This is the mail archive of the
libc-alpha@sources.redhat.com
mailing list for the glibc project.
Re: [patch] Matsushita AM33/2.0 port
- From: Roland McGrath <roland at redhat dot com>
- To: Jakub Jelinek <jakub at redhat dot com>
- Cc: Alexandre Oliva <aoliva at redhat dot com>,libc-alpha at sources dot redhat dot com
- Date: Thu, 24 Jun 2004 17:46:16 -0700
- Subject: Re: [patch] Matsushita AM33/2.0 port
> Maybe we could add nptl to nptl/sysdeps/unix/sysv/linux/Implies and
> linuxthreads to linuxthreads/sysdeps/unix/sysv/linux/Implies.
> Then if you configure --enable-add-ons=myport,nptl also
> myport/sysdeps/nptl
> would be searched up and similarly for linuxthreads.
I would prefer to have add-ons' sysdeps trees structured so the sysdeps
tree of the main tree and all kosher add-ons can be unioned without
conflicts. This makes it easier to figure out reading things, and also
makes it possible to take things in/out of the main tree and/or join
multiple add-ons into one.
It's hard to do that with straight Implies files in the add-ons. Another
idea is to use add-ons names on the end of the sysdeps search pattern.
Then --enable-add-ons=myport,nptl would find
myport/sysdeps/unix/sysv/linux/CPU/nptl
This is even more obfuscation of where to find the code. If we do this, I
think we should go with the convention of putting nptl code in the
unadorned myport/sysdeps/... dir and linuxthreads files in
myport/sysdeps/.../linuxthreads, where they will take priority when
linuxthreads is enabled. Then there is less gratuitous separation later on
when linuxthreads is eventually dropped from the tree.
Thanks,
Roland