From: Igor Mammedov <imammedo@redhat.com>
To: Michal Privoznik <mprivozn@redhat.com>
Cc: jingqi.liu@intel.com, tao3.xu@intel.com, qemu-devel@nongnu.org,
ehabkost@redhat.com
Subject: Re: [PATCH 3/3] numa: Initialize node initiator with respect to .has_cpu
Date: Fri, 29 May 2020 17:09:48 +0200 [thread overview]
Message-ID: <20200529170948.5bdb3316@redhat.com> (raw)
In-Reply-To: <d3039c47e4ce1118bc7c4d4f51da6412c6669339.1590753455.git.mprivozn@redhat.com>
On Fri, 29 May 2020 15:33:48 +0200
Michal Privoznik <mprivozn@redhat.com> wrote:
> The initiator attribute of a NUMA node is documented as the 'NUMA
> node that has best performance to given NUMA node'. If a NUMA
> node has at least one CPU there can hardly be a different node
> with better performace and thus all NUMA nodes which have a CPU
> are initiators to themselves. Reflect this fact when initializing
> the attribute.
It is not true in case of the node is memory-less
> Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
> ---
> hw/core/numa.c | 14 +++++++++-----
> 1 file changed, 9 insertions(+), 5 deletions(-)
>
> diff --git a/hw/core/numa.c b/hw/core/numa.c
> index 338453461c..1c9bc761cc 100644
> --- a/hw/core/numa.c
> +++ b/hw/core/numa.c
> @@ -136,11 +136,15 @@ static void parse_numa_node(MachineState *ms, NumaNodeOptions *node,
> numa_info[nodenr].node_memdev = MEMORY_BACKEND(o);
> }
>
> - /*
> - * If not set the initiator, set it to MAX_NODES. And if
> - * HMAT is enabled and this node has no cpus, QEMU will raise error.
> - */
> - numa_info[nodenr].initiator = MAX_NODES;
> + /* Initialize initiator to either the current NUMA node (if
> + * it has at least one CPU), or to MAX_NODES. If HMAT is
> + * enabled an error will be raised later in
> + * numa_validate_initiator(). */
> + if (numa_info[nodenr].has_cpu)
> + numa_info[nodenr].initiator = nodenr;
> + else
> + numa_info[nodenr].initiator = MAX_NODES;
> +
> if (node->has_initiator) {
> if (!ms->numa_state->hmat_enabled) {
> error_setg(errp, "ACPI Heterogeneous Memory Attribute Table "
next prev parent reply other threads:[~2020-05-29 15:11 UTC|newest]
Thread overview: 17+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-05-29 13:33 [PATCH 0/3] Couple of HMAT fixes Michal Privoznik
2020-05-29 13:33 ` [PATCH 1/3] qapi: Make @associativity, @policy and @line of NumaHmatCacheOptions optional Michal Privoznik
2020-05-29 15:25 ` Igor Mammedov
2020-05-29 13:33 ` [PATCH 2/3] numa: Allow HMAT cache to be defined before HMAT latency/bandwidth Michal Privoznik
2020-05-29 14:59 ` Igor Mammedov
2020-05-29 15:22 ` Michal Privoznik
2020-05-29 13:33 ` [PATCH 3/3] numa: Initialize node initiator with respect to .has_cpu Michal Privoznik
2020-05-29 15:09 ` Igor Mammedov [this message]
2020-05-29 15:24 ` Michal Privoznik
2020-06-01 8:10 ` Michal Privoznik
2020-06-02 8:00 ` Tao Xu
2020-06-03 9:16 ` Michal Privoznik
2020-06-05 1:52 ` Tao Xu
2020-06-11 14:42 ` Michal Privoznik
2020-05-29 20:11 ` [PATCH 0/3] Couple of HMAT fixes no-reply
2020-05-29 20:42 ` no-reply
2020-06-01 11:14 ` Michal Privoznik
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20200529170948.5bdb3316@redhat.com \
--to=imammedo@redhat.com \
--cc=ehabkost@redhat.com \
--cc=jingqi.liu@intel.com \
--cc=mprivozn@redhat.com \
--cc=qemu-devel@nongnu.org \
--cc=tao3.xu@intel.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).