[cgl_discussion] Re: sanity check for KGDB?

Andy Pfiffer andyp at osdl.org
Fri Oct 4 10:28:12 PDT 2002


George,

FYI: I tried your diff against the head of the tree (whatever it looked
like yesterday afternoon) and it went in cleanly.  I did not try to test
kgdb (something that went into the tree yesterday seems to have broken
the compile of aic7xxx_linux.c...oh look, it seems to be fixed now...)

As far as a config lockout for kdb -or- kgdb (but not both), Randy
suggested that we might look at the USB UHCI config option as an
example.  I believe the relevant sample looks like this:

    comment 'USB Controllers'
    if [ "$CONFIG_USB_UHCI_ALT" != "y" ]; then
       dep_tristate '  UHCI (Intel PIIX4, VIA, ...) support'
    CONFIG_USB_UHCI $CONFIG_USB
    fi
    if [ "$CONFIG_USB_UHCI" != "y" ]; then
       dep_tristate '  UHCI Alternate Driver (JE) support'
    CONFIG_USB_UHCI_ALT $CONFIG_USB
    else
       define_bool CONFIG_USB_UHCI_ALT n
    fi

Could we do something similar, say CONFIG_KERNEL_DEBUGGER, and then have
mutually exclusive options (as above) for CONFIG_KERNEL_KDB and
CONFIG_KERNEL_KGDB?

Andy





More information about the cgl_discussion mailing list