Re: Z shell signal handling

[Bart Schaefer <schaefer@xxxxxxxxxxxxxxxx>]

On May 17,  3:15am, Vincent Stemen wrote:
} Subject: Re: Z shell signal handling
}
} There are a couple potential problems I can see with the sh approach,
} that I am guessing zsh and bash were trying to fix.

The biggest problem is that C library routines like malloc() are not
re-entrant, and are heavily used by the shell.  The issue is not with
race conditions, etc., at the level of shell function execution, but
deep within the shell implementation, possibly in places that are not
within the shell implementer's control.

When PWS speaks about whether it's "safe" to enable signals, he means
that if it's NOT safe, the shell is going to corrupt memory, possibly
crash, or even open holes in system security.  Behaving the way that
the script programmer expects is secondary to these considerations.