[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-Id: <20091205115936.d3a7a6ec.sfr@canb.auug.org.au>
Date: Sat, 5 Dec 2009 11:59:36 +1100
From: Stephen Rothwell <sfr@...b.auug.org.au>
To: Emese Revfy <re.emese@...il.com>
Cc: tony.luck@...el.com, geoffrey.levand@...sony.com,
davem@...emloft.net, tglx@...utronix.de, joerg.roedel@....com,
muli@...ibm.com, jdmason@...zu.us, jbarnes@...tuousgeek.org,
dwmw2@...radead.org, torvalds@...ux-foundation.org,
linux-kernel@...r.kernel.org
Subject: Re: [PATCH 05/31] Constify struct dma_map_ops for 2.6.32 v1
Hi Emese,
On Fri, 04 Dec 2009 23:14:08 +0100 Emese Revfy <re.emese@...il.com> wrote:
>
> --- a/arch/powerpc/kernel/vio.c
> +++ b/arch/powerpc/kernel/vio.c
> @@ -601,6 +601,7 @@ static void vio_dma_iommu_unmap_sg(struct device *dev,
> vio_cmo_dealloc(viodev, alloc_size);
> }
>
> +/* cannot be const */
> struct dma_map_ops vio_dma_mapping_ops = {
The problem with leaving this not const is that a pointer to it is later
assigned to an archdata::dma_ops which you have made into a const pointer
earlier in this patch. This will generate a warning which is fatal in
arch/powerpc as we compile that directory with -Werror.
--
Cheers,
Stephen Rothwell sfr@...b.auug.org.au
http://www.canb.auug.org.au/~sfr/
Content of type "application/pgp-signature" skipped
Powered by blists - more mailing lists