2017-08-30 08:16:56

by Chiwu Yung

[permalink] [raw]
Subject: (Kernel Bug report) Make kaby lake cpu crash Ethernet

Hello Sir, i am using cpu i7 7700k kaby lake, i have no overclock the
CPU but i think i7 7700k come across with Ethernet crash issue. Don't
know why it will keeping crash.
I tested the kernel since 4.8 - 4.12 almost 2-3months and i can see
that the issue is still exist.
Kernel error then the Ethernet is stop responded after that machine
crash, i need to force reboot the computer.

PS: there are no cpu overclock , latest Bios ver. Centos7 with 4.12.8 kernel

Thank you for your help!

--------------------------------------------------------
Error Log
--------------------------------------------------------
enp:s 6: Reset adapter unexpectedly[47438.8192131 e1808e 8000:88:11.6
enp8s31f6: Reset adapter unexpectedly
[112822.759818] el008e 8800:88:1f.6 enp0s3116: Reset adapter
unexpectedly 1112838.119047] el888e 8808:80:1f.6 enp8s31f6: Reset
adapter unexpectedly
[112852-967086] e1800e 8000:00:11.6 enp8s31f6: Reset adapter
unexpectedly 1128194.823488] e1000e 0008:80:1f.6 enp8s31f6: Reset
adapter unexpectedly
[128213.991461] e1808e 0000:00:11.6 enp8s31f6: Reset adapter unexpectedly
[128228.839481] 0.800e 0000:08:1f.6 enp8s31f6: Reset adapter unexpectedly
[128283.111491] el880e 0808:08:11.6 enp0s31f6: Reset adapter unexpectedly
[123287.815631] e1800e 0000:08:11.6 enp8s31f6: Reset adapter
unexpectedly 1124878.8237221 el888e 0088:00:1f.6 enp0s3116: Reset
adapter unexpectedly
[124878.843883] kernel BUG at drivers/net/ethernet/intel/e1880e/netdev.c:38111
[124878.843832] invalid opcode: 8088 [*1] SHP
[124878.843847] Modules linked in: vfat fat x86_pkg temp_thermal
intel_powerclamp coretemp kvm_intel kvm irgbypass crctledif_pclmul
crc32_pclmul ghash_clmulni_intel pcbc aesni_intel hci_uart crypto_simd
btbcm glue_helper iltOwdt cryptd btq ca iTCO_vendor_support btintel
mxm_wmi intel_cstate intel_rapl_perf ac_i881 pcspkr shpchp sg joydev
bluetooth input_leds wmi ecdh_generic mei_me rfkill mei
intel_lpss_acpi acpi_als intel_lpss intel_pch_thermal mfd_core
kfifo_buf acpi_pad ndustrialio tcp_bbr sch_fg ip_tables xfs libcrc32c
sd_mod i915 i2c_algo_bit crc32c_intel drm_kms_helper syscopyarea
e1000e sysfillrect ahci sysimgblt fb_sys_fops libahci ptp pps_core
libata drm video i2c_hid dm_mirror dm_region_hash dm_log dm_mod dax
[124878.844060] CPU: 8 PID: 25374 Comm: kworker/0:2 Tainted: G W
4.12.8-1.e17.elrepo.x86_64 *1 [124878.844890] Hardware name: To Be
Filled By O.E.M. To Be Filled By 0.E.M./2170 Pro4, BIOS P7.38
11/28/2016
[124878.844127] Workqueue: events e1000_reset_task [e1000e]
1124878.844146] task: Ifff88869e482e80 task.stack: ff1fc90820044800
1124878.844170] RIP: 0010:e1080_flush_desc rings+8x2ea/8x2f8 [e1000e]
[124878.844191] RSP: 0018:f11fc900213047d68 EFLAGS: 00810202
[124878.844289] MX: 0000000000000884 RBX: ffff88084b42c8c0 RCX: 0080000000080099
[124878.844233] RDX: 0000000000000084 RSI: 8000008000800246 RDI:
0000880000000246
[124878.844256] RBP: 11ffc90028047d98 R08: 0000080000088002 R09:
ffffc90020047d1c
[124878.844288] R18: 00008888808008fe R11: 88000008008083c7 R12:
ffff8808517a3880
[124878.844384] R13: 111f88884b42cd78 R14: 000000003103f01a R15:
0000000004008000
[124878.844328] FS: 888888808e88eee0(088e) GS:aff888875c0000010000]
knlGS:0008808808000000
[124878.844354] CS: 0010 DS: 0000 ES: 0000 C1t8: 0000000088058033
[124878.844374] CR2: 08007f7031eb8808 CR3: 8080080001c09000 CR4:
00000000003406f0
[124878.844397] DR8: 0000808800000008 DR1: 0000000080000000 DR2:
0800000000000000
[124878.844421] DR3: 0080000000880888 DR6: 80000880fffeO110 DR7:
0000008000008400
[124878.844444] Call Trace: 1124878.844468] e1880e_reset*8x4b4/8x750 [e1000e]
[124878.844479] el080e_down+8xle3/0x210 [e1808e]
[124878.844497] el080e_reinit_locked+0x31/0x60 [e1880e]
[124878.844518] e1008_reset_task+8x32/0x68 [e1800e]
[124878.844536] process_one_work+8x149/8x368
[124878.844552] worker_thread+8x4d/8x3c0
[124878.8445661 kthread+8x109/0x148
[124878.844579] ? rescuer thread+8x300/0x380
[124878.844594] ? kthread_park+8x68/0x60
[124878.844689] ? do_syscall_64+0x67/0x158
[124878.844624] ret from fork+Ox25/0x30
[124878.844637] Code: 98 e9 Be ff ff ff 4c 89 of e8 d3 fc ff ff Of if
00 e9 14 fe ff ff 31 c0 45 31 f6 66 41 89 44 24 20 e9 71 fe ff ff e8
36 73 ed e0 <Of> IA Of 11 40 00 Of 11 44 00 00 55 48 89 e5 41 57 41 56
41 55
[124878.844716] RIP: e1008_flush_desc_rings+Bx2ea/Ox2f8 [e1000e] RSP:
ffffc98020047d68
[124878.844760] ---[ end trace 4ce09c8c366f79e7 ]-
[124878.844777] Kernel panic - not syncing: Fatal exception
[124878.844825] Kernel Offset: disabled
[124878.844833] ---[ end Kernel panic - not syncing : Fatal exception


2017-08-30 08:24:05

by Greg Kroah-Hartman

[permalink] [raw]
Subject: Re: (Kernel Bug report) Make kaby lake cpu crash Ethernet

On Wed, Aug 30, 2017 at 04:16:54PM +0800, Chiwu Yung wrote:
> Hello Sir, i am using cpu i7 7700k kaby lake, i have no overclock the
> CPU but i think i7 7700k come across with Ethernet crash issue. Don't
> know why it will keeping crash.
> I tested the kernel since 4.8 - 4.12 almost 2-3months and i can see
> that the issue is still exist.
> Kernel error then the Ethernet is stop responded after that machine
> crash, i need to force reboot the computer.
>
> PS: there are no cpu overclock , latest Bios ver. Centos7 with 4.12.8 kernel
>
> Thank you for your help!
>
> --------------------------------------------------------
> Error Log
> --------------------------------------------------------
> enp:s 6: Reset adapter unexpectedly[47438.8192131 e1808e 8000:88:11.6
> enp8s31f6: Reset adapter unexpectedly
> [112822.759818] el008e 8800:88:1f.6 enp0s3116: Reset adapter
> unexpectedly 1112838.119047] el888e 8808:80:1f.6 enp8s31f6: Reset
> adapter unexpectedly
> [112852-967086] e1800e 8000:00:11.6 enp8s31f6: Reset adapter
> unexpectedly 1128194.823488] e1000e 0008:80:1f.6 enp8s31f6: Reset
> adapter unexpectedly
> [128213.991461] e1808e 0000:00:11.6 enp8s31f6: Reset adapter unexpectedly
> [128228.839481] 0.800e 0000:08:1f.6 enp8s31f6: Reset adapter unexpectedly
> [128283.111491] el880e 0808:08:11.6 enp0s31f6: Reset adapter unexpectedly
> [123287.815631] e1800e 0000:08:11.6 enp8s31f6: Reset adapter
> unexpectedly 1124878.8237221 el888e 0088:00:1f.6 enp0s3116: Reset
> adapter unexpectedly
> [124878.843883] kernel BUG at drivers/net/ethernet/intel/e1880e/netdev.c:38111

That driver is not in the kernel.org releases, so there's nothing we can
do about it. Please contact the authors of the driver you are using
here, wherever you downloaded it from, and they can help you out.

good luck!

greg k-h

2017-08-30 09:33:32

by Andrey Melnikov

[permalink] [raw]
Subject: Re: (Kernel Bug report) Make kaby lake cpu crash Ethernet

Chiwu Yung <[email protected]> wrote:
> Hello Sir, i am using cpu i7 7700k kaby lake, i have no overclock the
> CPU but i think i7 7700k come across with Ethernet crash issue. Don't
> know why it will keeping crash.
> I tested the kernel since 4.8 - 4.12 almost 2-3months and i can see
> that the issue is still exist.
> Kernel error then the Ethernet is stop responded after that machine
> crash, i need to force reboot the computer.

> PS: there are no cpu overclock , latest Bios ver. Centos7 with 4.12.8 kernel

> Thank you for your help!

> --------------------------------------------------------
> Error Log
> --------------------------------------------------------
> enp:s 6: Reset adapter unexpectedly[47438.8192131 e1808e 8000:88:11.6
> enp8s31f6: Reset adapter unexpectedly
> [112822.759818] el008e 8800:88:1f.6 enp0s3116: Reset adapter
> unexpectedly 1112838.119047] el888e 8808:80:1f.6 enp8s31f6: Reset
> adapter unexpectedly
> [112852-967086] e1800e 8000:00:11.6 enp8s31f6: Reset adapter
> unexpectedly 1128194.823488] e1000e 0008:80:1f.6 enp8s31f6: Reset
> adapter unexpectedly
> [128213.991461] e1808e 0000:00:11.6 enp8s31f6: Reset adapter unexpectedly
> [128228.839481] 0.800e 0000:08:1f.6 enp8s31f6: Reset adapter unexpectedly
> [128283.111491] el880e 0808:08:11.6 enp0s31f6: Reset adapter unexpectedly
> [123287.815631] e1800e 0000:08:11.6 enp8s31f6: Reset adapter
> unexpectedly 1124878.8237221 el888e 0088:00:1f.6 enp0s3116: Reset
> adapter unexpectedly

Check memory. Note driver name stranformation in logs:

'el008e 8800:88:1f.6 enp0s3116' vs 'el888e 8808:80:1f.6 enp8s31f6'
'e1800e 8000:00:11.6 enp8s31f6' vs 'e1000e 0008:80:1f.6 enp8s31f6'
'e1808e 0000:00:11.6 enp8s31f6' vs '0.800e 0000:08:1f.6 enp8s31f6'
'el880e 0808:08:11.6 enp0s31f6' vs 'e1800e 0000:08:11.6 enp8s31f6'