John Plocher <John.Plocher at sun.com> wrote:

> [Roland and I are bouncing a proposal back and forth...]
>
> Garrett D'Amore wrote:
> > 4) analysis of big-picture performance issues, if any (does 64-bit run 
> > faster, or slower?  maybe run a some test of boot time analysis.)
>
> IIRC, the SPARC impact was "doesn't run faster" combined with
> "takes more disk space", rather than "runs significantly slower".

This is correct, but NV_89 still does not offer the minimum amount of 64 bit 
binaries. In former times, we did neither have ls(1) nor find(1) available as 
64 bit binary, so it was impossible to list files that need a 64 bit binary to 
even stat(2) the file. We now have ls(1) as 64 bit binary, so we are able to 
list the files. We still miss touch(1) in the list of 64 bit binaries
in order to be able to correct the time stamps for problematic files.

Back to the current case: there are programs that are known to be faster on 
sparc if compiled in 64 bit. This is in general true for all programs that 
know how to scan memory more efficiently if in 64 bit mode. A right grep(1)
implementation could e.g. benefit from this.

On amd64, all 64 bit binaries typically (with a few exceptions) run ~ 30% 
faster than their 32 bit couterparts.

Shouldn't we take take care of this difference? Is there no way to treat amd64
different from sparcv9?


> > 5) a rough estimate in the increase of the size of the media/miniroot 
> > for the SPARC port
>
> While interesting, this is IMO not an architectural issue, not part
> of this case (which isn't proposing to make a complete 64-bit
> miniroot), and with the way we are going with IPS and the size of
> disks (CD, DVD and rust) this is probably not a practical concern
> even if it ends up doubling the size (which I doubt it will).

I would be interested in a better aligned installed system. I do not care
that much how the installation is done unless the change could double the 
install speed.

J?rg

-- 
 EMail:joerg at schily.isdn.cs.tu-berlin.de (home) J?rg Schilling D-13353 Berlin
       js at cs.tu-berlin.de                (uni)  
       schilling at fokus.fraunhofer.de     (work) Blog: 
http://schily.blogspot.com/
 URL:  http://cdrecord.berlios.de/old/private/ ftp://ftp.berlios.de/pub/schily

Reply via email to