Date:	Mon, 16 Feb 2015 12:53:15 +0100
From:	Peter Maloney <peter.maloney@...ckmann-consult.de>
To:	linux-kernel@...r.kernel.org
Subject: Re: CentOS 7.0, e1000e driver issue/bug - "Detected Hardware Unit
 Hang:" "Reset adapter unexpectedly"

FYI this seems fixed since 2015-02-12 when I ran this fix:

    sudo ethtool -K com gso off gro off tso off

Which I found at:
http://serverfault.com/questions/616485/e1000e-reset-adapter-unexpectedly-detected-hardware-unit-hang

Here's a log of how many times it happened each day, 0 times since the
fix on 2015-02-12, always at least once every day, so I think it's
conclusive.

    12 2015-01-18
    17 2015-01-19
    10 2015-01-20
    11 2015-01-21
    21 2015-01-22
    17 2015-01-23
    20 2015-01-24
    20 2015-01-25
    16 2015-01-26
    15 2015-01-27
    20 2015-01-28
    14 2015-01-29
    10 2015-01-30
    20 2015-01-31
    21 2015-02-01
    20 2015-02-02
     5 2015-02-03
    11 2015-02-04
    67 2015-02-05
    14 2015-02-06
    22 2015-02-07
    27 2015-02-08
    16 2015-02-09
     8 2015-02-10
    39 2015-02-11
    24 2015-02-12



On 01/27/2015 11:27 AM, Peter Maloney wrote:
> Hi, I have a problem on a machine running CentOS 7.0, where the
> kernel/e1000e reports things like "Detected Hardware Unit Hang:" and
> "Reset adapter unexpectedly". The kernel version is
> 3.10.0-123.13.2.el7.x86_64.
>
> I had a similar issue years ago with the same machine running openSUSE
> 12.3 with kernel 3.7.10, and downgrading to 3.4.47 fixed it completely.
> At that time, I found this bug reported in fedora, marked as WONTFIX due
> to the fedora release hitting EoL
> https://bugzilla.redhat.com/show_bug.cgi?id=785806 and the dmesg output
> looks similar. And recently I found this old bug for CentOS 6.
> http://bugs.centos.org/view.php?id=6517 to which I replied but haven't
> seen any activity there for a week.
>
> Years ago on openSUSE 12.3 with kernel 3.7.10, this would eventually
> make the network fail completely requiring a reboot. So far (up 12 days)
> the machine with 3.10.x hasn't been disconnected long enough to be
> noticeable.
>
> I have seen
> http://git.kernel.org/cgit/linux/kernel/git/davem/net-next.git/commit/?id=09357b00255c233705b1cf6d76a8d147340545b8
> as mentioned in the fedora bug page, and it appears to already be
> applied to this kernel.
>
> *Here are the details for the machine with the problem:
>
> *root@...hine:~ # lsb_release -a
> LSB Version: :core-4.1-amd64:core-4.1-noarch
> Distributor ID: CentOS
> Description: CentOS Linux release 7.0.1406 (Core)
> Release: 7.0.1406
> Codename: Core
>
> root@...hine:~ # uname -a
> Linux machine.bc.local 3.10.0-123.13.2.el7.x86_64 #1 SMP Thu Dec 18
> 14:09:13 UTC 2014 x86_64 x86_64 x86_64 GNU/Linux
>
> root@...hine:~ # lspci -v
> ...
> 00:19.0 Ethernet controller: Intel Corporation 82579LM Gigabit Network
> Connection (rev 05)
>         Subsystem: Super Micro Computer Inc Device 1502
>         Flags: bus master, fast devsel, latency 0, IRQ 42
>         Memory at dfd00000 (32-bit, non-prefetchable) [size=128K]
>         Memory at dfd25000 (32-bit, non-prefetchable) [size=4K]
>         I/O ports at f020 [size=32]
>         Capabilities: [c8] Power Management version 2
>         Capabilities: [d0] MSI: Enable+ Count=1/1 Maskable- 64bit+
>         Capabilities: [e0] PCI Advanced Features
>         Kernel driver in use: e1000e
>
>
> Here is dmesg right after boot and plugging in network after booted:
>
> [ 368.697841] e1000e 0000:00:19.0 com: Detected Hardware Unit Hang:
>   TDH <3d>
>   TDT <67>
>   next_to_use <67>
>   next_to_clean <39>
> buffer_info[next_to_clean]:
>   time_stamp <1000106e9>
>   next_to_watch <3d>
>   jiffies <100010c60>
>   next_to_watch.status <0>
> MAC Status <40080083>
> PHY Status <796d>
> PHY 1000BASE-T Status <3800>
> PHY Extended Status <3000>
> PCI Status <10>
> [ 370.696960] e1000e 0000:00:19.0 com: Detected Hardware Unit Hang:
>   TDH <3d>
>   TDT <67>
>   next_to_use <67>
>   next_to_clean <39>
> buffer_info[next_to_clean]:
>   time_stamp <1000106e9>
>   next_to_watch <3d>
>   jiffies <100011430>
>   next_to_watch.status <0>
> MAC Status <40080083>
> PHY Status <796d>
> PHY 1000BASE-T Status <3800>
> PHY Extended Status <3000>
> PCI Status <10>
> [ 372.695807] e1000e 0000:00:19.0 com: Detected Hardware Unit Hang:
>   TDH <3d>
>   TDT <67>
>   next_to_use <67>
>   next_to_clean <39>
> buffer_info[next_to_clean]:
>   time_stamp <1000106e9>
>   next_to_watch <3d>
>   jiffies <100011c00>
>   next_to_watch.status <0>
> MAC Status <40080083>
> PHY Status <796d>
> PHY 1000BASE-T Status <3800>
> PHY Extended Status <3000>
> PCI Status <10>
> [ 374.694933] e1000e 0000:00:19.0 com: Detected Hardware Unit Hang:
>   TDH <3d>
>   TDT <67>
>   next_to_use <67>
>   next_to_clean <39>
> buffer_info[next_to_clean]:
>   time_stamp <1000106e9>
>   next_to_watch <3d>
>   jiffies <1000123d0>
>   next_to_watch.status <0>
> MAC Status <40080083>
> PHY Status <796d>
> PHY 1000BASE-T Status <3800>
> PHY Extended Status <3000>
> PCI Status <10>
> [ 374.710096] ------------[ cut here ]------------
> [ 374.710124] WARNING: at net/sched/sch_generic.c:259
> dev_watchdog+0x270/0x280()
> [ 374.710128] NETDEV WATCHDOG: com (e1000e): transmit queue 0 timed out
> [ 374.710131] Modules linked in: binfmt_misc act_police cls_basic
> cls_flow cls_fw cls_u32 sch_fq_codel sch_tbf sch_prio sch_htb sch_hfsc
> sch_ingress sch_sfq xt_CHECKSUM ipt_rpfilter xt_stat
> istic xt_CT xt_connlimit xt_realm xt_addrtype xt_comment xt_recent
> xt_nat ipt_REJECT ipt_MASQUERADE ipt_ECN ipt_CLUSTERIP ipt_ah xt_set
> ip_set ipt_ULOG xt_LOG nf_nat_tftp nf_nat_snmp_basic n
> f_conntrack_snmp nf_nat_sip nf_nat_pptp nf_nat_proto_gre nf_nat_irc
> nf_nat_h323 nf_nat_amanda ts_kmp nf_conntrack_amanda nf_conntrack_sane
> nf_conntrack_tftp nf_conntrack_sip nf_conntrack_pro
> to_udplite nf_conntrack_proto_sctp nf_conntrack_pptp
> nf_conntrack_proto_gre nf_conntrack_netlink nf_conntrack_netbios_ns
> nf_conntrack_broadcast nf_conntrack_irc nf_conntrack_h323 xt_TPROXY
> nf_defrag_ipv6 xt_time xt_TCPMSS xt_tcpmss xt_sctp
> [ 374.710177] xt_policy xt_pkttype xt_physdev xt_owner xt_NFQUEUE
> xt_NFLOG nfnetlink_log xt_multiport xt_mark xt_mac xt_limit xt_length
> xt_iprange xt_helper xt_hashlimit xt_DSCP xt_dscp xt_dccp xt_conntrack
> xt_connmark xt_CLASSIFY xt_AUDIT xt_state iptable_raw iptable_nat
> nf_nat_ipv4 nf_conntrack_ipv4 nf_defrag_ipv4 iptable_mangle nfnetlink
> nf_nat_ftp nf_nat nf_conntrack_ftp nf_conntrack iptable_filter ip_tables
> 8021q garp stp mrp llc sg iTCO_wdt iTCO_vendor_support coretemp kvm
> crct10dif_pclmul crc32_pclmul crc32c_intel ghash_clmulni_intel
> aesni_intel lrw gf128mul glue_helper ablk_helper cryptd serio_raw pcspkr
> i2c_i801 lpc_ich mfd_core shpchp ipmi_si video ipmi_msghandler
> acpi_cpufreq mperf ext4 mbcache jbd2 raid1 sd_mod crc_t10dif
> crct10dif_common mgag200 syscopyarea sysfillrect ahci sysimgblt
> [ 374.710232] drm_kms_helper libahci ttm libata drm igb e1000e dca
> i2c_algo_bit i2c_core ptp pps_core dm_mirror dm_region_hash dm_log dm_mod
> [ 374.710247] CPU: 1 PID: 0 Comm: swapper/1 Not tainted
> 3.10.0-123.13.2.el7.x86_64 #1
> [ 374.710251] Hardware name: Supermicro X9SCL/X9SCM/X9SCL/X9SCM, BIOS
> 2.0b 09/17/2012
> [ 374.710254] ffff88022fc83d90 a08d8f9572a8441c ffff88022fc83d48
> ffffffff815e232c
> [ 374.710259] ffff88022fc83d80 ffffffff8105dee1 0000000000000000
> ffff8802209b4000
> [ 374.710264] ffff880220f60e80 0000000000000001 0000000000000001
> ffff88022fc83de8
> [ 374.710268] Call Trace:
> [ 374.710271] <IRQ> [<ffffffff815e232c>] dump_stack+0x19/0x1b
> [ 374.710285] [<ffffffff8105dee1>] warn_slowpath_common+0x61/0x80
> [ 374.710291] [<ffffffff8105df5c>] warn_slowpath_fmt+0x5c/0x80
> [ 374.710298] [<ffffffff81088681>] ? run_posix_cpu_timers+0x51/0x840
> [ 374.710313] [<ffffffff814f0d20>] dev_watchdog+0x270/0x280
> [ 374.710318] [<ffffffff814f0ab0>] ? dev_graft_qdisc+0x80/0x80
> [ 374.710323] [<ffffffff8106d236>] call_timer_fn+0x36/0x110
> [ 374.710328] [<ffffffff814f0ab0>] ? dev_graft_qdisc+0x80/0x80
> [ 374.710333] [<ffffffff8106f2ff>] run_timer_softirq+0x21f/0x320
> [ 374.710339] [<ffffffff81067047>] __do_softirq+0xf7/0x290
> [ 374.710345] [<ffffffff815f435c>] call_softirq+0x1c/0x30
> [ 374.710352] [<ffffffff81014cf5>] do_softirq+0x55/0x90
> [ 374.710356] [<ffffffff810673e5>] irq_exit+0x115/0x120
> [ 374.710361] [<ffffffff815f4d35>] smp_apic_timer_interrupt+0x45/0x60
> [ 374.710366] [<ffffffff815f369d>] apic_timer_interrupt+0x6d/0x80
> [ 374.710368] <EOI> [<ffffffff814835a2>] ? cpuidle_enter_state+0x52/0xc0
> [ 374.710380] [<ffffffff814836d5>] cpuidle_idle_call+0xc5/0x200
> [ 374.710386] [<ffffffff8101bc7e>] arch_cpu_idle+0xe/0x30
> [ 374.710393] [<ffffffff810b47e5>] cpu_startup_entry+0xf5/0x290
> [ 374.710399] [<ffffffff815d028e>] start_secondary+0x1c4/0x1da
> [ 374.710403] ---[ end trace feb9f00b67f36ca1 ]---
> [ 374.710420] e1000e 0000:00:19.0 com: Reset adapter unexpectedly
> [ 378.560296] e1000e: com NIC Link is Up 1000 Mbps Full Duplex, Flow
> Control: Rx/Tx
>
>
>
> And hours later, it repeats (no trace after first time) about 6-11 times
> per day every day:
>
> [12881.408092] e1000e 0000:00:19.0 com: Detected Hardware Unit Hang:
>   TDH <63>
>   TDT <7d>
>   next_to_use <7d>
>   next_to_clean <60>
> buffer_info[next_to_clean]:
>   time_stamp <100bfef24>
>   next_to_watch <63>
>   jiffies <100c012b9>
>   next_to_watch.status <0>
> MAC Status <40080083>
> PHY Status <796d>
> PHY 1000BASE-T Status <3800>
> PHY Extended Status <3000>
> PCI Status <10>
> [12881.414206] e1000e 0000:00:19.0 com: Reset adapter unexpectedly
> [12885.520180] e1000e: com NIC Link is Up 1000 Mbps Full Duplex, Flow
> Control: Rx/Tx
>
>
> I scanned through all the machines here to see if any others use e1000e,
> and found only one, which has no known issues, but doesn't have as much
> network traffic.
>
> *Here are the details for the only other e1000e machine I have, without
> problems:*
>
> 09:00.0 Ethernet controller: Intel Corporation 82574L Gigabit Network
> Connection
>         Subsystem: Super Micro Computer Inc Device 0000
>         Kernel driver in use: e1000e
>         Kernel modules: e1000e
>
> Linux machine2 3.2.0-55-generic #85-Ubuntu SMP Wed Oct 2 12:29:27 UTC
> 2013 x86_64 x86_64 x86_64 GNU/Linux
>
> Distributor ID: Ubuntu
> Description:    Ubuntu 12.04.4 LTS
> Release:        12.04
> Codename:       precise
>
> --
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@...r.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at  http://www.tux.org/lkml/
>


-- 

--------------------------------------------
Peter Maloney
Brockmann Consult
Max-Planck-Str. 2
21502 Geesthacht
Germany
Tel: +49 4152 889 300
Fax: +49 4152 889 333
E-mail: peter.maloney@...ckmann-consult.de
Internet: http://www.brockmann-consult.de
--------------------------------------------

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

Powered by blists - more mailing lists