From: David Gibson <david@gibson.dropbear.id.au>
To: Andrew Jones <drjones@redhat.com>
Cc: peter.maydell@linaro.org, ehabkost@redhat.com,
qemu-devel@nongnu.org, agraf@suse.de, qemu-arm@nongnu.org,
qemu-ppc@nongnu.org, pbonzini@redhat.com, imammedo@redhat.com,
dgibson@redhat.com
Subject: Re: [Qemu-arm] [PATCH RFC 02/16] vl: smp: add checks for maxcpus based topologies
Date: Tue, 14 Jun 2016 11:28:52 +1000 [thread overview]
Message-ID: <20160614012852.GF4882@voom.fritz.box> (raw)
In-Reply-To: <1465580427-13596-3-git-send-email-drjones@redhat.com>
[-- Attachment #1: Type: text/plain, Size: 3002 bytes --]
On Fri, Jun 10, 2016 at 07:40:13PM +0200, Andrew Jones wrote:
> smp_parse computes missing smp options. Unfortunately cores and
> threads are computed by dividing smp_cpus, instead of max_cpus.
> This is incorrect because the topology doesn't leave room for
> hotplug. More unfortunately, we can't change it easily, as doing
> so would impact existing command lines. This patch adds a warning
> when the topology doesn't add up, and then checks that the topology
> at least computes when sockets are recalculated. If not, then it
> does fail.
>
> Adding the new failure is justified by the fact that we don't
> store the number of input sockets, and thus all consumers of
> cpu topology information recalculate it. If they choose to
> (correctly) calculate it based on maxcpus, then we need to
> guard them against building topologies which provide more cpu
> slots than are the maximum allowed cpus.
>
> Signed-off-by: Andrew Jones <drjones@redhat.com>
Hmm.. this makes sense to me. Except that I've never been clear if
sockets= was supposed to match initial cpus or maxcpus.
> ---
> vl.c | 19 +++++++++++++++++++
> 1 file changed, 19 insertions(+)
>
> diff --git a/vl.c b/vl.c
> index 7b96e787922f9..8d482cb1bf020 100644
> --- a/vl.c
> +++ b/vl.c
> @@ -1227,6 +1227,7 @@ static void smp_parse(QemuOpts *opts)
> unsigned sockets = qemu_opt_get_number(opts, "sockets", 0);
> unsigned cores = qemu_opt_get_number(opts, "cores", 0);
> unsigned threads = qemu_opt_get_number(opts, "threads", 0);
> + bool sockets_input = sockets > 0;
>
> /* compute missing values, prefer sockets over cores over threads */
> if (cpus == 0 || sockets == 0) {
> @@ -1269,6 +1270,24 @@ static void smp_parse(QemuOpts *opts)
> exit(1);
> }
>
> + if (sockets_input && sockets * cores * threads != max_cpus) {
> + unsigned sockets_rounded = DIV_ROUND_UP(max_cpus, cores * threads);
> +
> + error_report("warning: cpu topology: "
> + "sockets (%u) * cores (%u) * threads (%u) != "
> + "maxcpus (%u). Trying sockets=%u.",
> + sockets, cores, threads, max_cpus, sockets_rounded);
> + sockets = sockets_rounded;
> +
> + if (sockets * cores * threads > max_cpus) {
> + error_report("cpu topology: "
> + "sockets (%u) * cores (%u) * threads (%u) > "
> + "maxcpus (%u)",
> + sockets, cores, threads, max_cpus);
> + exit(1);
> + }
> + }
> +
> smp_cpus = cpus;
> smp_cores = cores;
> smp_threads = threads;
--
David Gibson | I'll have my music baroque, and my code
david AT gibson.dropbear.id.au | minimalist, thank you. NOT _the_ _other_
| _way_ _around_!
http://www.ozlabs.org/~dgibson
[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 819 bytes --]
next prev parent reply other threads:[~2016-06-14 1:36 UTC|newest]
Thread overview: 77+ messages / expand[flat|nested] mbox.gz Atom feed top
2016-06-10 17:40 [Qemu-arm] [PATCH RFC 00/16] Rework SMP parameters Andrew Jones
2016-06-10 17:40 ` [Qemu-arm] [PATCH RFC 01/16] vl: smp_parse: cleanups Andrew Jones
2016-06-14 1:15 ` David Gibson
2016-06-10 17:40 ` [Qemu-arm] [PATCH RFC 02/16] vl: smp: add checks for maxcpus based topologies Andrew Jones
2016-06-14 1:28 ` David Gibson [this message]
2016-06-14 6:43 ` [Qemu-arm] [Qemu-devel] " Andrew Jones
2016-06-10 17:40 ` [Qemu-arm] [PATCH RFC 03/16] hw/smbios/smbios: fix number of sockets calculation Andrew Jones
2016-07-11 14:23 ` Igor Mammedov
2016-06-10 17:40 ` [Qemu-arm] [PATCH RFC 04/16] hw/core/machine: Introduce pre_init Andrew Jones
2016-06-14 1:30 ` David Gibson
2016-06-14 5:58 ` [Qemu-arm] [Qemu-devel] " Andrew Jones
2016-07-14 20:10 ` Eduardo Habkost
2016-07-15 6:26 ` Andrew Jones
2016-06-10 17:40 ` [Qemu-arm] [PATCH RFC 05/16] hw/core/machine: add smp properites Andrew Jones
2016-06-14 2:00 ` David Gibson
2016-06-14 6:08 ` [Qemu-arm] [Qemu-devel] " Andrew Jones
2016-06-15 0:37 ` David Gibson
2016-06-15 7:11 ` Andrew Jones
2016-07-14 20:18 ` Eduardo Habkost
2016-07-15 6:29 ` Andrew Jones
2016-06-10 17:40 ` [Qemu-arm] [PATCH RFC 06/16] vl: move smp parsing to machine pre_init Andrew Jones
2016-06-13 17:04 ` Paolo Bonzini
2016-06-13 20:35 ` [Qemu-arm] [Qemu-devel] " Andrew Jones
2016-06-14 8:17 ` Paolo Bonzini
2016-06-14 11:39 ` [Qemu-arm] " Andrew Jones
2016-06-14 11:53 ` [Qemu-arm] " Paolo Bonzini
2016-06-14 14:03 ` [Qemu-devel] " Andrew Jones
2016-06-14 14:05 ` [Qemu-arm] " Paolo Bonzini
2016-06-15 0:51 ` David Gibson
2016-06-15 7:19 ` Andrew Jones
2016-06-15 0:43 ` [Qemu-arm] " David Gibson
2016-06-10 17:40 ` [Qemu-devel] [PATCH RFC 07/16] qom/cpu: make nr-cores, nr-threads real properties Andrew Jones
2016-06-11 6:54 ` [Qemu-arm] " Thomas Huth
2016-06-12 13:48 ` Andrew Jones
2016-06-14 2:12 ` David Gibson
2016-06-14 6:19 ` Andrew Jones
2016-06-15 0:56 ` David Gibson
2016-07-14 20:07 ` Eduardo Habkost
2016-07-15 6:35 ` Andrew Jones
2016-07-15 9:11 ` Igor Mammedov
2016-07-15 16:10 ` [Qemu-devel] QOM: best way for parents to pass information to children? (was Re: [PATCH RFC 07/16] qom/cpu: make nr-cores, nr-threads real properties) Eduardo Habkost
2016-07-15 16:10 ` [Qemu-arm] QOM: best way for parents to pass information to children? (was Re: [Qemu-devel] " Eduardo Habkost
2016-07-15 16:30 ` [Qemu-devel] QOM: best way for parents to pass information to children? (was " Andreas Färber
2016-07-15 16:30 ` [Qemu-arm] QOM: best way for parents to pass information to children? (was Re: [Qemu-devel] " Andreas Färber
2016-07-15 17:43 ` [Qemu-devel] QOM: best way for parents to pass information to children? (was " Eduardo Habkost
2016-07-15 17:43 ` [Qemu-arm] QOM: best way for parents to pass information to children? (was Re: [Qemu-devel] " Eduardo Habkost
2016-07-15 18:38 ` [Qemu-arm] [Qemu-devel] QOM: best way for parents to pass information to children? (was " Igor Mammedov
2016-07-15 21:33 ` Eduardo Habkost
2016-07-16 15:30 ` Andrew Jones
2016-07-19 11:54 ` Eduardo Habkost
2016-07-18 7:23 ` Igor Mammedov
2016-07-19 11:59 ` Eduardo Habkost
2016-07-19 12:21 ` Paolo Bonzini
2016-07-19 13:29 ` Igor Mammedov
2016-07-19 13:39 ` Paolo Bonzini
2016-06-10 17:40 ` [Qemu-devel] [PATCH RFC 08/16] hw/core/machine: set cpu global nr_cores, nr_threads in pre_init Andrew Jones
2016-06-10 17:40 ` [Qemu-arm] [PATCH RFC 09/16] hw/i386/pc: don't use smp_cores, smp_threads Andrew Jones
2016-07-14 20:33 ` [Qemu-devel] " Eduardo Habkost
2016-06-10 17:40 ` [Qemu-arm] [PATCH RFC 10/16] hw/ppc/spapr: " Andrew Jones
2016-06-14 3:03 ` David Gibson
2016-06-14 6:23 ` [Qemu-arm] [Qemu-devel] " Andrew Jones
2016-06-15 0:59 ` David Gibson
2016-06-15 7:34 ` Andrew Jones
2016-06-10 17:40 ` [Qemu-arm] [PATCH RFC 11/16] target-ppc: don't use smp_threads Andrew Jones
2016-06-10 17:40 ` [Qemu-devel] [PATCH RFC 12/16] hw/arm/virt: rename *.smp_cpus to *.cpus Andrew Jones
2016-06-10 17:40 ` [Qemu-arm] [PATCH RFC 13/16] hw/arm/virt: don't use smp_cpus, max_cpus Andrew Jones
2016-06-10 17:40 ` [Qemu-arm] [PATCH RFC 14/16] hw/arm/virt: stash cpu topo info in VirtGuestInfo Andrew Jones
2016-07-14 20:43 ` [Qemu-arm] [Qemu-devel] " Eduardo Habkost
2016-07-15 6:40 ` Andrew Jones
2016-06-10 17:40 ` [Qemu-devel] [PATCH RFC 15/16] smbios: don't use smp_cores, smp_threads Andrew Jones
2016-07-14 20:51 ` [Qemu-arm] " Eduardo Habkost
2016-07-15 6:45 ` Andrew Jones
2016-06-10 17:40 ` [Qemu-arm] [PATCH RFC 16/16] sysemu/cpus: bye, bye " Andrew Jones
2016-06-11 6:42 ` [Qemu-arm] [Qemu-devel] [PATCH RFC 00/16] Rework SMP parameters Thomas Huth
2016-06-12 13:58 ` Andrew Jones
2016-06-12 14:03 ` Andrew Jones
2016-07-14 9:16 ` Andrew Jones
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=20160614012852.GF4882@voom.fritz.box \
--to=david@gibson.dropbear.id.au \
--cc=agraf@suse.de \
--cc=dgibson@redhat.com \
--cc=drjones@redhat.com \
--cc=ehabkost@redhat.com \
--cc=imammedo@redhat.com \
--cc=pbonzini@redhat.com \
--cc=peter.maydell@linaro.org \
--cc=qemu-arm@nongnu.org \
--cc=qemu-devel@nongnu.org \
--cc=qemu-ppc@nongnu.org \
/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).