On Mon, Mar 01, 2004 at 10:25:02PM -0600, David Young wrote:
>On Mon, Mar 01, 2004 at 11:01:48PM -0500, Andrew Brown wrote:
>> On Mon, Mar 01, 2004 at 12:46:07PM -0600, David Young wrote:
>> >On Mon, Mar 01, 2004 at 10:11:53PM +1100, Daniel Carosone wrote:
>> >> On Sun, Feb 29, 2004 at 08:08:04PM -0600, David Young wrote:
>> >> > I've added sysctls for ath(4) in ath.* (that's right, top-level). I
>> >> > don't intend for this to be permanent. Where do the sysctls belong?
>> >> > I believe it was Dan Carosone who suggested,
>> >> > 
>> >> > 	driver.ath.<global variable>
>> >> > 	driver.ath.0.<instance variable>
>> >> 
>> >> driver.ath.0 or dev.ath0 seems fine
>> >> 
>> >> btw, these sysctl's appear, with values in them, in a running GENERIC
>> >> even though i have no ath(4) attached.  An oversight?
>> >
>> >I feel that ath.debug and ath.hal.debug (shows w/ options AH_DEBUG)
>> >belong. I think that ath.outdoor influences device attachment, so it
>> >should stay, also.
>> 
>> well...perhaps, but should they show up even if the device is not
>> attached?
>
>I was not clear, sorry.  If I am debugging device attachment, but the
>ath.debug sysctl only appears *after* attachment, it is too late to
>frob it.

that occurred to me last night, too.

then it occurred to me that you can't frob anything via sysctl until
userland is running, which takes place after auto-configuration, which
is where the device will be found.

unconfuse me?

-- 
|-----< "CODE WARRIOR" >-----|
codewarrior@daemon.org             * "ah!  i see you have the internet
twofsonet@graffiti.com (Andrew Brown)                that goes *ping*!"
werdna@squooshy.com       * "information is power -- share the wealth."