On Fri, Apr 14, 2006 at 10:04:10AM -0700, Bart Smaalders wrote:
> ....
> But to tell you the truth, that's not a very interesting part of
> the space to me.  Sun4m shipped w/ 75 Mhz cpus max (discounting the old
> HyperBrokens). They were obsoleted when UltraSparc shipped in
> 1995, 11 years ago.  Constraining all the software to run on that
> hardware would be like only supplying software that ran on 486
> machines; doable but clunky and slow.
> 
> For me, having the OpenSource software unable to take advantage
> of the last 10 years of evolution in instruction sets and
> 6 years of evolution in operating system features is a non-starter.



FYI: blastwave packages are constrained to support solaris 8 and
sparcv8, as a *minimum*. Some packages can and do support more advanced
architectures, *if* there is a noticable benefit.

We take advantage of isaexec to do this, and -R$ISALIST.


http://www.blastwave.org/standards/build.html

mentions specific examples of when a maintainer should consider providing 
64bit binaries. but the same logic applies for whether or not to supply an
sparcv8plus+vis executable:

"If an executable is noticably(10%+) faster "


Currently, the following CSW packages already provide a sparcv8plus+vis
optimized binary, in addition to the standard sparcv8 ones,according to
 http://www.blastwave.org/filesearch

mpg123
lame
radiance
mplayer
fox

This does not include libraries that may be cpu-optimized. We have many
more packages that offer sparcv8plus+vis optimized libraries, alongside the
plain sparcv8 ones.

Usually, there is more benefit in compiling the libraries that way, than
the top-level executables :-)



Similarly.. there are certain features that only exist in solaris 10.
We *are* supporting those through various mechanisms, usually involving a
run-time postinstall adjustment to the package.
(ie: SMF support)


Reply via email to