All of lore.kernel.org
 help / color / mirror / Atom feed
From: Andi Kleen <ak@linux.intel.com>
To: speck@linutronix.de
Subject: [MODERATED] Re: [PATCH v2 3/8] MDSv2 5
Date: Thu, 13 Dec 2018 12:48:28 -0800	[thread overview]
Message-ID: <20181213204828.GM25620@tassilo.jf.intel.com> (raw)
In-Reply-To: <CAHk-=wjBe-FxaEe7HVaa-6kXLL1ApPd7yn9BtEwQX2bz6243zQ@mail.gmail.com>

> This is actually fairly ok, because if the verw *really* doesn't do
> anything, it also isn't very expensive (ie on the order of just 40
> cycles instead of 400 cycles).  So I'd actually be inclined to say
> "let's do that for the *initial* patches and back-ports" just to give
> any vmware users the protection.

I think that's fine

40 cycles will be hard to measure even in the most extreme 
syscall micro benchmarks, and is unlikely to impact
anything real.

> Longer term (ie after a few months), the proper sequence ends up
> taking X86_FEATURE_MB_CLEAR into account, so this is basically just an
> artificial "it takes a while for that feature to show up in a vmware
> cluster even after the fixes have been applied" issue.

Seems odd to somehow tie our release cycle to VMware.

This is something the distribution vendors would need to do
anyways, as VMWare is mostly used in Enterprise environments
which are likely mostly on SLES, RHEL and somesuch, and
are unlikely to run raw kernel.org kernels.

*Iff* we really care about this oddity, it would be best to ignore
MB_CLEAR forever.

AFAIK in KVM software sequences should work, as long as the administrator
does not set a totally bogus CPUID (like no AVX on Sandy Bridge,
or AVX with a CPU model that doesn't support AVX). We can
add some documentation to make this clear.

> 
> Note that vmware *really* really doesn't want that software sequence
> in any case. Not only does it not work reliably in virtualization
> anyway, it's likely that vmware cpu's *have* gotten the microcode
> updates, they just don't show the cpuid bits. So the software version
> is basically slower and less effective. Just don't do it.

I don't think it has been worked out yet, but the reason
I added the software sequences is that some CPUs may not
be able to do the VERW in microcode updates.

In this case the only choice is to use software sequences,
and VMware will have to deal with it somehow.

But I can remove them again until that is settled.

-Andi

  reply	other threads:[~2018-12-13 20:50 UTC|newest]

Thread overview: 59+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2018-12-10 17:53 [MODERATED] [PATCH v2 0/8] MDSv2 8 Andi Kleen
2018-12-10 17:53 ` [MODERATED] [PATCH v2 1/8] MDSv2 4 Andi Kleen
2018-12-11 14:14   ` [MODERATED] " Paolo Bonzini
2018-12-12 21:22   ` Konrad Rzeszutek Wilk
2018-12-12 21:28     ` Andi Kleen
2018-12-12 21:25   ` Konrad Rzeszutek Wilk
2018-12-10 17:53 ` [MODERATED] [PATCH v2 2/8] MDSv2 1 Andi Kleen
2018-12-10 22:49   ` [MODERATED] " Jiri Kosina
2018-12-11  0:03     ` Andi Kleen
2018-12-11  0:13     ` Kanth Ghatraju
2018-12-11  2:00       ` Andi Kleen
2018-12-11  5:36       ` Jiri Kosina
2018-12-11 10:03       ` Borislav Petkov
2018-12-12 21:31         ` Konrad Rzeszutek Wilk
2018-12-12 21:43           ` Andi Kleen
2018-12-12 22:17           ` Borislav Petkov
2018-12-12 22:40             ` Konrad Rzeszutek Wilk
2018-12-12 22:45               ` Borislav Petkov
2018-12-13 15:15                 ` Andrew Cooper
2018-12-13 16:52                   ` Borislav Petkov
2018-12-10 17:53 ` [MODERATED] [PATCH v2 3/8] MDSv2 5 Andi Kleen
2018-12-10 23:00   ` [MODERATED] " Linus Torvalds
2018-12-11  0:03     ` Andi Kleen
2018-12-11  0:43       ` Linus Torvalds
2018-12-11  1:33         ` Linus Torvalds
2018-12-11  2:12           ` Andi Kleen
2018-12-11  2:20           ` Linus Torvalds
2018-12-11  3:25             ` Andi Kleen
2018-12-11 17:55               ` Linus Torvalds
2018-12-11 18:10                 ` Borislav Petkov
2018-12-11 18:21                 ` Linus Torvalds
2018-12-11 18:26                   ` Borislav Petkov
2018-12-11 19:47                   ` Andi Kleen
2018-12-11 21:22                   ` Thomas Gleixner
2018-12-12 14:02               ` [MODERATED] " Paolo Bonzini
2018-12-12 17:58                 ` Andi Kleen
2018-12-12 18:47                   ` Linus Torvalds
2018-12-13 19:44                     ` Linus Torvalds
2018-12-13 20:48                       ` Andi Kleen [this message]
2018-12-13 20:56                         ` Linus Torvalds
2018-12-15  0:30                         ` Andi Kleen
2018-12-11  2:10         ` Andi Kleen
2018-12-11  0:09     ` Andrew Cooper
2018-12-10 17:53 ` [MODERATED] [PATCH v2 4/8] MDSv2 0 Andi Kleen
2018-12-12 21:45   ` [MODERATED] " Konrad Rzeszutek Wilk
2018-12-12 22:09     ` Andi Kleen
2018-12-12 22:36       ` Konrad Rzeszutek Wilk
2018-12-10 17:53 ` [MODERATED] [PATCH v2 5/8] MDSv2 7 Andi Kleen
2018-12-11  0:33   ` [MODERATED] " Andrew Cooper
2018-12-12 18:05     ` Andrew Cooper
2018-12-12 21:41   ` Konrad Rzeszutek Wilk
2018-12-12 22:12     ` Andi Kleen
2018-12-10 17:53 ` [MODERATED] [PATCH v2 6/8] MDSv2 3 Andi Kleen
2018-12-11  0:37   ` [MODERATED] " Andrew Cooper
2018-12-11  0:46     ` Luck, Tony
2018-12-11  1:02       ` Andrew Cooper
2018-12-11  1:53       ` Andi Kleen
2018-12-10 17:53 ` [MODERATED] [PATCH v2 7/8] MDSv2 6 Andi Kleen
2018-12-10 17:53 ` [MODERATED] [PATCH v2 8/8] MDSv2 2 Andi Kleen

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=20181213204828.GM25620@tassilo.jf.intel.com \
    --to=ak@linux.intel.com \
    --cc=speck@linutronix.de \
    /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.