This is the mail archive of the
newlib@sources.redhat.com
mailing list for the newlib project.
Re: libgloss/arm/libcfunc.c: alarm [PATCH]
- From: Shaun Jackman <sjackman at gmail dot com>
- To: Jeff Johnston <jjohnstn at redhat dot com>
- Cc: newlib at sources dot redhat dot com
- Date: Mon, 11 Jul 2005 16:08:06 -0700
- Subject: Re: libgloss/arm/libcfunc.c: alarm [PATCH]
- References: <7f45d9390505191414271badb2@mail.gmail.com> <42D2EB68.1090906@redhat.com>
- Reply-to: Shaun Jackman <sjackman at gmail dot com>
On 7/11/05, Jeff Johnston <jjohnstn@redhat.com> wrote:
> Originally, I was anticipating some of the other ARM maintainers/users
> might speak up. I have no real objection to moving alarm to libnosys,
> but have you considered the ramifications of your change? Unless the
> compiler generates a -lnosys reference, you will break linking for
> others. If -lnosys was added to the appropriate specs files that use
> libcfunc.o, that could solve the problem. I have no idea if anybody
> links without the spec file. ARM folks??
>
> -- Jeff J.
Removing the alarm stub from libcfunc.c will break the build for the
users that only require a stub definition of alarm, but it also
allows the users that need a working copy of alarm to redefine it.
Meanwhile, the stub users can either add -lnosys to their link or add
an alarm() {} stub to their own project. Chances are good that if
their successful build depends on a stub of alarm, their build also
depends on other stubs not defined by newlib (like pause for example)
and already have a scheme in place for providing them.
Cheers,
Shaun