From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <linux-kernel-owner@vger.kernel.org>
Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand
	id S1754839AbbHFQD1 (ORCPT <rfc822;w@1wt.eu>);
	Thu, 6 Aug 2015 12:03:27 -0400
Received: from mga11.intel.com ([192.55.52.93]:10994 "EHLO mga11.intel.com"
	rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP
	id S1751235AbbHFQD0 (ORCPT <rfc822;linux-kernel@vger.kernel.org>);
	Thu, 6 Aug 2015 12:03:26 -0400
X-ExtLoop1: 1
X-IronPort-AV: E=Sophos;i="5.15,623,1432623600"; 
   d="scan'208";a="537198866"
Message-ID: <55C3854A.3000401@linux.intel.com>
Date: Thu, 06 Aug 2015 09:03:22 -0700
From: Dave Hansen <dave.hansen@linux.intel.com>
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:31.0) Gecko/20100101 Thunderbird/31.8.0
MIME-Version: 1.0
To: Dave Hansen <dave@sr71.net>, Ingo Molnar <mingo@kernel.org>,
        Linus Torvalds <torvalds@linux-foundation.org>
CC: Peter Anvin <hpa@zytor.com>, Denys Vlasenko <dvlasenk@redhat.com>,
        Thomas Gleixner <tglx@linutronix.de>, linux-kernel@vger.kernel.org,
        Andy Lutomirski <luto@kernel.org>, bp@alien8.de,
        Peter Zijlstra <a.p.zijlstra@chello.nl>, fenghua.yu@intel.com,
        x86@kernel.org
Subject: Re: [PATCH] x86, fpu: correct XSAVE xstate size calculation
References: <20150728172143.6DDFECA7@viggo.jf.intel.com> <20150805103227.GA3233@gmail.com> <55C21EFC.3060802@sr71.net> <20150806071545.GB2194@gmail.com> <CA+55aFxzOj-Ee=DN-_3CMeDeYVsmvmmgoxd3hp4MpRSp+og7AQ@mail.gmail.com> <20150806082746.GA10974@gmail.com> <55C3758A.50005@sr71.net>
In-Reply-To: <55C3758A.50005@sr71.net>
Content-Type: text/plain; charset=windows-1252
Content-Transfer-Encoding: 7bit
Sender: linux-kernel-owner@vger.kernel.org
List-ID: <linux-kernel.vger.kernel.org>
X-Mailing-List: linux-kernel@vger.kernel.org

Here's a correction about what we do today, and it's an important point.
 The code I was ripping out in the patch was *really* just for the
XSAVES/compacted case, does *NOT* do anything today since we've disabled
XSAVES.

So perhaps the title here should be:

	[PATCH] x86, fpu: correct XSAVES xstate size calculation

/* Option 1, what we have today */

	if (!cpu_has_xsaves) {
		cpuid(0xD0, 0, &total_blob_size, ...);
		return;
	}
	/*
	 * This breaks if offset[i]+size[i] != offset[i+1]
	 * or if alignment is in play.  Silly hardware breaks
	 * this today.
	 */
	for (i = 0; i < nr_xstates; i++) {
		if (!enabled_xstate(i))
			continue;
		total_blob_size += xstate_sizes[i];
	}