linux-kernel - Re: [RFC][PATCH 1/2] mm: cma: Export cma symbols for cma heap as a module

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]

Date:   Mon, 28 Oct 2019 15:26:28 -0700
From:   John Stultz <john.stultz@...aro.org>
To:     Sandeep Patil <sspatil@...gle.com>
Cc:     lkml <linux-kernel@...r.kernel.org>,
        Laura Abbott <labbott@...hat.com>,
        Benjamin Gaignard <benjamin.gaignard@...aro.org>,
        Sumit Semwal <sumit.semwal@...aro.org>,
        Liam Mark <lmark@...eaurora.org>,
        Pratik Patel <pratikp@...eaurora.org>,
        Brian Starkey <Brian.Starkey@....com>,
        Andrew Morton <akpm@...ux-foundation.org>,
        Chenbo Feng <fengc@...gle.com>,
        Hridya Valsaraju <hridya@...gle.com>,
        dri-devel <dri-devel@...ts.freedesktop.org>,
        "Andrew F . Davis" <afd@...com>, Yue Hu <huyue2@...ong.com>,
        Mike Rapoport <rppt@...ux.ibm.com>,
        Alistair Strachan <astrachan@...gle.com>
Subject: Re: [RFC][PATCH 1/2] mm: cma: Export cma symbols for cma heap as a module

On Mon, Oct 28, 2019 at 1:03 PM John Stultz <john.stultz@...aro.org> wrote:
>
> On Mon, Oct 28, 2019 at 12:12 PM <sspatil@...gle.com> wrote:
> > On Fri, Oct 25, 2019 at 11:48:33PM +0000, John Stultz wrote:
> > > --- a/kernel/dma/contiguous.c
> > > +++ b/kernel/dma/contiguous.c
> > > @@ -31,6 +31,7 @@
> > >  #endif
> > >
> > >  struct cma *dma_contiguous_default_area;
> > > +EXPORT_SYMBOL(dma_contiguous_default_area);
> >
> > I didn't need to do this for the (out-of-tree) ion cma heap [1].
> > Any reason why you had to?
>
> Its likely due to the changes I made in the separate
> non-default-CMA-region patch set. Earlier I had gotten away with just
> your change, but in testing before I sent this series, I hit the build
> error and quickly added the export before sending.

And just to clarify this point, I was mistaken and it wasn't the
non-default-CMA-region patch set, but the fact that I'm using
dev_get_cma_area(NULL) to get the default CMA area to register, rather
then trying to register every CMA area via cma_for_each_area() as is
done in ION.  I actually could probably drop the cma_for_each_area
export for the dmabuf heaps usage.

thanks
-john