xen-devel.lists.xenproject.org archive mirror
 help / color / mirror / Atom feed
From: Andrew Cooper <andrew.cooper3@citrix.com>
To: Jan Beulich <jbeulich@suse.com>
Cc: Xen-devel <xen-devel@lists.xenproject.org>,
	"Wei Liu" <wl@xen.org>, "Roger Pau Monné" <roger.pau@citrix.com>
Subject: Re: [PATCH] libx86: Introduce x86_cpu_policy_calculate_compatible() with MSR_ARCH_CAPS handling
Date: Wed, 5 May 2021 16:18:10 +0100	[thread overview]
Message-ID: <03b94402-6233-b6a9-6621-064b5c6bce26@citrix.com> (raw)
In-Reply-To: <d9e03aa4-bbf4-a1ef-bfab-2803e707f498@suse.com>

On 05/05/2021 16:00, Jan Beulich wrote:
> On 05.05.2021 16:50, Andrew Cooper wrote:
>> On 05/05/2021 15:48, Jan Beulich wrote:
>>> On 05.05.2021 16:29, Andrew Cooper wrote:
>>>> Technically, MCXSR_MASK is also a hard blocker to migration, but we
>>>> don't even have that data in a consumable form, and we just might be
>>>> extremely lucky and discover that it is restricted to non-64-bit CPUs.
>>> "it" being what here? The value's presence / absence in an {F,}XSAVE
>>> image? Or the precise value of it?
>> The precise value of it.
> Not sure whether DAZ is new enough, but relatively sure MM isn't.
>>   Migrating across the boundary where the
>> default changed will cause {F,}RSTOR instructions to #GP.
> That's understood. Is there actually anything standing in the way
> of treating MXCSR_MASK like yet another feature flag group?

Well - we'd need to find somewhere to put it.  It doesn't fit in
architectural CPUID or MSR information.

We could add a 3rd category of information in "a cpu policy", and that
is always an option.

However, this issue hasn't been a problem so far, is only for very
legacy CPUs at this point, and isn't high on my list of worries.


      reply	other threads:[~2021-05-05 15:18 UTC|newest]

Thread overview: 12+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-05-04 21:31 Andrew Cooper
2021-05-05  6:39 ` Jan Beulich
2021-05-05 12:15   ` Andrew Cooper
2021-05-05 12:32     ` Jan Beulich
2021-05-05 10:04 ` Roger Pau Monné
2021-05-05 12:37   ` Andrew Cooper
2021-05-05 13:02     ` Roger Pau Monné
2021-05-05 14:29       ` Andrew Cooper
2021-05-05 14:48         ` Jan Beulich
2021-05-05 14:50           ` Andrew Cooper
2021-05-05 15:00             ` Jan Beulich
2021-05-05 15:18               ` Andrew Cooper [this message]

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:

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=03b94402-6233-b6a9-6621-064b5c6bce26@citrix.com \
    --to=andrew.cooper3@citrix.com \
    --cc=jbeulich@suse.com \
    --cc=roger.pau@citrix.com \
    --cc=wl@xen.org \
    --cc=xen-devel@lists.xenproject.org \
    --subject='Re: [PATCH] libx86: Introduce x86_cpu_policy_calculate_compatible() with MSR_ARCH_CAPS handling' \


* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link

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).