[mercury-users] problem with signals

Pieter Laeremans pieter.laeremans at student.kuleuven.ac.be
Tue Mar 11 06:42:47 AEDT 2003

On Thu, 2003-03-06 at 15:04, Fergus Henderson wrote:
> one has an explanation for this ?
> By default, the Mercury compiler will invoke the C compiler with
> `-ansi', which will give you an ANSI/ISO C compliant set of standard
> header files.  In standard ISO C, the <signal.h> header file does not
> provide sigemptyset() etc.  If you want to make use of Posix extensions,
> such as sigemptyset() etc., you need to use the `--no-ansi-c' option to
> mmc and/or the `--no-ansi' option to mgnuc.
> This mess is needed because Posix chose to standardize the same header
> names as ANSI/ISO C, but with different contents.

Thx this fixes the compilation problem.  However another problem
emerged.  The sigprocmask I used to block for signals didn't had an
effect.  I wondered why now I think I now. 

ps -e showed me that therre were to processes of my mercury program. 
Although I never spawn any thread somewhere. And signals are only
delivered to one thread. 

Does the mercury runtime in someway spawns a new thread when a signal
gets delivered to program ?

Or does anyone have an idea how I can detect a which moment the second
thread emerges ?

thanks in advance,


mercury-users mailing list
post:  mercury-users at cs.mu.oz.au
administrative address: owner-mercury-users at cs.mu.oz.au
unsubscribe: Address: mercury-users-request at cs.mu.oz.au Message: unsubscribe
subscribe:   Address: mercury-users-request at cs.mu.oz.au Message: subscribe

More information about the users mailing list