On Thu, Jul 09, 2015 at 10:18:17PM -0000, [email protected] 
wrote:
> Author: bdubbs
> Date: Thu Jul  9 15:18:16 2015
> New Revision: 16226
> 
> Log:
> Update to openssl-1.0.2d.
[...]
> +<!-- Parallel build for version 1.0.2d at -j8 works for me without the 
> patch. 
> +     But it only reduces build time from 1.0 to 0.7 SBU - Bruce

Yes, I've just tried that on my i3 : with make -j4 the cpu monitor
on icewm's taskbar rarely exceeds 25% active.  Watching 'top' the
total for the CPU percentages mostly maxed out at 100 (instead of
an ideal 400), but there was one brief spell which was a _little_
higher.

No idea what was wrong with my attempt to fix up the patch, so I'm
glad we don't need it - thanks for trying without it.

Just on a whim, I just now tried make -j4 test : all passed, but
again it did not look as if it was using significantly more than 1 CPU.
But I hesitate to suggest removing the -j1 from make test unless
somebody with a more powerful machine is minded to test if it will
work for them.  And in any case, I'm not convinced that -j >1 is
giving any real benefit (I'm very surprised that you saw such a big
change in the SBU - unless the caches were warm after the first
build).

ĸen
-- 
This one goes up to eleven!
-- 
http://lists.linuxfromscratch.org/listinfo/blfs-book
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page

Reply via email to