Mail Archives: cygwin/1997/06/17/20:33:12
In article <01BC7B05 DOT DA103740 AT gater DOT krystalbank DOT msk DOT ru>,
Sergey Okhapkin <sos AT prospect DOT com DOT ru> wrote:
>Chris Faylor wrote:
>>
>> #ifdef POSIX_SIGNALS
>> struct sigaction act;
>>
>> act.sa_handler = (SIGNAL_HANDTYPE) handler;
>> sigemptyset(&act.sa_mask); /* only block sig while in handler
>*/
>> act.sa_flags = 0;
>> # ifdef SA_INTERRUPT /* SunOS 4.x */
>> if (interact)
>> act.sa_flags |= SA_INTERRUPT; /* make sure system calls are not
>restarte
>> d */
>> # endif
>> sigaction(sig, &act, (struct sigaction *)NULL);
>>
>> ----
>>
>> Where the 'sig' is 'SIGCHLD'.
>>
>> It appears that 'sig' is blocked while within the handler. Could that be
>> what is going wrong?
>
>Cygwin.dll blocks the signal while executing signal's handler. If arrived
>signal is blocked, it will be added to process's sig_pending mask. On exit
>from any handler all pending signals are raised again. See exceptions.cc
>for details. Probably, we have to change sig_pending from a bit mask to an
>array of counters to avoid possible losts of a signals. What do you think
>about?
That sounds right to me. Otherwise you lose multiple arrival of signals.
I did take the time to debug this yesterday, and I am mystified as to why
it is failing. When I set a breakpoint on the signal handler, it is
hit every time a subprocess dies, and 'zsh' continues on as it should.
When I remove the breakpoint zsh 'hangs' waiting for SIGCHLD. Weird.
--
http://www.bbc.com/ cgf AT bbc DOT com "Strange how unreal
VMS=>UNIX Solutions Boston Business Computing the real can be."
-
For help on using this list (especially unsubscribing), send a message to
"gnu-win32-request AT cygnus DOT com" with one line of text: "help".
- Raw text -