In message <20040919211630.GA24628@antioche.eu.org>, Manuel Bouyer writes:
>On Sun, Sep 19, 2004 at 10:25:56PM +0200, Jochen Kunz wrote:
>> On Sun, 19 Sep 2004 21:06:12 +0200
>> Manuel Bouyer <bouyer@antioche.eu.org> wrote:
>> 
>> > I don't know if it is safe to call without BUS_DMA_NOWAIT here.
>> > It needs to be called from a valid thread context to be.
>> I don't know how the PCMCIA / cardbus stuff works exactely. But isn't
>> there a kernel thread watching the slots for insert / removal events and
>> calling the appropriate MI attach / detach routines of the driver? Maybe
>
>I think there is one, yes.
>
>> it is possible to pass a flag to the MI attach routine to signal the
>> attach code that it is save to sleep during attach?
>
>This should be possible, but if the problem is getting physically contigous
>pages, this may not be enouth.
>

Hmm -- this is the same problem with have with USB umass devices.  It 
would be good to find a more general solution.

		--Steve Bellovin, http://www.research.att.com/~smb