linuxppc-dev.lists.ozlabs.org archive mirror
 help / color / mirror / Atom feed
From: Paul Mackerras <paulus@ozlabs.org>
To: "Aneesh Kumar K.V" <aneesh.kumar@linux.vnet.ibm.com>
Cc: kvm@vger.kernel.org, kvm-ppc@vger.kernel.org, linuxppc-dev@ozlabs.org
Subject: Re: [PATCH 07/13] KVM: PPC: Book3S HV: Adjust host/guest context switch for POWER9
Date: Sat, 19 Nov 2016 15:02:32 +1100	[thread overview]
Message-ID: <20161119040232.GB29462@fergus.ozlabs.ibm.com> (raw)
In-Reply-To: <8737ioony4.fsf@linux.vnet.ibm.com>

On Fri, Nov 18, 2016 at 08:05:47PM +0530, Aneesh Kumar K.V wrote:
> Paul Mackerras <paulus@ozlabs.org> writes:
> 
> > Some special-purpose registers that were present and accessible
> > by guests on POWER8 no longer exist on POWER9, so this adds
> > feature sections to ensure that we don't try to context-switch
> > them when going into or out of a guest on POWER9.  These are
> > all relatively obscure, rarely-used registers, but we had to
> > context-switch them on POWER8 to avoid creating a covert channel.
> > They are: SPMC1, SPMC2, MMCRS, CSIGR, TACR, TCSCR, and ACOP.
> 
> We don't need to context-switch them even when running a power8 compat
> guest ?

They physically don't exist on the P9 chip, so how could we
context-switch them?  They certainly can't be used as a covert
channel.

Accesses to them will be a no-op for the guest in privileged
(supervisor) mode (i.e., mfspr won't modify the destination
register), which could be confusing for the guest if it was expecting
to use them.  SPMC1/2 and MMCRS are part of the "supervisor" PMU,
which we have never used.  I think CSIGR, TACR and TCSCR are part of a
facility that was never completely implemented or usable on P8, so
nothing uses them.  ACOP is used in arch/powerpc/mm/icswx.c in
conjunction with accelerators.  There might be a problem there, but in
any case, with no physical ACOP register present there's no way to
save/restore it.

Paul.

  reply	other threads:[~2016-11-19  4:23 UTC|newest]

Thread overview: 32+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2016-11-18  7:28 [PATCH 00/13] KVM: PPC: Support POWER9 guests Paul Mackerras
2016-11-18  7:28 ` [PATCH 01/13] powerpc/64: Add some more SPRs and SPR bits for POWER9 Paul Mackerras
2016-11-18  7:28 ` [PATCH 02/13] powerpc/64: Provide functions for accessing POWER9 partition table Paul Mackerras
2016-11-18 14:27   ` Aneesh Kumar K.V
2016-11-19  4:19     ` Paul Mackerras
2016-11-19  6:35       ` Aneesh Kumar K.V
2016-11-21  2:14         ` Paul Mackerras
2016-11-19  0:45   ` Balbir Singh
2016-11-19  4:23     ` Paul Mackerras
2016-11-18  7:28 ` [PATCH 03/13] powerpc/powernv: Define real-mode versions of OPAL XICS accessors Paul Mackerras
2016-11-18  7:28 ` [PATCH 04/13] KVM: PPC: Book3S HV: Don't lose hardware R/C bit updates in H_PROTECT Paul Mackerras
2016-11-18  7:28 ` [PATCH 05/13] KVM: PPC: Book3S HV: Adapt to new HPTE format on POWER9 Paul Mackerras
2016-11-19  0:38   ` Balbir Singh
2016-11-21  2:02     ` Paul Mackerras
2016-11-18  7:28 ` [PATCH 06/13] KVM: PPC: Book3S HV: Set partition table rather than SDR1 " Paul Mackerras
2016-11-19  1:01   ` Balbir Singh
2016-11-18  7:28 ` [PATCH 07/13] KVM: PPC: Book3S HV: Adjust host/guest context switch for POWER9 Paul Mackerras
2016-11-18 14:35   ` Aneesh Kumar K.V
2016-11-19  4:02     ` Paul Mackerras [this message]
2016-11-18  7:28 ` [PATCH 08/13] KVM: PPC: Book3S HV: Add new POWER9 guest-accessible SPRs Paul Mackerras
2016-11-18  7:28 ` [PATCH 09/13] KVM: PPC: Book3S HV: Adapt TLB invalidations to work on POWER9 Paul Mackerras
2016-11-18 14:41   ` Aneesh Kumar K.V
2016-11-18 21:57     ` Benjamin Herrenschmidt
2016-11-19  4:14       ` Paul Mackerras
2016-11-19  4:41         ` Benjamin Herrenschmidt
2016-11-19  4:13     ` Paul Mackerras
2016-11-18  7:28 ` [PATCH 10/13] KVM: PPC: Book3S HV: Use msgsnd for IPIs to other cores " Paul Mackerras
2016-11-18 14:47   ` Aneesh Kumar K.V
2016-11-19  3:53     ` Paul Mackerras
2016-11-18  7:28 ` [PATCH 11/13] KVM: PPC: Book3S HV: Use OPAL XICS emulation " Paul Mackerras
2016-11-18  7:28 ` [PATCH 12/13] KVM: PPC: Book3S HV: Use stop instruction rather than nap " Paul Mackerras
2016-11-18  7:28 ` [PATCH 13/13] KVM: PPC: Book3S HV: Treat POWER9 CPU threads as independent subcores Paul Mackerras

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=20161119040232.GB29462@fergus.ozlabs.ibm.com \
    --to=paulus@ozlabs.org \
    --cc=aneesh.kumar@linux.vnet.ibm.com \
    --cc=kvm-ppc@vger.kernel.org \
    --cc=kvm@vger.kernel.org \
    --cc=linuxppc-dev@ozlabs.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).