From: "Jan Beulich" <JBeulich@suse.com>
To: Andrew Cooper <andrew.cooper3@citrix.com>
Cc: Boris Ostrovsky <boris.ostrovsky@oracle.com>,
Xen-devel <xen-devel@lists.xen.org>
Subject: Re: [PATCH 03/10] x86/cpuid: Handle leaf 0x1 in guest_cpuid()
Date: Wed, 22 Feb 2017 00:16:23 -0700 [thread overview]
Message-ID: <58AD48D7020000780013C8F6@prv-mh.provo.novell.com> (raw)
In-Reply-To: <140e0be7-c927-13ae-e938-057defcb9db5@citrix.com>
>>> On 21.02.17 at 18:29, <andrew.cooper3@citrix.com> wrote:
> On 21/02/17 17:20, Jan Beulich wrote:
>>
>>>>> The final 8 bits are the initial legacy APIC ID. For HVM guests, this was
>>>>> overridden to vcpu_id * 2. The same logic is now applied to PV guests, so
>>>>> guests don't observe a constant number on all vcpus via their emulated or
>>>>> faulted view.
>>>> They won't be the same everywhere, but every 128th CPU will
>>>> share values. I'm therefore not sure it wouldn't be better to hand
>>>> out zero or all ones here.
>>> There is no case where 128 cpus work sensibly under Xen ATM.
>> For HVM you mean. I'm sure I've seen > 128 vCPU PV guests
>> (namely Dom0-s).
>
> You can physically create PV domains with up to 8192 vcpus. I tried
> this once.
>
> The NMI watchdog (even set to 10s) is unforgiving of some the
> for_each_vcpu() loops during domain destruction.
>
> I can also still create workloads in a 64vcpu HVM guest which will cause
> a 5 second watchdog timeout, which is why XenServers upper supported
> vcpu limit is still 32.
Which does not contradict what I've said: I didn't claim 8k-vCPU
guests would work well, but I'm pretty convinced ones in the
range 128...512 have reasonable chances of working. And we
both know the situation sadly is worse for HVM ones.
Jan
_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel
next prev parent reply other threads:[~2017-02-22 7:16 UTC|newest]
Thread overview: 65+ messages / expand[flat|nested] mbox.gz Atom feed top
2017-02-20 11:00 [PATCH 00/10] x86/cpuid: Remove the legacy infrastructure Andrew Cooper
2017-02-20 11:00 ` [PATCH 01/10] x86/cpuid: Disallow policy updates once the domain is running Andrew Cooper
2017-02-21 16:37 ` Jan Beulich
2017-02-20 11:00 ` [PATCH 02/10] x86/gen-cpuid: Clarify the intended meaning of AVX wrt feature dependencies Andrew Cooper
2017-02-21 16:40 ` Jan Beulich
2017-02-21 16:41 ` Andrew Cooper
2017-02-21 16:47 ` Jan Beulich
2017-02-21 16:53 ` Andrew Cooper
2017-02-21 17:07 ` Jan Beulich
2017-02-21 17:12 ` Andrew Cooper
2017-02-21 17:17 ` Jan Beulich
2017-02-21 17:42 ` Andrew Cooper
2017-02-22 7:13 ` Jan Beulich
2017-02-20 11:00 ` [PATCH 03/10] x86/cpuid: Handle leaf 0x1 in guest_cpuid() Andrew Cooper
2017-02-21 16:59 ` Jan Beulich
2017-02-21 17:13 ` Andrew Cooper
2017-02-21 17:20 ` Jan Beulich
2017-02-21 17:29 ` Andrew Cooper
2017-02-22 7:16 ` Jan Beulich [this message]
2017-02-20 11:00 ` [PATCH 04/10] x86/cpuid: Handle leaf 0x4 " Andrew Cooper
2017-02-21 17:16 ` Jan Beulich
2017-02-21 17:35 ` Andrew Cooper
2017-02-22 7:23 ` Jan Beulich
2017-02-22 7:55 ` Andrew Cooper
2017-03-10 16:27 ` [PATCH v2 " Andrew Cooper
2017-03-13 12:03 ` Jan Beulich
2017-03-13 12:51 ` Andrew Cooper
2017-03-13 13:05 ` Jan Beulich
2017-03-13 13:24 ` Andrew Cooper
2017-03-13 13:36 ` Jan Beulich
2017-02-20 11:00 ` [PATCH 05/10] x86/cpuid: Handle leaf 0x5 " Andrew Cooper
2017-02-21 17:22 ` Jan Beulich
2017-02-20 11:00 ` [PATCH 06/10] x86/cpuid: Handle leaf 0x6 " Andrew Cooper
2017-02-21 17:25 ` Jan Beulich
2017-02-21 17:40 ` Andrew Cooper
2017-02-21 17:44 ` Andrew Cooper
2017-02-22 7:31 ` Jan Beulich
2017-02-22 8:23 ` Andrew Cooper
2017-02-22 9:12 ` Andrew Cooper
2017-02-22 9:26 ` Jan Beulich
2017-02-27 14:30 ` Andrew Cooper
2017-03-10 16:32 ` [PATCH v2 " Andrew Cooper
2017-03-13 12:04 ` Jan Beulich
2017-02-20 11:00 ` [PATCH 07/10] x86/cpuid: Handle leaf 0xa " Andrew Cooper
2017-02-22 9:11 ` Jan Beulich
2017-02-20 11:00 ` [PATCH 08/10] x86/cpuid: Handle leaf 0xb " Andrew Cooper
2017-02-22 9:16 ` Jan Beulich
2017-02-22 10:22 ` Andrew Cooper
2017-02-22 10:37 ` Jan Beulich
2017-02-27 15:05 ` Andrew Cooper
2017-03-10 16:44 ` [PATCH v2 " Andrew Cooper
2017-03-13 12:13 ` Jan Beulich
2017-02-20 11:00 ` [PATCH 09/10] x86/cpuid: Drop legacy CPUID infrastructure Andrew Cooper
2017-02-22 9:19 ` Jan Beulich
2017-02-20 11:00 ` [PATCH 10/10] x86/cpuid: Always enable faulting for the control domain Andrew Cooper
2017-02-22 9:23 ` Jan Beulich
2017-02-22 10:00 ` Andrew Cooper
2017-02-22 10:10 ` Jan Beulich
2017-02-27 15:10 ` Andrew Cooper
2017-02-28 9:31 ` Jan Beulich
2017-03-10 17:10 ` Andrew Cooper
2017-03-13 11:48 ` Jan Beulich
2017-03-14 15:06 ` Wei Liu
2017-03-14 15:13 ` Jan Beulich
2017-03-14 16:05 ` Wei Liu
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=58AD48D7020000780013C8F6@prv-mh.provo.novell.com \
--to=jbeulich@suse.com \
--cc=andrew.cooper3@citrix.com \
--cc=boris.ostrovsky@oracle.com \
--cc=xen-devel@lists.xen.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 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.