/tx ring parameters

This patch adds ethtool support to get the number of message buffers
configured for reception/transmission, which may also depends on
runtime configurations such as the 'rx-rtr' flag state.

Link: https://lore.kernel.org/all/20220108181633.420433-1-dario.binacchi@amarulasolutions.com
Signed-off-by: Dario Binacchi <dario.binacchi@amarulasolutions.com>
[mkl: port to net-next/master, replace __sw_hweight64 by simpler calculation]
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>

can: flexcan: add ethtool support to change rx-rtr setting during runtime

This patch adds a private flag to the flexcan driver to switch the
"rx-rtr" setting on and off.

"rx-rtr" on  - Receive RTR frames. (default)
               The CAN controller can and will receive RTR frames.

       On some IP cores the controller cannot receive RTR
       frames in the more performant "RX mailbox" mode and
       will use "RX FIFO" mode instead.

"rx-rtr" off - Waive ability to receive RTR frames. (not supported on all IP cores)
               This mode activates the "RX mailbox mode" for better
       performance, on some IP cores RTR frames cannot be
       received anymore.

The "RX FIFO" mode uses a FIFO with a depth of 6 CAN frames.
The "RX mailbox" mode uses up to 62 mailboxes.

Link: https://lore.kernel.org/all/20220107193105.1699523-6-mkl@pengutronix.de
Signed-off-by: Dario Binacchi <dario.binacchi@amarulasolutions.com>
Co-developed-by: Dario Binacchi <dario.binacchi@amarulasolutions.com>
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>

can: flexcan: add more quirks to describe RX path capabilities

Most flexcan IP cores support 2 RX modes:
- FIFO
- mailbox

Some IP core versions cannot receive CAN RTR messages via mailboxes.
This patch adds quirks to document this.

This information will be used in a later patch to switch from FIFO to
more performant mailbox mode at the expense of losing the ability to
receive RTR messages. This trade off is beneficial in certain use
cases.

Link: https://lore.kernel.org/all/20220107193105.1699523-5-mkl@pengutronix.de
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>

can: flexcan: rename RX modes

Most flexcan IP cores support 2 RX modes:
- FIFO
- mailbox

The names for these modes were chosen to reflect the name of the
rx-offload mode they are using.

The name of the RX modes should better reflect their difference with
regards the flexcan IP core. So this patch renames the various
occurrences of OFF_FIFO to RX_FIFO and OFF_TIMESTAMP to RX_MAILBOX:

| FLEXCAN_TX_MB_RESERVED_OFF_FIFO -> FLEXCAN_TX_MB_RESERVED_RX_FIFO
| FLEXCAN_TX_MB_RESERVED_OFF_TIMESTAMP -> FLEXCAN_TX_MB_RESERVED_RX_MAILBOX
| FLEXCAN_QUIRK_USE_OFF_TIMESTAMP -> FLEXCAN_QUIRK_USE_RX_MAILBOX

Link: https://lore.kernel.org/all/20220107193105.1699523-4-mkl@pengutronix.de
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>

can: flexcan: allow to change quirks at runtime

This is a preparation patch for the upcoming support to change the
rx-rtr capability via the ethtool API.

Link: https://lore.kernel.org/all/20220107193105.1699523-3-mkl@pengutronix.de
Signed-off-by: Dario Binacchi <dario.binacchi@amarulasolutions.com>
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>

can: flexcan: move driver into separate sub directory

This patch moves the flexcan driver into a separate directory, a later
patch will add more files.

Link: https://lore.kernel.org/all/20220107193105.1699523-2-mkl@pengutronix.de
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>

can: mcp251xfd: introduce and make use of mcp251xfd_is_fd_mode()

This patch replaces the open coded check, if the chip's FIFOs are
configured for CAN-FD mode, by the newly introduced function
mcp251xfd_is_fd_mode().

Link: https://lore.kernel.org/all/20220105154300.1258636-14-mkl@pengutronix.de
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>

can: mcp251xfd: move ring init into separate function

This patch moves the ring initialization from the mcp251xfd core file
into a separate one to make the driver a bit more orderly.

Link: https://lore.kernel.org/all/20220105154300.1258636-13-mkl@pengutronix.de
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>

can: mcp251xfd: move chip FIFO init into separate file

This patch moves the chip FIFO initialization from the mcp251xfd core
file into a separate one to make the driver a bit more orderly.

Link: https://lore.kernel.org/all/20220105154300.1258636-12-mkl@pengutronix.de
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>

can: mcp251xfd: move TEF handling into separate file

This patch moves the TEF handling from the mcp251xfd core file into a
separate one to make the driver a bit more orderly.

Link: https://lore.kernel.org/all/20220105154300.1258636-11-mkl@pengutronix.de
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>

can: mcp251xfd: move TX handling into separate file

This patch moves the TX handling from the mcp251xfd core file into a
separate one to make the driver a bit more orderly.

Link: https://lore.kernel.org/all/20220105154300.1258636-10-mkl@pengutronix.de
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>

can: mcp251xfd: move RX handling into separate file

This patch moves the RX handling from the mcp251xfd core file into a
separate one to make the driver a bit more orderly.

Link: https://lore.kernel.org/all/20220105154300.1258636-9-mkl@pengutronix.de
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>

can: mcp251xfd: mcp251xfd.h: sort function prototypes

The .c files in the Makefile are ordered alphabetically. This patch
groups the function prototypes by their corresponding .c file and
brings the into the same order.

Link: https://lore.kernel.org/all/20220105154300.1258636-8-mkl@pengutronix.de
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>

can: mcp251xfd: mcp251xfd_handle_rxovif(): denote RX overflow message to debug + add rate limiting

A RX overflow usually happens during high system load. Printing
overflow messages to the kernel log, which on embedded systems often
is outputted on the serial console, even increases the system load.

To decrease the system load in these situations, denote the messages
to debug level and wrap them with net_ratelimit().

Link: https://lore.kernel.org/all/20220105154300.1258636-7-mkl@pengutronix.de
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>

can: mcp251xfd: mcp251xfd_open(): make use of pm_runtime_resume_and_get()

With patch

| dd8088d5a896 PM: runtime: Add pm_runtime_resume_and_get to deal with usage counter

the usual pm_runtime_get_sync() and pm_runtime_put_noidle()
in-case-of-error dance is no longer needed. Convert the mcp251xfd
driver to use this function.

Link: https://lore.kernel.org/all/20220105154300.1258636-6-mkl@pengutronix.de
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>

can: mcp251xfd: mcp251xfd_open(): open_candev() first

This patch exchanges the order of open_candev() and
pm_runtime_get_sync(), so that open_candev() is called first.

A usual reason why open_candev() fails is missing CAN bit rate
configuration. It makes no sense to resume the device from PM sleep
first just to put it to sleep if the bit rate is not configured.

Link: https://lore.kernel.org/all/20220105154300.1258636-5-mkl@pengutronix.de
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>

can: mcp251xfd: add missing newline to printed strings

This patch adds the missing newline to printed strings.

Fixes: 55e5b97f003e ("can: mcp25xxfd: add driver for Microchip MCP25xxFD SPI CAN")
Link: https://lore.kernel.org/all/20220105154300.1258636-4-mkl@pengutronix.de
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>

can: mcp251xfd: mcp251xfd_tef_obj_read(): fix typo in error message

This patch fixes a typo in the error message in
mcp251xfd_tef_obj_read(), if trying to read too many objects.

Link: https://lore.kernel.org/all/20220105154300.1258636-3-mkl@pengutronix.de
Fixes: 55e5b97f003e ("can: mcp25xxfd: add driver for Microchip MCP25xxFD SPI CAN")
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>

can: mcp251xfd: remove double blank lines

This patch removes double blank lines from the driver.

Link: https://lore.kernel.org/all/20220105154300.1258636-2-mkl@pengutronix.de
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>

can: janz-ican3: initialize dlc variable

Clang static analysis reports this problem
janz-ican3.c:1311:2: warning: Undefined or garbage value returned to caller
        return dlc;
        ^~~~~~~~~~

dlc is only set with this conditional
if (!(cf->can_id & CAN_RTR_FLAG))
dlc = cf->len;

But is always returned.  So initialize dlc to 0.

Fixes: cc4b08c31b5c ("can: do not increase tx_bytes statistics for RTR frames")
Link: https://lore.kernel.org/all/20220108143319.3986923-1-trix@redhat.com
Signed-off-by: Tom Rix <trix@redhat.com>
Acked-by: Vincent Mailhol <mailhol.vincent@wanadoo.fr>
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>

Merge branch 'ena-capabilities-field-and-cosmetic-changes'

Arthur Kiyanovski says:

====================
ENA: capabilities field and cosmetic changes

Add a new capabilities bitmask field to get indication of
capabilities supported by the device. Use the capabilities
field to query the device for ENI stats support.

Other patches are cosmetic changes like fixing readme
mistakes, removing unused variables etc...
====================

Link: https://lore.kernel.org/r/20220107202346.3522-1-akiyano@amazon.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

net: ena: Extract recurring driver reset code into a function

Create an inline function for resetting the driver
to reduce code duplication.

Signed-off-by: Nati Koler <nkoler@amazon.com>
Signed-off-by: Arthur Kiyanovski <akiyano@amazon.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

net: ena: Change the name of bad_csum variable

Changed bad_csum to csum_bad to align with csum_unchecked & csum_good

Signed-off-by: Nati Koler <nkoler@amazon.com>
Signed-off-by: Arthur Kiyanovski <akiyano@amazon.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

net: ena: Add debug prints for invalid req_id resets

Add qid and req_id to error prints when ENA_REGS_RESET_INV_TX_REQ_ID
reset occurs.

Switch from %hu to %u, since u16 should be printed with %u, as
explained in [1].

[1] - https://www.kernel.org/doc/html/latest/core-api/printk-formats.html

Signed-off-by: Arthur Kiyanovski <akiyano@amazon.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

net: ena: Remove ena_calc_queue_size_ctx struct

This struct was used to pass data from callee function to its caller.
Its usage can be avoided.

Removing it results in less code without any damage to code readability.
Also it allows to consolidate ring size calculation into a single
function (ena_calc_io_queue_size()).

Signed-off-by: Shay Agroskin <shayagr@amazon.com>
Signed-off-by: Arthur Kiyanovski <akiyano@amazon.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>