All of lore.kernel.org
 help / color / mirror / Atom feed
From: Eduardo Habkost <ehabkost@redhat.com>
To: Babu Moger <babu.moger@amd.com>
Cc: geoff@hostfission.com, kvm@vger.kernel.org, mst@redhat.com,
	kash@tripleback.net, mtosatti@redhat.com, qemu-devel@nongnu.org,
	pbonzini@redhat.com, Jiri Denemark <jdenemar@redhat.com>,
	rth@twiddle.net
Subject: Re: [PATCH v13 3/5] i386: Enable TOPOEXT feature on AMD EPYC CPU
Date: Mon, 11 Jun 2018 18:10:26 -0300	[thread overview]
Message-ID: <20180611211026.GA7451@localhost.localdomain> (raw)
In-Reply-To: <20180611205030.GW7451@localhost.localdomain>

On Mon, Jun 11, 2018 at 05:50:30PM -0300, Eduardo Habkost wrote:
[...]
> > +        /* TOPOEXT feature requires 0x8000001E */
> > +        if (env->features[FEAT_8000_0001_ECX] & CPUID_EXT3_TOPOEXT) {
> > +            x86_cpu_adjust_level(cpu, &env->cpuid_min_xlevel, 0x8000001E);
> > +        }
> 
> I suggest moving this hunk to a separate patch.  I'm not 100%
> sure yet if this will require compat_props code to disable
> auto-xlevel-increase on older machine-types.

The problem here is that:
  $QEMU -machine pc-i440fx-1.3 -cpu Opteron_G4,+topoext
currently results in xlevel=0x8000001A, since QEMU 1.3.

(The same applies to all machine-types between 1.3 and 2.12)

I was hoping that we could declare topoext as non-migration-safe,
but I believe libvirt will already include "topoext" when using
"host-model" if the host CPU supports TOPOEXT.  Jiri, can you
confirm that?

We can address that with a "x-topoext-auto-xlevel" property, set
to true on all CPU models by default, and disabled by
PC_COMPAT_2_12.

The code would become:

    if (cpu->topoext_auto_xlevel && env->features[FEAT_8000_0001_ECX] & CPUID_EXT3_TOPOEXT) {
        x86_cpu_adjust_level(cpu, &env->cpuid_min_xlevel, 0x8000001E);
    }

Or, we could simply declare that "-cpu Opteron_G4,+topoext" will
never increase xlevel automatically (on any machine-type), and
change the code above to:

    if (cpu->auto_topoext && env->features[FEAT_8000_0001_ECX] & CPUID_EXT3_TOPOEXT) {
        x86_cpu_adjust_level(cpu, &env->cpuid_min_xlevel, 0x8000001E);
    }

-- 
Eduardo

WARNING: multiple messages have this Message-ID (diff)
From: Eduardo Habkost <ehabkost@redhat.com>
To: Babu Moger <babu.moger@amd.com>
Cc: mst@redhat.com, marcel.apfelbaum@gmail.com, pbonzini@redhat.com,
	rth@twiddle.net, mtosatti@redhat.com, qemu-devel@nongnu.org,
	kvm@vger.kernel.org, kash@tripleback.net, geoff@hostfission.com,
	Jiri Denemark <jdenemar@redhat.com>
Subject: Re: [Qemu-devel] [PATCH v13 3/5] i386: Enable TOPOEXT feature on AMD EPYC CPU
Date: Mon, 11 Jun 2018 18:10:26 -0300	[thread overview]
Message-ID: <20180611211026.GA7451@localhost.localdomain> (raw)
In-Reply-To: <20180611205030.GW7451@localhost.localdomain>

On Mon, Jun 11, 2018 at 05:50:30PM -0300, Eduardo Habkost wrote:
[...]
> > +        /* TOPOEXT feature requires 0x8000001E */
> > +        if (env->features[FEAT_8000_0001_ECX] & CPUID_EXT3_TOPOEXT) {
> > +            x86_cpu_adjust_level(cpu, &env->cpuid_min_xlevel, 0x8000001E);
> > +        }
> 
> I suggest moving this hunk to a separate patch.  I'm not 100%
> sure yet if this will require compat_props code to disable
> auto-xlevel-increase on older machine-types.

The problem here is that:
  $QEMU -machine pc-i440fx-1.3 -cpu Opteron_G4,+topoext
currently results in xlevel=0x8000001A, since QEMU 1.3.

(The same applies to all machine-types between 1.3 and 2.12)

I was hoping that we could declare topoext as non-migration-safe,
but I believe libvirt will already include "topoext" when using
"host-model" if the host CPU supports TOPOEXT.  Jiri, can you
confirm that?

We can address that with a "x-topoext-auto-xlevel" property, set
to true on all CPU models by default, and disabled by
PC_COMPAT_2_12.

The code would become:

    if (cpu->topoext_auto_xlevel && env->features[FEAT_8000_0001_ECX] & CPUID_EXT3_TOPOEXT) {
        x86_cpu_adjust_level(cpu, &env->cpuid_min_xlevel, 0x8000001E);
    }

Or, we could simply declare that "-cpu Opteron_G4,+topoext" will
never increase xlevel automatically (on any machine-type), and
change the code above to:

    if (cpu->auto_topoext && env->features[FEAT_8000_0001_ECX] & CPUID_EXT3_TOPOEXT) {
        x86_cpu_adjust_level(cpu, &env->cpuid_min_xlevel, 0x8000001E);
    }

-- 
Eduardo

  parent reply	other threads:[~2018-06-11 21:10 UTC|newest]

Thread overview: 72+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2018-06-08 22:56 [PATCH v13 0/5] i386: Enable TOPOEXT to support hyperthreading on AMD CPU Babu Moger
2018-06-08 22:56 ` [Qemu-devel] " Babu Moger
2018-06-08 22:56 ` [PATCH v13 1/5] i386: Add support for CPUID_8000_001E for AMD Babu Moger
2018-06-08 22:56   ` [Qemu-devel] " Babu Moger
2018-06-11 20:54   ` Eduardo Habkost
2018-06-11 20:54     ` [Qemu-devel] " Eduardo Habkost
2018-06-11 21:02     ` Moger, Babu
2018-06-11 21:02       ` [Qemu-devel] " Moger, Babu
2020-06-02 17:52   ` Eduardo Habkost
2020-06-02 17:52     ` Eduardo Habkost
2020-06-03 23:57     ` Babu Moger
2020-06-03 23:57       ` Babu Moger
2020-06-04 14:06     ` Babu Moger
2020-06-04 14:06       ` Babu Moger
2020-06-04 20:54       ` Eduardo Habkost
2020-06-04 20:54         ` Eduardo Habkost
2020-06-05 16:21         ` Babu Moger
2020-06-05 16:21           ` Babu Moger
2018-06-08 22:56 ` [PATCH v13 2/5] i386: Introduce auto_topoext bit to manage topoext Babu Moger
2018-06-08 22:56   ` [Qemu-devel] " Babu Moger
2018-06-11 20:25   ` Moger, Babu
2018-06-11 20:25     ` [Qemu-devel] " Moger, Babu
2018-06-11 20:46     ` Eduardo Habkost
2018-06-11 20:46       ` [Qemu-devel] " Eduardo Habkost
2018-06-11 20:59       ` Moger, Babu
2018-06-11 20:59         ` [Qemu-devel] " Moger, Babu
2018-06-11 21:05       ` Eduardo Habkost
2018-06-11 21:05         ` [Qemu-devel] " Eduardo Habkost
2018-06-11 21:20         ` Moger, Babu
2018-06-11 21:20           ` [Qemu-devel] " Moger, Babu
2018-06-08 22:56 ` [PATCH v13 3/5] i386: Enable TOPOEXT feature on AMD EPYC CPU Babu Moger
2018-06-08 22:56   ` [Qemu-devel] " Babu Moger
2018-06-11 20:50   ` Eduardo Habkost
2018-06-11 20:50     ` [Qemu-devel] " Eduardo Habkost
2018-06-11 21:01     ` Moger, Babu
2018-06-11 21:01       ` [Qemu-devel] " Moger, Babu
2018-06-11 21:10     ` Eduardo Habkost [this message]
2018-06-11 21:10       ` Eduardo Habkost
2018-06-12 16:29       ` Moger, Babu
2018-06-12 16:29         ` [Qemu-devel] " Moger, Babu
2018-06-12 17:40         ` Eduardo Habkost
2018-06-12 17:40           ` [Qemu-devel] " Eduardo Habkost
2018-06-12 18:38           ` Moger, Babu
2018-06-12 18:38             ` [Qemu-devel] " Moger, Babu
2018-06-12 19:05             ` Eduardo Habkost
2018-06-12 19:05               ` [Qemu-devel] " Eduardo Habkost
2018-06-12 19:46               ` Babu Moger
2018-06-12 19:46                 ` [Qemu-devel] " Babu Moger
2018-06-13 16:52                 ` Moger, Babu
2018-06-13 16:52                   ` [Qemu-devel] " Moger, Babu
2018-06-13 17:18                   ` Eduardo Habkost
2018-06-13 17:18                     ` [Qemu-devel] " Eduardo Habkost
2018-06-13 18:10                     ` Moger, Babu
2018-06-13 18:10                       ` [Qemu-devel] " Moger, Babu
2018-06-13 18:12                       ` Moger, Babu
2018-06-13 18:12                         ` [Qemu-devel] " Moger, Babu
2018-06-13 18:17                       ` Eduardo Habkost
2018-06-13 18:17                         ` [Qemu-devel] " Eduardo Habkost
2018-06-13 18:21                         ` Moger, Babu
2018-06-13 18:21                           ` [Qemu-devel] " Moger, Babu
2018-06-13 18:49                           ` Eduardo Habkost
2018-06-13 18:49                             ` [Qemu-devel] " Eduardo Habkost
2018-06-14  0:58                             ` Moger, Babu
2018-06-14  0:58                               ` [Qemu-devel] " Moger, Babu
2018-06-08 22:56 ` [PATCH v13 4/5] i386: Verify and enable topoext feature if supported Babu Moger
2018-06-08 22:56   ` [Qemu-devel] " Babu Moger
2018-06-11 20:51   ` Eduardo Habkost
2018-06-11 20:51     ` [Qemu-devel] " Eduardo Habkost
2018-06-11 21:01     ` Moger, Babu
2018-06-11 21:01       ` [Qemu-devel] " Moger, Babu
2018-06-08 22:56 ` [PATCH v13 5/5] i386: Remove generic SMT thread check Babu Moger
2018-06-08 22:56   ` [Qemu-devel] " Babu Moger

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=20180611211026.GA7451@localhost.localdomain \
    --to=ehabkost@redhat.com \
    --cc=babu.moger@amd.com \
    --cc=geoff@hostfission.com \
    --cc=jdenemar@redhat.com \
    --cc=kash@tripleback.net \
    --cc=kvm@vger.kernel.org \
    --cc=mst@redhat.com \
    --cc=mtosatti@redhat.com \
    --cc=pbonzini@redhat.com \
    --cc=qemu-devel@nongnu.org \
    --cc=rth@twiddle.net \
    /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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.