Onboard eth0 drops almost every boot randomly

Guellekelch · 2024-05-04 20:13:57

Hi,

I just installed arch a few days ago (first time I use linux at home). From time to time my ethernet connection just stops working. I finally figured out to look into the dmesg and saw this:

[   78.620590] igc 0000:09:00.0 eth0: PCIe link lost, device now detached
[   78.620598] ------------[ cut here ]------------
[   78.620600] igc: Failed to read reg 0xc030!
[   78.620617] WARNING: CPU: 0 PID: 9 at drivers/net/ethernet/intel/igc/igc_main.c:6643 igc_rd32+0x8d/0xa0 [igc]
[   78.620627] Modules linked in: snd_seq_dummy snd_hrtimer snd_seq hid_logitech_hidpp mousedev btusb btrtl intel_rapl_msr btintel intel_rapl_common btbcm btmtk bluetooth ecdh_generic crc16 hid_logitech_dj kvm_amd amdgpu kvm irqbypass crct10dif_pclmul crc32_pclmul snd_hda_codec_hdmi polyval_clmulni amdxcp snd_usb_audio snd_hda_intel polyval_generic drm_exec snd_intel_dspcfg gf128mul snd_usbmidi_lib gpu_sched snd_intel_sdw_acpi ghash_clmulni_intel sha512_ssse3 snd_ump drm_buddy snd_hda_codec vfat sha1_ssse3 snd_rawmidi asus_nb_wmi i2c_algo_bit eeepc_wmi fat aesni_intel drm_suballoc_helper snd_seq_device snd_hda_core asus_wmi drm_ttm_helper mc crypto_simd snd_hwdep ledtrig_audio ttm platform_profile cryptd igc snd_pcm i8042 drm_display_helper sp5100_tco ptp snd_timer sparse_keymap ccp serio rapl wmi_bmof cec pcspkr i2c_piix4 k10temp pps_core snd cfg80211 soundcore joydev gpio_amdpt gpio_generic rfkill mac_hid i2c_dev crypto_user loop dm_mod nfnetlink ip_tables x_tables hid_generic usbhid btrfs blake2b_generic libcrc32c
[   78.620690]  crc32c_generic xor raid6_pq crc32c_intel nvme sha256_ssse3 sr_mod nvme_core xhci_pci cdrom xhci_pci_renesas nvme_auth nvidia_drm(POE) nvidia_uvm(POE) nvidia_modeset(POE) video wmi nvidia(POE)
[   78.620706] CPU: 0 PID: 9 Comm: kworker/0:1 Tainted: P           OE      6.8.9-arch1-1 #1 b1154d98cdfe0792477fb31da8bf552e17e27d3c
[   78.620708] Hardware name: ASUS System Product Name/ROG STRIX X670E-A GAMING WIFI, BIOS 1905 02/05/2024
[   78.620710] Workqueue: events igc_watchdog_task [igc]
[   78.620717] RIP: 0010:igc_rd32+0x8d/0xa0 [igc]
[   78.620724] Code: 48 c7 c6 48 f9 dc c4 e8 c1 98 5a db 48 8b bb 28 ff ff ff e8 05 da 1a db 84 c0 74 bc 89 ee 48 c7 c7 70 f9 dc c4 e8 43 b8 b0 da <0f> 0b eb aa b8 ff ff ff ff e9 20 5b 95 db 0f 1f 44 00 00 90 90 90
[   78.620725] RSP: 0018:ffffc245c0147dc8 EFLAGS: 00010286
[   78.620727] RAX: 0000000000000000 RBX: ffff9da088104cb8 RCX: 0000000000000027
[   78.620729] RDX: ffff9da7be4219c8 RSI: 0000000000000001 RDI: ffff9da7be4219c0
[   78.620730] RBP: 000000000000c030 R08: 0000000000000000 R09: ffffc245c0147c58
[   78.620731] R10: ffffc245c0147c50 R11: 0000000000000003 R12: ffff9da088104000
[   78.620732] R13: 0000000000000000 R14: ffff9da0a46d9d80 R15: 000000000000c030
[   78.620733] FS:  0000000000000000(0000) GS:ffff9da7be400000(0000) knlGS:0000000000000000
[   78.620735] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[   78.620736] CR2: 00001e4c00c1b0c0 CR3: 00000004c7820000 CR4: 0000000000f50ef0
[   78.620738] PKRU: 55555554
[   78.620739] Call Trace:
[   78.620741]  <TASK>
[   78.620742]  ? igc_rd32+0x8d/0xa0 [igc a6aac84d96cc5465ff6105ee0df378744db3746c]
[   78.620749]  ? __warn+0x81/0x130
[   78.620754]  ? igc_rd32+0x8d/0xa0 [igc a6aac84d96cc5465ff6105ee0df378744db3746c]
[   78.620761]  ? report_bug+0x171/0x1a0
[   78.620764]  ? prb_read_valid+0x1b/0x30
[   78.620768]  ? srso_alias_return_thunk+0x5/0xfbef5
[   78.620771]  ? handle_bug+0x3c/0x80
[   78.620774]  ? exc_invalid_op+0x17/0x70
[   78.620776]  ? asm_exc_invalid_op+0x1a/0x20
[   78.620781]  ? igc_rd32+0x8d/0xa0 [igc a6aac84d96cc5465ff6105ee0df378744db3746c]
[   78.620787]  ? igc_rd32+0x8d/0xa0 [igc a6aac84d96cc5465ff6105ee0df378744db3746c]
[   78.620793]  igc_update_stats+0x8a/0x6d0 [igc a6aac84d96cc5465ff6105ee0df378744db3746c]
[   78.620800]  igc_watchdog_task+0xa0/0x4b0 [igc a6aac84d96cc5465ff6105ee0df378744db3746c]
[   78.620808]  process_one_work+0x178/0x350
[   78.620811]  worker_thread+0x30f/0x450
[   78.620815]  ? __pfx_worker_thread+0x10/0x10
[   78.620817]  kthread+0xe5/0x120
[   78.620820]  ? __pfx_kthread+0x10/0x10
[   78.620823]  ret_from_fork+0x31/0x50
[   78.620826]  ? __pfx_kthread+0x10/0x10
[   78.620829]  ret_from_fork_asm+0x1b/0x30
[   78.620834]  </TASK>
[   78.620835] ---[ end trace 0000000000000000 ]---

So when I understand this correct then something with my intel ethernet drivers might be not correct?
It happens at super random times, but most often in the first few minutes after boot.
Normally I reboot but today I tried the following

echo "1" > /sys/bus/pci/rescan

After that I got the network back.

Do you have suggestions what I could try to stop it from happening?

seth · 2024-05-05 06:49:00

Could be aspm, the problem is "[ 78.620590] igc 0000:09:00.0 eth0: PCIe link lost, device now detached", the interesting part happens before that.
(Also because "echo "1" > /sys/bus/pci/rescan" suggests this is more a bus than a device issue)

In case there's a parallel windows installation, see the 3rd link below. Mandatory.
Disable it (it's NOT the BIOS setting!) and reboot windows and linux twice for voodo reasons.

Guellekelch · 2024-05-05 11:29:48

Today it happened after 250 seconds. Unfortunately in the dmesg there is nothing happening immediately before the "eth0 detached" line:

[   13.827102] usb 3-7: Product: AURA LED Controller
[   13.827103] usb 3-7: Manufacturer: AsusTek Computer Inc.
[   13.827105] usb 3-7: SerialNumber: 9876543210
[   13.841956] hid-generic 0003:0B05:19AF.000C: hiddev99,hidraw11: USB HID v1.11 Device [AsusTek Computer Inc. AURA LED Controller] on usb-0000:0f:00.0-7/input2
[   14.391848] block nvme1n1: No UUID available providing old NGUID
[   19.452956] systemd-journald[427]: /var/log/journal/754b4a54dcf94f1f933614aa7b7fe4cb/user-1000.journal: Journal file uses a different sequence number ID, rotating.
[   20.445570] logitech-hidpp-device 0003:046D:407F.0009: HID++ 4.2 device connected.
[   28.132019] Bluetooth: hci0: Device setup in 20320636 usecs
[   28.132025] Bluetooth: hci0: HCI Enhanced Setup Synchronous Connection command is advertised, but not supported.
[   28.407451] Bluetooth: hci0: AOSP extensions version v1.00
[   28.407457] Bluetooth: hci0: AOSP quality report is supported
[  250.567420] igc 0000:09:00.0 eth0: PCIe link lost, device now detached
[  250.567429] ------------[ cut here ]------------
[  250.567430] igc: Failed to read reg 0xc030!
[  250.567448] WARNING: CPU: 0 PID: 9 at drivers/net/ethernet/intel/igc/igc_main.c:6643 igc_rd32+0x8d/0xa0 [igc]

Thanks for the suggestion I will try it out soon.

seth · 2024-05-05 11:32:01

Try "pcie_aspm=off", https://wiki.archlinux.org/title/Kernel_parameters

Guellekelch · 2024-05-15 17:57:13

I stumbled upon this discussion:
https://www.reddit.com/r/buildapc/comme … Findex.php

Apparently this is an issue with my ASUS mainboard... I went for the systemd service "solution" which scans the dmesg and just restarts the eth if it is happening.

Arch Linux

#1 2024-05-04 20:13:57

Onboard eth0 drops almost every boot randomly

#2 2024-05-05 06:49:00

Re: Onboard eth0 drops almost every boot randomly

#3 2024-05-05 11:29:48

Re: Onboard eth0 drops almost every boot randomly

#4 2024-05-05 11:32:01

Re: Onboard eth0 drops almost every boot randomly

#5 2024-05-15 17:57:13

Re: Onboard eth0 drops almost every boot randomly

Board footer