POSIX threads implementation
Chris Faylor
cgf@cygnus.com
Mon Aug 16 10:27:00 GMT 1999
On Mon, Aug 16, 1999 at 07:16:54PM +0200, Pascal Vittone wrote:
>I am used to work with Solaris' POSIX threads. I tried to use CYGWIN's
>threads but I had a lot of problem. So, I read the code. And I think
>there is some problems: when a process receives a signal, all the
>threads go out of the blocking calls; an exited thread can be chosen to
>process a signal; pthread_sigmask modifies process's signal mask;
>signals associated to exceptions are not always processed by the
>generating thread; ...
That sounds correct. The current signal mechanism is not completely
correct in its signal delivery.
>Is my analysis correct? Are these problems known? Are some of them
>features? Do you provide a description of your thread support
>implementation? Is someone currently working on threads
>implementation? Should I try to correct myself these problems?
No one is working on threads, AFAIK. John Fortin has submitted some
patches but I don't know if he is actively working on this.
The only information available regarding pthreads is in the winsup
directory.
-chris
More information about the Cygwin-developers
mailing list