[BUG] AB-BA deadlock between net and led-trigger module

[Shiji Yang]

The OpenWrt community reports that sometimes devices fail to start[1]
on 5.15 kernel. After further tracking, this is caused by a AB-BA
deadlock which can be reproduced in at least 5.15, 6.6, 6.12 and latest
6.18 LTS kenrel.

Stack tracing on 6.12 kernel:
```
Task1 "netifd" is used to start/restart the network:

[ 1361.967916] task:netifd          state:D stack:0     pid:4743  tgid:4743  ppid:1      flags:0x08100000
[ 1361.977269] Stack : 00000001 00000001 00000006 800bf464 817a0cb0 800b67ac 00000001 83261b20
[ 1361.985668]         83261a54 00000000 83261aac 000007ef 00000000 80c04d74 00000000 00000002
[ 1361.994138]         83aba760 80ce0000 83261af8 00000002 83261af8 80cd0000 00000002 80d8058c
[ 1362.002582]         80cc0000 809bae70 00000002 80d80000 80cd0000 80d80568 00000001 00000000
[ 1362.011033]         809baecc 83261af8 00000002 80d80560 80d80568 809bb11c 809c09c8 806614d0
[ 1362.019484]         ...
[ 1362.021942] Call Trace:
[ 1362.024380] [<809ba6fc>] __schedule+0x504/0xc28
[ 1362.028974] [<809bae70>] schedule+0x50/0x190
[ 1362.033251] [<809bb11c>] schedule_preempt_disabled+0x1c/0x34
[ 1362.038958] [<809c0b94>] rwsem_down_write_slowpath+0x240/0x7f8
[ 1362.044789] [<809c11c0>] down_write+0x74/0x90
[ 1362.049207] [<8054e4b8>] led_trigger_register+0x5c/0x1fc		<-- Trying to get lock "triggers_list_lock" via down_write(&triggers_list_lock);
[ 1362.054536] [<80662830>] phy_led_triggers_register+0xd0/0x234
[ 1362.060329] [<8065e200>] phy_attach_direct+0x33c/0x40c
[ 1362.065489] [<80651fc4>] phylink_fwnode_phy_connect+0x15c/0x23c
[ 1362.071480] [<8066ee18>] mtk_open+0x7c/0xba0
[ 1362.075849] [<806d714c>] __dev_open+0x280/0x2b0
[ 1362.080384] [<806d7668>] __dev_change_flags+0x244/0x24c
[ 1362.085598] [<806d7698>] dev_change_flags+0x28/0x78
[ 1362.090528] [<807150e4>] dev_ioctl+0x4c0/0x654			<-- Hold lock "rtnl_mutex" by calling rtnl_lock();
[ 1362.094985] [<80694360>] sock_ioctl+0x2f4/0x4e0
[ 1362.099567] [<802e9c4c>] sys_ioctl+0x32c/0xd8c
[ 1362.104022] [<80014504>] syscall_common+0x34/0x58


Task2 "led" is used to set the led-trigger "netdev" for a GPIO LED:

[ 1362.110308] task:led             state:D stack:0     pid:4943  tgid:4943  ppid:1      flags:0x08100002
[ 1362.119656] Stack : 809bf818 80ce3ce4 80d47fa8 fffff000 00000000 809bf840 00000001 800bb264
[ 1362.128115]         00000000 00000000 80ce0000 80ce0000 00000000 80c04d7c 00000000 00000002
[ 1362.136565]         83b50d20 80ce0000 834e1cec 80d8c750 00000002 80cd7380 80da5f2c 00000000
[ 1362.144962]         00000000 809bae70 00000000 00000000 80d8c750 80d8c750 00000001 00000000
[ 1362.153412]         809baecc 00000002 80cd7380 80d8c74c 00000000 809bb11c 00000000 834e1cec
[ 1362.161868]         ...
[ 1362.164327] Call Trace:
[ 1362.166835] [<809ba6fc>] __schedule+0x504/0xc28
[ 1362.171385] [<809bae70>] schedule+0x50/0x190
[ 1362.175651] [<809bb11c>] schedule_preempt_disabled+0x1c/0x34
[ 1362.181361] [<809bdd48>] __mutex_lock+0x310/0x940
[ 1362.186132] [<809be394>] mutex_lock_nested+0x1c/0x28
[ 1362.191101] [<806c2640>] register_netdevice_notifier+0x60/0x168	<-- Trying to get lock "rtnl_mutex" via rtnl_lock();
[ 1362.197073] [<805504ac>] netdev_trig_activate+0x194/0x1e4
[ 1362.202490] [<8054e28c>] led_trigger_set+0x1d4/0x360			<-- Hold lock "triggers_list_lock" by down_read(&triggers_list_lock);
[ 1362.207511] [<8054eb38>] led_trigger_write+0xd8/0x14c
[ 1362.212566] [<80381d98>] sysfs_kf_bin_write+0x80/0xbc
[ 1362.217688] [<8037fcd8>] kernfs_fop_write_iter+0x17c/0x28c
[ 1362.223174] [<802cbd70>] vfs_write+0x21c/0x3c4
[ 1362.227712] [<802cc0c4>] ksys_write+0x78/0x12c
[ 1362.232164] [<80014504>] syscall_common+0x34/0x58
```

When the above two tasks are created at the same time, there is a
probability that it will cause the network and LED to fail to
initialize.

[1] https://github.com/openwrt/openwrt/issues/18472

Regards,
Shiji Yang