Date: Wed, 22 Apr 2015 17:52:04 -0700 From: Adrian Chadd <adrian@freebsd.org> To: "freebsd-arch@freebsd.org" <freebsd-arch@freebsd.org> Subject: Re: RFT: numa policy branch Message-ID: <CAJ-VmonCp7VDWrSXhiQ5PwcCogPM8NG6tDjQRy8osUQw=uUYKQ@mail.gmail.com> In-Reply-To: <CAJ-VmokPd=CUAfqmjWPns%2Bpj6zKbpF55tDn2_u8JPNzaK7F1Pw@mail.gmail.com> References: <CAJ-VmomL9hZZHPtZ3%2BTdujHmo5UQfFhm59vQKUbxW%2B%2B-TGobmg@mail.gmail.com> <CAJ-VmokPd=CUAfqmjWPns%2Bpj6zKbpF55tDn2_u8JPNzaK7F1Pw@mail.gmail.com>
next in thread | previous in thread | raw e-mail | index | archive | help
On 21 April 2015 at 20:03, Adrian Chadd <adrian@freebsd.org> wrote: > OH, and the branch: > > https://github.com/erikarn/freebsd/tree/local/adrian_numa_policy Hi! Update: * the whole setup/copy process for thread and proc domain policies is slightly less dirty now; * the phys layer now checks domain policy in this order:: thread -> proc -> default; so now setting a proc policy will take action for all threads in that proc that don't have a more specific domain policy; * numactl is slightly less terrible to use. Todo: * for correctness, I should call the free methods on the domain policy whenever a thread/proc is destroyed. * .. write manpages for all of this. * Test on AMD NUMA systems - who has one I can poke at? -adrian
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?CAJ-VmonCp7VDWrSXhiQ5PwcCogPM8NG6tDjQRy8osUQw=uUYKQ>