From mboxrd@z Thu Jan 1 00:00:00 1970 From: Julien Grall Subject: Re: [PATCH] xen: arm: bitops take unsigned int (Was: Re: [PATCH] xen/arm64: disable alignment check) Date: Mon, 12 May 2014 14:44:44 +0100 Message-ID: <5370D04C.20203@linaro.org> References: <1398589848-14731-1-git-send-email-murzin.v@gmail.com> <1398761926.4457.14.camel@kazak.uk.xensource.com> <1399641896.561.9.camel@kazak.uk.xensource.com> <1399652387.561.58.camel@kazak.uk.xensource.com> <5370BB45.5000608@citrix.com> <1399897136.22097.12.camel@kazak.uk.xensource.com> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Return-path: Received: from mail6.bemta14.messagelabs.com ([193.109.254.103]) by lists.xen.org with esmtp (Exim 4.72) (envelope-from ) id 1WjqXD-0007pv-U4 for xen-devel@lists.xenproject.org; Mon, 12 May 2014 13:44:52 +0000 Received: by mail-ee0-f45.google.com with SMTP id d49so4622934eek.18 for ; Mon, 12 May 2014 06:44:48 -0700 (PDT) In-Reply-To: <1399897136.22097.12.camel@kazak.uk.xensource.com> List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Sender: xen-devel-bounces@lists.xen.org Errors-To: xen-devel-bounces@lists.xen.org To: Ian Campbell , Julien Grall Cc: Vladimir Murzin , Tim Deegan , Stefano Stabellini , David Vrabel , Jan Beulich , "xen-devel@lists.xenproject.org" List-Id: xen-devel@lists.xenproject.org On 05/12/2014 01:18 PM, Ian Campbell wrote: > On Mon, 2014-05-12 at 13:15 +0100, Julien Grall wrote: >> Hi Ian, >> >> On 05/09/2014 05:19 PM, Ian Campbell wrote: >>> (Just adding the other ARM guys...) >>> >>> On Fri, 2014-05-09 at 14:24 +0100, Ian Campbell wrote: >>>> On Tue, 2014-04-29 at 09:58 +0100, Ian Campbell wrote: >>>>> On Tue, 2014-04-29 at 08:38 +0100, Vladimir Murzin wrote: >>>>>> On Mon, Apr 28, 2014 at 11:43 AM, Ian Campbell wrote: >>>>>>> But I also wanted confirmation that the problematic instruction was >>>>>>> generated by gcc and not by some handcoded asm somewhere which we hadn't >>>>>>> properly fixed. >>>>> >>>>>> I believe it comes form test_bit (xen/include/asm-arm/bitops.h). >>>>> >>>>> Ah, then I think this code needs fixing too. Probably switching to >>>>> unsigned int * throughout would work, what do you think? >>>> >>>> I finally managed to upgrade to a new enough kernel to trigger this. >>>> >>>> This Works For Me(tm), along with the Linux patch "xen/events/fifo: >>>> correctly align bitops" which is queued for 3.15 Linus (but not sent >>>> yet?) >>>> >>>> 8<------------------- >>>> >>>> From aa6afe6520ea22241fb0ce430ef315c49a73867f Mon Sep 17 00:00:00 2001 >>>> From: Ian Campbell >>>> Date: Thu, 8 May 2014 16:13:55 +0100 >>>> Subject: [PATCH] xen: arm: bitops take unsigned int >>>> >>>> Xen bitmaps can be 4 rather than 8 byte aligned, so use the appropriate type. >>>> Otherwise the compiler can generate unaligned 8 byte accesses and cause traps. >>>> >>>> Signed-off-by: Ian Campbell >>>> --- >>>> xen/include/asm-arm/bitops.h | 37 +++++++++++++++++++------------------ >>>> 1 file changed, 19 insertions(+), 18 deletions(-) >>>> >>>> diff --git a/xen/include/asm-arm/bitops.h b/xen/include/asm-arm/bitops.h >>>> index 0a7caee..25f96c8 100644 >>>> --- a/xen/include/asm-arm/bitops.h >>>> +++ b/xen/include/asm-arm/bitops.h >>>> @@ -18,13 +18,14 @@ >>>> #define __set_bit(n,p) set_bit(n,p) >>>> #define __clear_bit(n,p) clear_bit(n,p) >>>> >>>> +#define BITS_PER_WORD 32 >> >> Can you define BITS_PER_WORD in asm-arm/config.h rather than here? > > For better or worse BITS_PER_BYTE is already defined in bitops.h and > since I've already run the majority of my pre-push commit checks on a > branch containing this fix (along with some other bits and bobs) I'm not > inclined to restart that process just for this change. No problem. I guess you plan to backport this patch for Xen 4.4? Regards, -- Julien Grall