INIT: id "7" respawning too fast : disabled for 5

[Pit]

Hi,

i have two ZS with five nics running very fine. One with beta12 the other one with beta13. Four nics are working in balancing mode powered by atheling’s patch.
Hardware: Jetway J7F2WE1GES-OLFLV with onboard nic + daughterboard MBz Jetway Adapter 02 3xGiga Lan (Realtek RTL-8110SC/8169SC) + Intel nic at raiserboard. Delock 54144 1GB flash module.

For traffic shaping i need a little more CPU power. So i bought some

Jetway J7F5M1G5-VHE_LF with onboard Giga Lan + daughterboard
MBz Jetway Adapter 02 3xGiga Lan (Realtek RTL-8110SC/8169SC) + Realtek 8169 Giga Lan at raiserboard. Delock 54144 flash module.

And now the very stange behavior:

With beta13 and beta12 this message appears at the shell:

INIT: id “7” respawning too fast : disabled for 5 minutes

With beta13 only the onboard nic works like expected. The daughterboard nics and the extra nic are recognized but refused to work. When connecting a patch cable the link is not recognized. After reboot the link is up but the nics refuses to work.

With beta12 the onboard nic and the daughterboard nics work fine. Only the extra nic refuses to work. I also tryed some old 3COM and COMPEX cards, but always the same behavior.

Has somebody an idea what is going wrong and how to fix it?

Thanks in advance for any help.

Regards
Pit

[ppalias]

Check for IRQ conflicts for a start. Are all NICs found in “lspci”? Are all modules loaded in “lsmod”?

[Pit]

My interrupts are:

CPU0
0: 68 IO-APIC-edge timer
1: 354 IO-APIC-edge i8042
9: 0 IO-APIC-fasteoi acpi
15: 3247 IO-APIC-edge ide1
16: 0 IO-APIC-fasteoi ETH00, ETH03
17: 14 IO-APIC-fasteoi firewire_ohci
18: 5193 IO-APIC-fasteoi ETH01, ETH04
19: 436 IO-APIC-fasteoi ETH02
NMI: 0 Non-maskable interrupts
LOC: 207416 Local timer interrupts
RES: 0 Rescheduling interrupts
CAL: 0 function call interrupts
TLB: 0 TLB shootdowns
TRM: 0 Thermal event interrupts
SPU: 0 Spurious interrupts
ERR: 0
MIS: 0
~

My lspci is:

00:00.0 Host bridge: VIA Technologies, Inc. CX700/VX700 Host Bridge (rev 03)
00:00.1 Host bridge: VIA Technologies, Inc. CX700/VX700 Host Bridge
00:00.2 Host bridge: VIA Technologies, Inc. CX700/VX700 Host Bridge
00:00.3 Host bridge: VIA Technologies, Inc. CX700/VX700 Host Bridge
00:00.4 Host bridge: VIA Technologies, Inc. CX700/VX700 Host Bridge
00:00.7 Host bridge: VIA Technologies, Inc. CX700/VX700 Host Bridge
00:01.0 PCI bridge: VIA Technologies, Inc. VT8237/VX700 PCI Bridge
00:0f.0 IDE interface: VIA Technologies, Inc. VX800 Serial ATA and EIDE Controller
00:11.0 ISA bridge: VIA Technologies, Inc. CX700/VX700 PCI to ISA Bridge
00:11.7 Host bridge: VIA Technologies, Inc. CX700/VX700 Internal Module Bus
00:13.0 Host bridge: VIA Technologies, Inc. CX700/VX700 Host Bridge
00:13.1 PCI bridge: VIA Technologies, Inc. CX700/VX700 PCI to PCI Bridge
01:00.0 VGA compatible controller: VIA Technologies, Inc. CX700/VX700 [S3 UniChrome Pro] (rev 03)
02:03.0 Ethernet controller: Compex RL100-ATX 10/100
02:04.0 Ethernet controller: Realtek Semiconductor Co., Ltd. RTL-8110SC/8169SC Gigabit Ethernet (rev 10)
02:05.0 Ethernet controller: VIA Technologies, Inc. VT6120/VT6121/VT6122 Gigabit Ethernet Adapter (rev 11)
02:06.0 Ethernet controller: Realtek Semiconductor Co., Ltd. RTL-8110SC/8169SC Gigabit Ethernet (rev 10)
02:07.0 Ethernet controller: Realtek Semiconductor Co., Ltd. RTL-8110SC/8169SC Gigabit Ethernet (rev 10)
02:08.0 FireWire (IEEE 1394): VIA Technologies, Inc. VT6306 Fire II IEEE 1394 OHCI Link Layer Controller (rev 80)

My modules are:

Module Size Used by
dummy1 5376 0
ipt_MASQUERADE 5632 2
xt_CONNMARK 5760 1
tun 13696 0
bridge 57752 0
llc 9228 1 bridge
iptable_nat 9224 1
xt_state 4352 4
xt_tcpudp 4992 11
iptable_mangle 5248 1
iptable_filter 5376 1
ip_tables 16144 3 iptable_nat,iptable_mangle,iptable_filter
x_tables 19076 6 ipt_MASQUERADE,xt_CONNMARK,iptable_nat,xt_state,xt_tcpudp,ip_tables
nf_nat_pptp 5632 0
nf_nat_proto_gre 4740 1 nf_nat_pptp
nf_nat_ftp 5248 0
nf_nat 21656 5 ipt_MASQUERADE,iptable_nat,nf_nat_pptp,nf_nat_proto_gre,nf_nat_ftp
nf_conntrack_ipv4 19720 8 iptable_nat,nf_nat
nf_conntrack_tftp 7060 0
nf_conntrack_pptp 8964 1 nf_nat_pptp
nf_conntrack_proto_gre 8064 1 nf_conntrack_pptp
nf_conntrack_irc 8604 0
nf_conntrack_sip 11156 0
nf_conntrack_ftp 11300 1 nf_nat_ftp
nf_conntrack_h323 51912 0
nf_conntrack 67668 15 ipt_MASQUERADE,xt_CONNMARK,iptable_nat,xt_state,nf_nat_pptp,nf_nat_ftp,nf_nat,nf_conntrack_ipv4,nf_conntrack_tftp,nf_conntrack_pptp,nf_conntrack_proto_gre,nf_conntrack_irc,nf_conntrack_sip,nf_conntrack_ftp,nf_conntrack_h323
dummy 5376 0
vfat 14976 0
fat 54432 1 vfat
ata_generic 10244 0
pata_via 13572 0
libata 165192 2 ata_generic,pata_via
dock 12304 1 libata
winbond_840 25096 0
via_velocity 34056 0
crc_ccitt 3968 1 via_velocity
firewire_ohci 20992 0
firewire_core 46016 1 firewire_ohci
crc_itu_t 3968 1 firewire_core
loop 20100 1

ETH00 and ETH03 are recognized but not working. Maybe because of no CPU cycles. The module for the Compex card is winbond_840.

To get more free interrupts i disabled serial ports, usb and sound.
But the nics still bites 16, 18 and 19. I can not find a setting to diable firewire.

[ppalias]

Can you assign IRQs manually rather than automatically from the system?

[Pit]

Idle timeout is not a good thing. One hour of typing for nirwana.

I’ll try it later again with copy and paste.

[Pit]

Hello ppalias,

thank you for responding.

No it is not possible to assign IRQs manually.

It is definitely a resource conflict. With strike matches in my eyes and a lot of coffee i think i found a way to get things working.

At first i tried several other mobo’s and recognized that the behavior of the same kind of mobo’s was different. At a deeper look i found two different series with the same name.

J7F5M1G5-VHE-LF and J7F5M1G5-VHE-PB.

LF shows the discribed behavior. PB was much harder. In the following i refer to the PB series board.

After installing the 3-giga-lan daughterboard and the riserboard nic nothing happens when switching on the machine. Only the fans whirl the air. The machine refuses to boot. This text is very short in comparision to the time i tryed to google for hints. f. e. : jetway j7f5 interrupt
At the Jetway site there was a BIOS update which sounds fine:

http://www.jetway.com.tw/jw/download_show.asp

BIOS /DOS 05 512KB

1.Change BIOS for PCI Riser Card.
2.Resolve some USB Storages can not start.
2008-9-10

This update was especialy for the LF. But: no risk no fun …..

After updating a clear cmos is necessary. A little wonder happens, the POST beeps and the machine boots into beta13. A cat /proc/interrupts shows that only the onboard and daughterboard nics are recognized properly. Lspci ends up with one unknown ethernet pci device.

To make the cmos know about the missing information i unpluged the daughterboard and start again. The second wonder: After the POST beep
the nics PXE handshake appears and the cmos knows the nic.

The complete machine boots now properly into beta13:

root@zeroshell root> cat /proc/interrupts
CPU0
0: 68 IO-APIC-edge timer
1: 16 IO-APIC-edge i8042
4: 8979 IO-APIC-edge serial
9: 0 IO-APIC-fasteoi acpi
15: 6202 IO-APIC-edge ide1
16: 1 IO-APIC-fasteoi ETH00, ETH02
17: 127 IO-APIC-fasteoi firewire_ohci
18: 94590 IO-APIC-fasteoi ETH03, ETH04
19: 0 IO-APIC-fasteoi ETH01
20: 0 IO-APIC-fasteoi uhci_hcd:usb2
21: 0 IO-APIC-fasteoi uhci_hcd:usb4
22: 0 IO-APIC-fasteoi uhci_hcd:usb3
23: 0 IO-APIC-fasteoi ehci_hcd:usb1
NMI: 0 Non-maskable interrupts
LOC: 2025324 Local timer interrupts
RES: 0 Rescheduling interrupts
CAL: 0 function call interrupts
TLB: 0 TLB shootdowns
TRM: 0 Thermal event interrupts
SPU: 0 Spurious interrupts
ERR: 0
MIS: 0

root@zeroshell root> lspci
00:00.0 Host bridge: VIA Technologies, Inc. CX700/VX700 Host Bridge (rev 03)
00:00.1 Host bridge: VIA Technologies, Inc. CX700/VX700 Host Bridge
00:00.2 Host bridge: VIA Technologies, Inc. CX700/VX700 Host Bridge
00:00.3 Host bridge: VIA Technologies, Inc. CX700/VX700 Host Bridge
00:00.4 Host bridge: VIA Technologies, Inc. CX700/VX700 Host Bridge
00:00.7 Host bridge: VIA Technologies, Inc. CX700/VX700 Host Bridge
00:01.0 PCI bridge: VIA Technologies, Inc. VT8237/VX700 PCI Bridge
00:0f.0 IDE interface: VIA Technologies, Inc. VX800 Serial ATA and EIDE Controller
00:10.0 USB Controller: VIA Technologies, Inc. VT82xxxxx UHCI USB 1.1 Controller (rev 90)
00:10.1 USB Controller: VIA Technologies, Inc. VT82xxxxx UHCI USB 1.1 Controller (rev 90)
00:10.2 USB Controller: VIA Technologies, Inc. VT82xxxxx UHCI USB 1.1 Controller (rev 90)
00:10.4 USB Controller: VIA Technologies, Inc. USB 2.0 (rev 90)
00:11.0 ISA bridge: VIA Technologies, Inc. CX700/VX700 PCI to ISA Bridge
00:11.7 Host bridge: VIA Technologies, Inc. CX700/VX700 Internal Module Bus
00:13.0 Host bridge: VIA Technologies, Inc. CX700/VX700 Host Bridge
00:13.1 PCI bridge: VIA Technologies, Inc. CX700/VX700 PCI to PCI Bridge
01:00.0 VGA compatible controller: VIA Technologies, Inc. CX700/VX700 [S3 UniChrome Pro] (rev 03)
02:03.0 Ethernet controller: Realtek Semiconductor Co., Ltd. RTL-8169 Gigabit Ethernet (rev 10)
02:04.0 Ethernet controller: Realtek Semiconductor Co., Ltd. RTL-8110SC/8169SC Gigabit Ethernet (rev 10)
02:05.0 Ethernet controller: VIA Technologies, Inc. VT6120/VT6121/VT6122 Gigabit Ethernet Adapter (rev 11)
02:06.0 Ethernet controller: Realtek Semiconductor Co., Ltd. RTL-8110SC/8169SC Gigabit Ethernet (rev 10)
02:07.0 Ethernet controller: Realtek Semiconductor Co., Ltd. RTL-8110SC/8169SC Gigabit Ethernet (rev 10)
02:08.0 FireWire (IEEE 1394): VIA Technologies, Inc. VT6306 Fire II IEEE 1394 OHCI Link Layer Controller (rev 80)

So far so good. In comparison to this my J7F2 boards assign nearly one irq for each device. Is there something serious when two ETHs share one irq?

For testing i gave ETH04 to my desktop and ETH03 with DHCP to my DSL router. Default Gateway was set automaticaly behind the curtain.
ETH03 pulled into the NAT box and and a lightning fast Gateway was build.

Can i expext this speed when all nics are on and four of them are balancing the WAN traffic? Or can i expext serious issues due to the resource sharing? Or is there a trick to persuade the APIC to spend more IRQs?

Regard

Pit

[ppalias]

Pit I don’t believe you will have any issues. The speeds that networks have now should not stress the irqs of a nowdays system. There were some cases were certain network cards, especially wireless, wouldn’t operate correctly if irq was shared, but I don’t think that this applies here.

[Pit]

Hello ppalias,
i don’t think to have such certain nics because they are from the same vendor and specially build for this board. The extra nic which causes my frustation has a similar chipset and use the same module.
Further tests show an irregular appearence of the ‘INIT: id “7” respawning …’ message again. Stress tests with heavy downloads results in a performance loss of 80% after a while.
So i asked my distributor and the vendor Jetway if there is a way to solve the irq conflict.
Report follows as soon as i get an answer.

Regards

Pit

[Author]

Posts