| To: | "'yyyyyyyyyyyyyyy@xxxxxx'" <yyyyyyyyyyyyyyy@xxxxxx> |
|---|---|
| Subject: | RE: set -e and SIGCHLD |
| From: | Bruce Korb <yyyyy@xxxxxxxxxxxxxxxxxxx> |
| Date: | Tue, 13 Mar 2001 16:28:48 -0800 |
| Cc: | yyyyyyyyyyyyyy@xxxxxxxxxxxxx |
> No. But if you are the programmer who sets the action for
> SIGCHLD to SIG_IGN, disregarding the POSIX warnings, then
> you are responsible for what the program does. One way of
> making everybody happy would be to set the action for
> SIGCHLD back to SIG_DFL after the fork but before the exec
> of any child.
You cannot do that, except perhaps with a wrapper around
the exec call. Re: "disregarding the POSIX warnings":
1. They are amazingly obscure and many (most?) people
even on this list did not fully understand consequences
2. There is _no_ conceivable use for passing SIG_IGN through
to child processes. The _only_ argument put forth for
it was "it is the standard". If something is not used,
has no purpose, causes problems and is misunderstood by
POSIX standards folks themselves, then why would anyone
possibly want to keep it as standard? Changing it can
hurt nobody and only help those who read man pages
instead of POSIX tomes. Change it, please.
|
| Previous by Date: | Re: set -e and SIGCHLD, Marc Aurele La France |
|---|---|
| Next by Date: | Re: set -e and SIGCHLD, Marc Aurele La France |
| Previous by Thread: | Re: set -e and SIGCHLD, Marc Aurele La France |
| Next by Thread: | Re: Re: set -e and SIGCHLD, David Korn |
| Indexes: | [Date] [Thread] [All Lists] |