From owner-svn-src-head@FreeBSD.ORG Wed Nov 11 02:39:21 2009 Return-Path: Delivered-To: svn-src-head@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34]) by hub.freebsd.org (Postfix) with ESMTP id 068F0106566C; Wed, 11 Nov 2009 02:39:21 +0000 (UTC) (envelope-from kuriyama@FreeBSD.org) Received: from scarlet.imgsrc.co.jp (scarlet.imgsrc.co.jp [202.235.195.9]) by mx1.freebsd.org (Postfix) with ESMTP id 9C9D48FC1C; Wed, 11 Nov 2009 02:39:20 +0000 (UTC) Received: from localhost (localhost [127.0.0.1]) by scarlet.imgsrc.co.jp (Postfix) with ESMTP id 3F8FB678CD; Wed, 11 Nov 2009 11:39:19 +0900 (JST) X-Virus-Scanned: IMG SRC scanner at mail.imgsrc.co.jp (scarlet) Received: from scarlet.imgsrc.co.jp ([127.0.0.1]) by localhost (scarlet.imgsrc.co.jp [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id vogpp4tPEet7; Wed, 11 Nov 2009 11:39:19 +0900 (JST) Received: from cebu.s2factory.co.jp (corsica.s2factory.co.jp [122.220.15.114]) by scarlet.imgsrc.co.jp (Postfix) with ESMTP id EE0C9678C6; Wed, 11 Nov 2009 11:39:18 +0900 (JST) Date: Wed, 11 Nov 2009 11:39:18 +0900 Message-ID: <7meio5g4yx.wl%kuriyama@s2factory.co.jp> From: Jun Kuriyama To: Ed Schouten In-Reply-To: <20091110160300.GD64905@hoeg.nl> References: <200911090254.nA92sG1G005921@svn.freebsd.org> <20091109225244.GB64905@hoeg.nl> <7mbpjbgt0n.wl%kuriyama@s2factory.co.jp> <20091110055910.GC64905@hoeg.nl> <7mvdhigb96.wl%kuriyama@s2factory.co.jp> <20091110160300.GD64905@hoeg.nl> User-Agent: Wanderlust/2.14.0 (Africa) Emacs/23.1 Mule/6.0 (HANACHIRUSATO) MIME-Version: 1.0 (generated by SEMI 1.14.6 - "Maruoka") Content-Type: multipart/mixed; boundary="Multipart_Wed_Nov_11_11:39:18_2009-1" Cc: svn-src-head@freebsd.org, svn-src-all@freebsd.org, src-committers@freebsd.org Subject: Re: svn commit: r199067 - in head/sys: amd64/amd64 i386/i386 X-BeenThere: svn-src-head@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: SVN commit messages for the src tree for head/-current List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Wed, 11 Nov 2009 02:39:21 -0000 --Multipart_Wed_Nov_11_11:39:18_2009-1 Content-Type: text/plain; charset=US-ASCII At Tue, 10 Nov 2009 17:03:00 +0100, ed@80386.nl wrote: > > You can try with hw.clflush_disable="1" in loader.conf. If you can > > boot with it, I must be mistaken something (anyway its my fault). > > For some reason that switch doesn't seem to have any effect. Okay, something wrong... > Apple MacBook3,1: > | Features=0xbfebfbff > > Dell SC440: > | Features=0xbfebfbff Both system has CLFUSH and SS, so CLFLUSH tweak should not executed. Can you test with these patches? Testing on only one of both system is enough. "patch-1" forces disabling CLFLUSH feature even if SS bit exists. "patch-2" forces no CLFLUSH tweak. I'd like to know with which patch your system can live. -- Jun Kuriyama // FreeBSD Project // S2 Factory, Inc. --Multipart_Wed_Nov_11_11:39:18_2009-1 Content-Type: application/octet-stream Content-Disposition: attachment; filename="patch-1" Content-Transfer-Encoding: 7bit Index: initcpu.c =================================================================== --- initcpu.c (revision 199067) +++ initcpu.c (working copy) @@ -176,16 +176,20 @@ * XXXKIB: (temporary) hack to work around traps generated when * CLFLUSHing APIC registers window. */ +#if 0 TUNABLE_INT_FETCH("hw.clflush_disable", &hw_clflush_disable); if (cpu_vendor_id == CPU_VENDOR_INTEL && !(cpu_feature & CPUID_SS) && hw_clflush_disable == -1) +#endif cpu_feature &= ~CPUID_CLFSH; /* * Allow to disable CLFLUSH feature manually by * hw.clflush_disable tunable. This may help Xen guest on some AMD * CPUs. */ +#if 0 if (hw_clflush_disable == 1) { cpu_feature &= ~CPUID_CLFSH; } +#endif } --Multipart_Wed_Nov_11_11:39:18_2009-1 Content-Type: application/octet-stream Content-Disposition: attachment; filename="patch-2" Content-Transfer-Encoding: 7bit Index: initcpu.c =================================================================== --- initcpu.c (revision 199067) +++ initcpu.c (working copy) @@ -176,16 +176,20 @@ * XXXKIB: (temporary) hack to work around traps generated when * CLFLUSHing APIC registers window. */ +#if 0 TUNABLE_INT_FETCH("hw.clflush_disable", &hw_clflush_disable); if (cpu_vendor_id == CPU_VENDOR_INTEL && !(cpu_feature & CPUID_SS) && hw_clflush_disable == -1) cpu_feature &= ~CPUID_CLFSH; +#endif /* * Allow to disable CLFLUSH feature manually by * hw.clflush_disable tunable. This may help Xen guest on some AMD * CPUs. */ +#if 0 if (hw_clflush_disable == 1) { cpu_feature &= ~CPUID_CLFSH; } +#endif } --Multipart_Wed_Nov_11_11:39:18_2009-1--