SH multilib removal possibility "-m1 -ml" & "-m2 -ml"

[tm]

Joel Sherrill wrote:
> 
> tm wrote:
> >
> > I've talked to some people, and evidently the SH1 and SH2 do not support
> > little-endian. The processors are hardwired for big-endian only.
> >
> > Can we remove the multilib configurations for these to improve build times
> > for the other SH targets?
> 
> If the sh1 and sh2 do not support little endian, this is certainly
> acceptable to the RTEMS community.  We just want to make sure
> sh1 and sh2 support do not disappear entirely.

Yep. The hardware definitely does not support little endian on
SH1 and SH2. Talked to the designer.

Toshi