AHA! So, we will loose all signals between the handler is called and python
can handle it! Because we call the signal handler signal only once, even if
the signal occured multiple times befroe the Python signal handler is called.
But you wrote about the SIGCHLD problem:
> Changing to System V semantics (requiring the Python handler to
> re-instate the handler) for all signals has the problem that there is
> a window within which signals may be lost (between the moment the C
> handler returns and the Python handler re-instates the handler).
Then I don't understand why it is a problem to delay the reinstallation of the
the signal handler until the signal is handled by python. Now, the C signal
handler reinstalls the signal immediately, which causes a problem for SIGCHLD
on SYSTEM V. But the reinstallation could be delayed until python handles the
signal, because the signal is regognized only once anyway. So it makes no sense
to reinstall the handler before python has handled it. Then the problem with
the SIGCHLD wouldd'nt be a problem anymore.
Michael
--
__________________________________________________________
**** ___ _ _ ___ _
****** | __) | \/ | | ) | | Michael Scharf
******** | _) | | | -< | |_ Tel: +49 6221 387 305 Fax: 517
* **** |___) |_||_| |___) |___) EMail: scharf@EMBL-Heidelberg.de
**** __________________________________________________________