Skip to main content ↑ ↓
Topic: Kernel trouble with my ethernet (Read 2656 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

Kernel trouble with my ethernet

Code: [Select]
[ 1031.036118] ------------[ cut here ]------------
[ 1031.036138] WARNING: CPU: 2 PID: 0 at net/sched/sch_generic.c:316 dev_watchdog+0x221/0x230
[ 1031.036140] NETDEV WATCHDOG: eth1 (e1000e): transmit queue 0 timed out
[ 1031.036142] Modules linked in: ipt_MASQUERADE nf_nat_masquerade_ipv4 iptable_nat nf_conntrack_ipv4 nf_defrag_ipv4 nf_nat_ipv4 nf_nat nf_conntrack iptable_filter ip_tables x_tables nls_iso8859_1 nls_cp437 vfat fat vboxvideo(O) ttm snd_hda_codec_hdmi mei_wdt snd_hda_codec_realtek snd_hda_codec_generic intel_rapl arc4 x86_pkg_temp_thermal intel_powerclamp coretemp rtl8723ae btcoexist kvm_intel iTCO_wdt rtl8723_common kvm rtl_pci irqbypass rtlwifi crct10dif_pclmul ghash_clmulni_intel iTCO_vendor_support ppdev aesni_intel aes_x86_64 lrw gf128mul i915 mac80211 snd_hda_intel glue_helper ablk_helper cryptd snd_hda_codec drm_kms_helper intel_cstate intel_rapl_perf cfg80211 psmouse drm mei_me pcspkr snd_hda_core mousedev btusb btrtl btbcm i2c_i801 btintel snd_hwdep i2c_smbus evdev bluetooth e1000e snd_pcm
[ 1031.036232]  input_leds intel_gtt r8169 syscopyarea sysfillrect mii sysimgblt ptp lpc_ich snd_timer fb_sys_fops snd crypto_user rfkill soundcore mei mac_hid i2c_algo_bit pps_core winbond_cir acpi_als kfifo_buf rc_core industrialio led_class shpchp parport_pc parport video thermal fjes fan button ac tpm_tis tpm_tis_core tpm ext4 crc16 jbd2 fscrypto mbcache sd_mod hid_generic usbhid hid serio_raw atkbd libps2 ahci libahci crc32_pclmul crc32c_intel libata xhci_pci ehci_pci xhci_hcd scsi_mod ehci_hcd usbcore usb_common i8042 serio
[ 1031.036312] CPU: 2 PID: 0 Comm: swapper/2 Tainted: G           O    4.9.59-1-lts #1
[ 1031.036314] Hardware name: CompuLab Intense-PC/Intense-PC, BIOS IPC_2.2.400.1 X64 08/31/2014
[ 1031.036320]  ffff88043e283d98 ffffffff812fa08d ffff88043e283de8 0000000000000000
[ 1031.036327]  ffff88043e283dd8 ffffffff8107cceb 0000013c3e297de8 0000000000000000
[ 1031.036332]  ffff88042b3c8000 0000000000000002 0000000000000001 ffff88042b3c8000
[ 1031.036338] Call Trace:
[ 1031.036342]  <IRQ> 
[ 1031.036353]  [<ffffffff812fa08d>] dump_stack+0x63/0x86
[ 1031.036362]  [<ffffffff8107cceb>] __warn+0xcb/0xf0
[ 1031.036367]  [<ffffffff8107cd6f>] warn_slowpath_fmt+0x5f/0x80
[ 1031.036375]  [<ffffffff81512961>] dev_watchdog+0x221/0x230
[ 1031.036381]  [<ffffffff81512740>] ? qdisc_rcu_free+0x40/0x40
[ 1031.036387]  [<ffffffff810ec443>] call_timer_fn+0x33/0x120
[ 1031.036391]  [<ffffffff810ec600>] expire_timers+0xd0/0x120
[ 1031.036396]  [<ffffffff810ec7a1>] run_timer_softirq+0x151/0x190
[ 1031.036403]  [<ffffffff810519e6>] ? lapic_next_deadline+0x26/0x30
[ 1031.036410]  [<ffffffff810faded>] ? clockevents_program_event+0xbd/0x110
[ 1031.036419]  [<ffffffff816002e7>] __do_softirq+0xd7/0x2d0
[ 1031.036426]  [<ffffffff81082966>] irq_exit+0xb6/0xc0
[ 1031.036432]  [<ffffffff81600122>] smp_apic_timer_interrupt+0x42/0x50
[ 1031.036437]  [<ffffffff815ff432>] apic_timer_interrupt+0x82/0x90
[ 1031.036439]  <EOI> 
[ 1031.036449]  [<ffffffff814b231e>] ? cpuidle_enter_state+0x13e/0x2b0
[ 1031.036455]  [<ffffffff814b24c7>] cpuidle_enter+0x17/0x20
[ 1031.036462]  [<ffffffff810c1e23>] call_cpuidle+0x23/0x40
[ 1031.036468]  [<ffffffff810c209c>] cpu_startup_entry+0x15c/0x240
[ 1031.036473]  [<ffffffff81050571>] start_secondary+0x151/0x190
[ 1031.036477] ---[ end trace 1c75b74fd3bbc1a0 ]---
[ 1031.036536] e1000e 0000:00:19.0 eth1: Reset adapter unexpectedly
[ 1032.812393] e1000e: eth1 NIC Link is Up 100 Mbps Full Duplex, Flow Control: Rx/Tx
this reatedly happens

Re: Kernel trouble with my ethernet

Reply #2 –
@mrbrklyn  Duuuuude... please use the code boxes to put stuff like that. Please. Forum etiquette. Simple stuff.

Re: Kernel trouble with my ethernet

Reply #3 –
Quote from: SGOrava – on
Blunt guess: this might be related to interrupted / failed update
https://artixlinux.org/forum/index.php?topic=200.msg1546;topicseen


OK - I did this.  And there is a new kernel upgrade this morning.  FWIW, this is the machine and it does have two different built in LAN connectors
Intense PC models specifications

http://www.fit-pc.com/web/products/intense-pc/

2x GbE LAN ports
LAN1: Intel 82579 GbE PHY (MAC integrated into the chipset) (RJ-45)
LAN2: Realtek RTL8111F-CG GbE controller (RJ-45)
LAN3-6: N/A

I am not even sure what modules I am looking at and there is no /etc/modules.conf even on the file system.

Re: Kernel trouble with my ethernet

Reply #4 –
@mrbrklyn
You'll probably need the realtek drivers from the linux-firmware package to get LAN2 going.
Not sure what Realtek is included/accessible in base kernel, but I find I have to grab firmware drivers to make work in the past.

Re: Kernel trouble with my ethernet

Reply #5 –
It that a standard pacman library?  I really need binary blobs to make the network card stable?

Re: Kernel trouble with my ethernet

Reply #6 –
Code: [Select]
[www3 ~]# sudo pacman -Ss firmware
system/b43-fwcutter 019-1
    firmware extractor for the b43 kernel module
system/efibootmgr 15-1 [installed]
    Tool to modify UEFI Firmware Boot Manager Variables
system/ipw2100-fw 1.3-8
    Intel Centrino Drivers firmware for IPW2100
system/ipw2200-fw 3.1-6
    Firmware for the Intel PRO/Wireless 2200BG
system/linux-firmware 20171009.bf04291-1 [installed]
    Firmware files for Linux
system/zd1211-firmware 1.5-1
    Firmware for the in-kernel26 zd1211rw wireless driver
extra/alsa-firmware 1.0.29-1 [installed]
    ALSA firmware package
extra/bluez-firmware 1.2-8
    Firmwares for Broadcom BCM203x and STLC2300 Bluetooth chips
extra/ovmf 1:r21243.3858b4a1ff-1
    Tianocore UEFI firmware for qemu.
community/bladerf 1.8.0-1
    Driver, userspace, fpga & firmware for the bladeRF SDR.
community/cura-binary-data 3.0.3-1
    Binary data (firmwares and such) for cura
community/dfu-util 0.9-1
    Tool intended to download and upload firmware using DFU protocol to devices
    connected over USB
community/fwupd 1.0.0-1
    A simple daemon to allow session software to update firmware
community/fwupdate 9-2
    Tools for using the ESRT and UpdateCapsule() to apply firmware updates
community/heimdall 1.4.2-1
    Tool suite used to flash firmware (aka ROMs) onto Samsung mobile devices
community/libuhd-firmware 3.10.2.0-1
    Universal Software Radio Peripheral (USRP) userspace firmware
community/sane-gt68xx-firmware 1.0-7
    gt68xx-based scanners firmwares
community/stlink 1.4.0-1
    Firmware programmer for STM32 STLINK v1/v2 protocol

Re: Kernel trouble with my ethernet

Reply #7 –
Quote from: mrbrklyn – on
Code: [Select]
[www3 ~]# sudo pacman -Ss firmware
...
system/linux-firmware 20171009.bf04291-1 [installed]
    Firmware files for Linux
...
That was the package I was alluding to. Looks like you have it installed. My best guess is to re-install the package, reboot, and then observe. If the problem persists we try something else. But we would know you have the latest firmware and could strike that off of the list of possibilities.

Re: Kernel trouble with my ethernet

Reply #8 –
Quote from: scottfurry – on
@mrbrklyn
You'll probably need the realtek drivers from the linux-firmware package to get LAN2 going.
Not sure what Realtek is included/accessible in base kernel, but I find I have to grab firmware drivers to make work in the past.



Actually, LAN2 is working.  It is the Intel Adapter which seems to be freaking out.  I'm using a 10/100 cat 5 cable on it with a 10/100 switch

Code: [Select]
www3 ~]# dmesg|grep eth1
[   10.313646] e1000e 0000:00:19.0 eth1: (PCI Express:2.5GT/s:Width x1) 00:01:c0:1a:f9:c5
[   10.313649] e1000e 0000:00:19.0 eth1: Intel(R) PRO/1000 Network Connection
[   10.313688] e1000e 0000:00:19.0 eth1: MAC: 10, PHY: 11, PBA No: FFFFFF-0FF
[   14.677822] IPv6: ADDRCONF(NETDEV_UP): eth1: link is not ready
[   16.203391] e1000e: eth1 NIC Link is Up 100 Mbps Full Duplex, Flow Control: Rx/Tx
[   16.203397] e1000e 0000:00:19.0 eth1: 10/100 speed: disabling TSO
[   16.203440] IPv6: ADDRCONF(NETDEV_CHANGE): eth1: link becomes ready
[  996.059388] NETDEV WATCHDOG: eth1 (e1000e): transmit queue 0 timed out
[  996.060077] e1000e 0000:00:19.0 eth1: Reset adapter unexpectedly
[  997.825577] e1000e: eth1 NIC Link is Up 100 Mbps Full Duplex, Flow Control: Rx/Tx
[ 1014.059529] e1000e 0000:00:19.0 eth1: Reset adapter unexpectedly

Re: Kernel trouble with my ethernet

Reply #10 –
Remember we were talking about the "possible" advantages of long named devices a while back?  Could this be an example of such an advantage?  Is it possible that some script has criss-crossed eth0/eth1 ?
What if you were to disable eth0, physically or virtually, and allow the problematic one to reset itself as eth0.  Then add the other card back as eth1.  It is probably a wild shot in the dark and a non-kosher way of doing things, but hey!
 

Re: Kernel trouble with my ethernet

Reply #11 –
Quote from: fungalnet – on
Remember we were talking about the "possible" advantages of long named devices a while back?  Could this be an example of such an advantage?  Is it possible that some script has criss-crossed eth0/eth1 ?
What if you were to disable eth0, physically or virtually, and allow the problematic one to reset itself as eth0.  Then add the other card back as eth1.  It is probably a wild shot in the dark and a non-kosher way of doing things, but hey!

At the moment it  looks like   ethtool -K eth1 tx off rx off - has fixed it the problem and made it stable.  I'm not sure what that even does.