[2.6.2-rc1] BTTV, RTC and I2C warning messages

From: Michael Buesch (mbuesch_at_freenet.de)
Date: 01/24/04

  • Next message: Manfred Spraul: "Re: [PATCH] [2.4] forcedeth network driver"
    To: kraxel@bytesex.org, frodol@dds.nl, greg@kroah.com
    Date:	Sat, 24 Jan 2004 20:03:39 +0100
    
    

    -----BEGIN PGP SIGNED MESSAGE-----
    Hash: SHA1

    Hi,

    I got these messages in the syslog while watching TV with
    the "tvtime" application.

    Jan 24 18:04:24 lfs kernel: bttv0: skipped frame. no signal? high irq latency? [main=1fd81000,o_vbi=1fd81018,o_field=1b0b6000,rc=1b0b601c]
    Jan 24 18:21:54 lfs kernel: bttv0: skipped frame. no signal? high irq latency? [main=1fd81000,o_vbi=1fd81018,o_field=1b0b6000,rc=1b0b601c]
    Jan 24 18:29:01 lfs kernel: bttv0: skipped frame. no signal? high irq latency? [main=1fd81000,o_vbi=1fd81018,o_field=d52b000,rc=d52b01c]
    Jan 24 18:41:09 lfs kernel: bttv0: skipped frame. no signal? high irq latency? [main=1fd81000,o_vbi=1fd81018,o_field=1b0b4000,rc=1b0b401c]
    Jan 24 18:59:21 lfs kernel: bttv0: skipped frame. no signal? high irq latency? [main=1fd81000,o_vbi=1fd81018,o_field=1b0b4000,rc=1b0b401c]
    Jan 24 19:27:27 lfs kernel: bttv0: skipped frame. no signal? high irq latency? [main=1fd81000,o_vbi=1fd81018,o_field=d569000,rc=d56901c]
    Jan 24 19:40:39 lfs kernel: bttv0: skipped frame. no signal? high irq latency? [main=1fd81000,o_vbi=1fd81018,o_field=1b0b4000,rc=1b0b4264]
    Jan 24 19:43:59 lfs kernel: i2c IR (Hauppauge): unknown key: key=0x3f raw=0x3fff down=1
    Jan 24 19:43:59 lfs kernel: i2c IR (Hauppauge): unknown key: key=0x3f raw=0x3fff down=0
    Jan 24 19:44:27 lfs kernel: rtc: lost some interrupts at 1024Hz.

    The "skipped frame" and "lost interrupt" messages were there
    forever in kernel 2.6 for me. (I never saw them in 2.4). I had
    several threads about them on lkml but nobody could really help me, so
    I began to ignore these, as it runs even with them pretty good.

    But the new message I got today was the "i2c IR (Hauppauge):" message
    Yes, my Happauge WinTV PCI has an IR connector for remote control.
    So kernel tells me, that someone pressed an unknown key on the
    remote control (I guess this from the message). But the curious thing is:
    There's no IR remote-control device connected to the Hauppauge
    connection pin for at least a year, now. :) I've no IR-device
    connected to any port.

    yes, the "rtc" message... . I think it has something to do with that
    tvtime uses RTC to display some picture or so. So tvtime accesses
    the bttv driver and the rtc driver at once. So my guess: Probably a
    race condition here? Probably in the kernel? Who knows... I don't.

    mb@lfs:~> cat /proc/interrupts
               CPU0
      0: 7284059 IO-APIC-edge timer
      1: 10495 IO-APIC-edge i8042
      2: 0 XT-PIC cascade
      4: 49597 IO-APIC-edge serial
      8: 7222782 IO-APIC-edge rtc
      9: 0 IO-APIC-level acpi
     12: 279485 IO-APIC-edge i8042
     14: 104363 IO-APIC-edge ide0
     15: 40 IO-APIC-edge ide1
     16: 1160911 IO-APIC-level eth0, nvidia
     17: 528152 IO-APIC-level bttv0
     19: 64980 IO-APIC-level uhci_hcd, Ensoniq AudioPCI
     20: 0 IO-APIC-level ohci_hcd
     21: 0 IO-APIC-level ohci_hcd
     22: 0 IO-APIC-level ehci_hcd
     23: 0 IO-APIC-level uhci_hcd
    NMI: 0
    LOC: 7284538
    ERR: 0
    MIS: 0

    Yes, somebody told me, that assigning a lower interrupt number to
    the bttv card removed the "lost interrupt" messages, but how to do that?
    I've tried to move the card around in the pci-slots, but it has always
    been assigned the same IRQ. I've tried to set some interrupt options in
    the bios, but that also didn't help. Is there any way to tell bttv to use
    for example IRQ 11 or something like that? Also loading bttv before
    the eth0 driver didn't work. As you see, bttv still has a higher IRQ.

    Thanks you for all your help and suggestions.

    - --
    Regards Michael Buesch [ http://www.tuxsoft.de.vu ]

    -----BEGIN PGP SIGNATURE-----
    Version: GnuPG v1.2.4 (GNU/Linux)

    iD8DBQFAEsGLFGK1OIvVOP4RAsUaAJ4kHsKRWLXFGIiSS7OtWzLlOYNmKwCgkBhR
    dw+QxPwCP/dhXbImazANp38=
    =bIBi
    -----END PGP SIGNATURE-----
    -
    To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
    the body of a message to majordomo@vger.kernel.org
    More majordomo info at http://vger.kernel.org/majordomo-info.html
    Please read the FAQ at http://www.tux.org/lkml/


  • Next message: Manfred Spraul: "Re: [PATCH] [2.4] forcedeth network driver"

    Relevant Pages

    • Re: [PATCH 4/6] RTC: Ratelimit "lost interrupts" message
      ... I got it a lot when using any sort of virtualization on my Core2Duo ... We at least jack up the rtc for fast clock timeouts with better granularity - so it can be a host kernel issue, although a guest kernel trying to get high frequency RTC will also be inaccurate, and inevitably will have unhidable interrupt lateness. ...
      (Linux-Kernel)
    • Re: Interrupts
      ... 14: 9143 IO-APIC-edge ide0 ... Interrupt: pin A routed to IRQ 169 ... Warum sieht der Kernel da diese hohen ...
      (de.comp.os.unix.linux.hardware)
    • [2.4.22-pre9] Unexpected IO-APIC (works)...
      ... but the kernel reports finding an unexpected IO-APIC: ... Enabling APIC mode: Flat. ... PCI->APIC IRQ transform: -> 16 ... 4: 1282 IO-APIC-edge serial ...
      (Linux-Kernel)
    • Re: IO-APIC on nforce2
      ... or patched io_apic.c and process.c and using kernel arg "idle=C1halt"? ... 12: 13499 IO-APIC-edge PS/2 Mouse ... with my new timer setup ... number of MP IRQ sources: ...
      (Linux-Kernel)
    • Re: VIA KT400 + Kernel 2.6.12 + IO-APIC + uhci_hcd = IRQ trouble
      ... >Research done with Alan shows that, on my system, the USB 2.0 controller seems ... >to generate interrupts on the IRQ line attributed to the USB 1.1 controller, ... With a 2.6.13-rc1-mm1 kernel, for example, ... 0: 201893429 IO-APIC-edge timer ...
      (Linux-Kernel)