Re: Design hole in nptl (and more general in POSIX threads)

Дана суб, 20-12-2003 у 06:47, Jamie Lokier је написао:
...
> > pthread_self (which isn't officially async-signal-safe, but geez)
> 
> On some implementations pthread_self has to search a data structure
> which is modified and hence locked whenever threads are created or
> destroyed.  It's not surprising that it's not async-signal-safe.  I
> think an older version of Glibc threads was like this on some
> architectures.

  As usual, last bit brings fresh air :).

  I hope this is good enough, As my suspension of other threads happens
in critical section shared with Fork() procedure which is only place
where I create new threads. 

  One point is destroying of threads, When this exactly happens? As
there is no manager thread and none of my threads (excepts suspendor)
currently creates new thread, no other thread can mess with mentioned
structures?

> 
> -- Jamie
> 

  dd