On Wed, 28 Aug 2024 at 18:05, Bert Karwatzki <spasswolf@xxxxxx> wrote:
>
> Am Mittwoch, dem 28.08.2024 um 16:20 +0100 schrieb Mike Lothian:
> > On Wed, 28 Aug 2024 at 09:17, Bert Karwatzki <spasswolf@xxxxxx> wrote:
> > >
> > > Am Mittwoch, dem 28.08.2024 um 00:38 +0100 schrieb Mike Lothian:
> > > > But not the follow on fix which fixes suspend...
> > > >
> > > > diff --git a/drivers/net/wireless/mediatek/mt76/mt7921/main.c
> > > > b/drivers/net/wireless/mediatek/mt76/mt7921/main.c
> > > > index 2e6268cb06c0..a85c19da77e5 100644
> > > > --- a/drivers/net/wireless/mediatek/mt76/mt7921/main.c
> > > > +++ b/drivers/net/wireless/mediatek/mt76/mt7921/main.c
> > > > @@ -1182,7 +1182,7 @@ static void mt7921_ipv6_addr_change(struct
> > > > ieee80211_hw *hw,
> > > > struct inet6_dev *idev)
> > > > {
> > > > struct mt792x_vif *mvif = (struct mt792x_vif *)vif->drv_priv;
> > > > - struct mt792x_dev *dev = mvif->phy->dev;
> > > > + struct mt792x_dev *dev = mt792x_hw_dev(hw);
> > > > struct inet6_ifaddr *ifa;
> > > > struct in6_addr ns_addrs[IEEE80211_BSS_ARP_ADDR_LIST_LEN];
> > > > struct sk_buff *skb;
> > > > @@ -1205,6 +1205,9 @@ static void mt7921_ipv6_addr_change(struct
> > > > ieee80211_hw *hw,
> > > > },
> > > > };
> > > >
> > > > + if (!mvif->phy)
> > > > + return;
> > > > +
> > > > read_lock_bh(&idev->lock);
> > > > list_for_each_entry(ifa, &idev->addr_list, if_list) {
> > > > if (ifa->flags & IFA_F_TENTATIVE)
> > > >
> > > > > > >
> > > > >
> > >
> > > That is interesting (or odd) because suspend (s2idle) works for me in next-
> > > 20240827 on this hardware:
> > >
> > > 04:00.0 Network controller [0280]: MEDIATEK Corp. MT7921K (RZ608) Wi-Fi 6E 80MHz
> > > [14c3:0608]
> > >
> > > are you using a different model?
> > >
> > > [ 272.780054] [ T848] wlp4s0: deauthenticating from 54:67:51:3d:a2:d2 by
> > > local choice (Reason: 3=DEAUTH_LEAVING)
> > > [ 273.067438] [ T3437] PM: suspend entry (s2idle)
> > > [ 273.151140] [ T3437] Filesystems sync: 0.083 seconds
> > > [ 273.157029] [ T3437] Freezing user space processes
> > >
> > > Can your the post the backtrace of your error (when commit 479ffee68d59c599f is
> > > applied)?
> > >
> > > Bert Karwatzki
> >
> > Apologies, it wasn't a suspend issue, it was when the router was restarted
> >
> > The error I saw was previously in this thread and someone pointed me
> > to the patch that fixed it - but that hasn't made it upstream yet
> >
> > Aug 01 08:59:33 quark kernel: BUG: kernel NULL pointer dereference,
> > address: 0000000000000008
> > Aug 01 08:59:33 quark kernel: #PF: supervisor read access in kernel mode
> > Aug 01 08:59:33 quark kernel: #PF: error_code(0x0000) - not-present page
> > Aug 01 08:59:33 quark kernel: PGD 0 P4D 0
> > Aug 01 08:59:33 quark kernel: Oops: Oops: 0000 [#1] PREEMPT SMP
> > Aug 01 08:59:33 quark kernel: CPU: 13 UID: 0 PID: 468 Comm:
> > NetworkManager Not tainted 6.11.0-rc1-tip+ #3200
> > 9c927d6f3c59d826d15d8e39c195392d1d16b8a8
> > Aug 01 08:59:33 quark kernel: Hardware name: Micro Computer (HK) Tech
> > Limited EliteMini Series/HPBSD, BIOS 1.02 03/28/2024
> > Aug 01 08:59:33 quark kernel: RIP: 0010:mt7921_ipv6_addr_change
> > Aug 01 08:59:33 quark kernel: Code: 41 57 41 56 41 54 53 48 83 e4 f0
> > 48 83 ec 50 48 8b 86 70 09 00 00 0f b6 8e 90 04 00 00 4c 8d ba 68 02
> > 00 00 49 89 d6 4c 89 ff <48> 8b 58 08 88 4c 24 04 66 c7 44 24 05 00 00
> > c6 44 24 07 00 66 c7
> > Aug 01 08:59:33 quark kernel: RSP: 0018:ffffc900069373b0 EFLAGS: 00010282
> > Aug 01 08:59:33 quark kernel: RAX: 0000000000000000 RBX:
> > ffff888106740920 RCX: 0000000000000000
> > Aug 01 08:59:33 quark kernel: RDX: ffff888106854800 RSI:
> > ffff88810bb35ca0 RDI: ffff888106854a68
> > Aug 01 08:59:33 quark kernel: RBP: ffffc90006937420 R08:
> > 0000000000000000 R09: ffff888104c98200
> > Aug 01 08:59:33 quark kernel: R10: ffffffff7fff0000 R11:
> > 0000000000000020 R12: 0000000000000002
> > Aug 01 08:59:33 quark kernel: R13: 0000000000000000 R14:
> > ffff888106854800 R15: ffff888106854a68
> > Aug 01 08:59:33 quark kernel: FS: 00007f4265049400(0000)
> > GS:ffff888c2df40000(0000) knlGS:0000000000000000
> > Aug 01 08:59:33 quark kernel: CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
> > Aug 01 08:59:33 quark kernel: CR2: 0000000000000008 CR3:
> > 0000000117250000 CR4: 0000000000350ef0
> > Aug 01 08:59:33 quark kernel: Call Trace:
> > Aug 01 08:59:33 quark kernel: <TASK>
> > Aug 01 08:59:33 quark kernel: ? __die_body+0x66/0xb0
> > Aug 01 08:59:33 quark kernel: ? page_fault_oops+0x39a/0x410
> > Aug 01 08:59:33 quark kernel: ? exc_page_fault+0x59/0xa0
> > Aug 01 08:59:33 quark kernel: ? asm_exc_page_fault+0x22/0x30
> > Aug 01 08:59:33 quark kernel: ? mt7921_ipv6_addr_change
> > Aug 01 08:59:33 quark kernel: ? __try_to_del_timer_sync
> > Aug 01 08:59:33 quark kernel: ieee80211_ifa6_changed+0x68/0x120
> > Aug 01 08:59:33 quark kernel: atomic_notifier_call_chain+0x45/0xc0
> > Aug 01 08:59:33 quark kernel: addrconf_ifdown+0x521/0x7d0
> > Aug 01 08:59:33 quark kernel: addrconf_notify+0x1ed/0x4a0
> > Aug 01 08:59:33 quark kernel: raw_notifier_call_chain+0x45/0xb0
> > Aug 01 08:59:33 quark kernel: __dev_notify_flags+0xf4/0x200
> > Aug 01 08:59:33 quark kernel: dev_change_flags+0x49/0x50
> > Aug 01 08:59:33 quark kernel: do_setlink+0x49b/0x1300
> > Aug 01 08:59:33 quark kernel: ? terminate_walk+0x6b/0x100
> > Aug 01 08:59:33 quark kernel: ? __nla_validate_parse
> > Aug 01 08:59:33 quark kernel: ? filename_lookup+0xc7/0x1b0
> > Aug 01 08:59:33 quark kernel: rtnl_newlink+0xb6a/0xde0
> > Aug 01 08:59:33 quark kernel: ? __wake_up_sync_key+0x51/0x80
> > Aug 01 08:59:33 quark kernel: ? scm_destroy+0xc/0x30
> > Aug 01 08:59:33 quark kernel: ? security_capable+0x38/0x50
> > Aug 01 08:59:33 quark kernel: rtnetlink_rcv_msg+0x2dd/0x330
> > Aug 01 08:59:33 quark kernel: ? select_task_rq_fair
> > Aug 01 08:59:33 quark kernel: ? rtnetlink_bind+0x30/0x30
> > Aug 01 08:59:33 quark kernel: netlink_rcv_skb+0xb5/0xf0
> > Aug 01 08:59:33 quark kernel: netlink_unicast+0x230/0x330
> > Aug 01 08:59:33 quark kernel: netlink_sendmsg+0x3b1/0x460
> > Aug 01 08:59:33 quark kernel: ____sys_sendmsg
> > Aug 01 08:59:33 quark kernel: ? chacha_block_generic+0x6a/0x130
> > Aug 01 08:59:33 quark kernel: ___sys_sendmsg+0x282/0x2a0
> > Aug 01 08:59:33 quark kernel: ? __fget_files+0x95/0xb0
> > Aug 01 08:59:33 quark kernel: __se_sys_sendmsg+0xf4/0x120
> > Aug 01 08:59:33 quark kernel: do_syscall_64+0x7e/0x130
> > Aug 01 08:59:33 quark kernel: ? pollwake+0x52/0x60
> > Aug 01 08:59:33 quark kernel: ? do_task_dead+0x50/0x50
> > Aug 01 08:59:33 quark kernel: ? __wake_up_locked_key+0x48/0x70
> > Aug 01 08:59:33 quark kernel: ? eventfd_write+0x193/0x1b0
> > Aug 01 08:59:33 quark kernel: ? syscall_exit_to_user_mode+0x93/0xc0
> > Aug 01 08:59:33 quark kernel: ? vfs_write+0xfa/0x3d0
> > Aug 01 08:59:33 quark kernel: ? __fget_files+0x95/0xb0
> > Aug 01 08:59:33 quark kernel: ? __fget_files+0x95/0xb0
> > Aug 01 08:59:33 quark kernel: ? ksys_write+0x8f/0xb0
> > Aug 01 08:59:33 quark kernel: ? arch_exit_to_user_mode_prepare+0x11/0x50
> > Aug 01 08:59:33 quark kernel: ? syscall_exit_to_user_mode+0x93/0xc0
> > Aug 01 08:59:33 quark kernel: ? do_syscall_64+0x8a/0x130
> > Aug 01 08:59:33 quark kernel: ? syscall_exit_to_user_mode+0x93/0xc0
> > Aug 01 08:59:33 quark kernel: ? do_syscall_64+0x8a/0x130
> > Aug 01 08:59:33 quark kernel: ? do_syscall_64+0x8a/0x130
> > Aug 01 08:59:33 quark kernel: ? do_syscall_64+0x8a/0x130
> > Aug 01 08:59:33 quark kernel: ? arch_exit_to_user_mode_prepare+0x11/0x50
> > Aug 01 08:59:33 quark kernel: entry_SYSCALL_64_after_hwframe+0x4b/0x53
> > Aug 01 08:59:33 quark kernel: RIP: 0033:0x7f4264d31fae
> > Aug 01 08:59:33 quark kernel: Code: 20 89 54 24 1c 48 89 74 24 10 89
> > 7c 24 08 e8 a9 75 f7 ff 41 89 c0 8b 54 24 1c 48 8b 74 24 10 b8 2e 00
> > 00 00 8b 7c 24 08 0f 05 <48> 3d 00 f0 ff ff 77 3a 44 89 c7 48 89 44 24
> > 08 e8 fd 75 f7 ff 48
> > Aug 01 08:59:33 quark kernel: RSP: 002b:00007ffff4b1afa0 EFLAGS:
> > 00000293 ORIG_RAX: 000000000000002e
> > Aug 01 08:59:33 quark kernel: RAX: ffffffffffffffda RBX:
> > 000055c35260c570 RCX: 00007f4264d31fae
> > Aug 01 08:59:33 quark kernel: RDX: 0000000000000000 RSI:
> > 00007ffff4b1afe0 RDI: 000000000000000d
> > Aug 01 08:59:33 quark kernel: RBP: 00007ffff4b1b050 R08:
> > 0000000000000000 R09: 0000000000000000
> > Aug 01 08:59:33 quark kernel: R10: 000000000000009d R11:
> > 0000000000000293 R12: 0000000000000004
> > Aug 01 08:59:33 quark kernel: R13: 0000000000000000 R14:
> > 0000000000000000 R15: 0000000000000000
> > Aug 01 08:59:33 quark kernel: </TASK>
> > Aug 01 08:59:33 quark kernel: Modules linked in:
> > Aug 01 08:59:33 quark kernel: CR2: 0000000000000008
> > Aug 01 08:59:33 quark kernel: ---[ end trace 0000000000000000 ]---
> > Aug 01 08:59:33 quark kernel: RIP: 0010:mt7921_ipv6_addr_change
> > Aug 01 08:59:33 quark kernel: Code: 41 57 41 56 41 54 53 48 83 e4 f0
> > 48 83 ec 50 48 8b 86 70 09 00 00 0f b6 8e 90 04 00 00 4c 8d ba 68 02
> > 00 00 49 89 d6 4c 89 ff <48> 8b 58 08 88 4c 24 04 66 c7 44 24 05 00 00
> > c6 44 24 07 00 66 c7
> > Aug 01 08:59:33 quark kernel: RSP: 0018:ffffc900069373b0 EFLAGS: 00010282
> > Aug 01 08:59:33 quark kernel: RAX: 0000000000000000 RBX:
> > ffff888106740920 RCX: 0000000000000000
> > Aug 01 08:59:33 quark kernel: RDX: ffff888106854800 RSI:
> > ffff88810bb35ca0 RDI: ffff888106854a68
> > Aug 01 08:59:33 quark kernel: RBP: ffffc90006937420 R08:
> > 0000000000000000 R09: ffff888104c98200
> > Aug 01 08:59:33 quark kernel: R10: ffffffff7fff0000 R11:
> > 0000000000000020 R12: 0000000000000002
> > Aug 01 08:59:33 quark kernel: R13: 0000000000000000 R14:
> > ffff888106854800 R15: ffff888106854a68
> > Aug 01 08:59:33 quark kernel: FS: 00007f4265049400(0000)
> > GS:ffff888c2df40000(0000) knlGS:0000000000000000
> > Aug 01 08:59:33 quark kernel: CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
> > Aug 01 08:59:33 quark kernel: CR2: 0000000000000008 CR3:
> > 0000000117250000 CR4: 0000000000350ef0
>
>
> this seems to be exactly the same error as I had (It can also be actively
> provoked by "nmcli radio wifi off"), but in my case commit 479ffee68d59c599 is
> enough to fix the issue. Can you please double check that your faulty kernel
> does not include this commit. If it does and you still get the error I can post
> the fix to the list.
>
> Bert Karwatzki
479ffee68d59c599 isn't enough to fix it here, I've been applying the
second patch since it was pointed out
