Re: Killing clones (threads)

Werner Almesberger (almesber@lrc.di.epfl.ch)
Thu, 14 Aug 1997 10:29:17 +0200 (MET DST)

mjr@iki.fi wrote:
> Hmm. Sounds otherwise good, but I think POSIX threads require that
> signals be sent to exactly one thread, and it should be one whose
> sigmask permits it.

Yup, that's basically what 1003.1 1996-07-12 says in 3.3.1.2 "Signal
Generation and Delivery". Also, pthread_kill (3.3.10) requires the
ability to kill individual threads (but I'm not entirely sure if this
also applies to killing from an "outside" process, i.e. if pthread_t
is only valid inside the same process. It probably is.).

Anyway, what about: high_bits = 0 means "pick one", alias high_bits = 1
to the "root" thread, and make high_bits = 0xffff the "kill all" ?

- Werner

P.S. Added "threads" to the subject. I missed the beginning of this thread
because I thought it was about Linux failing miserably on specific
"clone" hardware :-(

-- 
  _________________________________________________________________________
 / Werner Almesberger, DI-LRC,EPFL,CH   werner.almesberger@lrc.di.epfl.ch /
/_IN_R_133__Tel_+41_21_693_6621__Fax_+41_21_693_6610_____________________/