I486_CPU and I586_CPU removed from GENERIC kernel [was Re: svn commit: r205307 - head/sys/i386/conf]

[John Baldwin]

On Friday 19 March 2010 12:27:21 pm Xin LI wrote:
> On 2010/03/18 23:33, Garrett Cooper wrote:
> > On Thu, Mar 18, 2010 at 6:16 PM, Xin LI <delphij at freebsd.org> wrote:
> >> Author: delphij
> >> Date: Fri Mar 19 01:16:53 2010
> >> New Revision: 205307
> >> URL: http://svn.freebsd.org/changeset/base/205307
> >>
> >> Log:
> >>  SSE is enabled by default about 5 years ago so there is no point 
pretending
> >>  that we support I486 and I586 CPUs in the GENERIC kernel, users wants 
these
> >>  support would have to build a custom kernel to explicitly disable SSE
> >>  anyways.
> >>
> >>  MFC after:    1 month
> >>
> >> Modified:
> >>  head/sys/i386/conf/GENERIC
> >>
> >> Modified: head/sys/i386/conf/GENERIC
> >> 
==============================================================================
> >> --- head/sys/i386/conf/GENERIC  Fri Mar 19 00:51:48 2010        (r205306)
> >> +++ head/sys/i386/conf/GENERIC  Fri Mar 19 01:16:53 2010        (r205307)
> >> @@ -18,8 +18,6 @@
> >>  #
> >>  # $FreeBSD$
> >>
> >> -cpu            I486_CPU
> >> -cpu            I586_CPU
> >>  cpu            I686_CPU
> >>  ident          GENERIC
> >
> >     1. UPDATING entry ?
> >     2. CC -current@ with the news?
> 
> Perhaps not, I was wrong on this: CPU_ENABLE_SSE would compile in the
> support for SSE, not enforcing it.  Our lib32 on the other hand already
> uses -i686 -sse -sse2 and -mmx so I'm just cutting the wrong foot I
> guess :-/

I believe the lib32 bits assume they will always run on an amd64-capable CPU 
in which case SSE2 is guaranteed to be present.  Similarly, I think the lib32 
variant of libc uses a different method of setting up TLS than the native i386 
version (I think the lib32 libc uses a GSBASE sysarch() directly vs what an 
i386 libc does, or at least this used to be true at one point in the past if 
not currently true).

-- 
John Baldwin