NetBSD-Bugs archive

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Old Index]

kern/55683: ixgX (X550) sometimes stop receiving



	Note: There was a bad value `' for the field `Confidential'.
	It was set to the default value of `yes'.

>Number:         55683
>Category:       kern
>Synopsis:       ixgX (X550) sometimes stop receiving
>Confidential:   yes
>Severity:       critical
>Priority:       high
>Responsible:    kern-bug-people
>State:          open
>Class:          sw-bug
>Submitter-Id:   net
>Arrival-Date:   Fri Sep 25 13:30:00 +0000 2020
>Originator:     Frank Kardel
>Release:        NetBSD 9.99.72
>Organization:
	
>Environment:
	
	
System: NetBSD gateway 9.99.72 NetBSD 9.99.72 (GATEWAY) #7: Thu Sep 10 13:58:53 UTC 2020 kardel@gateway:/src/NetBSD/cur/src/obj.amd64/sys/arch/amd64/compile/GATEWAY amd64
Architecture: x86_64
Machine: amd64
>Description:
	I am observing receive outages on my X550 interfaces. Every once in a while
	one of the 4 interface becomes deaf. I see very high input error counts on
	these interfaces. An ifconfig down/up brings the interfaces back to life.

State observed when no reception is done:

# sysctl -a
net.interfaces.ixg3.rcvq.drops = 0
net.interfaces.ixg3.sndq.len = 0
net.interfaces.ixg3.sndq.maxlen = 2046
net.interfaces.ixg3.sndq.drops = 0
hw.acpi.wake.ixg3 = 0
hw.ixg3.rx_processing_limit = 256
hw.ixg3.tx_processing_limit = 256
hw.ixg3.debug = 0
hw.ixg3.num_rx_desc = 2048
hw.ixg3.num_queues = 4
hw.ixg3.fc = 0
hw.ixg3.enable_aim = 1
hw.ixg3.advertise_speed = 0
hw.ixg3.txrx_workqueue = 0
hw.ixg3.dmac = 0
hw.ixg3.wol_enable = 1
hw.ixg3.wufc = 0
hw.ixg3.force_10_100_autonego = 0
hw.ixg3.eee_state = 1
hw.ixg3.q0.interrupt_rate = 83333
hw.ixg3.q0.txd_head = 27
hw.ixg3.q0.txd_tail = 27
hw.ixg3.q0.rxd_nxck = 1696
hw.ixg3.q0.rxd_head = 1696
hw.ixg3.q0.rxd_tail = 1695
hw.ixg3.q1.interrupt_rate = 100000
hw.ixg3.q1.txd_head = 547
hw.ixg3.q1.txd_tail = 547
hw.ixg3.q1.rxd_nxck = 1568
hw.ixg3.q1.rxd_head = 1567
hw.ixg3.q1.rxd_tail = 1567
hw.ixg3.q2.interrupt_rate = 35714
hw.ixg3.q2.txd_head = 870
hw.ixg3.q2.txd_tail = 870
hw.ixg3.q2.rxd_nxck = 386
hw.ixg3.q2.rxd_head = 386
hw.ixg3.q2.rxd_tail = 385
hw.ixg3.q3.interrupt_rate = 31250
hw.ixg3.q3.txd_head = 359
hw.ixg3.q3.txd_tail = 359
hw.ixg3.q3.rxd_nxck = 1715
hw.ixg3.q3.rxd_head = 1715
hw.ixg3.q3.rxd_tail = 1714

# ifconfig -vm ixg3
ixg3: flags=0x8843<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST> mtu 1500
	capabilities=0x7ff80<TSO4,IP4CSUM_Rx,IP4CSUM_Tx,TCP4CSUM_Rx>
	capabilities=0x7ff80<TCP4CSUM_Tx,UDP4CSUM_Rx,UDP4CSUM_Tx,TCP6CSUM_Rx>
	capabilities=0x7ff80<TCP6CSUM_Tx,UDP6CSUM_Rx,UDP6CSUM_Tx,TSO6>
	enabled=0x7ff80<TSO4,IP4CSUM_Rx,IP4CSUM_Tx,TCP4CSUM_Rx>
	enabled=0x7ff80<TCP4CSUM_Tx,UDP4CSUM_Rx,UDP4CSUM_Tx,TCP6CSUM_Rx>
	enabled=0x7ff80<TCP6CSUM_Tx,UDP6CSUM_Rx,UDP6CSUM_Tx,TSO6>
	ec_capabilities=0xf<VLAN_MTU,VLAN_HWTAGGING,JUMBO_MTU,VLAN_HWFILTER>
	ec_enabled=0x7<VLAN_MTU,VLAN_HWTAGGING,JUMBO_MTU>
	address: X----X
	media: Ethernet autoselect (1000baseT full-duplex)
	status: active
	supported Ethernet media:
		media none
		media 1000baseT mediaopt full-duplex
		media 100baseTX mediaopt full-duplex
		media 10baseT mediaopt full-duplex
		media autoselect
	input: 26783760 packets, 17739001536 bytes, 4049916 multicasts, 1214457 errors, 45046 queue drops
	output: 20557722 packets, 10152762352 bytes, 320814 multicasts
	inet xx.xx.xx.34/29 broadcast xx.xx.xx.39 flags 0
	inet xx.xx.xx.35/32 broadcast xx.xx.xx.35 flags 0
	inet xx.xx.xx.36/32 broadcast xx.xx.xx.36 flags 0
	inet xx.xx.xx.37/32 broadcast xx.xx.xx.37 flags 0
	inet6 fe80::ec4:7aff:fefa:3979%ixg3/64 flags 0 scopeid 0x4

# intrctl list
interrupt id       CPU0      CPU1      CPU2     CPU3  device name(s)
ioapic0 pin 9         0*        0         0        0  acpi SCI
ioapic0 pin 4         0*        0         0        0  com0
ioapic0 pin 3   2155553*        0         0        0  com1
msi0 vec 0            0*        0         0        0  ismt0
msix1 vec 0           0*        0         0        0  ahcisata0 port0
msix1 vec 1           0*        0         0        0  ahcisata0 port1
msix1 vec 2           0*        0         0        0  ahcisata0 port2
msix1 vec 3           0*        0         0        0  ahcisata0 port3
msix2 vec 5    56028281*        0         0        0  ahcisata1 port5
msix2 vec 6           0*        0         0        0  ahcisata1 port6
msi3 vec 0          276*        0         0        0  xhci0
msix4 vec 0    36162181*        0         0        0  ixg0 TXRX0
msix4 vec 1           0  10552694*        0        0  ixg0 TXRX1
msix4 vec 2           0         0  13329746*       0  ixg0 TXRX2
msix4 vec 3           0         0         0  9439699* ixg0 TXRX3
msix4 vec 4           1*        0         0        0  ixg0 link
msix5 vec 0     7496431*        0         0        0  ixg1 TXRX0
msix5 vec 1           0   1226473*        0        0  ixg1 TXRX1
msix5 vec 2           0         0    700512*       0  ixg1 TXRX2
msix5 vec 3           0         0         0  1770846* ixg1 TXRX3
msix5 vec 4           1*        0         0        0  ixg1 link
msix6 vec 0     1015066*        0         0        0  ixg2 TXRX0
msix6 vec 1           0    246663*        0        0  ixg2 TXRX1
msix6 vec 2           0         0    209309*       0  ixg2 TXRX2
msix6 vec 3           0         0         0   138600* ixg2 TXRX3
msix6 vec 4           1*        0         0        0  ixg2 link
msix7 vec 0    23012456*        0         0        0  ixg3 TXRX0
msix7 vec 1           0   5872758*        0        0  ixg3 TXRX1
msix7 vec 2           0         0   3033676*       0  ixg3 TXRX2
msix7 vec 3           0         0         0  5857198* ixg3 TXRX3
msix7 vec 4           3*        0         0        0  ixg3 link
ioapic0 pin 23        0*        0         0        0  ichsmb0

# uptime
 1:07PM  up 14 days,  1:17, 1 user, load averages: 0.11, 0.11, 0.12

# dmesg (shortened)
Copyright (c) 1996, 1997, 1998, 1999, 2000, 2001, 2002, 2003, 2004, 2005,
    2006, 2007, 2008, 2009, 2010, 2011, 2012, 2013, 2014, 2015, 2016, 2017,
    2018, 2019, 2020 The NetBSD Foundation, Inc.  All rights reserved.
Copyright (c) 1982, 1986, 1989, 1991, 1993
    The Regents of the University of California.  All rights reserved.

NetBSD 9.99.72 (GATEWAY) #7: Thu Sep 10 13:58:53 UTC 2020
	kardel@gateway:/src/NetBSD/cur/src/obj.amd64/sys/arch/amd64/compile/GATEWAY
...
ppb5 at pci0 dev 22 function 0: Intel C3000 PCIe Root Port for LAN 0 (rev. 0x11)
ppb5: PCI Express capability version 2 <Root Port of PCI-E Root Complex> x1 @ 2.5GT/s
pci6 at ppb5 bus 6
pci6: i/o space, memory space enabled, rd/line, wr/inv ok
ixg0 at pci6 dev 0 function 0: Intel(R) PRO/10GbE PCI-Express Network Driver, Version - 4.0.1-k
ixg0: clearing prefetchable bit
ixg0: device X550EM A
ixg0: NVM Image Version 0.05 ID 0x8, NVM Map version 1.16, OEM NVM Image version 0.06, ETrackID 80000877
ixg0: for TX/RX, interrupting at msix4 vec 0, bound queue 0 to cpu 0
allocated pic msix4 type edge pin 1 level 6 to cpu1 slot 0 idt entry 96
ixg0: for TX/RX, interrupting at msix4 vec 1, bound queue 1 to cpu 1
allocated pic msix4 type edge pin 2 level 6 to cpu1 slot 1 idt entry 97
ixg0: for TX/RX, interrupting at msix4 vec 2, bound queue 2 to cpu 2
allocated pic msix4 type edge pin 3 level 6 to cpu1 slot 1 idt entry 97
ixg0: for TX/RX, interrupting at msix4 vec 3, bound queue 3 to cpu 3
allocated pic msix4 type edge pin 4 level 6 to cpu1 slot 1 idt entry 97
ixg0: for link, interrupting at msix4 vec 4, affinity to cpu 0
ixg0: Using MSI-X interrupts with 5 vectors
ixg0: Ethernet address X----X
ixg0: PHY OUI 0x000ac2, model 0x002a, rev. 0
ixg0: feature cap 0xfa0<TEMP_SENSOR,LEGACY_TX,FDIR,MSI,MSIX,EEE>
ixg0: feature ena 0xc20<TEMP_SENSOR,MSIX,EEE>
ixg1 at pci6 dev 0 function 1: Intel(R) PRO/10GbE PCI-Express Network Driver, Version - 4.0.1-k
ixg1: clearing prefetchable bit
ixg1: device X550EM A
ixg1: NVM Image Version 0.05 ID 0x8, NVM Map version 1.16, OEM NVM Image version 0.06, ETrackID 80000877
allocated pic msix5 type edge pin 0 level 6 to cpu1 slot 1 idt entry 97
ixg1: for TX/RX, interrupting at msix5 vec 0, bound queue 0 to cpu 0
allocated pic msix5 type edge pin 1 level 6 to cpu1 slot 1 idt entry 97
ixg1: for TX/RX, interrupting at msix5 vec 1, bound queue 1 to cpu 1
allocated pic msix5 type edge pin 2 level 6 to cpu1 slot 2 idt entry 98
ixg1: for TX/RX, interrupting at msix5 vec 2, bound queue 2 to cpu 2
allocated pic msix5 type edge pin 3 level 6 to cpu1 slot 2 idt entry 98
ixg1: for TX/RX, interrupting at msix5 vec 3, bound queue 3 to cpu 3
allocated pic msix5 type edge pin 4 level 6 to cpu1 slot 2 idt entry 98
ixg1: for link, interrupting at msix5 vec 4, affinity to cpu 0
ixg1: Using MSI-X interrupts with 5 vectors
ixg1: Ethernet address X----X
ixg1: PHY OUI 0x000ac2, model 0x002a, rev. 0
ixg1: feature cap 0xfa0<TEMP_SENSOR,LEGACY_TX,FDIR,MSI,MSIX,EEE>
ixg1: feature ena 0xc20<TEMP_SENSOR,MSIX,EEE>
ppb6 at pci0 dev 23 function 0: Intel C3000 PCIe Root Port for LAN 1 (rev. 0x11)
ppb6: PCI Express capability version 2 <Root Port of PCI-E Root Complex> x1 @ 2.5GT/s
pci7 at ppb6 bus 7
pci7: i/o space, memory space enabled, rd/line, wr/inv ok
ixg2 at pci7 dev 0 function 0: Intel(R) PRO/10GbE PCI-Express Network Driver, Version - 4.0.1-k
ixg2: clearing prefetchable bit
ixg2: device X550EM A
ixg2: NVM Image Version 0.05 ID 0x8, NVM Map version 1.16, OEM NVM Image version 0.06, ETrackID 8000087c
allocated pic msix6 type edge pin 0 level 6 to cpu1 slot 2 idt entry 98
ixg2: for TX/RX, interrupting at msix6 vec 0, bound queue 0 to cpu 0
allocated pic msix6 type edge pin 1 level 6 to cpu1 slot 2 idt entry 98
ixg2: for TX/RX, interrupting at msix6 vec 1, bound queue 1 to cpu 1
allocated pic msix6 type edge pin 2 level 6 to cpu1 slot 3 idt entry 99
ixg2: for TX/RX, interrupting at msix6 vec 2, bound queue 2 to cpu 2
allocated pic msix6 type edge pin 3 level 6 to cpu1 slot 3 idt entry 99
ixg2: for TX/RX, interrupting at msix6 vec 3, bound queue 3 to cpu 3
allocated pic msix6 type edge pin 4 level 6 to cpu1 slot 3 idt entry 99
ixg2: for link, interrupting at msix6 vec 4, affinity to cpu 0
ixg2: Using MSI-X interrupts with 5 vectors
ixg2: Ethernet address X----X
ixg2: PHY OUI 0x000ac2, model 0x002a, rev. 0
ixg2: feature cap 0xfa0<TEMP_SENSOR,LEGACY_TX,FDIR,MSI,MSIX,EEE>
ixg2: feature ena 0xc20<TEMP_SENSOR,MSIX,EEE>
ixg3 at pci7 dev 0 function 1: Intel(R) PRO/10GbE PCI-Express Network Driver, Version - 4.0.1-k
ixg3: clearing prefetchable bit
ixg3: device X550EM A
ixg3: NVM Image Version 0.05 ID 0x8, NVM Map version 1.16, OEM NVM Image version 0.06, ETrackID 8000087c
allocated pic msix7 type edge pin 0 level 6 to cpu1 slot 3 idt entry 99
ixg3: for TX/RX, interrupting at msix7 vec 0, bound queue 0 to cpu 0
allocated pic msix7 type edge pin 1 level 6 to cpu1 slot 3 idt entry 99
ixg3: for TX/RX, interrupting at msix7 vec 1, bound queue 1 to cpu 1
allocated pic msix7 type edge pin 2 level 6 to cpu1 slot 4 idt entry 100
ixg3: for TX/RX, interrupting at msix7 vec 2, bound queue 2 to cpu 2
allocated pic msix7 type edge pin 3 level 6 to cpu1 slot 4 idt entry 100
ixg3: for TX/RX, interrupting at msix7 vec 3, bound queue 3 to cpu 3
allocated pic msix7 type edge pin 4 level 6 to cpu1 slot 4 idt entry 100
ixg3: for link, interrupting at msix7 vec 4, affinity to cpu 0
ixg3: Using MSI-X interrupts with 5 vectors
ixg3: Ethernet address X----X
ixg3: PHY OUI 0x000ac2, model 0x002a, rev. 0
ixg3: feature cap 0xfa0<TEMP_SENSOR,LEGACY_TX,FDIR,MSI,MSIX,EEE>
ixg3: feature ena 0xc20<TEMP_SENSOR,MSIX,EEE>
>How-To-Repeat:
	Run current or older (I see this for ~1.5 years now) on a 
	SuperMicro C3000 platform.
>Fix:
	? - analyse error conditions?

>Unformatted:
 	
 	



Home | Main Index | Thread Index | Old Index