cvs commit: src/sys/compat/linux linux_socket.c

[John Baldwin]

On Wednesday 22 March 2006 03:49, Alexander Leidinger wrote:
> Alexander Leidinger <netchild at FreeBSD.org> wrote:
> 
> > netchild    2006-03-21 21:56:04 UTC
> >
> >  FreeBSD src repository
> >
> >  Modified files:
> >    sys/compat/linux     linux_socket.c
> >  Log:
> 
> >  - make some functions globally visible on alpha
> 
> According to the CVS history someone tried those functions on alpha in the
> past, but they failed to work correctly. Unfortunately the cause of the
> breakage isn't recorded. Since then a lot of commits where done and the
> problem may or may not exists anymore. If someone is able to actually test
> the linuxolator on alpha (I have some doubts that we have a linux_base which
> works on alpha *and* is still supported by the distributor and as such is
> also safe from a security point of view), I would be pleased to hear about
> success or failure. And in case of a failure I may have some ideas how to
> fix it on alpha...

This is probably because Linux/alpha tends to be compatiable with Tru64
(OSF/1) rather than Linux/i386.

> If nobody is using the linuxolator on alpha with a recent FreeBSD (either
> because everything is used on x86/amd64 hardware, and/or because there's no
> usable linux_base for alpha), I think it would be best to remove the linux
> part from alpha. It would make the code in compat/linux cleaner.

Or just undo what you did above and leave it as it is. :)  I've done various
things on the compat code over and haven't been really bothered by the
#ifdef __alpha__'s.

-- 
John Baldwin <jhb at FreeBSD.org>  <><  http://www.FreeBSD.org/~jhb/
"Power Users Use the Power to Serve"  =  http://www.FreeBSD.org