Recovery possible from signal aborted write(2) ?

[Henry Spencer]

> By the way, is there any reason to not have u directly readable?
> It's been that way for about two years here at cithep, but I think
> only two or three people know it!

If you are absolutely, positively, stunningly confident that the
data structures in u are *never* *ever* going to change in even the
smallest way, then there is no reason not to have it readable.  If
you wish to preserve the option to change those data structures
without having people scream, then neither u nor /dev/kmem should be
readable at all.  I think you can tell which I prefer.
-- 
				Henry Spencer @ U of Toronto Zoology
				{allegra,ihnp4,linus,decvax}!utzoo!henry