From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from e06smtp16.uk.ibm.com (e06smtp16.uk.ibm.com [195.75.94.112]) (using TLSv1 with cipher DHE-RSA-AES256-SHA (256/256 bits)) (No client certificate requested) by lists.ozlabs.org (Postfix) with ESMTPS id 589591A0DCC for ; Fri, 16 Jan 2015 20:43:40 +1100 (AEDT) Received: from /spool/local by e06smtp16.uk.ibm.com with IBM ESMTP SMTP Gateway: Authorized Use Only! Violators will be prosecuted for from ; Fri, 16 Jan 2015 09:43:36 -0000 Received: from b06cxnps4076.portsmouth.uk.ibm.com (d06relay13.portsmouth.uk.ibm.com [9.149.109.198]) by d06dlp02.portsmouth.uk.ibm.com (Postfix) with ESMTP id 472FA219005F for ; Fri, 16 Jan 2015 09:43:32 +0000 (GMT) Received: from d06av10.portsmouth.uk.ibm.com (d06av10.portsmouth.uk.ibm.com [9.149.37.251]) by b06cxnps4076.portsmouth.uk.ibm.com (8.14.9/8.14.9/NCO v10.0) with ESMTP id t0G9hXhF56688698 for ; Fri, 16 Jan 2015 09:43:33 GMT Received: from d06av10.portsmouth.uk.ibm.com (localhost [127.0.0.1]) by d06av10.portsmouth.uk.ibm.com (8.14.4/8.14.4/NCO v10.0 AVout) with ESMTP id t0G9hWxP024393 for ; Fri, 16 Jan 2015 02:43:33 -0700 Message-ID: <54B8DD44.1020402@de.ibm.com> Date: Fri, 16 Jan 2015 10:43:32 +0100 From: Christian Borntraeger MIME-Version: 1.0 To: Michael Ellerman Subject: Re: [PATCH 1/8] ppc/kvm: Replace ACCESS_ONCE with READ_ONCE References: <1421312314-72330-1-git-send-email-borntraeger@de.ibm.com> <1421312314-72330-2-git-send-email-borntraeger@de.ibm.com> <1421363369.23332.8.camel@ellerman.id.au> In-Reply-To: <1421363369.23332.8.camel@ellerman.id.au> Content-Type: text/plain; charset=utf-8 Cc: linux-arch@vger.kernel.org, kvm@vger.kernel.org, x86@kernel.org, linux-kernel@vger.kernel.org, kvm-ppc@vger.kernel.org, linux-mm@kvack.org, xen-devel@lists.xenproject.org, linuxppc-dev@lists.ozlabs.org List-Id: Linux on PowerPC Developers Mail List List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Am 16.01.2015 um 00:09 schrieb Michael Ellerman: > On Thu, 2015-01-15 at 09:58 +0100, Christian Borntraeger wrote: >> ACCESS_ONCE does not work reliably on non-scalar types. For >> example gcc 4.6 and 4.7 might remove the volatile tag for such >> accesses during the SRA (scalar replacement of aggregates) step >> (https://gcc.gnu.org/bugzilla/show_bug.cgi?id=58145) >> >> Change the ppc/kvm code to replace ACCESS_ONCE with READ_ONCE. >> >> Signed-off-by: Christian Borntraeger >> --- >> arch/powerpc/kvm/book3s_hv_rm_xics.c | 8 ++++---- >> arch/powerpc/kvm/book3s_xics.c | 16 ++++++++-------- >> 2 files changed, 12 insertions(+), 12 deletions(-) >> >> diff --git a/arch/powerpc/kvm/book3s_hv_rm_xics.c b/arch/powerpc/kvm/book3s_hv_rm_xics.c >> index 7b066f6..7c22997 100644 >> --- a/arch/powerpc/kvm/book3s_hv_rm_xics.c >> +++ b/arch/powerpc/kvm/book3s_hv_rm_xics.c >> @@ -152,7 +152,7 @@ static void icp_rm_down_cppr(struct kvmppc_xics *xics, struct kvmppc_icp *icp, >> * in virtual mode. >> */ >> do { >> - old_state = new_state = ACCESS_ONCE(icp->state); >> + old_state = new_state = READ_ONCE(icp->state); > > These are all icp->state. > > Which is a union, but it's only the size of unsigned long. So in practice there > shouldn't be a bug here right? This bug was that gcc lost the volatile tag when propagating aggregates to scalar types. So in theory a union could be affected. See the original problem ( http://marc.info/?i=54611D86.4040306%40de.ibm.com ) which happened on union ipte_control { unsigned long val; struct { unsigned long k : 1; unsigned long kh : 31; unsigned long kg : 32; }; }; Christian