lockup on interrupt - VxWorks

This is a discussion on lockup on interrupt - VxWorks ; Hi, I'm trying to get an ethernet PCMCIA card driver running on a PowerPC running VxWorks 5.3. The attach routine calls intConnect() and sysIntEnable(5). As soon as I call usrNetInit, the processor (or at least serial port I'm using kermit ...

+ Reply to Thread
Results 1 to 3 of 3

Thread: lockup on interrupt

  1. lockup on interrupt

    Hi,

    I'm trying to get an ethernet PCMCIA card driver running on a PowerPC
    running VxWorks 5.3. The attach routine calls intConnect() and
    sysIntEnable(5). As soon as I call usrNetInit, the processor (or at
    least serial port I'm using kermit through), locks up.

    I hacked pcmattach to not enable the interrupt. Then pcmattach
    succeeds and retrieves and prints the MAC address from the card. Then
    as soon as I either type sysIntEnable(5) or plug in the ethernet cable
    (if I already had sysIntEnable'd), the processor locks up.

    It doesn't look like the ISR is being called at all. I put a
    logMessage("hello") as the first inst, never saw anything - but then I'm
    not sure I would.

    Any ideas? How do you go about debugging a problem like this?

    Thanks.
    Alex

  2. Re: lockup on interrupt

    Dear Alex:
    What I usually do is to start with a scope and convince myself that the
    interrupt line I think is asserting is truly the interrupt line that is
    asserting. Sometimes, what the hardware is doing and what the software
    indicates it is doing are not quite the same thing. Interrupt lines are one
    of those where counting from zero rather then one between schematic
    designers and software engineers sometimes leads to trouble.

    I sometimes setup a push button on a PCI or PCMCIA extender card so I
    can push a button and see a logMsg be printed. With a button, one can
    disconnect from the network interface if desired.

    Charles Krinke (former section 357 employee, many years ago)


    "Alex Murray" wrote in message
    news:de8dgq$5ls$1@nntp1.jpl.nasa.gov...
    > Hi,
    >
    > I'm trying to get an ethernet PCMCIA card driver running on a PowerPC
    > running VxWorks 5.3. The attach routine calls intConnect() and
    > sysIntEnable(5). As soon as I call usrNetInit, the processor (or at
    > least serial port I'm using kermit through), locks up.
    >
    > I hacked pcmattach to not enable the interrupt. Then pcmattach
    > succeeds and retrieves and prints the MAC address from the card. Then
    > as soon as I either type sysIntEnable(5) or plug in the ethernet cable
    > (if I already had sysIntEnable'd), the processor locks up.
    >
    > It doesn't look like the ISR is being called at all. I put a
    > logMessage("hello") as the first inst, never saw anything - but then I'm
    > not sure I would.
    >
    > Any ideas? How do you go about debugging a problem like this?
    >
    > Thanks.
    > Alex




  3. Re: lockup on interrupt

    Thanks Charles;

    I'll give that a try.

    Alex

    cfk wrote:
    > Dear Alex:
    > What I usually do is to start with a scope and convince myself that the
    > interrupt line I think is asserting is truly the interrupt line that is
    > asserting. Sometimes, what the hardware is doing and what the software
    > indicates it is doing are not quite the same thing. Interrupt lines are one
    > of those where counting from zero rather then one between schematic
    > designers and software engineers sometimes leads to trouble.
    >
    > I sometimes setup a push button on a PCI or PCMCIA extender card so I
    > can push a button and see a logMsg be printed. With a button, one can
    > disconnect from the network interface if desired.
    >
    > Charles Krinke (former section 357 employee, many years ago)
    >
    >
    > "Alex Murray" wrote in message
    > news:de8dgq$5ls$1@nntp1.jpl.nasa.gov...
    >
    >>Hi,
    >>
    >> I'm trying to get an ethernet PCMCIA card driver running on a PowerPC
    >>running VxWorks 5.3. The attach routine calls intConnect() and
    >>sysIntEnable(5). As soon as I call usrNetInit, the processor (or at
    >>least serial port I'm using kermit through), locks up.
    >>
    >> I hacked pcmattach to not enable the interrupt. Then pcmattach
    >>succeeds and retrieves and prints the MAC address from the card. Then
    >>as soon as I either type sysIntEnable(5) or plug in the ethernet cable
    >>(if I already had sysIntEnable'd), the processor locks up.
    >>
    >> It doesn't look like the ISR is being called at all. I put a
    >>logMessage("hello") as the first inst, never saw anything - but then I'm
    >>not sure I would.
    >>
    >>Any ideas? How do you go about debugging a problem like this?
    >>
    >> Thanks.
    >>Alex

    >
    >
    >


+ Reply to Thread