linux-kernel - Re: block: DMA alignment of IO buffer allocated from slab

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]

Message-ID: <20180925065517.GA4868@ming.t460p>
Date:   Tue, 25 Sep 2018 14:55:18 +0800
From:   Ming Lei <ming.lei@...hat.com>
To:     Matthew Wilcox <willy@...radead.org>
Cc:     Bart Van Assche <bvanassche@....org>,
        Andrey Ryabinin <aryabinin@...tuozzo.com>,
        Vitaly Kuznetsov <vkuznets@...hat.com>,
        Christoph Hellwig <hch@....de>,
        Ming Lei <tom.leiming@...il.com>,
        linux-block <linux-block@...r.kernel.org>,
        linux-mm <linux-mm@...ck.org>,
        Linux FS Devel <linux-fsdevel@...r.kernel.org>,
        "open list:XFS FILESYSTEM" <linux-xfs@...r.kernel.org>,
        Dave Chinner <dchinner@...hat.com>,
        Linux Kernel Mailing List <linux-kernel@...r.kernel.org>,
        Jens Axboe <axboe@...nel.dk>, Christoph Lameter <cl@...ux.com>,
        Linus Torvalds <torvalds@...ux-foundation.org>,
        Greg Kroah-Hartman <gregkh@...uxfoundation.org>
Subject: Re: block: DMA alignment of IO buffer allocated from slab

On Mon, Sep 24, 2018 at 09:44:21PM -0700, Matthew Wilcox wrote:
> On Mon, Sep 24, 2018 at 09:10:43PM -0700, Bart Van Assche wrote:
> > On 9/24/18 8:28 PM, Matthew Wilcox wrote:
> > > [ ... ] Because if we have to
> > > round all allocations below 64 bytes up to 64 bytes, [ ... ]
> > Have you noticed that in another e-mail in this thread it has been explained
> > why it is not necessary on x86 to align buffers allocated by kmalloc() on a
> > 64-byte boundary even if these buffers are used for DMA?
> 
> Oh, so drivers which do this only break on !x86.  Yes, that'll work
> out great.

It shouldn't break !x86 because ARCH_KMALLOC_MINALIGN handles that.

Thanks,
Ming