Re: AC_SYS_RESTARTABLE_SYSCALLS

[Johan Danielsson]

David Oleszkiewicz <address@hidden> writes:

> AC_SYS_RESTARTABLE_SYSCALLS
>    If the system automatically restarts a system call that is interrupted
> by a signal, define HAVE_RESTARTABLE_SYSCALLS.

[...]

> which would seem to indicate that if the system automatically does not
> restart system calls but rather wait() returns -1 (and presumably sets
> errno to ERESTARTSYS or EINTR) the test would exit successfully meaning
> no_default_restartable_syscall.

Is there a difference in the documented behaviour, and your
description of the same? The keyword here is "automatically".

> This result having no indication of whether the system is _able_ to
> support automatically restarting system calls.

The test does not claim to do that either.

> what am i missing.

That it's documented in the "Obsolete Macros" section (at least in
Autoconf 2.52):

 - Macro: AC_SYS_RESTARTABLE_SYSCALLS
     If the system automatically restarts a system call that is
     interrupted by a signal, define `HAVE_RESTARTABLE_SYSCALLS'. This
     macro does not check if system calls are restarted in general-it
     tests whether a signal handler installed with `signal' (but not
     `sigaction') causes system calls to be restarted. It does not test
     if system calls can be restarted when interrupted by signals that
     have no handler.

     These days portable programs should use `sigaction' with
     `SA_RESTART' if they want restartable system calls.  They should
     not rely on `HAVE_RESTARTABLE_SYSCALLS', since nowadays whether a
     system call is restartable is a dynamic issue, not a
     configuration-time issue.

/Johan