LKML Archive on lore.kernel.org
help / color / mirror / Atom feed
From: Ravikiran Thirumalai <kiran@scalemp.com>
To: Yinghai Lu <yhlu.kernel@gmail.com>
Cc: Andi Kleen <andi@firstfloor.org>, Ingo Molnar <mingo@elte.hu>,
	Andrew Morton <akpm@linux-foundation.org>,
	Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
	shai@scalemp.com
Subject: Re: [PATCH] x86_64: make amd quad core 8 socket system not be clustered_box v2
Date: Mon, 25 Feb 2008 19:39:45 -0800	[thread overview]
Message-ID: <20080226033945.GJ3143@localdomain> (raw)
In-Reply-To: <86802c440802251405w7025a8b2na5fde4dcbf56067e@mail.gmail.com>

On Mon, Feb 25, 2008 at 02:05:45PM -0800, Yinghai Lu wrote:
>On Mon, Feb 25, 2008 at 11:08 AM, Ravikiran Thirumalai
><kiran@scalemp.com> wrote:
>> ...
>>  Andi, Yes.  AMD based vSMPowered systems uses clustered APICs, and this
>>  check base on cpu vendor id is not good :(.
>
>please check if you happy with
>
>http://lkml.org/lkml/2008/2/24/273
>

I don't quite understand the purpose of the patch to begin with.  Looking at
the current x86 git tree, apic_is_clustered_box is used only to determine if
tsc is synchronized on the platform.  The AMD docs  imply that TSC's are not
guaranteed to be synced across cores between nodes -- Opteron BKDG for
family 10h, Section 2.9.4:

<quote>
Note: Timers associated with different CPU cores in the same processor
increment at the same rate. Timers associated with different CPU cores
in different processors increment at slightly different rates if (1) they
are located on different nodes and (2) CLKIN for these nodes is derived from
different, non-synchronized oscillator sources.
</quote>

But that is not what the x86 tree does (with your patches) -- it looks for the
X86_FEATURE_CONSTANT_TSC at unsynchronized_tsc() and assumes a synchronized
clock.  Huh!??  Am i missing something here?  X86_FEATURE_CONSTANT_TSC
is set from CPUID Fn8000_0007 -- TscInvariant bit, which implies TSC is
not affected by change in PM states.  This does not talk about whether CLKIN
for different packages are from synchronized/non synchronized oscillator
sources in the above quote.

Thanks,
Kiran

  reply	other threads:[~2008-02-26  3:39 UTC|newest]

Thread overview: 48+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2008-02-21 10:58 [PATCH] x86_64: make amd quad core 8 socket system not be clustered_box Yinghai Lu
2008-02-22 12:25 ` Andi Kleen
2008-02-22 19:02   ` Yinghai Lu
2008-02-22 19:00     ` Andi Kleen
2008-02-22 19:04       ` Yinghai Lu
2008-02-22 19:07         ` Andi Kleen
2008-02-22 19:07           ` Yinghai Lu
2008-02-22 19:10             ` Andi Kleen
2008-02-23  8:55               ` Yinghai Lu
2008-02-24  5:48       ` [PATCH] x86_64: make amd quad core 8 socket system not be clustered_box v2 Yinghai Lu
2008-02-24  7:50         ` Ingo Molnar
2008-02-24 12:29         ` Andi Kleen
2008-02-24 23:00           ` Yinghai Lu
2008-02-25  1:52             ` Yinghai Lu
2008-02-25  2:32             ` Yinghai Lu
2008-02-25  5:36               ` [PATCH] x86_64: for apic_is_clustered_box for vsmp v2 Yinghai Lu
2008-02-25  6:43                 ` [PATCH] x86: vSMP selection in config Yinghai Lu
2008-02-26 19:40                   ` Kconfig configuration restore bug [Was: x86: vSMP selection in config] Sam Ravnborg
2008-02-27  2:59                     ` Roman Zippel
2008-02-29  4:09                     ` [PATCH 1/3] fix recursive dependencies Roman Zippel
2008-02-29  5:05                       ` Yinghai Lu
2008-02-29 13:22                         ` Roman Zippel
2008-02-29 17:40                           ` Sam Ravnborg
2008-02-29 20:05                             ` Ingo Molnar
2008-02-29 20:04                       ` Ingo Molnar
2008-02-29  4:10                     ` [PATCH 2/3] fix choice dependency check Roman Zippel
2008-04-28 21:08                       ` Sam Ravnborg
2008-02-29  4:11                     ` [PATCH 3/3] add named choice group Roman Zippel
2008-02-26 20:05                   ` [PATCH] x86: vSMP selection in config Sam Ravnborg
2008-02-26 21:03                     ` Yinghai Lu
2008-02-25  5:39           ` [PATCH] x86_64: make amd quad core 8 socket system not be clustered_box v2 Yinghai Lu
2008-02-25 19:08           ` Ravikiran Thirumalai
2008-02-25 22:05             ` Yinghai Lu
2008-02-26  3:39               ` Ravikiran Thirumalai [this message]
2008-02-26  3:46                 ` Andi Kleen
2008-02-26  4:05                   ` Ravikiran Thirumalai
2008-02-26  5:27                     ` Yinghai Lu
2008-02-26 18:42                       ` Ravikiran Thirumalai
2008-02-26 19:00                         ` Yinghai Lu
2008-02-26 20:32                           ` Ravikiran Thirumalai
2008-02-26 21:09                             ` Yinghai Lu
2008-02-26 21:10                             ` Yinghai Lu
2008-02-26 21:24                               ` Ravikiran Thirumalai
2008-02-26 23:16                                 ` Yinghai Lu
2008-02-26 23:31                                   ` Ravikiran Thirumalai
2008-02-26 23:41                                     ` Yinghai Lu
2008-02-22 19:08     ` [PATCH] x86_64: make amd quad core 8 socket system not be clustered_box Yinghai Lu
2008-02-22 18:59       ` 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=20080226033945.GJ3143@localdomain \
    --to=kiran@scalemp.com \
    --cc=akpm@linux-foundation.org \
    --cc=andi@firstfloor.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mingo@elte.hu \
    --cc=shai@scalemp.com \
    --cc=yhlu.kernel@gmail.com \
    --subject='Re: [PATCH] x86_64: make amd quad core 8 socket system not be clustered_box v2' \
    /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

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).