interrupt trouble with 8139too, r128

From: Stefan Schwarzer (sts_at_luzot.com)
Date: 04/10/05

  • Next message: David Schwartz: "Re: genereate random numbers in kernel" 
    Date: 10 Apr 2005 00:47:10 GMT

    I have got an Athlon PC with two pci network cards that are handled by
    the 8139too driver and ATI Rage 128 PF/PRO graphics that share
    an interrupt with one of the NICs. My PC runs a fairly clean debian
    sarge (KDE desktop) on top of a custom 2.6.10 kernel.
    Things run fine normally, but start going wrong when I log out and
    the the X server shuts down and restarts. On restart,
    the kernel disables an unhandled interrupt (the one shared
    between the graphics card and one of the NICs). The graphics keeps working
    afterwards, but the NIC on that interrupt remains quiet. I'd like to
    file a bug report somewhere, but I have some doubts whether

            - this is an issue with the network driver which may
                    not really like having to deal with 2 cards,
            - or rather a problem with the r128 driver of XFree86, since
                    things go wrong during re-initialization of that piece,
            - or my (configuration) stupidity/a broken BIOS, ACPI
                    implementation
            - what else could it be?

    I attach the relevant part of the kernel log.
    I'd be happy if someone could point out some additional tests (or docs
    to read) to find out which component causes the problem.

    Thanks a lot

    stefan dot schwarzer at gmx dot net.

    --------------------------------------------------------------------------

    8139cp: 10/100 PCI Ethernet driver v1.2 (Mar 22, 2004)
    8139cp: pci dev 0000:00:0c.0 (id 10ec:8139 rev 10) is not an 8139C+ compatible chip
    8139cp: Try the "8139too" driver instead.
    8139cp: pci dev 0000:00:0d.0 (id 10ec:8139 rev 10) is not an 8139C+ compatible chip
    8139cp: Try the "8139too" driver instead.
    8139too Fast Ethernet driver 0.9.27
    PCI: Found IRQ 10 for device 0000:00:0c.0
    PCI: Sharing IRQ 10 with 0000:01:05.0
    eth0: RealTek RTL8139 at 0xf0d1c000, 00:e0:7d:b5:87:b2, IRQ 10
    eth0: Identified 8139 chip type 'RTL-8139C'
    PCI: Found IRQ 5 for device 0000:00:0d.0
    eth1: RealTek RTL8139 at 0xf0d1e000, 00:c0:26:87:4f:56, IRQ 5
    eth1: Identified 8139 chip type 'RTL-8100B/8139D'// NIC 2 (eth1)
    PCI: Found IRQ 11 for device 0000:00:0e.0 // sound
    PCI: Sharing IRQ 11 with 0000:00:0a.0 // isdn active
    (... stuff deleted ...)

    PCI: Found IRQ 10 for device 0000:01:05.0 // graphics
    PCI: Sharing IRQ 10 with 0000:00:0c.0 // NIC 1 (eth0)
    [drm] Initialized r128 2.5.0 20030725 on minor 0: ATI Technologies Inc Rage 128PF/PRO AGP 4x TMDS
    mtrr: base(0xd4000000) is not aligned on a size(0x1f00000) boundary
    agpgart: Found an AGP 2.0 compliant device at 0000:00:00.0.
    agpgart: Putting AGP V2 device at 0000:00:00.0 into 1x mode
    agpgart: Putting AGP V2 device at 0000:01:05.0 into 1x mode
    [....... logout here ........]
    irq 10: nobody cared!
     [<c01322ea>] __report_bad_irq+0x2a/0x90
     [<c0131c60>] handle_IRQ_event+0x30/0x70
     [<c01323dc>] note_interrupt+0x6c/0xd0
     [<c0131dd0>] __do_IRQ+0x130/0x160
     [<c0104b19>] do_IRQ+0x19/0x30
     [<c010309a>] common_interrupt+0x1a/0x20
     [<c011bcb0>] __do_softirq+0x30/0x90
     [<c011bd36>] do_softirq+0x26/0x30
     [<c0131c24>] irq_exit+0x34/0x40
     [<c0104b1e>] do_IRQ+0x1e/0x30
     [<c010309a>] common_interrupt+0x1a/0x20
    handlers:
    [<f0d4dca0>] (rtl8139_interrupt+0x0/0x190 [8139too])
    Disabling IRQ #10
    mtrr: base(0xd4000000) is not aligned on a size(0x1f00000) boundary
    agpgart: Found an AGP 2.0 compliant device at 0000:00:00.0.
    agpgart: Putting AGP V2 device at 0000:00:00.0 into 1x mode
    agpgart: Putting AGP V2 device at 0000:01:05.0 into 1x mode

  • Next message: David Schwartz: "Re: genereate random numbers in kernel"

    Relevant Pages

    • redhat 7.3 installation
      ... Linux agpgart interface v0.99 Jeff Hartmann ... agpgart: Maximum main memory to use for agp memory: 202M ... PCI: PCI BIOS revision 2.10 entry at 0xf0900, ... Serial driver version 5.05c with MANY_PORTS MULTIPORT SHARE_IRQ SERIAL_PCI ISAPNP enabled ...
      (RedHat)
    • Re: Geforce2go slow: worth upgrading?
      ... If you've built your kernel with agpgart support, ... "cannot use AGP because of $ERROR, ... > Your problems do not concern the nvidia drivers, ... If you try to run with Driver ...
      (comp.os.linux.x)
    • Re: etch 965Q video 16:10
      ... The vesa driver is the installation default. ... It seems to be problem of the video memory I need for 1920X1200. ... But to get the additional memory I need agpgart. ... A second problem is the PCI scan if found in the Xorg log: ...
      (Debian-User)
    • via agp vs. agpgart? 2.4.29
      ... There used to be via_agp.o which seems to be integrated in agpgart now. ... than it on a i810 card which doesn't specify it's agp generation. ... have an 1106:0305 Host Bridge, ... 00:01.0 PCI bridge: VT8363/8365 ...
      (comp.os.linux.hardware)
    • Re: 64 bit ati driver problem under FC3 x86_64 (long)
      ... >>of the driver should work without AGPGART. ... > built-in agp module... ... kernel, then load the module mentioned by the referenced website? ...
      (Fedora)