Message-ID: <CAKwvOd=u=ysrQfEDEHgqHAEd8kATehoX56CJqp-LymCUv+hUtA@mail.gmail.com>
Date:   Thu, 4 Oct 2018 16:12:17 -0700
From:   Nick Desaulniers <ndesaulniers@...gle.com>
To:     Nathan Chancellor <natechancellor@...il.com>
Cc:     broonie@...nel.org, linux-spi@...r.kernel.org,
        LKML <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH v2] spi: spi-ep93xx: Change dir type in ep93xx_spi_dma_{finish,prepare}

On Thu, Oct 4, 2018 at 2:47 PM Nathan Chancellor
<natechancellor@...il.com> wrote:
>
> On Thu, Oct 04, 2018 at 02:32:48PM -0700, Nick Desaulniers wrote:
> > On Wed, Oct 3, 2018 at 7:41 PM Nathan Chancellor
> > <natechancellor@...il.com> wrote:
> > >
> > > Clang warns when one enumerated type is implicitly converted to another.
> > >
> > > drivers/spi/spi-ep93xx.c:342:62: warning: implicit conversion from
> > > enumeration type 'enum dma_transfer_direction' to different enumeration
> > > type 'enum dma_data_direction' [-Wenum-conversion]
> >
> > I'm just trying to think what this change would look like if this
> > driver ONLY used `enum dma_data_direction`?  There's an additional
>
> I can work up a proof of concept patch but it was kind of convoluted
> when I tried before this patch because dmaengine_prep_slave_sg and the
> direction member in struct dma_slave_config expect an enum of type
> dma_transfer_direction so there are two explicit casts needed and I
> thought there were one or two more places that it complained.

If you made the attempt already and found this to be a better
solution, I trust your judgement.
Reviewed-by: Nick Desaulniers <ndesaulniers@...gle.com>

>
> This was the cleanest solution to me but I'm happy to rework it in any
> way that I need to.
>
> > warning in the header that a return value is messed up there, too:
> >
> > In file included from drivers/spi/spi-ep93xx.c:34:
> > ./include/linux/platform_data/dma-ep93xx.h:88:10: warning: implicit
> > conversion from
> >       enumeration type 'enum dma_data_direction' to different
> > enumeration type 'enum
> >       dma_transfer_direction' [-Wenum-conversion]
> >                 return DMA_NONE;
> >                 ~~~~~~ ^~~~~~~~
> >
>
> This is fixed in the dma tree and that function isn't used in this file:
> https://git.kernel.org/pub/scm/linux/kernel/git/vkoul/slave-dma.git/commit/?id=9524d6b265f9b2b9a61fceb2ee2ce1c2a83e39ca

Heh, I should have remembered!

>
> > >         nents = dma_map_sg(chan->device->dev, sgt->sgl, sgt->nents, dir);
> > >                 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^~~~
> > > ./include/linux/dma-mapping.h:428:58: note: expanded from macro
> > > 'dma_map_sg'
> > > #define dma_map_sg(d, s, n, r) dma_map_sg_attrs(d, s, n, r, 0)
> > >                                ~~~~~~~~~~~~~~~~          ^
> > > drivers/spi/spi-ep93xx.c:348:57: warning: implicit conversion from
> > > enumeration type 'enum dma_transfer_direction' to different enumeration
> > > type 'enum dma_data_direction' [-Wenum-conversion]
> > >                 dma_unmap_sg(chan->device->dev, sgt->sgl, sgt->nents, dir);
> > >                 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^~~~
> > > ./include/linux/dma-mapping.h:429:62: note: expanded from macro
> > > 'dma_unmap_sg'
> > > #define dma_unmap_sg(d, s, n, r) dma_unmap_sg_attrs(d, s, n, r, 0)
> > >                                  ~~~~~~~~~~~~~~~~~~          ^
> > > drivers/spi/spi-ep93xx.c:377:56: warning: implicit conversion from
> > > enumeration type 'enum dma_transfer_direction' to different enumeration
> > > type 'enum dma_data_direction' [-Wenum-conversion]
> > >         dma_unmap_sg(chan->device->dev, sgt->sgl, sgt->nents, dir);
> > >         ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^~~~
> > > ./include/linux/dma-mapping.h:429:62: note: expanded from macro
> > > 'dma_unmap_sg'
> > > #define dma_unmap_sg(d, s, n, r) dma_unmap_sg_attrs(d, s, n, r, 0)
> > >                                  ~~~~~~~~~~~~~~~~~~          ^
> > > 3 warnings generated.
> > >
> > > dma_{,un}map_sg expects an enum of type dma_data_direction but this
> > > driver uses dma_transfer_direction for everything. Converting to
> > > dma_data_direction would be desirable but there are a few shared
> > > structures that expect dma_transfer_direction so it is just simpler to
> > > change the parameter here. dma_transfer_direction and dma_data_direction
> > > are different sizes but this driver only uses the 1 and 2 values which
> > > mean the same thing so this change is safe.
> > >
> > > Signed-off-by: Nathan Chancellor <natechancellor@...il.com>
> > > ---
> > >
> > > v1 -> v2:
> > >
> > > * Fix escaped hash symbols for '#define' lines
> > >
> > >  drivers/spi/spi-ep93xx.c | 6 ++----
> > >  1 file changed, 2 insertions(+), 4 deletions(-)
> > >
> > > diff --git a/drivers/spi/spi-ep93xx.c b/drivers/spi/spi-ep93xx.c
> > > index f1526757aaf6..189fc2225b69 100644
> > > --- a/drivers/spi/spi-ep93xx.c
> > > +++ b/drivers/spi/spi-ep93xx.c
> > > @@ -256,8 +256,7 @@ static int ep93xx_spi_read_write(struct spi_master *master)
> > >   * in case of failure.
> > >   */
> > >  static struct dma_async_tx_descriptor *
> > > -ep93xx_spi_dma_prepare(struct spi_master *master,
> > > -                      enum dma_transfer_direction dir)
> > > +ep93xx_spi_dma_prepare(struct spi_master *master, int dir)
> > >  {
> > >         struct ep93xx_spi *espi = spi_master_get_devdata(master);
> > >         struct spi_transfer *xfer = master->cur_msg->state;
> > > @@ -359,8 +358,7 @@ ep93xx_spi_dma_prepare(struct spi_master *master,
> > >   * Function finishes with the DMA transfer. After this, the DMA buffer is
> > >   * unmapped.
> > >   */
> > > -static void ep93xx_spi_dma_finish(struct spi_master *master,
> > > -                                 enum dma_transfer_direction dir)
> > > +static void ep93xx_spi_dma_finish(struct spi_master *master, int dir)
> > >  {
> > >         struct ep93xx_spi *espi = spi_master_get_devdata(master);
> > >         struct dma_chan *chan;
> > > --
> > > 2.19.0
> > >
> >
> >
> > --
> > Thanks,
> > ~Nick Desaulniers



-- 
Thanks,
~Nick Desaulniers

Powered by blists - more mailing lists