Wolfgang Solfrank wrote:

> I certainly didn't mean to say that "I'll change tsleep to work in an
> interrupt context".  To the contrary, I answered to Lennart's suggestion:
> "Yes, it's probably a good idea to make tsleep panic if called from
> an interrupt handler".

You didn't :-)

I was actually replying to an earlier message still in my mind, when
someone said "I'll fix it, then." That just came back and bit me! Sorry
for confusion.

	Neil