Re: "cpuset -n prefer:?" --what values for "?" are supposed to be allowed? (only 1 is, despite two numa domains)

From: Clay Daniels Jr. <clay.daniels.jr_at_gmail.com>
Date: Wed, 11 Sep 2019 01:28:30 -0500
Mark, this is what I get on my machine:

root_at_new:~ # cpuset -g
pid -1 mask: 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15
pid -1 domain policy: first-touch mask: 0
root_at_new:~ #  cpuset -l0 -n prefer:0 COMMAND
cpuset: COMMAND: No such file or directory
root_at_new:~ # cpuset -l0 -n prefer:2 COMMAND
cpuset: setdomain: Invalid argument
root_at_new:~ # cpuset -l0 -n prefer:1 COMMAND
cpuset: setdomain: Invalid argument

>From dmesg:
FreeBSD 13.0-CURRENT r351901 GENERIC amd64
CPU: AMD Ryzen 7 3700X 8-Core Processor(3600.08-MHz K8-class CPU)

Similar,
Clay

On Wed, Sep 11, 2019 at 12:58 AM Mark Millard <marklmi_at_yahoo.com> wrote:

> In a context with:
>
> # cpuset -g
> pid -1 mask: 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17,
> 18, 19, 20, 21, 22, 23, 24, 25, 26, 27
> pid -1 domain policy: first-touch mask: 0, 1
>
> I get:
>
> # cpuset -l0 -n prefer:0 COMMAND
> cpuset: setdomain: Invalid argument
>
> # cpuset -l0 -n prefer:2 COMMAND
> cpuset: setdomain: Invalid argument
>
> But one prefer:? value does allow the COMMAND
> to run:
>
> # cpuset -l0 -n prefer:1 COMMAND
>
> This seem odd to me. Am I missing something?
>
> For reference: I'm using a ThreadRipper 1950X
> with a head -r351227 based context for this
> activity. The above happens to have been run
> in a Windows 10 Pro HyperV session, instead
> of in a native-boot of the same media. (A
> native-boot would have had 32 CPUs.)
>
>
> ===
> Mark Millard
> marklmi at yahoo.com
> ( dsl-only.net went
> away in early 2018-Mar)
>
> _______________________________________________
> freebsd-current_at_freebsd.org mailing list
> https://lists.freebsd.org/mailman/listinfo/freebsd-current
> To unsubscribe, send any mail to "freebsd-current-unsubscribe_at_freebsd.org"
>
Received on Wed Sep 11 2019 - 04:28:44 UTC

This archive was generated by hypermail 2.4.0 : Wed May 19 2021 - 11:41:21 UTC