[RFC PATCH glibc] pthread_setspecific: Provide signal-safety across keys

[Joseph Myers]

On Tue, 17 Oct 2017, Mathieu Desnoyers wrote:

> @c pthread_setspecific @asucorrupt @ascuheap @acucorrupt @acsmem
> @c   a level2 block may be allocated by a signal handler after
> @c   another call already made a decision to allocate it, thus losing
> @c   the allocated value.  the seq number is updated before the
> @c   value, which might cause an earlier-generation value to seem
> @c   current if setspecific is cancelled or interrupted by a signal
> @c  KEY_UNUSED ok
> @c  calloc dup @ascuheap @acsmem
> Associate the thread-specific @var{value} with @var{key} in the calling thread.
> @end deftypefun
> 
> We'd need to edit the part about level2 block and allocation.

That part is *comments* explaining to people reading the manual source 
code why the particular user-visible @safety{} annotations were determined 
from the library sources.

It's true we should keep those up to date when the library sources change.  
But my point was that there should be *user-visible* documentation - not 
just comments in the manual sources - of any safety guarantees we choose 
to provide beyond POSIX.

-- 
Joseph S. Myers
joseph@codesourcery.com