if_em panic

[M. Warner Losh]

In message: <43C718EF.9050301 at elischer.org>
            Julian Elischer <julian at elischer.org> writes:
: Huang wen hui wrote:
: 
: >hi,
: >when I load if_em on TP42P got this panic:
: >  
: >
: 
: looks like the first interrupt occurs before the driver has finished
: setting itself up..
: (just a first impression).
: Scott'll probably nail it.
: 
: I'm guessing it works fine if compiled in..

Fast interrupts are not masked during device configuration, but slow
interrupts are.  This is why people are seeing this now.  Even if the
em device itself didn't interrupt, it is possible (likely) that other
devices that share an interrupt with it have...

Warner