Disabling IRQ - eth0

Wed Mar 16 13:39:47 UTC 2005

Just a follow up note:

The box has been up for two days with 0 errors in RX or TX.

Ifconfig:

eth0      Link encap:Ethernet  HWaddr 00:04:76:30:79:CE
          inet addr:10.9.9.145  Bcast:10.9.9.255  Mask:255.255.255.0
          inet6 addr: fe80::204:76ff:fe30:79ce/64 Scope:Link
          UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1
          RX packets:835546 errors:0 dropped:0 overruns:0 frame:0
          TX packets:15984 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:1000
          RX bytes:96549066 (92.0 MiB)  TX bytes:7391287 (7.0 MiB)
          Interrupt:5 Memory:eff00000-eff10000

lo        Link encap:Local Loopback
          inet addr:127.0.0.1  Mask:255.0.0.0
          inet6 addr: ::1/128 Scope:Host
          UP LOOPBACK RUNNING  MTU:16436  Metric:1
          RX packets:1416 errors:0 dropped:0 overruns:0 frame:0
          TX packets:1416 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:0
          RX bytes:1515308 (1.4 MiB)  TX bytes:1515308 (1.4 MiB) 

Problem solved.

Thanks all for the help.

-----Original Message-----
From: fedora-list-bounces at redhat.com
[mailto:fedora-list-bounces at redhat.com] On Behalf Of Andrew Leichter
Sent: Monday, March 14, 2005 9:45 AM
To: For users of Fedora Core releases
Subject: RE: Disabling IRQ - eth0

Thank you to those whom have replied.  I did not make it clear in my
last message that I did try booting with acpi=off last time.  It did not
seem to address the problem.

In regards to the bellow response:

You hit the nail right on the head.  That's exaclty the card I have.  To
be specific:

lspci |grep Ether
08:08.0 Ethernet controller: Broadcom Corporation NetXtreme BCM5700
Gigabit Ethernet (rev 12)

Ethtool eth0
        Supported ports: [ MII ]
        Supported link modes:   10baseT/Half 10baseT/Full
                                100baseT/Half 100baseT/Full
                                1000baseT/Half 1000baseT/Full
        Supports auto-negotiation: Yes
        Advertised link modes:  10baseT/Half 10baseT/Full
                                100baseT/Half 100baseT/Full
                                1000baseT/Half 1000baseT/Full
        Advertised auto-negotiation: Yes
        Speed: 1000Mb/s
        Duplex: Full
        Port: Twisted Pair
        PHYAD: 1
        Transceiver: internal
        Auto-negotiation: on
        Supports Wake-on: g
        Wake-on: d
        Current message level: 0x000000ff (255)
        Link detected: yes

ifconfig
eth0      Link encap:Ethernet  HWaddr 00:04:76:30:79:CE
          inet addr:10.9.9.145  Bcast:10.9.9.255  Mask:255.255.255.0
          inet6 addr: fe80::204:76ff:fe30:79ce/64 Scope:Link
          UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1
          RX packets:1123606 errors:2174570 dropped:0 overruns:0 frame:0
          TX packets:19185 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:1000
          RX bytes:132443161 (126.3 MiB)  TX bytes:4675634 (4.4 MiB)
          Interrupt:5

lo        Link encap:Local Loopback
          inet addr:127.0.0.1  Mask:255.0.0.0
          inet6 addr: ::1/128 Scope:Host
          UP LOOPBACK RUNNING  MTU:16436  Metric:1
          RX packets:2539 errors:0 dropped:0 overruns:0 frame:0
          TX packets:2539 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:0
          RX bytes:2367064 (2.2 MiB)  TX bytes:2367064 (2.2 MiB)

netstat -an -ieth0
Kernel Interface table
Iface       MTU Met    RX-OK RX-ERR RX-DRP RX-OVR    TX-OK TX-ERR TX-DRP
TX-OVR Flg
eth0       1500   0  1120903 2174570      0      0    19133      0
0      0 BMRU

I removed the tg3 driver per your suggestion and loaded the broadcom
driver.  The above stats are with the tg3 driver.  The below stats are
with the bcm5700 driver.  With the tg3 driver I immediatley began
getting receiving errors.  So far I am still clean 5 min after reboot.
Below is the ifconfig after the new driver install:

eth0      Link encap:Ethernet  HWaddr 00:04:76:30:79:CE
          inet addr:10.9.9.145  Bcast:10.9.9.255  Mask:255.255.255.0
          inet6 addr: fe80::204:76ff:fe30:79ce/64 Scope:Link
          UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1
          RX packets:582 errors:0 dropped:0 overruns:0 frame:0
          TX packets:61 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:1000
          RX bytes:67593 (66.0 KiB)  TX bytes:7197 (7.0 KiB)
          Interrupt:5 Memory:eff00000-eff10000

lo        Link encap:Local Loopback
          inet addr:127.0.0.1  Mask:255.0.0.0
          inet6 addr: ::1/128 Scope:Host
          UP LOOPBACK RUNNING  MTU:16436  Metric:1
          RX packets:1368 errors:0 dropped:0 overruns:0 frame:0
          TX packets:1368 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:0
          RX bytes:1505152 (1.4 MiB)  TX bytes:1505152 (1.4 MiB)

If it stay up until tomorrow I'll know it's fixed.  I'll post again
whether or not it stayed up.  Thanks for your help!

Andrew

-----Original Message-----
From: fedora-list-bounces at redhat.com
[mailto:fedora-list-bounces at redhat.com] On Behalf Of Randy Kelsoe
Sent: Sunday, March 13, 2005 12:34 PM
To: For users of Fedora Core releases
Subject: Re: Disabling IRQ - eth0

Andrew Leichter wrote:

>
>After disabling USB in the BIOS (this is a server and I don't need it),

>the OS moved eth0 to IRQ #5 and it is not sharing this IRQ with other 
>devices.  There are no other devices sharing any IRQ's in the whole 
>system.  The system stays connected to the network for about 6 hours 
>until the nic stops responding again.  I now receive this set of
>messages:
>
>Mar 11 14:44:05 linux1 kernel: irq 5: nobody cared! (screaming
>interrupt?)
>Mar 11 14:44:05 linux1 kernel: irq 5: Please try booting with acpi=off 
>and report a bug Mar 11 14:44:05 linux1 kernel:  [<02107fa9>]
>__report_bad_irq+0x3a/0x77 Mar 11 14:44:05 linux1 kernel:  [<02108530>]

>note_interrupt+0x19e/0x1c4 Mar 11 14:44:05 linux1 kernel:  [<02108ac7>]

>do_IRQ+0x24d/0x309 Mar 11 14:44:05 linux1 kernel:  [<022a007b>] 
>pci_conf2_read+0xc6/0x1dd Mar 11 14:44:05 linux1 kernel:  [<02124f00>]
>__do_softirq+0x2c/0x79 Mar 11 14:44:05 linux1 kernel:  [<021096ee>] 
>do_softirq+0x46/0x4d Mar 11 14:44:05 linux1 kernel:
>======================= Mar 11 14:44:05 linux1 kernel:  [<02108b77>]
>do_IRQ+0x2fd/0x309 Mar 11 14:44:05 linux1 kernel:  [<0210403b>]
>default_idle+0x23/0x26 Mar 11 14:44:05 linux1 kernel:  [<0210408c>]
>cpu_idle+0x1f/0x34 Mar 11 14:44:05 linux1 kernel:  [<023a86bb>]
>start_kernel+0x216/0x219 Mar 11 14:44:05 linux1 kernel: handlers:
>Mar 11 14:44:05 linux1 kernel: [<4290649e>] (tg3_interrupt+0x0/0x1c0
>[tg3])
>Mar 11 14:44:05 linux1 kernel: Disabling IRQ #5
>

Just a wild guess, but the broadcom 5700 gets detected as a tigon? card
and uses the tg3 driver. Broadcom has linux drivers available, and they
offer a .src rpm that can easily be rebuilt. Can you do a 'lspci |grep
Ether' and see if you have a broadcom NIC and let us know?

Other than that, high quality NICs are cheap these days, so you might
try replacing the one you have to see what happens. I guess you have
checked you network, and tried a different  switch/router port? What
does  'ethtool eth0'  tell you? What does 'ifconfig' and  'netstat -an
-ieth0' tell you about your error rate?

--
fedora-list mailing list
fedora-list at redhat.com
To unsubscribe: http://www.redhat.com/mailman/listinfo/fedora-list

--
fedora-list mailing list
fedora-list at redhat.com
To unsubscribe: http://www.redhat.com/mailman/listinfo/fedora-list