Skip to content
Prev 23950 / 63421 Next

"try"ing to understand condition handling (interrupts)

On Tue, 20 Feb 2007, Ross Boylan wrote:

            
This has nothing to do with thread safety.  It has everything to do
with what is safe to do within signal handlers on the one hand and OS
specific variations on what happens when signals interrupt a system
call on the other.
Experiments like this are useful but don't expect results you get to
be reliable in the future.  What the call stack as shown by sys.xyz
looks like in particular at handling time may change (probably not
much for calling handlers but most likely for exiting ones).

Best,

luke