> Is that the single case you need to do this though? What about SIGIO?
> Let's think about this now, before we end up having to kludge things more.

I have no idea yet of what happens with SIGIO, but I guess we'll run
into the same problem: calling the usaual signal code path is
unappropriate since it will filter out exceptions when the signal mask
filter the signal.

The easy fix is to call trapsignal, but it has some drawbaks on the
performance front. Doing so for a traced process is not such a problem
as you might not expect full performance when you debug, bug SIGIO is
another story. What do you think?

Maybe we should ifdef COMPAT_MACH in kernel code to call trapsignal
instead of the regular code path?

-- 
Emmanuel Dreyfus
Il y a 10 sortes de personnes dans le monde: ceux qui comprennent 
le binaire et ceux qui ne le comprennent pas.
manu@netbsd.org