CPUTYPE?= in make.conf

tech-lists tech-lists at zyxst.net
Sun Oct 6 00:02:15 UTC 2019


Hi,

On Fri, Oct 04, 2019 at 07:16:26PM +0200, Dimitry Andric wrote:

>This is not a hardware failure, but totally expected behavior.  You can
>only run poudriere builds for CPUs of an equal, or earlier generation,
>and even then you will have to ensure that the target CPU does not
>accidentally support instructions from your host CPU.

I was surprised to find that the sandybridge (Xenon E5-2690) would not compile for
CPUTYPE?=bonnell (this is intel atom N450) as the N450 is a lot (2010) older.

But the haswell compiles for CPUTYPE?=bonnell fine.

This is what made me wonder if there's some kind of fault with the
Xenon E5-2690. Or maybe it just can't use CPUTYPE?= properly? Does its use
require any special capability of the host chip?

If I don't put any CPUTYPE?= in poudriere's make.conf on the Xenon, 
everything compiles as expected without error. Maybe I should just do this ;)

*** although I'm going to continue compiling for btver1 on the haswell as the
resulting system feels more responsive.

thanks,

-- 
J.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 833 bytes
Desc: not available
URL: <http://lists.freebsd.org/pipermail/freebsd-hackers/attachments/20191006/94b624d0/attachment.sig>


More information about the freebsd-hackers mailing list