On Tue, 15 Apr 1997, Ingo Molnar wrote:
> the attached patch adds more signals to Linux. I'm kindof stuck how to do
> it right, thats why i dump it here, suggestions would be more than welcome
> at this stage ...
> - A bigger problem seems to be that sa->sa_mask has to be a bitfield too,
> which in turn makes it a more natural solution to split extended signals
> and 'normal' signals totally? I didnt do it this way because such dual
> APIs are a rather microsoftish approach, but maybe it's still the
> cleanest one ;)
I believe we will eventually see a new API appear (from POSIX) for
signals. The sa_mask is a big enough problem to make the 'extend' bit
approach a dead-end. The whole thing smells of poll() vs. select().
Glad to see you are using the slab-allocator :)
Regards,
markhe
------------------------------------------------------------------
Mark Hemment, Unix/C Software Engineer (Contractor)
markhe@nextd.demon.co.uk http://www.nextd.demon.co.uk/
"Success has many fathers, failure is a B**TARD!" - anon
------------------------------------------------------------------