linux-kernel - Re: cpu2000(both float and int) 13% regression with 2.6.28-rc1

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:	Fri, 31 Oct 2008 11:02:01 +0100
From:	Ingo Molnar <mingo@...e.hu>
To:	"H. Peter Anvin" <hpa@...or.com>
Cc:	"Pallipadi, Venkatesh" <venkatesh.pallipadi@...el.com>,
	"Siddha, Suresh B" <suresh.b.siddha@...el.com>,
	"Zhang, Yanmin" <yanmin_zhang@...ux.intel.com>,
	LKML <linux-kernel@...r.kernel.org>,
	Roland McGrath <roland@...hat.com>,
	Hiroshi Shimamoto <h-shimamoto@...jp.nec.com>,
	Yinghai Lu <yinghai@...nel.org>
Subject: Re: cpu2000(both float and int) 13% regression with 2.6.28-rc1


* H. Peter Anvin <hpa@...or.com> wrote:

> Pallipadi, Venkatesh wrote:
> > 
> > We figured out that this is not related to signals. But to this mismerge here
> > 
> >>  +#define X86_FEATURE_AMDC1E  (3*32+21) /* AMD C1E detected */
> >> + #define X86_FEATURE_XTOPOLOGY       (3*32+21) /* cpu topology enum extensions */
> > 
> > I had earler sent a patch to fix this.
> > http://marc.info/?l=linux-kernel&m=122341178202930&w=2
> > 
> > But, somehow I don’t see this patch either in Linus's git or in tip.
> > 
> > ingo, hpa: Can you push that patch along.
> > 
> 
> OK, thought it already had been.  I will investigate later tonight.

i've picked it up into x86/urgent - see the commit below.

	Ingo

--------------->
>From 2576c9991758e431b73e374f6019d6e1e12a8d36 Mon Sep 17 00:00:00 2001
From: Venki Pallipadi <venkatesh.pallipadi@...el.com>
Date: Tue, 7 Oct 2008 13:33:12 -0700
Subject: [PATCH] x86: fix AMDC1E and XTOPOLOGY conflict in cpufeature

Impact: fix xsave slowdown regression

Fix two features from conflicting in feature bits.

Fixes this performance regression:

   Subject: cpu2000(both float and int) 13% regression with 2.6.28-rc1
   http://lkml.org/lkml/2008/10/28/36

Reported-by: "Zhang, Yanmin" <yanmin_zhang@...ux.intel.com>
Bisected-by: "Zhang, Yanmin" <yanmin_zhang@...ux.intel.com>
Signed-off-by: Venkatesh Pallipadi <venkatesh.pallipadi@...el.com>
Signed-off-by: Ingo Molnar <mingo@...e.hu>
---
 arch/x86/include/asm/cpufeature.h |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

diff --git a/arch/x86/include/asm/cpufeature.h b/arch/x86/include/asm/cpufeature.h
index f73e95d..cfdf8c2 100644
--- a/arch/x86/include/asm/cpufeature.h
+++ b/arch/x86/include/asm/cpufeature.h
@@ -91,7 +91,7 @@
 #define X86_FEATURE_11AP	(3*32+19) /* "" Bad local APIC aka 11AP */
 #define X86_FEATURE_NOPL	(3*32+20) /* The NOPL (0F 1F) instructions */
 #define X86_FEATURE_AMDC1E	(3*32+21) /* AMD C1E detected */
-#define X86_FEATURE_XTOPOLOGY	(3*32+21) /* cpu topology enum extensions */
+#define X86_FEATURE_XTOPOLOGY	(3*32+22) /* cpu topology enum extensions */
 
 /* Intel-defined CPU features, CPUID level 0x00000001 (ecx), word 4 */
 #define X86_FEATURE_XMM3	(4*32+ 0) /* "pni" SSE-3 */
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/