On Sat, Oct 27, 2018 at 11:18:37PM +0100, Ken Moffat via blfs-dev wrote:
> On Sat, Oct 27, 2018 at 06:02:29AM +0100, Ken Moffat via blfs-dev wrote:
>
> 2. The other idea I got was to try -jN on ./mach build. It needs
> to come after the target, ./mach -j4 build is quickly spat out.
> This appears to prevent my Ryzen 3 falling back to 1 core for long
> periods of time.
>
Trying with -j4 (and sometimes -j5) on other machines (all are
BLFS-8.3 with llvm-6.0 and updated rustc etc)
intel haswell i7 from taskset -pc 4-7 $$ : disastrous, effectively
only one core for the whole build.
intel haswell i7 with all cores available, -j4 19 SBU
intel sandybridge i3, -j4 33 SBU similar to normal
amd phenom (4 real cores, but solow) -j4 18 SBU, -j5 slightly
slower.
To recap: on the ryzen3 I need to specify some -j value to stop it
mostly dropping back to pretending there is only 1 core. With -j4
on 8.3 I got down to 24SBU, with -j5 I hit 20 SBU.
The ryzen is the only machine with llvm-7, and that is definitely
slower (as well as bigger).
I've updated my bug report at mozilla, although I doubt anything
will come from that. Now I'm still thinking about how to word this
without straying into Al Viro's excellent terminology.
ĸen
--
Is it about a bicycle ?
--
http://lists.linuxfromscratch.org/listinfo/blfs-dev
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page
